Making Scenes: Global Perspectives on Scenes in Rock Art 9781789209211

Table of contents :
CONTENTS
List of Illustrations
Preface
Introduction. Behind the Scenes: Did Scenes in Rock Art Create New Ways of Seeing the World?
Chapter 1. Scenes and Non-Scenes in Rock Art
Chapter 2. The Possible Signifi cance of Depicted Scenes for Cognitive Development
Chapter 3. Event Depiction in Rock Art: Landscape-Embedded Plan-View Narratives, Decontextualized Profile “Scenes,” and Their Hybrid Instances
Chapter 4. Defining “Scenes” in Rock Art Research: Visual Conventions and Beyond
Chapter 5. Putting Southern African Rock Paintings in Context: The View from the Mirabib Rock Shelter, Western Namibia
Chapter 6. Scenic Narratives of Humans and Animals in Namibian Rock Art: A Methodological Restart with Data Mining
Chapter 7. Between Scene and Association: Toward a Better Understanding of Scenes in the Rock Art of Iran
Chapter 8. Music and Dancing Scenes in the Rock Art of Central India
Chapter 9. Hunting and Havoc: Narrative Scenes in the Black Desert Rock Art of Jebel Qurma, Jordan
Chapter 10. Making a Scene: An Analysis of Rock Art Panels from the Northwest Kimberley and Central Desert, Australia
Chapter 11. Scene but Not Heard: Seeing Scenes in a Northern Australian Aboriginal Site
Chapter 12. A Comparison of “Scenes” in Parietal and Non-Parietal Upper Paleolithic Imagery: Formal Differences and Ontological Implications
Chapter 13. Scene Makers: Finger Fluters in Rouffignac Cave, France
Chapter 14. Maps in Prehistoric Art
Chapter 15. Scenes in the Paleolithic and Levantine Art of Eastern Spain
Chapter 16. New Insights into the Analysis of Levantine Rock Art Scenes Informed by Observations on Western Arnhem Land Rock Art
Chapter 17. Rules of Ordering and Grouping in the Pitoti, the Later Prehistoric Rock Engravings of Valcamonica (BS), Italy: from Solitary Figures through Clusters, Graphic Groups, and Scenes to Narrative
Chapter 18. Finding Order out of Chaos: A Statistical Analysis of Nine Mile Canyon Rock Art
Chapter 19. Interpreting Scenes in the Rock Art of the Canadian Maritimes
Chapter 20. The “Black Series” in the Hunting Scenes of Cueva de las Manos, Río Pinturas, Patagonia, Argentina
Epilogue. Is There More to Scenes than Meets the Eye?
INDEX

Citation preview

Making Scenes

MAKING SCENES Global Perspectives on Scenes in Rock Art

=== •

•

Edited by

Iain Davidson and April Nowell

berghahn NEW YORK • OXFORD www.berghahnbooks.com

Published in 2021 by Berghahn Books www.berghahnbooks.com © 2021 Iain Davidson and April Nowell

All rights reserved. Except for the quotation of short passages for the purposes of criticism and review, no part of this book may be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system now known or to be invented, without written permission of the publisher.

Library of Congress Cataloging-in-Publication Data Names: Davidson, Iain, 1948– editor. | Nowell, April, 1969– editor. Title: Making scenes : global perspectives on scenes in rock art / edited by Iain Davidson and April Nowell. Other titles: Making scenes (Berghahn Books) Description: New York : Berghahn, 2021. | Includes bibliographical references and index. | Summary: “Dating back to at least 50,000 years ago, rock art is one of the oldest forms of human symbolic expression. Geographically, it spans all the continents on Earth. Scenes are common in some rock art, and recent work suggests that there are some hints of expression that looks like some of the conventions of western scenic art. In this unique volume examining the nature of scenes in rock art, researchers examine what defines a scene, what are the necessary elements of a scene, and what can the evolutionary history tell us about storytelling, sequential memory and cognitive evolution among ancient and living cultures?”—Provided by publisher. Identifiers: LCCN 2020045534 (print) | LCCN 2020045535 (ebook) | ISBN 9781789209204 (hardback) | ISBN 9781789209211 (ebook) Subjects: LCSH: Rock paintings--Themes, motives. | Petroglyphs. | Composition (Art) Classification: LCC N5310 .M26 2021 (print) | LCC N5310 (ebook) | DDC 759.01/13—dc23 LC record available at https://lccn.loc.gov/2020045534 LC ebook record available at https://lccn.loc.gov/2020045535

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library

ISBN 978-1-78920-920-4 hardback ISBN 978-1-78920-921-1 ebook

==== •

•

•

CONTENTS

List of Illustrations Preface Margaret Conkey Introduction. Behind the Scenes: Did Scenes in Rock Art Create New Ways of Seeing the World? Iain Davidson and April Nowell

vii xvii

1

Chapter 1. Scenes and Non-Scenes in Rock Art Iain Davidson

16

Chapter 2. The Possible Significance of Depicted Scenes for Cognitive Development Livio Dobrez

32

Chapter 3. Event Depiction in Rock Art: Landscape-Embedded Plan-View Narratives, Decontextualized Profile “Scenes,” and Their Hybrid Instances Patricia Dobrez Chapter 4. Defining “Scenes” in Rock Art Research: Visual Conventions and Beyond Madeleine Kelly and Bruno David

51 67

Chapter 5. Putting Southern African Rock Paintings in Context: The View from the Mirabib Rock Shelter, Western Namibia Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe

75

Chapter 6. Scenic Narratives of Humans and Animals in Namibian Rock Art: A Methodological Restart with Data Mining Tilman Lenssen-Erz, Eymard Fäder, Oliver Vogels, and Brigitte Mathiak

90

Chapter 7. Between Scene and Association: Toward a Better Understanding of Scenes in the Rock Art of Iran Ebrahim Karimi Chapter 8. Music and Dancing Scenes in the Rock Art of Central India Meenakshi Dubey-Pathak and Jean Clottes

107 122

Chapter 9. Hunting and Havoc: Narrative Scenes in the Black Desert Rock Art of Jebel Qurma, Jordan Nathalie Østerled Brusgaard and Keshia A. N. Akkermans

134

Chapter 10. Making a Scene: An Analysis of Rock Art Panels from the Northwest Kimberley and Central Desert, Australia June Ross

150

Chapter 11. Scene but Not Heard: Seeing Scenes in a Northern Australian Aboriginal Site Madeleine Kelly, Bruno David, and Josephine Flood

162

vi

CONTENTS

Chapter 12. A Comparison of “Scenes” in Parietal and Non-Parietal Upper Paleolithic Imagery: Formal Differences and Ontological Implications Elisabeth Culley

179

Chapter 13. Scene Makers: Finger Fluters in Rouffignac Cave, France Leslie Van Gelder and April Nowell

194

Chapter 14. Maps in Prehistoric Art Pilar Utrilla, Carlos Mazo, Rafael Domingo, and Manuel Bea

207

Chapter 15. Scenes in the Paleolithic and Levantine Art of Eastern Spain Valentín Villaverde

223

Chapter 16. New Insights into the Analysis of Levantine Rock Art Scenes Informed by Observations on Western Arnhem Land Rock Art Inés Domingo Sanz Chapter 17. Rules of Ordering and Grouping in the Pitoti, the Later Prehistoric Rock Engravings of Valcamonica (BS), Italy: from Solitary Figures through Clusters, Graphic Groups, and Scenes to Narrative Craig Alexander, Alberto Marretta, Thomas Huet, and Christopher Chippindale

240

259

Chapter 18. Finding Order out of Chaos: A Statistical Analysis of Nine Mile Canyon Rock Art Jerry D. Spangler and Iain Davidson

277

Chapter 19. Interpreting Scenes in the Rock Art of the Canadian Maritimes Bryn Tapper and Oscar Moro Abadía

295

Chapter 20. The “Black Series” in the Hunting Scenes of Cueva de las Manos, Río Pinturas, Patagonia, Argentina Carlos A. Aschero and Patricia Schneier Epilogue. Is There More to Scenes than Meets the Eye? Iain Davidson and April Nowell

310 327

==== •

•

•

ILLUSTRATIONS

Figures Figure 0.1. Map showing dots for the location of chapters (by number) that have a regional focus.

2

Figure 1.1a–g. (a) Elk at Warrior Ridge, Nine Mile Canyon. Copyright Iain Davidson. (b) The Grand Panel of Horses at Ekain. Copyright Jesús Altuna and Koro Mariezkurrena. (c) More than one species at Halo Cave, Lower Pecos, Texas. Copyright SHUMLA. (d) The Panel of Horses, Chauvet Cave. Copyright Jean Clottes. (e) Main panel, Cueva del Niño, Albacete. Copyright Iain Davidson. (f) Different weathering of images of bighorn sheep, Nine Mile Canyon, Utah. Copyright Iain Davidson. (g) Paintings of different animals done at different times, Serra da Capivara, Brazil. Copyright Livio and Patricia Dobrez.

20

Figure 1.2. Main panel at Cueva del Niño, Albacete, Spain. Copyright Iain Davidson.

21

Figure 1.3a–b. (a) Image of a horse from Parpalló (No 17757B in Villaverde 1994). Copyright Valentín Villaverde. (b) Image of Pooh, Eeyore and Christopher Robin by E. H. Shepard from Winnie the Pooh (Milne 1926, 52).

24

Figure 1.4. The Grand Panel of Horses at Ekain. Copyright Jesús Altuna and Koro Mariezkurrena.

24

Figure 1.5. Image of two or three lions from Chauvet Cave showing the interactions between them through the slight difference in the orientation of the head. Copyright Jean Clottes.

25

Figure 1.6. The Great Hunt Panel, Nine Mile Canyon, Utah. Copyright Iain Davidson.

27

Figure 2.1. “Peak action.” Copyright David L. Minick. Total Access.

36

Figure 2.2. Niaux (model), Parc de la Préhistoire, Tarascon-sur-Ariège, France. Copyright Livio Dobrez.

39

Figure 2.3. Daraki-Chattan, Madhya Pradesh, India. Copyright Livio Dobrez.

39

Figure 2.4. Tandjiesberg, Orange Free State, South Africa. Copyright Livio Dobrez.

40

Figure 2.5. Pressa Canyon, Texas, USA. Copyright Livio Dobrez.

41

Figure 3.1. Illustration of trace-view emu hunt depiction, central Australia. Basedow (1925, Fig. 44).

51

Figure 3.2. Scene. Noukloof Mountains, Namibia. Copyright Patricia Dobrez.

55

Figure 3.3. Charlie Marshall Tjungurrayi, Warlpiri people. Witchetty Grubs (maku) at Kunatjarri, 1984. Synthetic polymer paint on canvas, 122 x 244 cm. National Gallery of Australia, Canberra. Purchased from Gallery Admission charges 1984. Copyright the artist. Licensed by Aboriginal Artists Agency Ltd.

57

Figure 3.4. Pedal narrative. Owens Valley, California. Copyright Patricia Dobrez.

59

Figure 3.5. Hybrid profile and plan-view narrative depiction. Writing-on-Stone, Alberta, Canada. Reduced from ink tracing on polyethylene by James D. Keyser, 1976.

62

viii

ILLUSTRATIONS

Figure 5.1. Plan view map of the Mirabib rockshelter. Copyright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

80

Figure 5.2. View of the painted boulder surface at Mirabib and D-Stretch enhancement of the painted rock surface. Copyright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

80

Figure 5.3. Two quadrupeds being led by an anthropomorphic figure entering a watermark and disappearing into a crack in the rock. Copyright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

81

Figure 5.4. Large quadruped being led by an anthropomorphic figure into a watermark. Note also a therianthropic figure above the large quadruped and various anthropomorphic and zoomorphic figures below. Coypright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

81

Figure 5.5. Complex panel with a bat-eared fox therianthrope (center-left) and various other anthropomorphic, zoomorphic, and therianthropic figures. Copyright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

82

Figure 5.6. Quadruped elements pictured in Figure 5.3 covered by dripping water during a rainstorm. Copyright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

84

Figure 6.1. Reproduction of rock paintings from Circus Gorge, Brandberg-Daureb (Pager 2006: 358–89) displaying a fair number of typical activities and interactions. The small numbers next to the figures are the reference numbers for retrieval of the figures in the catalogue. Colors: red, brown, black, ochre and white; size of largest human at right (no. 183) 26 cm. Copyright Heinrich-Barth-Institut.

91

Figure 6.2. Hierarchical interaction of the main motifs; for a definition of the terms Agens and Patiens see text. (Graphic artwork O. Vogels)

98

Figure 6.3. Non-hierarchical interaction of the main motifs “in mutual attendance” and “coordinated action.” The former is any unspecifiable action of two actors directed at one another; the latter are three or more actors who display comparatively complex activity in the very same manner. (Graphic artwork O. Vogels)

99

Figure 6.4. Visualization of interactions as a network. The circles are a proportionate rendering of the quantity of interacting figures of every motif, and the breadth of lines is a proportionate rendering of the extent of interactions. The short loops indicate interaction within the same motif category. (Graphic artwork E. Fäder)

100

Figure 6.5. Primary scenic foci of the main motifs. They are listed from top to bottom in declining complexity, i.e., in terms of scenic information “goal of specialized action” is the most complex; “most elaborate painting” is the least. (Graphic artwork O. Vogels)

101

Figure 7.1a–b. (a) Shmsali rock shelter, western Iran (Azandaryani et al. 2015: 63). Copyright E. H. Azandaryani). (b) Domab, Esfahan, Central Iran. Copyright Ebrahim Karimi.

112

Figure 7.2a–d. (a) and (b) Petroglyphs with hunting themes, Ernan Mount, Yazd, central Iran. Copyright Ebrahim Karimi. (c) Human depictions startling game in a group hunting. From the Akbarnama, made probably in India in 1604 CE. Metropolitan Museum of Art, n.d., “Akbar Hunting with Cheetahs,” Folio from the Akbarnama, Copyright status: Public Domain. Credit: Theodore M. Davis Collection, Bequest of Theodore M. Davis, 1915. (d) Domab, Esfahan, central Iran. Copyright Ebrahim Karimi.

113

ILLUSTRATIONS

ix

Figure 7.3a–d. (a) Saravan, southeast Iran (Sarhaddi, 2013: 5), image credits: F. Sarhaddi, https://www.ancient-asia-journal.com/articles/10.5334/aa.12312/). (b) Gharghab, central Iran (Naserifard 2016: 386, Copyright M. Naserifard). (c) A leaping ibex, Jorbat, eastern Iran (Vahdati 2010: 49, Copyright A. A. Vahdati). (d) Teymare, central Iran (Naserifard 2009, 2016, Copyright M. Naserifard).

115

Figure 7.4a–b. (a) Teymare, central Iran (Naserifard, 2016, Copyright M. Naserifard). (b) Koochari, central Iran. Copyright Ebrahim Karimi.

116

Figure 7.5. The battle scene of Kal-Jangal (Photo credit: SKCHTO.ir).

117

Figure 8.1. Map showing the main regions of rock art in India.

123

Figure 8.2a–d. (a) Mythological scene of a warrior fighting a horned lion. Between them is a man who has just lost his head and his arms. Churna (Pachmari, Madhya Pradesh). Historic. (b) In this complex panel of Chaturbujnath Nala B17 (Madhya Pradesh), at top left a line of people wearing elaborate headgear and wielding axes and shields are dancing to the sound of big drums. Neolithic. Enhanced with DStretch (-lre). (c) At Bhimbetka 4 (Madhya Pradesh), to the right of a warrior armed with a bow and arrows in front of a mounted horse, two men with elaborate headgear are dancing while playing small drums connected by a wooden or bamboo bar. Historic. Enhanced with Dstretch (-lre). (d) At Urden 29 (Madhya Pradesh), a crouching man is playing a harp while a big man standing at left is playing an instrument and dancing. Two dancers are holding hands in the middle and two others, at right, are dancing in a sitting position. Four of those characters have open mouths (singing?). Historic. Enhanced with Dstretch (-lre).

124

Figure 8.3a–d. (a) At Kharwai 5 (Madhya Pradesh), two long rows of dancers are converging toward a harp player at left. Below them are smaller people probably also dancing. Historic. Enhanced with Dstretch (-lre). (b) Samardha (Madhya Pradesh). At top left, two men sitting are playing harps. Below them a long row of eight dancers holding each other are facing left where another harp player is sitting. The white bodies of the eight dancers have been completed and enhanced with brownish paint (long hair, neck and upper part of the body, skirts). Historic. (c) At Lekhamada 1 (Chhattisgarh), two harps players face each other at left. They are faced by seven dancers who are holding each other by the waist. Historic. Enhanced with Dstretch (-lre). (d) Chaturbujnath Nala B17 (Madhya Pradesh). A strange scene involves several characters. To the left, a man with a head with raised hair is playing a big harp in front of an elongated woman with the same head holding something and dancing. To their right are two extraordinary big figures. The first one, with four dots on her face has got the same hair as the other two. Is it singing/ dancing? The long vertical motif attached to its right cannot be determined as a human. Medieval. Enhanced with Dstretch (-lre).

126

Figure 8.4a–d. (a) Churna (Madhya Pradesh). At right and facing left, a man painted in solid red is playing a long pipe. Another man painted in the same way, at left, seems to be holding the same kind of pipe and perhaps dancing. There are many superimpositions on this panel. Historic. Enhanced with Dstretch (-lre). (b) At Tapka Pani (Madhya Pradesh), a drummer and a pipe4 are playing for a row of dancers holding each others’ shoulders. The dancer at the end of the row at left is brandishing a pipe or a stick. Historic. Enhanced with Dstretch (-lre). (c) At Chiklod (Madhya Pradesh), two men are playing drums behind a mounted elephant and numerous warriors. Enhanced with Dstretch (-lre). (d) Penganwa 4 (Madhya Pradesh). Badly preserved images of cymbal players. Historic.

127

Figure 8.5a–d. (a) Urden 19 (Madhya Pradesh). Two women and a child with decorated bodies, dancing with their arms outspread. Medieval. Enhanced with Dstretch (-lre). (b) Singar Pathar (Chhattisgarh). Two yellow and red rows of dancers facing right in the same stance, with bent arms held in front and knees half bent. The many handprints are from a later time. Mesolithic? Enhanced with Dstretch (-lre). (c) At Kathotiya (Madhya Pradesh), the top row of a scene with many

x

ILLUSTRATIONS

dancers and musicians could represent a puberty dance. Enhanced with Dstretch (-lre). (d) Sita Lekhni 4 (Chhattisgarh). Three rows of numerous slender dancers with bodies in an S shape face right (middle row) and left (upper and lower rows). Mesolithic.

128

Figure 8.6a–d. (a) Bori (Madhya Pradesh). Elegant dancers whose eyes are represented. They are inside a tent (denoted by white outlines) where quivers, swords, and shields are hanging. Medieval. (b) Bori (Madhya Pradesh). This long shelter features many dancing scenes belonging to different periods. This scene, painted white with brownish infilling, shows numerous dancers with elongated headgear. Below, two dancers are holding a bow and arrows, a wild boar, and a big fish. Mesolithic. (c) Bori (Madhya Pradesh). On this panel we can see three dancing scenes belonging to different periods (Mesolithic, Historic, and Medieval). (d) Belkandhar 1 (Madhya Pradesh). White woman dancing and singing. Historic.

129

Figure 8.7a–d. (a) Hathitol (Madhya Pradesh). S-bodied dancers with elaborate headgear holding hands. Mesolithic. (b) Bhimbetka (Madhya Pradesh). Mesolithic dancers painted in green. Enhanced with Dstretch (-lab). (c) At Kathotiya (Madhya Pradesh), a big bison, with a smaller animal inside its body, is superimposed over a row of green dancers. (d) Chaturbujnath Nala B11 (Madhya Pradesh). A number of dancers holding a stick in each hand to make music. Most of them seem to wear two bison horns as a headdress. Historic. Enhanced with Dstretch (-lre).

131

Figure 9.1. Location of the Jebel Qurma area in northeastern Jordan, in the Black Desert or harra. Copyright Jebel Qurma Archaeological Landscape Project.

135

Figure 9.2. An engraved panel from Jebel Qurma with a view of the desert landscape in the background. The panel features three Safaitic inscriptions, two of which refer to ‘the she-camel’. On the left a carnivoran, four ibex, and an ostrich are depicted. In the scene on the right, a human figure holds a camel that is nursing her infant. Copyright Jebel Qurma Archaeological Landscape Project.

136

Figure 9.3. Stacked bar chart showing the different types of scenes and the number of times they occur in the Jebel Qurma corpus. Scenes that could clearly be identified are displayed in dark gray; scenes that were unclear are displayed in light gray. The “Unknown” category features compositions that are clearly scenes, but of which the type is unknown. Copyright Nathalie Østerled Brusgaard and Keshia Akkermans.

137

Figure 9.4a–d. (a) A cooperative hunting scene depicting three archers (on the left) and two dogs (on the right) hunting a herd of bovids, either ibex or gazelles. The bovids have striped patterning on their bodies. Copyright Jebel Qurma Archaeological Landscape Project. (b) A solitary hunting scene in which an archer, upside down, faces a male wild ass. The seven dots on the right are a common motif in the rock art. The inscription on the right is made using the same technique as the archer and refers to the wild ass (“the ass”). Copyright Jebel Qurma Archaeological Landscape Project. (c) A rider on an equid hunts a lone oryx, possibly with the help of an archer (bottom right). The inscription, only partially visible here, refers to “the she-ass.” The equid is therefore probably a female mule. It also has patterning on its body. Copyright Jebel Qurma Archaeological Landscape Project. (d) Two canids surround a flock of ostriches. The inscription states the carver’s name. Copyright Jebel Qurma Archaeological Landscape Project.

139

Figure 9.5a–d. (a) A male horse and its rider, armed with a lance, face a man with a bow and arrow. The inscription refers to “the horse(man).” Copyright Jebel Qurma Archaeological Landscape Project. (b) A possible raiding scene. One camel rider touches the dromedary camel with his or her lance/spear, possibly “claiming” the dromedary as booty. Copyright Jebel Qurma Archaeological Landscape Project. (c) A conflict scene that might depict raiding. Two thinly incised archers holding shields are on the bottom left. One appears to be holding the dromedary camel by a long lead rope. The dromedary’s foreleg is hobbled, indicated by the raised leg. A figure on an

ILLUSTRATIONS

xi

equid and another thinly incised archer appear to be the attackers. The rider also has a bow and a shield and a quiver on his or her back. It is the only figure on horseback with a bow. Copyright Jebel Qurma Archaeological Landscape Project. (d) Another possible raiding scene. A figure holds the dromedary camel and a bow, facing toward the two riders on horseback. Both riders have a lance in their hands. Another archer appears to be standing on the dromedary’s back or next to it, perhaps trying to ward off the attackers. A smaller dromedary stands underneath the larger one. On the left is a small hunting scene of an archer hunting an oryx. Copyright Jebel Qurma Archaeological Landscape Project.

141

Figure 9.6a–e. (a) Stacked bar chart showing the types of weapons, including combinations, and number of anthropomorphic figures holding them. The unidentifiable objects (“uncl.”) are also included. Figures mounted on an animal are displayed in light gray, and figures on foot are indicated by dark gray. Copyright Nathalie Østerled Brusgaard and Keshia Akkermans. (b) Depiction of an archer with a double-recurve composite bow, which is characterized by its “m” shape. Also note the quiver carried on the figure’s back, with the arrows sticking out from the top. Copyright Jebel Qurma Archaeological Landscape Project. (c) Figure with a spear/lance seated on an equid. This object is most likely to be a lance due to its length in relation to the figure holding it. To the right a figure is depicted holding the equid by a lead rope. Copyright Jebel Qurma Archaeological Landscape Project. (d) Rider on an equid holding a lance and carrying a sword at the waist (the sword is traced in dark gray for clarity). Note the way the sword is carried at the waist and the short diagonal bar crossing the shaft. The rider is hunting an oryx. Copyright Jebel Qurma Archaeological Landscape Project. (e) Example of a small round shield with a cross motif on it. The other object the anthropomorph is holding is probably a spear rather than a sword, based on the context of this figure (depicted in a scene of several anthropomorphs hunting/attacking a lion). Copyright Jebel Qurma Archaeological Landscape Project.

143

Figure 10.1. Scene of Gwion figures, Mitchell River, Kimberley. Copyright June Ross.

153

Figure 10.2. A successful hunting scene, Dynamic Figures, Lawley River, Kimberley. Copyright June Ross.

154

Figure 10.3. Track scene, Atalpi, George Gill Range, Central Australia. Copyright June Ross.

156

Figure 10.4. Gestural kangaroo track lines, Kathleen Domes, George Gill Range, Central Australia. Copyright June Ross.

157

Figure 11.1. The Lightning Brothers at Yiwarlarlay 1. Photo by Bruno David.

165

Figure 11.2a–c. Rock art motifs at Yiwarlarlay 1. (a) Dreaming beings (animals) hunted by the Lightning Brothers. (b) Garlaja (younger brother). (c) Ngajamberle, the Lightning Brothers’ dog. Figures 11.2 a and 11.2b by Bruno David. Figure 11.2c copyright Josephine Flood.

169

Figure 11.3a–e. Rock art motifs at Delamere 3, (a) Abraded human feet. (b) Pecked human feet. (c) Painted boomerang motifs. (d) Walanu (bush potatoes). (e) View inside of the rock shelter at Delamere 3, featuring the Rainbow Serpent. Copyright Josephine Flood.

170

Figure 11.4a–f. (a) The break in the sandstone where Yagjagbula and Ganayanda were caught. (b) The split caused by the lightning. (c) and (d) The sandy plain at Yiwarlarlay, where the Lightning Brothers fought. (e) A frog who came to watch the fight. (f) Ngalanjarri (the water rock), the metamorphosed form of wiyan. Figures 11.4a, b, d by Bruno David. Figures 11.4c, e, f copyright Josephine Flood.

172

Figure 12.1. The “shaft scene” in Lascaux Cave that depicts a man with a bird-like head facing a charging and disemboweled bison with a spear through its side. Lascaux 4, Motignac, Dordogne, France. Copyright Traumrune, Wikimedia Commons, CC BY-SA 3.0.

183

xii

ILLUSTRATIONS

Figure 12.2. The rhinoceros on the Lion Panel in Chauvet Cave with multiple horns that are immediately adjacent to one another but do not converge at a single point. Ptilpv25, Wikimedia Commons, CC BY-SA 4.0.

185

Figure 12.3. The “sorcerer” at Les Trois-Frères consistent with the ritual use of masks and transformational capacities assigned to ritual practitioners. Wellcome Library, London, Wikimedia Commons, CC BY-SA 4.0.

189

Figure 13.1a–b. (a) Flutings from the ceiling of Chamber E, Rouffignac Cave, southern France, appear to be organized into specific units or images with clear spaces between them. (b) flutings from Chamber A1 of the same cave show meandering lines that some researchers have referred to as snake images (Nougier 1958; Barrière 1982: 205), water images (1977) or male symbols (Leroi-Gourhan 1958). Copyright Leslie Van Gelder.

195

Figure 13.2. Tectiforms 35 and 36, Voie Sacrée, Rouffignac Cave. Copyright Leslie Van Gelder.

198

Figure 13.3. Chamber G4 Panel, Rouffignac Cave. Copyright Leslie Van Gelder.

199

Figure 13.4. Rhinoceros Panel, Rouffignac Cave. Copyright Leslie Van Gelder.

200

Figure 13.5. Saiga Panel, Grande Fosse, Rouffignac Cave. Copyright Leslie Van Gelder.

201

Figure 13.6. Navigational lines, Voie Sacrée, Rouffignac Cave. Copyright Leslie Van Gelder.

202

Figure 13.7. Patriarch Panel, Galerie H. Breuil, Rouffignac Cave. Copyright Leslie Van Gelder.

202

Figure 13.8. Panel of dots at Chauvet Cave, France, dating to approximately 34,000 years ago, described variously as a bison, a rhinoceros (Mohen 2002), or a mammoth (Clottes 2010). The image was created by several people dipping their palms into pigment and then pressing their hands onto the cave wall to form a coherent image. Copyright Professor Jean Clottes.

204

Figure 14.1. Map from Block 1 of the cave of Abauntz and surrounding landscapes.

209

Figure 14.2a–c. (a) Tracing of the palimpsest of the cave of Gargas (after Barrière 1976). (b) Selected details of the Gargas palimpsest. (c) Map of the real course of the Garonne and Neste next to the location of the cave (OpenStreetMaps).

211

Figure 14.3a–b. (a) Satellite image of the Pyrenees and location of the caves (Gargas and Fuente del Trucho) (Wikimedia Commons). (b) Tracing of Panel XV. Hand stencils could mark locations of caves, while dot series would represent passages or routes through the Pyrenees.

213

Figure 14.4a–c. (a) General plan of the Lascaux Cave (from Leroi-Gourhan 1984: 181); (b) Detailed plan of the “Cabinet des felides” of Lascaux Cave (modified after Leroi-Gourhan and Allain 1979: 61); (c) Tracing of the decorated bone of La Rochette (after Beinhauer, 1986: 142).

216

Figure 14.5a–b. (a) (top). Location of the decorated block in Tito Bustillo. (b) (bottom). Detail of the block and dots and lines depicted on it. Their location on the block’s surface seems to have a distribution pattern quite similar to the one observed for major decorated caves in the territory (after Ruiz 2007).

217

Figure 15.1a–c. (a) Parpalló, tracing of Plaquette 16079A. Lower Solutrean. The anatomical details and the disposition of the two animals, a stag and a doe, lead to the identification of a precoupling scene. (b) Parpalló, tracing of Plaquette 16182. Middle Solutrean. Maternal scene of a doe suckling her fawn. The internal filling of the fawn by means of paired traces showcases the whitish spots found on young deer. (c) Detail of the previous piece. Double representation of the head of the fawn: one complete but not reaching the maternal abdomen and the other partial but reaching the mother’s body. Copyright Valentín Villaverde.

226

Figure 15.2a–b. (a) Parpalló, plaquette 16341. Middle Solutrean. Lynx jumping toward the neck of a goat. (b) Parpalló, plaquette 17757A. Evolved Solutrean. A doe walking with her two young.

ILLUSTRATIONS

xiii

The three animals rest on the same imaginary ground and are engraved using the same technique of multiple tracing. Copyright © Valentín Villaverde.

228

Figure 15.3a–d. (a) An archer integrated into the hunting scene of de La Palla rock shelter. Copyright Archivo del IVCR+i. (b) Tracing of the complete scene of de La Palla rock shelter (the temporal narrative must be read from bottom to top; for more details see the text) (Hernández Pérez et al. 1988). (c) Rock shelter VIII of Cingle de la Mola Remigia. DStrech enhancement (ybk filter) of photograph in 15D, with a scene in which a fallen individual is carried in the arms of his colleague (d) Photograph of the previous image. Copyright Valentín Villaverde.

231

Figure 15.4a–b. (a) La Araña, second cavity. Tracing of the hunting scene of a herd of ibex (Hernández Pacheco 1924). (b) Photographic detail of the previous scene. Copyright Archivo del IVCR+i.

233

Figure 15.5a–c. (a) Cova dels Cavalls. Tracing of the complete hunting scene of a herd of deer. Human archers were added at different times, transforming the character of the scene (Martínez Valle and Villaverde 2002). (b) Archer 25 of the previous tracing. This is one of the archers situated at the left of the scene, awaiting the arrival of the herd. (c) Detail of the deer herd. Two fawns with typical whitish spots are represented in the center of the herd, surrounded by several females. Copyright © Valentín Villaverde.

234

Figure 16.1. Geographic distribution of (1) Macroschematic Art (MA), (2) Levantine Rock Art (LRA), and (3) Schematic Art (SA) in Iberia. LRA is exclusive to the eastern side of the Iberian Peninsula, MS is limited to Alicante province, and SA appears throughout Iberia. Base map created by NASA, MODIS / LANCE. Public domain.

241

Figure 16.2. Archer from la Covatina site (Vilafranca). Motion is suggested by the exaggerated length of the figure’s stride, the sense of wind blowing as suggested by the figure’s clothing and the diagonal plane of representation. Copyright Inés Domingo Sanz.

243

Figure 16.3. Las Monteses site (Jalance). Is the figure on the left holding a hafted stone axe? Copyright Inés Domingo Sanz.

244

Figure 16.4. Ancient hunting tactics as depicted in LRA. Copyright Inés Domingo Sanz.

247

Figure 16.5. Evolution of styles and subject matters in LRA. Copyright Inés Domingo Sanz.

248

Figure 17.1. A solitary figure. Detail of figured surface, Seradina 12. Copyright Alberto Marretta. The motif is a human, with the manner of depiction of a common variant. Notice the faint marks above the figure, which are perhaps eroded traces of pecking. It may be that this was not a solitary figure but actually one with a companion, now weathered away, leaving the appearance of a solitary figure. That there are so very few solitary figures on this vast surface with its many hundred figures underlines how rarely figures are solitary.

263

Figure 17.2. A figure cluster. Detail of figured surface, Seradina 12. Copyright Alberto Marretta. Notice how the cluster is surrounded by an extensive area without figures, that is, the area of the cluster is defined. Notice that within the cluster are three quite similar human figures, each depicted in much the same way, holding in one outstretched arm a spear with its blade uppermost, and in the other a round shield depicted as a semicircle. Notice (below) that an array is defined as a group of similar figures spaced reasonably and regularly. Perhaps the three figures in the cluster make an array. The authors are very much aware that care will be required in the use of their analytical classes, and that some adjustment of their definitions will likely be necessary.

265

Figure 17.3. An array of figures. Detail of figured surface Seradina 12. Copyright Alberto Marretta. Amongst the figures are many images of humans depicted in the same manner: arms stretched out and curving down, legs well apart with feet turned out. These form an array. Other

xiv

ILLUSTRATIONS

subjects present—a human on horseback, dogs, a spear, humans holding a spear, two opposed warriors—are not part of the array but are within the same part of the surface.

267

Figure 17.4a–b. (a) Graphic design: statue-stele, Bagnolo 2. (After E. Anati, 1990, Fig. 224.). (b) Graphic design: rock art on bedrock, “monumental composition.” Plas Roccia 1: Capitello dei Due Pini, (After S. Casini, ed., 1994). Figures in graphic designs, examples of the two variants as distinguished in the research literature: the statue-stele on a small, rounded, portable boulder, often of broadly cylindrical shape; and the “monumental composition,” a comparable design differing only in that it is placed on bedrock, or on a large detached boulder that offers a large surface comparable to a bedrock panel.

268

Figure 17.5. Scenes. Detail of figured surface Seradina 12. Copyright Alberto Marretta. In this small area are several separate scenes worth describing and defining in a necessarily lengthy caption. The two scenes of ploughing (center-right and below) are contiguous-figure scenes, in which the whole scene with its leading human with mattock, the two beasts they lead, the yoke and shaft of the plough, the digging arm of the plough, and the ploughing human behind the plough are all depicted with a single continuous line. At top left are two pairs of opposed warriors, adjacent-figure scenes in which the figures are in a symmetrical opposition. At top right is a scene both symmetrical and sequential: two dogs on the left confront a dog on the right. Here, on a few square centimeters of a single surface, one sees the complexity in scenes, which is both our duty and joy as researchers to understand in a fitting rationality.

270

Figure 17.6. Narrative? Detail of figured surface Seradina 12. Copyright Alberto Marretta. Two adjacent-figure scenes: deer-hunting narratives. Also other figures variously placed, some in configurations that our other five classes may usefully define.

272

Figure 18.1. The typological norm for the Fremont Complex in Nine Mile Canyon, as suggested by Schaafsma, is a trapezoidal or rectangular anthropomorph. These actually represent a minority of the anthropomorphs here. Photo courtesy of the Colorado Plateau Archaeological Alliance.

280

Figure 18.2. The Sandhill Crane Site at the mouth of Currant Canyon exhibits numerous augmentations over time, as evidenced by superimposition and different levels of re-patination. Photo courtesy of the Colorado Plateau Archaeological Alliance.

282

Figure 18.3. The Family Panel near Devils Canyon is typical of most Nine Mile clusters in that it features fewer than twenty images, the images are arranged so as to avoid overlap with other figures, and identical re-patination suggests the images were made over a relatively short period of time. Photo courtesy of the Colorado Plateau Archaeological Alliance.

283

Figure 18.4. The unusually large, detailed elk figure at Warrior Ridge is unique in that no similar animal representations have yet been identified in the absence of anthropomorphic or geometric representations. Jerry Spangler shows how large the image is. Also shown is a panel above the elk that is probably not part of the same act of making. Copyright Iain Davidson.

287

Figure 18.5. This is one of more than a dozen different scenes depicting conflict at Warrior Ridge, especially above the large anthropomorph. Note that despite the high number of interacting images, there appears to be little overlap of images. The image also shows patterns of dots in rows, squares, rectangles, or polygons that are extremely common in Nine Mile Canyon but rare outside of it. Image courtesy of the Colorado Plateau Archaeological Alliance.

290

Figure 18.6a–b. (a) Backpacker figures are extremely rare outside of Nine Mile Canyon and might be considered an identifier unique to the canyon. They are found near Desbrough Canyon in direct association with several trapezoidal anthropomorphs. Photo courtesy of the Colorado Plateau Archaeological Alliance. (b) There are more than two hundred figures in Nine Mile Canyon where undulating lines and spirals terminate with a head featuring what appear to be horns (and occasionally antlers). The horned snake motif is found throughout North America but is especially

ILLUSTRATIONS

xv

common among Puebloan peoples, who associate it with the Water Clan. Photo courtesy of the Colorado Plateau Archaeological Alliance.

291

Figure 19.1. View from the position of the petroglyph boulder to the southeast, along the southern shore of French Lake, Oromocto River, New Brunswick. Copyright Bryn Tapper and Oscar Moro Abadía.

298

Figure 19.2. The French Lake petroglyph boulder artificially lit using the Highlight-Reflectance Transformation Imaging technique. Copyright Bryn Tapper and Oscar Moro Abadía.

299

Figure 19.3. Palimpsest of petroglyphs from McGowan Lake, Nova Scotia (see Molyneaux 1989: 20, Fig. 8.1) Copyright Ethnology Collection, Nova Scotia Museum (Ref. BcDg-4, 1,177).

301

Figure 19.4. Hunting scene from Fairy Bay, Kejimkujik Lake. The quadruped appears to have been rendered variously as a moose, a caribou, or even a horse in the past. Copyright Brynn Tapper and Oscar Moro Abadía.

302

Figure 19.5. Canoe motif, with two occupants, juxtaposed with a serpent motif at Fairy Bay, Kejimkujik Lake. The canoe is partially obscured by modern graffiti. The inset shows the line drawing interpretation of the two motifs discussed. Copyright Bryn Tapper and Oscar Moro Abadía.

304

Figure 19.6. A reproduction of the petroglyph tracing of a small sailing vessel depicted in association with a serpent motif with bifurcated tongue and teeth-filled jaws from Fairy Bay, Kejimkujik Lake. Copyright Bryn Tapper and Oscar Moro Abadía.

305

Figure 20.1. Northwestern Santa Cruz archeological sites, Patagonia, Argentina. Copyright Carlos Aschero.

311

Figure 20.2. View of Rio Pinturas Canyon, Santa Cruz. Copyright Carlos Aschero.

312

Figure 20.3. Interception of prey with lasso-balls with handle (bottom left). Paredón de las Escenas, Cueva de las Manos. Copyright Carlos Aschero.

318

Figure 20.4. Animal drive during a hunt, with the hunter with a long single arm. Paredón de las Escenas, Cueva de las Manos, trat. Enhanced with DStretch. Copyright Carlos Aschero.

320

Figure 20.5. One hunter and sixteen guanacos running toward the left. Alero de las Escenas, Cueva de las Manos. Copyright Carlos Aschero.

321

Figure 21.1. Hunting scene (detail) from Leang Bulu’ Sipong 4 (Photo credit Ratno Sardi) with permission from Maxime Aubert and Adam Brumm.

330

Tables Table 9.1. The types of scenes depicted in the Jebel Qurma rock art and their frequency.

137

Table 9.2. The types of objects and combination of objects held by anthropomorphs on foot and held by riders.

142

Table 11.1. Motif counts and techniques at Yiwarlarlay, by site.

166

Table 12.1. Perceptual entailments and visual markers of narrative events.

181

Table 12.2. Number of assemblages from portable and parietal contexts evaluated in study.

187

Table 12.3. Number of scenes identified in portable and parietal contexts with mean number of scenes per image assemblage (x¯).

188

Table 12.4. Number of scenes identified on all pre-Magdalenian and Magdalenian portable and parietal contexts with mean number of scenes per image assemblage (x¯).

188

xvi

ILLUSTRATIONS

Table 12.5. Percentage of scenes depicting ritualized humans, other humans, or only animals.

189

Table 13.1. List of individual finger fluters documented at Rouffignac Cave.

197

Table 15.1. Parpalló animal associations. PM: Pre-Magdalenian levels (25,000-17,000 BP), M: Magdalenian levels (17,000-12,000 BP), DA: Number of associated animals, AR: Number of represented animals, %: Percentage of AA in relation of the number in DR.

229

Table 18.1. Gross rock art data based on upper, middle and lower survey blocks and tabulated by images per site and images per cluster (panel).

281

Table 18.2. Rock art site location related to distance above the floodplain.

285

Table 18.3. Relative rock art complexity per rock art panel for each of the five survey blocks.

286

Table 18.4. Tabulations of image types by broad categories (quadrupeds, anthropomorphs, and abstract unknown) for each of the five survey blocks.

288

Table 18.5. Tabulations of quadruped images by identifiable animal for each of the five survey blocks.

289

Table 18.6. Tabulations of anthropomorphic images by image shape for each of the five survey blocks.

289

Table 18.7. Tabulation of abstract/unknown images by broad categories (rectilinear, curvilinear, and other).

290

Table 18.8. Tabulation of non-anthropomorph and non-zoomorphic images that are repetitive throughout the canyon organized by survey block.

292

Table 20.1. Distribution of Black series in Cueva de las Manos sites. Table created by Carlos Aschero.

315

Table 20.2. Representations of biomorphology of animals and humans. Table created by Patricia Schneier.

317

Table 20.3. Frequency of motif types within the site. Table created by Paricia Schneier.

319

==== •

•

•

PREFACE Reimagining Scenes in Image-Making Practices Margaret W. Conkey

This is one of those volumes that ask one to go back to the proverbial square one and reimagine everything—from one’s approach and presuppositions to the images we study, no matter what time or place they were made, and what, if any, generalizations we can make about even a well-known/understood corpus of archaeological or ethnographic materials at hand. Not everyone can do this, but, taken together, these chapters give us some excellent, if not sometimes challenging, guidelines for getting started on this now necessary reimagining. Perhaps the most critical takeaways from the chapters in this volume include the following: (1) the archaeology of “art” and image making can never be finalized or fully realized, especially in terms of western epistemologies and ontologies; (2) the vocabulary of our research, such as the term “scenes,” is irrevocably more complicated, but this is its strength; and (3) we must always try to expand our scope of reference, add examples, and embrace diversity in how to think about things. Let me push a bit on these. First, on the one hand, our audiences, if not our colleagues, always demand answers to the question “what does it mean?” Indeed, this has been the single most frequent query in the presentations I have given over the years, usually on the images of the European Upper Paleolithic. But on the other hand, as discoveries and research grow, expand, and accumulate, it becomes increasingly less likely that we can answer that question, and certainly not with an encompassing, accommodating narrative. Perhaps, I have often thought, the audiences expect that such things can be settled with materials from Europe (even though the images of the Upper Paleolithic are not within the so-called European traditions of “art” despite the unfortunate collapse of our geographies with cultures), whereas the “meanings” of image-making practices from places like Africa or the Pacific Islands, might not be so easily understood. Despite the obvious illogical thinking here, it remains that there is an increasing distance between the possibilities of “nailing down” meanings and the expansion of research, documentation, and interpretive possibilities. To me, this is neither surprising nor unwelcome. It has become wonderfully less possible to nail things down, and for all sorts of equally wonderful reasons. We researchers have got to learn not only to live with ambiguity, but to successfully communicate that fact to our colleagues and audiences. Anyone who expected to get a catalogue of scenes in image-making practices out of this volume, or to be assured that we can now be shown many scenes—despite the long-standing assumption that they are few in certain image-making traditions (e.g., the European Paleolithic arts)—will be disappointed indeed. Rather, we are confronted, in often detailed and specified ways, with the understanding that what constitutes a scene is hardly straightforward and any meanings to be derived therefrom are elusive or just puzzling. But this is to be celebrated: people and image-makers, past or recent, were not robots in following a narrow, mutually comprehensible cultural norm. What people “do” with images varies among and within groups small or large. The intentionalities, manifestations, actions, and agencies are not just set in varied cultural contexts, but the images themselves are often open systems that engage with forms, locations, media, cultural expectations, and parameters, even if they fall within what we might designate a “tradition.” This is, to me, liberating! We no longer have to show that there is an understandable and simple coherence; rather, we may argue that maybe, just maybe, things are now more complicated, more human, more creative and innovative, and this complication is itself integral to the story, to the significations, and to the processes of making images. To “get” an account along these lines, read on!

xviii

PREFACE

Second, my suggestion that our vocabularies are not immune to challenge and mutability is perhaps a core contribution of this volume. Our editors are clear when they relate to the readers that they had one intention in mind: to challenge and “show otherwise” the assumption that there are no or few “scenes” in the European Paleolithic arts. Yet once one gets engaged with what one needs to do and to think about in approaching this issue, the only way to do it fully is to open it up to other notions or definitions of scene, to other case studies, other approaches, other sets of images, and so on. This is a courageous move, and the results here are thus more informative and enlightening even if there is not, and nor can there be, a single definition, a constrained formula, or even that within any one so-called tradition, there may be varied and inconsistent illustrations. After all, the very exercise of just asking different researchers working with different materials from a large range of cultural contexts and time periods has shown here there can be no one definition of what a scene is, which in itself confounds the very inquiry into the question. But this sets the table for a more focused and well-structured research program. We will learn more, and we will learn new things, generating more questions that emerge from the reframings required. Where did the term scenes come from, anyway? Whose vocabulary are we using, and how does it lead into only certain kinds of research and thus only certain kinds of “answers” or observations? This is, of course, not limited to inquiry into scenes; instead it extends into all of our analytical categories and terminologies. No one needs much sophistication in epistemology or ontology to recognize that this is part of research, but to see how it plays out in this particular case of scenes is more than just instructive in assessing our methodologies. It has expanded not just how we know what we know (epistemology) but also how we think about our research, what it can lead to, and how we can expand our frameworks by asking just what are we assuming about the “ways of being in the world” of those we study and of ourselves. So, lastly, what does this kind of a volume and collection of chapters “mean”—not only for what such systems and practices of the visual arts of other groups in the deep and recent past and present, but also for how we reimagine the research at hand, the research of the future, and the research and “findings” of our intellectual predecessors? Here we have a double bonus—yes, in multiple ways the chapters in this collection are revelatory about image-making practices in many different contexts, but at the same time they are also a palpable lesson on the intellectual capacities and capabilities of our own probing into the visual worlds of our subjects. I learned several new ways to think about such a basic tool as the recording of images and to reflect on why the notion of a scene is not an unquestionable analytical given. If we cannot all agree on the viability or the varieties of the term scene, in what other ways can we engage each other? Here the very diversity of approaches and differences of opinions have provided a bridge to a very much more nuanced, illuminated, and generative set of insights into not just the subject matters at hand—the image-making—but also ways to grow our research capabilities and possibilities, which are crucial to a healthy, viable future for the radiating trajectories of our fields. The scene is complicated and complicating, but that is its strength, as shown with rigor and vigor herein. Margaret W. Conkey is the Class of 1960 Professor Emerita of Anthropology at the University of California, Berkeley, where she has taught since 1987. Her fields of research include ongoing survey and excavation of Upper Paleolithic sites in the French Midi-Pyrénées, and her scholarship has taken up issues of prehistoric social geography, the production and interpretation of prehistoric (especially Paleolithic) arts and visual culture, and the feminist practice of archaeology, including pioneering collaborative work on archaeology and the study of gender.

==== •

•

•

Introduction BEHIND THE SCENES Did Scenes in Rock Art Create New Ways of Seeing the World? Iain Davidson and April Nowell

Introduction One of the motivations for this collection of papers was articulated by one of us in an earlier publication that was an exploration of Paleolithic images of animals (Davidson 2017a, 22): It seems likely that there is an argument to be developed here about the emergence of the ‘Western’ styles of scene representation (which is by no means confined to Western rock art traditions). Just as the emergence of naturalism through the application of perspective is said to have created new ways of representing and seeing the world during the European Renaissance, so changes in the ways images of animals were represented with other animals probably testify to changes in the ways people saw the world.

The initial intent was to explore the question of scenes in the Paleolithic broadly, but then the question was expanded to include rock and cave art from later periods. It has been traditional to state that there are few representations of scenes in the Upper Paleolithic Cave art of Western Europe. Davidson (Ch. 1) reviews some of the ways the absence of scenes in Paleolithic art has been represented in textbooks over the last sixty years or more. In general, it has persistently proved to be true that scenes do not appear to be common in the art on the cave walls. On the other hand, Davidson (Ch. 1), Culley (Ch. 12) and Villaverde (Ch. 15) demonstrate that the view is distorted by the concentration on cave art to the neglect of portable art that is contemporary with it. Van Gelder and Nowell (Ch. 13) show also that the distortion derives from emphasizing representations of animals at the expense of other markings on the cave walls. When attention is turned to images engraved on bones or on plaquettes of stone or to more nuanced understandings of what constitutes a scene, scenes are not so rare. This suggested that the presence or absence of scenes might help reveal how the image making was used by the societies of the artists. Importantly, recent work by Fritz, Tossello, and Lenssen-Erz (2013) has addressed the problem of the lack of conventional scenes in cave art, identifying some instances where animals seem to have been represented with the ground on which they would be seen. The project, then, had its beginnings with one particular definition of how a scene might be recognized and has morphed, through the successive definitions by different authors in the book, into a broader discussion of scenes in rock art. The hope is that our broadening can contribute to correcting ideas about scenes that took hold early and have persisted despite general knowledge of exceptions that proved those ideas wrong. Kelly and David (Ch. 4) outline one history of the concept of scenes in rock art, and Lenssen-Erz and colleagues (Ch. 6) also address that history. Rock art is often said to be universal among modern humans, though it is slightly more difficult to define “universal” and to document that. The book includes material from a period of sixty thousand years and from six continents (Figure 0.1). It includes contributions from Europe (Davidson, Ch. 1; Culley, Ch. 12; Villaverde, Ch. 15; Domingo Sanz, Ch. 16; Van Gelder and Nowell, Ch. 13; Alexander, et al. Ch. 17), Asia (Brusgaard and Akkermans, Ch. 9; Karimi, Ch. 7; Dubey-Pathak and Clottes, Ch. 8), Africa (McCall et al, Ch. 5; Lenssen-Erz et al, Ch. 6), Australia (Ross, Ch. 10; Domingo Sanz, Ch. 16; Kelly et al., Ch. 11),

2

IAIN DAVIDSON AND APRIL NOWELL

FIGURE 0.1. Map showing dots for the location of chapters (by number) that have a regional focus.

North America (Spangler and Davidson, Ch. 18; Tapper and Moro, Ch. 19) and South America (Aschero and Schneier, Ch. 20). If nothing else, this account emphasizes the gaps in the coverage (e.g., Europe [other than France and Iberia], East and Southeast Asia, Africa beyond the south, northern South America), which is a good thing. We sought to fill some of these gaps, but for a variety of reasons the authors we invited to achieve this were unable to contribute to this volume. The necessary place to start is with considerations of the definitions of scenes in rock art. Most authors devote some space to this task, while some address it more explicitly. Once a definition has been established, there remains the question about how it is possible to interpret the significance of scenes in rock art in so many parts of the world, when the images were generated by so many people and so many cultures that have not survived. Rather than accepting that there are universal trends in the evolution of production of images, given the great diversity of location, culture and time, we must begin by trying to interpret particularities. Jean Clottes (2009) introduced a framework for looking at the meanings of rock and cave art, and Davidson (2012b) expanded it. This extended framework shapes this introduction to the volume. Finally, the question of whether the book contributes to an assessment of the role of scenes in people’s perceptions of the world will be for others to judge, but we conclude this introductory essay with some consideration of that question.

Definitions of Scenes Davidson’s (2017a) initial attempt to think about scenes produced a classification of different combinations of images that is reproduced to some degree in Chapter 1 and discussed again by Spangler and Davidson in Chapter 18. Others have used other definitions, and these are introduced in this section. Davidson (Ch. 1) introduced a definition that might be seen as a minimal one: a scene can be identified from a set of images in spatial proximity to each other from which, without any knowledge other than the images themselves, an observer can infer actions taking place among the actors represented in the images.

BEHIND THE SCENES

3

This definition, in all its simplicity, may miss the point made by Livio Dobrez (Ch. 2) that there could be circumstances in which an individual image constitutes a scene, an argument well summarized by Karimi (Ch. 7). Similarly, Kelly et al. (Ch. 11) give examples of what they regard as individual images that constitute scenes. That claim is easier to accept when there are living informants, or cultural knowledge derived from them, to address the meaning of images, but for images beyond living memory or oral history of their production, such interpretations will always be difficult. The richness of the ethnography available to Kelly et al. can seduce the observer into thinking that all rock art everywhere involved similar relationships between images, sites and landscapes. That may be true, but one of the functions of archaeological analysis is to use the evidence of the material remains to understand how such relationships came about. Several authors define what they mean by scenes. We collect some of those definitions here, but rather than try to produce a synthesis of what a scene “is,” we recognize that a workable definition will depend on the cultural context within which a researcher is working and on the questions asked—which only means that it would be confusing for readers to allow the authors to proceed without specifying a definition. Villaverde (Ch. 15) cites Delporte’s (1981) five-part scheme of what does and does not constitute a scene: 1. chance or accidental association (more common in cave art assemblages and on portable art pieces, and which entails an accumulation of different animal species within the same graphic space without obvious stylistic or thematic connection); 2. repetitive superimposition (where various animals of the same species share a graphic space, but cannot be interpreted with certainty as representing a herd or group, that is to say, they cannot be identified as a scene); 3. narrative association, or what could be termed as the configuration of a scene given that the individuals represented are taking part in a shared activity such as in scenes of hunting or mating; 4. geometric association (typically friezes where animals are often shown in line, or confrontations where the same or different species are shown opposed to each other. Both of these might be interpreted as scenes—the friezes as migrations of animals, and confrontations may be typical of fights between males of certain species during the rutting season); and 5. thematic association (rare examples in which the action cannot be specified, even when the intentionality of the artist is clear).

Villaverde concludes that scenes can only be identified if there is evidence of “discernible and identifiable action.” Livio Dobrez (Ch. 2) discusses what he believes is “the most fundamental idea of a scene as a perceived event . . . [and] the necessity of tying [a] depicted event firmly to a real event.” Importantly, he documents the neurological evidence that brains react similarly to their perceptions of both an event and the depiction of it. He argues that there are universal characteristics through which scenes can be interpreted. Patricia Dobrez (Ch. 3) tackles the thorny question of cross-cultural expectations by contrasting the unreflective judgments of Western-oriented anthropological observers Spencer and Gillen among the Aranda with an example, also from the Aranda, described by Basedow (this example is also discussed by Ross in Ch. 10). Where Spencer and Gillen had said that there were no scenes in Aranda art, Basedow found them in a plan view of the traces of an event—an emu hunt—defined by the footprints of the animals involved. Kelly and David (Ch. 4) engage in a search for definitions of scenes in a number of art and archaeology works, finding a scarcity of such definitions, which should be considered surprising. The chapter discusses the definitions by Dobrez (2011) and Lenssen-Erz (1992), whose more recent considerations are contained in Chapters 2 and 6 of this book, respectively. Kelly and David point out that formal definitions of scenes and the ways in which they mediate narratives through imagery tend to be biased toward Western visual conventions. They argue that in many places, there is indigenous knowledge that interprets the local rock art in ways that differ from those of naïve Western observers. While recognizing that those indigenous cultures may have “changed considerably since the art was made,” Kelly and David argue that the art carries special meaning to local people who do not engage with the Western traditions of interpretation; hence, the indigenous knowledge should not be ignored. This stands in contrast to Livio Dobrez’s assertion of universalist characteristics. It may also undermine any criteria by which Davidson (Ch. 1) proposes to identify agency in the images. Lenssen-Erz and his colleagues (Ch. 6) take a slightly different approach. Lenssen-Erz (1992), who has been working on the problem of scenes for more than a quarter of a century, has concluded that there is little agreement on the definition of a scene, but recognizes that narration is an important part of some defi-

4

IAIN DAVIDSON AND APRIL NOWELL

nitions. Therefore the authors review another set of definitions in Chapter 6 to identify scenes, interaction of figures, and narratives, along pragmatic lines. Ross (Ch. 10) makes a further point about the composition of scenes, observing that it is important to consider whether the scene was composed substantially in a single episode. She documents a panel in the Kimberley region where the unity of composition can be established through detailed chemical analysis of the ochres. That contrasts with an analysis by Villaverde (Ch. 15), who describes how he can document panels where an apparent scene was created by separate images produced at different times. Alexander and colleagues (Ch. 17) approach the definition of scenes from first principles, beginning with definition of “figures” (rather than images) and going on to recognize clusters of figures, and then figures in array, in graphic designs and in scenes, and finally a narrative. In a further elaboration of the question about the reasons why images might be clustered on a panel, the authors suggest that the prior use of a panel to produce figures might itself have attracted the production of others.

The Studies in This Volume, Considered within the Framework of Meanings In what follows, we consider the framework used by Jean Clottes (2009) and Davidson (2012b) to organize our discussion of the chapters in the book. First, all rock art studies deal to greater or lesser degree with the question of the content of the art itself, and, second, in seeking to go beyond the empirical towards an understanding of meaning, many rely on knowledge of ethnography that is either local to the art sites (Vinnicombe 1976) or gleaned from some supposed association with ethnography from elsewhere (Clottes and Lewis-Williams 1998). There have been two extreme positions about rock art studies: (1) that it is the specialty of those who are interested only in the rock art and their views of how it was used; and (2) that rock art is primarily of interest as art. The studies in this volume avoid these extremes and, in the third part of the framework, attempt to relate the art and the sites where it occurs to the archaeology that was contemporary with its production. The fourth and fifth approaches, which are fundamental to this book, concern how the rock art relates to the evolution of behavior in a particular region and how rock art varies between regions in relation to evolving behavior.

Scenes as Products of “the Art Itself” It is fundamental to the approaches in this volume that the first element of the framework of analyzing rock and cave art is to look at the art itself. Lenssen-Erz and his colleagues (Ch. 6) exemplify this approach. It is formal and analytical, and has involved mining painstakingly documented data from more than seventeen thousand painted figures in Namibia. This empirically driven approach epitomizes the identification of scenes from “the art itself.” One point that emerges from their analysis is that most scenes show cooperative interactions among people, and that the scenes where humans and animals interact are not explicitly of hunting. Scenes in rock and cave art may involve images of interactions between animals or between humans, between humans and animals, or more rarely, between animals and their environment. One approach has been to discuss the associations and juxtapositions of the individual images in a grouping (e.g., Davidson 2017a), an approach used by Spangler and Davidson (Ch. 18). Further examination of the issue in this volume renders this approach insufficient. Not only do scenes in rock and cave art involve different ways of considering combinations of animal images, but they also concern interactions or other connections between humans and other animals, as shown, for example, in Iran in the chapters by Karimi (Ch.7), and by Brusgaard and Akkermans (Ch. 9) in Jordan. One question that merits discussion is that of what is and what is not depicted. In general, rock art depicts little of the environment of its agents. In most of the imagery at issue here, especially that involving animals, there is very little attempt to portray the animals in a context of landscape or vegetation (see discussion of “Features of Landscapes” in Davidson 2017b). In most places there was little representation of plants, whether in landscapes or separately, except for one well-documented case in Australia (Veth et al. 2018). This observation can lead to somewhat false comparisons between post-Renaissance Western art and rock

BEHIND THE SCENES

5

and cave art, because surveys of the more recent art have not concentrated on the rather more fundamental issues that preoccupy scholars studying rock and cave art. Both landscape and vegetation have featured in European art for two thousand years, and with particular realism for six hundred years. This really is a distinctive difference from rock and cave art. Furthermore, there also exist images of people and animals that are not placed in a context of either landscape or vegetation, which might be considered a similarity with the subject matter of rock and cave art. Still, it may be better to point to the diversity of content of postRenaissance art, rather than concentrate on the different approaches to the context of landscape or vegetation. Identification of scenes is relatively simple in some rock art: some combination of animals and humans appear to be interacting with each other. This would apply to Karimi’s (Ch. 7) discussion of hunting scene petroglyphs from Iran, leaning heavily on recognition criteria defined by Dobrez (e.g., 2011, 2012). Karimi also points out that some other group images do not qualify as scenes. Brusgard and Akkermans (Ch. 9) describe textual and pictorial petroglyphs, including scenes of hunting, conflict, combat, and interactions between animals such as mating and nursing. There are different categories of scenes, depending on whether or not humans are represented. Villaverde (Ch. 15) considers art from two different contexts in eastern Iberia: the Paleolithic art on plaquettes (also referenced by Davidson, Ch. 1) and the later Levantine rock shelter art (also referenced by Domingo Sanz, Ch. 16). Villaverde (Ch. 15) presents evidence of scenes found on stone plaquettes from Parpalló, in eastern Iberia. He makes the important point that these scenes are limited, as if they were “snapshots of a well-known ethological reality” without any sense of the past or the future of the animals involved. A similar argument could be applied to the “swimming reindeer” scene from Lascaux (Aujoulat 2004, 177, 180-82) and may therefore be generalizable more widely to other European Paleolithic art. When Villaverde turns to Levantine art, he finds a much more episodic representation involving animals, people and tracks. Alexander and colleagues (Ch. 17) consider very large sets of rock art in Europe at Valcamonica in Italy and produce a hierarchy of definitions from first principles, beginning with “figure” (rather than the more generally used “image”) and leading to “group” (but not composition). For these authors, “contiguous figure scenes” are recognized when they allow the viewer to “reasonably think” they reflect a “real-world spatial order.” In other scenes, called “adjacent figure scenes,” figures are juxtaposed to appear to oppose each other, as in a fighting or hunting episode—again appearing to reflect a real-world spatial order. Aschero and Schneier (Ch. 20) offer a detailed analysis of the “Black Series” of paintings from Cueva de las Manos in Argentina, a site better known for its hand stencils. The paintings show people hunting guanaco using a set of weapons: the dart and spear-thrower, and two types of throwing stone or bola. They also reveal different tactics of the hunt, interpreted as driving the animals and intercepting them in traps created by the local topography. These panels all involve quite conventional understandings of what a scene is. Others depict scenes with rather more nuance. McCall and colleagues (Ch. 5) describe a painted rock shelter in which rain drips onto paintings that are images appearing to be associated with rainmaking in other contexts. This led the authors to a nuanced understanding of the variation of scenes and non-scenes as elements of southern African rock art. Scenes, they argue, may refer to supernatural phenomena or to real-world historical events, or they may have been part of rituals meant to influence real-world events, or they may have been part of the repetitive rituals in specialized places. In this way, the authors seek to shift the discussion away from meaning (especially the restrictions caused by suggesting that shamanism is an explanation for everything [McCall 2007]) and toward inferences about social and ritual contexts of the production of art. Utrilla (Ch. 14) discusses several recent discoveries that can be interpreted as representations of the landscape around sites, as if on a map. Many chapters show that this art was made in a context of some other archaeology situated in time and space. There might also be some convergence between the possible representation of maps and the account by Aschero and Schneier (Ch. 20) of the use of topography to guide the movements of prey. Ross (Ch. 10) shows that even without appeal to ethnography, the art of the Kimberley in northwestern Australia has enough detail to show aspects of interactions between participants in ceremonies that would be unknown without the evidence of the art. The imagery can likewise be interpreted in terms of gender roles in social and economic activity. Significantly, this is a result of associations between the individual images, going beyond anything that could be inferred from individual motifs. The nuance appears when

6

IAIN DAVIDSON AND APRIL NOWELL

Ross compares this art with art in Central Australia that is composed almost entirely of geometric and other non-representational images. As discussed below, when informed by the relevant ethnography, even these can be interpreted as scenes despite their minimal representational content. Kelly et al. (Ch. 11), also discussing Australia, show that the mythology associated with rock art in the Wardaman region promises a much richer narrative than could possibly be derived from a study limited to the images themselves, whatever their interpretation as scenes. One of the reasons for this is that the associations among the images extend beyond individual sites to provide a mythological narrative about the whole landscape. It would be very difficult to create such an account based on the images alone, once the link between the images and the culture of the painters and storytellers has been broken. In similar vein, Tapper and Moro (Ch. 19) describe the petroglyph tradition of eastern Canada and show how the choice of places and content relates to an ontology of shamanism and vision quests. In their view, much as with the Australian cases described by Kelly and colleagues, the scene does not consist of the images on the rocks alone but was included in the whole complex of places and social networks across the landscape. Finally, on this topic, Van Gelder and Nowell (Ch. 13) introduce another element in the repertoire of scenes by looking at the finger flutings of Rouffignac Cave in France. These authors, who argued that these finger flutings were a common practice in the representation of scenes, discovered that there was repetition between fluted motifs, interaction between others, and a common practice of fluting production by different individuals in yet others. The flutings, therefore, represent by proxy the interactions between the individuals who made the marks on the cave walls.

Interpretation of Rock Art: How Can We Use Ethnography of Modern People, and Does the Presence of Scenes Make a Difference? Many years ago, Wobst (1978) argued that it is erroneous to assume that ethnographers are able to understand everything about a culture because they can ask questions of their informants. Wobst (1978) warned of what he called an informant’s “worm’s eye view.” In other words, he argued that it is impossible for people to know everything about the culture in which they live. Some knowledge will remain inaccessible for social, religious, economic, or age- and sex-related reasons. All people, he continued, bring with them a particular agenda or bias or understanding of the world, which colors the information they share with anthropologists and archaeologists. The further away someone is (literally or metaphorically) from the behavior under study, the more stereotyped the response will be. Wobst makes a valid point. Nonetheless, contemporary or descendant communities remain an important source of information for archaeologists. Rock and cave art are seductive in that they seem to provide material for a narrative about the past that includes the representation of people interacting with animals, sometimes using the artefacts that can be found in the archaeology—but it can be a dangerous seduction. As the chapters here show, there is still plenty of room for selectivity in the subject matter of the art, exemplified by the attribution of scenes of hunting in eastern Iberian Levantine Art to the pastoral and agricultural Neolithic (Domingo Sanz Ch. 16). Thus it is difficult to understand the relationship between what was culturally salient generally and what was represented in the art. What was the cultural salience of rock art within the whole set of relationships of any particular culture? In the absence of consistency in such relationships, it can be adventurous to move from the modern ethnography back to the interpretation of the selected representations of the past. Moreover, ethnographic accounts of artistic systems have revealed many layers of interpretation in some societies that use art as a means of introducing young people to important cultural knowledge (see, e.g., an Australian case [Morphy 1991] and the review of it in Davidson [1995]). Art is ultimately a system of symbols in which the meanings are not plainly matched one-to-one to the images, so attempts to use modern knowledge to interpret ancient art may be quite prone to fallacy. Moreover, as Davidson (2012c) has shown, long sequences of art production, where they can be matched with the archaeology from the same time periods, can, within a system of iconography that has not changed much, show significant changes of meaning—even if we cannot tell what those meanings were. That is, within the one system there can be cultural continuity and cultural differences through time. Villaverde (Ch. 15) discusses the uses of ethnography, starting from the difficulty of translating from a known cultural context to another that is not known. In doing so, he recognizes the problems arising from

BEHIND THE SCENES

7

the occurrence of “scenes from different time periods, and social and economic contexts,” not just in the transfer from the modern ethnography to the past. Rather, as he deals both with Paleolithic art and with the post-Pleistocene Levantine art from more or less the same region, this is particularly important because the modern comparisons could apply to either of the non-modern contexts. Domingo Sanz (Ch. 16) has a different take on the same issue when considering those changes in the art from the same region: she appeals to her knowledge of the art and history of the Arnhem Land region of Northern Australia. The Aboriginal people of that region were invaded (at least) twice by outsiders, once from Sulawesi and once from Europe. Domingo Sanz points to changes in the art of both regions from an earlier predominant focus on the animal world to a later one where scenes of human interactions were more important. She suggests that the ethnographic evidence is valuable as a guide to the historical changes in the Iberian case, where the original fisher-hunter-gatherers of Paleolithic Iberia confronted invading farmers and agriculturalists. Dubey-Pathak and Clottes (Ch. 8) document several sites in India with undoubted representations of scenes, often with images of musical performance. Such performances are well documented in the region. As musical instruments are known to have been in use since the early appearance of modern humans in Europe (Conard, Malina, and Münzel 2009), and given that cave art occurs at similar dates (Quiles et al. 2016), we might expect there would be scenes of people dancing or playing musical instruments in that earliest cave art. But there are not. The flute player engraved at Les Trois Frères in the Pyrenees has often been interpreted as an image of a person playing music, but such images are rare and do not appear to be part of scenes of people affected by (i.e., dancing to) music. However unsurprising it is that there are scenes of people dancing and playing music in this body of art, the big lesson from the Indian examples is that depiction of music playing is not a universal phenomenon. What stimulates and what constitutes a scene are still culturally determined in different ways in different contexts, temporally and spatially—an observation that tends to undermine the use of ethnographic analogies from culturally different societies. In similar fashion, Brusgaard and Akkermans (Ch. 9) state that while there are no scenes of music making in their study of rock art from the Jebel Qurma in Jordan east of Azraq, there are such scenes in other parts of the Black Desert. At Wadi Salma¯ northeast of Azraq, for example, there are images of people playing flutes, drums, and lyres (al-Manaser 2018). The lesson of these two examples may be that care is warranted when applying generalizations about rock art beyond the immediate context of production. In support of the appeal to take indigenous approaches to rock art seriously, we recognize that there are many differences between scenes in Western art and almost all rock and cave art. We feel more comfortable appealing to a local ethnography than we would using assumptions deriving from more unexamined intuitions from our own cultural values. Clottes’s original list placed great emphasis on the ethnographic analysis of “hunter-gatherers elsewhere in the world,” an approach we have suggested needs to be treated with extreme caution because of the unstated assumptions about the applicability of remote ethnographies. Some authors, both in this volume and elsewhere, have leaned on ethnographic information more than others. By the same token, while most authors have a background in archaeological research and acknowledge its problems with the weight of the history of investigation of art, it is always possible that the study of archaeological art may be undertaken by scholars with a background in art historical studies. This approach is represented here by separate studies by Livio Dobrez (Ch. 2) and Patricia Dobrez (Ch. 3). Art historical studies can have problematic intuitions of their own. The issue is always about the cultural difference between the producers of the art, with their conventions for selecting themes and styles of representation, and the modern observer, whose conventions differ from the artists’ own traditions, whether in archaeology or art history. Whether these different conventions are those of archaeological classification or those of a viewer imbued with the sophistication of interpreting modern art, both sensitivities are different from the cultural conventions of the original producers. The trap here is to think that the cultural conventions of modern people from the same region may have more in common with those of the producers of very old art. As one of us pointed out previously, being French had nothing to do with the reasons the Abbé Breuil had any particular insight into the twenty thousand-year-old cave art he studied in France (Davidson 1994). One of the objectives of archaeology is to document the cultural changes that have occurred between such remote pasts and the present. Much rock art around the world has survived beyond the memories of the people who made it and the cultures they lived in, but some is still remembered in the oral traditions of ongoing cultural practice. Kelly,

8

IAIN DAVIDSON AND APRIL NOWELL

David, and Flood (Ch. 11) discussed an example from Australia that involved multiple records of oral history collected over more than sixty years by multiple non-Aboriginal people. The indigenous perspective is that images from different sites made up a narrative scene about relationships across the landscape. This would be difficult to interpret without the intimate association between the Aboriginal ethnography and those particular sites and images. Karimi (Ch. 7) states that there is no ethnographic context for the rock art from Iran that he discusses, but he does try to connect the later phases to scenes in Islamic art and notes that some panels have Islamic inscriptions. He also points out that there are some accounts of relatively recent hunters engraving hunting scenes on their own tombstones before their deaths. He furthermore notes that among nomadic pastoralists in Iran, art is also produced by painting tents. This is a crucial reminder of the significance of all the art that has not survived because it was produced on perishable materials. Importantly, Karimi mentions several interpretations of imagery in Iranian rock art that have been said to be shamanistic. In the absence of ethnographic context, he suggests that this is not a strong argument. The conclusion, therefore, is that ethnographic information is fundamental to intuitive interpretations of rock art. It is commonplace for observers to interpret images from their own experience of similar images, and if that experience included ethnographic knowledge, we include that. At some level that is only to be expected, but when trying to understand the historical or archaeohistorical meaning, that interpretation must be accompanied by an analysis of the nature of the assumptions about the links between the particular ethnography and that particular body of art. Without that analysis, there is a strong risk that the account of the rock art will be a statement about the preconceptions of the art rather than a contribution to a study of the past. Art that includes scenes is less difficult understand, but at the same time it remains problematic. The main point is that, depending on the definition of scenes—and here we adopt Davidson’s (Ch. 1) definition of the inference of actions through the agency of the individual participants—it is easy to describe the interactions between individual representations of animals. Several independent observers agree about the mating scene of the lions in Chauvet Cave discussed by Davidson (Ch. 1). That much is easy and allows some understanding without ethnographic knowledge. It still requires modern knowledge of the behavior of lions. But that clarity does not lead to greater understanding of why those images and that scene were painted in the cave.

Extracting Meaning of Rock Art Scenes from Comparisons with the Other Archaeology One of the frustrations of the archaeology associated with rock art is that there is often a divide in interests between rock art specialists and other archaeologists. The reasons for this are various and often relate to the coarseness of the chronology of rock art and the relative refinement of the chronology of other aspects of archaeology. Yet it is also highly probable that the people who were responsible for the other archaeology lived in societies that produced and used the rock art. Thus, relating the rock art aspect of those past societies to the other aspects should be an important goal. Sometimes the art can be dated by the presence of distinctive images from a particular period. Distinctive artifacts place the production of the art of Valcamonica in the Iron Age (Alexander et al. Ch. 17). Karimi (Ch. 7) says that the Iranian rock art he studies can be fixed in broad chronological terms because images of hunting dogs and horse-riding characteristic of the Bronze Age are also depicted on ceramics of the period. On the other hand, the connection between the rock art and other archaeology of the region is unpredictable, particularly because of the hunting scenes. The Bronze Age was also a period of agriculture and pastoralism—to say nothing of metalworking. Similarly, Domingo Sanz (Ch. 16) points out that the rock art of the eastern part of Iberia, known as Levantine Rock Art (LRA) represents, for the most part, wild animals and hunting scenes. Yet for a variety of reasons, the art is dated to the Neolithic period or later, during which time the economy was based principally on agriculture and pastoralism. As with the Iranian example, the social context of the rock art is not straightforwardly a representation of the daily activities of the people who made it. Brusgaard and Akkermans (Ch. 9) point out that the prevalence of hunting scenes “challenges preconceived notions” of activities in societies dependent on pastoralism rather than hunting for food. In addition, the types of weapons represented provide opportunities for comparison with artifacts known and dated in the archaeological record. By extension, what is represented in the rock art can provide insight into

BEHIND THE SCENES

9

the material culture that does not survive in the regional archaeological record. In all three of these cases, the art emphasized behavior that was not necessarily mainstream in the society, but that is not all: it also implies that the role of art in the society was not just to represent mundane daily activities—rather, at some level it was selective about what was important. The rock art of Nine Mile Canyon (Spangler and Davidson, Ch. 18) includes many varied panels, some featuring scenes with animals only, others with animals interacting with humans, sometimes in the hunt, and some scenes in which animals are interacting with humans who are dressed in animal headdresses. The authors considered the location of different categories of art sites, and the presence or absence of scenes, in relation to other types of archaeological sites and the topography of the canyon. This is a study of the art of a particular narrow period, but in other places it has been possible to consider the representation of animals in rock art in the context of environments transformed by the climatic changes after the Last Maximum Low Sea Level and peak aridity (David 2004). It seems to be universally the case that the representation of animals and the contexts in which they are shown is not a straightforward sampling of animals in any specific aspect of the artists’ lives (Vinnicombe 1972).

Scenes in the Context of Changes through Time and the Evolution of Behavior in a Region If the early art in the caves of Western Europe appears to lack scenes although scenes are present in later art in that region (broadly speaking), there might presumably be a change through time. The only chapters to address such changes explicitly are those by Villaverde (Ch. 15) and by Domingo Sanz (Ch. 16), both of whom deal in one way or another with Levantine Art of Eastern Iberia. There are studies in the book that involve the earliest rock art (Van Gelder and Nowell, Ch. 13) and Culley (Ch. 12), other art that was produced within the last millennium (Spangler and Davidson, Ch. 18, Moro and Tapper, Ch. 19); and still other art perhaps produced within living memory (Ross, Ch. 10; Domingo Sanz, Ch. 16). The art of eastern Iberia demonstrates within a restricted region both a change through archaeological time, and a change from an instantaneous vision of the world to a vision that incorporated the passage of time within the lifetime of the artists. Villaverde (Ch. 15) shows that for at least 5000 years after 20 thousand years ago, scenes represented moments in time—a freeze frame of an instantaneous interaction between the agents in the scene. About 5000-10000 years later, several panels in the stylistically different, and chronologically later Levantine art tell the story of the successive stages of a hunt. Rather like a graphic novel or a photo composite by David Hockney (e.g., Jerry Diving Sunday Feb. 28th 1982 http://www.daily artmagazine.com/david-hockney-photographs/), the juxtaposition of successive individual images has the capacity to create a unified whole. What Villaverde demonstrates here is that in a quite restricted region of eastern Iberia, there was a change from an approach to scenes involving freezing a moment in time to one showing the unfolding of a succession of events within a connected sequence. Domingo Sanz (Ch. 16) showed how there might be a parallel in the process of change in representation between the art exemplified by Villaverde’s account and completely unconnected art in Arnhem Land, Australia where it has been possible to document the changes in the art in the context of the historical social circumstances of the people who made it. The extension of the analogy from one continent to another and from one time period to another with no connection in time or space may seem adventurous, but the comparison is worth making because it highlights the sorts of social pressures that can lead to changes in rock art style—and in both examples can be easily documented to have suffered social changes. In Chapter 7, Karimi uses his model of identification of scenes, derived from the work of Livio Dobrez, to consider the art from different periods in Iran. As he points out, even though it is clear that there are panels produced at different times, identified through the differential weathering of petroglyphs, it is less certain which particular time periods were represented, except when there are clear links to Islamic art or inscriptions. Brusgaard and Akkermans (Ch. 9) point out that the themes of the rock art are often repetitive, formulaic and non-random. By implication the rock art is not a reflection of all of the society that produced it, but a deliberately selective representation of life in the region. Kelly, David and Flood (Ch. 11) show that the images and scenes in the rock art of Wardaman country in Northern Australia should not be considered in isolation, but instead are connected to each other in

10

IAIN DAVIDSON AND APRIL NOWELL

the spiritual framework of the Dreaming cosmology of the region marked by Songlines that described the travels of the Lightning Brothers across the landscape between these sites (Merlan 1989). A fundamental issue, therefore, would be consideration of when the cosmology of the Dreaming came into existence in a recognizably modern form (David 2002, 2006). This issue is made much more difficult as the mythology that connects images with landscapes incorporates images that, on archaeological evidence, were created at different times. Archaeologists concentrating only on images from later periods may miss an important point about the persistence of the images at places. The persisting importance of places can produce a narrative associated with an image from an earlier time period because the makers of later art could see those earlier images just as archaeologists can. Tapper and Moro (Ch. 19) discuss changes in the petroglyphs of eastern Canada over the period of the appearance of colonialists. They emphasize the importance of ritual performance to the pre-contact rock art ultimately with the goal of social benefit. By contrast, post-contact rock art emphasizes individual and communal biographies, and blended narratives of the changes occurring with fragmenting cultural memory and tradition. There is a more widespread tradition of rock art being used to record the otherness of strange people colonizing new lands as in the maritime representations in rock art of Australia and elsewhere (Kolpakov and Shumkin 2012, O’Connor and Arrow 2008) as well as the appearance of foreign animals such as bulls (Clegg 1984) and horses with riders (Chaloupka 1979), in which the earliest animals represented have similarities with representations of the best approximation to those animals in the endemic fauna. Whatever else this signifies, it shows that the producers of the rock art were not programmed to produce a limited range of images. Aschero and Schneier (Ch. 20) show that the long panel they analyze contains images that must be construed to have been produced at different times. This demonstrates that once a site was chosen as a place to produce art, it may have been revisited to add other images. This may be one of the game-changing aspects of rock art. Rock art images once produced had a permanence or persistence that was not likely with drawings in the ground or on many other media, such that they survived to be seen by people who may have had no connection to the narrative that was originally connected to the images. The site may, therefore, have attracted the production of more images on a subsequent occasion. In the history of art, one of the major developments in graphic representation of animals and humans was to place them in the context of their environment. This is not the place to chronicle the history of representations of invented and, later, naturalistic landscapes as either the image itself or the context for representing humans or other animals (see, for example Gombrich 1995, 113–14, 143-47, 153, 239). In Europe, realistic representations of landscape appear to be part of the Renaissance revolution due to Durer or to Altdorfer (Gombrich 1995, 345, 355–6) four or five hundred years ago. Whatever one’s views about scenes in rock art, it is a dominant feature of all rock art everywhere in the world that physical features of the environment, inanimate rocks and mountains or animate plants, are almost never represented. That might be a suitable topic for further research. Nevertheless, it has recently been suggested that, a change through time, an innovation, occurred during the last five thousand years of Upper Paleolithic cave art in a tiny area of central France. At three sites less than 150 km apart there were images that can be interpreted as representing the “ground-line” on which animals walked (Fritz, Tosello, and Conkey 2016). As with so much of the inventiveness and sophistication of Upper Paleolithic art, this innovation did not apparently contribute to later artistic traditions. Nevertheless, it might be taken as an indication of the cultural processes that led from using images of people and animals to material representations of them in context. That does not mean that there was no spoken narrative about the environmental context, but that such narratives did not have material form or persist after the death of the cultural tradition. We note that in two books about landscape and archaeology, both of them concentrating to greater or lesser degree on rock art, the discussion of landscape concerns the situation of sites in the landscape (Chippindale and Nash 2004; David 2002). There is an exception, in which Arcà (2004) discussed possible images of landscape elements in the art of Mont Bego in France (described by Alexander et al., Ch. 17, as the “twin” of their subject, Valcamonica). In that example, the supposed landscape elements seem to have been objects of interest to the artists rather than features of landscapes within which other agents carried out their actions. Having said that, it is important to ask how people perceived their environments at any time, even if it will be difficult to provide an answer. One avenue to explore this is offered by Utrilla and her colleagues (Ch. 14)

BEHIND THE SCENES

11

who describe engravings on rocks that appear to be interpretable as maps of the immediate environment of the site. In her survey of other examples, it turns out that images that can be interpreted as maps are rare throughout early art, and some are probably overinterpreted. The simple reason for this is that maps are conventionalized ways of representing the three dimensions of space in two dimensions. Once the cultural conventions of mapping are lost, it becomes quite difficult to interpret the two-dimensional image as a representation of that three-dimensional model. As with the cultural processes that led to some representation of the ground surface, it is possible that images such as these contributed to an awareness of the possibility of talking about environmental features spatially.

Comparing Rock Art Scenes on Broad Spatial Scales When treating a subject such as rock art, it is an inevitable tendency to concentrate on what is there and ignore the questions raised by what is not there. As discussed earlier, we tend to concentrate on the images of animals, people, or artifacts and pay less attention to the absence of landscapes. We also tend to concentrate on the art in the places where it has been found and pay less attention to the places where it has not been found. In the back of our minds, we suspect, most people have an idea that art may have been produced on perishable media such as wood or skins and that the absence was not real in the past. We also know that there is variation in a single time period between different places. With the benefit of archaeological knowledge, we can compare the presence of geometric markings at Blombos with their absence outside southern Africa 75,000 years ago (Henshilwood, d’Errico, and Watts 2009), we can note the similarities between the wall paintings of Chauvet Cave in France and the carved bones of the Swabian Jura sites while also noting differences in detail, and we can routinely note the absence of similar images in other parts of the world from that time period. We understand that people at the Apollo XI site in Namibia produced paintings of animals on stone plaquettes (Rifkin, Henshilwood, and Haaland 2015) a few thousand years earlier than those in Les Mallaetes and Parpalló in eastern Iberia (Villaverde Bonilla 1994), but surely without any connection. All of these examples contrast with the absence of similar practices elsewhere at the same time. The classic case of comparisons across broad spatial scales (other than that between different ends of the Mediterranean) is that by Ross (2013), who discussed the regional variations in rock art across northern Australia and the multiple factors that account for differences. Ingold (2000) has sought to explain some of the differences between art styles in different regions by appealing to the ways in which the ontologies of different peoples are represented through the different forms of art. Ross (Ch. 10) returns to this comparative approach across Australia, contrasting the art of the Kimberley, which includes many anthropomorphic and other iconic images, with the art of Central Australia, in which the dominant motifs are often called geometric, consisting of “circles and circle variants, arcs, sinuous lines, linear motifs and tracks.” Even without ethnographic guidance, compositional analysis readily leads to the identification of scenes, such as gendered activities of ceremony or food-getting among the iconic representations in the Kimberley. Ross goes on to apply compositional analysis to identify scenes among the geometric images of Central Australia, but, she warns, “without the addition of ethnographic documentation, scenes would remain unrecognized” (p. 155). The ethnography recorded by Spencer and Gillen (1899) shows that even the most stylized quasi-geometric signs can be interpreted as involving human and ancestral agency, which would be unknowable without the ethnography. From the point of view of comparison across broad spatial scales, these very different sets of imagery are found on opposite sides of one of the most remarked-upon cultural boundaries in Aboriginal Australia: that between the diverse language groupings in the Kimberley and the relatively homogenous widespread Pama-Ngyungan languages of most of the rest of Australia (Bouckaert, Bowern, and Atkinson 2018). There has been a tendency to concentrate on the art of particular regions that show a consistent style. There has been less attention to the variation between regions at any particular time and the reasons for the sorts of boundedness that they imply. The question of boundedness of rock art regions arises also on smaller scales. Brusgaard and Akkermans (Ch. 9) point out that there is variation in the subject matter within quite small distances in eastern Jordan. Spangler and Davidson (Ch. 18) find something similar within Nine Mile Canyon. The question of variation across space is, therefore, complicated. One of the most common generalizations about Paleolithic art, after noting the absence of scenes, is that the animals depicted do not correspond straightforwardly to the animals

12

IAIN DAVIDSON AND APRIL NOWELL

in the diet or in the environment. Yet, through Rice and Paterson’s (1985, 1986) region-by-region surveys it is possible to show that even that generalization is an oversimplification (Davidson 1999) and regional variation is the norm. We should expect that there is/was regional variation at different scales within and between regions in the presence or absence of scenes and of the role of those presences and absences in the societies that made them.

Scenes and Ways of Seeing the World The chapters in this volume have canvassed many different ways of seeing scenes. For Davidson (Ch. 1), any combination of images in which the subjects appear to be exercising agency constitutes a scene. Van Gelder and Nowell (Ch. 13) show that the predominantly non-figurative finger flutings of the Upper Paleolithic cave of Rouffignac allow identification of agency through analysis of the “intimate” acts of individuals and their interactions—a freezing of the relationships between the people making the marks. At some stage and in some places, such marks came to show a resemblance with real objects in the world and became pictures with iconic resemblance (Davidson 2013), yet, as Spangler and Davidson (Ch. 18) show, art that has representations of individual animals seems to have been rare. As Villaverde (Ch. 15) pointed out, discussing a site with abundant images of individual animals engraved or painted on stone plaquettes (Villaverde Bonilla 1994), some scenes may be no more than snapshots of moments of interaction—freezing actions external to the artists. Later in the same region of Iberia, some of the Levantine Art went beyond such snapshots to create an ongoing narrative by showing a succession of moments in a longer event—a freeze-frame representation that we can interpret as a narrative. Scenes, therefore, take the creation of art from individual marks, through the interaction of marks made by different individuals, to the production of pictures of objects “out there” and on to the representation of interactions between those individuals. This progression suggests the way in which the representation of scenes breaks the nexus between the artist and the observer and creates the possibility of the art communicating aspects of narratives through time without the need for a commentary informed by the producers’ intentions. An image of an individual animal may have been readily identifiable as an image of such an animal, just as it is to observers in the modern world, though for some images context was nevertheless necessary for identification (Morphy 1991: 158). Any meaning or interpretation depends and depended on the statements of the producer of the art, and the cultural context that linked the producer and the observer. But in scenes involving interactions between the represented individuals (animals and/or humans), whatever the complexities of meaning imparted between producers and observers, there was always a simple interpretation of the interactions. One of us pointed out that the creation of rock art alters the way people relate to their landscape, thus affecting other aspects of their behavior, including the nature of decisions about foraging (Davidson 2012a). The proposition was that simply marking a place in the landscape changed the values associated with that place because of the persistence of the images. Here we might go further and suggest that rock art with scenes affects the nature of the values because of the greater transparency of the narrative at the place. Kelly, David, and Flood (Ch. 11) describe a further stage of this value creation, in which sites in different parts of the landscape create or define links between those different places to allow a narrative about behavior on a wider scale. At each stage of this process, the rock art becomes a deeper part of the engagement with the environment and with the other people using it. Through this engagement, the mythology allows the rock art images to be agents in the landscape and not just in the rock art site. And through that agency in the landscape, elements of the landscape themselves become agents in the creation story. Much of this is derived from, and hence possibly specific to, the cosmologies of Australian Songlines and different accounts of the creation of the landscape in different regions. Nevertheless, reduced to these elements, it is conceivable that there is a process which might be more broadly applicable. What this collection hints at is a set of broader contexts of importance for rock art in the relations between people and their environments. These do not necessarily arise from the traditional preoccupations with identification of image content and the definition of image style boundaries in space and time. Also, they do not arise from speculations about the “meaning” of art in terms of the quasi-ethnographic categories of structuralism, shamanism, or other fashionable -isms in the study of people without history. Instead,

BEHIND THE SCENES

13

we suggest that the production of art consisting of scenes in which individual images could be seen to have agency allowed them to be seen as such after the times of production. Independent of the original storytelling and ritual with which they were produced, observers unversed in the stories or rituals could interpret the agency and tell a story, just as the authors in this book have done. As one of us argued elsewhere (Davidson, 2020) this is a major step toward art being intrinsic to the images, and not just a feature of the cultural context of their production. Iain Davidson has worked on Spanish Upper Palaeolithic Art, archaeology and ethnography of Northwest Queensland, Australian rock art, colonization of Sahul, language origins, and cognitive evolution. He has taught in England, Australia, the United States, and Chile, and has worked with ten different Aboriginal groups and undertaken major archaeological consultancy for many of Australia’s leading industries. His publications include 4 books, 7 edited volumes, more than 60 chapters and 80 articles in journals. He is a Fellow of the Australian Academy of Humanities and was awarded the Rhys Jones Medal of the Australian Archaeological Association 2010. He held the Chair of Australian Studies at Harvard University from 2008–09. April Nowell is a Paleolithic archaeologist and Professor and Chair of Anthropology at the University of Victoria. She directs an international team of researchers in the study of Lower and Middle Paleolithic sites in Jordan and collaborates with colleagues on the study of Pleistocene rock art in Australia and France. She is known for her publications on cognitive archaeology, Paleolithic art, the archaeology of children, and the relationship between science, pop culture, and the media. She is the author of Growing Up in the Ice Age: Fossil and Archaeological Evidence of the Lived Lives of Plio-Pleistocene Children (2021). Her co-edited volumes include Stone Tools and the Evolution of Human Cognition (2010) with Prof. Iain Davidson, Archaeology of Night: Life After Dark in the Ancient World (2018) with Dr. Nancy Gonlin, and the forthcoming Culturing the Body: Prehistoric Perspectives on Identity and Sociality with Dr. Benjamin Collins.

References al-Manaser, A. 2018. “Traditional Music or Religious Ritual? Ancient Rock Art Illumined by Bedouin Custom.” In To the Madbar and Back Again, ed. Laïla Nehmé and Ahmad Al-Jallad, 81–95. Leiden: Brill. Arcà, A. 2004. “The Topographic Engravings of Alpine Rock-Art: Fields, Settlements and Agricultural Landscapes.” In The Figured Landscapes of Rock-Art. Looking at Pictures in Place, ed. C. Chippindale and G. Nash, 318–349. Cambridge: Cambridge University Press. Aujoulat, N. 2004. Lascaux. Le geste, l’espace et le temps. Ed. J. Clottes. Paris: Seuil. Bouckaert, R. R., C. Bowern, and Q. D. Atkinson. 2018. “The Origin and expansion of Pama-Nyungan Languages across Australia.” Nature Ecology & Evolution 2(4) 741–49. doi: 10.1038/s41559-018-0489-3. Chaloupka, G. 1979. “Pack-Bells on the Rock Face: Aboriginal Paintings of European Contact in North-Western Arnhem Land.” Aboriginal History 3(1): 92–95. Chippindale, C., and G. Nash, eds. 2004. The Figured Landscapes of Rock-Art. Looking at Pictures in Place. Cambridge: Cambridge University Press. Clegg, J. 1984. “Pictures of Bulls and Boats: Some Evidence of Prehistoric Perceptive Processes.” In Under the Shade of a Coolibah Tree, ed. Richard A. Hutch and Peter G. Fenner, 219–38. Lanham, MD: University Press of America. Clottes, J. 2009. “Sticking Bones into Cracks in the Upper Palaeolithic.” In Becoming Human: Innovation in Prehistoric Material and Spiritual Culture, ed. C. Renfrew and I. Morley, 195–211. Cambridge: Cambridge University Press. Clottes, J., and J. D. Lewis-Williams. 1998. The Shamans of Prehistory. Trance and Magic in the Painted Caves. New York: Harry N. Abrams. Conard, N. J., M. Malina, and S. C. Münzel. 2009. “New Flutes Document the Earliest Musical Tradition in Southwestern Germany.” Nature 460: 737–40. David, B. 2002. Landscapes, Rock-Art and the Dreaming. An Archaeology of Preunderstanding. London: Leicester University Press. ———. 2004. “Rock-Art and the Experienced Landscape: The Emergence of late Holocene Symbolism in North-East Australia.” In The Figured Landscapes of Rock-Art. Looking at Pictures in Place, ed. C. Chippindale and G. Nash, 153–181. Cambridge: Cambridge University Press. ———. 2006. “Archaeology and the Dreaming: Toward an Archaeology of Ontology.” In Archaeology of Oceania: Australia and the Pacific Islands, ed. I. Lilley, 48–68. Carlton, Vic: Blackwell.

14

IAIN DAVIDSON AND APRIL NOWELL

Davidson, I. 2020. “Marks, Pictures, and Art. Their Contribution to Revolutions in Communication.” Journal of Archaeological Method & Theory 27(3), 745–70. Davidson, I. 1994. “Comment on Taçon and Chippindale.” Cambridge Archaeological Journal 4:234–37. ———. 1995. “Paintings, Power, and the Past: Can There Ever Be an Ethnoarchaeology of Art? [Ancestral Connections: art and an Aboriginal system of knowledge. (H. Morphy)].” Current Anthropology 36(5): 889–92. ———. 1999. “Symbols by Nature: Animal Frequencies in the Upper Palaeolithic of Western Europe and the Nature of Symbolic Representation.” Archaeology in Oceania 34: 121–31. ———. 2012a. “Four Questions about Foraging Models and the Process of Colonisation.” Australian Archaeology 74:19–20. ———. 2012b. “Symbolism and Becoming a Hunter-Gatherer.” L’art pléistocène dans le monde / Pleistocene art of the world / Arte pleistoceno en el mundo, Actes du Congrès IFRAO, Tarascon-sur-Ariège, septembre 2010, Symposium « Signes, symboles, mythes et idéologie. . . ». N° spécial de Préhistoire, Art et Sociétés, Bulletin de la Société Préhistorique Ariège-Pyrénées LXV-LXVI:Book: 292–293, CD: 1689–1705. ———. 2012c. “What a Carry On? Portable Art and Changes of Symbolic Meaning.” L’art pléistocène dans le monde / Pleistocene art of the world / Arte pleistoceno en el mundo, Actes du Congrès IFRAO, Tarascon-sur-Ariège, septembre 2010, N° spécial de Préhistoire, Art et Sociétés, Bulletin de la Société Préhistorique Ariège-Pyrénées LXV-LXVI:Book 268–69, CD:. 1559–70. ———. 2013. “Origins of Pictures: An Argument for Transformation of Signs.” In Origins of Pictures: Anthropological Discourses in Image Science, ed. K. Sachs-Hombach and J. R. J. Schirra, 15–45. Cologne, Germany: Halem. ———. 2017a. “Images of Animals in Rock Art: Not Just “Good To Think”.” In Oxford Handbook of the Archaeology and Anthropology of Rock Art, ed. B. David and I. J. McNiven. Oxford, UK: Oxford University Press. ———. 2017b. “Paleolithic Art.” In Oxford Bibliographies in Anthropology, ed. J. Jackson. New York, NY: Oxford University Press. Delporte, H. 1981. L’objet d’art préhistorique. Paris, FR: Réunion des Musées Nationaux. Dobrez, Livio. 2011. “Looking at Our Looking: A New Approach to the Definition of a Rock Art Scene.” American Indian Rock Art 37: 251–267. ———. 2012. “Towards a more Rigorous Definition of Terms: Are There Scenes in European Palaeolithic Art?” L’art pléistocène dans le monde / Pleistocene art of the world / Arte pleistoceno en el mundo, Actes du Congrès IFRAO, Tarascon-sur-Ariège, septembre 2010, Symposium « Signes, symboles, mythes et idéologie. . . ». N° spécial de Préhistoire, Art et Sociétés, Bulletin de la Société Préhistorique Ariège-Pyrénées LXV-LXVI:Book: 316–17, CD: 1837–51. Fritz, C., T. Lenssen-Erz, G. Sauvet, M. Barbaza, E. Lopez-Montalvo, G. Tosello, and M. Azema. 2013. “L’expression narrative dans les arts rupestres: approches théoriques.” Les Dossiers d’archéologie (358): 38–45. Fritz, C., G. Tosello, and M. W. Conkey. 2016. “Reflections on the Identities and Roles of the Artists in European Paleolithic Societies.” Journal of Archaeological Method and Theory 23(4): 1307–32. Gombrich, E. H. 1995. The Story of Art. 16th ed. London, UK: Phaidon. Henshilwood, C. S., F. d’Errico, and I. Watts. 2009. “Engraved Ochres from the Middle Stone Age Levels at Blombos Cave, South Africa.” Journal of Human Evolution 57(1): 27–47. Ingold, T. 2000. “Ch. 7. Totemism, Animism and the Depiction of Animals.” In The Perception of the Environment: Essays on Livelihood, Dwelling and Skill, ed. T.Ingold, 111–31. London, UK: Routledge. Kolpakov, E. M., and V. Y. Shumkin. 2012. “Boats in the Rock Art of Kanozero and Northern Europe.” Archaeology, Ethnology and Anthropology of Eurasia 40(1): 76–81. doi: 10.1016/j.aeae.2012.05.009. Lenssen-Erz, T. 1992. “Coherence—A Constituent of ‘Scenes’ in Rock Art.” Rock Art Research 9(2): 87–105. McCall, G. S. 2007. “Add Shamans and Stir? A Critical Review of the Shamanism Model of Forager Rock Art Production.” Journal of Anthropological Archaeology 26(2): 224–233. Merlan, F. 1989. “The Interpretive Framework of Wardaman Rock Art: A Preliminary Report.” Australian Aboriginal Studies 1989(2): 14–24. Morphy, H. 1991. Ancestral Connections: Art and an Aboriginal System of Knowledge. Chicago, IL: University of Chicago Press. O’Connor, S., and S. Arrow. 2008. “Boat Images in the Rock Art of Northern Australia with Particular Reference to the Kimberley, Western Australia.” In Islands of Inquiry: Colonisation, Seafaring and the Archaeology of Maritime Landscapes, ed. Geoffrey Clark, B. F. Leach, and S. O’Connor, 397–409. Canberra, AU: ANU Press. Quiles, A, H. Valladas, H. Bocherens, E. Delqué-Kolicˇ, E. Kaltnecker, J. van der Plicht, J.-J. Delannoy, V. Feruglio, C. Fritz, J. Monney, M. Philippe, G. Tosello, J. Clottes, and J.-M. Geneste. 2016. “A High-Precision Chronological Model for the Decorated Upper Paleolithic Cave of Chauvet-Pont d’Arc, Ardèche, France.” Proceedings of the National Academy of Sciences 113(17): 4670–75. doi: 10.1073/pnas.1523158113. Rice, P. C., and A. L. Paterson. 1985. “Cave Art and Bones: Exploring the Interrelationships.” American Anthropologist 87: 94–100. Rice, P. C., and A. L. Paterson. 1986. “Validating the Cave Art—Archeofaunal Relationship in Cantabrian Spain.” American Anthropologist 88: 658–67. Rifkin, R. F., C. S. Henshilwood, and M. M. Haaland. 2015. “Pleistocene Figurative art mobilier from Apollo 11 Cave, Karas Region, Southern Namibia.” South African Archaeological Bulletin 70(201): 113–125.

BEHIND THE SCENES

15

Ross, J. 2013. “A Continent of Nations: The Emergence of New Regionally Distinct Rock Art Styles across Australia.” Quaternary International 285(0): 161–71. doi: http://dx.doi.org/10.1016/j.quaint.2012.01.007. Spencer, W. B., and F. J. Gillen. 1899. Native Tribes of Central Australia. London, UK: Macmillan. Veth, P., C. Myers, P. Heaney, and S. Ouzman. 2018. “Plants before Farming: The Deep History of Plant-Use and Representation in the Rock Art of Australia’s Kimberley Region.” Quaternary International 489: 26–45. Villaverde Bonilla, V. 1994. Arte Paleolítico de la Cova del Parpalló. 2 vols. València: Museu de Prehistòria de València. Vinnicombe, P. 1972. “Myth, Motive and Selection in Southern African Rock Art.” Africa 42(3): 192–204. ———. 1976. People of the Eland. Rock Paintings of the Drakensberg Bushmen as a Reflection of Their Life and Thought. Pietermaritzburg: University of Natal Press. Wobst, H. M. 1978. “The Archaeo-Ethnology of Hunter-Gatherers or the Tyranny of the Ethnographic Record in Archaeology.” American Antiquity 43(2): 303–309. doi: 10.2307/27

==== •

•

•

1. SCENES AND NON-SCENES IN ROCK ART Iain Davidson

The Problem of Scenes in Rock Art One of the constant descriptors of Paleolithic cave art1 has been that there was little or no representation of scenes. The purpose of this chapter and of those by other authors in the book is to examine that proposition and explore the implications of the presence or absence of scenes in both the Paleolithic art of Western Europe and rock and cave art elsewhere in the world. It will be important, in considering this, to understand that we are dealing only with images on rock surfaces produced before written accounts could record any description of what the producers of the art may have intended. The issue becomes one of how we observers can interpret sets of images without either the benefit of experience of the production of those images or a cultural connection to the traditions within which the art was produced. This book will present a number of approximations to what is meant by a scene. The first requirement will be a definition of “scene” relevant to rock or cave art (or its mobile equivalents), particularly in the context of the absence of any account contemporary with the production of the art. None of the obvious dictionary definitions is quite suitable, but the closest one seems to be related to works written for the theater or cinema; Merriam-Webster2 defines a scene as “a division of an act in a play during which the action takes place in a single place without a break in time.” The important part here is that action takes place. For the purposes of this chapter, a scene can be identified from a set of images in spatial proximity to each other from which, without any knowledge other than the images themselves, an observer can infer actions taking place among the actors represented in the images. Much of the discussion of the presence or absence of scenes in Paleolithic cave art has not had the benefit of such a definition and has relied instead on intuition (but Villaverde’s Chapter 15 makes productive use of distinctions defined by Delporte). Moreover, the general statements about absence of scenes as a characteristic of the art belies the knowledge that some scenes have been recognized. Some Paleolithic scenes are in fact very famous, including those from Parpalló in eastern Spain (Pericot García 1942) (Figures 1 and 2 in Chapter 15), which were executed on small slabs or plaquettes of stone (and one on bone), and those from Lascaux in southern France, which were produced on the walls inside a deep cave. Contrary to the popular image beloved of cartoonists who show stick figure humans hunting herds of animals, scenes were not an important part of Paleolithic cave art, or so it has been said. Yet the same is not said about rock art elsewhere. One of the most famous examples of a hunting scene (in the post-Paleolithic Levantine Art of eastern Spain) is discussed by Villaverde in Chapter 15 (and his Figure 5). Without undertaking an exhaustive study, it is possible to find some statement by leading authorities about absence of scenes from cave art in almost every decade since the 1950s. This is the more remarkable since Parpalló and Lascaux were both published in the 1940s and these are two of the sites with the best known scenes. According to Ucko and Rosenfeld (1967: 41), citing Maringer and Bandi from fourteen years earlier: “It has been commonly stated that one of the distinctive characteristics of this art was that it was not concerned with scenes of grouping.” Two decades later, Bahn and Vertut (1988: 120) pointed out that “‘scenes’ are very hard to identify in Paleolithic art, since without an informant it is often impossible to prove ‘association’ of figures rather than simple juxtaposition.” The key words here are grouping, informant, association and juxtaposition.

SCENES AND NON-SCENES IN ROCK ART

17

In 1996, Clottes and Lewis-Williams (1998: 52), with their combined expertise in Paleolithic cave art and South African rock art, were able to say: “Given such a naturalist approach to the fauna, one might expect to find in the caves a multitude of scenes representing wildlife. Yet this is not the case, though narrative scenes including two or more animals are no doubt more numerous than has been recorded.” The emphasis had shifted somewhat to naturalism and narrative and a concession that there are probably more scenes than the stereotype would allow. Guthrie (2005: 61), concerned with a literal interpretation of the images, noted that “while there are few actual ‘scenes’ in Paleolithic art, large mammals are sometimes portrayed in groups. Interestingly, these group scenes are of animals that are social today, so we can take this portrayal as literal.” To complete this brief survey, Clottes (2011: 239), writing principally about the reasons behind cave art in France and Spain, described the panel that includes images of a person in front of a bison in the Well at Lascaux (Davenport and Jochim 1988) as follows: “Le Puits de Lascaux est célèbre par la présence d’une des rares scènes évidentes, même si elle reste mystérieuse, de l’art pariétal paléolithique.”3 In reality, many panels contained at Lascaux have elicited suggestions that the arrangements of individual animal images seem to go beyond association or juxtaposition, but they do not seem to form scenes as modern Western observers would like them to be. These extracts show that the ideology about the lack of scenes has been very strong for at least fifty years, although commentators have hedged the way they say it. People express their caution in different ways. They were “not concerned with scenes of grouping,” scenes are “very hard to identify, expectations of scenes are wrong, and there are “few actual scenes.” Through this use of language, the problem has been displaced onto “grouping,” ease of identification, our expectations, or definitions of what an actual scene might be. In making these statements over at least sixty years, the successive authors revealed some of the expectations of what a scene might be and some of the features that might assist in recognizing one. We have an expectation that informants whose lives intersect with the context of production of the art would disambiguate its “scene-iness.”4 This chapter and the whole book aim to clarify some of these issues about scenes in rock and cave art. The belief that scenes are relatively rare has probably inhibited the discussion of cases that quite likely involve recognizable scenes. As Villaverde shows in his chapter, there are scenes from different periods on several other plaquettes from Parpalló, but they have rarely been emphasized, perhaps because there is so little expectation that scenes existed. In Chapter 12, about European Paleolithic Art, Culley discusses a difference between the conventions for scenes in wall art and mobile art in the Upper Paleolithic of Western Europe.

Recent Discoveries in Scenes There is no doubt that the situation of Paleolithic art is changing. For example, Fritz, Tosello, and Conkey (2016) suggested that it may be possible to identify some images in European Paleolithic art as representing the land surface on which animals were standing, and that this might establish the conditions for representing both the figure and the ground. Nomade and colleagues (2016) also offered an interpretation of marks in Chauvet Cave as representations of a volcano that erupted during the period of use of the cave. Several parts of the cave have walls marked in a variety of ways that do not seem representational: most obviously in the Alcove of the Lions (Fritz and Tosello 2007: Figure 18), some of which seems related to preparing the rock surface for painting. I am skeptical, given that there is so much other modification of the cave wall surfaces at Chauvet Cave, that particular sets of marks could be singled out for interpretation in this way. In Spain, Utrilla and colleagues (2009) suggested that the combination of images on two engraved blocks from Abauntz Cave (in Navarra) could be interpreted as a map of the landscape of the site where they were found, and that the superimpositions of the animal images constitute a scene. Utrilla et al. discuss these and other examples of “maps” in Chapter 14. Historically, the idea that old art might include both animals and elements of their environments, in particular water, was discussed at length by Marshack (1977). A further example of images that might be interpreted as a representation of environmental context is the plaquette from Molí del Salt (in Tarragona) that has been interpreted as representing a camp of seven huts (García-

18

IAIN DAVIDSON

Diez and Vaquero 2015). Given the breakthrough in interpretation of the representation of huts in the art, it would be possible to point to the half circles on the Abauntz image as similarly representing camps. In these four examples there is at least an indication that the tide has turned toward interpreting Paleolithic images as scenes. In light of these examples, we may suppose that representations of the landscape and of scenes are not as scarce as they have been said to be, and that one of the impediments to our ability to recognize them has been the expectation that landscapes and scenes are absent. In reality, it may be that the poverty is in the way Paleolithic art has been looked at, rather than in its content. The issue is important because, as we shall see (in both this and many other chapters in the book), scenes are well known in other examples of rock art, and one question is whether there is a contrast between an impoverished Paleolithic cave art and a far richer rock art elsewhere, and if so, why. Did practice about the depiction of scenes change through time, or can we work out whether (1) there were cultural rules that determined whether scenes were ever produced, or (2) sometimes scenes were appropriate and other times not, depending on the cultural context, or (3) quite possibly, there was some mix of these options? How can we identify scenes? What were the conditions of production of rock and cave art images, and how might they be construed to constitute a scene? Is it indeed no more than “grouping?” What are our expectations? Once we have thought about these questions, it may be possible to consider the implications of the presence or absence of “scenes that meet our expectations” and maybe ask whether there are other groups of images that constitute scenes that do not meet our expectations.

The Composition of Panels of Images in Rock Art Some of the elements of the definition introduced earlier appear straightforward, but they need some elaboration. First, there is an assumption that the elements of a panel and hence of the scene on that panel can be separated into almost independent images; that is, that it is possible to discuss the individual animals represented and hence the relations between them. This would clearly not always be true of, for example, nineteenth-century Impressionist landscapes, where the whole picture has to be considered together, but it probably does often apply to rock and cave art. Studies of the modern production of art within the market economy are not straightforwardly relevant to what we are talking about here (Davidson 1995), not least because the function of the picture is to be transferred from studio to saleroom to gallery or home. One of the genuine scarcities in rock and cave art is representation of landscapes (but see Utrilla’s chapter on maps), so the question of independent images may not generally be a problem. Moreover, the panels stay where they were executed. Second, it would be desirable to define what is meant by spatial proximity, though it might be different for each situation. The chapters in this volume make various attempts to deal with that issue. The image of a lone elk at Warrior Ridge in Nine Mile Canyon (see Spangler and Davidson in Chapter 18) is situated about two meters below a small panel of small human and animal figures that represent a scene apparently unrelated to the elk. Intuitively, there is no relation between the large elk and the small figures, and the only reason for saying so is the absence of spatial proximity. Third, the combination of an “observer” and “any other knowledge” goes to the heart of what rock and cave art is about. Nobody in the present day has any of what I will call “knowledge untainted by time” about what was intended by the people who made the images. With or without knowledge of the circumstances of production, we cannot escape the fact that an artist or artists made the images, and, being cultural actors, they had some cultural constraints about how they produced them. Although there may have been restrictions about access to the images, there were also contemporary observers who may or may not have been privy to the thoughts of the artists. Even in the absence of those observers contemporary with the production, the art is viewed by present-day people whose cultural backgrounds differ from those of both the artist and other observers. What present-day people of educated Western background might require for the identification of a scene might be quite different from what people from a different or an earlier culture might have required. This inevitably shifts the emphasis away from both the performance of the production of the images and the relationships between the artist and the audience, and onto the art as an object and its

SCENES AND NON-SCENES IN ROCK ART

19

contents. Finally, as this chapter strongly emphasizes, there are important questions about how it is possible to infer interactions among the individual images within what may be a scene. The fundamental question concerns how panels of cave or rock art were composed. Every panel must have begun with a single image. Perhaps, if several artists were involved, there was more than one in a very short period of time, but even that extension is limited by the physical requirements for making marks on a panel of restricted size. Probably some panels were more often accumulated over longer periods of time, sometimes within an ongoing cultural practice, sometimes not. The most detailed study of this sort of composition is from the Panel of the Horses at Chauvet Cave, where it has been possible to disentangle a sequence of image making, rock surface modification, and successive image production and superimposition in meticulous detail (Fritz and Tosello 2007). Unfortunately, the resolution of radiocarbon dating is not precise enough to be sure of the length of time from the beginning of creation of the panel to the final phase visible today. Were the successive images of animals produced as additions to a rock panel by artists of a single culture in a relatively short time, or were they made at different times by people from different cultural situations? Villaverde addresses the question of additions to pre-existing panels in Chapter 15. We know from the situation of panels of petroglyphs with different states of weathering that images were sometimes added to a panel in actions sufficiently separated in time to allow weathering of the rock surface. There is a significant difference between the length of time required to allow weathering and that required to make a panel into a scene. That would suggest that less extreme separations in time could have existed, such that combinations visible today were not necessarily intended at the time of the first image. Or were panels initially planned and produced as a composition of animals and other signs? These fundamental questions are often overlooked by many approaches to rock and cave art. The answers are crucial to the assumptions involved in studies that consider whole sites using structural (Leroi-Gourhan 1965) or statistical analysis (e.g., Sauvet et al. 2000). In rare cases, as at the ~4,000-year-old White Shaman Mural in Texas (Boyd 2016), it has been possible, through meticulous observation including detailed documentation by digital microscopy of sequences of paint application, to show that a large panel was produced as a single composition. One startling confirmation of this was that many deer across the panel had black dots at the ends of their antlers, which upon examination proved to have been painted in each case before the antlers, strongly suggesting that the composition of the panel was planned (Boyd 2016: 38–44). Studies of rock art that are capable of such detailed exegesis are still very rare, but the questions raised apply to most rock and cave art panels. Images of individual animals are relatively rare in all rock and cave art, though I have not done a systematic survey. They are more abundant on bone or stone plaquettes and less so in rock and cave art. I do not think that an individual animal, such as the elk from Warrior Ridge in Nine Mile Canyon, Utah, can straightforwardly be said to be part of a scene (Spangler 2013: 88). In Chapter 18, Spangler and I present some statistics on the nature of panels in Nine Mile Canyon that bear only a single image. In practice, questions of intent in composition are very difficult to answer, so a more analytical approach may be warranted. The first approximation presented here can no doubt be improved upon, but it is important to make a start. I identified seven categories of images or image combinations as follows (Figure 1.1): Category A: single animals (Nine Mile Canyon—NMC); Category B: several animals produced at the same time, juxtaposed, single species (Ekain); Category C: several animals produced at the same time, juxtaposed, more than one species (Pecos); Category D: several animals produced at the same time but superimposed, same species (Chauvet horses); Category E: several animals produced at the same time, superimposed, more than one species (Niño); Category F: several animals done at different times, single species (NMC); Category G: several animals done at different times, more than one species (Serra da Capivara, Brazil). At Cueva del Niño in Spain (Almagro Gorbea 1971; Davidson and García Moreno 2013), there is a panel that illustrates some of these points (Figure 1.2). The dominant figure on the panel is a large male deer (stag) with prominent antlers, marked in the figure as image (a). Slightly above and to the right is another stag that is substantially stylistically similar to the larger one, particularly in the curves of the rump and of the antlers (b). To the left of the large deer are two small images interpreted as juvenile deer (c, d). In addition, there are two symmetrically placed images of Spanish ibex, one to the lower left (e) and one to the right of the panel (f), that are less visible in this photo. The ibex image on the left is on a different plane

20

IAIN DAVIDSON

One species One animal

A

Several animals same time, juxtaposed

Several animals, superimposed

More than one species

B

C

D

E

Several animals different time

G

F

FIGURE 1.1a–g. (a) Elk at Warrior Ridge, Nine Mile Canyon. Copyright Iain Davidson. (b) The Grand Panel of Horses at Ekain. Copyright Jesús Altuna and Koro Mariezkurrena. (c) More than one species at Halo Cave, Lower Pecos, Texas. Copyright SHUMLA. (d) The Panel of Horses, Chauvet Cave. Copyright Jean Clottes. (e) Main panel, Cueva del Niño, Albacete. Copyright Iain Davidson. (f) Different weathering of images of bighorn sheep, Nine Mile Canyon, Utah. Copyright Iain Davidson. (g) Paintings of different animals done at different times, Serra da Capivara, Brazil. Copyright Livio and Patricia Dobrez.

of the rock surface and is now covered with a crust not present on the rest of the images, so it could be from a different time. The fact that there are pairs of similar images of three types of animal suggests some possibility of composition of the panel, even if it is not constructed as a scene according to modern Western criteria. Nevertheless, Fortea Pérez (1978: 137–38) analyzed the panel in a different way, seeking parallels in art of different styles from different time periods. Although his dating was anchored to a relative chronology derived from images on plaquettes from the stratified site of Parpalló, in this instance such considerations serve principally to illustrate the difficulty, within earlier traditions of stylistic chronology, of even identifying a set of images on a rock wall that could be considered a panel, let alone a scene. If the images belonged to styles from different periods, they might have been added without any consideration of the whole composition. There are two other clear images on the panel: the forequarters of a hind (g) and of a horse (h), both apparently painted before the large stag, as it superimposes them. The composition was created after the hind

SCENES AND NON-SCENES IN ROCK ART

21

FIGURE 1.2. Main panel at Cueva del Niño, Albacete, Spain. Copyright Iain Davidson.

and the horse were painted on the panel, indicating that the question of the interaction between composed groupings and superimposition is not straightforward. Just as close reading of the panels at Chauvet Cave demonstrates that those panels were carefully composed (e.g., Fritz and Tosello 2001), so careful reading of other sites in this way would produce a convincing argument that composition was the norm rather than the exception. Why, then, does the dogma of the rarity of scenes persist? What is it that distinguishes a composed panel or panels from a scene? How can a modern observer determine that the figures on a panel represent “actions taking place among them?” What are the implications about the context of production of the art, if some art was produced to show actions and other art was not?

Agency The answer, I propose, lies in the question of agency. The question of who the agents are is crucial to the identification of images as scenes. A similar point was made by L. Dobrez (2012: CD–1839) in a discussion of two bison painted facing in opposite directions on Panel 6 in the Salon Noir of Niaux cave: “A scene is best thought of as a visual narrative: it means that the two bison are doing something.” The account that follows differs from the detailed exploration of Art and Agency by Gell (1998), who was principally concerned with developing an anthropological theory of art for which there are or were living informants. He does not seem to have been particularly concerned about rock or cave art, referring (p. 33) to “the famous hand-prints which occur beside the cave paintings of Lascaux, Altamira, etc.” in the context of them as an index or “congealed ‘trace’ of the artist’s creative performance.” Gell defined anthropological theories as concerned with the “social relationships” that “occupy a certain biographical space, over which culture is picked up, transformed and passed on, through a series of life stages” (p. 11). It is this concern with

22

IAIN DAVIDSON

short-term interests that makes it so difficult for such anthropologists to connect with the rest of the anthropological disciplines. The problem Gell would have faced in dealing with rock and cave art is the persistence of the images on the surface of the rock beyond the completion of the life stages he was interested in. Indeed, the essence of the understanding of the role of material culture in cultural heritage is about such persistence in the real lives of real people. Anthropology should not forget that the culture that dominates the biographies of individual actors is grounded in its history, and that the material records of that history strengthen memories of creative performance that pervade those life stages (e.g., Severi 2015). There are two fundamental agents in rock and cave art (actually any art, but let us not go there). The primary agent (A1) is (or was) the producer, who may or may not have told others, at the time of production, what has been produced and how the images relate to a narrative. The producer was, of course, also the artist (and the performer), but we should be careful about implying that there was a role for an individual as a specialist in the production of art without arguing the case. Consider the situation in present-day Australia where we are familiar with the context of interaction between Dreaming Songs and sand drawings. In that context the drawings are illustrative of the actions of the Ancestors in the Song, and the drawings are themselves ephemeral, however indestructible the Ancestors (Green 2014). In this case it would be possible to argue that the producer is a specialist in the ritual knowledge rather than in the production of the sand drawings. I suggest this is an appropriate analogy for one of the contexts of rock and cave art production— provided we can demonstrate that non–present-day art also involved ritual (e.g., Ross and Davidson 2006). For most purposes in this discussion, we are dealing with images that have some iconic resemblance to the objects they represent. It is, by now, a truism that creating such a resemblance to three-dimensional objects through the production of marks on a (nearly) two-dimensional surface is an achievement with its own cultural rules (Gombrich 1952; Nowell 2015), and so is the interpretation of such marks as a representation of an object visible in the environment (Serpell and Deregowski 1980). The A1 primary agent is particularly important when the images produced were primarily geometric: the meanings could only be construed by knowing the songs or stories of the A1 agent, however constrained by culture the meanings of non-iconic marks may be (Munn 1973). As Munn showed, simple geometric figures are often deliberately polysemic—the simple images stand for different things in different contexts. A circular shape could mean any or all of the following: nest, hole, water hole, fruits and yams, tree, hill, prepared food, fire, upright fighting stick, painting material, billy can, egg, dog (when curled up in camp) or circling (as e.g., dancing around), or any encircling object (Munn 1973: 67). The disambiguation of semiotic uncertainty comes from singing or telling the story. In consequence, a group of geometric figures could constitute a scene, but in the absence of the A1 agent, it would be impossible to construe a narrative that is the same as that of the producer. It would probably be possible to construe some sort of narrative knowing the general cultural rules of such signs, but not to know whether that narrative coincided with that of the A1 producer of the art. The secondary agent (A2) was (and is) any observer of the images. That includes any participants of the social process that produced the images, as well as, for all purposes, modern people. Gell (1998: 21–23) refers to the thing on which an agent acts as a “patient.” The images are, therefore, patients with respect to an agent. In rare cases present-day observers can infer the necessity of an observer at the time of the act of making the art by using the evidence of the difficulty-to-impossibility of seeing the marks that were made (e.g., Fowles and Arterberry 2013). We may or may not have witnessed the producer telling the story or singing the song, but in the end for us, and for many more, the image or images stand alone without the story or song, or with memories of it that are either very imperfect or enhanced by the ritual context of production. There are two other candidates for agency in art. The first is the question of the agency of the art itself (A3). Around thirty years ago, Conkey (1989) wrote that “the meaningfully constituted material record is not an ‘expression’ or ‘reflection,’ nor even a ‘record’ but an active, constructing, constituting agency, which does not express meaning, but produces it.” Unfashionably, I argued (Davidson 1997: 128) that she may have had some argument for such a position among the participants in the performance of production of the art, but not for non-participants. “The material record, being an object and not a person, cannot ever be an agent in the process of producing meaning. People, may, through material things, actively construct and constitute something. . .” My view has not changed, and is not swayed by Robb (2005: 106) referring to discussion of the agency of things as “a theological issue.” What is important, many authors have argued (Bar-

SCENES AND NON-SCENES IN ROCK ART

23

nard, Davidson, and Byrne 2016; Malafouris 2013; Schiffer 1999), is that the behavior of human agents who are interacting with the material world differs from their behavior when they are only interacting with other humans. Gell (1998: 21–22) recognized the primacy of the philosophical position that only humans can be agents with respect to a patient, but went on to describe how agency can move between agent and patient by virtue of the contextual attitudes of real human agents “in the fleeting contexts and predicaments of social life, during which we certainly do, transactionally speaking, attribute agency to cars, images, buildings, and many other non-living, non-human, things.” This is, of course, an error, a misunderstanding of metaphor—I call it the illusion of the agency of things—since the attibution can only be done by the agency of a human. Cars, images, and buildings do not have agency that can attribute agency to humans, transactionally speaking or in any other way. Gell (1998: 6), of course, understood this in rejecting the notion that the work of art “participates in a ‘visual’ code for communication of meaning.” Nevertheless, in pursuit of an anthropological theory of art, Gell (5) suggested that “there could be a species of anthropological theory in which persons or ‘social agents’ are, in certain contexts, substituted for by art objects.” I do not think such a device is needed in this discussion of scenes. The second additional candidate here is that of the agency (A4) represented among the images that may constitute a scene. My argument is that the agency apparently identifiable among the images—what we might call the illusion of the representation of agency—makes a composition of images into a scene and, incidentally, makes it possible for different observer agents (A2) to construct similar stories about a shared possible meaning of the art. Note that the A4 agency depends on A2 agents to recognize conventions of representation that allow A4 agency to be identified. Indeed, in my aesthetics (that is to say, the aesthetics of a Western, white, male archaeologist of a certain age and educational and cultural background), this issue is essential to the ability to construct a narrative from images. There may, however, be or have been aesthetics in which the absence of agents as we identify them among the images allows greater mysticism in narrating the actions of agents, such that scenes could have been talked about by producers or knowledgeable observers in relation to images that do and did not themselves appear to be agentful. Chapter 2 by Livio Dobrez gives some consideration to the possibility of changes through time in such aesthetics. The site of Parpalló south of Valencia in eastern Iberia contains, stratified in layers dated between 27ka cal BP and 13ka cal BP (Bofinger and Davidson 1977), more than five thousand stone plaquettes with more than 750 images of animals (the detailed numbers are given in Chapter 15 by Villaverde and in Villaverde Bonilla 1994). (For a summary of the publications about Parpalló and its enduring significance as a site, see Davidson 2012b). Among many images of horses at Parpalló I have always seen a resemblance between an image from a layer dated to about 18ka cal BP (Figure 1.3a) and E. H. Shepard’s drawings of the stuffed animal character Eeyore, anthropomorphized as the friend of the equally anthropomorphized stuffed animal character Winnie the Pooh (Milne 1926: 52) (Figure 1.3b). Trivial though that comparison appears to be, it reminds us that in the book illustration, Eeyore is part of a narrative, made so by the A4 agency of Christopher Robin (pinning his tail on).5 It is difficult to define agency (A4) in the Parpalló horse, though this one may be easier than some others. Why is that? In many European cave sites, such as the Grand Horse panel at Ekain, there are several animals represented in rather stereotyped ways as static (Figure 1.4). Dobrez (2012: CD–1844) cites similar panels of static animals at Font de Gaume, Rouffignac, Peche Merle, Cognac, Niaux, Covalanas and Tito Bustillo. Such stereotypes can be analyzed in a way that allows them to be called caricature (Cheyne, Meschino, and Smilek 2009). More to the point, they also seem to be examples of the representation of an image of a horse. That is to say, they are not representations of individual horses but examples of the way a horse was represented in that time, place, context, and culture (for discussion of the stylistic conventions of the Parpalló horse drawings, see Pigeaud 2007)—a theme that is widespread within the history of art (Gombrich 1952) and for which there are well-known examples in the rock and cave art of Australia, as Aboriginal artists initially struggled to produce images of animals they had never seen before by representing them as distorted versions of those they did know (Chaloupka 1979; Clegg 1984). A word of caution is needed here. Evans (2010: 57) documented that the Kunwinijku language of Arnhem Land, northern Australia, has different verbs to describe the motion of different species of kangaroo and wallaby. Without knowledge of

24

IAIN DAVIDSON

FIGURE 1.3a–b. (a) Image of a horse from Parpalló (No 17757B in Villaverde 1994). Copyright Valentín Villaverde. (b) Image of Pooh, Eeyore and Christopher Robin by E. H. Shepard from Winnie the Pooh (Milne 1926, 52). Copyright Penguin Random House.

FIGURE 1.4. The Grand Panel of Horses at Ekain. Copyright Jesús Altuna and Koro Mariezkurrena.

SCENES AND NON-SCENES IN ROCK ART

25

the language, it would be impossible to attribute differences among images to variation in the gaits of such similar species. The importance of the active interaction between the individual images is well understood by looking at the first part of the panel on the left wall of the Salle du Fond at Chauvet Cave that illustrates two (or three) lions side by side (Figure 1.5). These animals are shown in an attitude that is easily interpreted (by several independent observers, Peter Jarman, personal communication, 2002 and 2017) as a scene that takes place toward the end of the mating process. Jarman6 (with years of study of African ecology, e.g., Jarman 1974) “spontaneously suggested a context (estrus and courtship; female solicitation) for the scene,” and still react[s] to it in that way. He further pointed out that “head-rubbing (one lion rubbing its cheek or upper face against another) is not confined to that situation, but is a general and frequent part of greeting.” Clottes and Azema (2005: 176) write that “Il s’agit du moment où la femelle, finalement consentante, marche tut près du mâle et se frotte contre lui.7“ To Jarman, “she is cosying up to the male and probably soliciting mating, which may already have occurred dozens of time in that oestrus.” The issue here is that by making minor differences in the way the lions were represented (the scrotum of the larger one, and a smaller animal without a scrotum; the difference from one image to the other in the angles between neck and body) the artist showed two animals apparently interacting with each other—the producer of the image appears to have shown the agency of the animals in the juxtaposition. The implication of this analysis is that the markers of the agency of the animals may be quite small differences between an image and its stereotype.

FIGURE 1.5. Image of two or three lions from Chauvet Cave showing the interactions between them through the slight difference in the orientation of the head. Copyright Jean Clottes.

26

IAIN DAVIDSON

The animals in European cave art are in spatial relationship to each other (inevitably), but it is often difficult to determine where the A4 agency lies within that relationship. What are the animals doing, and how do observers, now or then, tell a story about them? I suggest that it is necessary to identify similar nuances of representation to identify the conventions that may have defined the nature of the (A4) agency of the animals in the juxtaposition. There is plenty of nuance about the way the horse image from Parpalló is represented in Figure 1.3a, but unlike Eeyore in Figure 1.3b, there is no patient to which the agency of the horse is applied—no equivalents of Winnie the Pooh or Christopher Robin (Milne 1926) (though note that the agency of the stuffed toys Pooh and Eeyore is only possible in the context of the anthropomorphizing story). On the other hand, in the Parpalló suckling scene (see Figure 15.1 in Chapter 15 by Villaverde), it is inescapable that there is an active relationship between the two animals most obviously represented. The larger animal is a deer without antlers, so possibly female; the smaller animal is in the (nuanced) position of a fawn suckling and thus has agency with respect to the patient mother. So it would seem that, in this analysis of scenes, an inference of scene-iness derives from the presence of more than one animal among which one is represented with nuance that distinguishes it from a second or other animals in such a way as to imply its interaction with the second. An observer can infer actions taking place among the creatures represented in the images by interpreting the nuances of different aspects of representation. That is to say, without any knowledge other than the images themselves. If we look at some famous panels where there are obviously scenes, such as those in the aptly named Warrior Ridge off Nine Mile Canyon (see Ch. 18), the juxtaposition of images of armed people with different body postures suggests the agency of people against people. These conventions of representation are perhaps more familiar to us. The chapters by Lenssen-Erz et al. (Ch. 6), and Ross (Ch.10) have much more to say about the representation of humans. In the Great Hunt Panel, in Cottonwood Canyon south of Nine Mile Canyon (also discussed in Ch. 18), there are thirty-six images of bighorn sheep in a grouping that suggests a natural winter aggregation of rams, ewes, lambs, and other age classes (Figure 1.6) (Matheny, Smith, and Matheny 1997: 85). Matheny and his colleagues documented many naturalistic aspects of the composition, relying on the standardizations of images of animals of different sexes and ages. The differences are fine—two rams with divided hooves, a rounding of the hindquarters for ewes—but these are nuances on what is generally a remarkably stereotyped body plan: rectangular bodies of similar proportions with all four legs parallel, and very similar aspect of the head and horns. There is substantial repetition of the form and invariance in the manner of representation, in contrast with the variation of representation of the deer in the scene at Cova dels Cavalls (Figure 5 in Ch.15). All the animals are facing in the same direction; none have indications of genitalia. There is little interaction among the animals. We might say that they have little agency in this situation; as with the images of the horses at Ekain, these appear to be images of the way a sheep was represented in that time, place, context, and culture. The term Gell (1998: 13–16) would use for images of sheep is indexes, a term he justifies both in terms of standard Peircean semiotics (the standard Peircean definitions of his basic semiotic categories sign, index, icon and symbol are reproduced in Davidson 2013) and in his own logic. Juxtaposed with these images of sheep are images of four armed hunters, represented in a way that suggests to a present-day viewer that the people are acting on the animals. They are not stereotyped: the figure furthest to the right is larger than the others and has legs bent at the knees, feet pointing downwards, clearly marked genitalia, arrow pointing upwards relative to the angle of the bow, broad shoulders, and a headdress that bends from the back of the head to the front. Below, to the left of this figure, is a smaller individual with straight legs, feet pointing forward, no genitalia, arrow horizontal, rounded shoulders and a headdress with two parallel elements bent back from the middle of the head. The two other figures carrying bows are less distinct and hence also different from the first two. One other figure below the large one does not carry a bow, and its feet are facing in the opposite direction, away from the animals. Another figure appears to be only the lower portion of a two-legged body. These human figures are doing things; they are the hunters. It does not matter that the sheep are somewhat stereotyped: the agency is among the hunters, and the sheep are patient with respect to them. It is almost as if there were two episodes of production of the panel: the earlier one with the stereotyped images of sheep, and later additions of the hunters, shield figures and one or two of the sheep. But this puzzle cannot be resolved because there is no differential weathering of the images. The

SCENES AND NON-SCENES IN ROCK ART

27

FIGURE 1.6. The Great Hunt Panel, Nine Mile Canyon, Utah. Copyright Iain Davidson.

panel as it exists is as it was when the last individual images were added, and it is what we have to interpret now. In Ch. 15, Villaverde discusses the transformation of panels in Levantine Art by the addition of figures sometime after the original painting. The presence in the center of the panel of another, larger image, which appears to be an armless (and unarmed) human but with a non-human (perhaps bison) headdress suggests that there is another dimension to the interaction. The armed humans have agency, but it is less clear that the animals do—indeed, by their stereotyped representation I would argue that they were probably represented in a ritual, perhaps related to the large human with the animal headdress. McCall and colleagues have more to say about ritual in Chapter 5. Ross analyzed means of identifying ritual in rock art (Ross and Davidson 2006). The principal characteristics of ritual behavior, derived from Rappaport’s (1999) detailed analysis are: (1) Invariance, (2) Repetition, (3) Specialized time, (4) Specialized place, (5) Stylized behavior/stylized form (6) Performance and participation, and (7) Form that can hold and transfer a canonical message. I would argue that there is enough evidence in the Great Hunt Panel to meet the first five criteria. The invariance and repetition of highly stylized form that allowed the inference of the specialized time of winter, together with the uniqueness of the location defined by the unique form of the panel itself. Meanwhile the setting of the site makes it very suitable for performance of the art and its rituals as well as participation by a group of observers. Following a previous precedent applied to the art of Parpalló (Davidson 2012a), I infer that there are sufficient grounds to suppose that the rituals carried and transferred a canonical message. I suggest that there may be another subtlety here: that the original ritual could have concerned the main panel, including the “human” figure

28

IAIN DAVIDSON

in the headdress, and that that panel was not to be interpreted as a scene except through the orally conveyed ritual. The addition of the agentful human figures to the right of the stylized animals would have turned it into a scene that could be interpreted without the ritual. This might suggest that the relationship between ritual and narrative is an important consideration.

Consequences There is a contrast between panels in which interactions among the images are easily interpreted in terms of the actions between the individual images—the agency of the fawn and hind at Parpalló, or of the lions at Chauvet—and others, such as single-image panels, where such agency, if it did exist, was contained in narratives that now are completely lost. The more common exception appears when there are interactions involving people as agents in the images—as in the Well at Lascaux or the Great Hunt Panel in Cottonwood Canyon. The representation of the agency of humans or animals is what distinguishes scenes in a panel from nonscenes. Images in which agency appears to be represented are more easily interpreted by observers separated from the context of their production. Images that do not appear to represent agency are difficult for observers to interpret, if they themselves were not participants in the production. It may be that for some panels it would be possible to construct arguments about the nature of ritual in the use of the panel, thus allowing inference about the context of production and the interaction between artists and audience. Several consequences of this analysis need to be explored cross-culturally in the past and the present to see whether there is any patterning. In particular, what is the evidence that panels from one time were different in this respect from those of another time? Is there evidence that particular panels changed through time? Is there other evidence that the cultural rules and contexts of rock or cave art production changed through time at any spatial scale? Did the appearance of scenes representing the agency of humans with respect to animals mark a change in the restrictions associated with the communication of ritual knowledge? To what extent did there exist panels that were transformed from ritual exercises into scenes by the addition of non-stereotyped figures? As interesting as these questions are, they are simple approximations to interpretation dependent on criteria created in a completely different (albeit informed) cultural context. What can be recognized as scenes in rock art depends on how agency is identified in traditions of visual interpretation that differ from those in our own culture. There quite possibly were no cultural rules by which scenes and narratives could be inferred from images that appear not to represent agency, but it is difficult to imagine how we could demonstrate that. One tantalizing speculation about these variations in rock and cave art would be to argue that just as the emergence of naturalism through the application of perspective is said to have created new ways of representing and seeing the world in the European Renaissance, so changes in the ways images of animals were represented with other animals (including humans) probably can testify to changes in the ways people of the deeper past saw the world. It might be possible to construct an argument that much of the art that does not represent agency instead injects agency into the panels through ritual or other exegesis. The emergence of art with agency heralded a very different social context for the art, making it secular and available for unrestricted view. It opened up the possibility of representation as an end in itself.

Acknowledgments I acknowledge the Musqueam, Squamish and Tsleil_Watuth (Tslay-wah-tooth) peoples, past and present, who are the First Nations people of the unceded traditional territory where the initial discussions that led to this publication took place. In improving this chapter, I received help, comment or advice from Jesús Altuna, Helen Arthurson, Jean Clottes, Elisabeth Culley, Peter Jarman, Koro Mariezkurrena, Ana Belén Marín Arroyo, April Nowell, June Ross, Jerry Spangler, and Valentín Villaverde. I hope I have not misinterpreted them here.

SCENES AND NON-SCENES IN ROCK ART

29

Iain Davidson has worked on Spanish Upper Palaeolithic Art, archaeology and ethnography of Northwest Queensland, Australian rock art, colonization of Sahul, language origins, and cognitive evolution. He has taught in England, Australia, the United States, and Chile, and has worked with ten different Aboriginal groups and undertaken major archaeological consultancy for many of Australia’s leading industries. His publications include 4 books, 7 edited volumes, more than 60 chapters and 80 articles in journals. He is a Fellow of the Australian Academy of Humanities and was awarded the Rhys Jones Medal of the Australian Archaeological Association 2010. He held the Chair of Australian Studies at Harvard University from 2008–09.

Notes 1. Ignore for now the question of using the word art for this material. 2. https://www.merriam-webster.com/dictionary/scene Downloaded 29 November 2017. 3. “The Shaft in Lascaux is renowned for the presence of one of the rare obvious scenes in Paleolithic parietal art, even though it retains most of its mystery” (Clottes 2016, 49). 4. John Robb introduced this neologism during our initial discussions to mean “its quality of representing a scene.” 5. Unlike my earlier argument that Picasso could have seen one of the Parpalló plaquettes before he painted Guernica, there is no possibility that E. H. Shepard saw this plaquette, as his drawing was done half a decade earlier than its discovery during the excavation of Parpalló. See Davidson, I. (2005). 6. In an email to the author dated 18 December 2017. 7. “This is the moment when the female, finally consenting, walks near the male and rubs against him.”

References Almagro Gorbea, M. 1971. “La cueva del Niño (Albacete). La cueva de la Griega (Segovia). Dos yacimientos de arte rupestre recientemente descubiertos en la Península Ibérica.” Trabajos de Prehistoria 28: 9–62. Bahn, P. G., and J. Vertut. 1988. Images of the Ice Age. Leicester, UK: Windward. Barnard, P. J., I. Davidson, and R. W. Byrne. 2016. “Toward a Richer Theoretical Scaffolding for Interpreting Archaeological Evidence Concerning Cognitive Evolution.” In Cognitive Models in Palaeolithic Archaeology, ed. T. Wynn and F. Coolidge, 45–67. Oxford: Oxford University Press. Bofinger, E., and I. Davidson. 1977. “Radiocarbon Age and Depth: A Statistical Treatment of Two Sequences of Dates from Spain.” Journal of Archaeological Science 4: 231–43. Boyd, C. E. 2016. The White Shaman Mural: An Enduring Creation Narrative in the Rock Art of the Lower Pecos. Austin: University of Texas Press. Chaloupka, G. 1979. “Pack-Bells on the Rock Face: Aboriginal Paintings of European Contact in North-Western Arnhem Land.” Aboriginal History 3(1): 92–95. Cheyne, J. A., L. Meschino, and D. Smilek. 2009. “Caricature and Contrast in the Upper Palaeolithic: Morphometric Evidence from Cave Art.” Perception 38(1): 100–108. Clegg, J. 1984. “Pictures of Bulls and Boats: Some Evidence of Prehistoric Perceptive Processes.” In Under the Shade of a Coolibah Tree, ed. R. A. Hutch and P. G. Fenner, 219–38. Lanham, MD: University Press of America. Clottes, J. 2011. Pourquoi l’art préhistorique? Malasherbes, France: Éditions Gallimard. Clottes, J., and M. Azéma. 2005. “Les images de félins de la grotte Chauvet.” Bulletin de la Société Préhistorique Française 102(1): 173–82. Clottes, J. 2016. What is Paleolithic art? Cave paintings and the dawn of human creativity, trans. O. Y. Martin and R. D. Martin. London, UK: The University of Chicago Press. Clottes, J., and J. D. Lewis-Williams. 1998. The Shamans of Prehistory: Trance and Magic in the Painted Caves. New York: Harry N. Abrams. Conkey, M. W. 1989. “The Structural Analysis of Paleolithic Art.” In Archaeological Thought in America, ed. C. C. LambergKarlovsky, 135–54. Cambridge: Cambridge University Press. Davenport, D., and M. A. Jochim. 1988. “The Scene in the Shaft at Lascaux.” Antiquity 62(236): 558–62. Davidson, I. 1995. “Paintings, Power, and the Past: Can There Ever Be an Ethnoarchaeology of Art? [Ancestral Connections: Art and an Aboriginal System of Knowledge. (H. Morphy)].” Current Anthropology 36(5): 889–92. ———. 1997. “The Power of Pictures.” In Beyond Art: Pleistocene Image and Symbol, ed. M. Conkey, O. Soffer, D. Stratmann, and N. G. Jablonski, 128–58. San Francisco: The California Academy of Sciences.

30

IAIN DAVIDSON

———. 2005. “The painting and the tree: symbolism in the Upper Palaeolithic. A tribute to a great Basque scholar.” Munibe (Antropologia-Arkeologia), 57, 197–205. ———. 2012a. “Symbolism and Becoming a Hunter-Gatherer.” L’art pléistocène dans le monde / Pleistocene art of the world / Arte pleistoceno en el mundo, Actes du Congrès IFRAO, Tarascon-sur-Ariège, septembre 2010, Symposium « Signes, symboles, mythes et idéologie. . . ». N° spécial de Préhistoire, Art et Sociétés, Bulletin de la Société Préhistorique Ariège-Pyrénées LXV-LXVI:Book, 292–93, CD: 1689–1705. ———. 2012b. What a Carry On? Portable Art and Changes of Symbolic Meaning. L’art pléistocène dans le monde / Pleistocene art of the world / Arte pleistoceno en el mundo, Actes du Congrès IFRAO, Tarascon-sur-Ariège, septembre 2010, N° spécial de Préhistoire, Art et Sociétés, Bulletin de la Société Préhistorique Ariège-Pyrénées LXV–LXVI: Book 268–69, CD: 1559–70. ———. 2013. “Origins of Pictures: An Argument for Transformation of Signs.” In Origins of Pictures: Anthropological Discourses in Image Science, ed. K. Sachs-Hombach and J. R. J. Schirra, 15–45. Cologne: Halem. Davidson, I., and A. García Moreno. 2013. “La excavación arqueologica de la Cueva del Niño (Ayna, Albacete) de 1973: secuencia estratigráfica y materiales.” Al-Basit 58: 91–117. Dobrez, L. 2012. “Towards a More Rigorous Definition of Terms: Are There Scenes in European Palaeolithic Art?” L’art pléistocène dans le monde / Pleistocene art of the world / Arte pleistoceno en el mundo, Actes du Congrès IFRAO, Tarascon-sur-Ariège, septembre 2010, Symposium « Signes, symboles, mythes et idéologie. . . ». N° spécial de Préhistoire, Art et Sociétés, Bulletin de la Société Préhistorique Ariège-Pyrénées LXV-LXVI:Book: p. 316–17, CD: 1837–51. Evans, N. 2010. Dying Words: Endangered Languages and What They Have to Tell Us. Malden, MA: Wiley-Blackwell. Fortea Pérez, J. 1978. “Arte Paleolítico del Mediterraneo español.” Trabajos de prehistoria 35: 99–149. Fowles, S., and J. Arterberry. 2013. “Gesture and Performance in Comanche Rock Art.” World Art 3(1): 67–82. Fritz, C., and G. Tosello. 2001. “Le secteur des Chevaux.” In La Grotte Chauvet. L’art des origines, ed. J. Clottes, 106–17. Paris: Seuil. ———. 2007. “The Hidden Meaning of Forms: Methods of Recording Paleolithic Parietal Art.” Journal of Archaeological Method and Theory 14: 48–80. Fritz, C., G. Tosello, and M. W. Conkey. 2016. “Reflections on the Identities and Roles of the Artists in European Paleolithic Societies.” Journal of Archaeological Method and Theory 23(4): 1307–32. García-Diez, M., and M. Vaquero. 2015. “Looking at the Camp: Paleolithic Depiction of a Hunter-Gatherer Campsite.” PLoS ONE 10(12):e0143002. Gell, A. F. 1998. Art and Agency: An Anthropological Theory. Oxford: Clarendon Press. Gombrich, E. H. 1952. Art and Illusion: A Study in the Psychology of Pictorial Representation. London: Phaidon. Green, J. 2014. Drawn from the Ground: Sound, Sign and Inscription in Central Australian Sand Stories. Cambridge: Cambridge University Press. Guthrie, R. D. 2005. The Nature Of Paleolithic Art. Chicago: The University of Chicago Press. Jarman, P. J. 1974. “The Social Organisation of Antelope in Relation to Their Ecology.” Behaviour 48(3–4): 215–66. Leroi-Gourhan, A. 1965. Préhistoire de l’art occidental. Paris: Mazenod. Malafouris, L. 2013. How Things Shape the Mind: A Theory of Material Engagement. Cambridge, MA: MIT Press. Marshack, A. 1977. “The Meander as a System: The Analysis and Recognition of Iconographic Units in Upper Palaeolithic Compositions.” In Form in Indigenous Art, Schematisation in the Art of Aboriginal Australia and Prehistoric Europe, ed. P. J. Ucko 286–317. Canberra: Australian Institute of Aboriginal Studies. Matheny, R. T., T. S. Smith, and D. G. Matheny. 1997. “Animal Ethology Reflected in the Rock Art of Nine Mile Canyon, Utah.” Journal of California and Great Basin Anthropology 19(1): 70–103. Milne, A. 1926. Winnie-the-Pooh. London: Methuen. Munn, N. 1973. Walbiri Iconography: Graphic Representation and Cultural Symbolism in a Central Australian Society. Ithaca, NY: Cornell University Press. Nomade, S., D. Genty, R. Sasco, V. Scao, V. Féruglio, D. Baffier, H. Guillou, C. Bourdier, H. Valladas, E. Reigner, E. Debard, J. F. Pastre and J.-M. Geneste. 2016. “A 36,000-Year-Old Volcanic Eruption Depicted in the Chauvet-Pont d’Arc Cave (Ardèche, France)?” PLoS ONE 11(1): e0146621. Nowell, A. 2015. “Learning to See and Seeing to Learn: Children, Communities of Practice and Pleistocene Visual Cultures.” Cambridge Archaeological Journal 25(4): 889–99. Pericot García, L. 1942. La cueva del Parpalló. Madrid: Consejo Superior de Investigaciones Científicas. Pigeaud, R. 2007. “Determining Style in Palaeolithic Cave Art: A New Method from Horse Images.” Antiquity 81(312): 409–22. Rappaport, R. A. 1999. Ritual and Religion in the Making of Humanity. Cambridge: Cambridge University Press. Robb, J. 2005. “Agency: A Personal View.” Archaeological Dialogues 11(2): 103–7. Ross, J. and I. Davidson. 2006. “Rock Art and Ritual: An Archaeological Analysis of Rock Art in Arid Central Australia.” Journal of Archaeological Method and Theory 13(4): 305–41. Sauvet, G., R. H. Layton, T. Lenssen-Erz, P. S. C. Taçon, and A. Wlodarczyk. 2000. “La structure iconographique d’un art rupestre est-elle une clef pour son interprétation?” Zephyrus 59: 97–110. Schiffer, M. B. 1999. The Material Life of Human Beings: Artifacts, Behavior, and Communication. London: Routledge.

SCENES AND NON-SCENES IN ROCK ART

31

Serpell, R., and J. B. Deregowski. 1980. “The Skill of Pictorial Perception: An Interpretation of Cross-Cultural Evidence.” International Journal of Psychology 15(1–4): 145–80. Severi, C. 2015. The Chimera Principle: An Anthropology of Memory and Imagination. Chicago: Hau Books and Carlo Severi. Spangler, J. D. 2013. Nine Mile Canyon. The Archaeological History of an American Treasure. Salt Lake City: The University of Utah Press. Ucko, P. J., and A. Rosenfeld. 1967. Palaeolithic Cave Art. London: Weidenfeld and Nicolson. Utrilla, P., C. Mazo, M. C. Sopena, M. Martínez-Bea, and R. Domingo. 2009. “A Palaeolithic Map from 13,660 cal BP: Engraved Stone Blocks from the Late Magdalenian in Abauntz Cave (Navarra, Spain).” Journal of Human Evolution 57(2): 99–111. Villaverde Bonilla, V. 1994. Arte Paleolítico de la Cova del Parpalló. València: Servei d’Investigació Prehistórica, Diputació de València

==== •

•

•

2. THE POSSIBLE SIGNIFICANCE OF DEPICTED SCENES FOR COGNITIVE DEVELOPMENT Livio Dobrez

Introduction: General Definition of “Scene,” Real and Depicted Is there anything special about scenes in art—in connection with human evolution? For those who want a definition of “art,” let us say: “any sort of formal arrangement, meaningful insofar as structured, which some of us decide to call art, whether before or after the fact, and possibly not without dissent.” However, it will help to orient the reader if I say at once that wider discussion of what we might mean by “art” (or associated terms such as “picture,” “depiction,” etc.) is not the issue here. For that I refer the reader to aspects of publications cited below and also to publications not otherwise cited (Dobrez 2007a, 2011a, 2013a). In this chapter the focus must be on the idea of a “scene.” Should we, then, be paying special attention to scenes in the early history of art? In any case, how are we to define a scene, the danger being, not least in rock art studies, that almost any figurative depiction may be described as a scene? I shall opt for what I take to be the most fundamental idea of a scene as a perceived event, stressing the necessity of tying this depicted event firmly to a real event. We know that we regard someone rushing to catch a bus as an event, and that a picture of someone rushing for the bus is likewise read as an event. Actually neurophysiology tells us that the connection between real and pictured is identifiable in the structures of the visual system (for a detailed account see Dobrez 2013b). We process a perceived X in the same neural area as a picture of X, that is, the same neuronal ensembles fire for both, though of course we do not confuse the two, since the information given in the two cases is not identical. In fact, while some neurons fire for the real thing and some for its depiction, some neurons, for example in connection with perceived motion, will fire indiscriminately for both. A single neuron has been isolated as firing for either the man catching the bus or a picture of the man catching the bus (Krekelberg, Vatakis, and Kourtzi 2005). This is astonishing, yet also perfectly in keeping with everyday observation: a perceived real galloping horse and a picture of the same both register as moving objects. And, more astonishingly, there is experimental evidence for it: Freyd and others have shown that we actually see movement in a still picture—and in a way that is measurable (Freyd 1983; Freyd and Finke 1984). I shall define a real as well as a depicted scene as an event, and one immediately, that is, unmediatedly, perceived. We do not need to infer that the horse is galloping. That would be too slow a process for evolution. No thinking is required: seeing is immediate or, as the perceptual psychologist Gibson (1950, 1966, 1979) would say, “direct.” In the broadest terms, an event or a scene simply amounts to something going on, something happening. This may be minimal (X sitting on a chair) or maximal (the Battle of Waterloo). So a scene requires action, that is, movement of some sort—even if, in the absolutely minimal case, that movement is arrested (to the query “what is X doing?” we reply “sitting,” so establishing it as a form of “doing”). Whatever we decide about borderline cases, it follows that depicted motion, for example a picture of a galloping horse, constitutes a scene. However, for interaction (over and above action) we require at least two objects, say, one horse chasing another, that is, doing something to it. At that point we perceive not merely movement but a cause-and-effect relation, and, as shown by Michotte’s (1963) perceptual experiments, this too is not a matter of making inferences but, as with motion, of immediate seeing (see Dobrez 2013b for discussion of Michotte). Again, the evolutionary rationale for direct rather than inferential visual processes is evident: you cannot wait for a chain of logic to inform you that the lion is charging.

THE POSSIBLE SIGNIFICANCE OF DEPICTED SCENES FOR COGNITIVE DEVELOPMENT

33

A Perceptual-Phenomenological Methodology Thus a picture of a scene involving either action or interaction, whether minimal or maximal, will be read like the real thing, with appropriate modification relating to the nature of the visual information supplied in each case. I emphasize our perception of the scene, real or depicted. Most rock art research takes perception of the rock art for granted without theorizing the issue of perception. For the most part it follows the logic of a subjective/objective binary that is commonplace in our culture, the assumption being that the two are mutually exclusive, that only objective (read experimental) data is reliable, and that subjective (read observational) data is to be regarded with skepticism. However, since I take philosophy seriously I am more inclined to be skeptical about the binary itself (2011b). Of course, experimental, that is, quantifiable evidence will settle certain issues (at least pending falsification). Likewise, everyday non-experimental observation cannot be relied upon without reservations. After all, it can get things wrong. Hence my ready appeal, on occasion, to perceptual psychology experiments, as well as to the substantial data currently available from work in neurophysiology. Yet in the end, with rock art, we continue to rely on plain observation, having no other choice. It seems to me as well neither to accept our observation uncritically nor uncritically to dismiss it, but rather to attempt to theorize the entire operation, perception included, so as to have some ideas about what we might be doing when we observe something, art in particular. I have called it “looking at our looking,” appealing to Husserl’s phenomenology (2001) as a strictly descriptive philosophical methodology for analyzing the structures of our mental activities—in the present case, what we do when we look at rock art. It is critical to grasp this definition, since it bears little resemblance to general usage in rock art studies, where “phenomenological” may mean no more than “experiential” and indeed may lead to easily parodied notions of “immersion” in the landscape as a supposed aid to understanding the hunter-gatherer past (see Dobrez 2016 response to Lowish). Some social sciences scholars have used the term judiciously; for example, Tilley (2004), though entirely via the French version of phenomenology practiced by Merleau-Ponty (1962), and Julian Thomas (1996), via the Heideggerian version (1962) (Heidegger unfortunately read through Derrida’s tendentious borrowing of his ideas). There are a number of other variations on the basics of phenomenology (with Sartre 1969 as the most famous), but my approach relies on the original Husserlian prototype, in particular as set out in the Logical Investigations (for more detail see Dobrez 1986, 2015a). It may be helpful to add that phenomenology of one sort or another is behind the reception theory that dominated hermeneutics in the last century, just as historical hermeneutics dominated the nineteenth century. This fact raises an important issue, since most rock art work continues to be directed to historical ends. At the same time, however, historical hermeneutics cannot assist my present argument. The historian observing, say, a picture of the Battle of Waterloo will raise issues of content (When? Where? Who?). But this procedure is of no use if we are looking for a definition of the term “scene.” It is not geared to answering questions such as, is this picture of Waterloo a scene, that is, a depicted event, and if so why or how is it a scene? For an answer to that sort of query we need to gloss over issues of content, shifting attention from historical to universalist considerations to ask what precisely it is that makes it a scene. I have already put my preliminary case in the above definition: a scene is a pictured event, seen in the same way as an event in life, though with appropriate adjustments (see Hagen 1986: 8 on the “equivalence of ordinary and pictorial information”). This means that a scene is seen directly, without inferential mediation; and the markers for it are depicted movement (see Dobrez 2011c for an account of this) and, where there is depicted interaction, depicted causality (Dobrez 2013b). In choosing to follow a universalist line on analysis of pictures, that is, not asking questions about context or content but about the ultimately biological perception of the object, I of course imply no criticism of cultural/historical procedures, which are critical to rock art studies. Rather my concern is to add one further way of dealing with the phenomenon of rock art. At the same time, I do not accept that cross-cultural studies provide any evidence that some humans perceive differently from other humans. Given the European colonial/imperial history of the last two centuries, it is unsurprising that Layton (1977) sharply criticized Dere˛gowski (1980, 1984) when the latter seemed to suggest something of the sort about African subjects (specifically on the basis of their perception of depicted perspective). For my part, I have commented on both the original thesis and the exchange, concluding that the claims Dere˛gowski actually makes about cultural

34

LIVIO DOBREZ

intrusion in visual perception are more modest than may at first appear (Dobrez 2010a). Nonetheless, my total skepticism about differences in the actual biology of seeing, as distinct from cultural variations (e.g., the treatment of depicted perspective) remains firm, though naturally highly idiosyncratic novel forms of depiction (such as photographs) sometimes require habituation. Leaving aside unfortunate political implications, claims of actual biological differences go against the premise of a phenomenology-based universalist reading of perception and, as explained below, against current knowledge of the neural substrates for processing visual information. But what precisely are the implications of a universalist approach? For Husserl these relate to claims for the validity of certain types of philosophical reflection that constitute his methodology. For someone in my position, seeking to apply Husserl a century later, the methodology requires, and indeed receives, support from biology (Dobrez 2013b). I cannot stress sufficiently, however, that my appeal to neural structures in this connection refers entirely to the visual system, not to the brain as a whole. We know that this system has remained much the same for c.20Ma, the time span separating us from Old World monkeys, because after we turned from the invasive brain mapping practiced on monkeys in the late twentieth century to non-invasive fMRI, which made it possible to map the human visual system, we found fundamental homologies (for an account of relevant experiments see 2013b). That indicates that the basic elements of visual perception, which must include event perception, are hardwired. Cultural change cannot alter what is hardwired, though this does not preclude superficial rewiring on an ongoing basis and, additionally, situations which might bring epigenetics into play. Naturally, biology and culture work together in real life, comfortably or otherwise. We see an event with our existing neural apparatus, but over and above that we understand its cultural significance on the basis of more or less specialized knowledge. So, while hardwired and non-hardwired elements interact at any point in our lives, biology and culture remain logically distinct. This logical distinction is one I maintain throughout the argument that follows. For present purposes a universalist approach of the sort I propose has some advantages. For a start, it (a) relates analysis of rock art to analysis of any other form of art. Equally importantly, it (b) relates art discourse to discourse of real-life situations, which in turn (c) allows depicted, like real, situations to be read in the context of evolution. Just as it matters to read events in life (the fact of something taking place even before we are sure of its precise nature), so it matters to read events in art, which is why we have no trouble reading a depicted event as an event, and why, at the most basic level, spotting a scene is an important operation. Since perception is biological before it becomes cultural, and since culture, unlike biology (usually), may change very quickly, we may understand that although the interpretation of rock art content calls for maximum caution, even for recent art, discussion in terms of reception, provided it keeps within boundaries of visual biology, is in principle reliable. Putting it simply: whatever its (inevitably variable) cultural meaning in any given case, a depicted circle remains just that and would have been seen as a depicted circle by its makers. But does this conclusion leave the researcher into visual universals anything to say, other than to repeat, echoing Gertrude Stein, “a circle is a circle is a circle . . . ”?

Visual Markers for a Depicted Scene The answer to the question above is that it leaves a great deal to say. Sometimes in collaboration with Patricia Dobrez, I have set out to outline a taxonomy of pictures on the basis of universalist assumptions. This has entailed identification of time/space ubiquitous types of art (not “styles”—a term better applied to culturespecific analysis), with special status accorded to rock art for its time-depth. The result has been analysis of images referred to as Canonicals, Narratives, and Performatives. Naturally the present argument concerns itself with Narratives, that is, pictures that tell a story—in short, scenes. What, then, are the visual markers for a scene and how exactly do they constitute a scene as a scene? Putting it another way, how do scenes work? For further details the reader is referred to earlier publications (2007b, 2008, 2010a, 2010/11, 2011b, 2011c, 2012, 2015b, 2015c, 2015d, 2016). In particular, I refer to an article for the online journal Arts (2013b) that aimed to give a substantially detailed account of how we see motion in a “still,” motion being, as argued above, the primary indication of the scenic. At the start I distinguished depicted movement from illusory movement (as found in, say, a Bridget Riley painting), since, contra Gombrich (1960) and pace

THE POSSIBLE SIGNIFICANCE OF DEPICTED SCENES FOR COGNITIVE DEVELOPMENT

35

ideas current since the Italian Renaissance, I do not think art has ever been primarily about the creation of illusions of reality. I also distinguished depicted scenes from symbolic signs. Thus a road sign of a stick man with a shovel is not read as a scene of roadwork but as a warning for the motorist to slow down. Signs of this sort depend entirely on cultural knowledge and so fall outside the scope of the present investigation. Over and above analysis of perceptual phenomena, I sought help from perceptual experiments, including Freyd’s mentioned above, and also Freyd (1987) which concluded that mental representations are themselves dynamic and that simple but very clever experiments can actually measure the extent to which the visual brain shifts that galloping horse in the picture. Likewise, I sought help from Johansson’s (1973) point-light experiments showing that we register motion before we are aware of what it is that is moving. This fact of registering motion before objects is neurally explained by Ungerleider and Mishkin’s (1982) discovery of a bifurcation of processing pathways, one more concerned with object specification and the other more with the specification of motion—the point being that the latter is faster than the former, and for sound evolutionary reasons. Since details of this neural geography are now well known, at least in broad outline (see the account in Dobrez 2013b), we are in a position to note that a scene, that is, depicted motion, will be first processed as to basic form in the striate cortex V1 (in the occipital lobe), then expeditiously passed on to the superior temporal V5 for motion processing (“galloping”) and, finally, to the inferior temporal TE for object processing (“horse”). Given that reading the scene requires that we register motion integrated with perception of what it is that is moving, connections between the superior temporal and inferotemporal paths will also be activated. The speed of the operation, involving feed-forward mechanisms, may at various points encourage us to grasp the essential nature of human perception as anticipatory. Here we may pose a blunt question: how, in this context of neural structures, are we to understand the fact that a picture may be taken as a stand-in or equivalent for the real thing? I noted above that neurons will fire for real or depicted objects or both, but there is a larger observation to be made. Quite simply the brain works by a near-enough–good-enough principle. This because it has been fashioned by its context over time and in varying circumstances, the main engine of the process being that of survivability. Thus the main issue is not so much “getting it right” as “not getting it disastrously wrong.” At the same time, and indeed in tune with this imperative, the brain is called upon to be entirely active in any operation, not least that of perceiving. The fallout for perception of pictures is that the mechanics of generating a scene, that is, reading a scene in a series of marks, depends first and last on operations performed by the visual brain. Seeing a scene is not a matter of passive stimulus reception; it is something the brain is required to be doing. This is not to say it may operate arbitrarily, which would be self-defeating. With respect to that vital event-perception as transferred to a scenic picture, there are rules to be observed. Briefly, scenes require movement markers in the form of asymmetries, both with respect to depicted figures (bent/ angled limbs) and overall panel structuring (imbalance, diagonal compositions). The former results in what is referred to as a “peak shift” (Ramachandran and Hirstein 1999)—or, in sports photography, “peak action” (Figure 2.1), where action is ideally shot at that maximal moment immediately prior to diminishment. No doubt the most common example of this is the depiction of a runner with legs outspread so as to almost form a single line, a convention found from the Spanish Levant to southern Africa to northern Australia. However, scenes also require the generation of represented dynamic space, without which motion will not register. Thus space must contract in front of the galloping horse and expand behind it. Yet there is no way of depicting this as such: working on the visual prompts, the viewer has to bring it into being. Indeed, the active nature of the operation is nicely illustrated by the fact that the viewer is by no means forced to see the horse as in motion; instead she is capable of throwing a switch and freezing the horse. I shall return to this point. Another factor that comes into play with scenes is profile depiction, which prompts the reading of motion in dynamic space across the viewer’s visual field—this inevitably being the type of motion generated by a scene. (Pictures which challenge movement across the field, e.g., by means of movement toward the viewer, constitute a different category of action, one I have termed Performative and proposed as incompatible with the scene (Dobrez 2013b).) Of course, in connection with figural orientation, profiles are necessary to sustain the generation of causal interaction: only profile figures prompt the sense of causality, one figure acting upon another. In this connection it should be added that distances between participants in depicted action must be optimal so as to sustain the sense of cause-and-effect relations. In real situations causality is specified by temporal intervals; in pictures it must be specified by spatial ones. Too great a distance between the spear

36

LIVIO DOBREZ

FIGURE 2.1. “Peak action.” Copyright David L. Minick. Total Access.

thrower and the target destroys a causal reading; too little distance confuses it. Perhaps underlining several of the above points is the further consideration that scenes are viewed as through a “window.” What this means is that they necessarily exclude the viewer. The viewer cannot enter the scene and the participants in the scene cannot exit. If the viewer enters or the participants exit they bring about the phenomenon known in theater and film as the “breaking of the fourth wall,” as when a character on the stage or in the story directly addresses the audience. The effect destroys the integrity of the scenic and substitutes a Performative interaction, namely one in which the figure acts not within the scene but on the viewer. Many depictions play with this effect, not least religious icons or celebrated propaganda posters (“Your Country Needs You!”). It is worth adding that the size of figures has a part to play in the generation of the scenic. Size in depictions is relative to several factors, but in general we may observe that where frontal Performatives and, frequently, Canonical images tend to be rather large in rock art, figures in scenes tend to be small. That ensures that they are, unlike other categories, more likely to be viewed in closeup (and not approached, as it were, from any distance), with the optimal viewpoint probably being at eye level—whereas classical Performative frontals will be viewed from below. Unlike large, confronting frontals, figures in a scene are not intended to loom over the viewer but rather to be observed from that contemplative “window” that implies non-involvement, and therefore security, for the viewer. Two further questions arise. What is it that integrates all these observations and, in a picture, integrates the many markers/prompts just listed? What determines how markers themselves will interact? Clearly what integrates a single figure in a scene is simply its state of being in motion; what integrates a larger composition involving interacting figures will be depicted motion and depicted causality. The Gestalt principle of Grouping has a role here, and in this connection I refer the reader to Lenssen-Erz (1992). But in my view the Gestalt principle by itself does not entirely account for full integration of individual figures into that visual package, “a scene”; concepts relating to motion and causality depiction are also required to explain the scenic effect. This is in view of the fact that the brain works with pre-packaged wholes (of which the scene is

THE POSSIBLE SIGNIFICANCE OF DEPICTED SCENES FOR COGNITIVE DEVELOPMENT

37

one visual example). It needs to avoid repeating unnecessary operations that might be time-consuming and thus bad for survival. So it may be presumed that the entire visual package of reading “event” or “scene” is hardwired and readily neurally available. There is no need to work it up from scratch bit by bit each time (Dobrez and Dobrez 2013a). On the question of relations between markers it suffices to say here that some will override others, some will in any given case be missing, so requiring others to compensate for them. The main point here being that, as I have argued, the perceptual operation in question is not a mechanical one and therefore is not reducible to additive factors. Quite simply, seeing a scene demands a perceptual judgement, a specific activity of the human brain, without which it will not work. Let me note once more, as an addendum to this section of my argument, that I have studiously avoided confusing universalist analysis with historically contingent explanations that might conceivably be put forward, the chief one being the idea that a scene has some connection with post-Renaissance perspectival conventions of “realism.” Nothing could be further from the truth. European art in the last few hundred years has featured many scenes in conjunction with perspectival conventions. But while you may fit a scene within those conventions, they do not suffice to define a scene, and scenes are perfectly possible, as demonstrated not least by rock art, without them. The above, then, suggests the nature of the constitution of a scene, of its visual markers or more precisely the type of mental operation required to constitute a scene as a scene. But might there be a case for arguing that the capacity to perform this operation constitutes us as “ourselves,” that is, constitutes human modernity? It must be stressed at once that, as pointed out above, the visual system has changed very little over a long period of time, though of course other areas of the cortex have changed a great deal. But this means that we are not talking about a human development in seeing. A monkey sees an event taking place just as I do. It registers in its visual areas, processing perceived movement (another monkey leaping to steal its fruit), as well as causality (awareness of cause/effect relations between itself, the other monkey and the fruit)—all the markers listed above. The relevant difference for the present discussion is that humans make exograms: we depict the events both we and the monkey see. I shall return to the question but here wish to place event-depiction in a more specific, stage-by-stage, perceptual context and eventually ask if an idea of cognitive development might emerge from that.

Perceiving “Composition” A depicted scene does not come out of nowhere, and we may isolate some of the stages that go into its making. The first would be the basic operation of seeing a composition. The first thing we do when we observe a picture is to turn attention to it. Attention is a not easily understood neural operation on which we rely at any perceptual point in our lives (and not merely perceptual). With a picture, attention takes the form of a framing operation (for more detailed treatment of perceptual “framing” see Dobrez 2010b). It is important to understand that this is a mental operation and one not requiring an actual material frame. Of course, in some historical cases a frame (of wood or some other support) is provided, such as to enclose the image and in this way underline the mental operation of framing. With rock art the image may be enclosed by a line or by the natural size/shape of the rock support. But for the most part, no form of overt or material frame comes into play, and in any case it is not necessary for the operation of framing. What framing amounts to is a conscious or unconscious attentional judgment that the image is set apart from its surroundings: framing constitutes the image as “in” and the rest as “out.” Putting it another way, it simply focuses attention on what is “in,” constituting the picture as such. This is not to say that the image has no effect on its surroundings, which in rock art will of course generally be of critical importance, not least for cultural/symbolic purposes. However, we are here dealing with the perceptual, not the cultural, and in that context the effect of the image—in its natural surroundings yet simultaneously declaring itself as different, i.e., a sign of human presence—is to organize those surroundings. Seeing the image as “different,” you see it in relation to what is around it (“not it”) and you see what is around it in relation to the image. In short, the image functions (what else?) as a sign of integrating mind. This is the most basic act of composition-making, and it need not be intentional, since it simply follows from the act of putting something non-natural in a natural context. (Actually, any perceptual act organizes its visual field in a comparable way, but this is not of key interest here.)

38

LIVIO DOBREZ

I intend to refer to any mark or marks as a composition, regardless of whether it was intended as such, and the reason for this is that composition is in the eye of the beholder rather than in the picture. Of course, the more complicated the image, the more likely it must be that the composition was intended. But since my approach here is independent of historical reconstruction, I need not concern myself with intention (on the hermeneutics of intention/reception see Dobrez 2011b). Some years ago Patricia Dobrez and I, walking in the Canberra botanical gardens, were brought up short by a fine scatter of white on the grey pavement remarkably like a Pollock painting but produced by bird droppings. We may assume the striking pattern was not intended, but it was certainly a composition since we read it as such, and no less a composition for our having to read it. Subsequently we participated in a zoological play experiment involving meerkats scurrying over paint so as to leave marks on sheets of canvas, and the same principle obtained here (see Dobrez 2016). I stress the above examples because the operation of making a mark whose effect is to organize surrounding space, initially doubtless as figure/ground, is not merely the logical first step of picture-making, but also in fundamental respects something that anticipates all further elaboration. In summary, then, the inclusion/ exclusion activity of framing generates pattern at its most basic, since it sets a mark in relation to its support and, in the case of rock art, in relation to the entire landscape, which visually extends the structuring effects of the image. (For more on this see Dobrez 2010b.)

Perceiving “Juxtaposition” and “Association” Patterning within the frame may be categorized into three types. The first would be a series of marks that need not resemble anything in particular, as in some twentieth-century abstracts. Pattern may be discerned in the formal relations of shape, color, texture, etc., but the composition will be either more or less loose. If removal of certain details makes little or no difference to the whole—though strictly speaking any change will make some difference, however minimal—we may think of the composition as guided by a general rule of “juxtaposition.” In order to establish necessary formal connections between the marks, which I term “associations,” we must meet the requirement that change to any detail has the effect of altering the rest. For example, in rock art we may give the marks the form of a mammoth or horse, in which case some of these marks will be salient for identification of mammoth (humps, line of trunk) or horse (curve of head and neck). Indeed, for most animals this will entail a particular cervico-dorsal pattern, and, for humans, stick-figure frontal configuration (see discussion of this under the rubric of “canonical form” in Dobrez and Dobrez 2013a, 2013b, 2014). Removal of hump will negate identification of, say, mammoth or bison, so lines essential to a particular canonical form constitute salient associations. Thus association takes patterning further than juxtaposition. Of course the two can coexist. For example, many rock art panels depict more or less identifiable motifs. But the overall composition does not stress their interconnectedness. Remove one motif, such as a bison, from a Niaux panel (Figure 2.2)—or a barramundi from an Arnhem Land panel—and other motifs will not be significantly altered. So while the compositional principle at motif level is one of association, it is one of juxtaposition for the panel as a whole. I think it helps to make this distinction between juxtaposition and association in order to bring out the steps that go into the perceptual constitution of a picture. In the past I have given examples of nested figures as associations (e.g., the Pedra Furada deer, Brazil), but lineups, whether of humans at Ubirr, Australia, or mammoths and rhinos at Rouffignac, France, or simply cupules at Daraki-Chattan, India, constitute the most obvious example (Figure 2.3). The point is that the brain always works to see something—whatever it happens to be—as a whole. In some instances it has to work very hard to find a pattern (in a “juxtaposition”) and in others less so (in an “association”). It goes without saying that an association is more likely to be intentional than a juxtaposition.

Perceiving “Scene” But what about the scene? I return to a Niaux example that strikes me as illuminating in this context. The two bison at bottom right (Figure 2.2) are facing each other. If we see that facing merely in the formal sense, then we have a good case for association: remove one bison and the other is altered—there is no longer any

THE POSSIBLE SIGNIFICANCE OF DEPICTED SCENES FOR COGNITIVE DEVELOPMENT

FIGURE 2.2. Niaux (model), Parc de la Préhistoire, Tarascon-sur-Ariège, France. Copyright Livio Dobrez.

FIGURE 2.3. Daraki-Chattan, Madhya Pradesh, India. Copyright Livio Dobrez.

39

40

LIVIO DOBREZ

“facing.” But if we see their facing in the sense of an activity, we still have an image association, but also something over and above that. The two bison are doing something. What, then, is the difference in seeing? It suffices to recall the definition of a scene given above. It is true that there is no depicted movement in the facing bison, so it is fair to say that, as a scene, this will be classed as borderline. But even so, if we read facing as doing, then motion, at least implied, will be read into the image: it is an action. Likewise causality: the two bison are interacting. Thus a depicted scene may be understood as a formal association of marks indicative of motion (in the above case only by implication), that is, of action—and motion relative-to, that is, interaction (Figure 2.4). Interaction generates a reading of causal relations, which of course can only happen where the scene includes more than one figure. Just as an association of marks is more plausibly read as intentional than is a set of juxtapositions, so markers of movement and, where appropriate, causality are more plausibly read as intentional than associations. However, the shift from juxtaposition to association to scene should not be thought of as quantitative. Rock art scholars distinguish between “simple” and “complex” images, but it must be realized that no amount of added complexity, i.e., more lines, will of itself transform a set of juxtaposed marks into necessary associations, or associations into scenes. Maynard (1979) thought of depicted motion as an element of “complexity,” and one can see the appeal of the idea. But it must be evident that no mere addition of detail will transform a static figure into a moving one, or a depicted action into an interaction. In each case we require more specific visual markers and above all we require these as prompts that will, as it were, throw the perceptual switch—the operation performed by the viewer—to say “something is happening . . . it’s a depicted event” (Figure 2.5). To grasp this, it suffices to conduct a simple experiment. Try to see the activity of Figure 2.5 (possibly a birthing) as static. Or—to return to the galloping horse—try to “freeze” it. You can do it. You can even try to see both scenes as going backwards. While this is impossible with a real event, it is doable with a picture. But you have to work at it because you are reading against the markers of motion, and it is easier to go with them than not: it is easier to see the birthing and the depicted horse as in motion rather than as frozen. Still, the other option is there, and it reminds us, as noted above, that seeing is not a passive activity. It is we, that is to say our brains, that respond to prompts—for instance by remaining free to refuse response—simply by throwing an attentional switch, focusing on the static surface qualities of the image, that is, those markers that inform us that it is not a real birthing or galloping horse at all, but only a picture.

FIGURE 2.4. Tandjiesberg, Orange Free State, South Africa. Copyright Livio Dobrez.

THE POSSIBLE SIGNIFICANCE OF DEPICTED SCENES FOR COGNITIVE DEVELOPMENT

41

FIGURE 2.5. Pressa Canyon, Texas, USA. Copyright Livio Dobrez.

Depicted Scene as a Cognitive Marker Granted that very particular mental operations are required for reading degrees of compositional structuring, from mere juxtaposition at one extreme to narrative depiction at the other, might we be in a position to argue for at least broadly equivalent or concomitant development in cognitive capacities? Do we require a certain level of cognition in order to depict events? And might we trace in rock markings development toward such a level of cognition, one that might define human modernity? To investigate the possibility, we have to turn to a history of art making, and it may be as well to open the discussion with some comments on notable attempts to outline such a history. I shall briefly consider a few examples, eventually with the aim of judging how my approach, very different in its philosophical origin from other approaches, might nonetheless fit with these. As a preamble, though, I must point out that the major divide in theories of the origins of art and proposals for developmental sequences leading to both art and mind as presently understood is between those favoring a deep-time genesis and those pointing to (in Pfeiffer’s 1982 expression) a “creative explosion,” possibly very recent. I think there is a case for deep-time perspectives, not least because the presumed sudden change constituting the emergence of modern humans has repeatedly been pushed back in time. It is not necessary to go into details of the discussion, which are not of central concern in this chapter, but it does require mention as well as a brief return to the issue in due course. To begin with, however, and with some introductory comment, I shall focus on the way in which a variety of theorizers have approached the issue of modernity with reference to the role of art and then discuss some conclusions they have drawn as to its connection with cognitive change. Is it possible to source the nature of such cognitive change to fundamental changes to the human brain? That depends on the time scale we are using, there being no evidence for recent developments of this kind. Of course we can point to differences between the human brain and that of our near relatives, the great apes, as

42

LIVIO DOBREZ

well as that of monkeys. Particularly with respect to the visual system, study of whose neural structures began with work on monkeys, there is a deal of available knowledge (outlined in Dobrez 2013b). With respect to still closer relatives, all we can say is that the brain increased in size on two relatively recent occasions, perhaps beginning with Homo habilis and, after that, culminating in Homo neanderthalensis (Mithen 1996), at which point it reached something like its present dimensions. (For a more detailed, if still provisional, chronology see Stringer 2011.) Mithen (1996) has speculated that very recent cognitive development has taken the form of greater interconnectedness between brain areas, which could be true, whatever we may think of his analogy with Romanesque and Gothic cathedrals. Donald (1991) suggested that the recent change had less to do with neurophysiology than with the coming of the exogram, namely external memory systems—of which art, more narrowly defined, would be an example, and with which art, more broadly defined, would coincide. There is doubtless something in the argument that the exogram revolution, however it came about, has had an immense impact. At the same time, Donald is clear on this: neither the change nor the “theoretic” phase it is supposed to have generated has any genetic basis. It may of course be argued that it has had epigenetic effects, or simply that its effects are of the kind envisaged by Mithen. More on these points later.

Theoretical Approaches to Art as a Marker of Cognitive Development Mention of exograms takes us directly to a consideration of art making. The genesis of this activity has been variously located. Davis (1986) simply says that initial marks must have been made by chance and have only subsequently been read as significant forms. Versions of this thesis have been influential. Davidson and Noble (1989; see also Davidson 2010, 2013), whose focus is as much on language as art, accept that early marks may have come about by accident, though they initially set out to fill in much of the developmental gap left by Davis, notably by appeal to tool making as a necessary prerequisite to the making of art and by stress on the further mediation of gestural communication. Bednarik (1995) reads the Davis proposition in the direction of an activity akin to doodling, and while accidental marking and doodling are distinct activities (the one happens without attention, the other in tandem with attention on a parallel object) the concept of doodling may help explain early marking (Watson 2008). Even if we are not altogether convinced of this, the doodling argument has the advantage of a possible connection, exploited by Bednarik (1984, 1995, 2003), with the phenomenon of phosphenes—basic forms, presumably associated with early visual processing in the striate cortex or V1 (Hodgson 2000) and found in rock art motifs. I think the phosphene argument is compelling, in part because it gives substance to the notion that ur-marks may have come about automatically, that is, without specific intent—failing which we come up against the problematical idea of art-before-art. Phosphenes may also offer an explanation for early patterning. In terms of the juxtaposition/ association distinction, chance marks would presumably come into the category of juxtapositions which, strictly speaking, exhibit no self-evident pattern and accordingly will not be recognized as early art (though it is always open to the viewer to read them as compositions, i.e., patterns in the viewer’s eye, necessarily borderline). What can be recognized would be a set of marks exhibiting certain formal associations and so structure, though of course not without debate. Bednarik (1984, 1995, 2003, 2008) and others have proposed various candidates for early art, but in general this is an area where little consensus is to be found (for a cautiously balanced account of the polemics see d’Errico et al. 2003). One way or another, spontaneous generation of fundamental patterned forms such as phosphenes suggests a possible mediating link between the chance mark and something that can be read as having specific form, that is, between the principle of juxtaposition and that of association. It should be added here that Lewis-Williams (Lewis-Williams 2002; Lewis-Williams and Dowson 1988) has used the phosphene idea in connection with his trance explanation for the genesis of patterned forms. Let me say as an aside that, in the light of acquaintance with the literature of mysticism, I think the focus on trance might be productive, though the thesis of a threefold progression from phosphene forms to iconicity (itself presented on the basis of modest evidence) is simplistic. Essentially, Lewis-Williams proposes a virgin birth for art, since the entire process is contained in the trance experience. Moreover, leaving aside the other problematical idea of a shamanic outlining of post-factum but still vivid visual aftereffects (a rewrite of the dubious “eidetic image” thesis: see Willcox 1984), the Lewis-Williams

THE POSSIBLE SIGNIFICANCE OF DEPICTED SCENES FOR COGNITIVE DEVELOPMENT

43

explanation simply defers the difficult question of genesis from that of art to that of trance. On the plus side, however, it does locate beginnings where in all probability they should be located, in initially automatic actions. Naturally there remains the question of how such actions are transformed into intentional art making. Davidson and Noble (1989) offer a more detailed possible sequence of steps than most, with emphasis on iconicity as a significant marker. In the terminology of the present article, iconicity corresponds to those sets of salient associations that allow identification of an image, not necessarily without ambiguities. I assume this logic would carry over to the identification of a scene, which might be expected to illustrate the culmination of the process in Davidson and Noble’s scheme. Something similar but less clearly worked out emerges from Halverson’s (1987) developmental model, which puts emphasis on increased structuring and, somewhat by-the-by, leads to the scene—by accident, I think, because Halverson talks of a shift towards greater “composition” (which he takes to be necessarily intentional), assuming in the event that the terms composition and scene are synonymous. In light of arguments put forth above I must regard this as misleading, but my main concern here is to make the point that rock art scholars who take the trouble to hypothesize about the origins of art and who frequently agree on very little seem to have in common the view that, starting with basic, possibly automatic or semiautomatic actions, art making develops in the direction of tighter structuring. Automatic or chance marks are understood as having no cognitive component, and structured marks as indicating a cognitive component, that is, an intention. At the same time there is a premise of sorts that evidence for a developmental shift in the direction of structure exists in the art record. Some of those mentioned above might identify the level of organization that defines a depicted event or scene as (currently) the endpoint of the process. Others would probably argue that the cognitive level presupposed by apparently simple associations is already as high as that required for a scene. Both may be right. I have deliberately avoided the criterion of simple/complex in favor of a development from loose (juxtapositional) to tight (associational to scenic) organization. But overall I accept the general thrust of the views just described. Moreover, it does not seem unreasonable to tentatively read depictive change as illustrating cognitive change understood as directed toward further neural integration. But this would not be a development of “seeing” as such, since a monkey’s reading of a real scene is not inferior to mine. So am I simply more adept at making and reading an externalized image, that is, one not restricted to the mind? And given that Donald and others think of the first intentionally patterned mark (read “exogram”) not as merely matched by cognitive development but as causing it, could expertise in the use of exograms have played such a vital role in turning us into modern humans? That would follow from the views of a number of theorizers cited above, though the thesis would not necessarily be expressed in the same way by all of them. If there is a larger difficulty here, it does not relate to possible skepticism about the importance of the exogram revolution but to interpreting how exactly it might have caused cognitive change, provisionally assuming it has done so. For me that resolves itself, at least in the first instance, to questions about possible kinds of changes in neural structure. Donald (1991) saw exograms as a development in cognitive hardware rather than software, though given his insistence on their non-genetic basis, it is hard not to conclude that he worked the computer metaphor beyond its limits. At any rate what he had in mind was a behavioral shift. Mithen (1996) argued for a modern brain, defined as having more interconnections, on the basis of cognitive models aimed at interpreting a behavioral shift. Since in both cases revolutionary change is not tied to a fundamental biological shift (it being too recent for that), we might simply conclude that the change is not as revolutionary as has been imagined. On the other hand, this surely underrates the sheer volume of change, which in the end suggests qualitative difference. One way of making sense of this presumed qualitative shift would be to push the genesis of the exogram back in time, say to some phase in that deep-time plateau of brain enlargement which Mithen postulates as encompassing habilis and neanderthalensis (Mithen 1996: 8). It must be the case that at that point, “seeing” itself, that is, the visual system, remains much the same as before. Equally, however, much else in the brain changes in the direction of what we eventually recognize as modernity. In this scenario the most recent exogram revolutions, evidenced by known behavioral change, would represent unfolding stages of an earlier genesis. They might in some instances evidence concomitant cognitive change, and in others simply be understood as expressing well established but previously unutilized or underutilized capacities. Either way, a long-term perspective for the exogram revolution seems less daunting in the light of regular discoveries

44

LIVIO DOBREZ

that continue to expand developmental time scales, for example the Jebel Irhoud remains of “moderns” at c.315,000 years (Hublin et al. 2017). In the above scenario I am of course attempting to visualize the nature of the neural shift required for a thesis of revolutionary change. If the coming of external memory systems were premised on that original increase in brain size (even bearing in mind that brain enlargement may never be the whole story), we would go some way to justifying Mithen’s idea of greater neural interconnection. After that initial change in brain size, neural connections might be understood as continuing to alter all the way to the present, though at this later stage within a range referrable to current knowledge of neuroplasticity (Doidge 2007). At this stage it would make sense to argue for a different rather than a greater neural connectedness. Instead of focusing on biological fundamentals, we would refer cognitive development to culture-driven behavioral factors that would define it simply in terms of “learning new skills.” (See examples from neuroplasticity, e.g., the violinist’s brain having neural connections different from those of the non-violinist.) Modernity, in such a perspective, would acquire deep-time roots. In terms of my own argument for depictions as tending towards greater compositional integration, the genesis of exograms would be at the point when juxtapositional or chance markings become intentional, that is, when the observer registers mark-associations. After that, via development of associational principles in the direction of the scenic, it would be less a matter of fundamental biological change (as a consequence of the bigger brain) than of learning varied depictive skills. Would the above contribute to proving the case for exograms as causing cognitive change? It might, to the extent of suggesting possibilities of relating behavioral arguments, and their “soft” rewiring of the brain, to gene-based, and thus more strictly evolutionary, neural structures. I remain agnostic on all these issues, being less concerned with the exogram as cause than with ways in which it might simply illustrate a development thesis. All in all, then, I see my specific idea of a pictorial-compositional criterion for cognitive capacity, put forward in line with but distinct from other theoretical models, as having to be treated with extreme caution. For a start, and returning to another, as yet unanswered, question posed above, is it actually illustrated in the visual record?

Practical Difficulties for Theories of Art’s Possible Role in Cognitive Development Certainly I know of no location in the world where rock art may be found to progress neatly from (1) more or less chaotic juxtapositions to (2) more organized associations to (3) the related but different associational level required for event depiction. However, given the massive imponderables of limited data, taphonomy, and cultural variables, who would expect it? At the same time, it may be that steps in this hypothetical progression are at least roughly perceptible in some places. To try to demonstrate more than this, using all the available data, would require a long and ultimately almost certainly inconclusive discussion. I shall content myself with a few comments, asking the reader to bear in mind that these are directed solely to the argument proposed above, in particular the concept of juxtaposed, associated, and scenic compositions. Thus the brief discussion of depictive sequences in rock art that follows is not primarily concerned with the content of the sequences or even their accuracy. I simply take my cue from standard textbook chronologies (as presently more or less accepted), my emphasis being entirely on the degree to which these might relate to my argument. Though discussion of relative sequences necessarily involves reference to absolute dating, this last is also not of itself a consideration here. As argued, juxtapositions pure and simple, as with line markings, are liable to be perceived as random and thus will not readily be read as art, and if they are so read, they will generate debate. Juxtapositional or partly juxtapositional compositions, on the other hand, as in some European Palaeolithic examples (Altamira, Niaux) will be readily recognized. This is because, though the overall arrangement of figures looks random, the figures themselves are structured by principles of association, that is, they consist of associated salient features that identify them as whatever they happen to be (horse, bison etc.). With scenes, the overall arrangement itself is structured, in this case to identify an event taking place. Hence my argument for a shift in the direction of tighter structuring, based on the logic of reading a picture, which, if observable chronologically in actual depictions, might suggest a comparable shift in the direction of cognitive modernity. The question, then, is do we find the kind of depictive sequences that concern me in

THE POSSIBLE SIGNIFICANCE OF DEPICTED SCENES FOR COGNITIVE DEVELOPMENT

45

the real-world record, if only in fragmentary form and despite the vagaries of data, taphonomy and so on just noted? In minimal terms: do depictive associations generally predate scenes? Unfortunately, in northern Australia, where they are most evident, what I call scenes are placed early in the chronologies of the two most exhaustive original researchers, preceded only by iconic “large naturalistic” (Chaloupka 1993) or “irregular infill” (Walsh 2000) figures. With respect to Arnhem Land, single-figure and multiple-figure interactive compositions (all scenes by my definition) are currently dated to the Palaeolithic, though not to the extent presumed by Chaloupka (May et al. 2018). Dating work on the Australian Kimberley is very much ongoing, but at this stage Walsh’s general scheme of “infill” and Gwion (formerly Bradshaw) figures as appearing early in the chronology remains (Ross et al. 2016). So all this amounts to associational logic (iconic figures) followed by scenes—helpful to my case—but both being closer to the start than to the end of a sequence. In India there are scenes in the celebrated areas of Madhya Pradesh, some of which (featuring green paint) were placed at the start of a stylistic chronology by the pioneering scholar Wakankar (Brooks and Wakankar 1976). Though the antiquity of cupules at Bhimbetka is disputed and the case for their antiquity at Daraki-Chattan is circumstantial, cupules, whether regarded as potential candidates for early dates or not, are significant for me since they include fine examples of lineups, motifs in association (see Misra, Mathpal, and Nagar 1977 for an early account of excavations and Kumar et al. 2005 as against Lorblanchet 1999 for the dating dispute). In southern Africa there are innumerable painted scenes, regarded as relatively recent, with (as a straightforward example of associated salient lines), the earliest known iconics, at the Apollo 11 shelter, Namibia, up to 26,000 years old (Lewis-Williams 1983). (The case for these iconics is especially supported by a reading of the least ambiguous two of the seven pieces of mudstone as a single broken image.) Thus we have a case of association preceding scenes. The best known engravings on a piece of ochre at Blombos, c.75,000 years (d’Errico 2003), may be read as a very early example of patterning. Remove some lines and the pattern changes or disappears—in this way determining the engraving as associational by my definition. All this might provide support for the developmental sequence put forward above, regardless of there being no historical/cultural connection between these temporally and geographically scattered examples. Still, I would not go so far as to suggest that such examples do more than offer a signpost. As regards Europe, I have argued for scenes in the Palaeolithic (Dobrez 2012), though not with reference to images usually identified as such by European researchers. For the classic Lascaux case see P. Dobrez in Ch.3. In general, I would simply observe that, by my definition, such images mostly qualify as suggestive associations of figures rather than scenes, since they do not include depiction of either motion or causality. Whether there are scenes or not at Chauvet (see Chauvet et al. 1995; Clottes et al. 2001) is debateable. It seems to me here that we need to be on guard so as not to conflate notions of “animation” (i.e., images as lively), or of “realism” (i.e., images as “lifelike”), with the notion of a scene. A picture may be animated and lifelike without amounting to a scene—for which depiction of motion and, with more than one figure, causality, are required. If we allow that there are scenes at Chauvet (which I would be prepared to do), then Europe fits poorly with my scheme. If we do not, however, this site falls in line with my hypothetical sequence, with juxtapositional logic (i.e., figures apparently randomly arranged) coupled with associational iconics (i.e., canonically recognizable figures) from Lascaux and Altamira to Niaux. After which the Spanish Levant scenes would emerge on cue, that is, late in the sequence (Villaverde et al. 2012). But Chauvet upsets this neat scheme and should be taken as evidence that we need to beware of easy packaging. Moreover there is a case for some scenic elements at Lascaux (e.g., the “Chinese” horses; the problematic “leaping” cow), as well as at Altamira (the bison on the ceiling bosses as either running or—a minimal motion option—resting). Parts of the Americas certainly contain scenes, including a great many in the Serra da Capivara, Brazil, relatively few in Canada and the USA. Gradin and, with modifications, others (see Dobrez and Dobrez 2014) have argued for the famous scene of a guanaco hunt as predating static canonical figures at Cueva de las Manos, Argentina. This of course reverses my sequence. In the US, the best candidates for scenes are Texas pictures in the Red Linear style, which Boyd (2003) has placed in the relatively recent “Late Archaic,” and the “Biographic” tradition investigated by Keyser (Keyser and Klassen 2001), which are very recent indeed. But it seems prudent to me to say that the sequence is too fragmented for the kinds of conclusions I require. A claim has been made for hand stencil and certain marks at three Spanish sites, marks that arguably qualify as associations, dated at 65,000 years (Hoffmann et al. 2018). But the date is disputed (Aubert,

46

LIVIO DOBREZ

Brumm, and Huntley 2018). In Sulawesi a hand stencil and the canonical depiction of animals (Taçon et al. 2014) are dated at close to 40,000 years, which puts them in my category of salient association of lines preceding scenic representation. However, still older images convincingly qualifying as scenic have recently been found (Aubert et al. 2019). Moreover, it would be rash to suppose that still older images, scenic or otherwise, will not turn up in due course—though there must be a point in time beyond which materials cannot last. All I need reiterate in connection with the question posed above (“might depictive associations generally predate scenic representation?”) is that overall, the picture in any part of the world is far from clear, and that this comes as no surprise. It would be astonishing if any rock art location, whether at the micro level of a single site or the macro level of a geographic area, offered a tailor-made illustration of the threefold shift I postulate, any more than it would provide a ready fit for the various theoretical models discussed above, from Davis to Halverson. The art record available to us is simply too patchy for this sort of synthesizing overview. Nonetheless, this does not constitute grounds for abandonment of the idea of depictive organization as a cognitive marker.

The Concept of “Trace” There is a further way of bolstering the developmental model, not by a route leading directly to the scenic, but by suggesting steps in the sequence that I have not so far considered. In addition to the three perceptual/ depictive categories so far mentioned, Patricia Dobrez (2013, 2014, 2015, 2017, 2018) has made the case for a fourth, that of “trace.” Since her focus in the following chapter is not on cognitive development, I may be permitted to mention arguments she has made elsewhere in my own context. Traces, such as those of hands and feet, uniquely mediate between nature and art. They are ubiquitous in rock art, and their genealogy goes all the way back to natural situations. The first handprint, say, on mud or sand, is arguably as old as any activity involving the hand and constitutes an iconic depiction from the start, though at that stage it is a depiction doubtless made by chance, in line with the Davis formula. Since the first print, hand images have continued to be made, as prints and stencils, up to the present. While depicted tracks, for example in the form of petroglyphs, are presumably a relatively recent phenomenon, tracks—also intrinsically recognizable as iconic—have been read and/or followed from the beginning and may, like handprints, be reproduced, by accident or by design. If Patricia Dobrez is correct in suggesting that depiction may have begun with ground drawing and in tandem with storytelling, after the manner still current in Australian desert regions, then the two key elements of the scene, namely iconic depiction and narrative, may be sourced to early operations focused on hand and track imagery, possibly in connection with associated marks of the phosphene type, such as lines and circles. The pedal narrative would then be the precursor of the scene, which of course requires precursors, with the difference that its perspective is oriented to plan view rather than to profile. Patricia Dobrez has given substance to this argument by offering a biological explanation, drawing both on Freyd’s representational momentum experiments and the phenomenon of “proprioception,” for the lure of trails, i.e., our tendency to follow a trail as it were automatically. Using authorities on gestural communication and the mirror-neuron work of Rizzolatti, Craighero and others, she has sought to explain the mechanisms in play in the making and subsequent perception of hand images. Not least of these is the process of selfothering entailed in print or stencil making: the hand is recognized as mine since I make it, but it also functions as separate from me, a sign of me in my absence. It may be that hand and track images, inevitably operating pars pro toto, have had a role in the development of self-consciousness. Very likely they are the first exograms. If these surmises are to any extent correct, the unique category of trace has played a critical role in the making of cognition we readily recognize as modern.

Conclusion This chapter poses the question: might there be some connection between depicted scenes and cognitive evolution? My approach has been to stand back as tactfully as possible, not merely regarding answers to the question but additionally regarding the vexed issue of what we may mean by “cognitive” and “cognitive evo-

THE POSSIBLE SIGNIFICANCE OF DEPICTED SCENES FOR COGNITIVE DEVELOPMENT

47

lution.” Accordingly, I have generally used the “softer” term development and pushed the sense of cognitive development no further than to mean a change in the direction of what we recognize as modern, that is, ourselves. Such a change, if epigenetic, might switch certain genes on or off and modify particular neural connections according to the dictum that “neurons that fire together wire together” (Edelman 2004: 29), though it would not alter hardwiring, for example the basic elements of our visual system, which is my main concern. Whether epigenetic or not, it might generate rewiring of a brain we know to be highly plastic, according to Edelman’s principle. In line with my open-ended approach to the term cognitive, I offer no specific definition of “mind.” My focus in this chapter is elsewhere. At the same time my leaning toward Gibson’s work (with, for those familiar with philosophy, its parallels with Heidegger’s “being-in-the-world”) might alert the reader to an “ecological” bias, that is, an understanding of humans as structured in the Gibsonian sense by engagement with their environment. Likewise, my stress on integration in pictures may also alert some to a holistic bias elaborated elsewhere (Dobrez and Dobrez 2013a). Finally, I require no strict definition of “art,” since my scheme tends to inclusion rather than exclusion. In short, for present purposes I can afford to put more rather than fewer candidates in the “art” category. In general, when I seek scientific support for phenomenology, I turn to neurophysiological knowledge rather than psychological models of mind, though I would not dispute the heuristic value of such models. Though sidestepping certain theoretical issues, I am, however, most concerned with precision in the definition of the subject of immediate concern: the scene. I open with reference to the notion of real and depicted events, observed not by inferential processes but directly, partly following Gibson but mostly for reasons related to evolutionary imperatives. After this general definition, which includes two key scenic elements, namely motion and causality, I turn to methodology, explaining the implications of a universalist, and for me ultimately biological, perspective derived from Husserlian analysis of the phenomenon in question, in this case pictures and the way we perceive them. More detailed analysis of scenic markers follows, introducing, perhaps most critically, consideration of scenic space and of the “window” principle. After which I tackle the perceptual stages of looking at a picture, these being composition (and the related matter of framing), then, depending on the depiction under observation, juxtaposition, association and, finally, scene. Having put the scenic center stage, I return to the original question of its possible link to cognitive development. First, I consider some notable scholars who have contributed to the discourse, focusing on the fact that despite deep disagreements, they may all be read as accepting (1) some sort of more or less automatic genesis for art, and (2) some idea of a progression toward greater organization in art, itself connected with cognitive development. Of course, I do not say that all the researchers I mention would put the matter in precisely this way. At any rate, combining consideration of previous theorizing with my own argument for the perception of pictures, I pose the original question, immediately pointing to some of the practical difficulties of applying the thesis to our inescapably partial knowledge of real-world rock art sites. Thus any answer to the question comes with caution. I end the chapter with a reference to Patricia Dobrez’s investigation of the phenomenon of “trace” in life and art—not because it feeds directly into the discussion of scenes (for which see her chapter in this volume), but because it identifies possible further mediating steps in a depictive/cognitive sequence. In short, it adds more, substantially original, detail to the story of how we receive pictures and, by extension, how we might have both begun and carried on the activity of making pictures. Livio Dobrez, originally a specialist in literature, has combined literary studies with art history and philosophy. He has taught courses on hermeneutics and Aboriginal writing/oral traditions and the research result has been a focus on rock art with emphasis on theory. He has published several books, as well as edited collections and many scholarly articles. Now retired, he was Head of English at the Australian National University on a number of occasions, as well as holding the Chair of Australian Studies (Humanities and Social Sciences) at Bond University for a time.

References Aubert, M., Brumm, A. and J. Huntley. 2018. “Early Dates for ‘Neanderthal Cave Art’ May Be Wrong.” Journal of Human Evolution 125: 215–217.

48

LIVIO DOBREZ

Aubert, M., R. Lebe, A. A. Oktaviana, M. Tang, B. Burhan, A. Jusdi, Hakim, Budianto, Zhao Jian-xin, I. M. Geria, and P. H. Sulistyarto. 2019. “Earliest Hunting Scene in Prehistoric Art.” Nature 576: 442–45. Bednarik, R. G. 1984. “On the Nature of Psychograms.” The Artefact 8: 27–32. ———. 1995. “Concept-Mediated Marking in the Lower Palaeolithic.” Current Anthropology 36(4): 605–16. ———. 2003. “The Earliest Evidence of Palaeoart.” Rock Art Research 20(2): 89–104. ———. 2008. “Cupules.” Rock Art Research 25(1): 61–100. Boyd, C. 2003. Rock Art of the Lower Pecos. College Station: Texas A & M University Press. Brooks, R., and V. S. Wakankar. 1976. Stone Age Painting in India. New Haven: Yale University Press. Chaloupka, G. 1993. Journey in Time. Chatswood, New South Wales: Heinemann. Chauvet, J.-M., E. B. Deschamps, and C. Hillaire. 1995. La Grotte Chauvet à Vallon-Pont-d’Arc. Paris: Seuil. Clottes, J. 2001. Return to Chauvet Cave: Excavating the Birthplace of Art; the First Full Report. Paris: Seuil. Davidson, I. 2010. “The Archaeology of Cognitive Evolution.” Cognitive Science 1(2): 214–29. ———. 2013. “Origins of Pictures: An Argument for Transformation of Signs.” In Origins of Pictures: Anthropological Discourses in Image Science, ed. K. Sachs-Hombach and J. Schirra, 16–46. Cologne: Halem. Davidson, I. and W. Noble. 1989. “The Archaeology of Perception: Traces of Depiction and Language.” Current Anthropology 30(2): 125–55. Davis, W. 1986. “The Origins of Image Making.” Current Anthropology 27(3): 193–202. Dere˛gowski, J. B. 1980. Illusions, Patterns and Pictures: A Cross-cultural Perspective. London: Academic Press. ———. 1984. Distortion in Art: The Eye and the Mind. London: Routledge and Kegan Paul. D’Errico, F., C. Henshilwood, G. Lawson, M. Vanhaeren, A.-M. Tillier, M. Soressi, F. Bresson, B. Maureille, A. Nowell, J. Lakarra, L. Backwell, and M. Julien. 2003. “Archaeological Evidence for the Emergence of Language, Symbolism, and Music: An Alternative Multidisciplinary Perspective.” Journal of World Prehistory 17(1): 1–70. Dobrez, L. 1986. The Existential and its Exits: Literary and Philosophical Perspectives on the Works of Beckett, Ionesco, Genet and Pinter. London: Athlone. ———. 2007a. “Against Deceit: Comment on D. Hodgson and P. A. Helvenston, ‘The Emergence of the Representation of Animals in Palaeoart.’” Rock Art Research 24(1): 113–116. ———. 2007b. “The Reception of Visual Representations.” In XXII International Valcamonica Symposium, 115–121. Capo di Ponte: Edizioni del Centro. ———. 2008. “A Rock Art Typology: Narrative and Non-narrative Figural Representation.” In Aesthetics and Rock Art III Symposium, ed. T. Heyd and J. Clegg, 29–33. Hampshire, UK: Ashgate. ———. 2010a. “‘Good to Think’ Representational Modalities.” In Global Rock Art FumdhaMentos IX, 284–94. São Raimundo Nonato, Brazil: Fundação Museu do Homem Americano. ———. 2010b. “Framing and the Organization of Space in Representation.” In Global Rock Art FumdhaMentos IX, 651–65. São Raimundo Nonato, Brazil: Fundação Museu do Homem Americano ———. 2010/11. “What Representations Tell Us about the Way We See (with Commentary by R. G. Bednarik and Reply).” Purakala: The Journal of Rock Art Society of India 20/21: 23–41. ———. 2011a. “Sex for its Own Sake: Comment on Varella et al., ‘Evolutionary Aesthetics and Sexual Selection in the Evolution of Rock Art Aesthetics.’” Rock Art Research 28(2): 166–68. ———. 2011b. “Rock Art, Perception and the Subject/Object Binary.” Rock Art Research 28(1): 71–83. ———. 2011c. “Looking at our Looking: A New Approach to the Definition of a Rock Art Scene.” In American Indian Rock Art 37, ed. M. Greer, J. Greer, and P. Whitehead, 251–64. Glendale, AZ: American Rock Art Association. ———. 2012. “Towards a More Rigorous Definition of Terms: Are There Scenes in European Palaeolithic Art?” In L’Art Pléistocène dans le monde, Actes du Congrès IFRAO, Tarascon-sur-Ariège, septembre 2010. Numéro spécial de Préhistoire, Art et Sociétés. Bulletin de la Société Préhistorique Ariège-Pyrénées, 65–66, ed. J. Clottes, 1837–51. ———. 2013a. “Births and Deaths: Comment on Moro Abadía, ‘Rock Art Stories: Standard Narratives and their Alternatives.’” Rock Art Research 30(2): 154–58. ———. 2013b. “The Perception of Depicted Motion.” Arts 2: 383–446. ———. 2015a. “Roman Ingarden and the Reader.” Ch.4 of Livio Dobrez, A Readable Introduction to Theories of Reading and Interpretation. DOI: 10.13140/RG.2.1.3952.0161 ———. 2015b. “Depicted Motion, Interaction and Causality in Rock Art.” In American Indian Rock Art 41, ed. J. D. Keyser and D. Keyser, 57–68. San Jose, CA: American Rock Art Research Association. ———. 2015c. “Evolutionary Perceptual Constants in Rock Art Motifs.” In XXVI Valcamonica Symposium, 101–106. Capo di Ponte, Italy: Edizioni del Centro. ———. 2015d. “Making Sense of Pictures.” Bollettino del Centro Camuno di Studi Preistorici 40: 31–47. ———. 2016. “Theoretical Approaches to Rock Art Studies (with Commentaries and Reply).” Rock Art Research 33(2): 143–66. Dobrez, L., and P. Dobrez. 2013a. “Rock Art Animals in Profile: Visual Recognition and the Principles of Canonical Form.” Rock Art Research 30(1): 75–90. ———. 2013b. “Canonical Form and the Identification of Rock Art Figures.” In American Indian Rock Art 39, ed. W. D. Hyder, 115–29. Glendale, AZ: American Rock Art Research Association.

THE POSSIBLE SIGNIFICANCE OF DEPICTED SCENES FOR COGNITIVE DEVELOPMENT

49

———. 2014. “Canonical Figures and the Recognition of Animals in Art and Life.” Boletín del Museo Chileno de Arte Precolombino 19(1): 9–22. Dobrez, P. 2013. “The Case for Hand Stencils and Prints as Proprio-performative.” Arts 2: 273–327. ———. 2014. “Hand Traces: Technical Aspects of Positive and Negative Hand-Marking in Rock Art.” Arts 3: 367–93. ———. 2015. “Their Proper Sphere: Situating Human Feet in Rock Art.” In Proceedings of the XIX International Rock Art Conference IFRAO 2015 (Cáceres, Spain), ed. H. Collado Giraldo and J. J. García Arranz. ARKEOS 37 (CD): 259–77. ———. 2017. “From Tracks to Gesture-Derived Inscription: An Australian Genealogy for ‘Tracks and Lines’ Petroglyphs.” Rock Art Research 34(2): 149–68. ———. 2018. “The Body as Intelligent Tool: The Making and Reading of Trace and Trace-Derived Forms in Rock Art.” Time and Mind: The Journal of Archaeology, Consciousness and Culture 11(1): 1–25. Doidge, N. 2007. The Brain that Changes Itself: Stories of Personal Triumph from the Frontiers of Brain Science. Melbourne: Scribe. Donald, M. 1991. Origins of the Modern Mind: Three Stages in the Evolution of Culture and Cognition. Cambridge, MA: Harvard University Press. Edelman, G. M. 2004. Wider Than the Sky: The Phenomenal Gift of Consciousness. New Haven, CT: Yale University Press. Freyd, J .J. 1983. “The Mental Representation of Movement When Static Stimuli Are Viewed.” Perception and Psychophysics 33: 575–81. ———. 1987. “Dynamic Mental Representations.” Psychological Review 94: 427–38. Freyd, J. J. and R. A. Finke. 1984. “Representational Momentum.” Journal of Experimental Psychology 10: 126–32. Gibson, J. J. 1950. The Perception of the Visual World. Boston: Houghton Mifflin. ———. 1966. The Senses Considered as Perceptual Systems. Boston: Houghton Mifflin. ———.1979. The Ecological Approach to Visual Perception. Boston: Houghton Mifflin. Gombrich, E. H. 1960. Art and Illusion: A Study in the Psychology of Pictorial Representation. London: Phaidon. Hagen, M. A. 1986. Varieties of Realism: Geometries of Representational Art. Cambridge, UK: Cambridge University Press. Halverson, J. 1987. “Art for Art’s Sake in the Palaeolithic.” Current Anthropology 28(1): 63–71. Heidegger, M. (1927) 1962. Being and Time. Trans. J. Macquarrie and E. Robinson. Oxford: Blackwell. Hodgson, D. 2000. “Art, Perception and Information Processing: An Evolutionary Perspective (with Commentaries and Reply).” Rock Art Research 17(1): 3–34. Hoffmann, D. L., C. D. Standish, M. García-Diez, P. B. Pettitt, J. A. Milton, J. Zilhão, J. J. Alcolea-González, P. Cantalejo-Duarte, H. Collado, R. de Balbín, M. Lorblanchet, J. Ramos-Muñoz, G.-C. Weniger, A. W. G. Pike, 2018. “U-Th Dating of Carbonate Crusts Reveals Neandertal Origin of Iberian Cave Art.” Science 359(6378), 912–915. Hublin, J.-J., A. Ben-Ncer, S. Bailey, S. Freidline, S. Neubauer, M. Skinner, I. Bergmann, A. Le Cabec, S. Benazzi, K. Harvati, and P. Guns. 2017. “New Fossils from Jebel Irhoud, Morocco and the Pan-African Origin of Homo sapiens.” Nature 546: 289–92. Husserl, E. (1900/1901) 2001. Logical Investigations], trans J. N. Findlay. London: Routledge. Johansson, G. 1973. “Visual Perception of Biological Motion and a Model for its Analysis.” Perception and Psychophysics 14: 201–11. Keyser, J., and M. Klassen. 2001. Plains Indian Rock Art. Seattle: University of Washington Press. Krekelberg, B., A. Vatakis, and Z. Kourtzi. 2005. “Implied Motion from Form in the Human Visual Cortex.” Journal of Neurophysiology 94: 4373–86. Kumar, G., R. G. Bednarik, A. Watchman, and R. G. Roberts. 2005. “The EIP Project in 2005: a Progress Report.” Purakala 14/15: 13–68. Layton, R. 1977. “Naturalism and Cultural Relativity in Art.” In Form in Indigenous Art: Schematisation in the Art of Aboriginal Australia and Prehistoric Europe, ed. P. J. Ucko, 33–43. London: Duckworth. Lenssen-Erz, T. 1992. “Coherence: a Constituent of ‘Scenes’ in Rock Art: The Transformation of Linguistic Analytical Models for the Study of Rock Paintings in Namibia (with Commentaries and Reply).” Rock Art Research 9(2): 87–105. Lewis-Williams, D. 1983. The Rock Art of Southern Africa. Cambridge: Cambridge University Press. ———. 2002. The Mind in the Cave: Consciousness and the Origins of Art. London: Thames and Hudson. Lewis-Williams, D., and T. A. Dowson. 1988. “The Signs of All Times: Entoptic Phenomena in Upper Palaeolithic Art.” Current Anthropology 29(2): 201–17. Lorblanchet, M. 1999. La Naissance de l’Art: Genèse de l’Art Préhistorique dans le Monde. Paris: Errance. May, S. K., I.G. Johnston, P. S. C. Taçon, I. Domingo, and J. Goldhahn, 2018. “Early Australian Anthropomorphs: Jabiluka’s Dynamic Figure Rock Paintings.” Cambridge Archaeological Journal 28(1): 67–83. Maynard, L. 1979. “The Archaeology of Australian Aboriginal Art.” In Exploring the Visual Art of Oceania: Australia, Melanesia, Micronesia, and Polynesia, ed. S. M. Mead, 83–110. Honolulu: The University Press of Hawaii. Merleau-Ponty, M. 1962. Phenomenology of Perception. Trans C. Smith. London: Routledge & Kegan Paul. Michotte, A. 1963. The Perception of Causality. Trans T. R. and E. Miles. London: Methuen. Misra, V. N., Y. Mathpal, and M. Nagar. 1977. Bhimbetka: Prehistoric Man and His Art in Central India. Pune: Poona University and Deccan College. Mithen, S. 1996. The Prehistory of the Mind: A Search for the Origins of Art, Religion and Science. London: Phoenix.

50

LIVIO DOBREZ

Pfeiffer, J. E. 1982. The Creative Explosion: An Inquiry into the Origins of Art and Religion. New York: Harper and Row. Ramachandran, V. S., and W. Hirstein. 1999. “The Science of Art: A Neurological Theory of Aesthetic Experience.” Journal of Consciousness Studies 6: 15–51. Ross, J., K. E. Westaway, M. Travers, M. J. Morwood, and J. Hayward. 2016. “Into the Past: A Step Towards a Robust Kimberley Rock Art Chronology.” PLoS ONE 11(8): e0161726. Sartre, J.-P. (1943) 1969. Being and Nothingness: An Essay on Phenomenological Ontology, trans. H. E. Barnes. London: Methuen. Stringer, C. 2011. The Origin of Our Species. London: Allen Lane. Taçon, P., N. H. Tan, S. O’Connor, X. Ji, G. Li, D. Curnoe, D. Bulbeck, B. Hakim, I. Sumantri, H. Than, I. Sokrithy, S. Chia, K. Khun-Neay, and S. Kong. 2014. “The Global Implications of the Early Surviving Art of Greater Southeast Asia.” Antiquity 88: 1050–64. Thomas, J. 1996. Time, Culture and Identity: An Interpretive Archaeology. London: Routledge. Tilley, C. 2004. The Materiality of Stone: Explorations in Landscape Phenomenology. New York: Berg. Ungerleider, L .G., and M. Mishkin. 1982. “Two Cortical Visual Systems.” In Analysis of Visual Behavior, ed. D. J. Ingle, M. A. Goodale, and R. J. W. Mansfield, 549–86. Cambridge, MA: The Massachusetts Institute of Technology Press. Villaverde, V., R. Martinez, P. Guillem, E. López, and I. Domingo. 2012. “What Do We Mean By Levantine Rock Art?” In The Levantine Question: Post-Palaeolithic Rock Art in the Iberian Peninsula, ed. J. L. García, H. Collado, and G. Nash, 81–115. Budapest: Archaeolingua. Walsh, G. 2000. Bradshaw Art of the Kimberley. Toowong, Queensland: Takarakka Nowan Kas. Watson, B. 2008. “Oodles of Doodles? Doodling Behaviour and its Implications for Understanding Palaeoarts (with Commentaries and Reply).” Rock Art Research 25(1): 35–60. Willcox, A. R. 1984. The Rock Art of Africa. London: Croom Helm.

==== •

•

•

3. EVENT DEPICTION IN ROCK ART Landscape-Embedded Plan View Narratives, Decontextualized Profile “Scenes,” and Their Hybrid Instances Patricia Dobrez

Introduction When in 1899 Spencer and Gillen (1968: 617) remarked on the seeming absence of “scenes” (their term) in Australian central desert rock art, they had in mind Arnhem Land representations “such as a kangaroo chase, or men spearing emus.” Had their focus not been merely on a dominant category in their own cultural tradition, they might have noticed a widespread practice of making marks that correspond to traces of activities left in the landscape. Basedow (1925) provides a central Australian illustration of a depicted emu hunt, which, in his Aranda example, also includes the hunter’s companion dog (Figure 3.1). The associated motifs involve a zigzag line representing the alternating foot motion of the hunter, along with dog and emu prints. Instead of the familiar profile depiction, we have in this case a plan view eminently suited to desert peoples who have ready-to-hand ground surfaces available for inscription. We might want to call such a composition a scene, since its content (an emu hunt) is identical to that of a European depiction of the kind Spencer and Gillen found worthy of the name. Content aside, however, the method of depiction is quite different from that of Spencer and Gillen’s examples. So when we presently consider “scenicity” as understood in recent European discourses of art and literature, it becomes apparent that a distinguishing terminology is desirable. Because there are important distinctions to be made, my preferred term in the Basedow central desert instance is “pedal narrative.” Where a depiction of emu and dog tracks in association with a conventionalized motion pathway in the form of a zigzag requires some degree of specialized knowledge for interpretation, there are, however, icons that are transparently readable. Because of our comprehension of our own bodies (P. Dobrez 2018), we have immediate access to some visual forms as iconic universals. These we understand from the inside, as it were. As Gallese (2017: 48) asserts, vision possesses a “haptic quality,” necessitating the abandonment of “purely visibilist,” that is, visually oriented, approaches to the way humans create and respond to images. Some universals exist in life in the form of trace images and in rock art as hand stencils and prints—of hands, in particular, but also of human feet. In connection with what we might call the “bodily self replication” thesis, we need to consider that the trace shapes produced by the hand and foot might have been the original catalysts enabling understanding of symbolic substitution. As transparently iconic, human foot and hand trace images give us access to iconicity itself. In so doing they present an affordance for use in person-to-person communication and, ultimately, for the construction of foundational symbolic systems (P. Dobrez 2013, 2015, 2017, 2018). When it comes to human feet, the images we largely find in the rock art record are rep-

FIGURE 3.1. Illustration of trace-view emu hunt depiction, central Australia. Basedow (1925, Figure 44).

52

PATRICIA DOBREZ

lications of tracks in the form of pecked images. Typically, human footprints present with animal tracks and are frequently sequenced in such a manner as to tell a story as in the emu hunt example. What, then, do we intend when we use the word scene, and why is “pedal narrative” a better usage in connection with perceived tracks sequencing? It is important to recognize at the outset that notions of scenic representation have come down to us from European discourses, in particular those relating to art and literature, discourses that conceptualize the human subject’s relationship with a landscape or a built environment in culturally specific ways. Consider some notable examples. Stimulated by theorizing about the nature of the real (Comte 1853; Feuerbach 1857), the English novelist George Eliot (1819–80) set out to imitate seventeenth-century Dutch painting achievements by depicting commonplace scenes (Scenes from Clerical Life), attending to pictorial detail (Adam Bede), and precisely situating her protagonists in place and time (The Mill on the Floss). As she urged in her review of Ruskin’s Modern Painters III, integrity in art was to be pursued through “a humble and faithful copy of nature” (Eliot 1856: 626). The effects she aspired toward have long since been regarded as hallmarks of the wider realist or naturalist movement in representation. Eliot regarded the realist aesthetic (for a discussion see P. Dobrez 1977) as expressing the highest aim of art and literature. Her companion G. H. Lewes (1858), the influential promoter of Comtean positivism in Britain, agreed. Anyone among their contemporaries who read a George Eliot novel, or one by any number of European writers (Balzac, Flaubert, Tolstoy, Zola), or who looked at a painting by Constable, Courbet, Millet, or Bastien-Lepage, or who read books on natural history, would have had been affected in some way by the idea of human beings as situated within a web of concrete relations. Above all, “man”—“a real being, the true Ens realissimum—man” (Feuerbach 1857: 6)—was at the center of things. Such attitudes eventually led to false expectations on the part of rock art researchers, for example the largely thwarted hopes of finding representations of humans and settings for figures in European Paleolithic art. More importantly, researchers were encouraged from the start to value portrayal of naturalistic detail. Finally, and relevantly for the present argument, realist assumptions prevented and continue to prevent recognition of the scene-like capacity of unfamiliar forms—in Spencer and Gillen’s case, sequenced tracks. Thus it remains unusual to use the term scene for tracks representation. “Pedal narrative” has the virtue of accurate description without contamination from realist presuppositions. In what follows I shall examine three categories of visual storytelling, that is, categories with equivalent functions, not all of which attract the label scene.

Rock Art’s Second Coming: Its Modern Moment of Rediscovery Considering the prominence that landscape painting had attained in Europe at the moment of (adapting a phrase) “the shock of the old,” initiated by the publication of images of allegedly prehistoric paintings (Sanz de Sautuola 1880), it is not surprising that it would take some time for sensibilities to adjust and for a new terminology to develop. If we take as a representative figure the influential art critic John Ruskin (1819–1900), barriers to understanding set up by an age that called itself “modern” come into sharp focus. Had Ruskin, surveying the history of European landscape painting in his monumental work, Modern Painters (1843–1860), been able to investigate the forms of prehistoric art, instead of turning to the first record that came to mind (in the absence of extant pictures of ancient landscapes, their evocations in the poetry of Homer), the shaper of contemporary taste would have had to take into account very different purposes and practices in pictorial representation. For an eager reading public increasingly interested in origins and “progress,” Ruskin’s third volume of Modern Painters (1856) demonstrated how different artists and historical periods select different aspects of the world to depict. Beyond Homer, Book III of Modern Painters contrasts the “then” of medieval art with the “now” of nineteenth-century landscape painting. The example of Ruskin’s wide survey of the Western art tradition drives home the problems Sanz de Sautuola faced in trying to overturn notions of the antique in art when he published the first account of Paleolithic rock art at Altamira. Throughout the five volumes of Modern Painters, where the baseline for comparison is classical landscape, the template for painting is invariably figures in a setting of some kind, an approach lending itself to discussion in terms of nineteenth-century notions of the scenic. What varies for Ruskin is the relative weight given to either figure or setting and the world view it might presuppose. Thus we are told that when reading the Odyssey “we shall always be struck by . . . the excessive similarity in

EVENT DEPICTION IN ROCK ART

53

the scenes,” owing to the “subservience of the whole landscape to human comfort” (Ruskin 1856 [1863]: 185). Medieval landscapes are presented as a shift from the agricultural bucolic to an aristocratic pursuit of dominance and pleasure. Nonetheless, priorities remain unchanged: landscape depictions of illuminated manuscripts epitomize the satisfaction of subordinating “flowers, castles, brooks, clouds, rocks . . . to the human figures in the foreground, and painted for no other end than that of explaining their adventures and occupations” (Ruskin 1856 [1863]: 208). As a Renaissance example, Poussin’s use of setting is remarked upon for its typical function as backdrop, with “the best pieces of it occur[ring] in fragments behind his figures” (Ruskin 1860 [1885]: 270). When it comes to his own nineteenth century, Ruskin points up principles founded on a “new affection for nature.” Landscape now dominates as a genre, but one undergoing a revolution as it foregrounds a fascination with “mutability.” Thus the “appearance of objects” is becoming “a subject of science with us,” and if “a characteristic name were needed for modern landscape art, none better could be invented than ‘the service of clouds’” (Ruskin 1856 [1863]: 254–55). The apotheosis of the nineteenth-century modern is of course J. M. W. Turner (1775–1851), in whose work Ruskin saw an elevation of the topographical to an art of the “impression.” The prioritizing of landscape that Ruskin admired in Turner is in tune with one of the great themes of the century of Darwin. Turner depicts humans and their works as subsidiary to natural forces. In this way, what was formerly a backdrop element in the representation of a scene becomes its chief subject matter. In view of his habitual attitudes, Ruskin, the arbiter of Victorian taste and a painter in his own right, as well as the foremost authority on the landscape tradition, would have been extremely puzzled by the 1880 publication of the black and white illustration of the Altamira polychrome ceiling, had he had occasion to take note of it. The prehistoric is still remote for his generation, as illustrated by Cormon’s fanciful 1882 painting (reassuring domestic setting, rude shelter and garb), “Returning from a Bear Hunt During the Stone Age” (Musée des Beaux-Arts, Carcasonne, France). In fact, Cormon’s picture seems to have been a sketch for another picture bought by the Museum of National Antiquities at Saint Germain-en-Laye, France’s leading institution for prehistoric material. This and other work made Cormon preeminent in the field of painting the ancient past. It also explains his commission to decorate the Paris Museum of Natural History (Nochlin 1971: 26, 250, Figure 9). In this context the unfamiliar spectacle of rock art would have thrown Ruskin completely off balance, as he, like his skeptical contemporaries, would have struggled to discover in the Altamira bison any representation at all of contextualizing landscape. The ceiling panel displays a measure of compositional coherence (Musquiz Pérez-Seoane 1998: 66). It is arresting in its painted color, engraved details of eyes, horns and neck hair, and the bulk of figures incorporating natural rock features. There is, however, little to suggest an association between the multiplied discrete images showing different animal postures—nothing here to recall the situated animal figures of a Constable or Cuyp, let alone an exotic image of prehistory after the manner of Cormon. With a paucity of prehistoric finds for comparison, the Altamira images appeared as probable fakes until their Paleolithic origin began to be acknowledged following Émile Cartailhac’s retraction of his opposition in “Mea Culpa d’un Sceptique” (1902). In short, Altamira challenged the entire European tradition of painting, understood by Ruskin’s contemporaries as figures in a setting or landscape—although in some rock art cases a visual echo of landscape was subsequently posited by scholars. For example, Leroi-Gourhan (1982: 28), who doubts that Paleolithic artists were at all interested in symbolic representation of “real space” (“The total absence of figures which could correspond to vegetable or mineral images means we cannot speak of ‘landscape’”), is, however, willing to entertain possible “ground surface” representation to suggest a connection between individual figures: “The imaginary ground surface line appears when at least two animals are following or confronting one another in such a way that their extremities (drawn or not) can be found on the same imaginary line” (26–28). In his discussion of the post-Paleolithic Spanish Levant, Beltrán (1982: 23–24) asserts that “landscape as such is never depicted,” but rather suggested by incorporation of natural features like cracks and crevices: “we find that the positioning of men, animals and objects has been adapted to the projections, hollows or contours of the rock to express surroundings, or a virtual landscape.” Comparable claims have been made for other cases, but these are universally regarded as the exception rather than the rule. As for a Turneresque foregrounding of atmospheric effects in scenic landscape depiction, rock art is more likely to personify natural forces than to directly illustrate them. In southern African San and Australian Aboriginal rock art, the atmospheric may take the form of a “rain-animal” (Lewis-Williams 1983: 20–21),

54

PATRICIA DOBREZ

Wardaman Lightning Brothers and Kimberley Wandjina figures (Flood 1997: 152–53, 294–98), or the Arnhem Land Lightning Man (Chaloupka 1993: 56–58). Such representations may possess suggestive features, such as the cloud-like appearance of Wandjinas (Crawford 1968: 28; Flood 1997: 296), but they are personifications rather than illustrations. Personification of weather events is of course evident in many traditions—for example in such figures as Indra brandishing his thunderbolt weapon, “the Atmosphere under his jurisdiction” (Hackin et al. 1964: 108), or Thor with his hammer, “ruler over the forces of weather” (Andersson 2016: 87; Sorensen 1999: 203). In summary so far, the nineteenth century expected events to be pictured as figures in a landscape, with greater or lesser emphasis on landscape, including atmospheric phenomena. We are left with the question: to what extent does the idea of some form of setting for figures need to be part of any definition of what we might regard as scenic representation, particularly in rock art? Should we be looking for settings at all? Despite its currency in the discourse, is scene perhaps not always the word we need?

Varying Definitions of a “Scene” Before answering the questions, I would like to complement the foregoing discussion of assumptions about scenic representation with some comments about varied usage of the term scene. The question then becomes: how might any one of the many definitions of a scene suggest itself as useful to rock art researchers? In general usage, snapshots of which non-technical dictionaries aim to record, understanding of the word seems to be roughly in line with Ruskin’s sense of “figures in a landscape.” It may stand for single or plural figures and/or objects and their settings, with some English dictionaries putting the emphasis on the place in which an event occurs. In this instance, attention is shifted from what is happening to its location, but the idea of an occurrence, albeit situated in some way, remains. I shall focus on the notion of an event in due course. It is generally understood that the content of a scene may be something witnessed, pictured, or imagined. In some dictionaries, the meaning of the word will be given as the place itself, regardless of the presence or absence of figures that might be read as engaged in an action. Secondary meanings frequently relate to stage performances, including their settings (as in “stage scenery”), followed by meanings relating to displays of emotion (“she made a scene”). Some dictionaries even give the word landscape as a synonym for scene. A miscellany of dictionaries from the period in which rock art commentary was becoming established as a scholarly discipline, which I have used as the basis of my general overview here—The Macquarie Dictionary (1987); Chambers 20th Century Dictionary (1983); The Australian Pocket Oxford Dictionary (1976); Webster’s New Twentieth Century Dictionary of the English Language (1959)—illustrates the diversity of lexicographical opinion. From the point of view of specific disciplines, the Glossary of a popular art history textbook that has run to many editions does not include the word scene; the only reference is to scenery under the heading of landscape, described as “without narrative content” (Kleiner 2016: 1153). In the Western landscape art tradition, “setting” is entirely possible on its own. Cognitive psychology goes further. Here the prevailing view of a scene is encapsulated in the following: “Scene Perception is the visual perception of an environment as viewed by an observer at any given time. It includes not only the perception of individual objects, but also things in their relative locations, and expectations about what other kinds of objects might be encountered” (Rensink 2000: 151). Thus, the typical “scene” used in perceptual experiments will be a set of objects, say, trash cans against a backdrop in the case of work on Boundary Extension (Intraub and Richardson 1989). Researchers investigate “scene gist,” examples being “seashore,” “farmyard,” or “shopping centre” (Rensink 2000: 154). On the other hand, the Rock Art Glossary (2010) defines the term thus: “a presumed depiction of a real or imagined episode involving more than one rock art motif.” In view of recent theorizing on the subject of rock art scenes, it would seem that the notion that more than one motif is required may be questioned. At the same time the term “episode” amounts to an imprecise gesture in the direction of event depiction. In what follows I shall emphasize the idea of an event or occurrence as central to any definition of rock art scenes. It seems clear that while we need not be averse to occasional speculation about the depiction of settings for events, the notion of specific landscape representation has been regarded as marginal to a discussion of rock art pictures, since few identifications have ever been made (Davidson 2017: 35). Event

EVENT DEPICTION IN ROCK ART

55

depiction is after all what Spencer and Gillen assumed as a definition, without putting it into words: referring to depicted activities like kangaroo chases and emu hunts, they knew a scene when they saw one. But might there be a more precise term than scene for pictures of this kind? If not settings, if not landscape, then what remains? Scene understood primarily as narrative. This has the advantage of allowing us to put aside the expectation of a depicted setting. In a number of articles whose purpose is to clarify what we mean by a scene, Livio Dobrez has chosen narrativity as his criterion, analyzing the scenic in terms of represented motion. Addressing critical operations of the human visual system when perceiving motion in still images, he refines a terminology capable of bringing into focus the capacity of a particular category of pictures to communicate a sense of something happening—without narrative content having to be inferred. This will not prevent him from also using the word scene to evoke the spectacle of, say, a Namibian rhinoceros “scattering a group of humans” (Figure 3.2) or fighting Injasuthi figures (L. Dobrez 2013: 426, 421). Movement/action/scene: these terms are viewed as broadly synonymous. All convey the idea of an unfolding narrative or of event depiction. I intend to follow the logic of this terminology in the rest of this chapter. One detail that follows from it is that there is no requirement for there to be more than one figure involved: a single man running is as much a narrative as several men running. Emphasis on pictorial narrative is in line with Leroi-Gourhan (1982: 38) who, without mentioning the term, effectively defined it when he wrote: “The translation into images of the movement of isolated or grouped beings is the only process which can suggest the passing of time.” This singling out of the representation of movement goes to the heart of what is implicitly understood by the word scene in the discourse of rock art studies. However, Leroi-Gourhan is less than rigorous in his application of the idea. Thus, while he is consistent (if arguably not entirely accurate) in insisting that “there are no scenes in the Paleolithic,” he fails to apply the criterion of movement to his identification of a scene in the Lascaux Shaft—interpreted as

FIGURE 3.2. Scene. Noukloof Mountains, Namibia. Copyright Patricia Dobrez.

56

PATRICIA DOBREZ

“a bison . . . disembowelled by a spear blow while tossing a man” (Leroi-Gourhan 1982: 38). Without initiation into the mythological world of the makers of the art, we cannot say what was meant by this collection of diverse motifs lacking depiction of movement. Whatever the degree of association, the figures themselves are static, so that one arrives at the interpretation of a struggle between a man and a bison only by inference. Likewise, perusal of the panel will probably dispel the notion that the motifs taken together may be regarded as an “explicit narrative,” a type defined by Keyser et al. (2013: 85) as “one in which the formal arrangement of figures clearly indicates that what is depicted is a narrative—a story, an event, something happening.” This amounts to saying that the composition’s narrative element is immediately grasped. A picture of a man digging may be intended to carry the meaning “Stop. Men at work.” But it is nevertheless understood perceptually as a man digging. Before considering types of narrativity, a brief digression is warranted on the positive and negative legacy of the “faithful study of nature”/”large vision of relations” tenets of the nineteenth-century Realist or Naturalist aesthetic (Eliot 1856: 626; Eliot 1860 [1993]: 247), the aim of which was to present figures, or in the case of literature, characters, in context. Art of this kind has provided a yardstick for the West’s aesthetic appreciation of rock art for over a century, as Chauvet, Brunel, and Hillaire (1995: 106), characterizing the images found at Chauvet Cave, demonstrate in their assertion that everyone who sees them remarks on the “naturalism” of the figures and the “veracity” of their postures. In some instances an assumption of a mimetic quality in rock art figures has led in useful directions, as with Azéma’s ethological studies of animal appearance, method of locomotion, and deportment with reference to depictions at numerous French sites (Azéma 2009, 2010). We may doubt, however, that an assemblage of bison at Fontanet (possibly agitated in a period of rut) constitutes a narrative whole (Azéma 2010: 133), as there is no indication of a unifying activity involving the figures. This assessment would be consistent with Leroi-Gourhan remarking of the Altamira ceiling that it “contains about twenty bisons in various postures without it being possible to establish a collective activity” (1982: 38). Still, it is perhaps not Azéma’s intention to go beyond the claim that each Fontanet image represents a specific characteristic, rather than activity suggestive of a narrative. The question then becomes: when does pictorial vivacity amount to narrative and when does it simply enhance canonical characterization as defining the appearance/typical deportment of an animal? These questions aside, the naturalism assumption has led to the identification of event depiction where there may be none. Beyond that of the Lascaux Shaft, another case in point is Ruspoli’s (1987: 145) labeling of a group of lions in Lascaux’s Chamber of the Felines a “scene,” and what is more, “a realistic one,” with the author holding back nothing in his interpretation of narrative detail. Everything in Ruspoli’s reading depends on the notion that the felines in question are depicted in association. But in fact there is very little to justify this assumption. The depiction of more than one figure might equally be explained by an artist’s (or multiple artists’) appreciation of the sheer power by which an iconic image may evoke a sense of presence. A further issue: some scholars, for example Leroi-Gourhan (1982: 31–34), have been tempted to base rock art chronologies on evidence for modern types of perspective associated with naturalist tenets. Behind Leroi-Gourhan stands Breuil with his positive evaluation of mastery of viewpoint (Breuil 1952). Nonetheless, in this context the last word should go to Leroi-Gourhan’s caution that “expressions like realism, schematism and naturalism . . . only respond partially to the needs of an analysis” (1982: 34).

Landscape-Embedded Pedal Narratives With preliminaries out of the way, we are now in a position to focus entirely on narrativity. As already foreshadowed, there is more than one way of depicting events in rock art. In this section I shall be examining one of them: the trace- or map-view variety of rock art inscription (P. Dobrez 2017, 2018) involving the sequencing of tracks images, frequently in association with other iconic forms. Reproducing impressions that occur in real life, and co-opting ancient skills of tracking conspecifics and other animals, tracks sequencing is a convenient way of storytelling that may be older than any other form of narrative depiction. Indeed, considering the clear selective advantage of an ability to track and share the facts of the hunt, it may well have marked the beginning of the human practice of recording events, thus constituting our first recorded narratives (P. Dobrez 2015: 259; 2018). There is evidence for the longevity of images relating to tracking

EVENT DEPICTION IN ROCK ART

57

in Australia, where the iconography of “tracks and lines” petroglyph assemblages, often found on patinated broadly horizontal rock surfaces, has important elements in common with the inscriptive dimension of a multimodal system of communication involving speech, gesture and drawing in the sand (Green 2014). This coordinated operation—one that exploits the full potential of the human body to engage in acts of communication—appears to possess the features of a foundational system whose combined speech, gestural, and inscriptive iconography is shaped to capture events through a re-enactment of actions that are not only witnessed in the moment but also leave visible traces, i.e., pictures on the ground. For detailed analysis of the gestural language element of this system see Kendon (1988), and of the inscriptive element Munn (1973). In sand-drawn form, the marks are ephemeral. In petroglyph and pictograph form, they are enduring. In previous discussions I have isolated key aspects of the drawn or pecked images—readable iconicity, directionality, perceived movement, scope for association with complementary icons—all of which need to be considered if we are to grasp their functioning as surrogates for real events (P. Dobrez 2015, 2017, 2018). In the first instance, however, it is iconicity which needs to engage us at the theoretical level, because without it, visualization of an agent is impossible: we might have the sense of something happening in terms of directionality and movement, but that is all. Iconicity gives us a protagonist for a story. Readable tracks deliver a human or animal engaged in some activity, as exemplified in this sand drawing–derived case paralleling petroglyph pedal narratives (P. Dobrez 2017) depicting a human looking for witchetty grubs (Figure 3.3). Given the quite recent turn to consideration of the role iconicity might have played in the story of multimodal communication (Levinson and Holler 2014; Vigliocco, Permiss, and Vinson 2014), it is puzzling that multimodality has been investigated as a dual system of speech and gesture alone, without due attention to the possible developmental role of inscription. It seems entirely relevant to consider the relationship between gestural iconicity and iconic inscription at a time when gesture, supplanting spoken language, is emerging as the key evolutionary player. (For relevant discussion of the theoretical turn to gesture origins see P. Dobrez 2017: 163–65). Levinson and Holler (2014: 3) write: “Where iconicity comes into its own is in gesture. Iconic gestures are the gestures that mimic motion, depict size, trace shapes or sketch spatial relations between things.” Clearly rock art research has an opportunity to contribute to this discussion, as it seeks to identify and date our earliest visual marks and speculate about their role in cognitive evolution. I shall return to the issue of iconicity in connection with rock art below. Here I merely note that we can be confident that rock art investigations should make an important contribution to establishing a multimodal model by posing

FIGURE 3.3. Charlie Marshall Tjungurrayi, Warlpiri people. Witchetty Grubs (maku) at Kunatjarri, 1984. Synthetic polymer paint on canvas, 122 x 244 cm. National Gallery of Australia, Canberra. Purchased from Gallery Admission charges 1984. Copyright the artist. Licensed by Aboriginal Artists Agency Ltd.

58

PATRICIA DOBREZ

the question: where might a “role for inscription” thesis sit with the “gesture first” or “co-speech” hypotheses gaining ground in evolutionary communications theory? I now come to a key issue for understanding the powerful affordance offered by pedal sequencing. This is the way in which tracks trails directly communicate “something happening.” Like hand icons, human foot icons are transparently recognizable. In their trace forms (in rock art these occur as stencils and prints), each has equal claim to arrest attention and to make their “presencing” or signature (P. Dobrez 2017: 150–51) meaning apparent in an immediate way. How so? It is part and parcel of our proprioceptive awareness of our action-ready bodies that we comprehend the shape of a hand or foot “from the inside,” as it were (P. Dobrez 2018: 51–53). As a consequence, we are able to connect via motor resonance with an external image of a hand or foot as something more than a visual percept. This is no doubt true for other body parts as well, but—apart from the occasional whole-body stencil that has been observed in Australia—in terms of human body images it is most often hands and feet that occur in rock art, encouraging us to investigate their role as foundational icons in communication systems involving pictorial signs. Like Robinson Crusoes of prehistory, we recognize human presence through the print of a foot, and where there is more than one, a story begins to unfold. Only very recently has neuroscience turned its attention to what Gallese (2017: 44), investigating the human tendency to create images, has described as the “intercorporeality” that is “the main source of the basic knowledge we entertain of others.” Gallese (2017: 43–44) sets out to challenge the theoretical position of, as he puts it, “Visual Imperialism,” arguing: “Our vision of the world is far more complex than the mere activation of the visual part of the brain. Neuroscience has shown that vision is multimodal: it encompasses the activation of motor, somatosensory, and emotion-related brain networks.” This understanding of neural interconnectedness must have implications for the communication multimodal model as we seek to understand how spoken language, gesture, and inscribed marks work together, especially in the common-ground situations of Australian desert Aborigines engaging in their traditional face-to-face exchanges. It is logical to think that our reading of animal tracks must have piggybacked on our proprioceptive understanding of human tracks. Animal tracks present in the same way, but they are not iconic universals. Each ecological situation offers a new range of animal prints to be identified and remembered. I have explained the modus operandi of pedal icons by drawing attention to the laboratory-evidenced phenomenon of our spontaneous ability to see movement in a motion-cued still image (P. Dobrez 2015, 2017, 2018). The cited experiments are those of Freyd (1983) and Freyd and Finke (1984), in which subjects are reported as experiencing a displacement in a forward direction of still images to which they have been briefly exposed. The notion of “representational momentum” or “implied motion” has been subsequently retested and interrogated (Hubbard 2010), and is now generally accepted: we do perceive motion in still images. If this much is granted, the question remains: how is an actual event involving a moving figure perceived in the case of tracks? My answer is that it is not a matter of inferring it through a process of introspective matching up of cues (e.g., extended legs for profile images) with remembered experience of visual event patterning. Although it is undeniable that the art of tracking developed as a detective, that is, inferential, skill, its basis is other, in that perception of the whole is immediately proprioceptive, involving all the senses that tell us where we are and what space we occupy. We lend our own body schema to the tracks images, so that we “see” the event through another modality altogether, namely our own multi-sense body awareness. This is not a matter of “illusion” as suggested by the (I would argue misnamed) “Rubber Foot Illusion” (italics mine), but of prosthetic incorporation resembling the extension of body awareness in tool use, a fact that suggests the tracks of other animals will be incorporated as readily as human tracks. In the case of visual input from pedal images, present body awareness and own motor memory contribute to the perception of a moving figure in the actually present contextualizing landscape. What is at stake here is the idea that an effective identification of bodily self and extra-bodily self is possible, and that this delivers a special kind of active perception that is capable, in the case of pedal images, of allowing the perceiver “to stand in the shoes” (colloquial language often strikes home) of the human or animal associated with the particular iconic form used to register the tracks. In other words, tracks imagery is not inert: it is possible to perceive, and mimic, that is, follow sequenced tracks motion rather than concluding there is depicted movement through a species of detective work. Perception of motion in rock art imagery is precisely that: perception, not inference. This is not to deny that there is strong evidence of reliance on painstakingly gathered information from which inferences about species, gait, velocity, age, and sex are made by expert trackers, but this not where the process of reading tracks begins.

EVENT DEPICTION IN ROCK ART

59

I wish to emphasize a last point relating to a major difference between trace/map-view pedal narratives and profile narratives, to which I turn below. Pedal narratives are situated narratives, embedded in topographies culturally understood as “country” (Myers 2002: 47). As such they may in many instances unfold within an absolute frame. Writing about Australian desert sand drawing, whose motifs correspond to rockart tracks and lines sites, Green (2014: 22) asserts that depicted objects such as a windbreak (drawn as an arc) and motion pathways “are expected to convey correct information about direction.” When it comes to pedal narratives, both the unmistakable tell-tale, real-life traces of events and their sand-drawn versions are situated in real world landscapes with their practical affordances for survival and long-accumulated mythic overwriting (Figure 3.4). Landscape-embeddedness is thus a major feature. Such narratives do not

FIGURE 3.4. Pedal narrative. Owens Valley, California. Copyright Patricia Dobrez.

60

PATRICIA DOBREZ

require contextualization in the form of a depicted setting, as the artist’s landscape is there to provide all that is needed in terms of locality, orientation, possible destinations, and so on. This applies, of course, to both ancient and more recent pedal depictions. The compositional similarities between the two systems, sand drawing and rock art, invite us to consider that rock art tracks narratives might be oriented in the same way. Although they have settings (in the real world), it would nonetheless be confusing to refer to them as scenes. “Pedal narratives” puts the emphasis nicely on what tracks-based depictions have in common with scenes, namely storytelling, and the attribute that marks them as different, that is, plan (or map) view.

Decontextualized Profile Narratives Throughout this discussion we have been dealing with two types of visual storytelling whose operations are somewhat different. A major advantage of comparing trace-form narratives with what is generally discussed in rock art studies under the heading of “Scenes,” is that we are able to see clearly, and perhaps for the first time, that a bias exists in the discourse toward the depiction of events in profile. The template of profile (including post-Renaissance ¾ view) event depiction reflects both art history and general usage, though usually in an unanalyzed form, and takes us back to the historical moment of re-engagement with rock art when Altamira was revealed to the public and the prevailing idea of a scene was the familiar type of painting encountered in the European landscape tradition. Confining inquiry to profile depiction keeps us from asking important questions about which type of representation might have come first, and whether one has representational advantages over the other. On the subject of the way in which profile narrative operates, there is no need to cover the same ground as Livio Dobrez, included in this volume. It suffices to recall that he has both held to the view that “a scene is best thought of as a visual narrative” (L. Dobrez 2012: 1839) and sought to explain the centrality of profile in such narrative. My own analyses of plan-view pedal narratives complement this. They share his perceptual approach to event depiction as an organic biological activity involving spontaneous processes, rather than as a kind of detective work on the part of the viewer (including the maker) by means of which images in motion are built up mechanistically piece by piece. On the question of the relative advantages of plan-view trace-form over profile depiction, we can say that the former clearly relates to a way of life in which human subjects are themselves embedded in landscape in such a way that they take their cues from it. At every moment, the fact of living in a given place (perhaps by analogy after the manner of indigenous Australians, for whom identity is inseparable from relationship with “country” [Myers 1986]) will frame their lives and likewise their representational activities. Consequently, their trace-form narratives are positioned in the landscape as belonging to it. Known environmental features such as watercourses and sand hills will constitute a concrete real-world setting for depicted events. Real-world landscape orientation has been confirmed at the petroglyph site of Panaramitee Hills, South Australia (Mott 1998), with the expectation that comparable studies will be undertaken in the future. The point is that, as noted above, where real and depicted dovetail there is no need for “settings” as such. On the other hand, and in contrast to pedal storytelling, profile narratives set their own frames. With these there would seem to be a predetermination that a vertical rather than horizontal surface will be chosen, so that the represented scene—understood as visual narrative— will open out like a window on a remembered or imagined event. Where one might expect a depiction of settings for figures—to compensate for the release of narrative profiles from actual defining landscape contexts—we find that this is not generally the case in rock art, though of course it is common in many historical art traditions. Either way, however, profile scenes are decontextualized. They separate themselves from their real-life settings. This does not mean that by drawing attention to profile narratives as windows —or, to use a contemporary metaphor, hypertext—we necessarily diminish the role played by topography. Landscape situations are always meaningful and often highly suggestive. Whatever and wherever they are, they will in all probability maintain a degree of relationship with a depicted story. However, profile depiction frames itself off from its location in a way that does not occur with pedal narratives. In saying this I do not mean to play down the role of what has been learned and previously imagined or acted out in ritual, as when an abbreviating pictogram triggers memory of a rich, culturally significant legendary exploit or myth. In other words, an acculturated viewer will supply a context where none

EVENT DEPICTION IN ROCK ART

61

is depicted. It is also important to note certain kinds of marginal rock art instances. If, for example, a crack or fissure in a rock surface is integrated into a profile composition, we may see this either as incorporation into the relative frame of the picture itself, in which case the natural feature assumes a new meaning distinct from its environmental one, or as recontextualization of the picture in real-world terms. It is perhaps a moot point whether we choose one or other option. On the larger question of comparison between embedded and decontextualized narrative, the differences are significant because they allow us to speculate that each option may offer a distinct kind of symboling affordance, and may indeed mark a different way of life or at least different aspects of a way of life. Whereas pedal narratives are characterized by direct involvement in a theater of re-enactment, profile narratives give scope for contemplative distance. I would like to end this brief discussion of profile narrative by returning to the multimodal communication model in order to raise the issue of origins. Above all, the model opens up the possibility of taking an integrated approach to the relationship between, on the one hand, the practice of telling stories via pedal narrative depictions, and on the other, via profile depictions. It stands to reason that the representational practices we encounter in the record require antecedents. If we are prepared to entertain the hypothesis that pedal narrative depiction in rock art is derived from the multimodal communication system that has been documented for Australian Aboriginal desert groups (Green 2014; Munn 1973), the Davidson and Noble (1989: 130) notion of “fixing” gesture in the case of profile depiction prompts a new question: how might profile narrative (the Davidson and Noble fixed gesture) sit with plan-view narrative, whether as a complementary dimension to ground inscription or, in some situations, a substitute for it? Which comes first is an important corollary question. Although we have scant material to work with here, it seems important to model alternative symboling practices, given the options available to a multimodal human brain. Though evidence of formative depictive practices has been lost along with the ephemeral media in which they arguably developed (Davidson and Noble 1989: 129), we are not prevented from speculating about the possible co-functioning of diverse narrative strategies. Davidson and Noble’s argument for the origin of iconic images in gestural mimicry is consistent with the multimodal communication model discussed above: “If we postulate that gesture and posture could indicate a bison by mimicry, then a trace [i.e., an image], as ‘frozen’ gesture, would be directly perceptible as looking like a bison” (1989: 130). The example offered brings to mind the emu head and neck images found at some Australian stencil sites, where the canonical features of the bird have been reproduced in the hand and arm posture of the artist. Here we are dealing with actual “trace” imagery, whereas the cervico-dorsal lines of horses and bison found as drawings at European cave sites would suggest imagistic conventions derived from trace versions. It is one thing to mimic, another to outline, and it might be argued that the former has logical priority. Genuine iconic trace forms may be present in Europe if we only knew how to look for them—at least in theory, either as stenciled hand and arm gesture forms or actual gestural icons fixed in a compliant material. At the same time, there are possible European candidates for the abbreviating and miniaturizing versions of pedal performance seen in Australian sand drawing in the form of extended lines and rows of dots to represent motion paths. Coexistence of plan view and profile would suggest co-functioning. In view of the strong suggestive links between sand drawing and Australian “tracks and lines” petroglyphs, I would argue that we need to look for comparable imageries and combinations around the world, particularly where we have known early dates. No matter which came first—gesture, vocalization or inscription (rock art’s potential contribution to the story)—it is apparent that our modern communication systems have deep-time roots and that there is support for “a long phylogenetic acquisition of layers”: “At the deepest stratum, one may assume a capacity for ad hoc gestural symbols ritualized from early elements of action sequences, as found among the present great apes” (Levinson and Holler 2014: 4). Levinson and Holler propose that gestural (involving pointing and “some capacity for iconic representation”) and vocalizing co-functioning should be seen as having “accumulated over the two and half million years that humans have been a cognitively advanced, tool-using species” (Levinson and Holler 2014: 4, 6). It is not a large stretch, as postulated by Davidson and Noble (1989), to imagine casual marking as an offshoot of gesture, wherein we may envisage a very early role for gestural fixing—this possibly in the context of a specific connection between toolmaking and the capacity to understand iconic marks: “the bodily systems engaged in the making of implements are exactly the ones employable as articulators of visible signs” (Davidson and Noble 1989: 130).

62

PATRICIA DOBREZ

If we also take into account hand and foot traces as possible iconic ur-marks, we have the makings of a contribution to the great symboling debate. A vertical surface offers an affordance for registering, say, the hump of a bison in mid-gestural flight. It also offers a horizonal view: a bison seen in picture profile, rather than identified through its plan-view tracks. What might this mean in terms of a sequence? Does plan-view representation necessarily precede the appearance of profile in the rock art sequence? Or, as suggested earlier, might not either actualization be simply explained in terms of affordable surfaces? Might not both plan and profile have appeared simultaneously?

Hybrid Profile and Map-View Narrative Forms One fact we can be certain about: the two forms do in fact exist side by side in a combinatory mode of storytelling, both in rock art and within other graphing media. The narrative capacity of both pedal and profile depiction is notably exploited in the “Biographic Tradition” art of the Great Plains of North America (Keyser and Klassen 2001; Klassen 1998). Here there is abundant available ethnography to throw light on culture-specific imageries and it has been deeply mined. Robe and Ledger art has been especially important in the decoding of details. What is notable is the occurrence of rock art panels (as well as Robe and Ledger graphics) that feature not only profile figures engaged in action, but also tracks trails presented in plan view (Figure 3.5). Given the busy-ness of many of the Plains Indian depictions in terms of the number of actions and interactions portrayed, and the degree to which ethnography can supply background information, it is not surprising that the language of scenicity is everywhere present in the literature. Consistent with other rock art commentary, the notion of a scene that is brought into play hinges once again on an appreciation of varied ways of depicting action. Indeed, Keyser and Klassen (2001: 224–25) define Biographic art as the art of event depiction combining diverse methods: “these scenes incorporate numerous pictorial devices to indicate action, movement, direction, and the passage of time.” If scenicity is not always immediately evident, it is because the Biographic Tradition is also characterized by elements of conventionalized shorthand. Other examples that come to mind are the narrative combination of profile bears and hunters with their tracks at Alta, Norway, and Crawford’s illustration (1968: 82, Figure 66) from the Kimberley, Western Australia, of a bi-angular composition depicting a kangaroo hunt (profile hunter and prey surrounded by plan-view kangaroo tracks). It is self-evidently possible to read these associations as active ones, illustrating an effective narrative strategy and one combining the two major plan- and profile-view categories. No doubt still further exFIGURE 3.5. Hybrid profile and plan-view narrative amples might be found around the world. Outside depiction. Writing-on-Stone, Alberta, Canada. Reduced rock art, plan-view footprint routes are a major from ink tracing on polyethylene by James D. Keyser, device in the dominantly profile art of Mexican codices (Brotherston 1995). Likewise, a combinatory 1976.

EVENT DEPICTION IN ROCK ART

63

approach is taken in contemporary Australian desert acrylics deriving from trace-view sand drawing but sometimes including profile figures. These instances bring into focus a practice much remarked upon in a very different context, that is, in relation to single figures. I mean the composite-view presentation for which Breuil used the term, now superannuated, of “perspective tordue” (twisted perspective). Such compositions, combining as they do profile and plan-view perspectives, have the advantage of promoting ease of recognition, regardless of the angle from which the object is viewed. Paradoxically, present-day viewers may have to adjust to bi- or multiangular narratives, while at the same time readily comprehending both profile- and plan-view elements. In a real-life setting such as that offered by the Hancock Library, Australian National University, they are quite capable of negotiating their way to the stacks by following footprints painted on the floor. In short, they can behave like trackers responding to a pedal narrative in actual landscape—in this case the winding corridors of the library. It is only when pedal narrative and profile depiction are brought together in one composition, as with American Biographic art, that the contemporary viewer registers a representational jarring—just as she does when viewing a figure simultaneously presented from more than one angle (so-called “twisted perspective”). This is because the present-day viewer has become habituated to post-Renaissance perspectival projection culminating in the nineteenth-century elevation of “realism” as outlined at the start of this chapter. The key point to bear in mind in this discussion of representational “hybridity,” as well as the entire case built up in this chapter, is that profile- and map-view would seem to indicate primary alternatives in the depiction of events.

Conclusion In its initial sections, this chapter considers various definitions of a scene, opting for scene as visual narrative or depicted event. In its later sections it investigates something that has not yet been adequately discussed in rock art studies, namely that there are two types of visual narrative fundamental to rock art. I open with an important question. How is “scene” understood in European nineteenth-century discourse at the time of the rediscovery of rock art and the launching of rock art studies? Taking the moment of Altamira’s reception as its type case, my discussion focuses on views of scenic representation current at the time. I look first of all at the historical context of a nineteenth-century “realist” aesthetic whose exemplar in English literature is the fiction of George Eliot. An appeal is also made to the work of Ruskin, the preeminent art theorist writing in English, whose views about the scenic were highly influential in his time. From this standpoint, a historical definition of the scenic emerges as figure(s) in a landscape, but with great, even sometimes exclusive, stress on landscape or setting, an attitude that culminated for Ruskin and others in a positive evaluation of Turneresque atmospherics in landscape art. Against this backdrop, I examine leading views as reflected in dictionary/glossary/Art History definitions up to the present and point to the expectations of rock art compositions they have set up. Nineteenth-century art provided templates for variations on the figure-and-backdrop notion of scenic representation, but, in sharp relief, rock art turned out to provide very few plausible instances of depicted settings for its figures. At the same time, the anthropologist Spencer at the turn of the twentieth century, steeped as he was in contemporary notions of the scenic, could read scenes in Arnhem Land rock art (where there are figures but not settings), but not in central Australian compositions of the kind illustrated by Basedow. This raises the question of why that might be. Are there not only scenes without depicted contexts, but also “scene-like” compositions that relate to but are distinct from scenes? The answer is apparent when “scene” is more fundamentally grasped less as figure(s) in landscape than as visual narrative. Moreover, it becomes difficult to resist when—picking up the difference Spencer saw (not analytically but with European nineteenth-century intuition) between Arnhem Land and central Australian compositions—we posit two equally fundamental and ubiquitous types of rock art, namely the profile and the pedal (respectively represented in the Spencer and the Basedow examples). Later sections of the chapter discuss these narrative types in turn, the pedal first, not only because in deriving from trace imagery it is likely to have the longer lineage, but also because it has so far not been examined by rock art scholars in connection with profile narrative. Such examination sheds light on the nature both

64

PATRICIA DOBREZ

of pedal narrative and the more familiarly scenic profile narrative. Both tell stories, that is, depict events, but they do so differently. Elsewhere I have argued (P. Dobrez 2018) that pedal narratives relate to hand images in rock art, because as trace images both are foundational forms. Indeed, trace images constitute a special representational category because they are not pictures-of something, but direct impressions or traces of an activity. Since pedal narratives are usually petroglyphs, they may be thought of as “trace-derived” (by comparison with stencils, for example). Overall, however, the point to be made is that trace/trace-derived rock art has not been fully theorized. Rock art pedal narratives take the form of plan-view representation and are incorporated into a real landscape; hence I describe them as “embedded.” Insofar as they relate to trace images, they serve a function of making something perceptually “present,” not by inferential logic but by the direct perception of a universally recognizable icon, the foot, or in the case of animals, the track. A track sequence (or trail) is not merely visually perceived as motion. It is here that I complement and add to Livio Dobrez’s perceptual-analysis approach, arguing that tracks are registered proprioceptively (as in the phenomenon of the “Rubber Foot Illusion”)—my hypothesis being that the tracker/viewer extends proprioception, or own body awareness, to incorporate other bodies (P. Dobrez 2017). This happens in a radical transformation of one’s body schema, comparable to the fused sensory inputs involved in driving a car or using a tool (P. Dobrez in press). In the case of sequenced tracks, my suggestion is that we extend our own body awareness to the animal/human we are tracking (P. Dobrez 2015). In short, there is an impulse to replicate the perceived motion of the trail—by following it. Thus the reading of pedal narratives is a more than visual exercise. It is multimodal, in the sense recognized by recent evolutionary communications theory—which, however, tends to focus on gesture and speech studies, overlooking the role of inscription provided by the example of rock art. By contrast with pedal narrative, profile narrative is described in this chapter as “inset” rather than “embedded.” It operates by the “window” principle of setting its own frame: we view the narrative as “decontextualized,” not of course divorced from its environment in terms of its cultural meaning, but perceptually framed off from it. I do not go into this in detail, since it is covered by Livio Dobrez in another chapter of the present volume. In the end, then, to what type of rock art composition are we to attach the term “scene”? My conclusion is that whereas profile action compositions are indeed scenes, we can use more precise terminology for them than the scenic. They are visual/pictorial narratives. Sequenced pedal compositions, while scene-like, are less scenes than a closely related type of visual narrative. Since these alternative depictive strategies, pedal and profile, have narrativity in common, it is not surprising that they may be combined in rock art. Noting that they instance composite view (Breuil’s “twisted perspective”), I have termed these combinations, which frequently occur, “hybrid”—offering the American Biographic tradition as the type case. Patricia Dobrez is an independent scholar currently working on the perception of rock art. She has held teaching and research positions (including that of lexicographer for the Oxford Australian National Dictionary project, ANU) at several Australian universities. Her publications include books and articles on literature and the visual arts. Her most recent publications have been for rock art journals: AIRA (USA), ARKEOS (Spain), the online journal Arts (Switzerland), the Boletín del Museo Chileno, Purakala (India), Rock Art Research (Australia), and Time & Mind (UK).

References Andersson, G. 2016. We Call Them Vikings. Halmstad: Maria Jansén. The Australian Pocket Oxford Dictionary. 1976. Melbourne: Oxford University Press. Azéma, M. 2009. Les Animaux Modèles: Aspect, Locomotion, Comportment. L’Art des Cavernes en Action, Tome 1. Paris: Editions Errance. ———. 2010. Les Animaux Figurés: Animation et Movement, l’Illusion de la Vie. L’Art des Cavernes en Action, Tome 2. Paris: Editions Errance. Basedow, H. 1925. The Australian Aboriginal. Adelaide, Australia: F. W. Preece and Sons. Beltrán, A. 1982. Rock Art of the Spanish Levant, trans. M. Brown. Cambridge: Cambridge University Press. Breuil, H. 1952. Quatre Cent Siècles d’Art Pariétal. Montignac, France: Centre d’Etudes et de Documentation Préhistorique. Brotherston, G. 1995. Painted Books from Mexico. London: British Museum Press.

EVENT DEPICTION IN ROCK ART

65

Cartailhac, E. 1902. “Les Cavernes Ornées de Dessins: La Grotte d’Altamira, Espagne; Mea Culpa d’un Sceptique.” L’Anthropologie 13: 348–54. Chaloupka, G. 1993. Journey in Time: The World’s Longest Continuing Art Tradition. Chatswood, NSW: Reed. Chambers 20th Century Dictionary. 1983. Edinburgh: W & R Chambers. Chauvet, J.-M., E. Brunel, and C. Hillaire. 1995. La Grotte Chauvet à Vallon-Pont-d’Arc. Paris: Editions du Seuil. Comte, A. 1853. The Positive Philosophy of Auguste Comte, trans. H. Martineau. London: John Chapman. Crawford, I. M. 1968. The Art of the Wandjina: Aboriginal Cave Paintings in Kimberley, Western Australia. Melbourne: Oxford University Press. Davidson, I. 2017. “Images of Animals in Rock Art: Not Just ‘Good to Think.’” In The Oxford Handbook of the Archaeology and Anthropology of Rock Art, ed. B. David and I. J. McNiven, 1–50. Oxford: Oxford University Press. Davidson, I., and W. Noble. 1989. “The Archaeology of Perception: Traces of Depiction and Language.” Current Anthropology 30(2): 125–55. Dobrez, L. 2012. “Towards a More Rigorous Definition of Terms: Are there Scenes in European Palaeolithic Art?” In L’Art Pléistocène dans le Monde, Actes du Congrès IFRAO, Tarascon-sur-Ariège, septembre 2010,. Numéro spécial de Préhistoire, Art et Sociétés. Bulletin de la Société Préhistorique Ariège-Pyrénées 65–66, ed. J. Clottes, 1837–51. ———. 2013. “The Perception of Depicted Motion.” Arts 2: 383–446. Dobrez, P. 1977. “George Eliot’s Concept of Self.” In Papers delivered at the A.V.S.A Conference held at the University of Sydney, August, 1975, ed. Alan Dilnot, 1–9. Wellington, New Zealand: Victoria University for the Australasian Victorian Studies Association. ———. 2013. “The Case for Hand Stencils and Prints as Proprio-performative.” Arts 2: 273–327. ———. 2015. “Their Proper Sphere: Situating Human Feet in Rock Art.” In Proceedings of the XIX International Rock Art Conference IFRAO 2015 (Cáceres, Spain), ed. H. Collado Giraldo and J. J. García Arranz, ARKEOS 37: 259–77. ———. 2017. “From Tracks to Gesture-Derived Inscription: An Australian Genealogy for ‘Tracks and Lines’ Petroglyphs.” Rock Art Research 34(2): 149–68. ———. 2018. “The Body as Intelligent Tool: The Making and Reading of Trace and Trace-Derived Forms in Rock Art.” Time and Mind 11(1): 41–65. ———. In press. “The Picturing of Weapons, Tools and other Objects at Australian Stencilled and Painted Rock Art sites.” In Weapons and Tools in Rock Art, ed. A. M. S. Bettencourt, M. Santos-Estevez, and H. A. Sampaio. Oxford: Oxbow. Eliot, G. 1856. “Art and Belles Lettres: Review of Modern Painters III.” The Westminster Review 65: 626–33. ———. (1860) 1993. The Mill on the Floss. London: Everyman. Feuerbach, L. 1857. The Essence of Christianity, trans. M. Evans [George Eliot]. New York: Calvin Blanchard. Flood, J. 1997. Rock Art of the Dreamtime: Images of Ancient Australia. Sydney: Angus and Robertson. Freyd, J. J. 1983. “The Mental Representation of Movement When Static Stimuli Are Viewed.” Perception & Psychophysics 33: 575–81. Freyd, J. J., and R. A. Finke. 1984. “Representational Momentum.” Journal of Experimental Psychology: Learning, Memory, and Cognition 10: 126–32. Gallese, V. 2017. “Visions of the Body. Embodied Simulation and Aesthetic Experience.” Aisthesis 1(1): 41–50. Green, J. 2014. Drawn from the Ground: Sound, Sign and Inscription in Central Australian Sand Stories. Cambridge: Cambridge University Press. Hackin, J., H. Clément, R. Linossier, H. de Wilman-Grabowska, C.-H. Marchal, H. Maspero, and S. Eliseev. 1964. Asiatic Mythology. New York: Crescent Books. Hubbard, T. 2010. “Approaches to Representational Momentum: Theories and Models.” In Space and Time in Perception and Action, ed. R. Nij-hawan and B. Khurana, 338–65. Cambridge: Cambridge University Press. Intraub, H., and M. Richardson. 1989. “Wide-Angle Memories of Close-up Scenes.” Journal of Experimental Psychology: Learning, Memory, and Cognition 15(2): 179–87. Kendon, A. 1988. Sign Languages of Aboriginal Australia: Cultural, Semiotic and Communicative Perspectives. Cambridge: Cambridge University Press. Keyser, J. D., and M. A. Klassen. 2001. Plains Indian Rock Art. Seattle: University of Washington Press. Keyser, J. D., L. A. C. Dobrez, D. Hann, and D. A. Kaiser. 2013. “How Is a Picture a Narrative? Interpreting Different Types of Rock Art.” American Indian Rock Art 39, ed. W. D. Hyder, 83–99. Glendale, AZ: American Rock Art Association. Klassen, M. A. 1998. “Icon and Narrative in Transition: Contact-Period Rock-Art at Writing-On-Stone, Southern Alberta, Canada.” In The Archaeology of Rock-Art, ed. C. Chippindale and P. Taçon, 42–72. Cambridge: Cambridge University Press. Kleiner, F. S. 2016. Gardner’s Art Through the Ages: A Global History, 15th ed. Boston: Cengage Learning. Leroi-Gourhan, A. 1982. The Dawn of European Art: An Introduction to Palaeolithic Cave Painting, trans. S. Champion. Cambridge: Cambridge University Press. Levinson, S. C., and J. Holler. 2014. “The Origin of Human Multi-modal Communication.” Philosophical Transactions of the Royal Society B Biological Sciences 369: 1–9. Lewes, G. H. 1858. “Realism in Art: Recent German Fiction.” Westminster Review 14: 488–518. Lewis-Williams, J. D. 1983. The Rock Art of Southern Africa. Cambridge: Cambridge University Press.

66

PATRICIA DOBREZ

The Macquarie Dictionary. 1987. 2nd ed. Sydney: The Macquarie Library. Mott, D. 1998. “Aboriginal Rock Engravings of the Panaramitee Hills.” BA (Honours) thesis. Adelaide, Australia: Flinders University of South Australia. Munn, N. D. 1973. Walbiri Iconography: Graphic Representation and Cultural Symbolism in a Central Australian Society. Ithaca, NY: Cornell University Press. Musquiz Pérez-Seoane, M. 1998. “Techniques, Procédés de Réalisation, Auteurs et Visées Artistiques des Peintures d’Altamira.” In Altamira, ed. A. Beltran, 59–87. Paris: Seuil. Myers, F. R. 1986. Pintupi Country, Pintupi Self: Sentiment, Place, and Politics among Western Desert Aborigines. Washington DC: Smithsonian Institution Press. ———. 2002. Painting Culture: The Making of Aboriginal High Art. London: Duke University Press. Nochlin, L. Realism. 1971. Harmondsworth: Penguin Books. Rensink, R. A. 2000. “Scene Perception.” In Encyclopaedia of Psychology, vol. 7, ed. A. E. Kazdin, 151–55. New York: Oxford University Press. Rock Art Glossary: A Multilingual Dictionary. 2010. Melbourne: Australian Rock Art Research Association. Ruskin, J. (1856) 1863 . Modern Painters Vol. III. New York: John Wiley. ———. (1860) 1885. Modern Painters Vol. V. New York: John W. Lovell. Ruspoli, M. 1987. The Cave of Lascaux: The Final Photographic Record. London: Thames and Hudson. Sanz de Sautuola, M. 1880. Breves Apuntes sobre Algunos Objetos Prehistóricos de la Provincia de Santander por Don Marcelino de Sautuola. Santander: Imprenta y Litografia de Telesforo Martinez. Sorensen P. M. 1999. “Religions Old and New.” In The Oxford Illustrated History of the Vikings, ed. P. Sawyer, 202–24. Oxford: Oxford University Press. Spencer, B. and F. J. Gillen (1899) 1968. The Native Tribes of Central Australia. New York: Dover. Vigliocco, G, P. Permiss, and D. Vinson. 2014. “Language as a Multimodal Phenomenon: Implications for Language Learning, Processing and Evolution.” Philosophical Transactions of the Royal Society London, B Biological Sciences 369(1651): 1–7. Webster’s New Twentieth Century Dictionary of the English Language. 1959. 2nd ed. New York: World Pub. Co.

==== •

•

•

4. DEFINING “SCENES” IN ROCK ART RESEARCH Visual Conventions and Beyond Madeleine Kelly and Bruno David

Defining “Scenes” in Rock Art Research The word “scene” originally derives from ancient Greek theater, where it refers to a setting or “sequence of dramatic action” (Oxford English Dictionary 2015: definitions 1, 2). Currently the term is used in a variety of ways in theater, film, and art. Of the fifteen definitions given in the Oxford English Dictionary, only one sufficiently refers to the visual arts: “A view or picture presented to the eye or mind of a place, incident, series of events” (Oxford English Dictionary 2015: definition 15). However, none of the specialist dictionaries consulted—including The Thames and Hudson Dictionary of Art Terms and The Oxford Dictionary of Art—define the term (Adeline 1967; Chilvers 2004, 2009; Clarke and Clarke 2010; Lucie-Smith 1993), nor do many leading rock art reference books (e.g., Chippindale and Nash 2004; Chippindale and Taçon 1998; Flood 1997; Layton 1992; Morwood 2002; Whitley 2001, 2011). Of the many rock art books we examined for this chapter, only Layton (1992) lists the term in the index. The general absence of “scene” in glossaries and indexes could lead us to think that the term is both well understood and of little importance in rock art discourse, but we may also arrive at a different conclusion (see below). Rock art researchers commonly require scenes to show juxtaposed motifs, yet there are several other words that also fulfill this role, including “gallery,” “panel,” “composition,” and “frieze.” The International Federation of Rock Art Organizations’ glossary (IFRAO 2017) and archaeologist Josephine Flood (1997: 353) both specify that a gallery is “a major site containing a large number of pictures” (a rock art “site” simply being a location that contains rock art). The spatial boundaries of a site are typically determined by the distances between the outermost extent of visible archaeological materials—such as stone artifacts on the ground or groups of images on a wall—or topographic features that house those archaeological assemblages in the landscape, such as rock overhangs or cave settings (e.g., Loendorf 2001: 59). A rock art site contains at least one or more rock art panels, each being a segment of a rock surface containing images (sometimes including particular rock configurations or non-anthropic markings). A panel is generally defined either by a change in the rock surface—e.g., a cleft or sudden change in direction of the rock—or by a concentration of motifs on a rock surface, so that spatially distinct groups of motifs may be considered separate panels (IFRAO 2017; Loendorf 2001: 61; Whitley 2011: 206). A long, continuous panel of rock art is often called a frieze (IFRAO 2017; Rosenfeld and Smith 2002: 106). Such definitions and uses of terms can vary from study to study according to the researcher’s judgment (Chippindale 2004: 102–115; Loendorf 2001: 61). These terms are usually perceived as clear descriptions defined primarily by either the natural landscape or the spatial distribution of motifs, rather than through a consideration of the motif content itself. Another term, “composition,” refers primarily to a set of images as intentionally composed on a rock surface, meaning that the term is frequently used interchangeably with “scene.” In relation to images, the Oxford English Dictionary defines a composition as “the action or art of disposing or arranging in due order the parts of a work of art, esp. of a drawing or painting, so as to form a harmonious whole” (Oxford English Dictionary 2015: definition 8). A similar definition is also used in the IFRAO glossary, which defines a composition as “the arrangement of parts of a rock art motif so as to form a meaningfully juxtaposed whole; cf. scene”; here “motif ” also acts as something equivalent to “panel.” These definitions focus on the making of an artistic ensemble, rather than on sets of images as isolated artifacts (see also May and Domingo Sanz

68

MADELEINE KELLY AND BRUNO DAVID

2010; Sauvet and Wlodarczyk 2008). The resulting composed imagery is sometimes considered a scene, as evident in the IFRAO glossary; rock art researchers sometimes refer to individual images also as compositions, for they too are usually made of differentiable parts. Flood (1997: 351) cites Robert Layton (1992: 187), for whom a composition is the sum of “meaningfully juxtaposed figures.” Such definitions have meant that the term “composition” is often used interchangeably with the term “scene” (cf. Balme and O’Connor 2015; Helskog 2004). Archaeologist Tilman Lenssen-Erz (1992) and art historian and literary researcher Livio Dobrez (2011a, 2012, 2016) have attempted to differentiate the two terms. For Lenssen-Erz (1992: 90), a composition is “made up of a number of elements placed together (i.e., com-posed) because the producer believes that they should be related to each other rather than separated from one another. . . . There is not necessarily a thematic coherence.” He implies that a composition relies on the intention of the artist (1992: 90). Dobrez (2011a: 75) echoes Lenssen-Erz’s view, defining a composition as “a series of marks which may be read as a unity. This may be because of textual, i.e., formal properties (the arrangement of marks), or contextual ones, i.e., the effect of the ambient frame (such as the rock face), or both.” He adds that while “it is generally taken to be intentional . . . this need not be the case,” meaning that a composition can be perceived “in the eye of the beholder.” Both Lenssen-Erz and Dobrez argue that a composition is a set of images that does not have to be meaningful, in contrast to scenes that do. Lenssen-Erz and Dobrez’s definitions differ from those of Flood, Layton and IFRAO, for each of whom a composition is intentionally meaningful. As neither Flood, Layton nor IFRAO compare the notion of “composition” with that of “scene,” their definitions leave open the possibility for “composition” and “scene” to be used interchangeably.

Lenssen-Erz’s Formal Scenes Lenssen-Erz (1992) attempts to clarify what is meant by “scene” and to distinguish the formal attributes that make a scene in rock art. Having analyzed San rock art from the Brandberg in Namibia—a body of rock art that features complex panels with “a high number of figures in comparatively dense concentration”— Lenssen-Erz argues that the essential factor in identifying a scene is its “coherence” (1992: 92). Measuring coherence involves determining formally similar motifs within compositions by examining their comparative structure (posture, action, size, elevation, and attributes) and content (species, type, color) (1992: 94–95). Lenssen-Erz adds three further formal requirements through which a scene can be “read”: 1. The “Subject or Theme,” which is determined by the motifs depicted (e.g., humans), 2. The “Focus,” being about the common action and interaction between the motifs depicted, and 3. The “Setting,” which includes “everything that is detectable in a scene but does not contribute to the focus,” yet may assist in defining a “temporal and spatial specification for that activity” (1992: 98). Lenssen-Erz thus defines a scene as a group of figures that are coherent (formally similar) and visibly interact in a manner that can be interpreted through their thematic content, focus, and setting. Under this definition, a composition can contain one or more scenes. Lenssen-Erz’s requirement of formally coherent scenes omits apparent scenes that contain motifs done in inconsistent styles, such as those that may have accumulated over time. Rock art can remain on surfaces for thousands of years and be exposed to generations of artists, creating complex panels and compositions. The analysis of coherence allows researchers to distinguish scenes in these situations. Carole Dudognon and Marcela Sepúlveda (2016) investigate several such complex panels in the south-central Andes, Chile, and determine patterns of superimposition among motifs and their formal similarities in order to group them into separate scenes that involve sets of interacting motifs with thematic and colorimetric coherence (Dudognon and Sepúlveda 2016: 142). Although Lenssen-Erz’s (1992) approach allows them to examine the buildup of consecutive scenes over time, it does not much lend itself to the transformation of individual scenes as more motifs are added (or modified). Here scenes are defined as synchronic creations on a decorated wall’s life history, not through time as new configurations are added. In contrast, Esther López-Montalvo’s (2013: 50, 51) study of scenes in the rock art of Upper Paleolithic Spain shows how older art can be incorporated into new scenes with thematic and formal incoherence (cf. David et al. 2017).

DEFINING “SCENES” IN ROCK ART RESEARCH

69

Lenssen-Erz addresses this issue of accumulation of motifs through time in a more recent paper co-written with Carole Fritz, Georges Sauvet, and others (Fritz et al. 2013). His three essential parameters for scenes— “Thème,” “Focus,” and “Cadre” (setting)—are restated, but now two types of scenes are distinguished. A “scène synchronique” (synchronic scene) involves multiple interacting figures sharing the same formal elements, whereas a “scène diachronique” (diachronic scene) accumulates over time and can consist of motifs with varying formal attributes (Fritz et al. 2013: 39, 41). Incorporating a diachronic definition allows for scenes that have been added to over time, yet maintains Lenssen-Erz’s requirements of theme, focus, and setting (see also Villaverde Ch. 15).

Dobrez’s Formal Scenes Dobrez (2008, 2011a, 2011b, 2012, 2015, 2016) also goes to some length to formally define and classify rock art scenes. He argues that “a scene is a depicted event, ‘something happening’” that can be perceived through a series of universal formal elements (Dobrez 2016: 9). Dobrez (2008: 29) takes a phenomenological approach, using his perceptual responses to rock art scenes to understand “how we read (in the sense of ‘interpret’) pictures on the basis of formal markers,” much as Lenssen-Erz does. Dobrez’s (2011a: 257) markers include “markers of movement,” “compositional asymmetries,” the “orientation of figures,” the “distance between figures,” the “size of figures” and their “profile depiction.” Each of these markers has requirements that need to be fulfilled for an image to be considered part of a scene. For example, the “distance between figures” must be close enough to allow “a reading of interaction,” but not too close, as this “confuses the eye” (Dobrez 2011b: 78). He recognizes that these formal markers are subjectively derived from perceptual information, although he also argues that his approach is based on neurophysiologically universal processes of perception and is thus “entirely a universalist one” (Dobrez 2016: 5). There is an inbuilt contradiction here: Whether or not perceptions are neurophysiologically universal, cultural conventions of the visual are not. While drawing “universalist typologies” from his perception of images, Dobrez acknowledges that cultural conventions impact the reception of images. An example of his recognition of cultural multivalence is the Panel of the Rhinoceroses in the Rear Chamber of Chauvet Cave (France), where, he argues, the black lines contouring the outlined figure of a rhinoceros are more likely to indicate multiple rhinoceroses, rather than a single rhinoceros’s movements, because the “indication of movement by [multiple surrounding lines] is a highly culture-specific modern convention” (Dobrez 2012: 8). The question of whether or not the painting of contoured lines to indicate movement really is a modern Western convention aside, Dobrez recognizes well that different cultures depict and read things in different ways (undermining the “universalist” and neurophysiological assumptions), although he does not explore how this might impact his classification of scenes. Ultimately, Dobrez’s notion of a scene is informed by Western visual conventions of form, allowing for the identification of scenes that conform to those conventions rather than to cultural conventions of scenic depiction informed by other kinds of criteria.

Reading Action through Interaction and Movement For Lenssen-Erz and Dobrez, scenes are visual narratives made readable through particular graphic conventions. Lenssen-Erz regards signs of interaction between motifs as the key, whereas for Dobrez it is the visible impression of movement, such as the presence of angled limbs to signify running (2016: 9, 14). Such indications of interaction or action are also used by other researchers when identifying scenes. For example, Delporte (1981: 59–71) distinguished five different types of associations between motifs on portable objects, one of which—“narrative” associations—has motifs performing definable actions. These are effectively action scenes, such as “a hunting scene,” For Australian Aboriginal rock art, every reference to a scene in Robert Layton’s (1992) Australian Rock Art: A New Synthesis concerns rock art images that formally display movement and interaction (Layton 1992: 187). Sally May and Inés Domingo Sanz (2010) also referred to the interaction of motifs in the rock art scenes they investigated. For them, a scene requires “motifs performing a common action . . . usually with

70

MADELEINE KELLY AND BRUNO DAVID

a defined time, that can be described even if the meaning and theme are unknown” (May and Domingo Sanz 2010: 37). May and Domingo (2010) worked in partnership with local Traditional Owners to interpret rock art scenes in western Arnhem Land (northern Australia). The initial selection of the scenes they studied, however, was approached with a predetermined notion of what a scene consists of. The presence of interacting motifs was a preformulated criterion, and local Aboriginal views were then being sought to determine the meaningfulness of the scene in question. What constitutes the concept of a scene was not the target of the enquiry (see also Domingo Sanz 2011; May et al. 2017). Michael Klassen’s (1998) study of petroglyphs at the Native American site Writing-On-Stone is in many ways similar. Klassen identified formal attributes that distinguish “iconic” from “narrative” rock art traditions. The “iconic” tradition consists of “static, frontal, symmetrical, non-interactive motifs,” while the “narrative” tradition “employs both profile and frontal representations, various perspectives,” where figures “are depicted interacting with other motifs” (Klassen 1998: 44, 45). Klassen refers to images of the “narrative” tradition as scenes (cf. Lenssen-Erz). He thus describes a set of human, horse, and tipi motifs at WritingOn-Stone as static and non-interactive, much like other “iconic” images. However, he cites ethnographic evidence that shows similar “scenes [that] depict the number of horses captured on a specific raid by the individual represented” (1998: 58), concluding that such graphic depictions “act as a ‘biographical’ record of a specific event” and therefore qualify as scenes (1998: 58). In this example, ethnographic evidence that the image depicts an event overrides a lack of formal “narrative” attributes, suggesting Klassen accepts that in a scene, narrative may be more essential than the presence of formal graphic portrayals of interaction. Despite the example of the “raid” scene above, Klassen does not explore the possibility that other, technically “iconic” motifs might, were further ethnographic information available, also display “specific events” and thereby meet the requirements of scenes. It appears that the initial selection of the “raid” scene was the result of the motifs’ formal similarities to each other in style, size, and orientation, as in Lenssen-Erz’s (1992) requirement of coherence. While this requirement is not why Klassen explicitly classified the composition as a scene, coherence seems nonetheless to have contributed to its selection for further analysis as a unified composition. Much like May and Domingo Sanz, Klassen used ethnography, but his method of attribution was grounded in formal criteria based in Western visual conventions. Jane Balme and Sue O’Connor (2015) provide an example similar to Klassen’s. They used historical records to suggest that each motif depicted in a rock art panel from the Kimberley region of northern Australia is part of a scene of “everyday life in the port of Derby” (2015: 77, 78). The “port scene” involves six motifs, each formally static and non-interactive. The scene is again consistent with Lenssen-Erz’s concept of coherence: the motifs were all made in a single color, they are close to each other, and they are all of similar size and style. Yet the composition evidently still qualifies as a scene only once local historical knowledge is considered. This suggests the identification of a graphic narrative derived from a combination of specific visual conventions and contextual historical knowledge.

One or More Motifs? Another commonly recognized criterion requires a scene to “involve more than one rock art motif ” (IFRAO 2017) (see also May and Domingo Sanz 2010: 37). Initially, Lenssen-Erz (1992: 93) also adhered to this requirement because multiple motifs need to be present to demonstrate coherence and interaction in his formal scenes. But Fritz et al. (2013: 42, 43) provide examples of scenes consisting of single motifs, such as Le Cheval Tombant (The Falling Horse) and La Vache Sautant (The Leaping Cow) at Lascaux in France. It appears that Le Cheval Tombant is interacting with its surroundings, a large crack in the rock; here the painted motif is not isolated but understood to interact with its setting, so that the latter becomes part of the scene. This is consistent with Fritz et al. (2013: 39), who note that “a scene describes a specific event that occurs at a given moment and involves a determined number of participants” (our translation from the French original), allowing single motifs to be treated as scenes. Dobrez also challenged the assumption that scenes must involve two or more figures. He argues that “the basic requirement for a scene is a visual marker . . . of ‘something happening’” and therefore can depict a single figure “such as a man running” (Dobrez 2012: 3; 2016: 9). Dobrez (2012: 5) notes that this basic

DEFINING “SCENES” IN ROCK ART RESEARCH

71

requirement allows for the classification of La Vache Sautant as a scene. However, in the same study he also argues that the basic narrative “the animal is doing such and such” is not “elaborate” enough to be considered a scene but is rather a “signifier” of a narrative that requires insider knowledge of “the story proper” (2012: 12). This implies that a scene must visually display a more complex narrative, though it remains ambiguous how complex it needs to be, and how that complexity can be identified. Such a condition would also mean that images requiring informed narratives would not in themselves constitute rock art scenes even if the oral narratives were known, thereby excluding Klassen’s “raid” scene and Balme and O’Connor’s “port scene.” While Dobrez does not clearly explain why “the animal . . . doing such and such” is not a scene but “a man running” is, he does, like Fritz et al. (2013), technically allow for scenes to consist of single motifs.

Background or Interference? The main material difference between rock art and so-called “fine art” is the lack of an original blank canvas and clear frame in the former. James Keyser and George Poetschat (2004: 118) argue that “rock surfaces are not neutral canvases, but themselves shapes of form and meaning” that “artists have incorporated . . . into their pictograph or petroglyph compositions.” May and Domingo Sanz (2010: 40) speculate that rock fissures crossing the scene they analyzed in western Arnhem Land may replicate features of the landscape in which the scene occurs, a concept supported by local Traditional Owners. Le cheval tombant is one of many Upper Paleolithic European images that incorporate pre-existing rock features into their compositions. Such associations between art motifs and rock surfaces are formal aspects that neither Lenssen-Erz nor Dobrez explicitly addresses in their considerations of scenes. Helskog (2004: 268) shows how fissures and crevasses were incorporated into the art in Scandinavia. A panel from Alta, Norway, contains several potential scenes. One motif of a double circle on a slanted surface is said to be of a bear den; three sets of bear tracks emerge from it (Helskog 2004: 273). One of the sets of tracks extends around the den; another, to the right (of the observer), and the third to the left, producing what he identifies as a “hunting scene.” Helskog derives further storylines for the panels from Saami ethnography. He suggests that the tracks going around the den point toward “higher parts of the terrain such as mountains to contact spirits in the upper world(s),” while the tracks traveling rightward lead down into a natural depression that contains a pool of water, which he argues “connects the bear with the spirits in the lower world” (Helskog 2004: 273). The tracks leading left travel along the bottom of a shallow depression in the panel, emerging on a sloping surface “where they are surrounded by armed figures and attacked” (Helskog 2004: 273.). Helskog describes a complex composition involving several small scenes, the bear’s activities in each location, and the larger scene of all the bear’s travels. The topography of the panel forms an integral element of the narrative scene he describes.

Formal and Informed Narratives In all the studies discussed above, narrative arises as the essential element that makes a scene, setting it apart from a composition. In some cases, the narrative is visually implicated by interaction of, or movement in, the rock art motifs, sometimes incorporating their surroundings as demarcated by features of the rock. In others, it is derived from ethnographic knowledge or historical context. Fritz et al. (2013: 45) argue that narrative is an intrinsic element of rock art’s meaningfulness, but that not all rock art narratives signify scenes. Here “scene” applies when images possess Lenssen-Erz’s criteria of theme, focus, and setting (Fritz et al. 2013: 42), but, like Dobrez (2012: 12), they do not treat compositions that “integrate several episodes of a narrative into a single image, ignoring the unity of time, place and action of each constitutive scene” as scenes, as these “are practically incomprehensible outside of an oral context” (2013: 44, our translation from the French original). On the other hand, David Lewis-Williams (1990) critiques Lenssen-Erz’s (1989) reliance on formal attributes, as “the concept of a pictorial ‘scene’ is a major assumption based on a Western ‘reading’ of the art” (Lewis-Williams 1990: 131), a critique also made by Iain Davidson (2017: 22). Where artworks appear to be devoid of visual expressions of movement, let us not assume that what the reader sees

72

MADELEINE KELLY AND BRUNO DAVID

as static imagery signals static mindsets. Images are, rather, multivalent in their expression, in which past events, present interactions, and knowledge each, and together, remain real and emergent.

Discussion and Conclusion Ultimately, in all these formulations a scene mediates a story through its imagery. This allows for the inclusion of informed knowledge in its initial identification. Relying primarily on formal characteristics recognized in Western visual conventions to define a scene in rock art when Indigenous knowledge is available for that same body of rock art excludes Indigenous visual conventions and silences Indigenous narratives. This is less a question of attributing knowledge to the creators of the art (or to their culture) by contemporary descendants (whose cultures may or may not have changed considerably since the art was made) than a recognition that the art is meaningful to more than contemporary Western users. Dobrez (2016: 6) argues that determination of specific formal characteristics of scenes can be based on a “universalist” approach to rock art. Rolando Vázquez (2011: 34) and Walter Mignolo (2009: 3) have noted that the Western establishment of “universal” categories creates a “monopoly over the real,” silencing Indigenous knowledge in the process. Dobrez (2016) acknowledges the limitations of his methodology, for his “universalist analysis cannot arrive at conclusions about the specific meanings of rock art” and “must remain at an appropriately general level,” unable to “comment on culture-specific aspects of art.” Yet he still posits Indigenous knowledge as unnecessary for the initial identification of rock art scenes, creating a hierarchy of epistemological processes in which Western positivism reveals the “real” types, while Indigenous epistemologies merely interpret it. This is not to say that Western formal analyses of rock art are not valuable. Formal analysis provides valuable information on stylistic patterning across space and through time, useful means by which to investigate actions and interactions in the world of the artist. But this does not mean that formal definitions of scenes are universal; they remain but one perspective on the subject. The privileging of Indigenous epistemologies and ontologies over or alongside Western perspectives allows for a reconsideration of rock art typologies, such as those articulated in the notion of a “scene.” Mignolo (2009: 15) argues that “de-linking” our thinking from Western ideals, as a form epistemic disobedience, lends itself to a rebalancing of dominant epistemologies. Likewise, Liam Brady and Amanda Kearney (2016: 3) note that incorporating Indigenous methodologies in rock art research allows for new conceptualizations of rock art and its place in the world. Working closely with social anthropologist John Bradley and members of the Yanyuwa community of the western Gulf of Carpentaria in northeastern Australia, rock art is revealed “not as a passive signifier of meaning, but instead as an entity that is active and capable of generating dialogs and responses that illustrate the meaningfulness of images today” (Brady, Bradley, and Kearney 2016: 28), a theme we take up in relation to Wardaman art of the Victoria River region of the Northern Territory in another chapter (Kelly et al. Ch. 11).

Acknowledgements We thank the Australian Research Council Centre of Excellence for Australian Biodiversity and Heritage project CE170100015 for funding this research, and the Monash Indigenous Studies Centre at Monash University for research support. Madeleine Kelly is an archaeologist currently completing her PhD at the Monash Indigenous Studies Centre at Monash University. She is also part of the Centre of Excellence for Australian Biodiversity and Heritage (CABAH). She is undertaking archaeological and anthropological research on the multivalent spatial and temporal relationships of rock art in Wardaman Country in the Northern Territory. She has also worked on excavations and rock art recording projects in the Kimberley and Victoria. Bruno David is an archaeologist at the Monash Indigenous Studies Centre, Monash University, and Chief Investigator with the Australian Research Council Centre of Excellence for Australian Biodiversity and Her-

DEFINING “SCENES” IN ROCK ART RESEARCH

73

itage. His most recent books include the Oxford Handbook of the Archaeology and Anthropology of Rock Art (Oxford University Press, 2018, co-edited with Ian J. McNiven) and Cave Art (Thames & Hudson, 2017). He is currently undertaking research on the archaeology of deep-time, spanning to recent cultural landscapes in close partnership with Indigenous Traditional Owners in East Gippsland and the Kimberley, Australia.

References Adeline, J. 1967. The Adeline Art Dictionary. New York: Frederick Ungar. Balme, J., and S. O’Connor. 2015. “A ‘Port Scene’: Identity and Rock Art of the Inland Southern Kimberley, Western Australia.” Rock Art Research 32(1): 75–83. Brady, L. M., and A. J. Kearney. 2016. “Sitting in the Gap: Ethnoarchaeology, Rock Art and Methodological Openness.” World Archaeology 48(5): 642–55. Brady, L. M., J. J. Bradley, and A. J. Kearney. 2016 “Negotiating Yanyuwa Rock Art: Relational and Affectual Experiences in the Southwest Gulf of Carpentaria, Northern Australia.” Current Anthropology 57(1): 28–52. Chilvers, I. 2004. The Oxford Dictionary of Art, 3rd ed. Oxford: Oxford University Press. ———. 2009. The Oxford Dictionary of Art and Artist, 4th ed. Oxford: Oxford University Press. Chippindale, C. 2004. “From Millimetre up to Kilometre: A Framework of Space and of Scale for Reporting and Studying RockArt in its Landscape.” In The Figured Landscapes of Rock Art: Looking at Pictures in Place, ed. C. Chippendale and G. Nash, 102–113. Cambridge: Cambridge University Press Chippindale, C., and G. Nash, eds. 2004. The Figured Landscapes of Rock Art: Looking at Pictures in Place. Cambridge: Cambridge University Press. Chippindale, C., and P. S. C. Taçon, eds. 1998. The Archaeology of Rock Art. Cambridge: Cambridge University Press. Clarke, M., and D. Clarke. 2010. The Concise Oxford Dictionary of Art Terms, 2nd ed. Online edition: Oxford University Press. David, B., J.-J. Delannoy, R. Gunn, L. M. Brady, F. Petchey, J. Mialanes, E. Chalmin, J.-M. Geneste, I. Moffat, K. Aplin, and M. Katherine. 2017. “Determining the Age of Paintings at JSARN-113/23, Jawoyn Country, Central-Western Arnhem Land Plateau.” In The Archaeology of Rock Art in Western Arnhem Land, Australia, Terra Australis 47, ed. B. David, P. S. C. Taçon, J.-J. Delannoy, and J.-M. Geneste, 371–422. Canberra: ANU Press. Davidson, I. 2017. “Images of Animals in Rock Art: Not Just ‘Good to Think’.” In The Oxford Handbook of the Archaeology and Anthropology of Rock Art, ed. B. David and I. J. McNiven. Oxford University Press Online: http://www.oxfordhandbooks.com .ezproxy.lib.monash.edu.au/view/10.1093/oxfordhb/9780190607357.001.0001/oxfordhb-9780190607357-e-36. Delporte, H. 1981. L’Objet d’Art Préhistorique. Paris: Réunion des Musées Nationaux. Dobrez, L. 2008. “A Rock Art Typology: Narrative and Non-Narrative Figurative Representation.” In Aesthetics and Rock Art III Symposium International Series 1818, ed. T. Heyd and J. Clegg, 29–33. Oxford: Archaeopress. ———. 2011a. “Looking at Our Looking: A New Approach to the Definition of a Rock Art Scene.” American Indian Rock Art 37: 251–67. ———. 2011b. “Rock Art, Perception and the Subject/Object Binary.” Rock Art Research 28(1): 71–83. ———. 2012 “Towards a More Rigorous Definition of Terms: Are There Scenes in European Palaeolithic Art?” In L’Art Pléistocène dans le Monde / Pleistocene Art of the World / Arte Pleistoceno en el Mundo, ed. J. Clottes, 1–15. Tarascon-sur-Ariège: Actes du Congrès IFRAO. ———. 2015. “Making Sense of Pictures.” Bolletino del Centro Camuno di Studi Preistorici 40: 59–75. ———. 2016. “Theoretical Approaches to Rock Art Studies.” Rock Art Research 33(2): 1–24. Domingo, I. S. 2011. “The Rock art Scenes at Injalak Hill: Alternative Visual Records of Indigenous Social Organisation and Cultural Practices.” Australian Archaeology 72: 15–22. Dudognon, C., and M. Sepúlveda. 2016. “Rock Art of the Upper Lluta Valley, Northernmost of Chile (South Central Andes): A Visual Approach to Socioeconomic Changes between Archaic and Formative Periods (6,000–1,500 years BP).” Quaternary International 491: 136–45. Flood, J. 1997. Rock Art of the Dreamtime: Images of Ancient Australia. New South Wales: Angus and Robertson. Fritz, C., T. Lenssen-Erz, G. Sauvet, M. Barbaza, E. Lopez-Montalvo, G. Tosello. and M. Azema. 2013. “L’Expression Narrative dans les Arts Rupestres: Approches Théoriques.” Les Dossiers d’Archéologie 358: 38–45. Helskog, K. 2004. “Landscape in Rock-Art: Rock-Carving and Ritual in the Old European North.” In The Figured Landscapes of Rock Art: Looking at Pictures in Place, ed. C. Chippendale and G. Nash, 265–88. Cambridge: Cambridge University Press. International Federation of Rock Art Organisations (IFRAO). 2017. IFRAO Glossary, ed. R. Bednarik. Retrieved 6 June 2016 from IFRAO: http://www.ifrao.com/ifrao-glossary/. Keyser, J. D., and G. Poetschat. 2004. “The Canvas as the Art: Landscape Analysis of the Rock-Art Panel.” In The Figured Landscapes of Rock Art: Looking at Pictures in Place, ed. C. Chippendale and G. Nash, 118–30. Cambridge: Cambridge University Press.

74

MADELEINE KELLY AND BRUNO DAVID

Klassen, M. 1998. “Icon and Narrative in Transition: Contact-Period Rock-Art at Writing-On-Stone, Southern Alberta, Canada.” In The Archaeology of Rock Art, ed. C. Chippindale and P. S. C. Taçon, 42–72. Cambridge: Cambridge University Press. Layton, R. 1992. Australian Rock Art: A New Synthesis. Cambridge: Cambridge University Press. Lenssen-Erz, T. 1989. “The Conceptual Framework for the Analysis of the Brandberg Rock Art Paintings.” In The Rock Paintings of the Upper Brandberg. Part I: Amis Gorge, ed. H. Pager, 361–70. Cologne: Heinrich Barth Institute. ———. 1992. “Coherence: A Constituent of ‘Scenes’ in Rock Art.” Rock Art Research 9(2): 87–105. Lewis-Williams, J. D. 1990. “Review: Documentation, Analysis and Interpretation: Dilemmas in Rock Art Research.” The South African Archaeological Bulletin 45(152): 126–36. Loendorf, L. 2001. “Rock Art Recording.” In The Handbook of Rock Art Research, ed. D. S. Whitley, 55–80. New York: Altamira Press. López-Montalvo, E. 2013. “La Construction Narrative dans l’Art du Levant Espagnol: Une Image du Passé.” Les Dossiers d’archéologie 358: 46–51. Lucie-Smith, E. 1993. The Thames and Hudson Dictionary of Art Terms. London: Thames and Hudson Ltd. May, S.K., and I. Domingo. 2010. “Making Sense of Scenes.” Rock Art Research 27(1): 35–42. May, S.K., I. Johnston, P. Taçon, I. Domingo, and J. Goldhahn. 2017. “Early Australian Anthropomorphs: Jabiluka’s Dynamic Figure Rock Paintings.” Cambridge Archaeological Journal 28(1): 67–83. Mignolo, W. D. 2009. “Epistemic Disobedience, Independent Thought and Decolonial Freedom.” Theory, Culture & Society 26(7– 8): 159–81. Morwood, M. J. 2002. Visions from the Past: The Archaeology of Australian Aboriginal Art. New South Wales: Allen and Unwin. Oxford English Dictionary. 2015. “Scene, n.” Oxford English Dictionary, 3rd ed. Online edition: Oxford University Press. Rosenfeld, A., and M. A. Smith. 2002. “Rock-Art and the History of Puritjarra Rock Shelter, Cleland Hills, Central Australia.” Proceedings of the Prehistoric Society 68: 103–24. Sauvet, G., and A. Wlodarczyk. 2008. “Towards a Formal Grammar of the European Palaeolithic Cave Art.” Rock Art Research 25(2): 165–72. Vázquez, R. 2011. “Translation as Erasure: Thoughts on Modernity’s Epistemic Violence.” Journal of Historical Sociology 24(1): 27–44. Whitley, D. S. 2001. The Handbook of Rock Art Research. New York: Altamira Press. ———. 2011. Introduction to Rock Art Research. Walnut Creek, CA: Left Coast Press.

==== •

•

•

5. PUTTING SOUTHERN AFRICAN ROCK PAINTINGS IN CONTEXT The View from the Mirabib Rock Shelter, Western Namibia Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe

Introduction Every now and then, you are in the right place at the right time to witness a special occurrence that helps you figure something out about rock art. This chapter is about one such experience that happened to us in the Central Namib Desert of western Namibia. In this chapter, we briefly consider the rock paintings at the Mirabib rock shelter—a reasonably well-known Later Stone Age (LSA) archaeological site located on an Alaskite granite koppie overlooking the gravel plains above the Kuiseb River drainage (Figure 0.1). Mirabib was first excavated and brought to the attention of the archaeological world by Beatrice Sandelowsky (1977) more than four decades ago. For the last several years, we have been involved in a project to reanalyze the results of these excavations, and we have also done further test excavations of our own in the interest of gathering samples for new dating efforts and other forms of technical analysis. On a late September day in 2013, we had the (mis)fortune of being present at Mirabib on the first day of a freak late winter rainstorm. When this happened, we noticed water dripping from the roof of the rock shelter onto the rock art panel, in particular on several specific images belonging to varieties that have been previously linked with San rainmaking beliefs and ritual practices (Challis, Hollman, and McGranaghan 2013; Challis, Mitchell, and Orton 2008; Deacon 1988; Dowson 1994, 1998; Lewis-Williams 1977, 1981a, 1981b; Lewis-Williams and Dowson 1988; Prins 1990; Solomon 1992, 2008, 2016; Sullivan and Low 2014; Yates, Golson, and Hall 1985). We think that the structure of this unique visual metaphor, in terms of the placement of painting elements relative to the naturally occurring dripline, holds potential clues about the views concerning southern African San rock art. This chapter combines our analysis of the rock art imagery present at Mirabib with evidence gained from our excavations and reanalysis of the collections resulting from Sandelowsky’s prior fieldwork. The rainmaking imagery articulates with other forms of archaeological evidence, pointing to a complex set of adaptations for living at a site in this hyper-arid region lacking a perennial water source. We propose that the act of painting itself was not just a narrative device used by ritual specialists to convey information about religiously significant constructs or activities. Instead, we suggest that it was intended to function as an instrumental aspect of ritual activities performed in order to enact real-world consequences—in this case, by bringing rain. Our analysis benefits from the analytical perspective at the heart of this book: the distinction between “scenes” and “non-scenes” (Davidson, this volume), as well as the implications of this distinction for our understanding of the social context of rock art production. Our analysis of the rock art at Mirabib explores the utility of scenes and non-scenes as analytical constructs in examining southern African rock paintings. Through our analysis, we consider some implications concerning the nature of ritual behavior associated with the rock paintings at Mirabib. Finally, we draw some implications for the broader application of this analytical perspective to the corpus of San rock paintings in southern Africa.

76

GRANT S. MCCALL, THEODORE P. MARKS, JORDAN WILSON, ANDREW G. SCHROLL, AND JAMES G. ENLOE

Shamans, Myths, and Rainmaking in Southern African San Rock Paintings The latter half of the twentieth century saw a sea change in the interpretation of San rock art based primarily on analyses of early ethnographic studies of southern San hunter-gatherers, especially the /Xam (Bleek 1924, 1933a, 1933b, 1935, 1936; Bleek and Lloyd 1911; Orpen 1874). The breakthrough occurred when rock art researchers, especially Vinnicombe (1972, 1976) and Lewis-Williams (1980, 1981a, 1981b, 1983a) began to make connections between rituals and religious beliefs described in the early San ethnographies and the scenes depicted in southern African rock art. This shift in perspective had the effect of undermining the canons of the art-for-art’s-sake perspective, which held that San rock art represented actual events and that it was produced for the sheer joy of making art (Cooke 1969; Lee and Woodhouse 1970; Willcox 1963). It also fostered the emergence of the shamanism interpretive framework, which has tended to view San rock art as representations of the experiences of shamans during ritually induced altered states of consciousness (ASCs; Blundell 1998; Dowson 2007; Lewis-Williams 1981a, 1981b; 1982, 1983a, 1983b, 2002, 2013; Lewis-Williams and Dowson 1988, 1990; Lewis-Williams and Pearce 2012, 2015; Ouzman 1998; see also McCall 2007). In brief outline, the shamanism interpretive framework is based on three main sets of observations: (1) early ethnographic accounts of the /Xam and other southern San groups that recount the experiences of ritual actors (later taken to be ecstatic shamans) in the spirit world; (2) more recent ethnographic accounts of San groups in the Kalahari, especially the Ju/’hoansi (formerly the !Kung), in which shamans have been thoroughly documented entering ASCs during trance dance healing rituals (Katz, Biesele, and St. Denis 1997; Guenther 1975; Heinz 1975; Katz 1982; Marshall 1962; Wiessner 2002; Wiessner and Larson 1979); (3) a more general literature arguing for the universality of a complex of shamanic rituals and beliefs involving ASCs among hunter-gatherers and other small-scale societies (Eliade 1964; Lewis 1971; Lewis-Williams 2002; Lewis-Williams and Dowson 1988; Lommel 1966). There have been objections to the shamanism interpretive framework, especially as it has been generically applied to hunter-gatherer rock art as a putatively global phenomenon (e.g., Bahn 1997). In addition, there has also been considerable controversy about whether San rock art refers to the ASC experiences of shamans or something else, such as the activities of mythological figures or the spirits of dead ancestors (Solomon 2013). Much of this controversy stems from disagreement about the meaning of certain crucial /Xam ethnographic accounts. On the one hand, the shamanist camp has argued that many of these stories refer to the experiences of shamans in ASCs. While there are no direct references to trance dances or other hallucination techniques in the early ethnographic accounts, shamanists have buttressed their arguments through references to more recent ethnographic evidence from the Kalahari, where the experiences of shamans in trance dance–induced ASCs have been securely documented in great detail. In this view, many of the surreal details of the /Xam ethnographic accounts are explained as metaphors for the hallucinatory experiences of shamans in ASCs. Hence, Lewis-Williams (1983b) began speaking of a “pan-San cognitive system” linking modern Kalahari San groups, such as the Ju/’hoansi, with the /Xam and other historically known southern San groups, and also the prehistoric San who produced rock art across the subcontinent. On the other hand, Solomon (1997, 2006, 2013) has criticized the shamanism interpretive framework on the grounds that there are no explicit descriptions of ecstatic shamanism in the /Xam and other southern San texts. Instead, she argues, these texts are better understood as stories about supernatural agents that take place outside of everyday human historical time, such as spirits, gods, and dead ancestors (see also Vinnicombe 1976). This impasse has been rather difficult to resolve for a number of reasons. For one thing, it has sometimes been argued that the hallucinatory experiences of ecstatic shamans were themselves key elements in structuring the broader religious and cosmological concepts of shamanic societies (Clottes and Lewis-Williams 1998; Lewis-Williams 2002, 2010; Lewis-Williams and Pearce 2005). In this way, even the qualities and activities of spirits, gods, mythological figures, etc., in the spirit world can be seen as the result of human hallucinations during shamanic rituals. For another thing, there is just no definitive way of disambiguating the early ethnographic texts describing southern San beliefs, which would be more directly historically relevant to rock art production. Furthermore, while modern Kalahari San groups have no apparent historical connection with southern African rock art traditions, our knowledge of their trance dance and other shamanic healing rituals is much clearer.

PUTTING SOUTHERN AFRICAN ROCK PAINTINGS IN CONTEXT

77

The subject of this chapter—rainmaking—exemplifies some of the ambiguities involved in using San ethnographic texts to interpret the meaning of rock paintings. Rainmaking has often been proposed both as a frequent focus of San religious rituals and as a common subject of southern African rock paintings. One commonly discussed variant of San rainmaking ritual involves the attraction of a “Rain Bull,” which is the spirit-world embodiment of rainstorms, with anatomical features mimicking the clouds and rain shafts (Challis, Hollman, and McGranaghan 2013; Challis, Mitchell, and Orton 2008; Deacon 1988; Dowson 1998; Lewis-Williams 1977, 1980, 1981a, 1981b; Lewis-Williams and Pearce 2004; Ouzman 1998; Solomon 1992, 2013, 2016; Sullivan and Low 2014). Religious beliefs surrounding the Rain Bull complex are described at length in the /Xam and other southern San ethnographies, and they would seem to constitute a key set of ritual concerns for San societies (Bleek 1924, 1933a, 1933b, 1935, 1936; Bleek and Lloyd 1911; Orpen 1874; see Schmidt 1979 for a review). These ethnographic descriptions of Rain Bull religious beliefs have often been applied to certain varieties of southern African rock paintings. In fact, Rain Bull imagery is exceptional in having been directly identified as such by living southern San informants with active connections with a rock painting tradition (LewisWilliams 1977; Schmidt 1979). Rain Bull–related rock paintings are often described as depicting large quadrupeds, sometimes with mixes of characteristics belonging to different animal species, being led by anthropomorphic figures holding charms. In the shamanic understanding of Rain Bull imagery, paintings are thought to represent shamans who have entered the spirit worlds through ecstatic trance rituals and who are leading the Rain Bull to desired places on the real-world landscape. Yet, as Solomon (2013) has argued, the /Xam and other southern San texts make no direct mention of trance dances or other forms of ecstatic experience in relation to the Rain Bull myths and rainmaking ritual practices. Instead, Solomon favors a scenario in which rainmakers would make appeals to helpful spirits of the dead and the like, who would then engage with the Rain Bull in the spirit world. Among the Ju/’hoansi, while there is incontrovertible evidence of ecstatic shamanic religious ritual, none of it apparently relates to rainmaking. As Marshall (1957, 1999) has noted, the Ju/’hoansi do not actually have very much by way of formal rainmaking or weather-controlling ritual, nor do they believe in the Rain Bull complex per se (see also Biesele 1978, 1993; Katz 1982; Katz, Biesele, and St. Denis 1997). For the Ju/’hoansi, the activities of shamans during trance dance rituals is purely limited to healing and not anything having to do with controlling the weather. Marshall (1999) states that the Nyae Ju/’hoansi no longer believe that shamans have the power to predict or control rain, but she points to one song that she thinks might be a vestige of earlier rainmaking ritual practices. Hence, as Solomon (2013) has suggested, while there is no ethnographic evidence linking the Rain Bull belief complex with ecstatic shamanism, there is good reason to doubt the universality of Rain Bull beliefs as an element of a putative “pan-San cognitive system” (sensu Lewis-Williams 1983b). With that said, research on Ju/’hoansi religious systems has offered some insight on the sources of the supernatural linkage between large game animals and rain control (Biesele 1978, 1993; Katz 1982; Katz, Biesele, and St. Denis 1997). As Biesele (1978) describes, there is a broad psychic connection between weather conditions, the productivity of the environment, and the social-psychological well-being of the group—with the presence of large game animals acting as an important component and a potential instrument of influence. Biesele (1978: 928–929) introduces the Ju/’hoansi spiritual concept of “n!au” in the following way: . . . n!au . . . , a complex of ideas relating atmospheric conditions, hunting, childbirth, and the great meat animals, is built on an opposition between the desirable cool, rainy weather and that season of the year when great heat, dryness, and nighttime cold conspire to make the !Kung hungry and uncomfortable. . . . Life and death, especially in a sparse environment like the Kalahari, are closely bound up with the vagaries of the weather. The !Kung have a set of beliefs which are both a way to comment on the weather and the foundation for certain attempts to influence it. In the n!au beliefs, men are thought to cause weather changes by their interactions with the great meat animals they kill.

As Biesele goes on to describe, n!au is also wrapped up in issues of gender and reproduction, though these subjects are beyond the scope of this chapter (see Solomon 1992, 2008 for related discussions of rock art imagery). More importantly, the n!au concept of the Ju/’hoansi has parallels in the religious cosmology of the

78

GRANT S. MCCALL, THEODORE P. MARKS, JORDAN WILSON, ANDREW G. SCHROLL, AND JAMES G. ENLOE

southern San, yet there are major differences in the ideas and ritual approaches to rainmaking held within these traditions. Hence, while Rain Bull imagery has been fairly securely identified in southern African rock paintings, there is considerable controversy in terms of what it all means. It could be the case, as Lewis-Williams (1977, 1980, 1981a, 1981b) and others (Challis, Hollman, and McGranaghan 2013; Challis, Mitchell, and Orton 2008; Dowson 1998; Lewis-Williams and Pearce 2004; Ouzman 1998) suggest, that Rain Bull imagery represents the hallucinatory experiences of shamans in ecstatic trances. It could also be the case that Rain Bull imagery refers to a complex of religious and cosmological beliefs about the relationship between mythological deities and spirits of the dead, as Solomon (1992, 1997, 2013) argues. Or it could be the case that the complex of Rain Bull religious beliefs is derived from the repeated experiences of shamans in ecstatic trances, which gradually became incorporated into a wider understanding of cosmology over time. And finally, it could be that Rain Bull rock painting imagery referred to different aspects of this ritual and belief system simultaneously, and that it was understood differently under varying circumstances. Even in this rare instance where we actually have a good idea about what the referent of a certain set of rock art imagery was, it still goes without saying that there is great ambiguity and potential multivalence in the understandings of rock painting imagery.

Scenes, Non-Scenes, and Ritual Context in Southern African Rock Art As the previous section demonstrates, southern African rock art researchers have spent a lot of time thinking about the meaning of rock art imagery. Much less attention has been paid to making inferences about the social and ritual context of the act of rock art production within prehistoric San societies. In this respect, the distinction at the heart of this book between “scenes” and “non-scenes” is a useful one. Very briefly, we follow Davidson (this volume) in defining a scene as “a set of images in spatial proximity to each other from which, without any knowledge other than the images themselves, an observer can infer actions taking place among the actors represented in the images.” A non-scene, therefore, is a set of images that lack this sort of agential relationship with one another. The issue of scenes in southern African rock paintings has been considered previously (Lewis-Williams and Pearce 2009), though not very much. The corpus of southern African rock paintings has long been noted for its tendency to occur in scenes as defined above. Thus, southern African San rock paintings have been variously referred to as “narrative” (e.g., Lewis-Williams 1982), “historical” (Willcox 1963), “realistic” (Campbell 1986), and so forth (see Vinnicombe 1972 for a review). This tendency can also be contrasted with certain other major prehistoric rock art traditions, such as the Upper Paleolithic cave paintings of Europe (Davidson, this volume), and even the parallel tradition of rock engraving in southern Africa (see Lenssen-Erz et al., this volume). There is no doubt that this scenic tendency of southern African rock paintings has influenced views of their function and social context. At the same time that Leroi-Gourhan (1967) was comparing Upper Paleolithic cave paintings to the iconography of churches, proponents of the art-for-art’s-sake position were discussing southern African rock paintings as primitive anecdotes of daily life (Cooke 1969; Willcox 1963; Lee and Woodhouse 1970). Perhaps owing to some such baggage from the art-for-art’s-sake era, later interpretive traditions have struggled with the question of why rock paintings were produced in the first place, as opposed to what they meant. While they are not nearly alone in this respect, the advocates of the shamanism perspective have always been somewhat vague about the contexts and purposes of San rock art. Yet when they have been explicit, they have tended to support a narrative function for rock paintings. One of the clearest expositions on this matter comes from one of Lewis-Williams’s (1982: 435) early contributions, in which he states: Sharing [emphasis added] is . . . an important part of the trance experience. During a curing dance !Kung men in trance sometimes direct their fellows’ attention to the “visions” they believe they can see. The combined apprehension of these visions enables them to unite their individual powers in the performance of their work. . . . The depiction in the rock shelters of such visions could have made for wider and more vivid sharing and so constituted a beneficial pooling of both experience and power.

PUTTING SOUTHERN AFRICAN ROCK PAINTINGS IN CONTEXT

79

In this view, rock art is seen as a tool for sharing the experiences of shamans during ecstatic trances.1 Hence, the “scene-iness” of San paintings is directly related to what we might call the sharing model of the shamanic rock art function. It is certainly not the case, however, that all southern African rock paintings are composed into scenes or that all rock art panels make coherent narrative sense; in fact, quite the contrary. Southern African rock painting panels tend to be formed by the accumulation of both scenic and non-scenic elements and various representational effects. While there are other dynamics at work, one phenomenon is particularly responsible for the non-scenic qualities of southern African rock painting panels: the superposition and juxtaposition of painting elements. As McCall (2007) has suggested, the results of these processes are palimpsests, in both the artistic and the archaeological senses of the term. These palimpsests are virtually never unitary scenes in their own right, though they often contain multiple different scenes within them. The superpositon of rock paintings was demonstrably structured by syntax-like ordering rules (Lewis-Williams 1972, 1974), yet it resulted in palimpsests that contained both superimposed isolated elements and scenes that do not add up to a single total scene in any way that we can detect. In a recent paper that is unique in explicitly considering the issue of scenes in southern African rock paintings, Lewis-Williams and Pearce (2009: 55) doggedly stand by the sharing model in dealing with the problems of superposition and non-scenic painting elements, stating: Neuropsychological studies of visual hallucinations have compared this experience to a slide or film show. . . . The overall effect created by the superimposition of an image over another thus replicated the shaman-artists’ own religious experience and way of seeing images projected onto surfaces.

We frankly doubt this explanation of the superposition of scenes and painting elements and think that it exposes an evident shortcoming inherent to the sharing model. In this scenario, even apparent non-scenes are still viewed as scenes, since they still convey the actual but hard-to-explain experiences of shamans in ASCs. We find it difficult to believe that all palimpsest accumulations of southern African rock paintings were composed for this particular narrative goal regarding the hallucinatory experiences of shamans. More importantly, it forecloses on the opportunity to learn about the social and ritual contexts of rock art, which may underlie the composition of non-scenic panels (Ross and Davidson 2006). In our view, the combination of scenes and non-scenes in the formation of rock art palimpsests amounts to a rarely acknowledged aspect of complexity in the southern African rock painting tradition. Yet making sense of this complexity requires shifting our focus from the meaning of rock art imagery as conveyed by coherent scenes to the social and ritual processes whereby rock art accumulated. Above all, the shamanic sharing model seems inadequate to the complexity of the cumulative admixture of scene and non-scene rock painting elements that is so common in the southern African San tradition. In this chapter, we focus on Mirabib as a case study in suggesting ways in which the ritual contexts of rock painting production may shed some light on this complexity while also pointing to alternative ways in which we might use rock paintings as a source of information about Holocene hunter-gatherer lifeways in southern Africa.

Mirabib’s Rock Art and Archaeology Mirabib is widely acknowledged as an important archaeological site, but this fact is not primarily due to the rock art present there. The rock paintings are limited to an exposed fracture plane of a single roof fall granite boulder (Figures 5.1 and 5.2) that is approximately 2.5m high and 3.5m wide. With the naked eye, we were able to identify about two dozen individual rock art elements, mostly anthropomorphic figures and quadrupeds. Using D-stretch© digital image enhancement software (Harman 2005), we are able to identify about another two dozen rock art elements.This image enhancement software was thus instrumental in making sense of this rock painting panel. The rock paintings at Mirabib are typical in being composed of a mix of scenes and non-scenes. The scenes present consist of five main concentrations of rock painting elements (Figures 5.3–5.6). The scenes have in common the representation of large ungulates, which have agential relationships with one another and

80

GRANT S. MCCALL, THEODORE P. MARKS, JORDAN WILSON, ANDREW G. SCHROLL, AND JAMES G. ENLOE

FIGURE 5.1. Plan view map of the Mirabib rockshelter. Copyright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

FIGURE 5.2. View of the painted boulder surface at Mirabib and D-Stretch enhancement of the painted rock surface. Copyright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

PUTTING SOUTHERN AFRICAN ROCK PAINTINGS IN CONTEXT

with adjacent human figures. In addition, these scenes are either depicted in relation to naturally occurring cracks in the rock face on which they are painted, or, as we will presently discuss further, they are placed in relation to driplines on the rock surface. One panel shows two large ungulate bulls facing downward and leftward, as if they were moving into a crack in the rock (Figure 5.3). These two bulls are being led by a human figure, which may be carrying a charm. A second panel also features a large human figure, this one carrying a much more evident charm, and leading a large ungulate bull, in this case likely a wildebeest (Connochaetes sp.; Figure 5.4). A third panel features two large ungulates facing toward a dripline, with some faded human figures also present. We take these three scenes as examples of Rain Bull imagery, discussed above. For the sake of brevity, suffice it to say that they are depictions of large ungulates, some of which seem to combine features of various animal species, being led by anthropomorphic figures carrying charms, and they are placed in relation to natural rock surface features, such as cracks and driplines—key characteristics that will figure prominently in our argument below. Next, a fourth panel, placed between a large crack running down the center of the rock face and a dripline, is characterized by the presence of an apparent human / bat-eared fox (Otocyon megalotis) therianthrope and two other therianthropic figures that are partially obscured by the dripline (Figure 5.5). This panel also features a smaller ungulate representation, probably a springbok (Antidorcas marsupialis), as well as a number of faded human and ungulate figures. There is also a quadruped at the top of this sub-panel that we take to be a feline, likely a leopard (Panthera pardis), which is rather unusual in the rock paintings of this region. Finally, a fifth panel is comprised of a jumble of apparent human figures, parallel vertical lines, and ungulate figures, again placed between the central crack and a dripline. All of these panels have in common some degree of superposition and/or juxtaposition, and most of them are quite crowded palimpsests of accumulated elements. In this way, Mirabib is like most southern African rock painting sites in that it combines scenes of related elements with non-scenic elements while also featuring dense superpositions of scenes and elements, resulting in palimpsests on which any clear narrative meaning inherently breaks down. Fortunately, as we will discuss below, some unusual aspects

81

FIGURE 5.3. Two quadrupeds being led by an anthropomorphic figure entering a watermark and disappearing into a crack in the rock. Copyright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

FIGURE 5.4. Large quadruped being led by an anthropomorphic figure into a watermark. Note also a therianthropic figure above the large quadruped and various anthropomorphic and zoomorphic figures below. Coypright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

82

GRANT S. MCCALL, THEODORE P. MARKS, JORDAN WILSON, ANDREW G. SCHROLL, AND JAMES G. ENLOE

of the context of these paintings may shed some light on the social and ritual processes through which these palimpsests accumulated. The broader archaeological record of the Mirabib rock shelter is a source of information that can shed some light on the social context in which the rock art was produced. Since 2013, we have been involved in a program of re-excavating Mirabib and reanalyzing the archaeological remains from earlier excavations. Our new dating assays revealed an LSA occupation of the Mirabib rock shelter dating from ~11 to 7ka, in addition to a pastoralist use of the site around 1.5ka (Marks 2018; see Sandelowsky 1977). The lithic artifacts at Mirabib broadly belong to the Wilton Industry, characterized mainly by microlithic stone tools produced primarily on hydrothermal vein quartz (Marks 2018). The bulk of the archaeological remains belong to strata dating to the last few centuries before the 7ka date, and we consider this to be a good estimate for the age of the rock paintings. If correct, this would make them a FIGURE 5.5. Complex panel with a bat-eared fox theri- bit older than the suspected age of the rock paintanthrope (center-left) and various other anthropomor- ings in the Brandberg massif (Kinahan 1991) and phic, zoomorphic, and therianthropic figures. Copyright would also place them at the early end of the known Grant S. McCall, Theodore P. Marks, Jordan Wilson, An- southern African LSA rock painting tradition time drew G. Schroll, and James G. Enloe. frame (e.g., Thackeray 1983), to the extent to which it is known. Unlike most MSA and LSA archaeological sites in the Namib and its peripheries, Mirabib lacks a perennial water source. Instead, the granite koppie on which Mirabib is located has a few shallow depressions that fill up with water in the aftermath of large rainfall events, referred to in Khoekhoegowab as “//gurab” (Hoernle 1925). This type of well is extremely important for modern pastoralists in the Namib Desert and its peripheries, and some of them are even fairly perennial—at least on historical time scales. However, the //gurab wells present on the Mirabib rock shelter koppie are apparently dry, except in the immediate aftermath of significant rainfall events. We suspect that Mirabib was primarily occupied by LSA foragers in the aftermath of significant rainfall events, when the Central Namib gravel plains experienced an explosion in plant productivity and therefore a surge in the presence of game animals. We have argued elsewhere that the ephemeral occupation of residential camps on the Central Namib gravel plains articulated with the more prevalent habitation of the Khan, Swakop, and Kuiseb River valleys, which provided perennial water sources as well as relatively rich riparian corridors of vegetation and game availability (Marks 2018). We believe there is good evidence that prehistoric populations tended to concentrate in the resource-rich river valleys throughout much of the year, occasionally (or perhaps seasonally) moving onto the gravel plains to take advantage of pulses of productivity following significant rainfall events (Marks 2018; McCall et al. 2011). Kinahan (1991) has also made the same argument for Holocene huntergatherer populations in the Brandberg. Among other things, Mirabib is famous for its phenomenal preservation of organic materials, which provide insight concerning the organization of LSA mobility patterns and settlement systems relative to the occupation of the Mirabib rock shelter. Both Sandelowksy’s and our excavations recovered items like vegetal twine, the wooden handles of composite tools, !nara melon (Acanthosicyos horridus) containers and seeds, and even animal hair. In addition, ostrich egg shell (OES) fragments were extremely abundant, with many large articulated pieces in the uppermost strata, apparently resulting from the breakage of whole eggs. Some of these whole eggs contained caches of lithics, twine, and other technological materials, and we take this phenomenon to indicate the use of a caching strategy. While we only find the durable remains of cached

PUTTING SOUTHERN AFRICAN ROCK PAINTINGS IN CONTEXT

83

items, we think the prime goal of this caching strategy was likely the caching of water, bearing in mind that Mirabib lacks a perennial water source. Caching water at Mirabib may have been a strategy for dealing with uncertainty in terms of water availability and for extending occupations in the absence of rainfall that might recharge ephemeral surface water features. This OES feature appears in one of the upper layers of the Mirabib, perhaps suggesting that it was part of a strategy that developed over time in response to the unique geographic and environmental context of Middle Holocene hunter-gatherer activities in the Central Namib Desert. On the one hand, if we are right about this feature representing water storage activities, then it underscores a concern for water availability, which articulates with the prevalence of rainmaking imagery in the Mirabib rock paintings. On the other hand, the development of a fairly specialized technological approach for coping with the hyper-arid conditions of the Central Namib Desert likely speaks to dynamics of foraging intensification among populations seeking to extend occupations at sites with limited water availability. In Australia, for example, Gould (1977) shows that modern hunter-gatherers living in the Central Desert tend to occupy sites lacking perennial water sources for as long as possible, depending mostly on ephemeral water resources and retreating to perennial water sources only when absolutely necessary (see also Thorley 2001). This foraging strategy allows hunter-gatherers to avoid the kind of resource overexploitation that is often caused by prolonged occupation of sites with perennial water availability, while also more effectively allowing for the exploitation of seasonal or sporadic pulses of environmental productivity, such those occurring in the aftermath of rainfall events. At Mirabib, we argue, huntergatherers developed a water caching strategy in order to prolong occupations away from perennial water sources located in the Kuiseb River valley as a strategy for maximizing foraging returns and, in Gould’s sense of the term, reducing the risk associated with the long-term occupation of sites with perennial water supplies.

A Rainy Day in the Central Namib Our breakthrough in thinking about the rock art at Mirabib came in the form of a freak weather event: a September storm system that swept northeastward across the Namib Desert, dropping several centimeters of rainfall in the space of two days. While rather inconvenient for those of us accustomed to living outdoors while working in the field, this rare event for the ecology and geography of the Central Namib gravel plains was illuminating. One of the unexpected things we observed was the fact that even this relatively modest rainfall event led to streams of water pouring over the dripline of the Mirabib cave, flowing onto the rock painting panel (Figure 5.6). Had we thought about it hard enough previously, we might have figured out that this would have happened by virtue of the locations of the water marks on the boulder. Yet upon seeing this phenomenon in action, it was clear that much of the rock art was positioned in a way such that it was covered by the flow of water when it rained. Figure 5.6 depicts the two quadrupeds shown in Figure 5.3, facing toward the aforementioned crack in the rock. The visual effects shown in Figure 5.6 constitute a complex metaphor and, as we will argue later, perhaps reflect the ritual context of rock art production having to do with rainmaking. Similarly, there are two other panels at Mirabib that include what we would diagnose as Rain Bull imagery and are positioned in relation to driplines that we suspect would also be covered by flowing water during heavier rainfall events. We strongly believe that these scenes were intentionally composed in order to construct a visual relationship between the flowing water and the imagery being depicted. We also suspect that the sporadic flow of water over this boulder during rainstorms explains why this rock surface was painted, while other eligible surfaces in the Mirabib rockshelter were not. Beyond the flow of water on and around the Mirabib rock paintings, this rainfall event also demonstrated that the many //gurab wells on the koppie filled with an ample supply of water. In addition, the surrounding gravel plains quickly exploded into life, first with a surge in the growth of grasses and other desert plants, and subsequently with an array of large ungulate fauna. In this way, the locality surrounding Mirabib did indeed experience a pulse of foraging productivity of the sort that was likely attractive to prehistoric foragers in the region. In a short space of time, a number of important things happened in concert with one another,

84

GRANT S. MCCALL, THEODORE P. MARKS, JORDAN WILSON, ANDREW G. SCHROLL, AND JAMES G. ENLOE

FIGURE 5.6. Quadruped elements pictured in Figure 5.3 covered by dripping water during a rainstorm. Copyright Grant S. McCall, Theodore P. Marks, Jordan Wilson, Andrew G. Schroll, and James G. Enloe.

all caused by the rainfall event that we experienced. This observation therefore provided us with some insight into the symbolism of the rock paintings at Mirabib. For us, the clear implication of the combined evidence from Mirabib is that the production of rock art related to issues having to do with rain. Without considering the content of the rock paintings for the moment, we note the following features of the site: (1) it lacks a perennial water source but has ephemeral water sources filled during rainfall events; (2) it experiences a spike in environmental productivity in terms of both

PUTTING SOUTHERN AFRICAN ROCK PAINTINGS IN CONTEXT

85

flora and fauna after rainfall events; (3) it has an expansive buried archaeological feature composed of what are likely broken water containers; and (4) it has a large accumulation of rock paintings on the only vertical surface in the rock shelter covered by dripping water during rainfall events. On top of the more general truism that water tends to be a major concern for hunter-gatherers living in hyper-arid environments like the Central Namib Desert (again, see Gould 1977 for ethnographic accounts), we conclude that much of the archaeological record in the Mirabib rock shelter is structured in relation to the anticipation of rainfall events.

Discussion The greatest implication of our experiences with the Mirabib rock paintings is that they raise problems for the shamanic sharing model of rock art production. There are a few central reasons for this, some likely obvious and others perhaps not. The manifest compositional patterning at Mirabib is complex in ways that make the sharing model unlikely. For one thing, the paintings placed in the dripline could not have been produced during the weather event that gives them their main effects. They must have been painted in some immediate temporal relationship with the rainfall that covered this portion of the rock face. Thus, even if the intent was to convey information about the sensations associated with the attraction of the Rain Bull in a shamanic trance—perhaps the feeling of being underwater—one would potentially have to wait a very long time for the next instances in which these paintings were covered by dripping water. Hence, for the sharing model to work, it would require shaman-artists to experience something having to do with Rain Bulls during an ecstatic trance, paint the scene of a human leading the Rain Bull(s), and then wait perhaps weeks, months, or years to complete the full effect when it rains. For us, this seems dubious. In contrast, we find this setup more similar to a solar alignment, such as the famous “sun dagger” at Chaco Canyon in the Southwestern US, where sunlight illuminates a spiral rock engraving on the summer solstice (Newman, Mark, and Vivian 1982). In that case, the full significance of the rock engraving occurs in the brief time when it is illuminated by the solstice sunlight, emphasizing the religious and ritual significance of that time and place. At Mirabib, a comparable situation occurs when it rains, covering certain of the rock paintings with flowing water and emphasizing the religious and ritual significance of that event in relation to the rock paintings. In contrast, and though this is obviously quite speculative, we might offer the following scenario as a contrast: In the Kalahari, we know that Ju/’hoansi beliefs surrounding the n!au concept stipulate that rainy weather can be induced by certain kinds of interactions with large game animals. We also recognize the strong formal similarities between certain rock painting scenes at Mirabib and the Rain Bull mythological complex of the /Xam and other southern San groups, as others have long noted elsewhere. Finally, we also know that in the Central Namib, rainfall events are tightly linked with the arrival of herds of large game animals. Putting these lines of thinking together, we wonder if rainmaking ritual specialists intentionally placed images of large game animals on portions of the rock face known to be covered by water during rainfall events, since large game animals were thought to attract storms, as well as the desirable social and emotional states with which rainy weather is associated within the n!au concept. Furthermore, blood, and in particular menstrual blood, is also believed to attract rainfall, according to the n!au belief system. Along these lines, Lewis-Williams and Dowson (1990) have noted historical references to the use of the blood from large game animals as a binding agent in San rock painting pigments and suggested that this enhanced the supernatural potency of the resulting images. It could also be that blood-based paints were used at Mirabib to attract flowing water and the rainfall that produced it. Certain lines of ethnohistoric research in places such as the Drakensberg/Maluti highlands of southeastern South Africa point to the profound diversity of rainmaking beliefs and practices among the San (Jolly 1986; Prins 1990). Though it is beyond the scope of this chapter to delve too deeply into this literature, we note that San sorcerers were widely respected for their power, including by neighboring farming societies. We also know that sorcerers sometimes engaged in ritual rainmaking activities in rock shelters, that some of them also painted in rock shelters, and that they may have painted as an aspect of rainmaking sorcery. Finally, Prins (1990) has demonstrated that sympathetic magic was a key aspect of these rainmaking techniques. Beyond pointing to the diversity of San ritual practices and spiritual beliefs that may have been asso-

86

GRANT S. MCCALL, THEODORE P. MARKS, JORDAN WILSON, ANDREW G. SCHROLL, AND JAMES G. ENLOE

ciated with rock art production, these accounts support our scenario of the production of rock paintings at Mirabib as an instrumental part of rainmaking ritual. Finally, some more conventional characteristics of the rock paintings at Mirabib warrant some consideration relative to the issue of ritual practice. Above all, these have to do with non-scenic rock painting compositions, discussed above, and especially non-scenic juxtapositions and superpositions. These phenomena are common at Mirabib, and once again, they raise problems for the shamanic sharing model. However, if we consider the possibility that the production of rock art itself was part of a ritual activity aimed at influencing real-world conditions, then repeated overpainting of rock surfaces in non-scenic ways takes on a different set of implications. In this respect, Ross and Davidson (2006) have emphasized the importance of the repetition of activities in specialized places and times as key aspects of rock art related to ritual performance. This repetition, they argue, leads to the accumulation of thematically consistent rock art imagery concentrated in particular spaces, often resulting in non-scenic juxtaposition and superposition. We feel that this insightful perspective holds great potential for making sense of the often confusing non-scenic superimposed imagery at southern African rock art sites, including Mirabib.

Final Thoughts In general, we think it is important to recognize that southern African rock art represents an incredibly varied set of phenomena likely relating to a wide range of social, ritual, and artistic concerns. Ignoring for now the categorical differences between the rock painting and rock engraving traditions (see Lessen-Erz et al., this volume), we can recognize, at a minimum, the following: (1) paintings of scenes that seem to refer to supernatural phenomena; (2) paintings of scenes that seem to refer to real-world historical events; (3) paintings of scenes that may have been instrumental elements of rituals aimed at influencing real-world conditions; (4) non-scene elements that may have resulted from the repetition of ritual activities in specialized places. As we have shown in this chapter, the sharing model is not adequate to the task of accommodating this diversity or explaining variability in the archaeological record of rock art in southern Africa, even though it may sometimes be correct in certain kinds of situations. Furthermore, by focusing our criticism on the sharing model, we do not necessarily mean to single out the shamanism framework of rock art interpretation. Very few researchers have been very explicit or thorough in considering how the production of rock art would have fit into the social behavior of prehistoric San societies, as opposed to what it all meant. In some senses, we seem to have failed to move on from the “founder’s effect” of the art-for-art’s-sake perspective, which viewed the content and meaning of rock art as more or less independent from the context in which it was produced. Further, it has not helped the situation that our best sources of interpretive information come in the form of narratives recorded by early ethnographers in southern Africa. In moving forward, we argue, we must develop an analytical framework for making inferences about the social and ritual contexts in which rock art was produced, rather than focusing exclusively on the issue of meaning. Our work at Mirabib speaks to the lifeways of Middle Holocene foragers in the Central Namib who developed sophisticated strategies for prolonging the economic uses of sites on the gravel plains in the absence of perennial water sources, and who apparently focused a great deal of ritual attention on rainmaking. There may be disagreement about what social functions these forms of ritual behavior played in the unique adaptation of LSA foragers in the Namib. For us, however, there is little question that these rainmaking activities occurred as a result of the rather extreme organizational adaptions to aridity seen in this region. For those who engaged in them, rainmaking rituals were every bit as real a strategy for coping with water scarcity as the material technologies of water storage that we have also discovered were. In this sense we feel that, like Ross and Davidson (2006), southern African rock art has much to offer our understanding of ritual activities of prehistoric San societies. Part of this shift in focus has to do with changes to our approach to the analysis of rock art panel content and composition, such as the characterization of scenes and non-scenes at the heart of this volume. Part of it also has to do with the development of a more holistic and integrative analytical framework for using rock art as a source of evidence with which to address our questions about the social and ritual lives of prehistoric hunter-gatherers.

PUTTING SOUTHERN AFRICAN ROCK PAINTINGS IN CONTEXT

87

Grant McCall received his Ph.D. in Anthropology from the University of Iowa and is an Associate Professor in the Department of Anthropology at Tulane University. He is also the Executive Director of the Center for Human-Environmental Research (CHER). Theodore Marks received his Ph.D. in Anthropology from the University of Iowa and is an instructor at the New Orleans Center for Creative Arts (NOCCA). Andrew Schroll is a Ph.D. student in the Department of Anthropology at Tulane University. James Enloe received his Ph.D in Anthropology from the University of New Mexico and is a Professor in the Department of Anthropology at the University of Iowa. Jordan Wilson is a Ph.D. student in the Department of Anthropology at the University of Arizona

Note 1. Lewis-Williams and Dowson (1990) later offer a somewhat different account of rock painting context, arguing that the painted rock surface was like a “veil” between the real world and spirit world, and that shamans painted rock surfaces as a way of making a kind of window onto the spirit world.

References Bahn, P.G. 1997. “Membrane and Numb Brain: A Close Look at a Recent Claim for Shamanism in Palaeolithic Art.” Rock Art Research 14: 62–67. Biesele, M. 1978. “Sapience and Scarce Resources: Communication Systems of the !Kung and Other Foragers.” Information 17(6): 921–47. ———. 1993. Women like Meat: The Folklore and Foraging Ideology of the Kalahari Ju/’hoan. Bloomington: Indiana University Press. Bleek, D. F. 1924. The Mantis and His Friends. Cape Town: Maskew Millar. ———. 1933a. “Beliefs and Customs of the /Xam Bushmen, Part 5: The Rain.” Bantu Studies 7: 297–312. ———. 1933b. “Beliefs and Customs of the /Xam Bushmen, Part 6: Rain-Making.” Bantu Studies 7: 375–92. ———. 1935. “Beliefs and Customs of the /Xam Bushmen, Part 7: Sorcerors.” Bantu Studies 9: 1–47. ———. 1936. “Beliefs and Customs of the /Xam Bushmen, Part 8: More on Charms and Sorcerors.” Bantu Studies 10: 131–62. Bleek, W. H. I., and L. C. Lloyd. 1911. Specimens of Bushman Folklore. London: George Allen. Blundell, G. 1998. “On Neuropsychology in Southern African Rock Art Research.” Anthropology of Consciousness 9(1): 3–12. Campbell, C. 1986. “Images of War: A Problem in San Rock Art Research.” World Archaeology 18(2): 255–68. Challis, S., J. Hollmann, and M. McGranaghan. 2013. “‘Rain Snakes’ from the Senqu River: New Light on Qing’s Commentary on San Rock Art from Sehonghong, Lesotho.” Azania: Archaeological Research in Africa 48(3): 331–54. Challis, S., P. Mitchell, and J. Orton. 2008. “Fishing in the Rain: Control of Rain-Making and Aquatic Resources at A Previously Undescribed Rock Art Site in Highland Lesotho.” Journal of African Archaeology 6(2): 203–18. Clottes, J., and J. D. Lewis-Williams. 1998. The Shamans of Prehistory: Trance and Magic in the Painted Caves. New York: Harry N. Abrams. Cooke, C. K. 1969. The Rock Art of Southern Africa. Cape Town: Books of Africa. Deacon, J. 1988. “The Power of a Place in Understanding Southern San Rock Engravings.” World Archaeology 20(1): 129–40. Dowson, T. A. 1994. “Reading Art, Writing History: Rock Art and Social Change In Southern Africa.” World Archaeology 25(3): 332–45. ———. 1998. “Rain in Bushman Belief, Politics and History: The Rock Art of Rain-Making in the South-Eastern Mountains, Southern Africa.” In The Archaeology of Rock-Art, ed. C. Chippindale and P. S. C. Tacon, 73–89. Cambridge: Cambridge University Press. ———. 2007. “Debating Shamanism in Southern African Rock Art: Time to Move on. . .” South African Archaeological Bulletin 62(185): 49–61. Eliade, M. 1964. Shamanism: Archaic Techniques of Ecstasy, trans. W. R. Trask. New York: Pantheon Books. Gould, R. A. 1977. Puntutjarpa Rockshelter and the Australian Desert Culture. New York: Anthropological Papers of the American Museum of Natural History.

88

GRANT S. MCCALL, THEODORE P. MARKS, JORDAN WILSON, ANDREW G. SCHROLL, AND JAMES G. ENLOE

Guenther, M. G. 1975. “The Trance Dancer as an Agent of Social Change among the Farm Bushmen of the Ghanzi District.” Botswana Notes & Records 7(1): 161–66. Harman, J. 2005. “Using Decorrelation Stretch to Enhance Rock Art Images.” American Rock Art Research Association Meeting, 27–29 May 2005. Sparks, Nevada. Heinz, H. J. 1975. “Elements of !Ko Bushmen Religious Beliefs.” Anthropos 70(1/2): 17–41. Hoernle, A. W. 1925. “The Social Organization of the Nama Hottentots of Southwest Africa.” American Anthropologist 27(1): 1–24. Jolly, P. 1986. “A First Generation Descendant of the Transkei San.” South African Archaeological Bulletin 41(143): 6–9. Katz, R. 1982. “Accepting ‘Boiling Energy.’” Ethos 10(4): 344–68. ———. 1985. Boiling Energy: Community Healing among the Kalahari !Kung. Cambridge, MA: Harvard University Press. Katz, R., Biesele, M., and V. St. Denis. 1997. Healing Makes Our Hearts Happy: Spirituality and Cultural Transformation among the Kalahari Ju/’hoansi. Rochester, VT: Inner Traditions/Bear & Co. Kinahan, J. 1991. Pastoral Nomads of the Central Namib Desert: The People History Forgot. Windhoek: Namibia Archaeological Trust. Lee, D. N., and H. C. Woodhouse. 1970. Art on the Rocks of Southern Africa. Cape Town: Purnell. Leroi-Gourhan, A. 1967. Treasures of Prehistoric Art. New York: H.N. Abrams. Lewis, I. M. 1971. Ecstatic Religion: An Anthropological Study of Spirit Possession and Shamanism. New York: Penguin. Lewis-Williams, J. D. 1972. “The Syntax and Function of the Giant’s Castle Rock-Paintings. South African Archaeological Bulletin 27(105/106): 49–65. ———. 1974. Superpositioning in a Sample of Rock-Paintings from the Barkly East District. South African Archaeological Bulletin 29(115/116): 93–103. ———. 1977. “Led by the Nose: Observations on the Supposed Use of Southern San rock art in Rain-Making Rituals.” African Studies 36(2): 155–60. ———. 1980. “Ethnography and Iconography: Aspects of Southern San Thought and Art.” Man 15(3): 467–82. ———. 1981a. Believing and Seeing: Symbolic Meanings in Southern San Rock Paintings. London: Academic Press. ———. 1981b. “The Thin Red Line: Southern San Notions and Rock Paintings of Supernatural Potency.” South African Archaeological Bulletin 36(133): 5–13. ———. 1982. “The Economic and Social Context of Southern San Rock Art.” Current Anthropology 23(4): 429–49. ———. 1983a. The Rock Art of Southern Africa. Cambridge: Cambridge University Press. ———. 1983b. “Introductory Essay: Science and Rock Art.” Goodwin Series 4: 2–13. ———. 2002. A Cosmos in Stone: Interpreting Religion and Society through Rock Art. Walnut Creek, CA: Rowman Altamira. ———. 2010. “The Imagistic Web of San Myth, Art and Landscape.” Southern African Humanities 22(1): 1–18. ———. 2013. San Rock Art. Athens: Ohio University. Lewis-Williams, J. D., and T. Dowson. 1988. “The Signs of All Times: Entoptic Phenomena in Upper Palaeolithic Art.” Current Anthropology 29(2): 201–45. ———. 1990. “Through the Veil: San Rock Paintings and the Rock Face.” South African Archaeological Bulletin 45(151): 5–16. Lewis-Williams, J. D., and D. G. Pearce. 2004. “Southern African San Rock Painting as Social Intervention: A Study of Rain-Control Images.” African Archaeological Review 21(4): 199–228. ———. 2005. Inside the Neolithic Mind: Consciousness, Cosmos, and the Realm of the Gods. London: Thames & Hudson. ———. 2009. “Constructing Spiritual Panoramas: Order and Chaos in Southern African San Rock Art Panels.” Southern African Humanities 21(1): 41–61. ———. 2012. “The Southern San and the Trance Dance: A Pivotal Debate in the Interpretation of San Rock Paintings.” Antiquity 86(333): 696–706. ———. 2015. “San Rock Art: Evidence and Argument.” Antiquity 89(345): 732–39. Lommel, A. 1966. Shamanism: The Beginnings of Art. New York: McGraw-Hill. Marks, T. P. 2018. “Bedtime for the Middle Stone Age: Land use, Strategic Foraging, and Lithic Technology at the End of the Pleistocene in the Namib Desert.” Ph.D. dissertation. Iowa City: University of Iowa. Marshall, L. 1957. “N!ow.” Africa 27(3): 232–40. ———. 1962. “!Kung Bushman Religious Beliefs.” Africa 32(3): 221–52. ———. L. 1999. Nyae Nyae !Kung Beliefs and Rites. Cambridge, MA: Peabody Museum Press. McCall, G. S. 2007. “Add Shamans and Stir? A Critical Review of the Shamanism Model of Forager Rock Art Production.” Journal of Anthropological Archaeology 26(2): 224–33. McCall, G. S., T. P. Marks, J. T. Thomas, M. S. Eller, S. W. Horn, R. A. Horowitz, K. Kettler, and R. Taylor-Perryman. 2011. “Erb Tanks: A Middle and Later Stone Age Rockshelter in the Central Namib Desert, Western Namibia.” PaleoAnthropology 2011: 398–421. Newman, E. B., R. K. Mark, and R. G. Vivian. 1982. “Anasazi Solar Marker: The Use of a Natural Rockfall.” Science 217(4564): 1036–38. Orpen, J. M. 1874. “A Glimpse into the Mythology of the Maluti Bushmen.” Cape Monthly Magazine 9: 1–13.

PUTTING SOUTHERN AFRICAN ROCK PAINTINGS IN CONTEXT

89

Ouzman, S. 1998. “Towards a Mindscape of Landscape: Rock-Art as Expression of World-Understanding.” In The Archaeology of Rock-Art, ed. C. Chippindale and P. S. C. Tacon, 30–41. Cambridge: Cambridge University Press. Prins, F. E. 1990. “Southern Bushmen Descendants in the Transkei: Rock Art and Rainmaking.” South African Journal of Ethnology 13(3): 110–16. Ross, J., and I. Davidson. 2006. “Rock Art And Ritual: An Archaeological Analysis of Rock Art in Arid Central Australia.” Journal of Archaeological Method and Theory 13(4): 304–40. Sandelowsky, B. H. 1977. “Mirabib: An Archaeological Study in the Namib.” Madoqua 10(4): 221–53. Schmidt, S. 1979. “The Rain Bull of the South African Bushmen.” African Studies 38(2): 201–24. Solomon, A. 1992. “Gender, Representation, and Power in San Ethnography and Rock Art.” Journal of Anthropological Archaeology 11(4): 291–329. ———. 1997. “The Myth of Ritual Origins? Ethnography, Mythology and Interpretation of San Rock Art.” South African Archaeological Bulletin 52(16): 3–13. ———. 2006. “San ‘Spirituality’ and Human Evolution: Eight Questions for Lewis-Williams and Pearce.” South African Archaeological Bulletin 61(184): 209–12. ———. 2008. “Myths, Making, and Consciousness: Differences and Dynamics in San Rock Arts.” Current Anthropology 49(1): 59–86. ———. 2013. “The Death of Trance: Recent Perspectives on San Ethnographies and Rock Arts.” Antiquity 87(338): 1208–13. ———. 2016. “Re-viewing the Sehonghong Rainmakers: Visual Interpretations and Copies of a Key South African Rock Art Motif.” South African Archaeological Bulletin 71(203): 27–35. Sullivan, S., and C. Low. 2014. “Shades of the Rainbow Serpent? A KhoeSan Animal between Myth and Landscape in Southern Africa: Ethnographic Contextualisations of Rock Art Representations.” Arts 3(2): 215–44. Thackeray, A. I. 1983. “Dating the Rock Art of Southern Africa.” Goodwin Series 4(1): 21–26. Thorley, P. 2001. “Uncertain Supplies: Water Availability and Regional Archaeological Structure in the Palmer River Catchment, Central Australia.” Archaeology in Oceania 36(1): 1–14. Vinnicombe, P. 1972. “Myth, Motive, and Selection in Southern African Rock Art.” Africa 42(3): 192–204. ———. 1976. People of the Eland: Rock Paintings of the Drakensberg Bushmen as a Reflection of Their Life and Thought. Pietermaritzburg, South Africa: University of Natal Press. Wiessner, P. 2002. “Hunting, Healing, and Hxaro Exchange: A Long-Term Perspective on !Kung (Ju/’hoansi) Large-Game Hunting.” Evolution and Human Behavior 23(6): 407–36. Wiessner, P., and F. T. Larson. 1979. “‘Mother! Sing Loudly for Me!’: The Annotated Dialogue of a Basarwa Healer in Trance.” Botswana Notes & Records 11(1): 25–31. Willcox, A. R. 1963. The Rock Art of South Africa. Johannesburg: Thomas Nelson and Sons. Yates, R., J. Golson., and M. Hall. 1985. “Trance Performance: The Rock Art of Boontjieskloof and Sevilla.” South African Archaeological Bulletin 40(142): 70–80.

==== •

•

•

6. SCENIC NARRATIVES OF HUMANS AND ANIMALS IN NAMIBIAN ROCK ART A Methodological Restart with Data Mining Tilman Lenssen-Erz, Eymard Fäder, Oliver Vogels, and Brigitte Mathiak

Introduction In the approach we present here we combine information theory and statistics with archaeology and linguistic concepts. The application of such a complex methodology requires a rich and sound database that we believe is available in the rock paintings of the Brandberg-Dâureb (Namibia) which seem to us particularly suited for deciphering meaning in rock art without having to rely on a narrowly presupposed cultural context first. The basis for our case study is the rich online Brandberg-Dâureb Database on rock art.1 These rock art data have been annotated using a textual methodology and include all rock paintings in an area of c. 300 km2 on this inselberg. This methodology implies that for purposes of data processing, every depiction has been transcribed into a formalized phrase capturing a predictable amount of information irrespective of what the figure represents (Lenssen-Erz 1994). In this mountain area there are 651 sites with more than 39,000 single figures, of which 17,416 figures have been organized in 4,842 scenes based on the textual methodology. Every single one of these pictures has been conventionally published in books at the same scale together with rich tabulated data (see online Brandberg-Dâureb Database) for each individual figure, scene, and site (Pager 1989–2006) (Figure 6.1). This database is unique in terms of both quantity and the approach, which aimed at data processing from the onset (Lenssen-Erz 1994) and therefore needed little pre-processing to allow for data mining. This chapter aims to make a general contribution to the understanding of scenes in rock art by defining testable criteria for their identification and for the information structure within scenes. This conceptualization of scenes enables us to make quantifications of scenes across the entire body of art, not only with regard to formal features but also on a semantic level. In view of the sheer mass of individual depictions, it is impossible for a human observer to discern internal structures or patterns, which is why we make use of data mining in our analyses. Regarding scenes, this analytical process is in its initial state. Here we present results from work in progress.

Scene Definitions in Rock Art Research In most rock art traditions worldwide and particularly in the representational rock art of southern Africa, the visual input at one moment, i.e., a scene, depicts the actors and not the material environmental features. Concentration on activity therefore seems most expedient for defining scenes. However, only a few sources in the literature actually clearly spell out a scenic concept (see below). Meanwhile, occasional awareness of the problem may result in a rather opaque, circular definition. For example, Smits (1971: 17) defines intentional scenes as “true groupings, depicting a scene that is originally conceived and purposely composed as such by the artist, and consisting of a number of paintings that really ‘belong together.’” Other authors provide a variety of definitions. Following the structuralist school of rock art research (Leroi-Gourhan 1983; Raphael 1945), Anati concludes that a scene made by archaic hunters may be composed of ideograms that indicate the verb, the adjective and the predicate, besides other attributes (Anati

SCENIC NARRATIVES OF HUMANS AND ANIMALS IN NAMIBIAN ROCK ART

91

FIGURE 6.1. Reproduction of rock paintings from Circus Gorge, Brandberg-Daureb (Pager 2006: 358–89) displaying a fair number of typical activities and interactions. The small numbers next to the figures are the reference numbers for retrieval of the figures in the catalogue. Colors: red, brown, black, ochre and white; size of largest human at right (no. 183) 26 cm. Copyright Heinrich-Barth-Institut.

1991: 67) leaving unresolved why there are a verb and a predicate but no subject. In contrast here is the short definition of “scene” in rock art offered by Bednarik (2003: 17): “Scene: a presumed depiction of a real or imagined episode involving more than one rock art motif.” A more precise definition is provided by Fritz et al. (2013: 40): “Une scène rupestre est une structure significative combinant des unités de moindre complexité que l’on nomme «motifs». Elle comporte un «thème», un «centre d’intérêt ou focus» vers lequel l’activité dominante est dirigée et un «cadre» qui fixe le contexte de l’action.”2 For Saharan rock art in particular, Holl (1995: 9) maintains that scenes depict sequences of actions and thus produce a localized narrative, in his words a “maximal theme.” The combination of various maximal themes constitutes a large narrative. In contrast to these arguably practical approaches used for scene determination, we suggest a formal method focusing on the interaction between figures and their formal Gestalt coherence (cf. Lenssen-Erz 1992, 1999, 2001). But whereas a scene can also be constituted even if none or few of these criteria apply, interaction between participants indisputably binds them together. Such interaction can be expounded in words, i.e., it can be narrated. A clearly narrative paradigm is also favored by May and Domingo Sanz (2010), who define a narrative “scene” in Australian rock art as involving two or more figures performing a joint, dynamic action through which they likely provide information about social and cultural practices (May and Domingo Sanz 2010:

92

TILMAN LENSSEN-ERZ, EYMARD FÄDER, OLIVER VOGELS, AND BRIGITTE MATHIAK

36). They also stress the need for homogeneity of motifs, including form, size, and technique, suggesting a scene is the result of a single artistic episode. They define a scene as follows: “A scene reflects an action, usually with a defined time, that can be described even if the meaning and theme are unknown” (May and Domingo Sanz 2010: 37). In his definition of a scene, Dobrez (2011a, 2011b) emphasizes the active character of perception, thus going beyond a concept like coherence. He understands a scene as principally linked to activity and consequently already capable of being constituted by a single figure, thus contradicting other authors cited here who maintain that a scene requires two or more actors. Moreover, Dobrez’s approach is also a rather theoretical deliberation that does not translate into a tool that other researchers can copy or adapt and implement for their research. The broadest consensus about characteristics of a scene points to a scene’s narrative potential as a general semantic quality (Dobrez 2011b; Fritz et al. 2013; Kechagia 1996; Lenssen-Erz 1992, 1999, 2001; May and Domingo Sanz 2010). Narrativity (Porter Abbott 2014) occurs when a depiction can be described in terms of an activity taking place. While this would seem to be comparatively straightforward, it does not help in defining the limits of a scene, particularly on panels with many figures. This “belonging together” is sometimes labeled “coherence” (Fritz et al. 2013; Lenssen-Erz 1992, 2001; Lewis-Williams 1992; May and Domingo Sanz 2010). Though so far only regarding the Brandberg-Dâureb rock art, these theoretical considerations have been developed into a formal method.

Key Terms in the Conceptualization of “Scenes” Narrativity If we are to consider rock art a vehicle for narration (e.g., Cole 2011; Dobrez 2011b; Fritz et al. 2013), we need to identify the means by which a pictorial narrative can be produced. In the words of Porr and Matthews (2016: 251) “a narrative can be understood as a rhetorical or communicative form that relates a number of events to each other within a temporal framework, in the sense of a succession of events.” Following definitions of narrativity in verbal communication (e.g., Barthes 1966; Ryan 2006; Sturgess 1992; see Rudrum 2005 for a list of a variety of definitions), which in general closely link a narrative to events, we have to look for similarly complex communicative entities in rock art. We suggest these are scenes (Lenssen-Erz 1989: 348–50, 365–69; 2001:141–92; cf,, Dobrez 2011b; May and Domingo Sanz 2010: 37). A scene in rock art is a meaningful structure combining several units of a lower degree of complexity, which in rock art literature conventionally are called motifs.

Motif The term scene describes the relation of individual figures to aggregations of figures, which has always been among the most relevant questions in rock art research (e.g., Davidson 2017), where large panels with many figures and layered palimpsests have always posed a problem for interpretation and understanding. In this study we too describe the individual figures that aggregate to form scenes in terms of “motifs,” for instance when dealing with fauna in a specific or generic term (giraffe or bovid [cf. Davidson 2017], but also human, woman, etc.). Each such figure contains a basic set of features that we can analyze in a standardized descriptive framework. A motif depicting a living being can be an actor in a narrative or in a scene. Data provided in this manner for the present research scheme are the basic ingredients of a linguistic proposition or a normalized indicative phrase with subject-predicate-object (e.g., woman–walk-carry–stick). The definition of motif that we use in our research is: Motif: a pictorial element that can be distinctly described in maximum resolution with a (preferably non-composite) single noun, in the case of living beings irrespective of the activity displayed.

Action of a motif and objects implied by it occur in a wide variety and therefore are not part of the category of motif. The motifs of a rock art tradition are like a clearly defined catalogue from which artists choose

SCENIC NARRATIVES OF HUMANS AND ANIMALS IN NAMIBIAN ROCK ART

93

single specimens for use in their works of art. If several such figures are combined, they would seem to form a narrative, which in our approach is largely defined through action and interaction. Other authors, too, link scenes with narrativity: Holl (2004: 9) has it that scenes depict “sequences of action” that produce a “localized narrative.” Davidson (2017: 22) also has in mind a basically narrative character when speaking about scenes in a “Western sense,” though his criteria for identifying scenic configurations—other than the motif— are based mainly on topological features. Besides the narrative content, we address compositional aspects of scenes through gestalt principles (for more on this see below). The definition of scene that we use in our research is: Scene: A moment in a specifiable event, constituted by a finite number of participants with a coherent gestalt.

According to this definition, the narrative “depth” of a scene is temporally restricted and thus unlikely to tell a whole story, that is, a sequence of events that may imply temporal depth, different places, and changing actors. More likely, a scene can display an epitomic event that represents or symbolizes a whole story. As an example, one may take the crucifixion scene common in Christian iconography: while it depicts only a specific moment (or short phase), for the initiated it epitomizes the complex happenings of a whole day (and more than that). Irrespective of such considerations, the term and concept of “scene” hold such a wide range of meanings and applications, and apparently proved so handy, that it has always been in use in writings about and analyses of rock art, where it most commonly appears as a hunting, fighting, domestic, or dancing scene. The term obviously has its origins in theater, in which the “Western” scene concept (Davidson 2017) goes back to Shakespearean times when the term scene (scena) first appeared in theatrical plays. “A ‘scene’, then, is what is seen at a given moment” (Smith 2013: 98). This “sets the scene,” as it were, for cognition and neuroscience, where scenes are also objects of research and where [a] scene can be defined as a view of the real-world environment from a particular vantage point. Scenes typically comprise background elements and multiple discrete objects arranged in a spatially licensed and semantically coherent manner (Henderson 2011: 593).

Obviously, in this use of scenes activities by humans do not play a role, but here too the visual input at one point is a defining criterion.

Coherence The term coherence has applications in several scientific fields. The common connotations derived from its everyday use across the disciplines (Toolan 2013) are also applicable to the questions addressed here. Coherence has been in use in text-linguistic studies for decades (Bußmann 1983: 537; Toolan 2013), being the feature that makes a text identifiable as a meaningful entity through the assessment of syntactic as well as semantic-pragmatic features. In his analysis of universals of texts, de Beaugrande (1980: 19, original emphasis) provided a definition that allows for a rather effortless transfer of this concept from text to pictorial information: . . . COHERENCE subsumes the procedures whereby elements of KNOWLEDGE are activated such that their CONCEPTUAL CONNECTIVITY is maintained and made recoverable. The means of coherence include: (1) logical relations such as causality and class inclusion; (2) knowledge of how events, actions, objects, and situations are organized; and (3) the striving for continuity in human experience. Cohesion is upheld by continual interaction of TEXT-PRESENTED KNOWLEDGE with PRIOR KNOWLEDGE OF THE WORLD. . . .

In psychology, coherence (Thagard 2000) is used “to help understand processes as diverse as word perception, discourse comprehension, analogical mapping, cognitive dissonance, and interpersonal impression formation” (Thagard and Verbeurgt 1998: 1). Despite the frequent usage of coherence in pathological contexts in the psychological literature, the authors’ further explanation shows that this understanding of coherence may also have some bearing for the understanding of scenes. As Thagard and Verbeurgt (1998: 2) note, “When we make sense of a text, a picture, a person, or an event, we need to construct an interpretation that fits with the available information better than alternative interpretations. The best interpretation is one

94

TILMAN LENSSEN-ERZ, EYMARD FÄDER, OLIVER VOGELS, AND BRIGITTE MATHIAK

that provides the most coherent account of what we want to understand. . . .” Not only does the concept of coherence that Thagard and Verbeurgt develop allow for linkage to pictorial scenes, but they also propose algorithms that make coherence measurable by providing formulae that allow for calculation, that is, an exhaustive inclusion of elements into a coherent whole, or use of neural network algorithms to compute connectionist networks of elements. A calculatory scheme may well point the way forward by diminishing idiosyncratic processes in research of scene identification.

Gestalt In our search for a concept that embodies important characteristics of scenic coherence, the notion of gestalt (Wertheimer 1923) has proved particularly apt because it can be adapted to help in the identification of scene participants and thus the delimitation of scenes: For social psychology, Aron Gurwitsch (1964) did much to resurrect a notion of gestalt perception for its importance to the assembly of everyday scenes, suggesting that perceptual data confront us with phenomenal aspects, features, and characters on account of their integration into a certain contexture or organized group of specific structure. By gestalt is meant ‘a unitary whole of varying degrees of detail, which, by virtue of its intrinsic articulation and structure, possesses coherence and consolidation and thus detaches itself as a closed unit from the surrounding field’ (Maynard 2005: 501).

Research into gestalt formation focuses on the perception and interpretation of grouped objects as well as on small entities within larger environments, and is still relevant today (Wagemans et al. 2012a, 2102b). So-called Gestalt Laws (Fitzek and Salber 1996) are particularly vital to the advertisement industry (e.g., Graham 2008), and outside of psychology (e.g., Wörgötter et al. 2004) they have also received quite some attention in computer-science and mathematical approaches (e.g., Elder and Goldberg 2002; Wen et al. 2010; Zhu 1999). In computer science there is also a conflation of gestalt with scenes, though the latter are normally understood as static configurations (landscapes, architecture; e.g., Grossberg and Huang 2009; Neumann and Terzic 2010; Oliva and Torralba 2001; see also Henderson 2011 as quoted above). A comparable kind of synthesis of the concept of an inanimate scene and gestalt is to be found in Kechagia’s (1995) semiotic approach to rock art aggregations, in which the scene is the spatial setting of the rock art site while the gestalt is the figures acting in a depiction. This scene analysis thus links to phenomenological approaches by emphasizing the visual and bodily experience the beholder of the art experiences at a site, so that moving within the site is understood as comparable to moving about in a scene in a theatrical play. In this concept, however, gestalt and scene are not intrinsically linked to one another. In the juncture of both concepts and in corroborating Dobrez (2011b: 82), who maintains that a scene ultimately is “the observer’s perception of something happening,” we postulate that the identification of scenes follows the “laws of Gestalt” (Fitzek and Salber 1996). These laws include assessment of elements in terms of proximity, similarity, continuity, closure, and periodicity as well as “common fate” (meaning common movement), or Prägnanz (i.e., good gestalt, pithiness) (Wertheimer 1923). Applying this to our concept of scenic coherence, we assess gestalt features of several figures in terms of whether they are identical, similar, or different. According to our concept, scenic coherence exists if several figures share (in order of decreasing importance) species, style, posture, action, color, size, elevation, and/or special features (for comprehensive explanations on this sequence see Lenssen-Erz 1992; 2001: 153–58). The more these features are identical or similar, the easier is it to recognize an aggregation of figures as having a consistent gestalt, that is, as forming a scene. This again agrees with Dobrez’s contention “that seeing a scene is an act of judgement” (2011b: 82). In our method of analysis, a scene provides three main fields of meaning—theme, focus, and setting. This model is borrowed from discourse analysis (e.g., de Beaugrande 1980), where these fields denote the main elements of the information structure. On this basis, a scene comprises given information with motifs that are known to the beholder, thus constituting the theme. The new information about a group of figures forming a scene is the focus, and anything new is primarily dependent on dynamic processes that again can only be activity. Therefore, the first and foremost important feature for the identification of a focus is the activity displayed by the scene participants. The third element of the scene information structure, the setting, is

SCENIC NARRATIVES OF HUMANS AND ANIMALS IN NAMIBIAN ROCK ART

95

comprised of a broad array of contextual information that can be anything that does not contribute to either theme or focus. For a further categorization, defined below are ten foci that make scenes quantifiable on a semantic level, though obviously such a restricted range of categories overrides details of interaction. Therefore, details of the action discernible in the art need to be analyzed on the level of concrete activities displayed by individual figures.

Methods At first glance, a strict transcription of almost unlimited information (rock art) into text (where a strictly formalized scheme with a restricted vocabulary forms the data set for each scene) poses a disadvantage in that compressing information into categories results in information loss. Moreover, such a formal approach merely appears to reveal how rock art is perceived by the investigator and not how its authors perceived and conceived it. The process of categorization is thus a two-level transcription—first from visual artifact to text, and then from original prehistoric language/culture into to a contemporary language/culture (cf. Jäger 2003). We should, however, take into consideration that the method of categorization follows an innate human capacity that forms the basis for the main principles of communication: speech. Although syntax is unique in each of the various languages, some main principles, such as subject, verb and object, are shared, allowing for translation/transcription of information from one language to another. In any of its actualizations, this universal phenomenon can be used to convey the information entities of a scene, which we have labeled theme (participants), focus (directionality), and setting (features surrounding a scene). Given the caveat that the original message remains concealed, probably forever, the archaeological task is to trace as much information as possible from the “artifact”—in this case rock art (scenes). To this extent, the archaeological method of categorization allows us to further transform limited text (categories) back into an encompassing set of meta-information. Regarding rock art, this reads as taking as much information as can reasonably be managed with respect to a single figure, such as species, body features, body posture, and action, along with scenic attributes of participating figures (theme), the main message of the respective scene (focus) and its surrounding context (setting). The linkage between these attributes allows for thousands of different possible, identifiable messages per figure. When aggregated on different scales, from the individual figure to a scene, site, or complete rock art region, these messages can be further contextualized regarding environment or location. The largest analytical advantage of transcribing information into a structured, text-based data set is that it allows for the identification of recurrent patterns, such as the most common behavior of the humans and animals that are its participants—or, in higher resolution, of men, women, children, therianthropes and the various animal species within the rock art. Identifying patterns here means, first of all, a search for a recurring behavior or elaboration of a particular motif, which, on a higher level of organization, stands for latent structures generating complex meanings. This search process is necessarily a computer-aided task, in other words, data mining. The online Brandberg-Dâureb Database is a particularly suitable source for data mining, as all entries in this database contain especially fine-grained annotation describing action. Data mining, a relatively new field in computer science, uses data-driven statistical analysis to find patterns and predict outcomes. In this context, we use it to analyze co-occurrences in an attempt to answer research questions that could not be answered without a rich data set like that of the Brandberg-Dâureb. Rather than looking at specific images and sites, we can analyze and visualize all 39,000 figures from all 651 sites at once, which allows us to distinguish between singular occurrences and encompassing patterns. Data mining is not a singular method but a bundle of methods united by a common goal. In contrast to pure statistics, it is typically not used to verify a hypothesis or test models; rather, it is used to generate hypotheses and construct models. Data mining is thus part of exploratory data analysis. Meanwhile, it allows us to look at the data in a new way. As a tool, data mining is often used to tackle the “Big Data” gathered automatically in large-scale experiments or through massive user input. The dataset of the Brandberg-Dâureb rock art cannot be considered Big Data, but it is too large to be analyzed by means of human observation. Absent additional knowledge of the context the images were made in, intuition can only fail to explain what we see and what it may mean

96

TILMAN LENSSEN-ERZ, EYMARD FÄDER, OLIVER VOGELS, AND BRIGITTE MATHIAK

in its wider context. Instead, with this formal approach, we seek a more objective method that allows us to formulate hypotheses about the wider contexts of the Brandberg-Dâureb rock art. Regarding the scenic information structure, ten foci were predefined using abductive reasoning (LenssenErz 1992, 2001). These foci occur in 120 variations of multiple combinations, always with one primary focus. In the Brandberg-Dâureb rock art, the foci are basically determined by a mimetic description of the participating figures: (1) “common fate” (to cite a gestalt concept), (2) “common direction of movement,” (3) “common direction of gaze” or “bifocal gaze,” or, if neither of these applies, (4) “multifocal gaze” can be attested; their most distinct complex type of culture-specific action (5) “goal of specialized action” or (6) “center of specialized action,” or else interaction (7) “goal of interaction,” (8) “center of interaction”; or the most apparent compositional focus within larger groups of figures lacking any relevant aspect of activity (9) “density center of the scene” that is, the spot where most figures accumulate without any clear activity, or (10) “most elaborate painting” (elaboration of attributes without any marked activity). These ten basic foci are again aggregated in three wider categories as being either directional, interactional, or compositional. The two compositional variants are the “most elaborate painting” (EP) and the “density center of the scene” (DC). Naturally, as it were, the latter is prevalent in animal scenes, since animals are mainly depicted in static mode, slightly dispersed and without any marked activity or body postures. The directional foci category assembles the unspecific dynamic activities as either unidirectional as “common direction of gaze” (CDG), or “movement” (CDM), or “bifocal” (BF) and “multifocal” (MF). Finally, the interactional foci encompassing the “goal of interaction” (GI) or “center of interaction” (CI) as well as types of cultural activity, namely their “center of specialized action” (CSA) or “goal of specialized action” (GSA). Interactions convey the most complex information on figures networked in scenes (Lewis-Williams 1983: 55). Preparation and modeling of rock art into digitally analyzable data (identification of scenes and transformation into text) is based on the concept of coherence discussed above. It has to be emphasized that data mining has been applied in a second analytical step, because this process reveals underlying structures that require a theoretical background and well-founded knowledge of the data to avoid false conclusions.

Data Mining Interactions in the Brandberg-Dâureb Rock Art Data mining is a fairly well established method in computational archaeology, where it is often used with regard to spatial data. Our method is different in that we use data mining of the detailed metadata of rock art images. A basic method of our data analysis is the search for co-occurrences of elements and the correlations implied therewith, which mainly aims at the interaction between actors. Along with the identification of scenes (see above), we have identified different qualities of interaction. Within the corpus of 39,074 single figures on record, 17,416 participate in 4842 scenes. However, only 3,501 of these are actually interacting. The other 13,915 figures in 3332 scenes are defined by a common direction of gaze or movement. The formal aspects of actions and body postures of single figures are regarded as objectively describing mimetic features that are distinctly recorded in the underlying propositional structure. Interactions are thus described within a framework determined by directionality and intensity. In the following section, we will present some examples of insights that can be gained by using data mining methodology. Within the Brandberg-Dâureb Database (BDDB), a figure interacting with another figure is defined by a relational link. The clearest directional and most intense interaction (25 percent of all interactions) is constituted by a (transitive) action of the arms, performed with or without a tool being directed at another figure. In such configurations the focus is on the targeted goal and defines this figure as the “patiens” (Lenssen-Erz 2001: 60), while the figure performing the action is termed “agens.” Another similarly intensive type of interaction exists where two agens-patiens interactions are mutually applied to each other (10 percent of all interactions). With such “mutual attendance,” the directional focus becomes bifocal. In this transformation, the hierarchical character of the agens-patiens relation (Lenssen-Erz 2001: 62) forms a stark contrast to its dissolution in the mutual attendance, where interaction gains a different quality by being simply bidirectional, not hierarchical. These figures indicate that interaction is normally conceived of as being on equal terms and not an expression of power relations where agency is connected to hierarchical dominance. This intense scenic arrangement of interactions is complemented by a broad array of weaker relations. There are directional interactions where a figure is attending to another, which means that the transitive

SCENIC NARRATIVES OF HUMANS AND ANIMALS IN NAMIBIAN ROCK ART

97

action is either less explicit or the directional coherence less pronounced (4 percent of all interactions). Naturally, a certain amount of arbitration is necessary in assigning the intensity of interaction; for example, we consider the manipulation of objects more complex than movement of arms without objects. Finally, there remains a category that is defined by absence of the directionality of action or body posture. These unspecific interactions however maintain a certain coherence in spite of a variety of actions being performed (8 percent of all interactions). This vague category of interactions avoids trivializing them as “miscellaneous” in the context of their important contribution to coordinated actions. Figures thus subsumed are depicted with a very strong coherence where one figure seems to be a copy of the other, and where more than two participants perform the same action and supplementary body posture (Lenssen-Erz 2001: 116, 164) that goes beyond biologically programmed mobility (as seen in animals) to depict a strong cultural expression. This type of activity obviously requires coordination and communication between the participants and is thus described as specialized action, denoting a culture-specific process. Going beyond the gestalt concept of common fate, this kind of alignment is a mimetic reminder of social harmonization of activity, as recalled in group dances or rituals (52 percent of all interactions). As can be seen in the graphs in Figures 6.2 and 6.3, each of the motifs that we have identified develops its own distinct, individual profile in all these parameters of activity. Regarding mutual attendance, the more numerous motifs (humans, women, men, animals) prefer to interact within their own group. This does not pertain to children (children do not play with one another in the Dâureb rock art) but does apply more conspicuously to imaginary beings such as “eared snakes” or therianthropes: they do not interact with each other but largely with natural animals in mutual interactions, avoiding hierarchical action (such as expressed in agens/patiens configurations). This is in contrast to humans, who interact with imaginary beings only in agens/patiens situations. Whereas men interact in all kinds of configurations with imaginary beings, women never interact with them. Interestingly, the intra-group relations are stronger among women than among men, while women’s role toward children appears at first glance to be reduced to a biological quality—most children are attended by women (37 percent) but also by humans (37 percent). This detail is indicative of gender roles not being performed in a “traditional” binary manner—as described above, “humans” are sexually zero-marked. Clearly marked men never interact with children, but humans with signs of male gender performance do (manifested, e.g., in humans carrying children on their shoulders). In applying this interaction model to the BDDB corpus of figures that could potentially participate in scenes, we must account for 30,734 humans and animals as well as all imaginary beings, since we define a scene as depicting an event, and thus requiring actors to perform this event. Approximately a third of these figures are solitary, while 17,416 figures form part of scenes. Three classes of features link the figures: hierarchic interactions (agens-patiens, attending to, attended by: 1,006 figures, 3.3 percent), non-hierarchic interactions (mutual attendance, unspecific interactions: 602 figures, 2 percent), and harmonized group actions (coordinated action: 1,873 figures, 5.8 percent). At this analytic level, the relevant figural motifs already display characteristic profiles of interaction, as a network visualization reveals (Figure 6.4, visualized with Cytoscape). Herein, the area of the category circle represents the number of the interacting figures in the category.3 The thickness of the arrows indicates the quantity of interactions.4 In order to normalize the representations, the interactions shown were referenced to the average number of participants in both groups involved. On the one hand this keeps the size of large groups from over-representing their amount of interaction, while on the other hand the interactions become visible in relation to their pairing combination and at the same time comparable to other pairings. Zero-marked humans without primary sexual markers are the most numerous human motif. They interact mainly with each other, though some directional interactions exist toward other motifs, especially animals, children, imaginary beings, and men. Most children are exclusively engaged in passive interaction with zero-marked humans and women. Over half of the animals taking part in interactions are the goal of directed interactions by men and zero-marked humans. Imaginary beings are the passive target of most multiple interactions, which are performed by men and zero-marked humans; at the same time, imaginary beings never interact with children or women. Sexually marked men have the highest rate of interaction with other motifs; however, they also have a very unspecific diversity. In the rare occasions of sexually marked men engaging with women, they never dominate the action.

98

TILMAN LENSSEN-ERZ, EYMARD FÄDER, OLIVER VOGELS, AND BRIGITTE MATHIAK

Human (910)

Woman (98)

Agens to Human (309) Agens to Woman (309) Agens to Man (54) Agens to Child (13) Agens to Imginary Being (11) Agens to Animal (104)

Man (144)

Child (16)

Agens to Human (309) Agens to Woman (309) Agens to Man (54) Agens to Child (13) Agens to Imginary Being (11) Agens to Animal (104)

Imaginary Being (20)

Animal (146)

Agens to Human (309) Agens to Woman (309) Agens to Man (54) Agens to Child (13) Agens to Imginary Being (11) Agens to Animal (104) 0%

10%

20%

Human (910)

Man (144)

Imaginary Being (20)

0%

10%

20%

Woman (98)

Child (16)

Animal (146)

FIGURE 6.2. Hierarchical interaction of the main motifs; for a definition of the terms Agens and Patiens see text. (Graphic artwork O. Vogels)

SCENIC NARRATIVES OF HUMANS AND ANIMALS IN NAMIBIAN ROCK ART

Human (910)

99

Woman (98)

Man (144)

Child (16)

Imaginary Being (20)

Animal (146)

Number of Figures in Scenes %

Coordinted Action %

Humans (11870) Women (1239) Men (1449) Children (23) Imaginary Beings (131) Animals (2611) 0%

20%

40%

60%

0%

20%

40%

60%

FIGURE 6.3. Non-hierarchical interaction of the main motifs “in mutual attendance” and “coordinated action.” The former is any unspecifiable action of two actors directed at one another; the latter are three or more actors who display comparatively complex activity in the very same manner. (Graphic artwork O. Vogels)

What do these findings reveal about the “subtext” of the rock art? At first glance there seems to be a rather clear division of labor, with a rather high rate of men displayed in a variety of activities while women take responsibility for the children but do not interact with animals or imaginary beings. However, it is apparent in the way women interact within female surroundings that they appear in larger groups, and that their activity is much more likely to be coordinated than that of men or other figures (Figure 6.3). The number of women participating in scenes represents rather exactly their ratio within the entire corpus of rock art—very much as in the case of men or zero-marked humans. However, if one looks at the performance of coordinated action, women clearly stand out as showing the highest frequency of coordination of all figures. This suggests that communication, a precondition for coordinated action, was integral to their activities.

100

TILMAN LENSSEN-ERZ, EYMARD FÄDER, OLIVER VOGELS, AND BRIGITTE MATHIAK

FIGURE 6.4. Visualization of interactions as a network. The circles are a proportionate rendering of the quantity of interacting figures of every motif, and the breadth of lines is a proportionate rendering of the extent of interactions. The short loops indicate interaction within the same motif category. (Graphic artwork E. Fäder)

This is further corroborated when we look at the foci and at which profile the motifs produce (Figure 6.5). In the first place, there is a basic pattern in which all motifs turn up in scenes with a common direction that is dynamic with human figures, generally static with animals, and even more static with imaginary beings. Only children are mainly characterized by interactive narratives. Apart from the directional foci, women are the strongest in the center of specialized action, which again mirrors their high ratio of participation in coordinated action (being a specific form of specialized action). Interestingly, the special status of the “imaginary beings” motif is emphasized by their participation in scenic foci that resemble neither a human pattern nor the pattern of animals: there are remarkably few directional foci, whereas the interactive foci are more equally used than with any other motif. Unsurprisingly, this transcription of the art into patterns of interaction or scenic foci underlines its principal character of being a means of communication that, as a rule, conveys a picture of a realistic, empirical life-world whose “population” acts in a way that can be revealed by our analysis. With this method, we are able to track patterns of behavior and activity that allow us to generate hypotheses about gender roles or the human-animal relationship on a statistically sound basis. These analyses, however, have only just begun.

Future research with data mining Looking at interactions between figures reveals where we would definitely expect a scene, but it might lead to some scenes that cannot be categorized as easily being missed. For the automatic detection of scenes, two properties— homogeneity and repeating pattern—can give us an indication of where to find them.

SCENIC NARRATIVES OF HUMANS AND ANIMALS IN NAMIBIAN ROCK ART

Human (11870)

101

Woman (1239)

goal of specialised action centre of specialised action goal of interaction centre of interaction common direction of movement common direction of gaze bi−focal multi−focal density centre of the scene most elaborate painting

Man (1449)

Child (23)

goal of specialised action centre of specialised action goal of interaction centre of interaction common direction of movement common direction of gaze bi−focal multi−focal density centre of the scene most elaborate painting

Imaginary Being (131)

Animal (2611)

goal of specialised action centre of specialised action goal of interaction centre of interaction common direction of movement common direction of gaze bi−focal multi−focal density centre of the scene most elaborate painting 0%

20%

40%

60% 0%

20%

40%

60%

FIGURE 6.5. Primary scenic foci of the main motifs. They are listed from top to bottom in declining complexity, i.e., in terms of scenic information “goal of specialized action” is the most complex “most elaborate painting” is the least. (Graphic artwork O. Vogels)

102

TILMAN LENSSEN-ERZ, EYMARD FÄDER, OLIVER VOGELS, AND BRIGITTE MATHIAK

From a mathematical point of view, we define homogeneity as the normalized measure of similarity inside the given group. Three elements are necessary: a model of the data set that provides information about the criteria that should be taken into consideration, a limited set of data within that particular model, and some training data. From there we can apply data mining methods that allow us to determine the accuracy with which we are able to predict scenes, and also to find the model that predicts the data with the highest accuracy. Hence, once we know the data of one of the figures in the scene, we can surmise the properties of the other figures more successfully. A figure with a certain head shape would lead us to the hypothesis that the other figures in the scene might have the same head shape, given that in all the other scenes figures tend to have the same head shape. This general approach is then applied to other gestalt features to dismiss non-relevant features and will likewise be able to include complex features of the figures that we have found to be useful in order to approximate gestalt. Our studies on duplicates provide an estimation of what to expect within a scene in terms of homogeneity. Looking at duplicates “manually,” as it were, after they had been detected from the data, we observed that similar figures in close proximity within the same panel tend to occur with higher frequency than expected. This is even true, though to a lesser degree, for panels that are fairly close together. While duplication with a higher complexity is repeated far more rarely in other panels, we can identify entire compositional sets as duplicates close by. An aspect not observed before in rock art research is that duplication can include mirrored figures of simple complexity as well as whole duplicate interactions of this type. In the light of interactions like mutual attendance and coordinated action, discussed above, it becomes necessary to inquire whether the gestalt principles previously applied for scene recognition might actually constitute a high order of mathematical homogeneity and can help us understand standardized structures more thoroughly. A manual inspection of duplicates reveals patterns of similarity for figures in close proximity within the same panel. For figures with a low complexity of action and feature configuration, duplicates can be found with high frequency on other panels and in other sites. Though this observation might seem trivial, it appears to repeat the concept of the harmonic interaction “coordinated action,” thus spreading a concept of homogeneity from basic compositions (scenes in panels) to reference entities of higher order (panels in sites, sites in gorges, and the figure corpus as a whole). On the other end of the duplication spectrum, figures with a high complexity of action and specific elaboration were very rarely encountered as duplicates at other locations. Apparently, duplicates of high complexity (and thus high homogeneity) contribute strongly to scenic coherence. Another aspect of duplication that was revealed through the data mining analysis is the apparent “mirroring” of figures that was observed very frequently in the bidirectional interaction and mutual attendance. All these indications point towards homogeneity as a good approximation for scenic coherence. It is evident that the corpus as a whole requires more mathematical investigation of the frequency of similarities in order to recognize scenes beyond the initial detection model (Lenssen-Erz 1992: 94; 2001: 366), which used a straightforward algorithm of weighted categories to determine similarity along gestalt principles. Our studies on duplicates lead us to a more general principle: We expect homogeneity of the figures to be greater for scenes than for non-scenes. Just as we consider a scene usually to be painted by one artist, so it also seems that homogeneity is generally greater within a scene than in the complete rock art corpus. We can use this principle to find groups that do not fit the generally expected level of homogeneity, as our theory would suggest that they are good candidates for a split into two or more scenes. An obvious interpretation of such a split would be that two different artists shared a space but did not refer to each other. This interpretation can be tested by looking at additional indicators not previously taken into consideration, such as superimpositions, in order to verify the validity of our approach. The advantage of data mining is that the training data does not have to be perfect. Once we have a good approximation of the detection of scenes, we can manually inspect the false positives and false negatives and have the algorithm find the mistakes in our training data for us. As a first approximation, we use the similarity of the pre-existing groups and refine from there, treating the task as a supervised distance metric learning problem (Yang and Jin 2006). Distance metric learning can be reformulated into a binary classification problem (cf. BarHillel et al. 2003) that can then be solved by any classifier, typically the Support Vector Machine (SVM) classifiers (Domeniconi and Gunopulos 2002).

SCENIC NARRATIVES OF HUMANS AND ANIMALS IN NAMIBIAN ROCK ART

103

Many statistical methods like cluster analyses and the like rely on computation of similarity of individuals (e.g., single rock paintings). But leveraging homogeneity has its limits. One important issue is that not all potentially relevant data can be encoded. In particular, relevant geometric properties like the relative position between figures is missing from the available database, as are other kinds of complex information. In addition, expressivity is limited to relations of positive or negative linear correlation, meaning that a certain feature raises a positive correlation between two figures while another may lower it. In general, such correlations may be insufficiently complex to provide an accurate description of the pattern present in the data. Since we are interested in understanding the underlying rules that drive algorithmic decision, we will also use an association rule miner (Hipp, Güntzer, and Nakhaeizadeh 2000). The association rules give us a statistically founded narrative as to why two figures are considered similar for these purposes. A completely different approach to scene detection is to find and analyze repeating patterns rather than looking at homogeneity as the only factor. This way, rather than finding similarity between figures in close proximity, we collect common elements occurring in multiple scenes, which are likely to be interpreted as semantically significant. This could provide insights into what role larger groups of humans or animals in the harmony of common fate fulfill at higher levels of aggregation, such as within and between sites, and in correlation with typical specialized activities and the scenes where these take place. Defining such themes can contribute toward the defining of the cultural significance of characteristic archaeological site profiles within the cognitive landscape. Given the richness of relational data, such as interaction between figures, we can model the data as a network and study the problem of pattern mining in graphs. Again, there are many pertinent works in bioinformatics (Shen-Orr et al. 2002) and linguistics (Bär 2012) which general approaches are similar. In general, the main challenge in applying these algorithms is that we do not know a priori which features are relevant to the repetition and which can be safely ignored. To solve this problem, we redo the work we have already done, in particular, the calculated homogeneity that already factors in to the individual features’ relevance to scene composition. Thus, it can be used as a weighted mutation table. Alternatively, we can use subgroup discovery (Herrera et al. 2011) to find interesting subgroups first and then use the generated rules to either exclude nonspecific features in general or look for repetition in the found subgroups. As zero-marked humans are the most frequent motif found in interactions and scenes as well as the most diversely distributed within themes, defining subgroups will help in establishing a typology of topics and their characterization. It is not yet clear what the social role of this apparent “third gender” without a sex category (Lenssen-Erz 1998) could have been or why it is represented with such abundance. Another challenge is based on repetition and other meta-patterns. A very common pattern in the data that we have not looked at yet from a data mining point of view, is that of common direction of gaze or movement, i.e., several figures walking or looking in the same direction. For aesthetic purposes, it does not matter much if there are four or five figures in the group, but if we do not treat the number of figures as a variable on the pattern, the results will multiply. A pattern of four figures walking will twice match the pattern of five figures walking (i.e., 1-2-3-4 or 2-3-4-5). It is not yet clear how much of an impact this phenomenon will have on the data or how many other features of regular expression learning, e.g., nested repetition and grouping (Fernau 2009), will be needed. However, substantiations on such phenomena would certainly open up completely new spheres of evidence in rock art research.

Conclusion Data mining in a rock art database can only be as good as the quality of its annotations. The BDDB is particularly well suited for such analyses because it is sufficiently rich and because its annotations are based on structural models of information exchange. While the data model for the single-figure database is oriented toward the structure of a linguistic proposition, the scenes in the Brandberg-Dâureb rock art have been identified using a data model inspired by discourse analysis, based on the assumption of a function of the art that was principally narrative. The identification and delimitation of scenes is achieved by implying gestalt laws and a particular concept of coherence. That said, the entire BDDB is available in open access, and any results presented here are open to independent scrutiny and testing.

104

TILMAN LENSSEN-ERZ, EYMARD FÄDER, OLIVER VOGELS, AND BRIGITTE MATHIAK

For the first steps in data mining, we chose to find correlations of individual action in 1:1 encounters of figures as related to scenic foci that on their part aim at abstracting concrete action on a metalevel of activity. While these investigations have just begun, the analyses already show that the clear majority of scenes are characterized by cooperation and harmony, largely due to the fact that most scenes only involve humans. In scenes of humans with animals, there is a high statistical significance of the human imposing an action on the animal, even though explicit hunting scenes are very rare. Another finding indicating an underlying structure of roles surfaces regarding specialized action, which is a collective term for activities that underlie cultural impact, such as culture-specific convention or communication. In these activities, there is a strong dominance of women—as long as this activity remains in the abstract, as it were, without being directed at a goal. In other cases, where there is a goal of specialized action like launching an arrow at prey (though this is rare), men become statistically more prevalent than women. The analyses completed so far show that a good part of the messages in rock art rest on an implicit level, while explicit pictures such as actual hunting or other clearly determinable activities are the exception. Whether rock art is the active manipulation of societal structures or instead their unconscious, passive reproduction cannot be resolved today, but data mining opens up possibilities for unveiling some of these latent structures. Tilman Lenssen-Erz has a master’s degree in African Studies and a PhD in Prehistory. Since 1986 he has been Head of Rock Art Research at the institute of African Archaeology at the University of Cologne, and since 2012 he has led the open online African Archaeology Archive, Cologne. He is co-leader of two projects combining indigenous knowledge with archaeology: “Tracking in Caves” and “Indigenous Knowledge and Archaeoinformatics.” His research always aims at cooperation with and training of local communities for capacity building. Eymard Fäder has worked on Egyptian prehistoric pottery and since 2002 has contributed to rock art research in Namibia at the institute of African Archaeology at the University of Cologne. He has been intensively involved in field research in Sudan and Namibia and has provided training for local guides in Namibia and Botswana. Since 2012 he has been a research associate at the open online African Archaeology Archive Cologne. Oliver Vogels has worked on the Upper Paleolithic of the Balkans as well as music archaeology and rock art in southern Africa. He is part of the research project “Indigenous Knowledge and Archaeoinformatics” (University of Cologne, Germany), which aims at informing models on hunter-gatherer mobility by Namibian hunters. Brigitte Mathiak earned her PhD in Computer Science and is now a professor of Digital Humanities at the University of Cologne, Germany, and speaker of the Data Center for the Humanities. Her research interests include machine learning, pattern recognition, and their applications in the Humanities. According to Google Scholar, she has published a total of seventy-three papers and talks, more than fifty of which went through the peer-review process, including in highly ranked venues such as WWW, ESWC, ECIR, and ICDM as well as journals, Springer lecture notes, and book chapters.

Notes 1. The Brandberg-Dâureb Database (BDDB) is fully accessible at http://datenportal.ianus-fdz.de/pages/collectionView .jsp?dipId=1672239#collectionFiles 2. “A scene in rock art is a meaningful structure combining entities of lesser complexity that are termed ‘motifs.’ The scene contains a ‘theme,’ a ‘centre of interest or focus,’ and the ‘setting,’ which denotes the context of the action” (translation by T. Lenssen-Erz). 3. Coordinated action was not included due to standardization issues, as this interaction does not pair participants—the number of interaction pairs of a given number of participants is their small Gauss sum. 4. Also note that directional interactions are represented in both orientations in order to visualize biases in respect to hierarchy.

SCENIC NARRATIVES OF HUMANS AND ANIMALS IN NAMIBIAN ROCK ART

105

References Anati, E. 1991. Felsbilder. Wiege der Kunst und des Geistes. Zürich: Bär. Bär, D., C. Biemann, I. Gurevych, and T. Zesch. 2012. “Ukp: Computing Semantic Textual Similarity by Combining Multiple Content Similarity Measures.” In Proceedings of the First Joint Conference on Lexical and Computational Semantics Volume 1: Proceedings of the Main Conference and the Shared Task, and Volume 2: Proceedings of the Sixth International Workshop on Semantic Evaluation. 435–40. Association for Computational Linguistics. BarHillel, A., T. Hertz, N. Shental, and D. Weinshall. 2003. “Learning Distance Functions Using Equivalence Relations.” In Proc. International Conference on Machine Learning, 11–18. Barthes, R. 1966. ”Introduction à l’analyse structurale des récits.” Communications 8: 1–27. BDDB. Brandberg-Dâureb-DataBase, IANUS Sammlung 2016-00002. DOI 10.13149/r5f.3fyhip-x Bednarik, R. G. 2003. Rock Art Glossary: A Multilingual Dictionary. Brepols: Turnhout. Bußmann, H. 1983. Lexikon der Sprachwissenschaft. Kröner: Stuttgart. Cole, N. 2011. “‘Rock Paintings Are Stories’—Rock Art and Ethnography in the Laura (Quinkan) Region, Cape York Peninsula.” Rock Art Research 28(1): 107–16. Davidson, I. 2017. “Images of Animals in Rock Art: Not Just ‘Good to Think.’” In The Oxford Handbook of the Archaeology and Anthropology of Rock Art, ed. B. David and J. McNiven, 1–55. Oxford Handbooks Online. De Beaugrande, R. 1980. Text, Discourse and Process: Toward a Multidisciplinary Science of Texts, 4th ed. Norwood, NJ: Ablex. Dobrez, L. 2011a. “Looking at Our Looking: A New Approach to the Definition of a Rock Art Scene.” American Indian Rock Art 37: 251–64. ———. 2011b. “Rock Art, Perception and the Subject/Object Binary.” Rock Art Research 28(1): 71–83. Domeniconi, C,. and D. Gunopulos. 2002. “Adaptive Nearest Neighbor Classification Using Support Vector Machines.” In Proceedings of the 14th International Conference on Neural Information Processing Systems: Natural and Synthetic, ed. T. G. Dietterich, S. Becker, and Z. Ghahramani, 665–72. Cambridge, MA: MIT Press. Elder, J. H., and R. M. Goldberg. 2002. “Ecological Statistics of Gestalt Laws for the Perceptual Organization of Contours.” Journal of Vision 2: 324–53. Fernau, H. 2009. “Algorithms For Learning Regular Expressions From Positive Data.” Information and Computation 207(4): 521–41. Fitzek, H. and W. Salber. 1996. Gestaltpsychologie. Darmstadt: Wiss. Buchgesellschaft. Fritz, C., T. Lenssen-Erz, G. Sauvet, M. Barbaza, E. López-Montalvo, G. Tosello, and M. Azéma. 2013. “L’expression narrative dans les arts rupestres: approches théoriques.” Les Dossiers d’Archéologie 358: 38–45. Graham, L. 2008. “Gestalt Theory in Interactive Media Design.” Journal of Humanities & Social Sciences 2(1): 1–11. Grossberg, S., and T.-R. Huang. 2009. “ARTSCENE: A Neural System for Natural Scene Classification.” Journal of Vision 9(4): 1–19. Henderson, J. M. 2011. “Eye Movements and Scene Perception.” In The Oxford Handbook of Eye Movements, ed. S. P. Liversedge, I. Gilchrist, and S. Everling, 593–603. Oxford Handbooks Online. Herrera, F., C. J. Carmona, P. González, and M. J. Del Jesus. 2011. “An Overview on Subgroup Discovery: Foundations and Applications.” Knowledge and Information Systems 29(3): 495–525. Hipp, J., G. Güntzer, and G. Nakhaeizadeh. 2000. “Algorithms For Association Rule Mining—A General Survey and Comparison.” ACM Sigkdd Explorations Newsletter 2(1): 5864. Holl, A. 1995. “Pathways to Elderhood: Research on Past Pastoral Iconography: The Paintings from Tikadiouine (Tassilin-Ajjer).” Origini 18: 69–113. Holl, A. F. C. 2004. Saharan Rock Art: Archaeology of Tassilian Pastoralist Iconography. Walnut Creek, CA: Altamira Press. Jäger, L. 2003. “Transkription—zu einem medialen Verfahren an den Schnittstellen des kulturellen Gedächtnisses.” TRANS. Internet-Zeitschrift für Kulturwissenschaften 15. Kechagia, H. 1995. “The Row and The Circle: Semiotic Perspective of Visual Thinking.” Rock Art Research 12(2): 109–16. Kechagia, H. 1996. Sémiotique visuelle et art rupestre: application sur un corpus de Namibie. Ph.D. dissertation. Toulouse, France: Université Le Mirail. Lenssen-Erz, T. 1989. “The Catalogue.” In The Rock Paintings of the Upper Brandberg, Part I—Amis Gorge, ed. H. Pager, 343–502. Cologne: Heinrich-Barth-Institut. ———. 1992. “Coherence: A Constituent of ‘Scenes’ in Rock Art.” Rock Art Research 9(2): 87–98. ———. 1994. “Facts or Fantasy? The Rock Paintings of the Brandberg, Namibia, and a Concept of Textualization for Purposes of Data Processing.” Semiotica 100(3/4): 169–200. ———. 1998. “The Third Gender: Human. Gender-Related Patterns of Activity in the Rock Paintings of the Brandberg, Namibia.” In The Proceedings of the Khoisan Cultures & Cultural Heritage Conference, Cape Town, ed. A. Banks, H. Heese, and C. Loff, 146–52. Cape Town: Infosource. ———. 1999. “The Empiricism Strikes Back: Textual Analyses of Namibian Rock Paintings regarding the Role of Women.” In NEWS 95 Congress Proceedings, ed. D. Seglie and P. Ricchiardi, 1–14. Pinerolo, Italy: CeSMAP.

106

TILMAN LENSSEN-ERZ, EYMARD FÄDER, OLIVER VOGELS, AND BRIGITTE MATHIAK

———. 2001. Gemeinschaft–Gleichheit–Mobilität. Felsbilder im Brandberg, Namibia, und ihre Bedeutung. Grundlagen einer textuellen Felsbildarchäologie. Cologne: Heinrich-Barth-Institut. Leroi-Gourhan, A. 1983. Le Fil du Temps. Paris: Fayard. Lewis-Williams, D. 1983. The Rock Art of Southern Africa. Cambridge: Cambridge University Press. ———. 1992. Vision, Power and Dance: The Genesis of a Southern African Rock Art Panel. Amsterdam: Universiteit van Amsterdam. May, S. K., and I. Domingo Sanz. 2010. “Making Sense of Scenes.” Rock Art Research 27(1): 35–42. Maynard, D. W. 2005. “Social Actions, Gestalt Coherence, and Designations of Disability: Lessons from and about Autism.” Social Problems 52(4): 499–524. Neumann, B., and K. Terzic. 2010. “Context-Based Probabilistic Scene Interpretation.” Proceedings IFIP AI, September 2010: 155–64. Oliva, A., and A. Torralba. 2001. “Modeling the Shape of the Scene: Holistic Representation of the Spatial Envelope.” International Journal of Computer Vision 42(3): 145–75. Pager, H. 1989–2006. The Rock Paintings of the Upper Brandberg, Parts I–VI. Cologne: Heinrich-Barth-Institut. Porr, M., and J. Matthews. 2016. “Thinking Through Story: Archaeology and Narratives.” Hunter Gatherer Research 2(3): 249–74. Porter Abbott, H. 2014. Narrativity: The Living Handbook of Narratology. Hamburg: Hamburg University Press. Retrieved 21 November 2017 from http://www.lhn.uni-hamburg.de/article/narrativity. Raphael, M. 1945. Prehistoric Cave Paintings. Washington DC: Pantheon. Rudrum, D. 2005. “From Narrative Representation to Narrative Use: Towards the Limits of Definition.” Narrative 13(8): 195–204. Ryan, M. L. 2006. “Semantics, Pragmatics, and Narrativity: A Response to David Rudrum.” Narrative 14(2): 188–96. Shen-Orr, S. S., R. Milo, S. Mangan, and U. Alon. 2002. “Network Motifs in the Transcriptional Regulation Network of Escherichia coli.” Nature Genetics 31: 64–68. Smith, B. R. 2013. “Scene.” In Early Modern Theatricality, ed. H. S. Turner, 93–113. Oxford Handbooks Online. Smits, L. G. A. 1971. “The Rock Paintings of Lesotho, Their Content and Characteristics.” South African Journal of Science SP2: 14–19. Sturgess, P. J. M. 1992. Narrativity and Its Definitions. Oxford: Clarendon Press. Thagard, P. 2000. Coherence in Thought and Action. Cambridge, MA: MIT Press. Thagard, P., and K. Verbeurgt. 1998. “Coherence as Constraint Satisfaction.” Cognitive Science 22(1): 1–24. Toolan, M. 2013. “Coherence.” In The Living Handbook of Narratology. Hamburg: Hamburg University Press. Accessed 21 November 2017 from http://www.lhn.uni-hamburg.de/article/coherence. Wagemans, J., J. H. Elder, M. Kubovy, S. E. Palmer, M. A. Peterson, M. Singh, and R. von der Heydt. 2012a. “A Century of Gestalt Psychology in Visual Perception I. Perceptual Grouping and Figure-Ground Organization.” Psychological Bulletin 138(6): 1172–217. Wagemans, J., J. Feldman, S. Gepshtein, R. Kimchi, J. R. Pomerantz, P. A. van der Helm, and C. van Leeuwen. 2012b. “A Century of Gestalt Psychology in Visual Perception II. Conceptual and Theoretical Foundations.” Psychological Bulletin 138(6): 1218–52. Wen, G., X. Pan, L. Jiang, and J. Wen. 2010. “Modeling Gestalt Laws for Classification.” In Proceedings of the 9th International Conference on Cognitive Informatics, ed. F. Sun, Y. Wang, J. Lu, B. Zhang, W. Kins, and I. A. Zadeh, 914–18. Los Alamitos, CA: IEEE Computer Society Press. Wertheimer, M. 1923. “Untersuchungen zur Lehre von der Gestalt II.” Psychologische Forschung 2: 301–50. Wörgötter, F., N. Krüger, N. Pugeault, D. Calow, M. Lappe, K. Pauwels, M. van Hulle, S. Tan, and A. Johnston. 2004. “Early Cognitive Vision: Using Gestalt-Laws for Task-Independent, Active Image-Processing.” Natural Computing 3: 293–321. Yang, L., and R. Jin. 2006. “Distance Metric Learning: A Comprehensive Survey.” Ph.D. dissertation. East Lansing: Michigan State University. Zhu, S. C. 1999. “Embedding Gestalt Laws in Markov Random Fields—A Theory for Shape Modeling and Perceptual Organization.” IEEE Transactions on PAMI 21(11): 1–44.

==== •

•

•

7. BETWEEN SCENE AND ASSOCIATION Toward a Better Understanding of Scenes in the Rock Art of Iran Ebrahim Karimi

Introduction: Rock Art Studies in Iran Rock art studies do not have a long history in Iran. The identification of a small number of petroglyph panels in 1962 in Iranian Baluchistan in the southeast of the country (Dessau 1960) and several clusters of pictographs in Kuhdasht in western Iran (Izadpanah 1969; McBurney 1969a, 1969b) marked the beginning of the history of rock art studies in the country. However, for many years rock art studies remained restricted to a number of site reports with limited efforts for primary dating of the images (e.g., Khaniki and Bashash 1994; Shahzadi 1997). The discovery of large concentrations of petroglyphs in Teymare in central Iran (Farhady 1998) was a milestone in the rock art studies of the country, which led to a considerable increase in the number of discoveries and publications on the rock art of Iran. The lack of reliable dating, the absence of direct ethnographic evidence, and the gulf between the discourse of rock art and the main body of archaeological studies are obstacles to reaching a better understanding of the Iranian rock art. However, recent studies have developed our knowledge about some aspects of rock art in Iran, including the distribution of sites across the country and their iconographic and stylistic features (Rafifar 2005; Vahdati 2011). Pictographs have been reported in different parts of the country, for example those identified in recent years in the Zagros Mountains of Fars Province (e.g., Ghasimi et al. 2016), but the main body of rock art in Iran is comprised of petroglyphs, as thousands of engravings have been reported throughout the country. Two characteristics of rock art in Iran are the abundance of hunting themes and the ibex motif, which is the most prevalent depiction in Iranian rock art (e.g., Ghasrian 2009; F. Karimi 2007). The principal subject matters in the rock art of the country include fight scenes between people, particularly in southeastern Iran (Moradi et al. 2013; Sarhaddi 2013; Shirazi 2015, etc.), and predators such as lions and leopards chasing ibex or deer. Other subjects include falconry and group dance in central Iran (Farhady 1998; Jamali 2015; Naserifard 2016). Petroglyphs in northeast Iran show similarities to those in central Asia (Vahdati 2010, 2011). Moradi et al. (2013) also point out the resemblances between rock art in the Marzbanik Valley in the southeast of Iran and the petroglyphs in Central Asia, Afghanistan, and Pakistan, as well as eastern India. In addition, rock art in northwest Iran reveals iconographic and subject-matter similarities to petroglyphs in Armenia’s Geghama Mountains and to rock art in Gobustan in Azerbaijan (Kazemi, Someeh, and Tahmasebi 2016; Rafifar 2005). Dating is one of the most important gaps in the rock art studies of Iran, as the main part of Iranian rock art is comprised of petroglyphs, which are not amenable to direct dating methods (Bednarik 1992, 2010: 217). Some dating suggestions have included use of style comparisons, iconography, or ancient inscriptions available on some panels across the country (Ghasrian, Khanmoradi, and Ghasimi 2014; E. Karimi, Taghva, and Kurdshuli 2016; Shirazi 2015). Although some rock art in Iran may be prehistoric in age, some of the suggested dates attributing Iranian rock art to prehistory (e.g., Farhady 1998; Naserifard 2009, 2016) are not based on sufficient evidence. For example, Farhady (1998) has not developed any strong argument for his dating suggestions about rock art that he identified in Teymare in central Iran. In one case, he suggests 10,000 BP to 15,000 BP for an engraved zoomorphic depiction resembling a stag in Qarqab in central Iran,

108

EBRAHIM KARIMI

based on analogy with the Paleolithic cave paintings in Europe (Farhady 1998: 219–20, 171–74). However, it does not seem that the depictions in Europe and the Qarqab’s stag show any significant similarities to each other. A recent proposal suggested the rock paintings of Abdozu (Ghasimi, Barfi, and Norouzi 2014) and PirBareh (Ghasimi et al. 2016) in the Fars province of Iran were Neolithic or Chalcolithic in age, a suggestion worthy of attention since the authors have provided a better basis for their dating suggestion. They compared the rock paintings with the depictions on prehistoric ceramics of the area, taking into account the archaeological context of the study region (Ghasimi, Barfi, and Norouzi 2014; Ghasimi et al. 2016). However, direct dating results are still needed to provide a stronger basis on which to place Iran’s rock art in a prehistoric context. Iconographic and style comparisons have revealed a degree of similarity between the petroglyphs of Jorbat (Rashidi Nejad et al. 2010) and Dasht-e-Tous (Bakhtiari 2009) in the northeast of Iran to the motifs of the fourth millennium BCE potteries of Bakun in Fars province. Some petroglyphs of Jorbat are similar to art of the late Bronze and early Iron Age traditions (second millennium BCE) of rock art in Central Asia (Vahdati 2010, 2011: 185). The Bronze Age is also suggested as the oldest possible date for some of the petroglyphs in Aps-e Golam and Kouhbodan-e Jor in southern Iranian Baluchistan, as their stylistic features demonstrate resemblances to the images on the ceramic vessels of the third millennium BCE (Shirazi and Soltani 2015). Based on the inscriptions identified on some petroglyph panels found across the country, these engravings are attributed to the historic periods between 25 BCE and 630 CE, which include the Parthian and Sassanid dynasties (Ghasrian, Khanmoradi, and Ghasimi 2014; Khaniki and Bashash 1994). Patination suggests that some of the rock art in Iran is not very old and probably was made during the Islamic period, as Islamic inscriptions are also reported along with rock art panels all over the country (E. Karimi and Ujang 2015; Khaniki and Khaniki 2015; Vahdati 2010; etc.). In some cases the Islamic inscriptions are even reported to have been made on the same panels that hold petroglyphs, such as those identified in Meymeh in central Iran (Khosrowzadeh, Aarab, and Bahraminia 2017) where some of the Islamic inscriptions seem to be even older than the rock art itself, suggesting a maximum Islamic age for the petroglyphs. On one hunting-themed panel identified in Qameshlu in central Iran, the images and the Arabic name Mohammad engraved on the panel show very similar patination, which suggests an Islamic age for the engravings (Karimi and Ujang 2015). In addition to the problems associated with dating, another issue is that the interpretation and the cultural context of Iranian rock art are still not adequately discussed. Farhady (1998) made the first concentrated effort to interpret rock art in Iran. Based on the subject matter, typology of images, and location of the sites in regions with hunting potential, Farhady (1998) suggested that the petroglyphs of Teymare were made by hunters. Appealing to ethnographic evidence, he suggested that some of the rock art in Iran might be related to fertility and the desire for rain. For instance, he mentioned that women of the Afshar and Ghara¯ei nomadic tribes in southern Iran used kashk, a traditional dairy product, to paint herds of sheep, shepherds, and sheepdogs on their black nomadic tents to promote the fertility of their livestock. Each year, according to Farhady (1998), the nomads use the first kashk of the year to paint images on the tents. However, this ethnographic evidence is not compatible with the main body of rock art in Iran because livestock and shepherds are not the usual subject matter in Iranian rock art, nor is tent art directly related to rock art. Rafifar (2005, 2007) suggests that some anthropomorphic images in the rock art of Arasbaran may depict shamans performing rituals, as they exhibit similarities to rock art in Armenia and Azerbaijan that is proposed to have been made by shamans (Rafifar 2005, 2007). Furthermore, Shidrang (2007) also suggests that one of the anthropomorphic depictions in the petroglyphs of Sorkheh Lizeh in Maiwaleh to the west of Iran resembles anthropomorphic motifs found in Central Asia and Siberia and, like those, may be related to shamanic and supernatural beliefs. However, the anthropomorphic image addressed by Shidrang (2007) has equivalents in the art of Iran itself as anthropomorphic depictions with similar style were made on the pendants or seal impressions of the late fifth to the early fourth millennium BCE from Tepe Giyan in western Iran (Caldwell 1976) and Susa in the southwest of the country (Le Breton 1957), and even in Tepe Gawra in northern Mesopotamia (Caldwell 1976). Rafifar (2005: 148–50) mentions that shamanistic rituals still exist in different parts of Iran and are mostly used to treat phobias. Rafifar (2005: 150–51) suggests that shamanism could have entered the Arasbarab region with nomadic tribes who immigrated to Iran by crossing the country’s northern boundaries. One major problem with attributing shamanism to the rock art in

BETWEEN SCENE AND ASSOCIATION

109

Arasbaran is that Rafifar (2005) does not discuss the ethnographic evidence in detail nor use it effectively to develop a strong shamanistic context for the rock art in Arasbaran. Ghasrian (2016) also proposes that rock art in Arasbaran is more similar to hunting scenes in other parts of Iran rather than to the rock art in neighboring countries to the north, such as Armenia and Azerbaijan, as Rafifar (2005) also suggests. Ghasrian (2016) notes that if we accept Rafifar’s (2005) suggested shamanistic interpretation of the petroglyphs in Arasbaran, then this same supposition should apply to the rock art of the other parts of the country, which is not valid in the context of rock art in Iran. Ghasrian (2016) proposes his own interpretation for rock art in Iran, which I discuss in more detail below. Although Ghasrian’s negative attitude toward attributing shamanistic interpretation to the rock art of the other parts of the country seems fair, it needs to be taken into consideration that Rafifar (2005) does not attribute shamanism to the rock art in other parts of the country. Rafifar (2005, 2007) discusses shamanism only in the context of the Arasbaran region, at a particular site, Songon, and in relation to particular “anthropomorphic” images depicted in a dancing-like stance that is unusual in the other parts of the country, a point that Ghasrian (2016) has not fully addressed. However, although shamanism may still be a possible approach to the meaning of at least some of the rock art in Iran, it still does not have a strong context in the country (Ghasrian 2016; Vahdati 2010). Reliable ethnographic evidence is required before inferences about shamanism can be made about Iranian rock art. Ethnographic evidence recorded from the Bisotun region in western Iran has revealed that even in recent decades, some local hunters continued to engrave hunting scenes on their own tombstones before their death to commemorate their hunting skills (Ghasrian 2016). Based on this evidence, Ghasrian (2016) proposes that the hunting themes in the rock art of Iran could have been made by hunters as testaments to their hunting skills. It has also been suggested that petroglyphs in some regions could have been made by nomadic tribes (Farhady 1998; Ghasimi 2007; F. Karimi 2007; Kazemi, Someeh, Tahmasebi 2016) and may be related to their migration routes (Ghasimi 2007) or mark pasture boundaries (Kazemi, Someeh, and Tahmasebi 2016; Sarkhosh, Nazari, and Sharbaf 2015) and the borders between tribes (Moradi et al. 2013). However, none of these suggestions have been adequately tested using independent data.

What Is a Rock Art Scene? There are two main different approaches to defining scenes in the context of rock art. The first definition, which was the typical view for many years, held that “action-interaction” or the presence of at least two images interacting together was the key feature (Bahn 1998; Clegg 1979; Lenssen-Erz 1989, 1992; Smits 1971: 17). Smits (1971) is among the pioneers who have undertaken to define what a scene may be in the context of rock art. He posits that scenes are groupings that the artist originally conceived and intentionally composed to include a number of images that “really belong together.” Lenssen-Erz (1992) emphasized that a scene is a combination of at least two figures communicating together. His publication on the topic in 1992 includes a section reviewing the previous discussions about the notion of scene in rock art studies (Lenssen-Erz 1992). More recently, a new approach to defining scenes was developed by Livio Dobrez (see Ch. 2, this volume). His approach is primarily based on the reception theory approach to perception grounded in phenomenology (Dobrez 2007, 2008, 2010, 2011a, 2011b, 2012, 2017) in combination with cognitive psychology and neurophysiology (Dobrez 2013, 2015a, 2015b, 2016). Dobrez (2007, 2008, 2016) holds that “action” is the key feature in establishing a scene, for which the presence of a single image with sufficient motion or action markers available to a viewer’s perception would be adequate. Dobrez defines a scene as a “depicted event” (2016: 9), an action, “something happening” or “something going on” (2008: 30; 2010: 285; 2016: 9; etc.), which he calls “narrative” (e.g., Dobrez 2008: 30; 2011a: 75). The idea of a single-image scene as proposed by Dobrez has broadened the scope of what a scene is and is even more compatible with the most recent theories of narrative, in which a “single action” is enough to create a narrative (e.g., Altman 2008: 11; Porter 2002: 12). Dobrez (2013: 413) posits the “interaction fallacy” as one of the problematic presumptions in rock art studies. Common in the study of rock art, the interaction fallacy states that interaction is required to represent an event or scene (Dobrez 2013: 413). Drawing on the work of the psychologist

110

EBRAHIM KARIMI

Michotte (1963), Dobrez (2013: 425) discusses how movement perception can occur in the absence of the perception of causality, arguing that “it is the perception of interaction over and above action that entails perception of causality.” Dobrez (2011a: 76; 2016: 9) concludes that a scene requires a depicted action, for which a “single image” suffices and the presence of two images is not necessarily required. Therefore, “two or more interacting” is an arbitrary definition that makes no sense (Dobrez 2011a: 76; 2016: 10). Dobrez (2007: 118; 2008: 30) argues that narrative action does not require group scenes: a “single figure” such as a graceful runner is enough to create a scene and establish us as beholders. According to Dobrez (2008: 30), the observer “sees” not only the visual object but also its “storyline: people running or hunting or dancing etc.” Focusing mainly on the way we perceive a scene (Dobrez 2011b: 252), he argues that this is a “visual constant, a “perceptual universal” (Dobrez 2015a: 61). Dobrez (2015a, 2016) states that three different types of scenes can be categorized as narrative: (1) A scene made of a single image showing action, such as a human figure in running position; (2) An interaction scene that depicts the result but not the agent of the action, as when an animal is hit by a spear; and (3) An interaction that comprises both the result and the agent that performed the action, such as a man striking an animal with a spear. Traditional rock art studies accept only this third type of scene as a scene, but according to Dobrez’s approach, based on the reception theory grounded in phenomenology and the determining role of action as the key and satisfactory element of a scene, types 1 and 2 also qualify as scenes (Dobrez 2015a: 63; 2016: 10). Dobrez (2016: 19) observes that “this approach has the advantage of not discriminating against depicted action in favor of depicted interaction, i.e., one-figure scene vs. multiple-figure scene. If the essential element is indeed “a depicted event,” then one figure running is as much a scene as two figures running.” Dobrez (2016: 19) argues that the distinction is not between what is or is not a scene, but between two types of scenes: “action/interaction ones” (Dobrez 2016). Another aspect of the notion of scene is the question of intention, because in the traditional view, a scene is regarded as an intended unit (Bahn 1998: 195), whereas Dobrez (2011a: 75) suggests it need not necessarily be so. Dobrez (2011a) argues that in rock art, there is a tendency to combine the meaning with intention, and therefore the meaning of an image is the meaning the maker intended (Dobrez 2011a: 71). He points out (2011a: 71, 81; 2011b: 253–54; 2013: 417) that there may be no access to the original intention of the rock art (2011a) and scene (2013). Further, the original intention may be inaccessible not only for the original audience but even for the maker of the rock art, who may have changed or altogether forgotten the intention behind the rock art (Dobrez 2011a: 71). He continues, suggesting that the meaning cannot be restricted to the original meaning and that objects hold other meanings that may be available to us (Dobrez 2011a). He differentiates between the contextual and the textual meaning, arguing that while we may know a lot about the context of an artwork such as the Mona Lisa, we still may not know what the artwork means (Dobrez 2011a: 71). He contends that although access to the meaning would be very limited, it is nonetheless there in the artifact. According to Dobrez, if an artifact is being discussed, then meaning in the artifact will not be incomplete at all, and in the case of a rock art figure, meaning will be placed in the visual markers that constitute the depiction (Dobrez 2011a: 71). However, if the representation pertains to a sign system or symbols, we still would be unable to read them without culture-specific information, which in most cases of rock art would not be available to us (Dobrez 2011a: 71). He continues, explaining how representational visual markers can or cannot be perceptual universals. Dobrez mentions that there are at least some visual markers that cannot be considered universal (Dobrez 2011a). Dobrez (2011a: 72) calls into question our ability to visually perceive ancient rock art images such as those in Chauvet without justifying them. He explains that in the example of Chauvet, the images are still there in the cave and we can still perceive them; however, Dobrez regards visual markers on their own as unable to provide us with contextual meaning, even though the markers perceivable through our eyes would still hold some degree of culture-specific information, offering us a degree of understanding of such images as the Paleolithic paintings in the example of Chauvet (Dobrez 2011a: 72). Although the visual markers in the images would be readable for us and make some sense to our understanding, we still cannot comprehend what, for example, the Chauvet Rhinos might have signified for their Paleolithic producers. Dobrez (2013: 417–18) furthermore observes that whether a composition is a scene and whether the maker intended to create a scene are different questions. To determine whether it was intended as a scene,

BETWEEN SCENE AND ASSOCIATION

111

we should ask the makers (Dobrez 2013: 417), so in the case of rock art there will be no access to the original creator of the art. Dobrez (2013: 418) continues, remarking that the proposition that a scene is a scene because it was intended to be so by the maker, “has it the wrong way around.” Instead, he argues, a scene is intended because it can be seen as a scene. Dobrez reasons that if a group of figures can be perceived as a scene, it is likely to have been intended as such; otherwise, the scene would be just a formal coincidence (Dobrez 2013: 418). As our visual system has not undergone any substantial change for an extensive period of time, the perceptual conformity between past and present would not be an appropriate question (Dobrez 2013: 418). He concludes that although a scene does not necessarily need to be intentional, it is more likely to be intended if it is perceived in this way by a beholder (Dobrez 2011a: 75; 2012: 1840). Dobrez mentions a natural landscape as an example of this situation: a beholder would structure the landscape or the relations between its components, whereas there is no such intention behind a natural landscape (Dobrez 2011a:75).

Scene Markers: How to Identify Scenes in the Context of Rock Art? Dobrez (2008: 29) attempts to suggest universal visual markers to identify rock art scenes. To identify scenes, we need to detect the basic criteria of a scene, which Dobrez (2010, 2011b) proposes as the following: (1) motion markers, (2) compositional asymmetries, (3) orientation or direction of images, (4) distance between representations, (5) profile depiction, and (6) relatively small size of images. His emphasis is mostly on motion markers, including the dynamic stance of images, speed or movement indicators such as angled and bent limbs, and extended legs (Dobrez 2007: 118; 2015a: 39). Asymmetrical and diagonal composition, as well as the juxtaposition of figures with expanding/contracting intervening space that generates compositional discontinuities expressing action, are among the markers that make a scene more active to the viewer’s perception (Dobrez 2008: 30; 2016: 4–5; etc.). Profile depiction plays a twofold role in establishing a scene. It makes the interaction between figures possible and prevents the beholder from entering the closed imaginational space of a scene and taking part in the narrative (Dobrez 2011b: 259; 2016: 5). The distance between figures also has a determining effect of establishing that “something is happening,” as too little space causes confusion and too much space may prevent the beholder from reading an interaction (Dobrez 2011a: 78). In addition to internal features, Dobrez discusses contextual markers including approach, location, and optimal viewpoint (Dobrez 2008; 2011a: 80). A scene is more likely to be seen at eye level and close up, rather than from a distance (Dobrez 2011b: 259). However, a composition does not need all the criteria mentioned so far to be interpreted as a scene, as the presence of only some of them would be sufficient to make a group of images into a scene (Dobrez 2011a, 2011b). Some markers may have a relatively strong effect that facilitates the perception of a scene, while some of the features may not be present in a panel at all. Some markers may be weak in the beholder’s perception and cause ambiguity in the perception of a scene (Dobrez 2011a: 79, 82; 2012: 1840). Conversely, a number of markers evidently available to perception would be adequate to compensate for the weakness of the others and create a typical scene (Dobrez 2011a: 79; 2011b: 259). “Juxtaposition” and “association” are two useful tools for defining the types of relations between representations. The terms are occasionally used interchangeably (Bahn 1998: 195), but Dobrez (2011a: 81; 2012: 1838–39) has effectively discriminated between them. In “juxtaposition” there is no apparent formal relation between depictions, as removing one does not have any effect on another and the images would remain the same, with or without the other (Dobrez 2011a, 2012). On the other hand, when two images are in “association,” they are connected, as removing one affects the other and images look different with or without each other (Dobrez 2012: 1839). Nevertheless, an association is still not adequate to create a scene (Dobrez 2012: 1839): a scene requires action as the key element to be established. In addition to the criteria and tools discussed so far, Jewitt and Oyama (2001: 141–42) suggest that one way to recognize scenes within narrative pictures is the presence of a line, typically an imaginary line implied by the components of a scene, such as a hand pointing toward someone or something. The line—usually a diagonal one called a “vector”—connects the participants of a scene (Jewitt and Oyama 2001: 141–42). A vector expresses a dynamic relation consisting of an action happening between participants (Jewitt and Oyama 2001). Actors, generally, produce the vectors; “goals” are the participants toward whom the vector is

112

EBRAHIM KARIMI

directed (Jewitt and Oyama 2001). When a scene is made of both an actor and a goal, it is “transactive,” indicating the action is taking place between two participants. However, a scene may be non-transactive, that is, made of only one actor and a vector (Jewitt and Oyama 2001). In rock art, a vector can be made of a spear or an arrow in the hands of a hunter (e.g., Figure 7.1b). However, it is important to note that in the context of rock art a vector may just indicate an association between two images, like the hunting theme depicted in Figure 7.1b, rather than a scene, which, according to Dobrez (2007, 2008), requires proper action or motion markers. In the next section, using the tools and criteria discussed above, I will investigate the presence of scenes in the rock art of Iran.

Scenes in the Rock Art of Iran The term “scene” can be found in many publications about rock art in Iran, particularly when dealing with panels displaying hunting themes (e.g., Farhady 1998; E. Karimi and Ujang 2015; Vahdati 2011). However, the presence of scenes and their formal features are topics that have not yet been discussed in the context of rock art in Iran. It seems that some of the examples that are referred to as scenes do not meet the required criteria of a scene and are probably better thought of as “associations” or “symbolic readings,” as Dobrez (2011b: 259–60) has put it. One example is a hunting-themed panel in the Shamsali rock shelter in southwestern Iran (Figure 7.1a) (Azandaryani et al. 2015). The composition is made of several juxtaposed zoomorphic motifs including ibex, a canine, and anthropomorphic figures, all depicted in profile. A horse rider with a hunting tool in hand is described as “a stylistic depiction of a horse rider who is hunting in a scene that is related to the other motifs” (Azandaryani et al. 2015: 62). The closest figures to the horse rider are two black ibex made in a different style and color, suggesting that they and the hunter probably do not belong to the same group of images. Although the hunter is very similar to the rest of the composition in terms of style and pigment, the distance between the horse rider and the ibex—particularly the ibex made with red pigment—has made the connection between them too weak, as it is not available to perception through mere observation. An “association” between the horse rider and the ibex can hardly be inferred, as the hunter is almost isolated from the rest of the panel. Therefore, neither a group scene nor a single image scene can be perceived in this composition, as all images, including the horse rider, are very static and show no motion or action markers. One of the horse rider’s hands is depicted in open position, but this is still not enough to make the image appear to be in motion. The tool in the hand of the horse rider signifies a hunter, yet there is no strong marker indicating an action such as running or throwing a spear. Even if the maker or makers of these images intended to narrate

a

b

FIGURE 7.1a–b. (a) Shmsali rock shelter, western Iran (Azandaryani et al. 2015: 63). Copyright E. H. Azandaryani). (b) Domab, Esfahan, Central Iran. Copyright Ebrahim Karimi.

BETWEEN SCENE AND ASSOCIATION

113

a hunting story, the composition has insufficient formal markers to make it readable for viewers without culture-specific information about the composition. Unlike the depiction of the hunter in Shmasali, which was isolated from the other depictions, in the example illustrated in Figure 7.1b from Domab in central Iran, the distance between the hunter and the ibex depiction is close enough to establish an association between the images. Removing one would alter the other, since eliminating either the archer or the ibex would eliminate the hunting theme, leaving no sense of action implied by the communicating figures. In addition to the suitable distance between motifs, the profile depiction of images, orientation of the motifs in a similar direction, and relatively small size of the images have all come together to effectively suggest that the act of hunting is taking place. Regardless, this combination of the hunter and the ibex is just an “association”: it is still not sufficient to qualify as a scene (Dobrez 2012: 1839) because the composition lacks the key feature of a scene, which is “action” or “motion” markers (Dobrez 2011b: 260). Although the action can still be read by means of other formal markers such as the arrow, which has created a vector showing the direction of action from the hunter toward the ibex, this grouping “signifies a scene,” rather than depicting it by means of perceptible formal action markers (Dobrez 2011b). Dobrez (2011b: 260) calls this “symbolic reading” and suggests that the depictions in such cases are meant to work as signs or symbols that are not accessible to perception. In the rock art of Iran, the lack of motion markers can also be seen in compositions made of multiple images with narrative content, such as a panel showing a hunting theme in Mount Ernan in Yazd in central Iran (Figure 7.2a) (Shahzadi 1997). The hunting theme, which Shahzadi (1997) described as a scene, portrays ibex, canines resembling dogs, a bowman, and two anthropomorphic images with arms open, proba

c

b

d

FIGURE 7.2a–d. (a) and (b) Petroglyphs with hunting themes, Ernan Mount, Yazd, central Iran. Copyright Ebrahim Karimi. (c) Human depictions startling game in a group hunting. From the Akbarnama, made probably in India in 1604 CE. Metropolitan Museum of Art, n.d., “Akbar Hunting with Cheetahs,” Folio from the Akbarnama, Copyright status: Public Domain. Credit: Theodore M. Davis Collection, Bequest of Theodore M. Davis, 1915. (d) Domab, Esfahan, central Iran. Copyright Ebrahim Karimi.

114

EBRAHIM KARIMI

ably driving the ibex. The images are not well integrated together, but the bowman, depicted in a shooting position and facing an ibex, makes the action of hunting readable. Nonetheless, this combination remains at the level of “association.” It does not qualify as a typical scene because all images are static and there is no motion marker to activate the scene. Another example of a group hunting from the same site (Figure 7.2b), Ernan, shows several anthropomorphic and zoomorphic images including ibex and possibly dogs accompanying hunters on a hunt. Like the previous example, the composition seems to narrate a particular event or story, but all depictions are static and lack movement markers. Culturally specific information retrieved from Islamic sources is very helpful in better understanding the content of these hunting themes and similar examples in the rock art of Iran. Hunting scenes have a long history in the ancient art of Iran and are especially prevalent in Sassanid art (224 to 651 CE) (Ghasrian 2009; Porada 2004). Scenes showing animals like dogs, cheetahs, and hawks taking part in hunting with humans have also been common subject matter in the Islamic art of Iran from the first to the recent Islamic centuries, which I will discuss in detail below (Azadbakht and Tavoosi 2012; Canby 1990; Gray 1961: 133). Islamic historic evidence also provides information about group hunting, known as Jargih, in which some people drove animals toward a certain area to make the hunting easier for the hunters (Pajooheshgar 2015: 136–39). This type of hunting was common among royal families of Iran during the Islamic and historic periods (Pajooheshgar 2015). Farhady (1998) has also pointed out panels showing Jargih hunting in Teymare. Humans depicted driving game in open-arm postures also appear in hunting scenes in the Persian Islamic paintings of India, showing royal hunting (Figure 7.2c) during the Mughal Empire (1526–1857 CE) (Eaton 1984; Metropolitan Museum of Art “Akbar Hunting with Cheetahs”). Anthropomorphic images are depicted beside dogs in the rock art of Ernan (Figures 7.2a, 7.2b). Another example in Persian paintings is a hunting scene in Zafar-na¯ma of Sharaf al-Din Ali Yazdi, Tabriz, that shows a human separating a hunting dog from an ibex (Gray 1961: 133). However, it should be taken into consideration that hunting was depicted in the ancient art of Iran for a substantial period of time, so hunting scenes in different parts of the country could have been made in different periods. In some cases in the rock art of Iran, Arabic names or inscriptions on panels with hunting themes, such as the inscription identified on a petroglyph panel in Qameshlu, feature patination very similar to that on the petroglyphs (E. Karimi and Ujang 2015), which suggests the petroglyphs may be of similar age. The panels showing group hunting in Ernan could also have been made during the Islamic period, as their components are similar to the Islamic examples in Persian paintings. Moreover, Islamic names engraved on the rock surfaces and an Islamic graveyard have also been reported in the Ernan region (Shahzadi 1997). Many group combinations representing hunting in the rock art of Iran are made of similar components, as in a panel showing a group hunting in Domab in central Iran (Figure 7.2d). In this case, the linear direction of the motifs is a key feature uniting the whole group. Without the first two ibex, there would be no connection between the hunter and the rest of the composition, including the third ibex and the human shape at the opposite side of the panel. The ibex have connected all of the images by filling the gap between the hunter, the third ibex, and the anthropomorphic image in an open-armed posture at the end of the panel (Figure 7.2d). In other words, the two human depictions are not in direct relation to each other; instead, they are communicating with the same object, that is, the ibex, which connect the whole group. Although the strong integrity among the components of the panel has produced a unified group of images narrating a hunting event, the composition still remains at the level of “symbolic reading” or the “borderline” of a scene (Dobrez 2011b), since this grouping still requires stronger motion markers to activate the scene effectively. Although the front leg of the hunter seems to be bent, it does not provide a strong sense of movement and suggests a position of targeting the ibex rather than indicating the sense of motion. The use of lines to connect the main components of compositions and indicate actions can be seen in many cases in the petroglyphs of central Iran (Farhady 1998; Naserifard 2016), as well as in the southeast of the country (Mohamadi et al. 2016; Sarhaddi 2013; etc.). Two different groups of images, a hunting scene and a fight between two archers, can be seen on the panel illustrated in figure 7.3a (Sarhaddi 2013: 4), where this technique is used to display the arrows fired between the fighters and also to show that the action of hunting is taking place. The lines play a twofold role in creating a narrative on this rock art panel: they work as signs signifying the action of shooting the arrows, and they provide a temporal dimension to the narratives, particularly in the fighting theme, in which the lines would signify the sense of time and the

BETWEEN SCENE AND ASSOCIATION

115

a

b

c

d

FIGURE 7.3a–d. (a) Saravan, southeast Iran (Sarhaddi, 2013: 5), image credits: F. Sarhaddi, https://www .ancient-asia-journal.com/articles/10.5334/aa.12312/). (b) Gharghab, central Iran (Naserifard 2016: 386, Copyright M. Naserifard). (c) A leaping ibex, Jorbat, eastern Iran (Vahdati 2010: 49, Copyright A. A. Vahdati). (d) Teymare, central Iran (Naserifard 2009, 2016, Copyright M. Naserifard).

duration of action. However, the compositions remain at the level of symbolic reading, as actions in both groupings are signified by the lines, rather than being depicted by formal markers such as motion indicators. Although a considerable part of rock art in Iran consists of static images without strong motion markers, dynamic scenes also exist in the country’s rock art. Motion markers such as extended legs and bent limbs (Dobrez 2007, 2015a) can be seen throughout the country in single images, interaction scenes made of two figures, and group scenes. Ibex, deer, felines such as lions and leopards, and horse riders are the most common images, and they are represented with strong motion markers suggestive of scenes. The horse rider of Gharghab in Golpayegan in central Iran (Jamali 2015; Naserifard 2016: 386) (Figure 7.3b) indicates a strong sense of motion, with the animal’s front bent legs suggesting a running stance. The leaping ibex of Jorbat (Figure 7.3c) in eastern Iran (Vahdati 2010) is a good example of an action scene made of a single figure. The profile depiction, the linear style of the image, its relatively small size, and the legs of the animal, depicted in the fully opened position that suggests the leaping stance of the ibex, all come together to create a dynamic scene out of a single zoomorphic representation. A composition made of two images interacting together is a typical form of scenes in the rock art of Iran. The panel illustrated in Figure 7.3d is an example of such a combination, in which a pedestrian bowman is shown in shooting position toward an ibex. Although this grouping is not very dynamic, the minimal motion markers of the ibex, the stance of the human figure in shooting position, and the arrow indicating the direction of action from the hunter toward the ibex intensify the sense of causality between the images and qualify the composition as a scene. Another type of subject matter showing action-interaction scenes consists of images of predators like cheetahs or leopards chasing ibex or deer. In a scene identified in Teymare (Figure 7.4a) (Naserifard 2016),

116

EBRAHIM KARIMI

the sense of movement is effectively indicated by the extended legs of the predator, probably a cheetah or a leopard, following a deer. The front legs of the predator are depicted in attack position toward the deer, which effectively directs the beholder’s eyes toward the result of the scene in order to follow the story line of the depicted narrative. Similar subject matter is evident in the hunting scenes depicted in Islamic paintings, particularly in those produced in India (e.g., Figure 7.2c) (Robinson 1976; Sims, Marshak, and Grube 2002: 66), which suggests a possible Islamic age for at least some panels showing wild predators following deer or ibex. Falconry is another subject matter suggesting scenes, and several examples depicting falconry and hunting scenes with the participation of hunting dogs have been identified in the rock art of central Iran (e.g., Figure 7.4b) (Jamali 2015; Naserifard 2016). In terms of the components of the scenes, they resemble the hunting and falconry scenes in the Islamic art of Iran (e.g., Canby 1990; Gray 1961: 158–59). The falconry scene of Koochari (Jamali 2015; Naserifard 2016) is composed of a human figure in a standing posture and a zoomorphic image resembling a cheetah, both on the back of a horse (Figure 7.4b). A bird-like figure, probaFIGURE 7.4a–b. (a) Teymare, central Iran (Naserifard, 2016, Copyright M. Naserifard). (b) Koochari, central bly a hawk, is illustrated on the horse rider’s hand. The animal’s bent front legs effectively put the scene Iran. Copyright Ebrahim Karimi. into motion. Cheetahs and falconry were common motifs on pottery and in paintings in the Islamic art of Iran from the first to the recent Islamic centuries (Azadbakht and Tavoosi 2012). Use of cheetahs in royal hunting was more common in the Persian paintings of India (Figure 7.2c), where, like the falconry scene of Koochari, cheetahs were even depicted behind horse riders during the Mughal Empire (1526–1857 CE) (Welch 1963). Remarkable examples of cheetahs on horses’ backs or behind horse riders can be seen on the Islamic pottery of the Samani period (819–999 CE) in Neishapur in northeast Iran (Azadbakht and Tavoosi 2012) and also on the thirteenth-century Blacas ewer in Mosul in northern Iraq. These scenes are very similar to the example identified in the rock art of Teymare (Figure 7.4b), suggesting an Islamic age for at least some of the rock art with this subject matter in Iran. Examples of highly active scenes can be seen in the rock art of Lakh-Mazar and Kal-Jangal in Brijand in eastern Iran (Khaniki and Bashash 1994) and Jorbat (Vahdati 2010, 2011) in the northeast of Iran. A battle scene between a human figure and a lion (Figure 7.5) can be mentioned as an outstanding example of a dynamic scene in a realistic style and dimension in the rock art of Iran. The stance of the lion in an attacking position and the partially bent front leg of the human shape are the action markers activating the scene. The “confrontation” between the human and the lion further intensifies the dynamism of the scene. According to the Pahlavi inscription made above the lion, the scene is probably Sassanid (224–651 CE) or Parthian (247 BCE–224 CE) in age and possibly portrays a king or a local ruler of the aforementioned dynasties (Ghasrian, Khanmoradi, and Ghasimi 2014; Hassani 2012; Khaniki and Bashash 1994). Scenes showing battles between humans (mostly kings) and wild animals such as lions have a long history in Middle Eastern art (Hassani 2012; Porada 2004). Such scenes had symbolic meaning, suggesting the power and courage of the kings or champions fighting with the wild animals (Hassani 2012).

BETWEEN SCENE AND ASSOCIATION

117

FIGURE 7.5. The battle scene of Kal-Jangal (Photo credit: SKCHTO.ir).

Conclusion Many compositions in the rock art of Iran reveal a strong sense of integrity and even narrate stories, particularly stories of hunting. In many cases, however, group combinations remain at the level of association or symbolic reading because action—the key element in establishing scenes—and motion markers are not effectively represented. Although in some cases actions are still readable through features other than visual movement markers, such as lines signifying fired arrows, this is not enough to qualify these groupings as typical scenes. Typical scenes with motion markers vary from highly active scenes, such as the leaping ibex of Jorbat and the lion-human confrontation in Kal-Jangal, to scenes with only minimal movement markers, for example

118

EBRAHIM KARIMI

the hunting scene in Figure 7.2c. Zoomorphic depictions are the main images; they usually bear visual motion markers and play the main role in activating both scenes and the flow of narrative in the rock art of Iran. Anthropomorphic figures are mostly depicted as static and rarely show motion markers. In most cases, their action markers are nothing more than a mere bodily stance, such as that of a pedestrian with bow and arrow in shooting posture. The presence of bows and arrows in the hands of hunters is one of the most common features “signifying” actions in the rock art of Iran. The arrows establish a vector line indicating the direction of action (Jewitt and Oyama 2001) from a hunter toward the animal and effectively illustrate the sense of action-interaction in hunting scenes. Good examples of dynamic scenes are found in the east of Iran in Jorbat, Kal-Jangal, and Lakh-Mazar, where the linear style of the motifs, fully opened legs of zoomorphic depictions, attack modes of felines, and confrontation of entities are some of the features used to create dynamic scenes. As discussed before, hunting is a highly prevalent subject matter in the rock art of Iran and was also a common subject in the pre-Islamic art of the country (Azadbakht and Tavoosi 2012; Porada 2004), suggesting that some hunting scenes in Iranian rock art may be pre-Islamic in age (Ghasrian 2009). However, from the first Islamic centuries until recent centuries, hunting subject matter was continuously developed in Islamic art on Islamic pottery (Azadbakht and Tavoosi 2012) and in paintings in Iran, India, and other neighboring countries (Gray 1961; Robinson 1976). Therefore, some these hunting scenes, particularly those showing falconry, dogs, and cheetahs, may be Islamic in age, given that these are common features of Islamic art. According to Ghasrian (2016), one possible function of the hunting scenes is to commemorate the exceptional hunting skills and abilities of the hunters who made them. This interpretation of hunting scenes is supported by ethnographic evidence reported from Bisotun in Kirmanshah in western Iran (Ghasrian 2016). The ethnographic evidence suggests that up until a few decades ago, the hunters of the Bisotun region and neighboring areas engraved hunting scenes on their own tombstones to record and commemorate their hunting skills. This tradition also existed in the other parts of the country among Lor people in Ilam, Loristan, and the Khuzistan provinces in western and southwestern Iran (Faizi and Faizi 2014; Ghasrian 2016; Mirdarikvandi 2016). Although this ethnographic evidence is not directly related to rock art, the iconographic and subjectmatter similarities between the scenes on tombstones and the hunting scenes in the rock art of Iran establish at least a primary connection between the ethnographic evidence and the hunting scenes in the rock art of Iran. Tombstones made in recent Islamic centuries (Faizi and Faizi 2014; Mirdarikvandi 2016) feature depictions of ibex, pedestrians targeting ibex with guns, horse riders, and hunters with swords in hunting fields with ibex. Based on these ethnographic and ethnohistoric accounts and other historic sources, such as the Islamic paintings discussed above, it may be inferred that one possible function of the hunting scenes could be to record the narratives of the hunting experiences of the hunters, and probably to highlight the social value that outstanding hunting skills brought to the hunters in their societies. This interpretation can also be supported by historic reports and archaeological evidence, as hunting was a usual theme in Sassanid art (Heydari et al. 2012), particularly in Sassanid bas-reliefs that show Sassanid kings participating in royal hunting events (Ghasrian 2009; Heydari et al. 2012). Using scene identification criteria derived from the work of Dobrez, it has been possible to characterize rock art made in Iran in several different periods. The main finding is that although the rock art of Iran features highly dynamic scenes with strong motion markers, many compositions introduced as scenes in the publications on the rock art of Iran so far lack strong visual action markers and do not qualify as scenes; instead, they can be categorized as “associations” and “symbolic reading” of scenes. Part of the rock art in Iran, including some hunting compositions with narrative content, fall into this category because they suggest action in ways other than visual markers.

Acknowledgments I am grateful to all copyright holders for their permissions to use their images in this chapter. I also thank April Nowell and Iain Davidson for their helpful comments to improve the chapter.

BETWEEN SCENE AND ASSOCIATION

119

Ebrahim Karimi is a PhD candidate of anthropology at the University of Victoria, Canada. He holds a BA and MA in Middle Eastern archaeology and has participated in several excavations in southwest and southcentral Iran. His main research interest is Middle Eastern rock art with a particular interest in the rock art of Iran, where he has conducted several fieldwork programs, resulting in the identification of several sites in different parts of that country. He has authored a number of publications on Iranian rock art. As a part of his PhD research project, Ebrahim recently conducted intensive fieldwork on the rock art of central Iran resulted in the identification of large number of petroglyphs.

References Altman, R. 2008. A Theory of Narrative. New York: Columbia University Press. Azadbakht, M., and M. Tavoosi. 2012. “Continuity of Motifs in Milver Mishes of Sassanian Era on Nishapur Pottery Designs in Samanid Period.” Negareh 7(22): 57–71. Azandaryani, E. H., H. Qolami, Y. Mohammadifar, and A. Razmposh. 2015. “Discovering New Rock Paintings at Shmsali and Gorgali Rock Shelters in Kohgiluye and Bouier Ahmad Province, Southern Iran.” Arts 4(2): 61–67. Bahn, P. G. 1998. The Cambridge Illustrated History Of Prehistoric Art. Cambridge: Cambridge University Press. Bakhtiari, M. S. 2009. “Survey and Investigation New Petroglyphs of Tous Plain.” Nashri-e Motaleat Bastanshenasi 1(1): 21–43. Bednarik, R. G. 1992. “A New Method to Date Petroglyphs.” Archaeometry 34(2): 279–91. ———. 2010. “Developments In Petroglyph Dating.” Rock Art Research 27(2): 217–22. The British Museum. (n.d.). The Blacas Ewer. Museum number: 1866,1229.61, Retrieved 3 January 2018, from http://www .britishmuseum.org/research/collection_online/collection_object_details.aspx?assetId=22816001&objectId=239367& partId=1. Caldwell, D. H. 1976. The Early Glyptic of Gawra, Giyan and Susa, and the Development of Long Distance Trade. Orientalia 45: 227–250. Canby, S. R. 1990. Persian Masters: Five Centuries of Painting. Bombay: Marg Publications. Clegg, J. 1979. A Possibly Universal Tendency In Art Composition. Paper presented at the Proceedings of the Valcamonica Symposium III, Capo di Ponte, Italy. Dessau, G. 1960. “Rock Engravings (Graffiti) from Iranian Baluchistan.” East and West 11(4): 258–66. Dobrez, L. 2007. “The Reception of Visual Representation,” in XXII International Valcamonica Symposium: Rock Art in the Frame of the Cultural Heritage of Mankind, Darfo Boario Terme, Italy. ———. 2008. “A Rock Art Typology: Narrative and Non-narrative Figurative Representation,” in Aesthetics and Rock Art III Symposium, Proceedings of the XV UISPP World Congress, Lisbon, 4–9 September 2006. ———. 2010. “Good to Think Representational Modalities.” Global Rock Art FumdhaMentos (9): 284–94. ———. 2011a. “Rock Art, Perception and the Subject/Object Binary.” Rock Art Research 28(1): 71–83. ———. 2011b. “Looking at Our Looking: A New Approach to the Definition of a Rock Art Scene.” American Indian Rock Art 37: 251–64. ———. 2012. “Toward a More Rigorous Definition of Terms: Are There Scenes in European Palaeolithic Art?” in L’Art Pléistocène dans le Monde, Bulletin de la Société Préhistorique de l’Ariège LXV-LXVI, special issue, Pleistocene Art of the World), 1837–51. Tarascon-sur-Ariège and Foix: France. ———. 2013. “The Perception of Depicted Motion.” Arts 2(4): 383–446. ———. 2015a. “Making Sense of Pictures.” BCSP 40, Bollettino del Centro Camuno di Studi Preistorici 40: 31–47. ———. 2015b. “Depicted Motion, Interaction and Causality in Rock Art.” American Indian Rock Art 41: 57–68. ———. 2016. “Theoretical Approaches to Rock Art.” Rock Art Research 33(2): 1–24. ———. 2017. “A Universalist Taxonomy for Pictures.” American Indian Rock Art 43: 1–10. Eaton, R. M. 1984. “AKBAR-NA¯MA.” In Encyclopædia Iranica, ed. E. Yarshater, 714–15. New York: Columbia University. Faizi, M., and N. Faizi. 2014. “Archeological Survey of Islamic Cemeteries in Abdanan, Ilam.” Ilam Culture 14(40/41): 46–67. Farhady, M. 1998. Museums in the Wind. Tehran, Iran: Allameh Tabatabaei University. Ghasimi, T. 2007. “A Survey Report of the Ouraman Rock Engravings.” Bastanpazhuhi 2(3): 71–80. Ghasimi, T., C. Barfi, and R. Norouzi. 2014. “Newly Found Pictograms from Abdozu Rockshelter, Firouzabad, Southern Zagros.” Rock Art Research 31(2): 199–204. Ghasimi, T., P. Ghasemi, A. Vafaei, and E. Ghezelbash. 2016. “Recent Rock Art Finds from North of Kavar in Fars, Iran.” International Journal of Archaeology 4(6-1): 8–21. Ghasrian, S. M. 2009. “Wild Goat (Ibex), Hunting and Rock Art Interpretation in Iran.” IFRAO Conference, 29 June–3 July, 2009, Piauí, Brazil. ———. 2016. “Writing Memento or Shamanism? Two Different Interpretations of Iran’s Rock Art.” Rock Art Research 33(1): 73–78.

120

EBRAHIM KARIMI

Ghasrian, S. M., M. Khanmoradi, and T. Ghasimi. 2014. “The Role of Engraved Inscriptions in the Dating of Iranian Rock Art.” Rock Art Research 31(1): 112–15. Gray, B. 1961. Treasures of Asia, Persian Painting. New York: The World Publishing Company. Hassani, M. 2012. “Kal- i Jangal Parthian Carved Stone.” Pazhoshesh Namey-e Tarikh 9(28): 1–16. Heydari, Y., R. Paziresh, P. Sorur, and H. Habibi. 2012. “Khosrow II’s Hunting Relief in Taq-e Bustan and Its Influence on Tang-e Vashi Relief of Fathali Shah Qajar.” The Scientific Journal of NAZAR Research Center (NRC) for Art, Architecture & Urbanism 9(21): 3–16. Izadpanah, H. 1969. “The Paintings of Ducheh Cave in Lorestan.” Persian Journal of Archaeology and Art 3: 53–57. Jamali, M. 2015. Golpayegan Petroglyphs, The Passage for the History. Qom, Iran: Aemeh. Jewitt, C., and R. Oyama. 2001. “Visual Meaning: A Social Semiotic Approach.” In Handbook of Visual Analysis, ed. T. V. Leeuwen and C. Jewitt, 134–56. London: SAGE Publications. Karimi, E., A. Taghva, and F. Z. Kurdshuli. 2016. “The Petroglyphs of Dasht-e-Morghab in the Fars Province of Iran.” Rock Art Research 33(1): 65. Karimi, E., and B. Ujang. 2015. The Petroglyphs of Qameshlu National Park, Central Iran.” Rock Art Research 32(1): 116–19. Karimi, F. 2007. “A New Insight into the Rock Engravings of Iran Based on Field Investigations.” Bastanpazhuhi 2(3): 20–34. Kazemi, M., H. N. Someeh, and F. Tahmasebi. 2016. “Study of Dash Complex, Petroglyphs in Meshginshahr, Northwest of Iran.” International Journal of Science and Research Methodology 5(1): 317–28. Khaniki, R. L., and R. Bashash. 1994. Research Articles: The Archaeology of Ancient Inscriptions: Rock Carvings of Lakh-Mazar, Birjand. Tehran: Cultural and Heritage Organization of Iran. Khaniki, R. L., and M. L. Khaniki. 2015. “The Rock Art of Khanik.” Second National Archaeological Congress of Iran, Birjand, University of Birjand. Khosrowzadeh, A., A. Aarab, and M. Bahraminia. 2017. “New Petroglyphs in Ziad Abad and Hassan Robat Plains (Isfahan Province, Iran).” Mediterranean Archaeology & Archaeometry 17(3): 215–24. Le Breton, L. The early period at Susa, Mesopotamian relations. Iraq 19 (2): 79-124. Lenssen-Erz, T. 1989. “The Conceptual Framework for the Analysis of the Brandberg Rock Paintings.” In The Rock Paintings of the Upper Brandberg, Part 1, ed. A. Gorge, 361–73). Federal Republic of Germany: Barth-Institut. ———. 1992. “Coherence: A Constituent of ‘Scenes’ in Rock Art: The Transformation of Linguistic Analytical Models for the Study of Rock Paintings in Namibia, with Comments and Reply.” Rock Art Research 9(2): 87–105. McBurney, C. B. M. 1969a. “On an Examination of Rock Art Paintings in the Kuh-Dasht Area.” Bastan Chenasi va Honar-e- Iran 3: 7–9. ———. 1969b. “Report on Further Excavations in the Caves of Kuhi-Dasht Area.” Bastan Shenasi va Honar-e- Iran (Review of Iranian Archaeology and Art) 3: 8–9 Metropolitan Museum of Art. (n.d.). Akbar Hunting with Cheetahs, Folio from the Akbarnama. Retrieved 3 January 2018 from https://www.metmuseum.org/art/collection/search/448467?rpp=30&pg=9&ft=Illustrated+manuscript&pos=264. Michotte, A. 1963. The Perception of Causality, trans. E. Miles. London: Methuen Young Books. Mirdarikvandi, J. 2016. “The Depictions of the Lur Tombstones in Lorestan.” Retrieved from the cultural and social website of Tezhgah: http://tezhgah.ir. [In Persian]. Mohamadi, M., E. H. Azandaryani, N. Nikfarjam, and M.S. Doabi. 2016. “New Discovery: Petroglyphs of Takht-e-Siah, Kerman Province, Southeast of Iran.” International Journal of Archaeology 4(2-1): 7–11. Moradi, H., H. D. Sarhaddi, M. Soltani, N. H. S. N. A. Rahman, and B. O. Chang. 2013. “Study and Typological Comparison of Petroglyphs in the Marzbanik Valley, Baluchestan, Iran.” Time and Mind 6(3): 331–50. Naserifard, M. 2009. Iran Petroglyphs (Ideogram Symbols). Khomeyn, Iran. ———. 2016. Iran Petroglyphs; Universal Common Language. Qom, Iran: Vasef Lahiji. Pajooheshgar, A. 2015. “Jargih,” in The Great Islamic Encyclopaedia, ed. M. Jaafari and A. Karimi, 136–39. Iran: The Center for Greater Islamic Encyclopaedia. Porada, E. 2004. Ancient Iran: The Art of Pre-Islamic Times, trans. Y. Majidzadeh. Tehran, Iran: University of Tehran. Porter, A. H. 2002. The Cambridge Introduction to Narrative. Cambridge: Cambridge University Press. Rafifar, J. 2005. The Rock Carvings Art in Arasbaran (Iran). Tehran, Iran: Iranian Cultural Heritage and Tourism Organization, Anthropology Research Center. ———. 2007. “Some Indications of Shamanism in Arasbaran Rock Carvings.” Documenta Praehistorica 34: 203–13. Rashidi Nejad, M., M. Delfani, and N. Mostafazadeh. 2010. “Introducing the Rock Carvings of the Jorbat in Jajarm, Iran.” Modares Archeological Research 1(2): 166–74. Robinson, B. W. 1976. Persian Paintings in the India Office Library: A Descriptive Catalogue. London: Sotheby Parke Bernet. Sarhaddi, F. 2013. “Pictograph and Petroglyphs of Saravan (Sistan-Baluchistan, Iran).” Ancient Asia 4: 3. Sarkhosh, A., S. Nazari, and M. Sharbaf. 2015. “Introduction to Newly Discovered Petroglyphs with Animal Motifs during an Archaeological Survey in Southern Regions of Fars (the Beyram Plain).” Historia i S´wiat 4: 15–30. Shahzadi, D. 1997. “The Petroglyphs of the Ernan Mountain in Yazd.” Archaeological Reports of Iran 1: 134–42. Shidrang, S. 2007. “Maiwaleh: New Discovered Rock Art from Kermanshah.” Bastanpazhuhi 2(3): 55–60. Shirazi, R. 2015. “The Petroglyphs of the Kajou.” Paléorient 42(1): 185–98.

BETWEEN SCENE AND ASSOCIATION

121

Shirazi, R., and M. Soltani. 2015. “The Evidence of Rock Art in Iranian Makran: Aps-e Golam and Kouhbodan-e Jor Petroglyphs, Qasr-e Qand.” Archaeology, Ethnology and Anthropology of Eurasia 43(2): 23–33. Sims, E., B. I. I. Marshak, and E. J. Grube. 2002. Peerless Images: Persian Painting and Its Sources. New Haven, CT: Yale University Press. Smits, L. G. 1971. “The Rock Paintings of Lesotho, Their Content and Characteristics.” South African Journal of Science Special Publication 2: 14–19. Vahdati, A. 2010. Stone Canvases: A Preliminary Report on the Study of Two Rock Art Complexes in North Khorasan Province, North Eastern Iran (Jorbat and Nargeslooy-e-Olya). Iran: Cultural and Heritage Organization of Khorasan-e-Shomali, Iran. ———. 2011. “A Preliminary Report on a Newly Discovered Petroglyphic Complex near Jorbat, the Plain of Jajarm, Northeastern Iran.” Paléorient 37(2): 177–87. Welch, S. C. 1963. The Art of Mughal India: Painting and Precious Objects. New York: Asia House Gallery.

==== •

•

•

8. MUSIC AND DANCING SCENES IN THE ROCK ART OF CENTRAL INDIA Meenakshi Dubey-Pathak and Jean Clottes

Introduction to Indian Rock Art Indian rock art has been under study for decades. The pioneers of research on Indian rock art were two Englishmen, Archibald Carlyle, who first noticed rock paintings in 1867, and John Cockburn (1883) who in 1881 visited painted sites in the Kaimur Range (Mirzapur district). C. A. Silberrad (1907), C. Anderson (1918), B. Allchin (1958), D. H. Gordon (1939) and M. Ghosh (1932) also worked on Indian rock art sites. From 1957 to 1987, V. S. Wakankar discovered many such sites all over India. In particular, he discovered and excavated Bhimbetka, where he revealed stratified sequences from the Lower Paleolithic to Early Historic times (Wakankar and Brooks 1976). Since 1990, the Journal of the Rock Art Society of India, called Purakala, has been published yearly. Only one rock art site so far, or rather a group of rock art sites at Bhimbetka in Madhya Pradesh, has been put on the World Heritage List of UNESCO, in 2003 (Dubey-Pathak 2014). Rock art is plentiful in India. Many thousands of painted or engraved shelters are known to exist all over the country and particularly in its center—Madhya Pradesh, the largest of the twenty-nine Indian states, which is located right in the middle of the country and is the place with the most rock art. Madhya Pradesh is crossed by two mountain ranges, Satpura and Vindhachal, which have numerous rock formations with shelters favorable to painting. We will draw most of our examples from Madhya Pradesh and a few from the neighboring state of Chhattisgarh, which was part of Madhya Pradesh until Chhattisgarh seceded in 2002 (see Figure 8.1). Central Indian rock art subjects are mostly humans, but many animals are shown as well (Mathpal 1984; Pandey 1993; Wakankar and Brooks 1976). Humans are generally dominant in the art, and they are nearly always present, even in the earliest paintings. They may be rigid stick figures or may, on the contrary, hold a dynamic and S-shape (mainly in Raisen and Sehore districts), seeming to be dancing. Other human figures have double lines for the body and arms, and sometimes an inner decoration, though this is seen far more often with animals. Humans often wield weapons, such as bows and arrows (already present in Mesolithic scenes), shields, variously tipped spears, or axes. They are often engaged in activities with other humans (dancing, fighting, parading, having sex, curing the sick, carrying loads, eating, sometimes interacting inside a house or a tent) or with animals (hunting, fishing; riding horses, elephants, or oxen; driving carts or chariots; drawing ploughs). The abundance of scenes of all sorts, taken from ordinary life or mythological, is a characteristic of Indian rock art (Figure 8.2a). Humans, however, have a double particularity: males and females can make music at any time of their choosing, and they make use of musical instruments. That they did so at a very early time is indisputable, as well-made bone or ivory flutes have been found in Germany in the Aurignacian (about 35,000 years ago at Hohle Fels). Not surprisingly, music and dance are practiced in all cultures. However, that does not mean that they necessarily play a significant part in visual art, i.e., in representations, which depend upon diverse cultural beliefs and practices. For example, we know very few instances of such representations in European Paleo-

MUSIC AND DANCING SCENES IN THE ROCK ART OF CENTRAL INDIA

FIGURE 8.1. Map showing the main regions of rock art in India.

123

124

MEENAKSHI DUBEY-PATHAK AND JEAN CLOTTES

d

a

b c

FIGURE 8.2a–d. (a) Mythological scene of a warrior fighting a horned lion. Between them is a man who has just lost his head and his arms. Churna (Pachmarhi, Madhya Pradesh). Historic. (b) In this complex panel of Chaturbujnath Nala B17 (Madhya Pradesh), at top left a line of people wearing elaborate headgear and wielding axes and shields are dancing to the sound of big drums. Neolithic. Enhanced with DStretch (-lre). (c) At Bhimbetka 4 (Madhya Pradesh), to the right of a warrior armed with a bow and arrows in front of a mounted horse, two men with elaborate headgear are dancing while playing small drums connected by a wooden or bamboo bar. Historic. Enhanced with DStretch (-lre). (d) At Urden 29 (Madhya Pradesh), a crouching man is playing a harp while a big man standing at left is playing an instrument and dancing. Two dancers are holding hands in the middle and two others, at right, are dancing in a sitting position. Four of those characters have open mouths (singing?). Historic. Enhanced with DStretch (-lre).

lithic cave art. The “Sorcerer” engraved in the Trois-Frères Cave (France), who seems to be playing a nose flute, is unique. By contrast, as we discuss below, such images abound in Central Indian rock art from the Mesolithic onward and particularly in the recent periods we have been studying (Clottes and Dubey-Pathak 2013; DubeyPathak 2013; 2014; Dubey-Pathak and Clottes 2017).

Chronology Since Vishnu Wakankar’s work in the 1970s, the color green has been regarded as a color belonging to the late Paleolithic because it was (rarely) used in early paintings that sometimes were covered over by later paintings in red or white. Several chronologies have been proposed for Indian rock art. They can mostly be classified in three main periods: the Prehistoric, the Protohistoric, and the Historic.

MUSIC AND DANCING SCENES IN THE ROCK ART OF CENTRAL INDIA

125

At the end of the Paleolithic or the beginning of the Mesolithic at the end of the Pleistocene, that is, in cultures that had not yet domesticated animals, the Prehistoric Period, also called the Mesolithic, lasted from perhaps 10,000–8,000 BC (?) to 2500 BC. No evidence exists of any paintings that could be dated to earlier times, except perhaps for the green figures mentioned above, whose antiquity has not yet been corroborated. The Protohistoric period, from perhaps 2500 BC to AD 300 is also called the Neolithic/Chalcolithic. It is characterized by depictions of agricultural activities, ploughing in particular, and domestic animals (cattle, mostly humped bulls), by pottery making, and by the presence of huts and chariots. Archers hunting are frequent. The images from that period are more difficult to attribute to a particular culture, as many of the motifs (men with bows and arrows, hunting scenes, dancing) existed before and would continue to be represented later. The third period, called Historic, lasted from AD 300 to 800 and even later, into the Late Historic or Medieval period, post-Gupta, where the chronology is more assured thanks to the presence of particular weapons (e.g., swords made in particular styles); that is, it would last from AD 800 to 1300. Some of the images can be attributed to Buddhists. The paintings differ from those of the earlier periods in both subject and style; they include inscriptions, symbols, various animals including camels, bullock carts, archers and scenes with warriors using shields, elephant and horse riders on sometimes heavily caparisoned animals, and warriors using metal weapons. There also are domestic scenes of men and women engaged in various activities inside houses or tents. Other fairly frequent scenes portray people collecting honey from huge natural hives, usually hanging from the branches of slender trees. The paintings are in a much better state of preservation than those belonging to earlier periods, on which they are often superimposed. In fact, painting in shelters has never stopped. Some “Historical” figures here and there can be attributed to the nineteenth century and others (hand stencils, crude figures) were even made in the past few decades in Deur Kothar and Dharkundi in the Rewa area east of Madhya Pradesh, Chaturbujnath Nala in the Mandsaur area in the west, Jaldafy, the Madhai area within the Pachmarhi area. In some areas (Betul), geometric motifs are still painted nowadays.

The Musical Instruments Represented We shall now consider the rock art of Central India, where musical instruments are quite frequently shown.

Drums We have found representations of drums at a number of rock art sites, all in Madhya Pradesh State, where we did most of our work. They are: Bhimbetka (five shelters) Samardha, Urden, Chiklod, Jawara, Karwai, Jhiri in Raisen district, Kathotiya, Budni in Sehore district, Hathitol in Panna district, Nachankoda in the Chhaterpur district, Abchand in Sagar district, Jhinjhari in Katni district, Chaturbujnath Nala (shelter B7) in Mandsaur district, Likhichhaj in Morena district, and five in Pachmarhi (Pareba Pahari, Saderi, Bori, Tapka Pani, Kanji Ghat 1) in Hoshangabad district. None of them can be ascribed to the Mesolithic; they all belong to more recent periods (i.e., the Neolithic and Historic). We have seen two kinds of drums in the art. A big cylindrical drum hangs from the neck and is held in the middle of the body. It can be used at both ends. Sometimes the player hits it with his (or her) hands; sometimes the drum is hit with a stick instead (Figure 8.2b). Most times it is impossible to determine the sex of the player, except in one case where a headless woman is recognizable because of her breasts. The second kind of drum consists of two separate, small, roundish drums, one on each side of the body, connected by a bar made of bamboo or of wood (Figure 8.2c). The body of the drum was generally made from a hollow piece of wood. The membrane stretched over the mouth of the instrument was made of animal hide. Dancers are often shown close to drum players.

Harps Harps are nearly as numerous as drums. We found them depicted at sites in the following states. Madhya Pradesh: Urden (two shelters, Figure 8.1d), Kathotiya, Firangi, Kharwai (Figure 8.3a), Samardha, Chaturbujnath Nala

126

MEENAKSHI DUBEY-PATHAK AND JEAN CLOTTES

a

b

d

c

FIGURE 8.3a–d. (a) At Kharwai 5 (Madhya Pradesh), two long rows of dancers are converging toward a harp player at left. Below them are smaller people probably also dancing. Historic. Enhanced with Dstretch (-lre). (b) Samardha (Madhya Pradesh). At top left, two men sitting are playing harps. Below them a long row of eight dancers holding each other are facing left where another harp player is sitting. The white bodies of the eight dancers have been completed and enhanced with brownish paint (long hair, neck and upper part of the body, skirts). Historic. (c) At Lekhamada 1 (Chhattisgarh), two harp players face each other at left. They are faced by seven dancers who are holding each other by the waist. Historic. Enhanced with Dstretch (-lre). (d) Chaturbujnath Nala B17 (Madhya Pradesh). A strange scene involves several characters. To the left, a man with a head with raised hair is playing a big harp in front of an elongated woman with the same head holding something and dancing. To their right are two extraordinary big figures. The first one, with four dots on her face has got the same hair as the other two. Is it singing/dancing? The long vertical motif attached to its right cannot be determined as a human. Medieval. Enhanced with Dstretch (-lre).

(five sites), Penganwa, Pachmarhi (five sites: Langi Hill, Batki Bundal, Nimbu Bhoj, Rajat Prapat, Khari Lane). Chhattisgarh: Lekhamada (two sites). The harps can mostly be ascribed to the Chalcolithic and Historic. The harps are open-ended and have different numbers of strings (from one to seven) (Figures 8.3b and 8.3c). Their frames and soundboards were made of flexible wood, the pegs were made from wood or gourd, and animal gut was used for strings. Harps are mostly associated with dancers, so their role is clear (Figures 8.2d–8.3d). Another type of harp has a long bamboo or wood frame with a gourd at each end. It is considered a single-string instrument and can be played with a hand or a bow. Some kinds of harps are still popular among the tribes.

Pipes We found sixteen rock art sites depicting pipe players: Bhimbetka, Jhiri, Jawara, Chiklod, Kathotiya, Budni, Panna, Sagar, Chhaterpur, Rewa, and in Pachmarhi (Nimbu Bhoj (two pipe players), Sambhar Jhil, Kanji Ghat 2, Churna (Figure 8.4a), Tapka Pani) and one in Urden 28. At Tapka Pani, there is also a drummer and dancers, one of whom may be wielding a big pipe (Figure 8.4b). Such pipes would be made of readily

MUSIC AND DANCING SCENES IN THE ROCK ART OF CENTRAL INDIA

127

b

a

d c

FIGURE 8.4a–d. (a) Churna (Madhya Pradesh). At right and facing left, a man painted in solid red is playing a long pipe. Another man painted in the same way, at left, seems to be holding the same kind of pipe and perhaps dancing. There are many superimpositions on this panel. Historic. Enhanced with Dstretch (-lre). (b) At Tapka Pani (Madhya Pradesh), a drummer and a pipe player are playing for a row of dancers holding each others’ shoulders. The dancer at the end of the row at left is brandishing a pipe or a stick. Historic. Enhanced with Dstretch (-lre). (c) At Chiklod (Madhya Pradesh), two men are playing drums behind a mounted elephant and numerous warriors. Enhanced with Dstretch (-lre). (d) Penganwa 4 (Madhya Pradesh). Badly preserved images of cymbal players. Historic.

available hollow tubes such as horns, bamboo, or bones. In early times, pipes and drums were also used to announce conflicts and warn other groups (Figure 8.4c).

Cymbals Cymbals are present, with dancers, in two Pachmarhi shelters (Penganwa 4 [Figure 8.3d] and Sambhar Jhil), where the images, though still distinct and understandable, are not very well preserved. At Penganwa 12, a group of dancers includes three big figures and at least eight smaller ones with a possible cymbal player. Five of the dancers, one of whom has a big penis, are in a row holding their arms upwards.

Dancing Scenes in the Art Representations of people dancing are particularly numerous in Central Indian rock art. Dances obviously played an important part in people’s lives, as they also do in art. The dancers are most often in a row, but sometimes they may be in a group (organized or not), or even alone or just in a couple.

128

MEENAKSHI DUBEY-PATHAK AND JEAN CLOTTES

Dancing in a Row, in a Group, or Solo In the hundred painted shelters with dancing scenes in the two states of Central India where we have mainly worked, we have found many rows of dancers painted in more than fifty different shelters, eight in the state of Chhattisgarh and the others in Madhya Pradesh. Many depictions of S-type green and red humans dancing in a long row belong to the Mesolithic period (Figures 8.7b and 8.7c). Most times there is a single row of dancers in varying numbers (from three [Figure 8.5a] to thirty-five). Occasionally one finds several rows of dancers on the same panel or in the same shelter. For example, in Singar Pathar (Chhattisgarh), we saw four different rows of dancers (Figure 8.5b). The first row includes at least thirty-five yellow dancers with bent knees facing right and arms half raised; above them, two bigger red dancers face left and five red ones face right. Just below that row, another row includes five red dancers facing right with their arms held horizontally and bent upwards. At least four of them have penises, suggesting this may be a puberty ceremony. Below them are at least sixteen red dancers facing left with their arms in the same posture as the dancers above. The style of these dancers indicates that they probably date to the Chalcolithic. A similar scene at Kathotiya (Madhya Pradesh) involves forty-eight dancers and musicians (harpers), all with erect penises. They are represented in two rows, one above the other. In the top row the dancers are smaller and have open mouths. They might be boys during a puberty ceremony (Figure 8.5c). The harper is standing between them. In the bottom row people dancing with another harper are bigger and look like men. In Sita Lekhni 4 (Figure 8.5d), also in Chhattisgarh, many small humans, not touching, are depicted in the same style and posture, with their bodies slightly bent and their short arms held forward. They are in three rows, with the top row having at least nineteen dancers, the middle row at least twenty-seven and the bottom row at least eight. The figures in the middle row face right; all the others face left. The other sites are a

c

d b

FIGURE 8.5a–d. (a) Urden 19 (Madhya Pradesh). Two women and a child with decorated bodies, dancing with their arms outspread. Medieval. Enhanced with Dstretch (-lre). (b) Singar Pathar (Chhattisgarh). Two yellow and red rows of dancers facing right in the same stance, with bent arms held in front and knees half bent. The many handprints are from a later time. Mesolithic? Enhanced with Dstretch (-lre). (c) At Kathotiya (Madhya Pradesh), the top row of a scene with many dancers and musicians could represent a puberty dance. Enhanced with Dstretch (-lre). (d) Sita Lekhni 4 (Chhattisgarh). Three rows of numerous slender dancers with bodies in an S shape face right (middle row) and left (upper and lower rows). Mesolithic.

MUSIC AND DANCING SCENES IN THE ROCK ART OF CENTRAL INDIA

129

Likha Pakhna 2, Majhingargh 2, and Badra 2 in Chhattisgarh, and Nagauri, Jhiri, Kathotiya (Figure 8.5c), Jawara, Chaturbujnath Nala B10, and Chaturbujnath Nala B6 in Madhya Pradesh. We found representations of people dancing in a group in twenty shelters (nineteen in Madhya Pradesh and one in Chhattisgarh). In many cases (Likha Pakhna, Bhimbetka, Putli Karar, Urden, Kathotiya, Chiklod, Jhiri, Jawara, Budni, Panna, Sagar, Vidisha, Chhaterpur, Rewa, Katni, Morena, Shivpuri, Chanderi, and Pachmarhi), several distinct scenes were painted. For example, at Bori (Figure 8.6a), eleven beautiful dancers were painted, each with big eyes. They are framed by white lines that might represent an open tent. Among them are at least three women (with small breasts, wearing long dresses). Three dancers are lying down. The dancers’ outlines are drawn in brownish red, and their bodies are drawn in off-white. The upper halves of the bodies are naked. From the waist down they are wearing skirts: long ones for women (distinguishable by their breasts) and shorter ones for men. All have their hair tied in buns. Six isolated quivers are hanging from the “tent.” These drawings are probably Historic (Medieval times). Among other Historic scenes, we can refer to two rows of dancers in Kathotiya holding sticks. The sticks are called dandas among the Gonds, who still practice that particular dance during their festivals. a

b

d

c

FIGURE 8.6a–d. (a) Bori (Madhya Pradesh). Elegant dancers whose eyes are represented. They are inside a tent (denoted by white outlines) where quivers, swords, and shields are hanging. Medieval. (b) Bori (Madhya Pradesh). This long shelter features many dancing scenes belonging to different periods. This scene, painted white with brownish infilling, shows numerous dancers with elongated headgear. Below, two dancers are holding a bow and arrows, a wild boar, and a big fish. Mesolithic. (c) Bori (Madhya Pradesh). On this panel we can see three dancing scenes belonging to different periods (Mesolithic, Historic, and Medieval). (d) Belkandhar 1 (Madhya Pradesh). White woman dancing and singing. Historic.

130

MEENAKSHI DUBEY-PATHAK AND JEAN CLOTTES

Other dancers, sometimes in the same shelter, date to the Mesolithic, judging from their style. Also in Bori (Pachmarhi), eight dancers have off-white and dark red bodies, elaborate headgear, and body ornaments for their legs and arms. Below them, two archers are in front of a charging wild boar with an intricately decorated body. To the right is a big fish painted with the same technique (Figure 8.6b). In the same shelter there is another Mesolithic scene of ten to twelve dancers with large, complex headgear and body ornaments. Still in Bori, a big panel (figure 8.6c) depicts three groups of dancers belonging to different periods. The first group (below left) is made up of five yellow figures: two couples dancing and, to their right, a fifth figure holding a stick. Weapons are seen next to them. The second, main group features a man at the top holding a sword and shield and walking left. In the middle of the panel, fourteen white figures are outlined in dark red. They are superimposed over earlier off-white paintings. Some have two big eyes. Eleven of them are going to the left and wield weapons (bows and arrows, shields and swords). One is running. Others seem to be dancing. Three persons are sitting without any weapons; next to them are bows, arrows and quivers. The third group is early (Mesolithic); it used to occupy the whole panel. Five stylized possible dancers are wearing headgear, and two humans are chasing a bull. The sites with dancers in groups are Likha Pakhna 2 in Chhattisgarh; in Madhya Pradesh, they are Pachmarhi (Bori, Pareba Pahari 1, Belkandhar 1, Churna, Burburi Lane, Madai) Bhimbetka 6, Firangi, Chaturbujnath Nala B4 and 19, Kathotiya, Putli Karar, Urden 28, Kharwai, Jawara, Panna, Sagar, and Budni. Isolated dancers (i.e., not in a row) are rare (six in all). Three are in pairs, at Kohbar in Chhattisgarh and at Deur Kothar and Belkandhar 2 in Madhya Pradesh. They are not touching each other, and they are all raising their arms. At Belkandhar 2, two off-white figures include an adult in a long dress (possibly a woman) and a long headdress turned to the right, next to a much smaller figure (possibly a child) on her right. At Deur Kothar the pair wear a four-pronged vertical headgear. Three dancers are represented alone at Chaturbujnath Nala B19, Putli Karar, Belkandhar 1. The most spectacular one, at Belkandhar 1 (Figure 8.5d) is a woman (identified by her breasts) turned to the left with a stole round her neck, one arm raised and the other down. Her mouth seems to be open, perhaps singing. Whatever their numbers, the dancers’ bodies are sometimes detailed, as at Urden 19 (Figure 8.5a), Urden 28 (Mesolithic geometric decoration inside the body), Urden 9, and Samardha. Early Mesolithic green dancers (Figures 8.7b and 8.7c) are shown in Bhimbetka, Kathotiya, Jawara, and Kharwai. The shape of the bodies may be a rectangle (Kabra Pahar, Badra 2) or a triangle, which may be simple (Madhay, Churna) or double (Chaturbujnath Nala B3, B5, B7, B11, Tapka Pani), or have vaguer outlines (Urden 40, Kohbar). Their inside is nearly always filled, exceptions being Kharwai C7 (Figure 8.3a) and Bhimbetka 4. A majority of dancers, though, are mere stick figures. They may be represented in four different postures, several of which can be found in the same shelter: facing the visitor, turned right, turned left, or facing each other. Many of the dancers are wearing an item of clothing at the waist that looks to be a skirt or a long dress, often with a band, sometimes with inside decoration (Bhurburi Lane, Samardha, Kharwai, Urden, etc). As to their heads, while some dancers have complex headgear, like at Hathitol in Raisen district (Figure 8.7a), others are surrounded with dots or short dashes (Bori, Chaturbujnath Nala B3). Buns are fairly frequent (Bhurburi Lane, Madhai, Rajat Prapat, Kharwai, Jawara, Urden, etc.). Sometimes headgear is represented by more or less important slanting lines that converge on the top of the head or body, as at Chaturbujnath Nala B5, B7, B8, B11, B13, Majingarh 2. Some of the dancers are represented with an open mouth, that is, singing while dancing (Bhurburi Lane, Belkandhar 1, Nishan Garh, Rajat Prapat, Chaturbujnath Nala B19). As to the paint used for those dancers, it is mostly red. A few are white or off-white (Badra 2, Gaddie, Bhurburi Lane, Belkandhar 1 and 2, Churna, Bori, Pareba Pahari). There are a few paintings of green dancers at Bhimbetka (Figure 8.7b), Jawara, Kathotiya (Figure 8.7c) and Kharwai, and some of light yellow ones at Churna and Bori. The dancers are often holding an object; sometimes they may all be holding it (Putli Karar 6, Hathitol Penganwa 4 in Raisen district Astachal, Bori, Madhai, Bhurburi Lane in Pachmarhi in Hoshangabad district, Chaturbujnath Nala B11 (Figure 8.7d), B3 and B19 in Mandsaur district. One of the most interesting examples is Kharwai B9, where all the characters hold a madhai, a bamboo stick adorned with peacock feathers all around its top, used for special ceremonies after Diwali (the figure at right also holds arrows in one hand). In the other cases, the object bearers are at one or both ends of the row. The objects held can be axes,

MUSIC AND DANCING SCENES IN THE ROCK ART OF CENTRAL INDIA

131

c

a

b

d

FIGURE 8.7a–d. (a) Hathitol (Madhya Pradesh). S-bodied dancers with elaborate headgear holding hands. Mesolithic. (b) Bhimbetka (Madhya Pradesh). Mesolithic dancers painted in green. Enhanced with Dstretch (-lab). (c) At Kathotiya (Madhya Pradesh), a big bison, with a smaller animal inside its body, is superimposed over a row of green dancers. (d) Chaturbujnath Nala B11 (Madhya Pradesh). A number of dancers holding a stick in each hand to make music. Most of them seem to wear two bison horns as a headdress. Historic. Enhanced with Dstretch (-lre).

sticks, shields, swords, branches, and in one case—at Chaturbujnath Nala B6—a stack of pots held atop their heads, as is done nowadays for Rajasthan folk dances called Kalbeliya dance.

Dances in Tribal Life Today, traditional dances are still an integral part of tribal life. In general, they are dedicated to gods and take place at various festivals. They may assume different forms according to the diverse tribes and to the innumerable occasions on which they are performed. We shall mention a few examples. In some cases, the dances have close links to the rock art. For example, in both Chhattisgarh and Madhya Pradesh, tribal members use rock art sites as places in which to sing religious songs and play drums, cymbals, and flutes, for example at Chaturbujnath Nala (Mandsaur), Gouri Dant (Sagar), or Patai Dongri in Baster Chhattisgarh. The Kalbeliya or Bhawai dance mentioned earlier is still popular at festivals such as Diwali, in important ceremonies (weddings) in Rajasthan, and in the art of Madhya Pradesh close to that state. Music is sometimes used for the precise purpose of entering a trance. One of us (MDP) saw Chhattisgarh Gond shamans ask a flute player to play particular music to induce trance. When it worked and they were in trance, the flute player and drummers went on playing nonstop for a long time. In the Bastar area in south of Chhattisgarh, the Muria Gonds also perform special Gaur dances for weddings and other festive events. They are very popular. Those who dance are dressed in their best. Men wear

132

MEENAKSHI DUBEY-PATHAK AND JEAN CLOTTES

elaborate headgear, always with bison horns topped with entwined feathers. Their faces are adorned with strings of cowrie beads. Very big cylindrical drums are played with one hand and a stick. Women next to them wear long, elegant and flashy dresses. Their heads are topped with feathers. They hold long rods that in the past were made of wood and now are made of metal iron, with very small bells at the top. As they dance and sing, they knock the rods against the ground to participate in the music. The Raut dance of the Ahirs (cattle herders) is performed only by men during the eleven days after Diwali, in order to honor their cattle. Some stand, playing traditional drums, flutes or cymbals. Others dance, crouching and holding sticks, decorated or not. In the Jabhua region of Madhya Pradesh, the Bhils have many dances that can be formalized, with dressed-up women dancing in a circle around the male musicians, or not. The women may dance with nothing in their hands or wave different sorts of scarves, whereas the men will be armed with big sickles, swords, or bows and arrows. Those dances are performed for weddings as well as local festivals and ceremonies. In the Pachmarhi area of the Satpura Hills of Madhya Pradesh, local tribes (Gonds and Korkus) perform many dances during different festivals of local tribes, for example the Danda, in which each dancer (only men do this) holds two sticks in each hand and knocks them together as they dance. One of the main forms of tribal art among the Korkus is funerary art, that is, mundas (carved wooden boards). The Korkus’s munda ceremonies must be organized at some time within the ten years following somebody’s death. A munda is a carved wooden “memory board,” so called because it is made and deposited in honor of the deceased and his/her memory. It must be deposited at the foot of a sacred tree where hundreds of such boards have piled up over time. The family and members of the tribe bring offerings of spirits and tobacco as well as chickens to be sacrificed during the ceremony. While all this is taking place, young men and women make music and dance nearby in a circle, having walked there from their village, singing, making music with several types of drums and cymbals, and dancing. In the south of Chhattisgarh, the Gonds likewise perform music and dances for funerals. After a person’s death, drums will be played for a whole day and night to inform the people of the village. After the arrival of relatives, friends, and local villagers, the dead body will be taken to the burial or cremation ground. Friends and villagers play drums, sing, and dance in honor of the deceased until they reach the burial/cremation ground. For ten days after the day of the person’s death, his or her relatives will cook his or her favorite food and keep it outside the house. On the tenth day they will invite the village to an expensive feast, where family members dance with dried animal carcasses on their shoulders to the sound of drums and other instruments. The procession then goes to the burial or cremation ground; providing occasion for other animal sacrifices and wine libations under the direction of their shaman. The memory board will be made in the years that follow. It used to be made out of wood, but nowadays it is made of stone or concrete. Its shape is more or less rectangular with a carved top. It is stuck vertically into the ground not far from the roadside. Both sides are painted with scenes from everyday life, among which music (drummers) and dancing have a prominent place. Though it is unsurprising that music and dances should be a crucial aspect of tribal culture, certain dances may also look far less “serious” to some outsiders, even if they are serious for the people concerned. Thus Elwin writes, about the Baiga tribe in the Dindori area: “They also greatly enjoy a dance that they call the tamasha. A boy makes a cloth penis of enormously exaggerated length and goes round and round with obscene movements and gestures, wagging the cloth at everyone he meets. Sometimes, all the boys dance together in a round, each holding his penis in his hand, while the little girls look on, screaming with affected embarrassment and delight” (Elwin 1951: 232).

Conclusion Considering that music and dance play such a prominent role in tribal life, it is not surprising that we should find them so often represented in rock art. Songs, music, and dance unite the participants with the spirit world. “Much tribal dancing has a serious ritual purpose and there are legends which ascribe its origins to the need of keeping the gods happy and amused” (Elwin 1951: 131). Therefore, in India one cannot separate music from any aspect of life, as its importance is reflected in every activity. Songs accompany ceremonial rit-

MUSIC AND DANCING SCENES IN THE ROCK ART OF CENTRAL INDIA

133

ual hero or Gatha tales, god/goddess worship, routine activities, game playing, courtship and love affairs, war preparations, hunting trips, healing ceremonies, shamanism and its various practices, and so on. The images represented in rock art may have different purposes, as we know from modern ceremonies that dances, for example, may be performed for any event in the year: curing rituals, weddings, births, mourning, before and after hunting, insuring prosperity, on auspicious days, et cetera. We have no way, though, of saying precisely whether a dance scene painted in one or another shelter is meant for any particular precise purpose. Meenakshi Dubey-Pathak is a specialist in the field of Indian Rock Art, which originated in the Mesolithic (more than 10,000 years ago) or even before. During her many years of fieldwork she has discovered dozens of new painted sites, mostly in Madhya Pradesh (especially the Pachmarhi area, which was the subject of her PhD), but also in Chhattisgarh, Rajasthan, and Ladakh. She has devoted nearly thirty years of her life to discovery, study, publication, exhibitions, workshops, and protection of Indian rock art. She has published many papers and three books: Rock Art of Pachmarhi Biosphere (Delhi, BR Publications, 2012), Des Images pour les Dieux: Art rupestre et Art tribal dans le Centre de l’Inde (with Jean Clottes, Éditions Errance, Arles, France, 2013), and Powerful Images: Rock Art and Tribal Art of Chhattisgarh (with Jean Clottes, Bloomsbury Publications, 2017). Her latest, Madhya Pradesh Rock Art and Tribal Art (with Jean Clottes) is in press. She is a Wakankar Senior Research Fellow and in 2014 was awarded the high honor of Chevalier des Arts et Lettres by the French Ministry of Culture and Communication. She has worked as an international expert on rock art with ICOMOS and UNESCO and is a member of the Bradshaw Foundation Advisory Board (England). She was awarded the Bourse du Patrimoine, 2019, by the Ministry of Culture, France. Jean Clottes studied at Toulouse University. He was Director of Prehistoric Antiquities for Midi-Pyrénées from 1971 onward, and after being a General Inspector for Archaeology at the Ministry of Culture (1991), where he became a Scientific Advisor for everything relating to prehistoric rock art, a position he held until his official retirement in July 1999. He is an International Expert for rock art with ICOMOS and UNESCO and has taught courses at the Universities of Toulouse (France), Neuchâtel (Switzerland), Gerona (Spain), Buenos Aires (Argentina, INAPL), Berkeley (USA), and Victoria (Canada). He has published or co-published twentyeight books, directed thirteen others and published more than five hundred papers. Six of his books (and many papers) have been published in English, including The Cave Beneath the Sea (Harry Abrams, 1996); The Shamans of Prehistory (with D. Lewis-Williams, Harry Abrams, 1998); World Rock Art (Getty Foundation, 2002); Chauvet Cave: The Art of Earliest Times (ed.) (The University of Utah Press, 2003); Cave Art (Phaidon, 2008) and Powerful Images. Chhattisgarh Rock Art and Tribal Art (with M. Dubey-Pathak, Bloomsbury, 2017). He is currently the editor of the International Newsletter on Rock Art (INORA).

References Allchin, B. 1958. “Morhana Pahar: A Rediscovery.” Man 58: 153–55. Anderson, C. W. 1918. “Rock Paintings in Sighanpur.” Journal of the Bihar and Orrisa Research Society 4(2): 298–306. Clottes, J., and M. Dubey-Pathak. 2013. Des Images pour les Dieux. Art rupestre et art tribal dans le Centre de l’Inde. Arles: Éditions Errance. Cockburn, J. 1883. A Short Account of the Petrographs in the Caves and Rock Shelters of the Kaimur Range in Mirzapore District. Proceedings of the Asiatic Society of Bengal, 52, 125–26. Dubey-Pathak, M. 2013. Rock Art of Pachmarhi Biosphere: Mesolithic to Historic Times. Delhi: B.R. Publishing. ———. 2014. “The Rock Art of the Bhimbetka Area in India.” Adoranten 2014: 5–22. Dubey-Pathak, M., and J. Clottes. 2017. Powerful Paintings: Rock Art and Tribal Art in Chhattisgarsh. Delhi: Bloomsbury. Elwin, V. 1951. The Tribal Art of Middle India. Bombay: Oxford University Press. Gordon, D. H. 1939. “Indian Rock Paintings.” Science and Culture 6: 1–33. Ghosh, M. 1932. Rock Paintings and Other Antiquities of Prehistoric and Later Times. Delhi, Memories of the Archaeological Survey of India, Vol. 24. Mathpal, Y. 1984. Prehistoric Rock Paintings of Bhimbetka, Central India. New Delhi: Indira Gandhi National Centre for the Arts. Pandey, S. K. 1993. Indian Rock Art. New Delhi: Aryan Books International. Silberrad, C. A. 1907. Rock Drawings of Banda District. Journal of the Royal Asiatic Society 5(3–1): 567–70. Wakankar, V. S., and R. R. Brooks. 1976. Stone Age Painting in India. Bombay: Taraporevala and Sons.

==== •

•

•

9. HUNTING AND HAVOC Narrative Scenes in the Black Desert Rock Art of Jebel Qurma, Jordan Nathalie Østerled Brusgaard and Keshia A. N. Akkermans

Introduction Pictorial and textual engravings can be found in vast numbers across the Black Desert of Northern Arabia, a basalt desert that stretches from southern Syria through northeastern Jordan into northern Saudi Arabia. The carvings were made by nomadic peoples inhabiting the desert in the late first millennium BC and early first millennium AD. The rock art is figurative in nature, depicting anthropomorphic figures such as archers and women, zoomorphic figures such as dromedaries, horses, lions, and ibex, as well as various geometric designs. The figures are depicted individually, accumulated on panels, and in scenes interacting with one another. The inscriptions, written in the Ancient North Arabian Safaitic script, are intrinsically linked to the pictorial engravings. A common composition is a rock art figure or scene associated with an inscription in which the author states his or her name and genealogy and “signs” the image. 1 Some texts also contain a narrative component in which the author states, for example, that he pastured his camels, migrated to another area, spent the winter in a particular place, or mourned the loss of a loved one. Based on these unique insights into the authors’ lives, the image emerges that these peoples were nomads who moved through the desert, subsisting at least in part on owning dromedaries and possibly ovicaprids and horses, built cairns for their dead, and worshipped a range of deities (Al-Jallad 2016; Macdonald 1992, 1993, 2006, 2012). However, many questions about these societies remain, in particular how they operated in the desert landscape, what the nature of their ideology was, and what the role of these desert carvings was. The potential of the rock art in addressing these issues remains underutilized. A few notable studies have been conducted on particular motifs, such as women (Macdonald 2012) and equids (Macdonald 2019). But other than these exceptions, little is known about the imagery, in part because of the lack of complete and systematic surveys of the petroglyphs. As a result, much remains to be investigated about the rock art and the insights it can provide into the societies that created these carvings (cf. Brusgaard 2019). One important, still neglected aspect is the narratives that have been created in the Safaitic rock art through the interaction between figures. The carvings depict anthropomorphs and zoomorphs together in scenes that appear to represent hunting, combat, and pastoral activities. This chapter presents the results of the first indepth study of these scenes depicted in Safaitic rock art, based on a dataset of rock art from the Jebel Qurma region of the Black Desert in northeastern Jordan. In particular, we focus on the two most common type of scenes—those of hunting and of conflict—discussing the composition of and patterns in these scenes. This chapter also looks at the weaponry depicted in the scenes and rock art in general, as detailed study of the representation of objects can provide valuable additional information on the material culture of past societies (May et al., 2017), especially in regions with few archaeological remains. Through this detailed study, we examine the patterns and themes in the scenes and use of weapons, investigating what they can tell us about both the rock art and the desert societies in the late first millennium BC and early first millennium AD.

Safaitic Engravings The Jebel Qurma region lies in the northeast of Jordan, approximately 30 km east of Azraq. It is part of the Black Desert, which is characterized by basalt-covered uplands, known locally as the harra, and surrounding

HUNTING AND HAVOC

135

limestone plains, or hamad (Figure 9.1). Since 2012, the Jebel Qurma Archaeological Landscape Project of Leiden University has been carrying out surveys and excavations in this region, investigating the archaeological remains, rock art, and inscriptions. The pictorial and textual engravings are “Safaitic,” which describes a pre-Islamic script and associated rock art from the late first millennium BC and early first millennium AD in Northern Arabia. The Safaitic engravings have been conventionally dated from the first century BC to the fourth century AD based on references to known historical events in some inscriptions; however, these can only be seen as a tentative guideline (Al-Jallad 2015: 18). Medieval and modern Arabic engravings and wusu¯m—late twentieth-century “tribal marks”—

FIGURE 9.1. Location of the Jebel Qurma area in northeastern Jordan, in the Black Desert or harra. Copyright Jebel Qurma Archaeological Landscape Project.

136

NATHALIE ØSTERLED BRUSGAARD AND KESHIA A. N. AKKERMANS

have been found in the Jebel Qurma region as well, but there have been no findings of clearly prehistoric rock art such as that known from the Black Desert area northeast of the Jebel Qurma region (cf. Betts 1987). The recent surveys of the Jebel Qurma region have led to the discovery and documentation of more than 4,500 individual Safaitic petroglyphs and more than 5,400 Safaitic inscriptions. In the rest of the basalt desert, over 40,000 inscriptions have been recorded since their discovery in 1858 (Al-Manaser and Macdonald 2017), and it is likely that almost as many rock art depictions exist. The content of the pictorial and textual engravings across the harra is quite homogeneous, although there does seem to be some stylistic diversity. More cross-regional comparisons are necessary to investigate this, especially for the rock art. The rock art depicts a large number of zoomorphic motifs and a few different anthropomorphic motifs. Geometric motifs occur as well, the most common being sets of lines or dots. Three-quarters of the figures in the Jebel Qurma corpus are zoomorphic. There are domestic animals, such as dromedary camels, equids, and dogs, of which the camel motif is by far the most common, and wild animals such as wild asses, oryx, ibex, ostriches, and lions (Brusgaard 2019). Anthropomorphic figures make up less than 10 percent of the figures; the majority of these are archers (i.e., figures holding a bow and arrow). The engravings occur in different types of compositions. The rock art depiction is commonly accompanied by an inscription stating the author’s name and referring to the image (Figure 9.2). The images can feature

FIGURE 9.2. An engraved panel from Jebel Qurma with a view of the desert landscape in the background. The panel features three Safaitic inscriptions, two of which refer to ‘the she-camel’. On the left a carnivoran, four ibex, and an ostrich are depicted. In the scene on the right, a human figure holds a camel that is nursing her infant. Copyright Jebel Qurma Archaeological Landscape Project.

HUNTING AND HAVOC

137

a single figure, such as a dromedary, or a composition of different anthropomorphic, zoomorphic, and/or geometric figures in an assemblage or scene. A scene is defined here as a composition that “reflects an action, usually with a defined theme, that can be described even if the meaning and theme are unknown” (May and Domingo Sanz 2010, 37). Therefore, in this study a scene is defined as two or more figures interacting with one another and reflecting an action, for example, hunting or fighting. Following these criteria, 168 scenes were identified in the Jebel Qurma corpus, featuring a total of 657 figures. Five different types of scenes could be recognized: hunting, conflict and combat, leading, nursing, and mating (Figure 9.3, Table 9.1). Two additional scenes could represent an erotic scene between anthropomorphs. Though there is one scene that might depict anthropo- TABLE 9.1. The types of scenes depicted in the Jebel Qurma rock art and morphs dancing, they could not be their frequency. identified as such with certainty. A % of total N number of scenes that have been Type Clear Unclear Total of scenes found in other Safaitic rock art 93 7 100 59.5% appear to feature music making Hunting (Macdonald 2012) and plough- Conflict and combat 24 3 27 16.1% ing (Ababneh 2005; Al-Manaser Leading 19 0 19 11.3% 2008), but none such have been 16 2 18 10.7% found so far in the Jebel Qurma Nursing region. In this chapter, we briefly Mating 1 0 1 0.6% discuss the leading, nursing, and 0 2 2 1.2% mating scenes before turning our Erotic attention to the two most com- Dancing 0 1 1 0.6% mon types of scenes: hunting and Total 153 15 168 100.0% conflict/combat.

FIGURE 9.3. Stacked bar chart showing the different types of scenes and the number of times they occur in the Jebel Qurma corpus. Scenes that could clearly be identified are displayed in dark gray; scenes that were unclear are displayed in light gray. The “Unknown” category features compositions that are clearly scenes, but of which the type is unknown. Copyright Nathalie Østerled Brusgaard and Keshia Akkermans.

138

NATHALIE ØSTERLED BRUSGAARD AND KESHIA A. N. AKKERMANS

Pastoral Scenes The leading, nursing, and mating scenes could all be categorized as pastoral scenes, scenes that depict herding narratives. The majority feature dromedary camels, either interacting with one another or interacting with a person. The leading (or holding) scenes depict a person leading or holding a domestic animal by what appears to be a lead rope. Most of these scenes–10 in total–feature a person holding or leading a dromedary camel. In many of these cases, the anthropomorphic figure appears to be holding rather than leading the camel as he or she is facing it. In five of the scenes, the camel also has his or her leg hobbled, indicated by a raised foreleg. There are also four scenes in which the person is holding or leading either a dromedary or an equid with a rider. 2 In one scene, a person is holding a female dromedary camel while a young camel nurses from it (Figure 9.2). The nursing scenes portray female zoomorphs, primarily camels, with a young between their legs with its head facing up towards the mother’s belly as if drinking from her. There are 15 of these nursing scenes and two possible ones. In nine of these cases, the mother’s udders are also depicted. In the scenes featuring camels, the mother is generally depicted in detail while the young is simple and lacks detail. Lastly, there is one scene that appears to depict mating between two dromedary camels. It shows a male camel, depicted with phallus, partially on top of a female camel.

Hunting Hunting scenes are classified by the depiction of humans hunting animals or animals hunting other animals (figure 9.4). They are the most common of all the scenes; there are 93 clear hunting scenes and an additional seven that probably depict hunting. There is a total of 467 figures featured in these scenes. Hunting scenes thus make up 59.5 percent of the scenes, making them the most common theme in the rock art. On average, hunting scenes feature a large number of figures compared to the other types of scenes; some scenes feature up to between 10 and 21 figures. The scenes show two different types of hunting with regard to the method used to hunt by humans or animals: solitary and cooperative hunting. Solitary hunting is classified as one predator (human or animal) hunting the prey. There are four variations on solitary hunting in the rock art: a solitary human, a solitary human on a mount (usually an equid), a human hunting with a dog, and a solitary animal hunting. Cooperative hunting is classified as two or more predators hunting the prey. There are four similar variations: two or more humans hunting, one or more human(s) hunting together with a person on a mount, two or more humans hunting with dogs, and two or more animals hunting together (Brusgaard 2019). Classifying the hunting scenes by the hunting method reveals that the type and number of prey vary depending on the type and number of hunters (Brusgaard 2019, Table 4.18). Altogether, solitary hunting scenes are more frequent than cooperative hunting scenes (77 scenes versus 22). The most commonly depicted type of hunting is the solitary human on foot. The second most common type is the single person on a mount, followed closely by solitary animals hunting. The solitary human hunter is almost always an archer and they are commonly depicted either hunting the wild ass or a bovid that could be either an ibex or gazelle.3 Interestingly, the wild ass is almost exclusively hunted by the solitary archer (Figure 9.4b). In comparison, the human on a mount (usually a horse or mule) is most frequently portrayed hunting an oryx (figure 9.4c). Most of the solitary animal predators are canids, but there are also a few more generic-looking carnivorans and one lion hunting. The most common prey is the ostrich, almost always depicted in flocks, followed by the ibex/gazelle. There are only six scenes featuring a solitary person hunting with a dog. In the majority of these scenes, the human and dog are hunting flocks of ostriches. There are no scenes of a single person hunting with more than one dog. The most common type of cooperative hunting depicted is animals hunting together; the majority of such scenes depict a pack of canids (Figure 9.4d). They are often shown hunting a flock of ostriches, but occasionally also ibex/gazelles and oryx. There are three scenes in which lions are hunting together. One depicts two lions attacking a camel, while the other two show two lions hunting a flock of ostriches. The second most common type of cooperative hunting is humans hunting together with dogs, depicted in seven scenes (Figure 9.4a). The majority of these scenes feature archers, usually two but sometimes three or four,

HUNTING AND HAVOC

139

a

b

c

d

FIGURE 9.4a–d. (a) A cooperative hunting scene depicting three archers (on the left) and two dogs (on the right) hunting a herd of bovids, either ibex or gazelles. The bovids have striped patterning on their bodies. Copyright Jebel Qurma Archaeological Landscape Project. (b) A solitary hunting scene in which an archer, upside down, faces a male wild ass. The seven dots on the right are a common motif in the rock art. The inscription on the right is made using the same technique as the archer and refers to the wild ass (“the ass”). Copyright Jebel Qurma Archaeological Landscape Project. (c) A rider on an equid hunts a lone oryx, possibly with the help of an archer (bottom right). The inscription, only partially visible here, refers to “the she-ass.” The equid is therefore probably a female mule. It also has patterning on its body. Copyright Jebel Qurma Archaeological Landscape Project. (d) Two canids surround a flock of ostriches. The inscription states the carver’s name. Copyright Jebel Qurma Archaeological Landscape Project.

together with one dog. In contrast to the single hunter with a dog, they are usually hunting bovids and rarely ostriches. In addition, there are five scenes of humans hunting on foot together with a person on a mount. In most cases, they are hunting oryx. Lastly, there is one scene depicting a large group of people on foot and one person on an equid attacking a lion. The least frequent of all the types of hunting scenes is the cooperative hunting between only humans on foot. There are only two scenes that depict this. Both scenes feature two people hunting one or two lions. It is also possible that these scenes depict people defending themselves from lions rather than hunting them. The most frequently hunted animal is the ostrich and the most common hunter is the archer. Canids are also depicted often—28 times. It is important to note that many of these may represent dogs as they look very similar to them, but because there is no human depicted, it cannot be said with certainty that it is a

140

NATHALIE ØSTERLED BRUSGAARD AND KESHIA A. N. AKKERMANS

domestic canid. Archers occur most frequently; they are depicted in 44 different scenes. They are followed by ostriches (29 scenes), oryx (24 scenes), and equids with riders (20 scenes). Dromedary camels, with or without rider, are rare in the hunting scenes.

Conflict and Combat Twenty-four scenes depict conflict and combat, and three scenes also appear to do so but are unclear. Two of these scenes differ from the rest in that they feature fighting between animals. Both depict two male camels that appear to be fighting. The camels are facing each other with their necks crossed in the manner seen when bull camels fight. The other twenty-four conflict and combat scenes all feature humans with weapons. These scenes can be broadly divided into two categories: scenes depicting general combat or conflict, and scenes depicting raiding. The first category features narratives of fighting that take place in contexts that are unclear. There are thirteen scenes in which this is the case. Two of the scenes depict people on horseback fighting each other, and one depicts two archers fighting each other. Ten scenes feature a person on a mount fighting one or more anthropomorphs on foot (Figure 9.5a). In the majority of these scenes, the person’s mount is an equid, either a horse or a mule. The riders are always either holding what could be a lance or spear (see Weaponry below) or apparently not holding any weapon. The second category is scenes that appear to depict raiding. The existence of raiding as an ‘activity” among the desert nomads is evidenced by the Safaitic inscriptions (cf. Al-Jallad 2015; OCIANA 2017). For example, ‘By Ngs² son of ʿm son of Grm and he was on a raid so and he was lying in wait’4 and ‘By ʾnʿm son of ʿbt. and he grieved for his raiding party’.5 In the Jebel Qurma corpus, there are several attestations of raiding (Della Puppa forthcoming). One of these is associated with an image depicting three riders with weapons on camelback. The camels appear as if in movement. Macdonald (1990) has also identified raiding scenes in Safaitic rock art based on the occurrence of scenes in which a person on a mount is touching a camel with his or her spear. Macdonald (1990) argues that these are not scenes depicting the hunting of camels, but the raiding of camels, whereby the victor touches his “booty” with a spear to indicate that it is his. This interpretation is the most plausible considering the epigraphic evidence for raiding and the broad time period to which this rock art belongs, within which wild camels no longer existed in Arabia and the dromedary camel had already long been domesticated (Almathen et al. 2016; Rosen and Saidel 2010). In the Jebel Qurma corpus, there is one scene very similar to the ones described by Macdonald. In this scene, two riders on dromedary camels are depicted alongside a male dromedary camel without rider. One of the riders is touching the male dromedary with what appears to be a spear (figure 9.5b). Following on the argument laid out by Macdonald (1990), this scene depicts a raid in which a camel is being claimed. According to the epigraphic evidence, raiding occurred among these desert societies and the pictorial engravings appear to support what is mentioned in the texts. Following on this evidence and the argument of Macdonald (1990), it may be possible to identify some of the combat and conflict scenes as representations of raids. There are eight scenes in which people are fighting each other around a dromedary camel. In several of these, the camel is also being held by an anthropomorph, as if being held back from the raiders (Figure 9.5c). In three of them, a person on horseback forms part of the attack (Figure 9.5d). Between three and nine anthropomorphs are involved in these scenes, either facing each other with weapons as if actively fighting, holding the camel, or riding the camel or equid. We propose that these are all scenes that portray a conflict or combat in the context of a raid. In all of these scenes, the anthropomorphic figures are depicted in small dimensions and lack any detail. The dromedary camel in the scenes are portrayed large, detailed, and in the centre of the scene. They therefore draw the visual focus of the scene.

Weaponry Many of the hunting and conflict/combat scenes feature the use of weaponry in them. They are not depicted prominently in the rock art and, like the anthropomorphs, are rarely depicted in detail. Yet a close study of the weapons can provide interesting insights. This has been demonstrated by notable works such as that by

HUNTING AND HAVOC

141

a

b

c

d

FIGURE 9.5a–d. (a) A male horse and its rider, armed with a lance, face a man with a bow and arrow. The inscription refers to “the horse(man).” Copyright Jebel Qurma Archaeological Landscape Project. (b) A possible raiding scene. One camel rider touches the dromedary camel with his or her lance/spear, possibly “claiming” the dromedary as booty. Copyright Jebel Qurma Archaeological Landscape Project. (c) A conflict scene that might depict raiding. Two thinly incised archers holding shields are on the bottom left. One appears to be holding the dromedary camel by a long lead rope. The dromedary’s foreleg is hobbled, indicated by the raised leg. A figure on an equid and another thinly incised archer appear to be the attackers. The rider also has a bow and a shield and a quiver on his or her back. It is the only figure on horseback with a bow. Copyright Jebel Qurma Archaeological Landscape Project. (d) Another possible raiding scene. A figure holds the dromedary camel and a bow, facing toward the two riders on horseback. Both riders have a lance in their hands. Another archer appears to be standing on the dromedary’s back or next to it, perhaps trying to ward off the attackers. A smaller dromedary stands underneath the larger one. On the left is a small hunting scene of an archer hunting an oryx. Copyright Jebel Qurma Archaeological Landscape Project.

Bradley (1997, 1998), May et al. (2017), and Ling and Cornell (2017). However, what many of these studies have in common is that there are archaeological finds for comparison (e.g., Bradley, 1997; Ling and Cornell, 2017) or valuable ethnohistorical information (e.g., May et al., 2017). Additionally, most investigated weaponry depictions are detailed and featured prominently in the rock art corpus, allowing for thorough analyses. This is not the case in the Safaitic rock art, and moreover there are few material finds of weaponry with which to compare, mainly due to the poor preservation conditions of the desert (cf. Akkermans and Brüning 2017). Still, the rock art corpus of the Jebel Qurma area is very suitable for a systematic study, as will be illustrated below. The depictions of weapons in the rock art can thereby provide information on a form of material culture that has perished over time.

142

NATHALIE ØSTERLED BRUSGAARD AND KESHIA A. N. AKKERMANS

In the Jebel Qurma rock art corpus, a total of 563 anthropomorphic figures are depicted, including those mounted on an animal. Of these, 271 are depicted holding an object, almost all of which are weapons. Fifty-eight of the objects could not be identified for various reasons. In general, neither the objects nor their users are depicted in much detail; therefore, identifying some of the objects, and especially weapons, can be difficult. However, it was possible to distinguish six material categories in the objects: “bow,” “lance/spear,” “shield,” “sword,” “lead ropes,” and “whips.” These types of objects can be seen being used on their own and combined with one another. Lead ropes and whips are used by humans only for or on animals, and are therefore not categorized as weapons and excluded from this study. The shields, although not technically a weapon, are considered part of the weaponry equipment as defensive gear. This leaves 257 figures holding a weapon, a combination of weapons, or an unclear object, the majority of which are probably weapons (Figure 9.6a). Weapons are depicted being used by both anthropomorphs on foot (120 figures) and figures mounted on an animal (79 figures) (Table 9.2). The majority of these are equids, probably either horses or mules/hinnies, but there are a few riders on dromedary camels. Most of the objects held by anthropomorphic figures are bows. Ninety-nine figures are handling a bow and arrow; only one is on horseback. In eight cases, the anthropomorphic figure is holding a bow and has a quiver on his/her back (Figure 9.6b). The shape of the bows is very uniform; the bows are consistently and without exception depicted in an “m” shape. This “m” shape is reminiscent of the shape of the so-called “double-recurve composite bow.” The earliest appearance of composite bows in Mesopotamia dates to the third millennium BC (Miller, McEwen, and Bergman 1986: 183). Literary evidence suggests that from the second half of the second millennium BC onward, the composite bow can also be observed in the Levant (Zutterman, 2003, 123). The shape of the double-recurve composite bow is the result of its structural composition: in order to withstand and adapt to the pressure and tension on the bow’s body when the bow is drawn, several different types of material were used, such as wood, bone, horn, and sinew (Bowden 2012: 44). The design and the choice of materials makes the composite bow much more efficient than the preceding technologies of the self-bow or the laminated bow (Loades 2016: 5). The fabrication process of double-recurve composite TABLE 9.2. The types of objects and combination of objects held by anthropomorphs on foot and held by riders. N of anthropomorphs on foot

% of anthropomorphs on foot with objects

86

58.1%

0

0.0%

Bow + lead rope

1

0.7%

0

0.0%

Bow + quiver

8

5.4%

0

0.0%

Bow + shield

3

2.0%

0

0.0%

Bow + quiver + shield

0

0.0%

1

0.9%

Lance/spear + lead rope

2

1.4%

0

0.0%

Shield

1

0.7%

0

0.0%

Shield + uncl. object

1

0.7%

0

0.0%

Lance/spear

1

0.7%

67

61.5%

Lance/spear + shield

6

4.1%

6

5.5%

Lance/spear + sword

0

0.0%

1

0.9%

Sword

5

3.4%

1

0.9%

Sword + shield

6

4.1%

3

2.8%

28

18.9%

30

27.5%

Total

148

100.0%

109

100.0%

No object

124

168

Total anthropomorphs on foot/riders

272

277

Object Bow

Uncl. object

N of riders

% of riders with objects

HUNTING AND HAVOC

143 b

a

c

e

d

FIGURE 9.6a–e. (a) Stacked bar chart showing the types of weapons, including combinations, and number of anthropomorphic figures holding them. The unidentifiable objects (“uncl.”) are also included. Figures mounted on an animal are displayed in light gray, and figures on foot are indicated by dark gray. Copyright Nathalie Østerled Brusgaard and Keshia Akkermans. (b) Depiction of an archer with a double-recurve composite bow, which is characterized by its “m” shape. Also note the quiver carried on the figure’s back, with the arrows sticking out from the top. Copyright Jebel Qurma Archaeological Landscape Project. (c) Figure with a spear/lance seated on an equid. This object is most likely to be a lance due to its length in relation to the figure holding it. To the right a figure is depicted holding the equid by a lead rope. Copyright Jebel Qurma Archaeological Landscape Project. (d) Rider on an equid holding a lance and carrying a sword at the waist (the sword is traced in dark gray for clarity). Note the way the sword is carried at the waist and the short diagonal bar crossing the shaft. The rider is hunting an oryx. Copyright Jebel Qurma Archaeological Landscape Project. (e) Example of a small round shield with a cross motif on it. The other object the anthropomorph is holding is probably a spear rather than a sword, based on the context of this figure (depicted in a scene of several anthropomorphs hunting/attacking a lion). Copyright Jebel Qurma Archaeological Landscape Project.

144

NATHALIE ØSTERLED BRUSGAARD AND KESHIA A. N. AKKERMANS

bows was a lengthy one, and the making of a proper composite bow would have likely demanded a considerable degree of patience, experience, and specialization on the part of craftsmen (Bowden 2012; Loades 2016; Miller et al. 1986). Taking these factors into account, it is likely that composite bows were made in a sedentary environment in large batches of several hundred at a time (Miller et al. 1986). The second largest weapon category is that of the lance/spear: eighty-two anthropomorphic figures are holding what could be a spear or lance. Interestingly, by far the majority of these are held by anthropomorphs seated on a mount (74 figures versus 9 on foot) (Figure 9.6c). As is already suggested by the name given to this weapon category, differences in length and mode of use suggest more than one subcategory of these “stick weapons.” According to Potts’s (1998: 18) criteria, “spear” refers to a “light projectile which could be thrown over a considerable distance at an enemy and for which the term ‘javelin’ is sometimes employed.” Conversely, a “lance” is “a much heavier and longer weapon, which, although it could be thrown at a short distance, was more commonly hand-held and used for thrusting in close combat” (Potts 1998: 183). The majority of the lances/spears depicted in the rock art range from fairly long to very long in relation to the figures using them and are therefore probably lances, whereas the shorter spears/lances are more likely to represent spears instead. In addition, the objects that are likely to be lances have a stronger association with figures on equids and camelback. The potential spears are most often held by figures on foot and are more often accompanied by a shield. Both the spears and the lances are used in hunting scenes as well as conflict/combat scenes. The last and most problematic weapon category is the category of “swords.” Due to the lack of detail in the anthropomorphic figures and weapon depictions, it is difficult to distinguish swords from the lance/spear category. One identifier might be that if the object is being held at the outer end instead of the middle and is relatively short in length, then it might be a sword. A more convincing depiction is a figure on horseback who is carrying a stick-like object at the waist (Figure 9.6d). In this depiction, a shorter bar is depicted diagonally crossing the shaft. This short bar probably portrays the cross guard of a sword. Further strengthening the assumption that this is indeed a sword are the location and position in which the object is carried on the body: the object is hanging from the hips as a sword would do when hanging from a sheath. Macdonald (2012: 282) has also observed that in Safaitic rock art, figures on horseback are rarely seen wielding a sword but occasionally appear with a sword at their belts. Based on these criteria, sixteen swords were identified in the rock art, but these results are tentative. Lastly, a total of twenty-seven figures are depicted with a shield in hand. Most shield use is accompanied with a weapon, the majority of which are lance/spears and swords. Despite the fact that it seems near impossible to fire a bow while holding a shield, three figures are carrying both a bow and a shield. All shields are small and round, but the patterning on the shields varies (Figure 9.6e). Most of the shields have patterns carved into them, ranging from cross-hatching, crosses, and radiating lines to feather-like carvings and circles. Cross-hatching might imply leather slabs, while the circles are most likely depictions of the shield boss or umbo, the system that attaches the grip of the shield to the shield using a convex, round piece of material in the center of the shield. Lines and cross-hatching are not limited to the shields; these patterns are also sometimes carved into the bodies of anthropomorphic and zoomorphic figures (Figure 9.4a and c) (Brusgaard 2019). It is therefore unclear whether the patterns are of a functional or decorative nature. Of the 199 weapon depictions, 126 of them are shown in the context of either a hunting or a conflict/ combat scene. In hunting scenes, the bow and arrow is the most prevalent weapon: 57 archers, 30 lance/ spear wielders, and only 4 (albeit questionable) swordsmen are shown in hunting scenes. Aside from three depictions of probable spear wielders on foot, all figures associated with lances/spears are riders, whereas none of the archers or swordsmen are mounted. The type of weapon used and its user appear to depend on the hunting technique being used, as described above. In the combat/conflict scenes, the preference for certain weapon types is less pronounced. Bows and lances/spears are depicted equally often, each with a total of fifteen depictions. Only one of the archers is mounted, while ten of the lance/spear wielders are mounted. Swords are depicted on only five occasions, all used by figures on foot. The more heterogeneous distribution in the combat/conflict scenes might be explained by taking into account that different weapon types are suitable depending on the distance of fighting. The sword is most useful at short range, thrusting weapons such as the lance are best at medium range, and the spear and bow and arrow are optimal weapons at long range.

HUNTING AND HAVOC

145

Discussion Reviewing the results of our study on the scenes and depictions of weapons in the rock art of the Jebel Qurma region, we can draw a number of conclusions. First, there are three dominant themes within the scenes. One theme is scenes of a pastoral or domestic nature, including young animals, mostly dromedaries, nursing from their mother and humans leading or holding animals, again mostly dromedaries. Scenes with a pastoral theme make up approximately 22 percent of the 168 scenes. The second theme is combat and conflict, which consists of scenes featuring horsemen fighting each other, a horseman fighting one or more anthropomorphs on foot, and scenes that appear to depict raiding. The third and most dominant theme is hunting, which includes humans hunting animals and animals hunting other animals. Both solitary and cooperative types of hunting are identifiable in the scenes. The former are more common. The second observation that can be drawn from the results is that out of all the weaponry depicted in the rock art, there are clearly two main weapons: the bow and arrow—most likely the composite bow—and the lance/spear. The majority of the latter are probably lances used by riders; probably only a minority are spears, used on foot. In total, excluding the unclear objects, there are 199 anthropomorphic figures holding weapons. Of these figures, 99 are holding a bow (or a combination of a bow and another weapon) and 83 of them are holding a lance/spear (or a combination of a lance/spear and another weapon). Swords are difficult to identify but may be depicted in at least six instances. Shields are usually seen in combination with another weapon. They are often carved with patterns, but the meaning of these patterns is still unclear. The majority of the depicted weapon use is shown in the context of a hunting or conflict/combat scene. The type of weapon used seems to depend on a number of factors: in the hunting scenes the bow is the primary weapon of choice, whereas in combat/conflict scenes the distribution of weapon types is much more equally dispersed. This brings us to our third observation, based on the hunting and combat/conflict scenes and the weaponry. Although there is some variation within these scenes and within the use of weapons, there are a number of clear patterns. There is an apparent distinction between the use of the bow and the use of the lance/ spear. The former is almost exclusively used by a person on foot. It is used in some combat and conflict scenes by humans fighting one another, but it is primarily depicted in hunting scenes. In these scenes, we tend to see the lone archer hunting a wild ass or an ibex/gazelle, or a group of archers hunting with or without the help of dogs. The lance/spear is almost exclusively used by a person on horseback, in which case these weapons probably depict lances. They are often used in combat/conflict scenes where the rider is facing a person on foot. Additionally, they are often used by the solitary hunter on horseback, who in the majority of cases is hunting a lone oryx. Regarding the scenes depicting animal hunters, there is a dominant theme as well: the solitary hunter and the pack hunters, the majority of which are canids, tend to be depicted hunting flocks of ostriches. Thus, distinct patterns are observable in the depictions of hunting and fighting in the Jebel Qurma rock art, from the type of hunter or fighter portrayed and his or her weapon, to the type of hunting technique and the prey being hunted. These findings provide interesting new insights into the Safaitic rock art and the people who carved it. Our study on the scenes and weapons depicted in the Jebel Qurma rock art shows that these petroglyphs fit well into the historical context of the area and furthermore provide new information, outlined below.

Historical Context The way in which weapons are used in the rock art at Jebel Qurma often coincides with descriptions in contemporary and later texts. Early Arabic poetry specifically mentions all three of the main weapon categories (bows, swords, and lances/spears) seen in the rock art corpus of the Jebel Qurma area (Schwarzlose 1886: 45). In particular, in pre-Islamic poetry (oral poetry thought to originally have been composed in the sixth and seventh centuries AD and later written down in the eighth century), the hunt is a frequent subject matter, with the archer and his faithful hounds and their prey, usually oryx or ostriches, playing important roles (Smith 1990; Stetkevych 1999). Additionally, the preference for lances as the primary weapon of choice while riding is well attested in Classical Greek and Roman writings (Gordon 1953; Potts 1998), so the

146

NATHALIE ØSTERLED BRUSGAARD AND KESHIA A. N. AKKERMANS

association between lance use and riders in conflict is not unexpected. However, it is somewhat surprising that the lance is also the primary weapon of mounted hunters and that the bow is used almost exclusively by people on foot. The unparalleled technical advantages of the double-recurve composite bow, especially on horseback, were already well known in ancient times (Loades 2016: 10). The composite bow could be held drawn for a longer time with less energy input, allowed for a more precise aim and greater range, and it was highly portable due to its relatively small size (Bowden 2012; Zutterman 2003). The recurring depictions of riders using lances and archers on foot, whether fighting or hunting, does not necessarily contradict what is known from classical sources but instead might therefore point to specific cultural conventions among the desert nomads. The frequent depiction of the composite bow in the rock art is interesting because the carvers were according to their activities in their texts, nomadic or at least semi-nomadic (cf. Macdonald 1992; 1993). Yet the fabrication of this type of bow requires careful handling and drying over considerable amounts of time. If the scenes are accurate depictions of nomadic life in the desert (see further below), it is worth considering whether they were able to make these bows themselves or, conversely, where they obtained these bows from and whether their doing so then indicates further types of interaction between the desert and the sedentary areas, as evidenced by the Safaitic texts (cf. Macdonald 2014). This issue requires further exploration. The three themes recognizable in the scenic compositions—pastoralism, hunting, and conflict—largely match what we know about this region from other sources. From the Safaitic inscriptions we know that conflict was part of the authors’ world view. The texts mention raiding parties, (cavalry) troops, enemies, and desire for plundered goods (Al-Jallad 2015; OCIANA 2017). The extent to which these are portrayed in the rock art, including different types of weapons and weapon use, reflects the importance of conflict and raiding in the ideology of the desert nomads. Furthermore, the several scenes featuring dromedaries in the central role, representing the raiding of these animals, speak to the importance of this animal (Brusgaard 2020). Like conflict, pasturing or activities associated with it, such as migrating with herd animals and watering, form a common theme in the inscriptions. The domestication of the dromedary camel and its introduction into an already semi-pastoral subsistence in the Arabian Peninsula after 3000 BP (Magee 2014) also makes it plausible that the desert societies of this region had a pastoral or semi-pastoral mode of subsistence. The depiction of seemingly domestic, pastoral scenes such as infant domestic animals nursing and people leading domestic animals is thus not surprising. Lastly, the predominance of hunting and the interaction with and between animals stands out as a theme in the Jebel Qurma rock art. Other than the signing of images of wild animals (e.g., “By [name] is the lion”), the inscriptions do not allude to this theme at all in the Jebel Qurma corpus and do so only very rarely in other corpora. Therefore, based on the epigraphic evidence, scholars trying to reconstruct the societies behind the Safaitic texts have done so primarily in terms of pastoralism and nomadism (cf. Macdonald 1992; 1993). The scenic compositions in the Jebel Qurma rock art in no way contradict this and in fact support it in many ways, but they do offer an additional insight into the ideological significance of hunting and wildlife to these peoples. In this respect the Safaitic rock art from Jebel Qurma finds comparison with the (roughly) contemporary Hismaic rock art from the southern Jordanian desert, which depicts two major themes: dromedary camels and hunting (Corbett 2010; King 1990).

Safaitic Rock Art and Its Carvers How we are to subsequently interpret these findings in terms of what the themes reflect and what they can tell us about the desert societies is a matter that requires further investigation on several levels. On the one hand, it will involve further comparison with archaeological findings from Jebel Qurma and the Black Desert. On the other, more discussion is needed on the question of what the images signified for their makers and what they portray. Whether we are, for example, to interpret the dominance of hunting as a reflection of subsistence strategies in the desert or socially significant activities depends on the extent to which these images express a degree of reality or a representation of the day-to-day reality of these societies. Of course, “we cannot expect to read it [rock art] as a mirror of society” (Walderhaug 1998: 298). However, it can inform

HUNTING AND HAVOC

147

us about past world views and aspects that were important to these societies. Scenes in particular provide insights into social organization, practices, and activities, many of which are inaccessible through other parts of the archaeological record (May and Domingo Sanz 2010: 35). For this reason, an in-depth study of scenes and themes in the rock art opens up new possibilities for investigating the social practices, subsistence strategies, and regional connections of the societies that carved these pictorial and textual engravings. For example, the abundance of hunting scenes in the rock art, and wild animals in general (cf. Brusgaard, 2019), may help to challenge preconceived notions of what a pastoralist society entails and what would have been important to the supposed pastoralist worldview. While this is a matter for further research, this study on scenes and weapons has already revealed some new insights into the Safaitic rock art and its makers. Most notably, it is clear that there are specific, recurrent patterns in the rock art compositions. This indicates that there were set rules to follow in what to portray and how. These subjects were selective, as not all aspects of (daily) life are depicted; for example, domestic human activities are not portrayed in the Jebel Qurma rock art, with the exception of two possible but very unclear sexual scenes. The scenes are thus by no means random depictions of narratives of interest to the individual carver, but a product of cultural and social norms. Interestingly, this matches what has been proposed for the Safaitic inscriptions by Al-Jallad (2015: 3), who argues that the texts are not forms of “unstructured self-expression” but highly formulaic and uniform communication, and that the subject matter of the inscriptions is limited and selective. This study on the rock art reveals similar insights, reaffirming the complementary nature of the two types of engravings. On a final note, these forms of expression and themes may have varied within the Black Desert region, so the Jebel Qurma rock art cannot be assumed to be representative of the entire area. Indeed, there is no evidence to suggest that the people who carved Safaitic engravings were all part of one cultural or ethnic community, as has already been emphasized elsewhere (Al-Jallad 2015; Macdonald 2009). As mentioned earlier, other corpora of engravings show images that appear to depict music making, dancing, and ploughing. There is no indication of the portrayal of these themes in Jebel Qurma.

Conclusion This study is part of ongoing research, and therefore many angles of investigation still need to be examined. However, we hope that with the data presented in this chapter we have revealed new insights into the scenes and material culture depicted in Black Desert rock art, an understudied rock art corpus. In particular, this study has revealed the occurrence of specific, recognizable patterns within the scenic compositions, including figures, interactions, and activities. Additionally, the dominant themes in the scenes and the use of weapons fit well with what we know of the historical context, while providing new insights into this period and region. Finally, we have also endeavored hereby to contribute to opening up new questions for debate in rock art research, both in the analysis of the various components of rock art imagery, such as weaponry, and in the study of scenes as a whole. Nathalie Brusgaard is a postdoctoral researcher at the University of Groningen. Nathalie earned a PhD from Leiden University for her work on the rock art of nomadic herders in the Black Desert of Jordan. Her research focused on how rock art can reveal past human-animal relationships and human-landscape interactions. Nathalie specializes in social zooarchaeology and rock art studies and has worked on numerous fieldwork projects in the Netherlands, Germany, Jordan, and the United States. Keshia Akkermans is a graduate student at Leiden University, majoring in the archaeology of the Near East. In 2017 she received a bachelor’s degree based on her thesis on weapon depictions in the rock art of the Jebel Qurma area in Jordan. She is currently pursuing a research master’s degree, focusing on the Late Bronze Age burial assemblages at Tell Sabi Abyad in Syria.

148

NATHALIE ØSTERLED BRUSGAARD AND KESHIA A. N. AKKERMANS

Notes 1. Female authors are very rare. In the known Safaitic corpora and in other Ancient North Arabian corpora of inscriptions, there are only a few rare examples of texts signed by women (Norris 2017). 2. Due to the difficulty in distinguishing between different types of equids, this study will use the generic term “equid” unless a specific member of this family, such as a horse or mule, can be recognized (cf. Brusgaard 2019; Littauer and Crouwel 1979; Macdonald 2019). The term “horseback” will be used as a general term for riders on any type of equid. 3. For an explanation of the identification of the animal motifs in the rock art see Brusgaard (2019). 4. ASWS 303 (Banı¯ 1999). 5. C 908 (Ryckmans 1950–1951).

References Ababneh, M. I. 2005. Neue safaitische Inschriften und deren bildliche Darstellungen, vol. 6. Aachen: Shaker Verlag. Akkermans, P. M. M. G., and M. L. Brüning. 2017. Nothing but Cold Ashes? The Cairn Burials of Jebel Qurma, Northeastern Jordan. Near Eastern Archaeology 80(2), 132–39. Al-Jallad, A. 2015. An Outline of the Grammar of the Safaitic Inscriptions, vol. 80). Leiden: Brill. Al-Jallad, A. 2016. An Ancient Arabian Zodiac. The Constellations in the Safaitic Inscriptions, Part II. Arabian Archaeology and Epigraphy 27: 84–106. Al-Manaser, A. Y. K. 2008. Ein Korpus neuer safaitischer Inschriften aus Jordanien, vol. 10. Aachen: Shaker Verlag. Al-Manaser, A. Y. K., and M. C. A. Macdonald. 2017. The OCIANA Corpus of Safaitic Inscriptions, Preliminary Edition. Edited by A. Y. K. Al-Manaser. The Online Corpus of the Inscriptions of Ancient North Arabia (OCIANA). Oxford: The Khalili Research Centre. Almathen, F., P. Charruau, E. Mohandesan, J. M. Mwacharo, P. Orozco-terWengel, D. Pitt, . . . B. De Cupere. 2016. “Ancient and Modern DNA Reveal Dynamics of Domestication and Cross-continental Dispersal of the Dromedary.” Proceedings of the National Academy of Sciences 113(24): 6707–12. Banı¯, A. 1999. “Dira¯sat nuqu¯š s. afawiyyah gˇadı¯dah min gˇanu¯b wa¯dı¯ sa¯rah/ al-ba¯diyah al-ʾurdunniyyah aš-šama¯liyyah.” Unpublished MA thesis, Yarmouk University. Retrieved 13 August 2018 from http://krc.orient.ox.ac.uk/ociana/corpus/ pages/OCIANA_0031861.html. Betts, A. V. G. 1987. “The Hunter’s Perspective: 7th Millennium BC Rock Carvings from Eastern Jordan.” World Archaeology 19(2): 214–25. Bowden, B. L. J. 2012. “The Origin and the Role of the Composite Bow in the Ancient Near East.” Ancient Warfare 5: 42–54. Bradley, R. 1997. Signing the Land: Rock Art and the Prehistory of Atlantic Europe. London: Routledge. Bradley, R. 1998. “Daggers Drawn: Depictions of Bronze Age Weapons in Atlantic Europe, in C. Chippindale and P. S. C. Taçon, ed., The Archaeology of Rock-Art, 130–45. Cambridge: Cambridge University Press. Brusgaard, N. Ø. 2019. Carving interactions: Rock Art in the Nomadic Landscape of the Black Desert, North-Eastern Jordan. Oxford: Archaeopress. Brusgaard, N. Ø. 2020. “Depicting the Camel: Representations of the Dromedary Camel in the Black Desert Rock Art of Jordan.” In Landscapes of Survival: The Archaeology and Epigraphy of Jordan’s Northeastern Desert, ed. P. M. M. G. Akkermans and A. Al-Jallad, 285–302. Leiden: Sidestone Press. Corbett, G. J. 2010. “Mapping the Mute Immortals: A Location and Contextual Analysis of Thamudic E/Hismaic Inscriptions and Rock Drawings from the Wa¯dı¯ Hafı¯r of Southern Jordan.” PhD. Dissertation, University of Chicago. Della Puppa, C. Forthcoming. “The Safaitic Scripts: An Ethno-palaeographic Investigation.” PhD dissertation, Leiden University. Gordon, D. H. 1953. “Swords, Rapiers and Horse-Riders.” Antiquity 27(106): 67–78. King, G. M. H. 1990. “Early North Arabian Thamudic E: A Preliminary Description Based on a New Corpus of Inscriptions from the Hisma Desert of Southern Jordan and Published Material.” PhD dissertation, SOAS University of London. Ling, J., and P. Cornell. 2017. “Violence, Warriors, and Rock Art in Bronze Age Scandinavia.” In Feast, Famine or Fighting? Multiple Pathways to Social Complexity, ed. R. J. Chacon and R. G. Mendoza, 15–33. Cham: Springer. Littauer, M. A., and J. H. Crouwel. 1979. Wheeled Vehicles and Ridden Animals in the Ancient Near East, vol. 1. Leiden: Brill Academic Publisher. Loades, M. 2016. The Composite Bow. Oxford: Osprey Publishing. Macdonald, M. C. A. 1990. “Camel Hunting or Camel Raiding?” Arabian Archaeology and Epigraphy 1(1): 24–28. Macdonald, M. C. A. 1992. “The Seasons and Transhumance in the Safaitic Inscriptions.” Journal of the Royal Asiatic Society (Third Series) 2(01): 1–11. doi:doi:10.1017/S1356186300001760 Macdonald, M. C. A. 1993. “Nomads and the H. awra¯n in the Late Hellenistic and Roman Periods: A Reassessment of the Epigraphic Evidence.” Syria 70(3/4): 303–413.

HUNTING AND HAVOC

149

Macdonald, M. C. A. 2006. “Burial between the Desert and the Sown: Cave-Tombs and Inscriptions Near Dayr al-Kahf in Jordan.” Damaszener Mitteilungen 15: 273–301. Macdonald, M. C. A. 2009. Literacy and Identity in Pre-Islamic Arabia. Surrey: Ashgate. Macdonald, M. C. A. 2012. “Goddesses, Dancing Girls or Cheerleaders? Perceptions of the Divine and the Female Form in the Rock Art of Pre-Islamic North Arabia.” In Dieux et déesses d’Arabie. Images et représentations. Actes de la table ronde tenue au Collège de France (Paris) les 1er et 2 octobre 2007, ed. I. Sachet and C. J. Robin, vol. 7, 261–97. Paris: De Boccard. Macdonald, M. C. A. 2014. “Romans Go Home? Rome and Other ‘Outsiders’ as Viewed from the Syro-Arabian Desert.” In Inside and Out: Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity, ed. J. H. F. Dijkstra and G. Fisher, 145–64. Leuven: Peeters. Macdonald, M. C. A. 2019. “Horses, Asses, Hybrids, and Their Use as Revealed in the Ancient Rock Art of the Syro-Arabian Desert.” In Equids in the Ancient Near East, Egypt, and Arabia: Proceedings of a Conference in Memory of Mary Aitken Littauer, ed. P. Raulwing K. M. Linduff and J. H. Crouwel, 149–168. Oxford: BAR Publishing. Magee, P. 2014. The Archaeology of Prehistoric Arabia: Adaptation and Social Formation from the Neolithic to the Iron Age. Cambridge: Cambridge University Press. May, S. K., and I. Domingo Sanz. 2010. “Making Sense of Scenes.” Rock Art Research: The Journal of the Australian Rock Art Research Association (AURA) 27(1): 35–42. May, S. K., D. Wesley, J. Goldhahn, M. Litster, and B. Manera. 2017. “Symbols of Power: The Firearm Paintings of Madjedbebe (Malakunanja II).” International Journal of Historical Archaeology 21: 690–707. Miller, R., E. McEwen, and C. Bergman. 1986. “Experimental Approaches to Ancient Near Eastern Archery.” World Archaeology 18(2): 178–95. Norris, J. 2017. “A Woman’s Hismaic Inscription from the Wa¯dı¯ Ramm Desert: AMJ 2/J. 14202 (Amman Museum).” Arabian Archaeology and Epigraphy 28(1): 90–109. OCIANA. 2017. The Online Corpus of the Inscriptions of Ancient North Arabia. Retrieved 28 August 2018 from http://krc.orient .ox.ac.uk/ociana/index.php/database Potts, D. T. 1998. “Some Issues in the Study of the Pre-Islamic Weaponry of Southeastern Arabia.” Arabian Archaeology and Epigraphy 9(2): 182–208. Rosen, S. A., and B. A. Saidel. 2010. “The Camel and the Tent: An Exploration of Technological Change among Early Pastoralists.” Journal of Near Eastern Studies 69(1): 63–77. Ryckmans, G. 1950–1951. Corpus Inscriptionum Semiticarum: Pars Quinta, Inscriptiones Saracenicae Continens, Tomus I, Fasciculus I, Inscriptiones Safaiticae. Paris: E Reipublicae Typographeo. doi:http://krc.orient.ox.ac.uk/ociana/corpus/pages/ OCIANA_0004113.html. Schwarzlose, F. W. 1886. Die waffen der alten Araber aus ihren dichtern dargestellt: ein beitrag zur arabischen alterthumskunde, synonymik und lexicographie, nebst registern. Hildesheim: Georg Olms Publishers. Smith, G. R. 1990. “Hunting Poetry (T. ardiyya¯t).” In Abbasid belles-lettres, ed. J. Ashtiany, T. M. Johnstone, J. D. Latham, and R. B. Serjeant, 167–184. Cambridge: Cambridge University Press. Stetkevych, J. 1999. “The Hunt in Classical Arabic Poetry: From Mukhad. ram “Qas. ¯ıdah” to Umayyad “T. ardiyyah.” Journal of Arabic Literature 30(2): 107–27. Walderhaug, E. M. 1998. “Changing Art in a Changing Society: The Hunters’ Rock-Art of Western Norway.” In The Archaeology of Rock-Art, ed. C. Chippindale and P. S. C. Taçon, 285–301. Cambridge: Cambridge University Press. Zutterman, C. 2003. “The Bow in the Ancient Near East.” Iranica Antiqua 38: 119–65.

==== •

•

•

10. MAKING A SCENE An Analysis of Rock Art Panels from The Northwest Kimberley and Central Desert, Australia June Ross

Introduction Many rock art studies in Australia and indeed, worldwide, have focused on the content of art assemblages: the individual motifs that together make up an engraved or painted assemblage. While earlier research frequently focused on quantitative analysis, which placed significance on numerically dominant motifs (e.g., Edwards 1966; Franklin 1991; Vinnicombe 1976), research over the past four decades has expanded to incorporate contextual analyses (e.g., Bradley 2000; David and Chant 1995; Ross 1997). Stylistic changes across space and through time have been evaluated against the social and environmental contexts in which rock art was produced in order to provide explanations for the form and content of assemblages. Rather than considering art as an “object,” art is seen as a “practice” (Conkey 1990: 5–17) intentionally created by individuals as a visual expression of aspects of their society. Despite these developments, the content of the assemblage remains central. Identification and analysis of the relationship between motifs is likely to broaden our understandings of rock art assemblages and inform us about the ways in which past societies viewed their world. In Australia, and perhaps more widely, analyses of what could be called “scenes” have been extremely limited in rock art studies (see Davidson, this volume). A number of factors have contributed to this neglect. In the recent past, researchers worldwide have reacted against previous speculative interpretations concerning the meaning of motifs, a shift largely driven by the integration of rock art studies into the discipline of archaeology (see Bahn 2010: 12–14 for discussion). A concomitant rejection of the indiscriminate use of ethnographic analogy has also fueled a change in research directions (Davidson 1995; Hiscock 2008; Wylie 1985), resulting in a general acceptance that the original meaning of motifs cannot be securely recovered. With these considerations in mind, many researchers have turned their attention away from the numerical composition of assemblages toward questions related to art as practice, the physical fabric of the art, and its communication qualities, alongside attempts to date and conserve assemblages. Wariness in identifying scenes in Australian rock art studies has been amplified by the lack of readily apparent iconic qualities across many regional assemblages, especially those in desert areas (see Dobrez 2017: 157 for discussion). Earlier ethnographic studies of art across different media in both the Kimberley and desert regions (e.g., Love 1930; Munn 1973; Spencer and Gillen 1899) demonstrated that a single motif (which may or may not have a loose iconic relationship to the object or place it represents) can hold many meanings at the same time, with the relevant meaning dependent on the age, gender, or seniority of the viewer or the context in which it has been produced. In these cases, the relationship between individual motifs cannot be securely ascertained, and the identification of the intentionality of compositions becomes problematic. In addition, the limits of current rock art dating techniques make it difficult to establish contemporaneity between motifs on a given panel, thus restricting the researcher’s ability to identify scenes with certainty. This problem is exacerbated in many sites across Australia where art panels frequently comprise multiple layers of superimposed motifs that add to the challenges of identifying intentional relationships between individual images. A careful analysis by R. G. Gunn (David et al. 2017: 247) identified as many as forty-nine different

MAKING A SCENE

151

layers of art on a single panel at the Nawarla Gabarnmang site in Arnhem Land, and three or more layers are common throughout the Kimberley and larger desert painting sites (Ross 2017; Travers 2015). Moreover, the physical boundaries across the rock substrate on which many assemblages have been produced cannot always be clearly defined, so the relationship between motifs becomes difficult to establish (Rosenfeld 1992). Pressure to document large bodies of rock art efficiently could contribute to the limited interest in scenes in Australian rock art studies and may be linked to the introduction, over the last decades, of handheld devices (e.g., Palm Pilots) programmed with drop-down menus that identify single/individual motifs. These devices are now commonly used by consultants and researchers to speed up the documentation of large bodies of rock art and enable instant downloads into databases that can be then be analyzed quickly and efficiently. Scenes can thus be missed. Finally, it may be that few scenes exist in the art assemblages, or at least that few are identified. As recently as 1985, Bob Layton (1985: 446), who documented rock art across the west Kimberley concluded that there are “no narrative scenes in Kimberley rock art.” In central Australia a century before, after undertaking extensive field recordings, the zoologist and anthropologist Baldwin Spencer and his co-author Frank Gillen reported that: . . . nowhere amongst these tribes, so far as we know, do we meet with any of the more complicated drawings depicting scenes such as a kangaroo chase or men spearing emus, or a corrobberee (sic) dance, such as are found amongst other tribes in the south and east parts of the continent (Spencer and Gillen 1899: 617).

In contrast, Grahame Walsh’s comprehensive study of the rock art in the Kimberley, undertaken over three decades, identified multiple scenes across the assemblage (Walsh 2000: 338–76). He flagged the potential of further studies of scenes, but cited both the exfoliation of some art panels and the time-consuming nature of analyses as contributing to the current perception of paucity (Walsh 2000: 370). So, what can an analysis of the relationship between motifs tell us about the way past societies viewed their world? A single animal motif, read literally, can provide information about animal species present, related climatic conditions, or the relative importance of species for diet or ritual. A single anthropomorphic figure can reveal information about items of clothing, hairstyles, or weapons not otherwise preserved in the archaeological record. Without associated motifs, single figures cannot tell us about the relationship between people, between people and animals, or between people and their environment. Three simple definitions of the term scene are relevant to the following analysis of the two regional rock art assemblages. The New Oxford Dictionary of English (1998: 1660) defines a scene as “the place where an incident in real life or fiction occurs or occurred; any view or picture, an incident or situation in real life.” Taken more generally, a scene then can include a location, an event, and/or an activity. Here we are concerned with the representation of scenes defined in this way, but does a scene have to represent one place, seen from a particular point, at a precise moment in time? If the spatial relationship between elements of the scene is not presented in a form that resembles the concept of reality made standard during the Renaissance, when the rules of perspective were adhered to, should the composition be excluded? Adoption of such a stand might discount modernist artworks generally accepted as scenes painted by internationally acclaimed artists such as Picasso or Miro, as well as the ancient scrolls produced in Asia with their multiple vanishing points (de la Croix and Tansey 1976) or the contemporary aerial map–like artworks painted by Australia’s Western Desert artists (Bardon 1991). Such paintings would be rejected if Renaissance-era relationships among time, space, and place relationship were prerequisite constituents needed to ensure their classification as scenes. As we cannot know the full range of artistic devices adopted by past societies to indicate an association between motifs, it might then be more useful to look at the relational composition of art panels without the limitations and semantic arguments that a tight definition of “scene” imposes. Instead, my focus will be on identifying the compositional principles that link motifs to create a scene alongside the physical limitations of the rock art panel itself, the relationship or association between motifs, and the relative style, pigment color, and chemical composition of associated motifs. In this chapter, I have selected examples from two very different rock art assemblages; the first from the Kimberley in the tropical northernmost reaches of Western Australia, and the second in the arid heart of Australia in the Central Desert region. Hundreds of painted art sites (Donaldson 2012a, 2012b, 2013) are

152

JUNE ROSS

located in rock shelters amongst the rugged and remote sandstone gorges, riverine corridors, and woodland plateaus that comprise the Kimberley region, a relatively well-watered area about the size of Spain. The long stylistic sequence (e.g., Travers and Ross 2016; Walsh 1994; 2000; Welch 1993) dominated by figurative motifs is known to have its origins in the late Pleistocene (Roberts et al. 1997; Ross et al. 2016), and unlike Upper Paleolithic art sites in Europe, motifs continued to be added to the assemblage as styles changed through time. Meanwhile the practice of repainting or “freshening” particular motifs continued into the very recent past (Love 1930: 1936; Woolagoodja 2007: 29–30). In the vast sand ridge deserts and inland ranges of the Central Desert, painted and engraved rock art sites are found adjacent to limited water sources, often ephemeral rock holes or subterranean soaks. The climate is characterized by the unpredictability of the rainfall; long periods of drought are commonplace. Numerically extensive engraved assemblages located in open gorges dominate, while smaller assemblages of painted art are found in shallow rock shelters and overhangs (Gunn 1995; 2000; Rosenfeld and Smith 2002; Ross 2005). Though dating of the arid zone assemblage has proved problematic (but see Smith, Watchman and Ross 2009), it too is likely to have its origins in the late Pleistocene. Rather than the distinct chronological stylistic changes seen in the Kimberley rock art assemblage, it is the enduring production of a limited core range of geometric and track motifs that characterizes the Central Desert assemblage (Ross 2005). Despite the dramatic differences in climatic conditions, the past inhabitants of these two areas share certain underlying features. Both populations were hunter-gatherers foraging widely in flexible descent groups known as clans. Each possessed a distinct body of spiritual beliefs with an associated set of sacred sites, water holes, rock shelters, and other geographic features imbued with the spiritual power of Ancestral Beings who crossed the landscape in the creation period (sometimes morphed under the term “the Dreaming” [Keen 2003]), leaving traces of their activities in topographic features, songs, and rock art (Layton 1985: 435). Despite these broad similarities in economic, social, and religious lifeways, the regions have different creation stories, different mythical ancestors, and different ceremonial activities. Furthermore, their rock art assemblages are strikingly different. In both regions, understanding of the meaning of motifs (at any time) in both regions is revealed to appropriate individuals at various stages of initiation as they gain in age and stature, with deeper meanings divulged according to seniority.

Kimberley Amongst the oldest rock art stylistic phases in the Kimberley are groups of finely executed, monochrome anthropomorphic figures known as Gwion figures (previously Bradshaw figures). While there are some clear differences between figures in this art phase, most share a number of characteristics, like elongated bodies depicted with jointed limbs, distinct waists, and well-developed musculature. Although fingers and feet are often evident, all figures lack facial features. Gwion figures are frequently depicted wearing ornate headdresses and body decorations on arms, ears, waists, and legs, with details of elaborate hair styles carefully represented. The figures can be seen carrying a range of accoutrements, such as paired boomerangs, dilly bags, spears, and spear-throwers. Analysis of more than two hundred rock art sites in the northwest Kimberley containing more than 10,000 motifs, undertaken as part of a broader archaeological project (Holt and Ross 2016; Ross et al. 2016; Travers 2015; Travers and Ross 2016), led to identification of a number of compositional principles commonly used to relate one motif to another, or one motif to a group of other motifs in the Kimberley assemblage. Principles used to create a scene include repetition, intersecting motifs, creation of a linear frieze, joining motifs to a line, surrounding motifs with a line, linking motifs with a line, or mirror imaging. Repetition and intersecting motifs are the compositional principles that have been used to create a scene that depicts fifteen Gwion figures (Figure 10.1) painted on the entire face of a smooth rock panel 4650 mm long and 1510 mm high. The hanging rock face was exposed when a large block of sandstone at the base of a cliff line subsided, creating a shallow shelter filled with the fallen roof slabs, thus making it unsuitable for habitation. The site is one of six that form a semicircular complex of art sites (known as Malauwarra) along the low cliff line, fronted by a sandy flat that runs out to the Mitchell River, a major freshwater source in the region. Optically Stimulated Luminescence of soil samples from an excavation undertaken in this complex

MAKING A SCENE

153

FIGURE 10.1. Scene of Gwion figures, Mitchell River, Kimberley. Copyright June Ross.

(LMR02d) returned a preliminary age estimate of 28,000 +/- 6000 BP from the strata in which the lowest stone artifacts were recovered, providing a minimum time depth for occupation in the Malauwarra basin. Analyses of the style, pigment, placement and content of the panel (Travers 2015) showed that all fifteen figures were painted using the same purple/brown (mulberry) pigment, indicating that all were painted in a single painting episode. Recent research on ochre sources and hues in the northwest Kimberley art assemblage (Huntley et al. 2015; Keats 2017) has shown that pigment color within panels, across sites, and through time vary markedly, making identification of specific pigment colors within a panel an ideal diagnostic tool for identifying individual painting episodes. Huntley and Keats have attributed pigment variation in the assemblage to the opportunistic collection of readily available surface and riverine lateritic gravels used to supplement quarried ochre sources. Accepting that the figures were painted in a single episode, then the positioning of each motif, including the intersections, is likely to be intentional. All fifteen figures are stylistically alike and painted using a similar technique, creating a two-tone effect by using some areas of thick pigment and infilling with areas of watery paint. The figures are all of different sizes, but they appear to have been placed in positions that complement the others, suggesting that the figures form a group engaged in a related activity. Seven figures face left, six figures face right, and two face the front. Figures are depicted with bent knees, with raised, bent or extended arms, and arched backs—all postures associated with energetic movement. If the feet are depicted at all, the toes are pointed downwards, suggesting that the figures may be dancing or jumping (Welch 2007). Figures are adorned with a range of body decorations including waist sashes, armlets, and large elaborate headdresses with feathers, pompoms, and tassels attached. Details of hairstyles include depictions of intricate knotting of lengths of hair at the back of the head (Figure 10.1, Gwion on the extreme right). In addition, ten of the figures are depicted holding paired boomerangs. In the ethnographic literature, paired boomerangs were used during ceremonies to beat out the rhythm of accompanying songs (Love 1936). Several of the fig-

154

JUNE ROSS

ures carry grass whisks that dancers today commonly replace with small bunches of leaves that are shaken to add to the drama of the performance. Analyses of northwest Kimberley rock art aimed at identifying features used by artists to flag the sex and/or gender roles of Gwion figures determined that figures depicted wearing elaborate headdresses, and armlets and holding boomerangs or spears were likely to be male (Holt 2014; Holt and Ross 2016). The figures on the panel at LMR02b are depicted with these features, so are therefore likely to represent male participants, perhaps performing a ceremonial dance or ritual. The ruggedness of the surrounding terrain would argue against the wearing of elaborate accoutrements for everyday activities, supporting the assertion that the figures are participating in some form of ritual performance. A second panel of Dynamic Figures (a style that appears in Kimberley rock art toward the end of the Gwion phase) depicts a secular activity rather than ceremonial one. Animated hunting and camp scenes are recurring subjects amongst the finely painted figures in this style (Holt 2014). Painted on the sloping ceiling of a low shelter, the isolated panel measures 305 mm long and 220 mm high and depicts two human figures with the spoils of a successful hunt (Figure 10.2) painted using the same dark red pigment. The inclusion of breasts on one figure indicates that the image represents a woman. The other is depicted carrying spears and boomerangs, implying that it is an image of a man (see above for diagnostic sexing features). The woman is portrayed carrying a kangaroo on her head, whereas the man walks ahead holding another kangaroo. Ethnographic records indicate that in the recent past, hunting was “the prerogative of the mature men” (Petri 1954: 16–17), but in this scene it is evident that the woman is playing an integral role in the hunting party. Analysis of the scene provides an opportunity to review the past status and gender roles of men and women. It may well be that gender roles have changed through time, with women and men both sharing in hunting activities in the past. Conversely, it may be that Kimberley ethnographic records, usually documented by men FIGURE 10.2. A successful hunting scene, Dynamic Figures, interviewing men (Lommel (1952, 1997; Love Lawley River, Kimberley. Copyright June Ross. 1936), downplay women’s role in the highly prestigious activity of providing large game. A review of additional panels of Dynamic Figures engaged in secular scenes have provided further insights into the economic and social practices of their society (Holt 2014). One documented Kimberley art panel appeared to depict a scene demonstrating the ability of the people represented by Dynamic Figures to hunt and spear the now extinct thylacine, the only member of the Thylacinidea family and the world’s largest marsupial (pouched) carnivore.1

Central Australia The painted, engraved, and drawn rock art assemblages of Central Australia contain few motifs other than tracks that appear to hold a strong iconic relationship to the object they represent. The dominant motifs in all techniques are circles and circle variants, arcs, sinuous lines, linear motifs, and tracks. Without a specific iconic form that suggests an immediate meaning, each motif can hold multiple meanings at the same time. (This, of course, is also true of motifs in other assemblages that do hold a close resemblance to the subjects they represent). The apparent lack of iconic form in Central Australian rock art (but see Munn [1973: 4] for an explanation of her use of the term “representational” to describe Walbiri symbols in the north of the region) provides an ideal means to create multivalent motifs, or what Munn (1973: 4) described as motifs

MAKING A SCENE

155

with “discontinuous meaning range,” acknowledging that meaning is not inherent in the motif but agreed to by individuals or societies. For example, a circle might represent a hill, a person, a water hole, or a breast, or several of these at the same time. The minimal nature of the conventional forms can be combined in unlimited variation to create multiple meanings (Munn 1973). The viewer interprets motifs according to the explanation given by the mark maker or by later instructors, dependent on the context in which the art occurs and the viewer’s gender, age, level of knowledge, and/or position in the society. Understanding this, how then can we identify what might be considered a scene in Central Australian rock art? Three compositional principles used to create scenes have been identified in the assemblage, including the creation of a series of track-lines, repetition of motifs, and linking of motifs by means of sets of parallel straight, dotted, or sinuous lines. Without the addition of ethnographic documentation, however, scenes would remain unrecognized. Ethnographic recordings undertaken amongst Central Australian Aborigines in the late nineteenth and early twentieth centuries (e.g., Basedow 1904; Mountford 1948, 1968, 1976; Spencer and Gillen 1899, 1927; Strehlow 1947, 1970) confirm that every prominent topographic feature and many smaller natural features hold some history relating to the mythical past, the Alcheringa (Dreaming), and the Ancestral Beings who traveled across the country creating the landscape, people, and resources. The meanings of the motif forms used in all visual and graphic mediums are believed to be “fixed in ancestral origins” and “do not represent events in the ordinary world” (Morphy 1989: 158). For Aboriginal Central Australians, the totemic landscape forms a firm basis for religion, for social order, and for established authority itself. Rock art is seen to provide tangible evidence of the Alcheringa. Spencer and Gillen (1899: 632, Figure 132) documented the contemporary meanings of a series of painted panels at Emily Gap (Nthwerrke) where the red sandstone cliffs form a steep-sided passage through the ranges. The gap is believed to be the place where, in the mythical past, the Caterpillar Ancestor (Udnirringita) sprang into being, the caterpillar being an important food resource for Aboriginal people in desert regions. The paintings have been published many times (e.g., Ross 2005: 220, Fig. 16.3 [available online on Academia.com]; Spencer and Gillen: 1899: 632, Fig. 132) and are believed to have sprung up spontaneously on the smooth vertical rock faces to mark the spot where the Ancestral Women painted themselves and stood peering up at a small cave high up on the opposite side of the gap, watching as the leader and his men performed a ceremony to ensure the ongoing supply of caterpillars. The paintings depict elements of this scene (Spencer and Gillen 1899: 632). A red motif above the main panel represents the forearm of a woman as she leans against the rocks looking upwards to the cave while a trail of three human tracks below represents the tracks of the women. It is the series of large red and white striped panels that represent the major aspects of the scene. The slanting line at the top of the panel also represents a woman shielding her eyes while looking upwards toward the cave, and the three dots represent the eggs of the caterpillar. On an adjacent striped panel, the central slanting lines represent the shoulders of the Alcheringa Ancestors. The striped design is also painted on the chests of men when they perform the sacred ceremonies relating to the Caterpillar Ancestor. The paintings at Emily Gap and many other major art sites throughout the arid zone include stylized motifs of a scene without incorporating all elements, participants, or objects in a form or a time/space relationship that could be deciphered without ethnographic documentation. Rather, parts of a scene can be spread over a number of discrete panels as they are at Emily Gap, and a scene may lack any of the usual diagnostic compositional principles that tie a scene together. Meanwhile, when considered from a Eurocentric perspective, the selection of the elements included in the painting is neither diagnostic nor comprehensive. At Emily Gap, motifs were limited to images representing the presence of women, their arms and actions, the Ancestors’ shoulders, the eggs of the caterpillar, and the ceremonial chest designs painted on the men. While the Central Australian graphic system provides tangible evidence of the Alcheringa, a verification of the actions of the creative ancestors, it remains opaque to those without the requisite knowledge to interpret the scenes.

Tracks The significance of tracks in the life of the Arrernte of Central Australia is underscored by the use of a special word, mbatja, for the making of footprints or tracks. Creating tracks is the common visual device used to signify the directional movement of a human, animal, or bird in both rock art and sand drawing. The shape

156

JUNE ROSS

of a footprint or track resembles that of the agent that made it and thus has iconic properties, but at the same time, it is also “an index of a presence in a particular place” (Green 2014: 17). Representing an entity in this way is not merely a matter of “economy of labor”; rather, it emphasizes the importance of the track in Aboriginal life (Elkin 1949–50: 134). Mothers teach their small children to draw the tracks of animals in the sand, and people learn to identify other individuals by their footprints alone. As in all hunter-gatherer societies, knowledge of tracks is a matter of everyday necessity. Linguist and anthropologist T. G. H. Strehlow (1951: 3) stated that “a Central Australian native used to read the ground like a book,” a surface perceived as rich with graphic meaning. Painted or engraved trails of tracks are unlikely to have a mundane meaning. Rather, their presence is perceived as a means of confirming the travels and activities of Ancestral Beings as they crossed the landscape in the Alcheringa. The trails may take the form of animal tracks or human footprints, as Ancestral Beings are considered to be transformative, imbued with the qualities of both, and able to transmute from one form to another. In this way, a series of tracks can be viewed as a scene. One such scene was documented at Atalpi (Ross 2003), a thirteen-meter–long sandstone shelter formed under a cliff face behind an ephemeral waterfall at the head of a narrow gorge on the western side of the George Gill Range in the south of the Central Desert. Although the rock hole formed below the site is large, it holds water for only a short time after rain. The site contains more than forty red hand stencils, engraved circles, and pecked cupules, but an isolated panel (measuring 1620 mm long and 480 mm high) painted on a hanging ledge on the rear wall differs from the rest of the art (Figure 10.3). It depicts two parallel gray, sinuous lines, possibly representing a snake (a significant Ancestral Being in this region) flanked by gray and yellow diagonal lines or offsets made using a finger dipped in paint. These offsets are interpreted by Traditional Owners as characters accompanying or possibly following the “snake” (Leo Abbott, pers comm. 1999). The direction of the offsets, identified by analyzing the finger marks, indicates that the figures are moving along the same route toward the left of the panel. While the exact meaning of the scene cannot be interpreted, it is likely to represent an activity undertaken by Ancestral Beings at Atalpi as they passed through the area. Clusters or vertical trails of bird tracks and trails of other species such as native cats, goannas and other lizards, possums, and snakes are also common amongst the painted and engraved assemblages throughout the region.

FIGURE 10.3. Track scene, Atalpi, George Gill Range, Central Australia. Copyright June Ross.

MAKING A SCENE

157

A different type of track scene was documented at a nearby site at Kathleen Domes, only a few kilometers north of Atalpi. On top of the range, the sandstone forms a series of dramatic rounded domes undercut at the base, forming shallow overhangs ideal for sheltering during extreme weather. Many contain evidence of recent habitation as well as small assemblages of hand stencils, drawn motifs, and painted rock art. Amongst the most commonly painted figures are roughly executed trails of tracks, mostly painted in a distinctive dark red pigment (Figure 10.4) and believed to depict the travels of the Ancestral Kangaroo (note that several species of macropod play a role in the mythology of the region—here the term Kangaroo is used in its generic form). Kangaroo tracks are depicted by sequences of paired parallel lines painted using fingers rather than a brush. Offsets, said to represent the distinctive way a kangaroo, always alert, stops to “look about” to monitor its surroundings (Leo Abbott, pers. comm. 1999) have been added to the trail. Single straight lines represent the tail when it is rested on the ground to balance the kangaroo when it pauses. The splatter of paint portrays the place where the Ancestral Kangaroo “finished up”—the place where it is believed to go into the ground to enable it to move on to another location where further mythological activities could take place. It is the meandering and spontaneous nature of these painted trails of tracks that differentiates them from other, more formal track lines like those at Atalpi. The trails at Kathleen Domes follow protruding ledges for meters (the longest extends for nearly ten meters) and frequently terminate in or circle niches in the rock face. The gestural and spontaneous nature of the painting technique suggests that the track lines are likely to be “co-speech graphics” in which storytelling and/or singing and painting are used in unison (Green 2014: 1), providing a mnemonic prompt that reminds participants of the mythological activities of Ancestral Beings relevant to the area, the artist, or both. In this way, they perform a role similar to that of sand drawing (Munn 1973: 121–34). Similar spontaneous or gestural trails have been documented in both the painted and the engraved assemblages across Central Australia, the engraved track lines sometimes comprising a series of just a few rough peck marks (Ross 2002). Based on superimposition analysis, erosional evidence, and state of patination of the roughly engraved trails, all are attributed to the most recent phase of rock art production (Ross 2002, 2003). Practices such as touching, rubbing, abrading, battering, pounding, or pecking significant rock surfaces have been documented at rock art sites across central Australia (Ross 2003). Interaction with particular rock surfaces is seen as a means to release the power held within the substrate (Mountford 1968, 1976). Munn (1973: 243) has argued that the Walbiri perceive the ground on which they produce their sand drawings as “a conduit to the metaphysical world of spirits.” It may well be that the rock substrate at art sites is also viewed in this way. The inclusion of a number of handprints (Figure 10.4) added as part of the Kathleen

FIGURE 10.4. Gestural kangaroo track lines, Kathleen Domes, George Gill Range, Central Australia. Copyright June Ross.

158

JUNE ROSS

Domes trail and other trails lends weight to the contention that the actual production of the spontaneous and gestural tracks may be another form of interaction with rock surfaces.

Conclusion The identification and analysis of a small sample of scenes from both the northwest Kimberley and the Central Desert illustrates both the potential and the limitations of such studies, and accentuates the differences between the assemblages. A literal interpretation of a scene depicting Gwion figures in the northwest Kimberley provides a depiction of the participants, their actions, attire, and demeanor during ceremonial life, all otherwise lost to archaeologists. Analysis of scenes depicting Dynamic Figures reveals the potential of these scenes to encode information on past gender roles and relative status, as well as social and economic activities. Animals targeted for hunting and the technology deployed can be identified. Significantly, the breadth of information obtained from analyses of Kimberley rock art scenes could not be gleaned from studying individual motifs alone: it is the association of motifs to each other that provides information on the relationships between them. Scenes in Central Australian rock art assemblages are more difficult to identify. Without specific ethnographic explanations, scenes and meaningful associations between motifs are impossible to recognize, as they are not necessarily limited to a single panel and only selected elements of the scene are depicted. Those that are depicted are highly stylized and may not be placed in a manner that reflects reality or relative size. The lack of recognizable compositional principles is exacerbated by the multivalent nature of the core vocabulary. Therefore, identification and interpretation of scenes remains problematic. Trails of tracks, so common in both the engraved and painted assemblages, can be analyzed as scenes. Ethnographic accounts document the activities of Ancestral Beings as they traveled across the country creating topographic features and establishing kinship and belief systems. Ancestral Beings are represented in graphic form and many transmute between human and animal form, so many trails of tracks can be reasonably interpreted as tangible evidence of the activities of specific mythical beings. “Orientation in space is a prime concern” (Myers 1986: 54) for desert Aborigines, so locations where mythological events occurred are highly significant, as are the directions from whence they came and went. The form of the trails of tracks provide some guidance to the actors in the traveling myths, and the direction of the fingermarks indicates the direction in which they were traveling. The simplified form of many trails of tracks obscures the finer details of the species depicted, so background knowledge of the art site and the traveling myth are needed in order to interpret details of the scene. Thus, any comprehensive understanding of Central Australian scenes depends upon ethnographic accounts or knowledgeable informants. Still, this is not to say that a broad understanding of the art system could not be enhanced by a study of the compositional principles that link motifs and panels, and panels to places, in order to further our understanding of these painted relationships.

Acknowledgments My thanks to Iain Davidson and April Nowell for their invitation. Field research was undertaken with the collaboration and support of the Native Title holders from the Wunambal Gaambera Aboriginal Corporation in the northwest Kimberley with special thanks to the Karadada and Peurmora families from the Kandiwal and Brremangurey graas. Research was funded by the Australian Research Council: Linkage Project (LP0991845) Change and Continuity: Archaeology, Chronology and Art in the Northwest Kimberley, Australia. I appreciate the generous support of the Linkage Partners: Kandiwal Aboriginal Corporation, Kimberley Foundation Australia, Western Australian Department of Environment and Conservation, Slingair, and Heliwork Pty. Ltd. In Central Australia, I would like to thank the Traditional Owners from the many sites I have recorded over two decades, especially Peter Bullah, Max and Clive Inkamala, Mavis and Herman Malbunka, Doug, Peter, Barry, and Leo Abbott, Geoffrey Oliver, Conrad Ratara, Bruce Breaden, Edward Ronji, Max Stuart, Sid Coulthardt, Jack, Syd, and Peter Kenny, Douglas Multa, and Dennis Ebartarintja. I appreciate

MAKING A SCENE

159

the assistance and financial support provided by Parks and Wildlife Service, Northern Territory and their staff, the Department of Environment and Heritage, and the Australian Institute of Aboriginal and Torres Strait Islander Studies. Finally, I would like to thank the many enthusiastic field assistants who have contributed to the research. June Ross Originally trained as a designer and art educator, June Ross taught at secondary school level before joining the museum sector. Over the past two decades she has lectured in the archaeology program at the University of New England, now as Adjunct Professor. Her enthusiasm for the challenges and romance of remote regions has seen her focus her research on the rock art of arid Australia, the subject of her PhD thesis, and more recently in the rugged Kimberley region, where she leads an ongoing broad-scale archaeological project. She has undertaken numerous rock art recording, dating, and conservation projects across Australia for both government agencies and Aboriginal Land Councils. Her publications reflect her interest in rock art, ritual, chronology, and context.

Note 1. For more information on the Thylacinidea family and the world’s largest marsupial (pouched) carnivore visit: https:// www.livescience.com/58753-tasmanian-tiger-facts.html.

References Bahn, P. G. 2010. Prehistoric Rock Art: Polemics and Progress. Cambridge: Cambridge University Press. Bardon, G. 1991. Papunya Tula: Art of the Western Desert. Ringwood: McPhee Gribble. Basedow, H. 1904. “Anthropological Notes Made on the South Australian Government North-West Prospecting Expedition, 1903.” Transactions and Proceedings of the Royal Society of South Australia 12–51. Bradley, R. 2000. An Archaeology of Natural Places. London: Routledge. Conkey, M. W. 1990. “Experimenting with Style in Archaeology: Some Historical and Theoretical Issues.” In The Uses of Style in Archaeology, ed. M. W. Conkey and C. Hastorf, 5–17. Cambridge: Cambridge University Press. David, B., and D. Chant. 1995. “Rock Art and Regionalism in North Queensland Prehistory.” Memoirs of the Queensland Museum 37(2): 358–528. David, B., J.-J. Delannoy, R. Gunn, E. Chalmin, G. Castets, F. Petchey, K. Aplin, M. O’Farrell, I. Moffat, J. Mialanes, J. M. Geneste, B. Barker, B. Sadier, M. Katherine, M. Manataki, and U. Pietrzak. 2017. “Dating Painted Panel E1 at Nawarla Gabarnmang, Central-Western Arnhem Land Plateau.” In The Archaeology of Rock Art in Western Arnhem Land, Australia, ed. B. David, P. Taçon, J. J. Delannoy, and J. M. Geneste, 245–301. Canberra: ANU Press. Davidson, I. 1995. “Paintings, Power and the Past: Can There Ever Be an Ethnoarchaeology of Art?” Current Anthropology 16(5): 889–92. De la Croix, H. and R. G. Tansey. 1976. Gardener’s Art Through the Ages. 8th ed. New York: Harcourt Brace Jovanovich. Dobrez, P. 2017. “From Tracks to Gesture-Derived Inscriptions: An Australian Genealogy for ‘Tracks And Lines’ Petroglyphs.” Rock Art Research 34(2): 149–68. Donaldson, M. 2012a. Kimberley Rock Art. Volume One, Mitchell Plateau Area. Wildrocks Press. ———. 2012b. Kimberley Rock Art. Volume Two, North Kimberley. Wildrocks Press. ———. 2013. Kimberley Rock Art. Volume Three, Rivers and Ranges. Wildrocks Press. Edwards, R. 1966. “Comparative Study of Rock Engravings in South and Central Australia.” Proceedings of the Royal Society of South Australia 90: 33–40. Elkin, A. P. 1949–50. “The Origin and Interpretation of Petroglyphs in Southeast Australia.” Oceania 20: 119–57. Franklin, N. 1991. “Explorations of the Panaramitee Style.” In Rock Art and Prehistory: Papers Presented at AURA Congress, Darwin 1988, ed. P. Bahn and A. Rosenfeld, 39–44. Oxford: Oxbow Books. Green, J. 2014. Drawn from the Ground: Sound, Sign and Inscription in Central Australian Sand Stories. Cambridge: Cambridge University Press. Gunn, R. G. 1995. “Regional Patterning in the Aboriginal Rock Art in Central Australia: A Preliminary Report.” Rock Art Research 12(2): 117–28. ———. 2000. “Central Australian Rock Art: A Second Report.” Rock Art Research 17(2): 111–26. Hiscock, P. 2008. Archaeology of Ancient Australia. Oxford: Routledge. Holt, D. A. 2014. “Sex and Gender: An Archaeological Analysis of Rock Art of the Northwest Kimberley.” MA thesis. Armidale: University of New England.

160

JUNE ROSS

Holt, D. A., and J. Ross. 2016. “Sex and Gender in Wanjina Rock Art, Kimberley, Australia.” Expression 11: 39–47. Huntley, J., M. Aubert, J. Ross, A. Brand, and M. J. Morwood. 2015. “One Colour, Two Minerals: A Study of a Mulberry Rock Art Pigment Quarry, Kimberley, Northern Australia.” Archaeometry 57(1): 77–99. Keats, S. A. 2017. “The Kimberley palette: the archaeology of Colour in Northwest Kimberley Ochre.” BA (Honours) thesis. Armidale: University of New England. Keen, I. 2003. “Aboriginal Economy and Society at the Threshold of Colonisation: A Comparative Study.” Before Farming 3(2): 1–24. Layton, R. 1985. “The Cultural Context of Hunter-Gatherer Rock Art.” Man 20(3): 434–53. Lommel, A. 1952. The Unambal: A Tribe in Northwest Australia, trans I. Campbell. Carnarvon: Takarakka Nowan Kas Publications. Love, J. R. B. 1930. “Rock Paintings of the Worora and Their Mythological Interpretation.” Journal of the Royal Society of Western Australia, 16: 1–24. ———. 1936. Stone Age Bushmen of Today: Life and Adventure among a Tribe of Savages in North-Western Australia. London: Blackie & Son Limited. Morphy, H. 1989. Aboriginal Art. London: Phaidon Press. Mountford, C. P. 1948. Brown Men and Red Sand. Melbourne: Robertson and Mullens Limited. ———. 1968. Winbaraku and the Myth of Jarapiri. Adelaide: Rigby Limited. ———. 1976. Nomads of the Australian Desert. Adelaide: Rigby Limited. Munn, N. 1973. Walbiri Iconography. Ithaca, NY: Cornell University Press. Myers, F. 1986. Pintupi Country, Pintupi Self: Sentiment, Place and Politics among Western Desert Aborigines. Washington DC: Smithsonian Institution Press. Pearsall, Judy (ed) 1998. The New Oxford Dictionary of English. Oxford, Clarendon Press. Petri, H. 1954. The Dying World in Northwest Australia, trans I. Campbell. Perth: Hesperian Press. Roberts, R. G., G. L. Walsh, A. Murray, J. Olley, R. Jones, M. J. Morwood, C. Tuniz, E. Lawson, M. Macphail, D. Bowdey, and I. Naumann. 1997. “Luminescence Dating of Rock Art and Past Environments Using Mud-Wasp Nest in Northern Australia.” Nature 387: 696–99. Rosenfeld, A., 1992. “Recent Developments in Australian Rock Art Studies.” Paper presented to the Fourth World Congress of Aegean Archaeologists. Hobart, Tasmania. Rosenfeld, A., and M. A. Smith. 2002. “Rock Art and the History of Puritjarra Rock Shelter, Cleland Hills, Central Australia.” Proceedings of the Prehistoric Society 68: 103–24. Ross, J. 1997. “Painted Relationships: An Archaeological Analysis of a Distinctive Anthropological Rock Art Motif in Northwest Queensland.” BA (Honours) thesis. Armidale: University of New England. ———. 2002. “Rocking the Boundaries, Scratching the Surface: An Analysis of the Relationship between Paintings and Engravings in the Central Australian Arid Zone.” In Barriers, Borders and Boundaries: Proceedings of the 2001 Australian Archaeological Association Annual Conference, Tempus 7, ed. S. Ulm, C. Westcott, J. Reid, A. Ross, I. Lilley, J. Prangnell, and L. Kirkwood, 83–91. Brisbane: University of Queensland. ———. 2003. “Rock Art, Ritual and Relationships: An Archaeological Analysis of Rock Art from the Central Australian Arid Zone.” PhD thesis. Armidale: University of New England. ———. 2005. “Rock Art of the Red Centre.” In 23 South: Archaeology and Environmental History of the Southern Deserts, ed. M. Smith and P. Hesse, 217–30. Canberra: National Museum of Australia Press. ———. 2017. “Kungkarangkalpa and the Art of Walinynga.” In Songlines: Tracking the Seven Sisters, ed. Margo Neale, 82–86. National Museum of Australia Press. Ross, J., K. Westaway, M. Travers, M. J. Morwood, and J. Hayward. 2016. “Into the Past: A Step toward a Robust Kimberley Rock Art Chronology.” PLOS ONE 11(8): e0161726. Smith, M., A. Watchman, and J. Ross. 2009. “Direct Dating Indicates Mid-Holocene Age for Archaic Rock Engravings in Arid Central Australia.” Geoarchaeology 24(20): 191–203. Spencer, B. and F. J. Gillen. 1899. The Native Tribes of Central Australia. London: MacMillan. ———. 1927. The Arunta: A Study of Stone Age People. London: MacMillan. Strehlow, T. G. H. 1947. Aranda Traditions. Melbourne: Melbourne University Press. ———. 1951. “Foreword.” In Modern Australian Aboriginal Art, ed. R. Batterbee. Sydney: Angus and Robertson. ———. 1970. “Geography and the Totemic Landscape in Central Australia: A Functional Study.” In Australian Aboriginal Anthropology: Modern Studies in the Social Anthropology of the Australian Aborigines, ed. R. M. Berndt, 92–140. Perth: University of Western Australia Press. Travers, M. 2015. “Continuity and Change: Exploring Stylistic Transitions in the Anthropomorphic Figures of the Northwest Kimberley Rock Art Assemblage and the Varying Contexts of Rock Art Production.” PhD thesis. Armidale: University of New England. Travers, M., and J. Ross. 2016. “Continuity and Change in the Anthropomorphic Figures of Australia’s Northwest Kimberley.” Australian Archaeology 82(2): 148–67. Vinnicombe, P. 1976. People of the Eland. Pietermaritzburg: University of Natal Press.

MAKING A SCENE

161

Walsh, G. L. 1994. Bradshaws: Ancient Rock Paintings of Northwest Australia. Carouge-Geneva: Edition Limitée. ———. 2000. Bradshaw Art of the Kimberley. Toowong: Takarakka Nowan Kas Publications. Welch, D. M. 1993. “The Early Rock Art of the Kimberley, Australia: Developing a Chronology.” In Time and Space Dating and Spatial Consideration in Rock Art Research, Papers, ed. J. Steinbring and A. Watchman, 13–21. AURA Publication No 8. ———. 2007. “Bradshaw Art of the Kimberley.” In Rock Art of the Kimberley, Proceedings of the Kimberley Society Rock Art Seminar, ed. M. Donaldson and K. Kenneally, 81–100. Perth: Kimberley Society. Woolagoodja, D. 2007. “Keeping the Wandjina Fresh.” In Rock Art of the Kimberley, Proceedings of the Kimberley Society Rock Art Seminar, ed. M. Donaldson and K. Kenneally, 29–31. Perth: Kimberley Society. Wylie, A. 1985. “The Reaction against Analogy.” In Advances in Archaeological Method and Theory, ed. M. J. Schiffer, 63–111. New York: Academic Press.

==== •

•

•

11. SCENE BUT NOT HEARD Seeing Scenes in a Northern Australian Aboriginal Site Madeleine Kelly, Bruno David, and Josephine Flood

Introduction Scenes are commonly identified in Western art history, where they typically refer to framed images containing subjects performing actions, or snapshots of landscapes. In rock art research, scenes are also sought, especially for their ability to provide insights into (usually non-Western, and past) social practices and associations between people, animals, plants, and objects. Yet the term “scene” is rarely explicitly defined, and it is often used interchangeably with other terms. This disregard for definition has been coupled with insufficient reflection on Western assumptions that underlie research, and on the role of Indigenous visual expressions in the artistic production of many scenic narratives. Two preeminent rock art researchers, Tilman Lenssen-Erz (1989, 1992; Fritz et al. 2013) and Livio Dobrez (2008, 2011a, 2011b, 2012, 2015, 2016), have bucked the trend by defining “scenes,” as Sally May and Ines Domingo Sanz (2010), Carole Dudognon and Marcela Sepúlveda (2017), and Iain Davidson (2017) have also done in more recent studies. Still, there usually remains little (if any) consideration of whether non-Western visual conventions are, and can be, accommodated in Western definitions of rock art scenes (see Kelly and David, this volume). How can cultural differences in visual conventions affect what we understand a scene to be in rock art? Our primary aim in this chapter is to explore this cross-cultural gap through Aboriginal narratives of a major Wardaman rock art site.

Archaeological Research in Wardaman Country Wardaman Country is situated in semi-arid sandy plains in tropical northern Australia. Lying between Wugimadgun (Scott Creek) and Mululeyn (Katherine River) to the north, and Wombila (Romula Knob) and Langaay (Victoria River) to the south (Merlan 1994a: 7), the region abounds in rock shelters, many of which contain rock art, especially paintings and engravings (see Flood 1997: 300–20; 2006). Despite considerable dispossession and direct violence from European colonial practices dating to the late 1800s and early to mid-1900s especially, members of the Wardaman community have for the most part remained on customary lands, and until the 1980s and 1990s largely continued to speak Wardaman language, maintaining a detailed knowledge of local history and traditions (David 2002: 81, 86; Merlan 1994b; Rose 1992: 9). Members of the Wardaman community continued to create rock art until about fifty years ago, and rock art sites persist to this day in Wardaman conceptions of the world. Understanding of aspects of Wardaman ontology and engagements with Country is crucial to understanding the art. Academic research in Wardaman Country began in 1912, when Baldwin Spencer (1914: 20, 54–55, 194–99) briefly recorded details of totemic organization and kinship affiliations. More recent research has focused on rock art recording and archaeological excavation (e.g., Arndt 1962; Barrett and Croll 1943: 45–9; Clarkson 2007; Cundy 1990; Davidson 1935, 1936; Harney 1943, 1959; Lewis and McCausland 1987; Mulvaney 1975: 184–89, 217, 234–43). Social anthropological, linguistic and historical research has been undertaken by Francesca Merlan (1989, 1994a, 1994b, 1998), Deborah Bird Rose (1992) and

SCENE BUT NOT HEARD

163

Darrell Lewis (Lewis and Rose 1988). However, the most intensive phase of rock art research was The Lightning Brothers Earthwatch project (1988–1992), which saw the systematic recording of some 47,000 motifs from over 200 rock art sites, along with associated Wardaman knowledge of both motifs and sites (Flood 1997: 300). Led by archaeologist Josephine Flood with contributions from other archaeologists and anthropologists, including Bruno David, Robin Frost, Bryce Barker, Ian McNiven, Val Attenbrow, Robert Gunn, Jackie Collins, and Francesca Merlan, the Lightning Brothers Project worked on-site with many members of the Wardaman community of all ages, including Elders Elsie Raymond Nonomarran and her family, Tarpot Ngamunugarri, Julai Blutja, Queenie Morgan Ngabijiji, Lily Gin.gina, Riley Birdum, Jessie Brown, Ruby Alison, Riley and Daisy Gimin (e.g., Flood and David 1994: 6; Merlan 1994a: 9). Contemporary Wardaman Elder Billie Yidumduma Harney was also involved in much discussion and was awarded the 1988 Australian Heritage Award for producing (with Film Australia, George Chaloupka and Josephine Flood) the film Land of the Lightning Brothers, which included Wardaman re-enactment of traditional ceremonies at Yiwarlarlay (Film Australia 1987). The Lightning Brothers Project found that the rock art consists primarily of paintings and engravings in a broad range of techniques, with little clear evidence of distinct rock art style sequencing through time, although research into this question is ongoing (David et al. 1999: 16; Flood 1997: 303). A paucity of clear style changes is probably due to the softness of the sandstone in this monsoonal setting, meaning that older artworks tend to weather away relatively quickly, despite excavated archaeological evidence for much longer cultural sequences dating back more than 10,000 years (Clarkson and David 1995: 29; David et al. 1995: 6; for further details of excavated sites and their cultural sequences, see e.g., Attenbrow, David, and Flood 1995; Clarkson 2007; David et al. 1990, 1994, 1995; McNiven, David, and Flood 1992; Watchman et al. 2000).

Wardaman Country As is the case with other Indigenous Australians, Wardaman relationships with Country cannot be reduced to geographic ownership or subsistence economies, as the land forms part of a more encompassing and more socially and existentially nuanced “system of relatedness” (Bradley and Johnson 2014/15: 19). For Wardaman people, such systems of relatedness revolve around the notion of buwarraja, “Dreaming.” While Wardaman notions of Dreaming relate to local world views, there is much cosmogonic equivalence with the same word used for cosmologies in other parts of Australia (David 2002: 72; Merlan 1989). Buwarraja concerns how the world operates, incorporating notions of beginnings when Ancestral Beings formed the land, all life within it, and the laws handed down through ancestral forces (Lewis and Rose 1988: 50). Wardaman ancestry and that of all animate and inanimate things is traced back to these Ancestral Beings and their creative acts (Flood and David 1994: 8; Rose 1992: 86). Of key importance is a logic of maternal inheritance known as ngurlu, which concerns relationships between a person and their related animal or plant taxa; ngurlu is sometimes referred to anthropologically as matri-totems, and in Aboriginal English as Dreamings, such as in “that’s my dreaming” (see Billy Yidumduma Harney in Allam 2014: 19–40; David 2002: 72; Merlan 1989, 1994a: 451; Rose 1992: 82). Ngurlu are associated with particular Ancestral Beings, and all people, animals, and plants that share a ngurlu are kin (Rose 1992: 82). Paternal Dreamings, or kuning, are Ancestral Beings that tie people to Country, with rights and responsibilities to patri-estates traced through one’s father and his kuning (David 2002: 72; Rose 1992: 82). Wardaman Country is divided into eleven patri-estates, each the domain of an Ancestral Being, such as gulirrida—peewee (magpie-lark, Grallina cyanoleuca)—in central Wardaman Country (Merlan 1994a: 6). The madin, “language” or “word” of peewees is mamundajgani, which is also the name of the people paternally linked to the estate of gulirrida Dreaming. Wardaman people inherit their mother’s kin and their father’s Country, although maternal links to Country and paternal kinship links also hold significance (Merlan 1989, 1994a). Through this complex web of paternal and maternal Dreamings and other kin relationships informed by the Dreaming, the people, animals, plants, and places of Wardaman Country are unified in a complex existential system that shapes Wardaman perceptions of and engagements with the world.

164

MADELEINE KELLY, BRUNO DAVID, AND JOSEPHINE FLOOD

The patri-estates of Wardaman Country are determined by major Dreaming sites and associated Ancestral Beings, rather than by cartographic boundaries. For example, the peewee estate—gulirrida Country— contains a number of large rock art sites that, according to Wardaman cosmology, are the ancestral peewees metamorphosed into the rock (Merlan 1989: 18). Traveling Ancestral Beings such as gorondolni the Rainbow Serpent1 are also present. They cut across patri-estates, roaming through Wardaman Country and the lands and waters of neighboring language groups (David 2002: 72; Flood 1997: 301). Places associated with localized and traveling Ancestral Beings can take many forms, including water holes, trees, boulders, and rock art shelters. According to Wardaman cosmology, these places are shaped (and not just “were shaped,” for the acts of the Dreaming are ever-present and ever-enduring rather than simply past events) through the creative acts of the Ancestral Beings, and are often material expressions of the Ancestral Beings themselves, transformed into features of the landscape during the creative era of the Dreaming (David 2002: 72; Flood 1997: 302; Merlan 1989: 16).

Rock Art and the Dreaming The constellation of Dreamings across the Wardaman landscape also creates a framework through which Wardaman people made and interpret rock art. Merlan (1989: 18) described a visit to two rock art sites near Yingalarri water hole in the north of Wardaman Country, where several large paintings dominate the walls. Wardaman Elders initially related the paintings to the large and visually similar paintings of the Lightning Brothers at the major southern Dreaming site of Yiwarlarlay. However, several senior Wardaman Elders later argued that this identification was incorrect, because although the images looked similar to those at Yiwarlarlay, here they were instead gulirrida (Merlan 1989: 18). The Elders reached this conclusion because gulirrida are a defining feature of the estate that contains the site and paintings in question. Furthermore, this area is not associated with the Lightning Brothers at all. A site’s location relative to the location of emplaced Dreamings, and thus relative to other sites that have those Dreamings, helps identify the significance of each place in the buwarraja landscape. Buwarraja is of central importance in Wardaman interpretations of rock art motifs. This is a key, defining framework through which the Wardaman understand the identity of the sites and their art (for a discussion of —bulawula art—that attributed by Wardaman Elders to marlarluga (“the old men/people”) or remembered artists; see Merlan (1989: 16), Flood (1997: 305–6) and David (2002: 72). The Wardaman stress that buwarraja images “are not of human origin, nevertheless they can be and should be painted over by people to make them new” (Merlan 1989: 17). This task of painting or repainting is ideally taken up by someone paternally linked to the place that contains the art and its associated Dreamings. Engravings are also mostly buwarraja. For example, Ngalanjarri, the place of the rain Dreaming, is a rock outcrop covered in abraded grooves near Yiwarlarlay. Merlan visited the outcrop in the 1980s in the company of several Wardaman who indicated that it was forbidden to stand on the rock or engrave further grooves into its surface, although they frequently do this elsewhere in Wardaman Country. Merlan’s Wardaman informants added that the grooves are called maburn, and that they were not made by people but created in the Dreaming, with people copying them to make cicatrices (scars) also called maburn on their bodies (Flood, David, and Frost 1992: 34; Merlan 1989: 17). While buwarraja rock art can be “made good” by the right people, its origins hark back to the Dreaming, so like the land, the art expresses the agency of both Wardaman human ancestors and Ancestral Beings, forming a complex web of relationships structuring and uniting the Wardaman physical-cultural-social landscape. Understanding the formal features and structure of buwarraja rock art has been a focus of research in Wardaman Country. Merlan (1989: 20), Flood (1997: 303–4) and David (2002: 73) note that major buwarraja images, which like features of the landscape are visual manifestations of Ancestral Beings or the results of their acts (Merlan 1989: 16), are often very large, centrally located within rock art sites, and depict human or animal figures. As Merlan (1989: 14) aptly concludes, “anything we would want to call ‘meaning’ of the art lies in the relation between images and their understood interpretation or sense within the terms of the contemporary narrative tradition.”

SCENE BUT NOT HEARD

165

Yiwarlarlay: Archaeology and Rock Art Yiwarlarlay is an isolated set of sandstone stacks among sandy plains in southern Wardaman Country. The outcrop contains several shelters richly imbued with rock art. There are three major shelters— Yiwarlarlay 1, Yiwarlarlay 2 and Delamere 3—and four minor ones. Archaeological excavations and dating of the on-wall art at Yiwarlarlay 1 and Delamere 3 indicate that people have been making rock art in this site complex since at least 3215–3554 cal BP (3160±60 BP, at 95.4 percent probability calibration with IntCal13 Curve Selection, on Calib 7.1) (Watchman et al. 2000: 323), but that at Yiwarlarlay 1 painting activity increased dramatically in the late nineteenth century, that is, the early European-contact period. Painting of the two large, striped anthropomorphs that dominate the main wall of Yiwarlarlay 1 began in the late 1800s (e.g., David 2002; David et al. 1990; McNiven, David, and Flood 1992). Josephine Flood, Bruno David, and Robin Frost recorded the rock art of Yiwarlarlay in 1989 with the assistance of Earthwatch volunteers. A total of 2046 motifs were identified from the seven sites in the Yiwarlarlay complex, with a “motif ” defined as an image made of an individual mark or set of marks of human origin. Seven hundred and fifty-four of the motifs are paintings, stencils, or prints. Red, white (including creamy whites), yellow, black, and pink colors were all used. According to Chaloupka (1978: 10), the white (barnjan) pigment sources are found near the site, the black (minyardin) is charcoal, the red (liwin) is from nearby Yerrlyn, and the yellow (gilirringa) is from the Yingalarri water hole. We will return to the significance of these colors below.

Yiwarlarlay 1 Yiwarlarlay 1 is a large, open shelter that contains the most rock art and the most dominating paintings in the site complex. The 7m-high main decorated wall contains the two large, striped anthropomorphs identified by Wardaman Elders as the Lightning Brothers (Figure 11.1). A large boulder opposite the main wall also features rock art. Overall, c. 55 m2 of Yiwarlarlay 1’s rock surfaces are painted and 32 m2 are engraved. Yiwarlarlay 1 contains 1257 motifs (61 percent of the motifs in the Yiwarlarlay site complex), all of which are engravings and/or paintings except for two grass prints (Table 11.1).

The main wall: Rock art techniques and motif types The main wall at Yiwarlarlay 1 has 645 motifs, consisting of 324 engravings, 319 paintings (nine of the engravings have paint infill and six of the paintings have engraved parts) and two small grass prints on the upper northeastern section of the wall. Most of the engravings are abraded grooves (either single or in sets), or pecked (and sometimes abraded) macropod tracks. A small number of other abraded or pecked animal tracks are also present, including in the shape of human feet,

FIGURE 11.1. The Lightning Brothers at Yiwarlarlay 1. Photo by Bruno David.

166

MADELEINE KELLY, BRUNO DAVID, AND JOSEPHINE FLOOD

TABLE 11.1. Motif counts and techniques at Yiwralarlay, by site. Rock Art Technique Engraving Site

Painted Engraving

Painting

Engraved Painting

Print

#

%

#

%

#

%

#

%

#

%

Yiwarlarlay 1

816

39.9

9

0.4

424

20.7

6

0.3

2

0.1

Yiwarlarlay 2

48

2.3

85

4.1

1

0.1

366

17.9

126

6.2

62

3.0

39

Yiwarlarlay 5

Stencil #

%

Total Motifs #

%

1257

61.4

3

0.1

137

6.7

1

0.1

507

24.8

1.9

101

4.9

5

0.2

5

0.2

Yiwarlarlay 6

38

1.9

38

1.9

Yiwarlarlay 7

1

0.1

1

0.1

2046

100

Delamere 3 Yiwarlarlay 4

Total

10

0.5

4

0.2

1292 (63.1%) 19 (0.9%) 718 (35.1%) 10 (0.5%)

3 (0.2%)

4 (0.2%)

birds, and echidna. The wall also features low numbers of drilled holes and pecked and abraded hollows. Generally, engravings appear on the lower sections of the wall. Eighty-nine of the paintings are anthropomorphs: nine have breasts and/or vulvas, suggesting females, and eight have penises, suggesting males. The remainder are either unsexed or are too faded to tell. The main wall also has 76 paintings of zoomorphs, most commonly macropods (kangaroos, wallabies, or bettongs), then echidnas, birds (including emus), snakes, and dogs, as well as a turtle, a stingray, a bat, and a fish. Boomerangs are the only material culture item depicted by themselves (i.e., not directly associated with anthropomorphs). Of the remaining paintings, 13 are non-figurative and 151 are indeterminate (their shapes being too damaged to determine).

The decorated boulder: Rock art techniques and motif types There are 612 rock art motifs on the boulder facing the main wall. Most (n = 501) are engravings, primarily abraded grooves occurring singly or in sets. The engravings also include nine drilled holes, ten pecked and in some cases abraded hollows, a single pecked macropod track, an abraded anthropomorph with a rayed headdress, and five pecked and abraded vulvas. While paintings are fewer (n=111) than engravings, they are larger than any single engraving and dominate the overall boulder surface area. The paintings are often superimposed over the engravings, with 45 percent of paintings being more than 20 cm long. Most of the paintings are anthropomorphs, generally without distinctive features, although nine have rayed headdresses and six have unusually large black eyes. These large-eyed anthropomorphs are not found elsewhere at Yiwarlarlay. There are also twenty-six zoomorphs, including birds, macropods, snakes, a bat, and a turtle. A further eleven non-figurative and forty-four indeterminate paintings are present.

Formal scenes at Yiwarlarlay 1 Under the criteria set by Lenssen-Erz (1992; Fritz et al. 2013), and Dobrez (2016), rock art scenes must depict narratives that are “readable” through the art’s formal characteristics, be it the depiction of interaction (Lenssen-Erz) or the insinuation of movement (Dobrez) (see Kelly and David this volume). At Yiwarlarlay 1, the main wall consists of a large, flat surface with three panels divided by large natural cracks. Motifs occur on either side of the cracks, but not across them, allowing for a convenient differentiation of wall panels. Facing the wall, Panel 3 on the right hand-side is the largest of the three. Panel 2 is to its

SCENE BUT NOT HEARD

167

left, and Panel 1 is further to the left again. While some panels have high densities of motifs, often juxtaposed or in partial superimposition, very few show any hint of interaction or movement. Further details of the art on each panel can be found in Kelly (2017). There is little on the main decorated wall that conforms to either Lenssen-Erz’s or Dobrez’s formal notion of a scene. A very few individual motifs (such as Motif #676, a running man on the main decorated wall at Yiwarlarlay 1, or a baby Rainbow Serpent apparently wriggling nearby at Delamere 3) signal a degree of motion across the observer’s field of view, although for Dobrez (2012: 12) a depiction of an “animal doing such and such” is not a sufficiently elaborate visual narrative to classify as a scene. Two sets of motifs nevertheless fulfil the requirements set out by Lenssen-Erz. The first set, in the right section of Panel 3, comprises a red female anthropomorph with a headdress (#669) and an upside-down male anthropomorph (#672) who appear to be copulating. The second features two red birds (#305, #351) appearing to follow one another in profile view, below the two large striped anthropomorphs. The paucity of formal scenes presents the rock art of Yiwarlarlay as rather unscenic and made up of many individual, more or less isolated static depictions (as befits local conventions of image-making).

The Lightning Brothers at Yiwarlarlay Let us now look at the rock art of Yiwarlarlay from a Wardaman perspective, and ask again about where scenes may lie. Rather than start with the art’s formal properties, we will begin with the Wardaman narrative that relates to the art. Janginyina-wuya (the two geckos) are the Lightning Brothers Yagjagbula and Jabirringi, the major Ancestral Beings in the Wardaman patri-estate at Yiwarlarlay (Merlan 1989: 20; 1994a: 359). The Lightning Brothers’ Dreaming story, recounted below, has been reported many times. The version below was recorded by Flood and David during the Lightning Brothers Project (David 2002: 75; Flood and David 1994: 10). Wardaman Elders Elsie Raymond Nonomarran (who was patrifiliated with Yiwarlarlay) and Tarpot Ngamunugarri told the story (David 2002: 75): Of the brothers Yagjagbula and Jabirringi, Yagjagbula is the younger. He is tall and handsome, while his older brother is short and not so attractive. Both brothers are of the Jabijin sub-section. Yagjagbula has a wife, Gulliridan, and Jabirringi is married to Ganayanda (some people say that Jabirringi has been promised to Ganayanda, but that both brothers could potentially have married her, given the correct sub-section affiliation). Each day one of the brothers goes hunting for food, bringing the day’s catch back to camp where it is shared by all. One day it is Yagjagbula who hunts; the next it is Jabirringi. One day Jabirringi returns from the hunt to hear Ganayanda whispering with Yagjagbula in a secluded break in the rock. Suspicious, he investigates and finds them copulating. Jabirringi throws a spear at Yagjagbula, who evades it. A fight breaks out, with each brother taking his position on the plains at Yiwarlarlay. Spears and boomerangs are thrown, the fury of the fight producing lightning and thunder. The lightning strikes the sandstone outcrop at Yiwarlarlay and splits the rock in two. The Frogs come up from the south to watch the fight, clapping their thighs rhythmically. Wiyan, the Rain, who has been heading north to the Yingalarri water hole, gets distracted while passing by Yiwarlarlay (at the same time, the Rainbow Serpent, Gorondolni, flashes at the Rain to warn it not to advance to Yingalarri). Wiyan metamorphoses into the Rain rock Ngalanjarri nearby. Eventually Yagjagbula hits Jabirringi across the forehead with his boomerang. Some say that Jabirringi’s headdress is knocked off, others say that Jabirringi is decapitated. In either case, Yagjagbula wins the fight. Jabirringi’s headdress metamorphoses into a rock that could be seen near Yiwarlarlay until it was allegedly stolen by European pastoralists, while the Lightning Brothers transform into the rock. Elsie Raymond Nonomarran had told a very similar version of the Dreaming story to Merlan (1978) a decade earlier, and Lily Gin.Gina also recounted some of the details in another account (Merlan 1994a: 538–40). Earlier still, Davidson (1936) recorded a version in the early years of the twentieth century, noting that the Lightning Brothers “Yagtchadbulla” (Yagjagbula) and “Tcabuinji” (Jabirringi) had fought over “Karnanda” (Ganayanda) (Davidson 1936: 112). While his account lacks some of the details found in the

168

MADELEINE KELLY, BRUNO DAVID, AND JOSEPHINE FLOOD

more recent versions told by Elsie Raymond Nonomarran, Tarpot Ngamunugarri and Lily Gin.Gina, the general narrative remains the same. Following Davidson, William Harney (1943, 1959) also described a dispute between the two brothers over the older brother’s wife. However, in a later study, Arndt (1962: 165) noted that the details of William Harney’s multiple accounts and his description of the site could not always be corroborated, nor did they quite match Davidson’s recordings or those taken by Arndt himself. Arndt (1962) undertook the most detailed study of Yiwarlarlay prior to the Lightning Brothers Project. His primary informant was Wardaman Elder Kulumput, who recounted the Lightning Brothers story. Kulumput’s version differs only slightly from that subsequently given by Elsie Raymond Nonomarran and Tarpot Ngamunugarri. Rather than using boomerangs, according to Kulumput, Jabirringi used a djugalutba (stone axe) that was so strong he could split whole trees with it (Arndt 1962: 167). He further explained that lightning strikes split trees as a result of Jabirringi’s actions with his axe. Arndt also explains the significance of the colors used in the paintings (see below). Kulumput reported the following additional details to Arndt (1962: 169–70): . . . the Lightning Brothers originally “camped” on the Victoria River, where several neighboring tribes were free to visit them. When the country and the people were divided between rival pastoral interests, it was no longer practical for the Wardaman people to visit the [original] Lightning Place. The Wardaman elders at Delamere Station decided that the Lightning Brothers could “camp” at the Rain Place near the homestead, so that they could be seen by the rising generation. Emu Jack, a contemporary of Kulumput, “dreamed” (visualized) the design and did the painting. The task was delayed by station and tribal duties and was not finished until he was in bush-retirement prior to his death “near the end of the Japanese war (1945).”

Paintings of the Lightning Brothers: Ancestral Beings in the rock Wardaman people who spoke with Davidson (1936), Arndt (1962), Merlan (1989), and Flood and David (1994) all identified the two striped anthropomorphs (#548, #568) that dominate the main decorated wall at Yiwarlarlay 1 as the Lightning Brothers. The larger of the two is Yagjagbula and the smaller is Jabirringi. Kulumput revealed that the vertical yellow and red stripes on the Lightning Brothers’ bodies represent the body markings of a red and yellow goanna, lardaja, who marked the trees that Jabirringi would split in future storms. The black pigment indicates exceptional strength. Men who created thunderstorms had to have strong ears and eyes to withstand the thunder and lightning flashes; hence, the Lightning Brothers have black ears and eyes. The Lightning Brothers also have a black band across their feet because their feet would ache from the powerful stomping men do to warm up for a fight. In the case of the Lightning Brothers, such stomping also produces the thunder in lightning storms. Since in Arndt’s version Jabirringi was the most enraged of the two brothers, he has stronger, thicker ankles. Other black bands across the Lightning Brothers’ bodies denote their sexual strength. Kulumput also identified the headdress worn by Yagjagbula as a cumulonimbus storm cloud shaped like the turbans of Afghan camel men, who commonly frequented the region in the early 1900s (Arndt 1962: 167–68). Wardaman understand the paintings of the Lightning Brothers at Yiwarlarlay to be buwarraja, expressions of the Dreaming. As noted above, rock art identified as buwarraja is generally either the result of an Ancestral Being’s actions, or the metamorphosed form of an Ancestral Being. In Wardaman cosmology, the paintings of the Lightning Brothers are such metamorphosed Ancestral Beings, as are the rich storylines they embody: they are not merely the representations of the Lightning Brothers, they are the brothers “sitting” in the rock, embodying the full Lightning Brothers story of Yiwarlarlay in the making (David 2002: 75; Merlan 1989: 21). Davidson’s (1936: 112) informants stated that the Lightning Brothers “exist at the current time, but frequent the totemic centre only during the rainy season. The rest of the year they are away visiting other sites.” The paintings are thus active presences at Yiwarlarlay, rather than passive recipients of meaning (contra Gell [1998]; for views toward non-human agency, see David (2002: 67–74) for the way things call on us to behave in certain ways; also see Brady, Bradley, and Kearney (2016) for the agency of the numinous among the Yanyuwa; see also I. Davidson, this volume). They actively interact in and with the surrounding landscape and with the Wardaman. These interactions between the Lightning Brothers and

SCENE BUT NOT HEARD

169

associated personages, as well as with descendant Wardaman who make living sense of the Dreaming story, lend themselves to recognition as scenes.

Buwarraja rock art at Yiwarlarlay 1 Wardaman Elders have identified many of the individual motifs at Yiwarlarlay 1 as buwarraja, although the identification of specific paintings sometimes varies, particularly those of the small striped anthropomorph (#643) under Jabirringi’s right arm (Figure 11.2b). Davidson (1936: 113) identified this figure as “Karnanda” (Ganayanda), Jabirringi’s wife, while Kulumput and other Wardaman men with him insisted the figure was Jabirringi’s stone axe (Arndt 1962: 167). The men “pointed out that men commonly carry stone-axes and similar articles by clamping them under the arm” (Arndt 1962: 167). During the Lightning Brothers Project, the small striped anthropomorph was identified as the Lightning Brothers’ unnamed garlaja (younger brother) (David 2002: 75; Merlan 1989: 20). While the identity of this motif may have changed over the years, or may have had different meanings to different Wardaman individuals, whether it be Ganayanda, Jabirringi’s stoneaxe or the Lightning Brothers’ garlaja, it has in each case played a role as an active agent in the Lightning Brothers narrative, actively connecting one image with another and with the broader narrative scene. While Arndt (1962) and Davidson (1936) say little of the other motifs surrounding the Lightning Brothers, Merlan (1989) and the Lightning Brothers Project’s recordings detail the role of some of the smaller motifs in the Lightning Brothers story. Merlan (1989: 20) identifies a small red female anthropomorph with headdress (#669) as Ganayanda (Figure 11.2c), as do Flood and David in unpublished recordings. Merlan (1989: 20) also identifies one of the female anthropomorphs on the decorated boulder opposite the main wall at Yiwarlarlay 1 as Gulliridan, Yagjagbula’s wife. A painted white dog with red outline (#550) near Yagjagbula’s feet was identified by Elsie Raymond Nonomarran as Ngajamberle, the Lightning Brothers’ dog (Merlan 1989: 20) (Figure 11.2c). She also remarked that although only one dog was depicted, the Lightning Brothers had a second dog named Modborronggo, later corroborated by another Wardaman Elder who was not present with Merlan and Elsie Raymond Nonomarran in the field. The other nearby zoomorphs at a

c

b

FIGURE 11.2a–c. Rock art motifs at Yiwarlarlay 1. (a) Dreaming beings (animals) hunted by the Lightning Brothers. (b) Garlaja (younger brother). (c) Ngajamberle, the Lightning Brothers’ dog. Figures 11.2a and 11.2b by Bruno David. Figure 11.2c copyright Josephine Flood.

170

MADELEINE KELLY, BRUNO DAVID, AND JOSEPHINE FLOOD

Yiwarlarlay 1 were identified by Wardaman informants during the Lightning Brothers Project as animals that the Lightning Brothers hunted (Figure 11.2a). The Wardaman ethnography gathered during the Lightning Brothers Project reveals that all these smaller motifs are interrelated through their participation in the Lightning Brothers story. They are the visual expression of the Lightning Brothers Dreaming, a visual narrative that together constitutes an overarching scene. That scene is not limited to a single panel, nor to a group of panels in a single rock shelter, but incorporates other nearby shelters and the broader landscape setting. Merlan (1989) thus also recorded several rock art motifs relating to the Lightning Brothers story at Delamere 3, on the other side of the stone outcrop a

b

d

c

e

FIGURE 11.3a–e. Rock art motifs at Delamere 3, (a) Abraded human feet. (b) Pecked human feet. (c) Painted boomerang motifs. (d) Walanu (bush potatoes). (e) View inside of the rock shelter at Delamere 3, featuring the Rainbow Serpent. Copyright Josephine Flood.

SCENE BUT NOT HEARD

171

to Yiwarlarlay 1. There, Wardaman informants identified a group of circular paintings as walanu (bush potatoes), one of the food items consumed by the Lightning Brothers (Merlan 1989: 20) (Figure 11.3d). In that same shelter, the two boomerang motifs are said to be the Lightning Brothers’ weapons (Figure 11.3c). Again, these groups of motifs are said to be buwarraja by Wardaman Elders: they are the metamorphosed objects that the Lightning Brothers used during the creation period known as the Dreaming, and they contain the agency of buwarraja just as Yagjagbula and Jabirringi do. Another metamorphosed Ancestral Being at Delamere 3 is a large, 2.5 m-long snake (#66) located on the ceiling in the back recess of the shelter (Figure 11.3e). Except for the Lightning Brothers themselves, the snake is the longest motif at Yiwarlarlay. Merlan (1989: 20) identified the motif as a rainbow serpent. Earlier, Kulumput had also identified the motif as a “baby” rainbow serpent “because it had ears” (Arndt 1962: 175). Kulumput recounted that “the ‘baby’ came to ‘look’ at Lightning from its home in a cave to the west where the ‘big’ Rainbow Snake, called Ku-rakan, lives” (Arndt 1962: 175). The baby rainbow serpent is, like the Lightning Brothers, a Dreaming Being “sitting” in the rock. Several large and small engraved human footprints on the boulders in front of Delamere 3 (Figures 11.3a, 11.3b) were identified as the footprints of the two Lightning Brothers as they moved around the shelter. As outcomes of the Lightning Brothers’ actions during the creation period, they too are buwarraja.

The Lightning Brothers landscape The landscape of Yiwarlarlay also forms an integral part of the Lightning Brothers story. As the overarching framework through which the Wardaman understand the workings of the world, buwarraja expresses itself in both rock art and broader features of the land. The Lightning Brothers landscape is no different: many physical features of the rock outcrops and plains play important roles in the Lightning Brothers Dreaming story. Yiwarlarlay is the home of the Lightning Brothers and the location of their fight. Wardaman people identify certain locations as places where particular events take place in the Lightning Brothers story: a recess in the rock, where Yagjagbula and Ganayanda were found copulating; the plains fronting the rock outcrop and art sites where the fight between the Lightning Brothers took place; a large split in the main rock outcrop caused when lightning struck as the brothers fought (David 2002: 75–78; David et al. 1991: 73–74) (Figure 11.4). A further site near Yiwarlarlay, Barrgu-ya (“at the fighting-stick”), contains several elongated rocks, identified by Elsie Raymond Nonomarran as the rocks Ganayanda threw in her efforts to stop the Lightning Brothers from fighting (Merlan 1989: 20). Each of these places contains an episode of the Lightning Brothers story. Their presence in the landscape is as much a part of and as constructive of the Lightning Brothers Dreaming story as are the footprints in stone or the paintings of Ancestral Beings on the walls. Some features of the landscape are also identified by Wardaman Elders as the bodies or body parts of other Ancestral Beings metamorphosed into rock. The frogs that came to watch the Lightning Brothers fight, clapping their thighs and chanting, are transformed into small sandstone outcrops spread across the adjacent plains (Figure 11.4e). Another nearby place is the metamorphosed wiyan (rain Dreaming), who sat down facing toward his Country and became Ngalabjarri (the water rock) (Merlan 1989: 21) (Figure 11.4f). Ngalabjarri is an isolated, imposing rock covered in abraded grooves (engravings). The entire rock and the grooves on it are buwarraja. It is strictly forbidden to abrade fresh grooves on the rock or to climb it; Wardaman Elders say that the rock would spin upwards and throw violators off. Elsie Raymond Nonomarran explained that the grooves on Ngalabjarri were not created by human hand but done during the Dreaming and then copied by the Wardaman when they make cicatrices on their skin. Like the Lightning Brothers and many other paintings at Yiwarlarlay, they share the same significance as the rock art in the Lightning Brothers narrative, being part of the landscape’s visual narratives and connecting the rock art motifs with one another and with the broader landscape. The totality, and the parts, are scenes of the Wardaman world.

172

MADELEINE KELLY, BRUNO DAVID, AND JOSEPHINE FLOOD

a

b

d

c

e

f

FIGURE 11.4a–f. (a) The break in the sandstone where Yagjagbula and Ganayanda were caught. (b) The split caused by the lightning. (c) and (d) The sandy plain at Yiwarlarlay, where the Lightning Brothers fought. (e) A frog who came to watch the fight. (f) Ngalanjarri (the water rock), the metamorphosed form of wiyan. Figures 11.4a, b, d by Bruno David. Figures 11.4c, e, f copyright Josephine Flood.

Scenes of the Lightning Brothers Yiwarlarlay’s rock art attains its meaningfulness through a rich narrative that interconnects the imagery with the landscape, individually and together rendering scenes of buwarraja activity. Yet a more formal analysis of the art fails to reveal more than a handful of isolated, tentative indications of actions and scenes, as if each image was positioned in isolation from a broader scenic narrative. Approaching scenes through a

SCENE BUT NOT HEARD

173

Wardaman ontological framework, in recognition of Wardaman interpretative conventions, allows us to develop new and arguably culturally more relevant notions of Wardaman scenes. The Lightning Brothers are the central actors in a broader landscape scene at Yiwarlarlay. They sit in the rock of their active Dreaming landscape. The paintings of the two brothers are themselves a scene that locates the Lightning Brothers story both before and after the brothers’ fight. These paintings are not mere depictions—passive signifiers of a narrative—but agentive Dreaming Beings with active gazes, continuing actions and interrelationships with the surrounding landscape and with contemporary members of the Wardaman community. A more applicable description might be that the paintings provide a view of the Lightning Brothers sitting in the rock. This scenic notion also applies in a Western setting, with the Oxford English Dictionary (2015: 15) defining a scene as “a view or picture presented to the eye or mind of a place, incident, series of events,” although as far as we know it has not previously been applied in quite this way to scenes in rock art. In Wardaman ontology, the view of the Lightning Brothers is fluid and changeable rather than static, allowing for the brothers’ continuing agency and actions both at the site and across the landscape. The Lightning Brothers’ agentive nature means that they continue to engage with Wardaman people, who address how the scene is to be perceived. The Lightning Brothers are themselves direct ancestors of many Wardaman people: one of Davidson’s unnamed informants was a brother to the Lightning Brothers, as he too was of the Jabijin sub-group. Because of this kinship relation, he was allowed to marry Ganayanda, who sits next to the Lightning Brothers. This kin affiliation with the Lightning Brothers, and with other Ancestral Beings at Yiwarlarlay, creates another scene: a scene of interaction between kin. Some Wardaman people, as descendants of the Lightning Brothers, constitute another interrelated part of the narrative of the Lightning Brothers Dreaming story, a very real location of connected human and extrahuman life. The scene thus becomes a continually shifting and multi-vocal view of interaction between kin and with non-kin. Many of the smaller motifs at Yiwarlarlay 1 and Delamere 3 articulate through their participation in the Lightning Brothers story, thus forming part of a wider scene. Motifs relate to each other within sites as well as across sites, through their joint participation in the Lightning Brothers Dreaming. For example, the engraved footprints, baby rainbow serpent, and boomerangs at Delamere 3 all interact with the Lightning Brothers at Yiwarlarlay 1 on the other side of the same rock outcrop. The footprints are a scene of the Lightning Brothers traveling around the rock, the rainbow serpent sits watching the Lightning Brothers, and the boomerangs thrown by the brothers come to rest, metamorphosing into the rock. All these motifs, from an archaeological perspective—and despite differences in their ages, form, and location—interact through their shared participation in the Lightning Brothers story embodied in the landscape.2 Like storyboards, each is a separate scene episode in the narrative of the Lightning Brothers Dreaming, and together they signal a complete interconnected scene of the Lightning Brothers Dreaming story.

Scenes in the landscape The broader physical landscape of Yiwarlarlay also forms part of the Lightning Brothers Dreaming narrative. The crack in the Yiwarlarlay outcrop is the result of the Lightning Brothers’ actions: it forms part of the narrative, as does the place where Yagjagbula and Ganayanda copulated and the plain where the brothers fought. These places are scenes of particular events, and while those locations may not themselves be human actors, they remain affective elements of the Lightning Brothers Dreaming in which the paintings are placed. The crack in the rock is not just a crack: in the Wardaman cosmology it is an active and recurrent expression of the Lightning Brothers’ fight. Thus a view of the crack is a view of an episode in the narrative of the Lightning Brothers, as is a view of the plain and the recess in the rock and the place named Barrgu-ya, which contains rocks thrown by Ganayanda as the Lightning Brothers fight. The Ancestral Beings wiyan and the frogs sit in the landscape as well, metamorphosed into boulders, watching the Lightning Brothers. By performing actions—watching, sitting, clapping their hands on their thighs—they thus form scenes, even by more standard archaeological definitions such as that of Lenssen-Erz.

174

MADELEINE KELLY, BRUNO DAVID, AND JOSEPHINE FLOOD

Scenes through time The ongoing, vital nature of buwarraja means that Western notions of temporal progression and chronological separation fail to adequately capture the affective nature of artistic events and how each is incorporated into a cumulative whole. Rock art researchers and archaeologists often attempt to determine the relative and absolute ages of rock art motifs in order to develop an understanding of the temporal sequence of motifs, while nonetheless understanding that each new phase of artwork adds to an already existing body of images. Yiwarlarlay is no exception. Archaeological evidence from radiocarbon-dated charcoal as well as ethnographic accounts indicate that the paintings of the Lightning Brothers were created within the past c. two hundred years (David et al. 1990). Patterns of superimposition further demonstrate that at least some of the smaller motifs at Yiwarlarlay 1 were done even more recently. However, many of the engravings are much older. AMS radiocarbon dating of environmental carbon captured in micro-laminae accumulated in and around engraved macropod tracks at Delamere 3 show that at least some of the engravings are well over three thousand years old, suggesting that other (undated) engravings at Yiwarlarlay are probably of a similar great age as well (Watchman et al. 2000). This archaeological evidence allows many of the paintings to be positioned in time, distinguishing particularly the Lightning Brothers painting of two hundred years from the engravings, which in some cases can be shown to have been done more than three thousand years ago, long before the paintings were made. Yet the engravings of human feet at Delamere 3 are the footprints of the Lightning Brothers, which in Western understandings should appear only after the paintings of the Lightning Brothers (with their painted feet) had themselves come into being at Yiwarlarlay 1. Here the temporality of the narrative tradition lies in an opposite order and a different ontological logic from that of archaeological time. This affects the temporality of the scenes themselves. These and other motifs—which, when seen in Western perspective, are chronologically separated and distinguished from one another—are chronologically and narratively united in their shared participation in the encompassing Wardaman system of buwarraja, a system that sees each image as having an active role in the Lightning Brothers Dreaming story. This brings a problem to the fore: archaeologists tend to explore when things were “originally” made, yet people remain engaged with these things from that point on. It is this fuller (hi)story of engagement that becomes highly significant during attempts to understand the art’s meaningfulness, including in archaeology. Wardaman scenes at Yiwarlarlay alert us to this conundrum.

Discussion Through Wardaman visual conventions, the entire Yiwarlarlay landscape, as well as several surrounding sites, is a scene of the Lightning Brothers Dreaming. The scene is made of many smaller, interrelated episodes and ongoing active presences that shape and change the narrative, interacting with the viewers and their time. While we have identified scenes involving single motifs in both our formal and ethnographically informed analyses, such as Motif #676, the running man on the main decorated wall at Yiwarlarlay 1; or the baby rainbow snake sitting in the rock at Delamere 3—in archaeology, scenes more commonly involve multiple images. The Wardaman scenes at Yiwarlarlay are part of a larger narrative, and scenes involving single motifs form part of larger scenes with multiple interacting motifs. Sven Ouzman (2001: 252) has elsewhere noted that “sight has long enjoyed eminence as the ‘sense of reason’” in Western epistemology. Indeed, the Western approaches reviewed above rely on visual criteria to determine interrelationships such as formal similarities between motifs. Wardaman interpretations, however, explicitly rely on informed knowledge about how things lie in buwarraja, how they are themselves Dreamings. David (2002: 74), Flood (1997: 303–4), and Merlan (1989: 20) thus found little correspondence between the formal appearance of Wardaman rock art motifs and their buwarraja identification with particular types of Dreaming Beings. This incongruity between the visual and the narrative, which is evident in the art and landscape of Yiwarlarlay, is one of the main reasons why formal (Western stylistic) notions of scenes cannot adequately de-

SCENE BUT NOT HEARD

175

scribe Wardaman visual narratives. For example, the motif identified as Ganayanda (#669) at Yiwarlarlay 1 has no internal decoration and is significantly smaller and more faded than the Lightning Brothers nearby. From a Western perspective, Ganayanda’s apparently depauperate formal characteristics often suggest a minor, almost extraneous figure in the broader scheme of things. However, Wardaman ethnography indicates that Ganayanda is a highly significant figure who provides the motivation for the Lightning Brothers’ fight and articulates critical kinship relations. She is directly related to both brothers, as the wife of one and the lover of the other, through their “skin” affiliations, which also express appropriate (and inappropriate) behaviors. It is the informed narrative rather than the formal characteristic that primarily determines how the images at Yiwarlarlay relate to each other. The pre-eminence of informed narrative in Wardaman interpretation also challenges the Western focus on imagery. Ouzman (2001: 238) further argues that because Western notions of reality are deeply dependent on imagery, “the apparently overwhelmingly visual nature of rock ‘art’ may blind us to uses of the imagery, the rock, the rock-art site and the landscape that do not obviously fall within the eye’s compass.” This reliance on imagery is evident in the way notions of scenes tend to be identified in archaeological endeavors. It is common in archaeology to restrict the identification of scenes to images or sets of images that can, through visual conventions of image making, be demonstrably argued to form part of narratives (see Kelly and David this volume). However, we now see that such definitions exclude key features of Wardaman narratives, such as the broader landscape that meaningfully locates the art and the relationships between the Ancestral Beings who are materialized in the art. This is more than erasing the canvas—as if it was neutral to the making of a scene—for it also silences the very setting that makes the scene possible in the first place, and in which the scene is set. Approaching the art through Wardaman notions of visual narratives, as informed by Wardaman ontology, presents what for many may be an unfamiliar notion of a scene that involves rock art, the landscape, and continuing exchanges with living Wardaman people.

Snapshots of the past or/and agents of the future In archaeology, scenes are sometimes treated as fixed “snapshots” of past events, and it is only through the formal characteristics of those frozen moments that particular actions can be identified by an outsider. But in Wardaman ethnography, the rock art motifs are not just images but also vital Beings who interact with, and impact on, each other and their surroundings. Nor is the notion of a scene as a “snapshot” of past events entirely compatible with Wardaman notions of buwarraja and time. Deborah Bird Rose (2004: 55) describes Australian Aboriginal notions of the Dreaming as “enduring yet contingent.” She reasons that “Dreaming and ordinary times are separated by something that could reasonably be called a temporal boundary. Dreamings walked the earth in human form, making, shaping, and bringing into being the world of form and difference, as well as the relationships that cross these boundaries. At some point, they changed over into the species they are now” (Rose 2004: 55). Yet, she argues, this boundary between the “enduring Dreaming” and ordinary time is “cross-cut”: “Dreaming action continues in the present in the bodies of all living things whose origins are in Dreaming. And Dreaming action continues in the present through ceremony, creation, song and other forms of creative memory and connection” (Rose 2004: 56). Rock art is located in such logic and takes on those properties of the Dreaming in the process. Wardaman conceptions of buwarraja, and of rock art, are thus located. At Yiwarlarlay, the Lightning Brothers Dreaming scenes express ever-present and ongoing past events that shape the site as it is seen today. The scenes are views of the enduring, living past that involve vital Beings. Hence the boulder Ngalanjarri, the water rock, is a scene of wiyan as it sits down and becomes Ngalanjarri in the Lightning Brothers’ fight (during the ever-present period of creation). Those who violate the surface of Ngalanjarri in the present day invoke active responses, so that the scene changes from one of Ngalanjarri sitting and watching to one of active response against the intruder. Wiyan is also an Ancestral Being with living descendants. The interactions between wiyan and his contemporary kin also form part of wiyan/Ngalanjarri’s narrative, creating a multivocal narrative that varies according to kin relationships. The multivocality of wiyan/Ngalanjarri’s narrative in turn alters the perception of scenes involving the Being, because scenes are the visual expression of those narratives. Scenes of wiyan/Ngalanjarri are thus multivalent; they express Dreaming narratives, actions, and interactions with kin. In their multivalency, scenes of wiyan/Ngalanjarri are also “multi-temporal” since they express past events, present-

176

MADELEINE KELLY, BRUNO DAVID, AND JOSEPHINE FLOOD

presence and agency for change into the future. The multivocal, multivalent, multi-temporal nature of wiyan/ Ngalanjarri’s scenes is likewise evident in many of the other scenes at Yiwarlarlay, which also express the presence of vital Beings.

Conclusion In the Wardaman language there is no word for “scene.” Most closely related would be some version of the word “see,” but in another form, such as “they saw it” (Francesca Merlan, pers. comm. 2018). Nor is there a word for “art” (Merlan 1989). By applying the concept of “scene” to Wardaman rock “art,” we thus translate the art via a double-distinction averse to the logic inherent in Wardaman art making and art reading. Vázquez (2011: 34) argues that the act of translation is erasure: “Translation is a process of selection, classification and appropriation that erases all that does not fit into the proper place of the already established epistemic territory”. He also notes that “translation enables the coming together of a plurality of movements and by turning difference into a site for struggle it comes to fracture, to challenge the forces of erasure of modernity’s epistemic territory” (Vázquez 2011: 41). Maintaining a plurality of perspectives and challenging of dominant epistemic structures has been a major aim of this chapter. In exploring Wardaman engagements with narrative in rock art, we have tried to question Western assumptions concerning the universality of formal typologies and explore other ways of perceiving “rock art scenes” through the vital stories they tell.

Acknowledgements BD and MK thank the Wardaman People for allowing renewed research on the Earthwatch archives as well as the Australian Research Council Centre of Excellence for Australian Biodiversity and Heritage project CE170100015 for funding, and the Monash Indigenous Studies Centre at Monash University for research support. Madeleine Kelly is an archaeologist currently completing her PhD at the Monash Indigenous Studies Centre at Monash University. She is also part of the Centre of Excellence for Australian Biodiversity and Heritage (CABAH). She is undertaking archaeological and anthropological research on the multivalent spatial and temporal relationships of rock art in Wardaman Country in the Northern Territory. She has also worked on excavations and rock art recording projects in the Kimberley and Victoria. Bruno David is an archaeologist at the Monash Indigenous Studies Centre, Monash University, and Chief Investigator with the Australian Research Council Centre of Excellence for Australian Biodiversity and Heritage. His most recent books include the Oxford Handbook of the Archaeology and Anthropology of Rock Art (Oxford University Press, 2018, co-edited with Ian J. McNiven) and Cave Art (Thames & Hudson, 2017). He is currently undertaking research on the archaeology of deep-time spanning to recent cultural landscapes in close partnership with Indigenous Traditional Owners in East Gippsland and the Kimberley, Australia. Josephine Flood studied Classical Archaeology at Cambridge but then emigrated from England to Australia in 1963 and switched to Australian Aboriginal studies. She completed an MA and then a PhD at the Australian National University, which she published in her book The Moth Hunters: Aboriginal Prehistory of the Australian Alps. She married, had three children, and wrote her book Archaeology of the Dreamtime. From 1979 till 1991 she worked as a deputy director of the Australian Heritage Commission in Canberra. Her work involved putting over two thousand Aboriginal sites on the Register of the National Estate, and she has continued her archaeological fieldwork, most recently with the Wardaman people of the Northern Territory. Three more books followed, including Rock Art of the Dreamtime and most recently, The Original Australians: The Story of the Aboriginal People (2nd edition, 2019). She is a Fellow of the Australian Academy of the Humanities and of the Emeritus Faculty of the Australian National University. In 2019 she was awarded Member of the Order of Australia (AM) for “significant service to archaeology and to the study of Indigenous culture.”

SCENE BUT NOT HEARD

177

Notes 1. In this chapter we do not capitalize or otherwise orthographically distinguish particular Ancestral Beings, such as rainbow serpents, from everyday animals of the environment, because in Wardaman cosmology all animals have numinous, Dreaming cosmogonic manifestations that connect the physical form with the spiritual. We resisted the temptation to capitalize specific Ancestral Beings to avoid deifying or turning them into High Gods, which would not fit well with Wardaman cosmology. 2. From an archaeological perspective, the addition of motifs through time signals that the associated stories would also have changed through time, whether simply by way of the story growing, or by way of fundamental shifts. It is for such reasons that a Wardaman perspective can only come from Wardaman narratives, as there is nothing in the art itself that can tell the contemporary Wardaman story without its storytellers.

References Allam, R. (director) 2014. “Episode 6.” Talking Language with Ernie Dingo. Melbourne: Informit. Arndt, W. 1962. “The Interpretation of the Delamere Lightning Painting and Rock Engravings.” Oceania 32(3): 163–77. Attenbrow, V., B. David, and J. Flood. 1995. “Mennge-Ya and the Origins of Points: New Insights into the Appearance of Points in the Semi-Arid Zone of the Northern Territory.” Archaeology in Oceania 30(3): 105–20. Barrett, C., and R. H. Croll. 1943. The Art of the Australian Aboriginal. Melbourne: The Bread and Cheese Club. Bradley, J. J., and S. Johnson. 2014/15. “We Sing Our Law, Is That still TEK?” Traditional Ecological Knowledge and Can the West Come to Know?” PAN: Philosophy, Activism, Nature 11: 1–22. Brady, L. M., J. J. Bradley, and A. J. Kearney. 2016. “Negotiating Yanyuwa Rock Art: Relational and Affectual Experiences in the Southwest Gulf of Carpentaria, Northern Australia.” Current Anthropology 57(1): 28–52. Chaloupka, G. 1978. Endangered Site: A Famous Example. Canberra: Unpublished report to the Australian Heritage Commission. Clarkson, C. 2007 Lithics in the Land of the Lightning Brother: The Archaeology of Wardaman Country, Northern Territory (Terra Australis 25). Online edition: ANU E Press. Clarkson, C., and B. David. 1995. “The Antiquity of Blades and Points Revisited: Investigating the Emergence of Systematic Blade Production South-West of Arnhem Land, Northern Australia.” The Artefact 18: 22–44. Cundy, B. 1990. “The Analysis of the Ingaladdi Assemblage: A Critique of the Understanding of Lithic Technology.” PhD dissertation. Canberra, Australia: Australian National University. David, B. 2002. Landscapes, Rock-Art, and the Dreaming: An Archaeology of Preunderstanding. London: Leicester University Press. David, B., I. McNiven, J. Flood, and R. Frost. 1990. “Yiwarlarlay 1: Archaeological Excavations at the Lightning Brothers Site, Delamere Station, Northern Territory.” Archaeology in Oceania 25(2): 79–84. David, B., I. McNiven, and J. Flood. 1991. “Archaeological Excavations at Yiwarlarlay I: Site Report.” Memoirs of the Queensland Museum 30(3): 373–80. David, B., I. McNiven, V. Attenbrow, J. Flood, and J. Collins. 1994. “Of Lightning Brothers and White Cockatoos: Dating the Antiquity of Signifying Systems in the Northern Territory, Australia.” Antiquity 68: 241–51. David, B., J. Collins, B. Barker, J. Flood, and R. Gunn. 1995. “Archaeological Research in Wardaman Country, Northern Territory: The Lightning Brothers Project 1990-91 Field Seasons.” Australian Archaeology 41: 1–8. David, B., M. Lecole, H. Lourandos, A. J. Baglioni Jr., and J. Flood. 1999. “Investigating Relationships between Motif Forms, Techniques and Rock Surfaces in North Australian Rock Art.” Australian Archaeology 48: 16–22. Davidson, D. S. 1935. “Archaeological Problems of Northern Australia.” The Journal of the Royal Anthropological Institute of Great Britain and Ireland 65: 145–83. ———. 1936. Aboriginal Australian and Tasmanian Rock Carvings and Paintings. Philadelphia: Memoirs of the American Philosophical Society 5. Davidson, I. 2017. “Images of Animals in Rock Art: Not Just ‘Good to Think.’” In The Oxford Handbook of the Archaeology and Anthropology of Rock Art, ed. B. David and I. J. McNiven. Online edition: Oxford University Press online. Retrieved 29 July 2017 from http://www.oxfordhandbooks.com.ezproxy.lib.monash.edu.au/view/10.1093/oxfordhb/9780190607357.001.0001/ oxfordhb-9780190607357-e-36. Dobrez, L. 2008. “A Rock Art Typology: Narrative and Non-Narrative Figurative Representation.” In Aesthetics and Rock Art III Symposium International Series 1818, ed. T. Heyd and J. Clegg, 29–33. Oxford: Archaeopress. ———. 2011a. “Looking at Our Looking: A New Approach to the Definition of a Rock Art Scene.” American Indian Rock Art 37: 251–67. ———. 2011b. “Rock Art, Perception and the Subject/Object Binary.” Rock Art Research 28(1): 71–83. ———. 2012. “Towards a More Rigorous Definition of Terms: Are There Scenes in European Palaeolithic Art?” In L’Art Pléistocène dans le Monde / Pleistocene Art of the World / Arte Pleistoceno en el Mundo, ed. J. Clottes, 1–15. Tarascon-sur-Ariège: Actes du Congrès IFRAO.

178

MADELEINE KELLY, BRUNO DAVID, AND JOSEPHINE FLOOD

———. 2015. “Making Sense of Pictures.” Bolletino del Centro Camuno di Studi Preistorici 40: 59–75. ———. 2016. “Theoretical Approaches to Rock Art Studies.” Rock Art Research 33(2): 1–24. Dudognon, C., and M. Sepúlveda. 2016. “Rock Art of the Upper Lluta Valley, Northernmost of Chile (South Central Andes): A Visual Approach to Socio-economic Changes between Archaic and Formative periods (6,000–1,500 years BP).” Quaternary International 1(10): 1–10. Film Australia. 1987. Land of the Lightning Brothers. Video. 27 minutes. Sydney. Flood, J. 1997. Rock Art of the Dreamtime: Images of Ancient Australia. Sydney: Angus and Robertson. ———. 2006. “Copying the Dreamtime: Anthropic Marks in Early Aboriginal Australia.” Rock Art Research 23(1): 239–46. Flood, J., and B. David. 1994. “Traditional Systems of Encoding Meaning in Wardaman Rock Art, Northern Territory, Australia.” The Artifact 17: 6–22. Flood, J., D. David, and R. Frost. 1992. “Dreaming into Art: Aboriginal Interpretations of Rock Engravings, Yingalarri, Northern Territory (Australia).” In Rock Art and Ethnography, ed. M. J. Morwood and D. R. Hobbs, 33–38. Melbourne: Australian Rock Art Research Association. Fritz, C., T. Lenssen-Erz, G. Sauvet, M. Barbaza, E. Lopez-Montalvo, G. Tosello, and M. Azema. 2013. “L’Expression Narrative dans les Arts Rupestres: Approches Théoriques.” Les Dossiers d’Archéologie 358: 38–45. Gell, A. 1998. Art and Agency: An Anthropological Theory. Oxford: Oxford University Press. Harney, W. E. 1943. Taboo. Sydney: Australasian Publishing Co. ———. 1959. Tales from the Aborigines. London: Australasian Publishing Co. Kelly, M. 2017. “A Piece of the Action: Locating Scenes in Rock Art research.” BA (Honours) Thesis. Clayton: Monash University. Lenssen-Erz, T. 1989. “The Conceptual Framework for the Analysis of the Brandberg Rock Art Paintings.” In The Rock Paintings of the Upper Brandberg. Part I: Amis Gorge, ed. H. Pager, 361–70. Cologne: Heinrich Barth Institute. ———. 1992. “Coherence: A Constituent of ‘Scenes’ in Rock Art.” Rock Art Research 9(2): 87–105. Lewis, D., and D. McCausland. 1987. “Engraved Human Figures and Faces from Wardaman Country, Eastern Victoria River District, Northern Territory.” Australian Aboriginal Studies 1: 67–79. Lewis, D., and D. B. Rose. 1988. The Shape of the Dreaming: The Cultural Significance of Victoria River Rock Art (Institute report series). Canberra: Aboriginal Studies Press for the Australian Institute of Aboriginal Studies. May, S. K., and I. S. Domingo. 2010 “Making Sense of Scenes.” Rock Art Research 27(1): 35–42. McNiven, I., B. David, and J. Flood. 1992. “Delamere 3: Further Excavations at Yiwarlarlay (Lightning Brothers Site), Northern Territory.” Australian Aboriginal Studies 1: 67–73. Merlan, F. 1978. “Report on Wardaman Request for Land.” Darwin: Unpublished report to the Northern Land Council. ———. 1989. “The Interpretive Framework of Wardaman Rock Art: A Preliminary Report.” Australian Aboriginal Studies 2: 14–24. ———. 1994a. A Grammar of Wardaman: A Language of the Northern Territory of Australia. Berlin: Mouton de Gruyter. ———. 1994b. “Narratives of Survival in the Post-Colonial North.” Oceania 65(2): 151–74. ———. 1998. Caging the Rainbow: Places, Politics, and Aborigines in a North Australian Town. Honolulu: University of Hawaii Press. Mulvaney, J. 1975. The Prehistory of Australia. Harmondsworth: Penguin Books. Ouzman, S. 2001. “Seeing Is Deceiving: Rock Art and the Non-visual.” World Archaeology 33(2): 237–56. Oxford English Dictionary. 2015. 3rd ed. Online edition: Oxford University Press. Accessed 30 June 2017 from http://www.oed .com.ezproxy.lib.monash.edu.au/view/Entry/172219?rskey=JERxqL&result=1&isAdvanced=false#eid. Rose, D. B. 1992. Dingo Makes Us Human: Life and Land in an Aboriginal Australian Culture. Melbourne: Cambridge University Press. ———. 2004. Reports from a Wild Country: Ethics for Decolonisation. Sydney: UNSW Press. Spencer, B. 1914. Native Tribes of the Northern Territory of Australia. Digitally printed 2010: University of Cambridge Press. Vázquez, R. 2011. “Translation as Erasure: Thoughts on Modernity’s Epistemic Violence.” Journal of Historical Sociology 24(1): 27–44. Watchman, A., B. David, I. McNiven, and J. Flood. 2000. “Micro-archaeology of Engraved and Painted Rock Surface Crusts at Yiwarlarlay (the Lightning Brothers Site), Northern Territory, Australia.” Journal of Archaeological Science 27: 315–25.

==== •

•

•

12. A COMPARISON OF “SCENES” IN PARIETAL AND NON-PARIETAL UPPER PALEOLITHIC IMAGERY Formal Differences and Ontological Implications Elisabeth Culley

Introduction The Eurasian Upper Paleolithic dates from approximately 45,000 to 10,000 years ago and is marked by an “explosion” in technological diversity and creative expression. Hominins began exploiting an unprecedented array of raw materials for the production of tools, jewelry (Douka et al. 2019), musical instruments (Conard, Malina, and Münzel 2009), and an extensive body of art. Images were painted, engraved, and carved in relief on rock shelters and cave walls, and many portable objects were also elaborately decorated or carved in the round (Clottes 2008; White 2003). Upper Paleolithic art of Europe includes some of the earliest known depictive imagery (Quiles et al. 2016), only slightly younger than found in Indonesia (Aubert et al. 2014; Aubert et al. 2018; Aubert et al. 2019), and is renowned for the use of foreshortening, shading, and other techniques that were not seen again until the Renaissance (Clottes 2008; White 2003: 229). Realistic representations of animals decorating rock walls and small objects at times even captured the age, sex, and seasonal attributes of Pleistocene fauna (Clottes 2008; Clottes, Garner, and Maury 1994; Guthrie 1984; White 2003) that would have populated the Pleistocene landscape. Yet the artists who were responsible for the images produced few scenes of animals engaged in specific actions, or of any event or happening. Such scenes appear largely restricted to portable art. Indeed, the lack of narrative depiction on shelters and cave walls is widely reported (Clottes 2001: 464; Delporte 1984; Halverson 1992; Levine 1957; White 2003: 100–101, Figure 69), but uncritically so. Researchers typically fail to specify what constitutes a scene in prehistoric art or how narrative depiction can be identified in Upper Paleolithic imagery (Dobrez 2011, 2012; Lenssen-Erz 1992: 89). A notable exception is the work of Marc Azéma, discussed below. Livio Dobrez (2011, 2012, 2015) has defined scenes as a universally hardwired perceptual category that constrains both the perception and depiction of visual experience. Dobrez further argues that scenes can be objectively identified in artistic representation by the presence of categorical markers. This chapter adopts Dobrez’s approach for a more formal analysis of the frequency and nature of narrative depiction in caves, on rock shelter walls, and on portable objects during the Upper Paleolithic. The study is not designed to be conclusive but to provide a forum for discussing both the limitations and advantages of defining scenes as a perceptual category for rock art research and to isolate possible trends in the distribution of scenes during the Upper Paleolithic for future analyses.

Scenes as a Perceptual Category For some psychologists, perceptual categories are “hardwired” mental representations comprising attributes that are fundamental to and shared by sets of similar phenomena and through which novel experiences are organized (Lakoff and Johnson 1999: 16–28; Mervis and Rosch 1981; Rosch and Lloyd 1978; Rosch et al.1976).1 Humans, non human primates, and many other animals rely on perceptual categories to quickly and accurately classify sensory experience at the level at which it is critical to distinguish between phenom-

180

ELISABETH CULLEY

ena in the natural environment, including among plants, animals, conspecifics and allospecifics, colors, and smells (Lakoff and Johnson 1999: 16–28; Mervis and Rosch 1981; Rosch and Lloyd 1978; Rosch et al.1976; Zentall et al. 2008). For example, mental representations of different kinds of animals are how the brain can immediately identify dogs as dogs—and distinguish dogs from goats, sheep, and cows—despite significant variation within each class of animal. The neurological structures effect perception of category membership in a gestalt or gestalt-like recognition of minimally salient attributes, or categorical markers (Dobrez and Dobrez 2013a: 117; Lakoff and Johnson 1999: 16–28; Mervis and Rosch 1981; Rosch and Lloyd 1978; Rosch et al. 1976). The basic principles of categorical perception have been defined through linguistic analyses, extensive experimental research, and now brain imaging studies (see Lakoff 1987; Lakoff and Johnson 1999; Mervis and Rosch 1981; and Zentall et al. 2008 for detailed reviews). Still, it is unclear how lived experience shapes the structure and activity of categorically responsive neurons (Prather et al. 2009). Our neurophysiology changes throughout childhood and adolescence, with structures undergoing both expansion and pruning that can affect hardwired representations and schema (Cantlon et al. 2011; Iglesias et al. 2005). Categories are also easily learned and can be population-specific (Goldstone 1994; Kay, Berlin, and Merrifield 1995; Prather et al. 2009; Roberson, Davies, and Davidoff 2000; Sloutsky 2010). However, perceptual categories are frequently based on the overall shape of class members and on the motor actions needed to engage with them (Lakoff and Johnson 1999: 16–28; Mervis and Rosch 1981: 92; Rosch et al.1976). The result is significant convergence among studied cultures and the expectation that many categories are common to all humans, due to shared physiology and sensory relationships with the natural world (Kay, Berlin, and Merrifield 1995; Lakoff and Johnson 1999: 16–28; Mervis and Rosch 1981; Rosch 1975, 1977, 1978). Culley (2006, 2008) found evidence of universal percepts and concepts in prehistoric populations by examining evidence for the ritualized use of metaphor in petroglyphs associated with Numic speakers in Nevada. Dobrez (2011, 2012, 2015: 36) argues that recurrent perceptual situations where survival hinges on quick and accurate perception would have selected for automated perceptual outcomes, and he specifically defines an “evolutionary imperative” as the distinction in categorical perception between activities that are directed toward us and those that are not. For example, it is critical for an observer to determine if an animal is focused on and coming toward him or her or if it is otherwise engaged and poses no threat. It is no less crucial to determine if a vehicle is approaching head-on or simply driving by. Dobrez (2011, 2012, 2013, 2015) draws on research from a number of disciplines to identify PERFORMATIVE EVENTS and NARRATIVE EVENTS as two perceptual categories with distinct defining attributes as the mechanisms through which we distinguish between threatening and non-threatening activities. PERFORMATIVE EVENTS (Dobrez 2013, 2015: 41–44) effect recognition of displays of aggression and other actions or movement that are directed away from event participant(s) and toward the perceiving agent(s). Dobrez (2013: 408–11 and 429–37, 2015: 41–44) has described performative action as “looming”; it demands attention and engages viewers by entering into the observational space. Performative action comes at you and creates a sense of immediacy. In sharp contrast, NARRATIVE EVENTS (scenes) (Dobrez 2011, 2012, 2013, 2015: 38–40) effect the recognition of happenings in which one or more actors is doing something that does not involve perceiving agents. Actors are oriented away from observers and engaged in an activity, with movement directed across the perceiving agents’ field of vision. Narrative action unfolds in front of us, as if in a different space and time, and demands explanation. It prompts us to consider what participants are doing, what precipitated the event, and what happened in the moments that followed. Scenes effect “imaginational space” (Dobrez 2012: CD1839–44). As perceptual categories, PERFORMATIVE and NARRATIVE EVENTS are not only mechanisms through which we perceive, recall, and imagine that something is happening, but also mechanisms through which we re-present those happenings to others in recognizable ways. As such, we can expect categorical perception to constrain the representation of performative and narrative action in even the earliest arts (Dobrez 2013: 411). Moreover, the categorical attributes necessary for the recognition of events composed of universal perceptual forms (e.g., frontal and profile views of event participants), instances of performative and narrative depiction can be identified over and above the actual contents of each image (e.g., children playing)

A COMPARISON OF “SCENES” IN PARIETAL AND NON-PARIETAL UPPER PALEOLITHIC IMAGERY

181

that is likely constrained by culturally specific values, forms, and stylistic conventions. That is to say, the fact of an event can be recognized by any observer, whether or not the details of the event itself are discernable (Dobrez 2013: 411–12). Dobrez provides a significant amount of evidence for both PERFORMATIVE and NARRATIVE EVENTS, as well as for the specific attributes that prompt the recognition of category members and enable their identification in artistic representation. He also discusses categorical attributes as they appear in imagery from around the world and the Upper Paleolithic, further clarifying expectations for performative and narrative depiction. Though a comprehensive review of his work is beyond the scope of this chapter, the attributes that are relevant to the current study of scenes in Upper Paleolithic portable and parietal imagery are summarized in the following section. Attention is given to the classification of ambiguous imagery, to classification decisions that vary from Dobrez’s, and more general methodological problems with using perceptual constants to identify narrative depiction.

Categorical Attributes of Narrative Depiction Categorical attributes are the formal traits that phenomena must exhibit to be perceived as a scene. If, at the most fundamental level, a NARRATIVE EVENT is “one or more actors doing something independent of observers,” then, at a minimum, the formal traits of narrative depiction must trigger the perception of an actively engaged agent, delimited from observational space. Recognition of formal traits depends, in turn, on their iconic representation. The attributes, or visual markers, that are used to identify scenes in Upper Paleolithic art are reviewed here in terms of the perceptual entailments of NARRATIVE EVENTS that each prompts: iconicity, agency, engagement, and coherence (Table 12.1).

Iconicity

TABLE 12.1. Perceptual entailments and visual markers of narrative events. Perceptual Entailment

Visual Markers

Iconicity

recognizable agents recognizable actions

Agency

one or more agents

Action

diagonal vectors contracted/extended space small elements visual echoes

Engagement

profile views facing agents directional gazes linking vectors minimal spacing between agents

Coherence

framing devices shared orientation, color, size, etc. of elements profile views facing agents directional gazes linking vectors minimal spacing between agents

Icons, indexes, and symbols are different types of representations (or signs) that convey meaning by respectively looking like, having a causal or proximal relationship to, or having a culturally determined relationship with the ideas they represent. An iconic representation of a tree thus brings to mind the idea of a tree by virtue of looking like a tree. “Iconicity” speaks to the degree of similarity a representation has to its referent, or its recognizability (Culley 2016: 7–32; Nöth 1990: 121–24; Peirce [1931] 1960). Detailing in iconic images can increase iconicity. For example, a portrait of George Washington is not just recognizable due to the claimed similarity to Washington’s actual facial features, but also to his hairstyle and clothing. Indeed, the recognizability of Washington as general or president is situated in the details of his location, the presence or absence of a military uniform, gray hair, and facial lines that indicate his age and social role. The increased iconicity effectively demarcates, not just a specific individual, but a specific individual in a specific place and time. Meanwhile, detailing of iconic images can also decrease similarity. A portrait that depicts a young General Washington through the representation of his uniform, for example, will look very similar to Washington when he was a young man, but less similar to the elder president. Icons, then,

182

ELISABETH CULLEY

are more effective at conveying their intended meaning when detailed and, consequently, are often highly variable or idiosyncratic. Symbolic reference is quite different. Although symbols may have iconic and indexical grounds, reference extends beyond similarities and causal relationships with other phenomena to more abstract ideas that have been determined by a social group (Culley 2016: 7–32; Deacon 2012; Peirce 1960[1931]). Symbols are often minimal in form and necessarily standardized in order to direct the mind away from historically situated details and toward negotiated cultural meanings (Culley 2016: 7–32; Deacon 2012; Peirce 1960). The United States presidential seal is a symbol that builds on yet extends beyond an iconic representation of an eagle to simultaneously reference the bird, notions of strength and freedom, the executive office, and the country at large. NARRATIVE EVENTS are, by definition, iconic representations of category members’ minimally salient, shared features, and narrative depictions are necessarily similar to each other, relative to those categorical attributes. Moreover, for an image to be perceived as a scene—as an agent doing something—agents and actions must be recognizable as agents and actions. They must be iconic representations of real world or mythic beings and activities; a depiction of horses running must have agents that look like horses in an active posture that looks like running. The perceptual demand for scenes to be iconic does not, however, preclude the integration of non-iconic elements into narrative imagery. In Upper Paleolithic art, many compositions with attributes that constitute NARRATIVE EVENTS include geometric signs superimposed over, or near to, event actors. The perceptual demand for iconicity in narrative depiction predicts some detailing of agents and action as a mechanism that increases similarity. We can expect with a scene of horses running, not just an outline of horses, but brown, black, or spotted horses running uphill, downhill, away from, or toward something and with the consequence of delimiting specificity of agency, time, and space. Detailing, in turn, predicts variability among scenes, even of the same subject matter. It must be emphasized that Dobrez has not identified detailing as a categorical attribute that prompts the gestalt recognition of scenes. In fact, the extent to which iconic images are embellished may be constrained by cultural preference and so cannot provide an objective visual marker of narrative depiction. Nevertheless, as a frequent consequence of iconicity and effective strategy for delimiting events in space—which are perceptual entailments of NARRATIVE EVENTS— both detailing and idiosyncrasy can be used cautiously to help identify scenes in artistic representation. It is also important to note that while detailing increases the likelihood a scene will be idiosyncratic, there is nothing that precludes the standardization of narrative depictions and the attachment of symbolic values to that standardized form. For example, any given painting of Washington crossing the Delaware River can become standardized and symbolically reference the revolutionary spirit that founded the United States, over and above the iconically referenced event. The implication that scenes and symbols are not mutually exclusive departs from Dobrez’s position. He (Dobrez 2011: 260) states that symbolic representations contrast with scenes because the latter unambiguously manifest a range of categorical attributes. He further describes some paintings that have ambiguous narrative attributes and suspected symbolic meaning as possibly transitioning into symbols, leaving little room for iconic and symbolic reference to co-occur. Others (Bahn and Vertut 1997: 140; Delporte 1984: 112–14) have also suggested narrative and symbolic depiction are incompatible. Dobrez (2012: CD1848–9, Figure 12) specifically discusses the “shaft scene” in Lascaux Cave that depicts a man with a bird-like head facing a charging and disemboweled bison with a spear through its side (Figure 12.1). The anthropomorph appears stiff, perhaps dead, and is falling backwards (Davenport and Jochim 1988). The panel is often cited as one of the rare examples of scenes in Upper Paleolithic cave art (e.g., Clottes 2008: 120–21; White 2003: 100–101, Figure 69), and the imagery does exhibit categorical attributes. However, Dobrez considers the visual markers too weak to prompt the recognition of a scene. He argues the panel is not seen as a NARRATIVE EVENT through the gestalt perception of categorical attributes, but inferred from symbolic content. Put another way, the intimation of a scene is not perceived but conceived through higher order cognitive processes. For Dobrez, the shaft scene is a symbolic representation of an event, not an iconic category member, and he cites notable similarity with other cave images as further indication of symbolic reference. The argument is unconvincing. The images in question are actually quite variable and, as idiosyncratic compositions, cannot support symbolic reference (Culley 2016: 7–32). Indi-

A COMPARISON OF “SCENES” IN PARIETAL AND NON-PARIETAL UPPER PALEOLITHIC IMAGERY

183

FIGURE 12.1. The “shaft scene” in Lascaux Cave that depicts a man with a bird-like head facing a charging and disemboweled bison with a spear through its side. Lascaux 4, Motignac, Dordogne, France. © Traumrune, Wikimedia Commons, CC BY-SA 3.0.

vidual elements that do appear standardized, such as the man’s bird-like face and a second spear below him (White 2003: 100–101, Figure 69), could have functioned as symbols. In this study, the question of whether an image can be perceived as a scene due to the presence of categorical attributes is treated as separate from whether those same or different traits can support symbolic reference. If an iconic image exhibits a range of categorical attributes for narrative events, it has been classified as a scene. The actual salience of the fact of an event that is defined by those visual markers and/or any symbolic meanings associated with it cannot be determined. Consequently, appeals to suspected symbolic value in the determination that an image is or is not a narrative depiction undermines the objectivity of Dobrez’s approach. In this analysis the “shaft scene” at Lascaux is classified as a scene. Dobrez’s appeal to suspected symbolic value in the determination that an image is or is not a narrative depiction introduces a degree of subjectivity that is contrary to the stated goals of his approach. The actual salience of the fact of an event as defined by visual markers and/or any symbolic meaning that was associated with it cannot be known without recourse to ethnographic information. Dobrez’s approach ultimately restricts us to a determination of whether or not an image could have been perceived as a member of the perceptual category of NARRATIVE EVENTS. In this study, any iconic image that exhibits a range of categorical attributes sufficient to prompt the minimum perceptual entailments of NARRATIVE EVENTS has been classified as a scene, including the shaft scene at Lascaux.

Agency Dobrez (2011: 253; 2012: CD1839, 2013: 413–15) argues that the perceptual category NARRATIVE EVENTS effects the recognition of one or more agents engaged in an activity (emphasis mine), and most of

184

ELISABETH CULLEY

the scenes that he identifies in the decorated caves have only one event participant (Dobrez 2012). Other researchers (Bahn 1998: 193–95; Dobrez 2013: 413; Lenssen-Erz 1992) implicitly or explicitly state that scenes must have at least two agents, despite countless situations in which an individual’s actions clearly constitute a happening (e.g., a bear fishing, a person dancing). Certainly, there are paintings in Western art that have only one agent and would be considered scenes under any model, such as Edward Hopper’s Girl at Sewing Machine and John William Waterhouse’s The Soul of the Rose.2 The present study adopts Dobrez’s position and counts among scenes images with only one agent that meet the criteria for narrative depiction. A likely outcome of defining images with just one agent as scenes will be a higher frequency of scenes in Upper Paleolithic art than typically recognized. There will also be potential problems with comparisons between such analyses and those using a more restricted understanding of narrative events.

Action The most fundamental perceptual entailment of narrative events—of seeing any happening—is seeing action, or motion; yet discerning activity in Upper Paleolithic art can be difficult. Indeed, Bahn (Bahn and Vertut 1997: 139) suggests the inability to tell what figures are doing—if they are doing—poses the biggest obstacle to the identification of scenes. Animals were frequently represented with animated postures, such as with splayed legs and raised heads, and can seem spirited and even powerful (Bahn and Vertut 1997: 139), but the figures nevertheless seem stiff and motionless, with no signs of weight bearing in the arms or legs. Despite markedly well-executed and realistic representations of animal anatomy and detailing, the figures are not naturalistic (Bahn and Vertut 1997:139–40; Dobrez 2012: CD1840–4). Some researchers have argued the animals are not just motionless, but dead, and were copied from corpses (Leason 1939). Most, however, are consistent with “canonical forms,” a perceptual category that effects the recognition of animals’ essential forms and qualities only (e.g., “horseness”) (Dobrez 2015: 37; Dobrez and Dobrez 2013a, 2013b, 2014; see also Halverson 1992). The appearance of action can be especially ambiguous in panels with multiple animated figures. It is often unclear whether adjacent figures just happened to be placed next to each other, or are meaningfully associated, with the removal of one qualitatively impacting the other (Dobrez 2012: CD1839; see also Bahn and Vertut 1997:140). In many instances of a recurring motif that shows two animals facing each other, the distance between figures appears to be the only perceptual clue as to whether they are actively engaged or not (Dobrez 2012: CD1839). Animals were also depicted in large and animated groups, including the famous herd-like clusters of aurochs, rhinoceros, horses, and lions seeming to emerge from, or enter into, the recesses of Chauvet Cave and the linear frieze of both real and mythical animals at Lascaux. Lines of a single species of animals were frequently engraved on portable objects, including bone and antler lissoirs and pierced batons, especially near the end of the Upper Paleolithic. In many, if not most, of the compositions with multiple animals, the figures still appear stiff and more canonical in form than in motion. Any sense of a happening appears situated in the dynamic nature of the compositions and not in the action of individual agents (Dobrez 2012: CD1840–3). In some cases, adjacent or clustered animals were created at different times and could not have been perceived as engaged in an activity or with each other. Similarly, an analysis (White 2003: 79, Figure 38) of a group of horses in Chauvet has shown their animated postures and facial expressions represent calm, aggressive, relaxed, and alert attitudes that would not occur in response to the same stimulus. The grouping more likely represents different aspects of a single horse, or of horseness, than an active herd. Many of the linear groupings are found on restricted linear supports, including at Lascaux where the only area suitable for painting in the Hall of the Bulls is linear, and on tools, where functional requirements limit the shape and size of supports. These groupings may be more a matter of practical constraint than meaningful interaction. Instead of relying on subjective assessments of whether figures are or are not actively engaged, Dobrez (2012: CD1840–4) 2013, 2015: 39) has used experimental work on how humans perceive motion and interaction to develop a list of visual markers that effect their perception in artistic representation (see also Nowell 2015). This brief summary is taken from Dobrez’s (2013) article on motion perception and its relationship to categorical perception and narrative events specifically, unless

A COMPARISON OF “SCENES” IN PARIETAL AND NON-PARIETAL UPPER PALEOLITHIC IMAGERY

185

otherwise stated. The markers are also discussed relative to the entailment of iconicity that is emphasized in this analysis. Diagonal vectors, especially as manifest in agents’ angled or extended limbs, effect recognition of action through their iconic relationship with bodies in motion and the transference of weight to generate force and momentum forward. The diagonal orientation of objects similarly signals the momentum and trajectory of movement through space. Contracted space in front of an agent, in tandem with extended space behind, effects the perception of motion through iconic representation of the consequences of movement forward: the reduction of space between agents and their end goal and the increase in space between agents and their point of departure. Smallness in size can increase the perception that agents and objects are moving and some distance away. Dobrez (2012: CD1844) gives some attention to the “multiplication of features,” a technique used in Western art to indicate motion, and cites Giacomo Balla’s painting, Dynamism of a Dog on a Leash (1912), as a prime example. Balla depicts a woman walking a dachshund with the woman’s legs, the leash, and the dog’s legs and tail represented multiple times to capture the successive phases of their rapid movements. The painting reflects Balla’s interest in chronophotographic studies of animals in motion that were produced at the end of the nineteenth century and are credited with introducing European artists to multiplication, blurring, and other techniques for representing movement (Salazar-Sutil and Melo 2016). In fact, multiplied features, or “visual echoes,” are relatively common in Upper Paleolithic art and are variously interpreted as representing a single animal in motion (Azéma 1992, 2005, 2006, 2008, 2011; Azéma and Rivière 2012) or multiple animals standing side by side. Dobrez (2012: CD1844) cautions that the multiplication of features to depict motion is specific to Western cultures and unlikely to have had salience during the Ice Age. He argues the visual markers cue the perception of animal herds. However, different phases of motion are visible to the human eye as repeating outlines of a moving object when motion is rapid, but not rapid enough for the phases to blur. For example, if an old-fashioned, coil-spring doorstop is bent sideways and then released, multiple outlines of its basic shape can be seen as the spring moves back and forth due to retinal persistence. The outlines overlap at their point of origin where the base of the spring is attached to the wall, with the extent of differentiation at the tips decreasing as momentum slows. The visual echoes are a function of the human perceptual system and, as such, cannot be dismissed as an indicator of motion in even the earliest arts (see Nowell 2015). A quick review of Upper Paleolithic imagery suggests two different types of multiplied features can be distinguished: visual echoes that have a common point of origin with the foregrounded agent, as with Balla’s Dog on a Leash, and those that do not. An example of the second case is the rhinoceros on the Lion Panel in Chauvet Cave (Figure 12.2), whose multiple horns are immediately adjacent to one another but do not converge at a single point. The implication is that the different types of visual echoes constitute distinct phenomena: iconic representations of multiple phases of motion and iconic representations of multiple actors, or herds. Determinations regarding the presence of motion in the images that were evaluated in the current study could be made without using visual echoes as an indicator, but further analysis of when and how humans see individual phases and of any patterning in feature multiplication in the Upper Paleolithic is warranted. At a minimum, researchers should be explicit re- FIGURE 12.2. The rhinoceros on the Lion Panel in Chauvet garding their identification of scenes from Cave with multiple horns that are immediately adjacent to one the use of visual echoes and the potential another but do not converge at a single point. Ptilpv25, Wikimedia Commons, CC BY-SA 4.0. impact of that practice on their results.

186

ELISABETH CULLEY

Engagement The action that characterizes scenes tends to be described as “interaction.” However, the term refers to reciprocal action and mutual influence, with the implicit requirement of two or more agents (but see Dobrez 2013: 413–15 on the “interaction fallacy” in rock art research). Narrative events include individuals engaged in activities by themselves, alongside one other, or with each other. “Engaged action” is used here to better accommodate individual agency and to emphasize the directionality of attention and motion that distinguishes narrative events from other kinds of happenings. One of the strongest visual indicators of engagement is the physical orientation of agents toward the goal of their actions and/or other event participants (Dobrez 2011: 258, 2012 CD1840). Engagement is consequently recognized through the profile view and gazes of agents, particularly when directed toward each other. The orientation of gazes, active limbs, and linear elements in general also creates linking vectors that visually connect actors and actions and prompt the recognition of engagement (Dobrez 2011, 2012: CD1840, 2015: 39). As previously suggested, recognition is also prompted by the close proximity of actors (Dobrez 2012: CD1840). The physical orientation required of event participants in the perception of engaged action directs their attention and actions across the observational field and excludes viewers from participation. Narrative action is thus mutually exclusive with performative action, as performative action moves from representational to observational space and is recognized through the frontal view of event participants (Dobrez 2011: 258, 2015: 39; see also Jewitt and Oyama 2001).

Coherence The visual markers and exclusivity of engagement also create and engender the perception of coherence among actors, actions, and other event elements. Coherence is further perceived through the similarity of elements, such as shared orientation, coloring, size, etc., as well as through framing devices, including isolation from unrelated imagery. Ultimately, narrative depictions are recognizable as coherent compositions of engaged activity and delimited imaginational space that provokes a story.

Identifying Scenes in Upper Paleolithic Art Researchers face a number of challenges when working with the Upper Paleolithic archaeological record, regardless of their analytical goals. A large percentage of sites were excavated prior to the advent of modern scientific recovery and dating methods, resulting in little to no chronological control over deep deposits that likely represent significant time depth and an array of behavioral changes. In other cases where the stratigraphic sequence and artifact provenience were appropriately documented, direct chronometric dates are few. There may be only one or two dates for sequences that clearly span tens of thousands of years. Instead of being directly dated, all of the artifacts from an entire site—or, when differentiated, from an individual stratigraphic layer—are assigned to one of four broad chrono-industrial time periods that are primarily based on the stylistic affiliations of stone tools: the Aurignacian (~40,000–30,000 BP), Gravettian (~33,000–20,000 BP), Solutrean (~22,000–17,000 BP), and Magdalenian (17,000–11,000 BP). These chrono-industrial periods are nevertheless problematic in their own right because some classifications are based on older excavations that had little control over site stratigraphy, with the integrity of date ranges, claimed stylistic differences, and their behavioral implications all suspect (Clark and Riel-Salvatore 2009). As a result, it is not unusual for Upper Paleolithic deposits, including those with decorated objects, to have uncertain, contested, revised, and/or no date assignment (see discussion in von Petzinger and Nowell 2011 for the impact of stylistic dating on our understanding of the development of European parietal art). The occurrence of parietal imagery at rock shelters that were primarily used as occupation sites has allowed researchers to establish correlations between stylistic attributes of the imagery and the major chrono-industrial periods (see the critique by von Petzinger and Nowell 2011). Although there is typically no definitive link between the time when rock shelters were occupied and when the art was made, some scholars claim that repeated co-occurrences between representational and lithic forms affords some confidence in

A COMPARISON OF “SCENES” IN PARIETAL AND NON-PARIETAL UPPER PALEOLITHIC IMAGERY

187

chronologies established through stylistic affiliations (Bahn and Vertut 1997: 62–63). In contrast, it is also worth noting that Clottes’s original dating of Chauvet was based on the stylistic chronology and was incorrect (in Chauvet, Deschamps, and Hillaire 1995) in assigning it to a period after the cave entrance had collapsed and the site could not be used (Clottes et al. 1995; Quiles et al. 2016). To Clottes’s credit, he accepted the dates when they came in. It remains the case that similarities between the content and style of portable art recovered from relatively well-dated contexts and imagery on both rock shelter and cave walls have been instrumental in dating the parietal art (Bahn 2016; Keyser 2001: 119). Fragments from the walls and ceilings of decorated rock shelters have also fallen into occupation layers, giving a minimum age for the artistic representations (Bahn and Vertut 1997: 61; Clottes 2001: 468; Keyser 2001: 118; Ucko and Rosenfeld 1967: 66–69). In some fortuitous cases such as at Chauvet Cave, the collapse of geological formations has sealed cave entrances and so provided the last possible date that Upper Paleolithic people could have entered (Bahn and Vertut 1997: 61; Clottes 2001: 468; Keyser 2001: 127). Direct dating of cave sites has steadily increased with the advent of new technologies in the 1990s, but the (typically) 14C dates are not without problems (Bahn and Vertut 1997: 76; Clottes 2001: 469–70; Rowe 2001: 139–44).3 The assessments are often of charcoal pieces or torch swipes found near the images and not of the art itself, and where charcoal pigments are dated directly, the source material being dated may be much older than the paintings made from it (Bahn and Vertut 1997: 76; Rowe 2001: 144). Where images have been dated directly, they are only one or two among hundreds that we know may reflect multiple episodes of image production over extended periods of time (Clottes 1993, 2001: 470; Pike et al. 2012; Rowe 2001: 148). There are also a number of discrepancies between chronometric and stylistic date assignments (Bednarik 1995; Pike et al. 2012; Valladas et al. 2001; von Petzinger and Nowell 2011, 2014: 51; White 2003: 78–79). Nevertheless, the dates have been broadly consistent (Bahn and Vertut 1997: 74; Clottes 1993, 2001: 470) and indicate that ongoing efforts will provide a chronology that is both more accurate and more precise. Given the current limitations on placing Upper Paleolithic art in time, researchers must acknowledge that we can only access gross trends in image production that loosely correspond to chrono-industrial periods as now defined and only determine cultural salience through rigorous testing of possible explanations for those trends. Identifying the distribution of scenes and any behaviorally meaningful patterning therein requires not only an objective definition and visual markers of narrative depiction, but also an “unpacking” of Upper Paleolithic art as a monolithic entity. One productive direction of research would be to shift attention from scenes as manifest in the record at large to the frequency of narrative depiction on portable objects, rock shelter walls, and cave walls with uncontroverted and relatively secure date assignments. Using such an approach, a large number of assemblages of imagery from all three contexts that are reasonably dated to each chrono-industrial period can be evaluated for statistically viable analyses of trends in narrative depiction and how those trends changed through time. Over and above chronological concerns, the accessibility of imagery is a problem for this approach. Ideally researchers evaluate the entire corpus of imagery from the archaeological contexts in question, not just the dramatic portable objects and friezes that are widely published. Where direct access to decorated artifacts and sites is not feasible, photographs, scans, and drawings must be of sufficient quality to identify iconic elements and compositions. In reality, the greatest challenge to understanding narrative depiction during the Pleistocene is developing appropriate data sets. This study is a preliminary analysis and is necessarily more limited than needed. Appropriate data were located in published documents and internet resources for 49 different image collections that can be reasonably attributed to distinct chrono-industrial peTABLE 12.2. Number of assemblages from portable and riods. These assemblages include the images on parietal contexts evaluated in study. decorated objects from 17 subsurface archaeoPortable Shelter Cave Total No. logical deposits, on the walls of 11 rock shelters Time Objects Walls Walls Assemblages that were used as habitation sites, and on the Period walls of 21 caves. Thirty-two sites in 11 regions Magdalenian 10 6 13 29 of France are represented. Table 12.2 presents Solutrean 1 1 2 4 the distribution of assemblages relative to asGravettian 2 0 5 7 semblage type and assigned date range, with 4 4 1 9 the under-representation of all but Magdale- Aurignacian

188

ELISABETH CULLEY

nian imagery immediately clear. Time vectored comparisons of narrative depiction are consequently best limited to only two contexts: pre-Magdalenian and Magdalenian. Table 12.3 shows the number of scenes that were identified in portable and parietal contexts and the mean number of scenes identified in each type of assemblage. Despite the small sample sizes, scenes were evident in even the earliest decorated caves and with some freTABLE 12.3. Number of scenes identified in portable quency. The complete absence of narrative depiction and parietal contexts with mean number of scenes on portable objects may be due to sampling, but is conper image assemblage (x¯). sistent with the lower numbers of decorated objects Portable Shelter Cave (Culley 2016), and specifically with iconic depictions, Time Period Objects Walls Walls that characterize earlier periods. Indeed, the distributions suggest that the representation of narrative Magdalenian 37 (3.7) 4 (.67) 18 (1.38) events on portable objects is predominately a MagdaleSolutrean 0 3 (3) 5 (2.5) nian phenomenon that only appears more widespread Gravettian 0 — 0 when there is a lack of chronological control. Aurignacian 0 0 7 (7.0) Despite the hint of patterning, the low and uneven distribution of evaluated assemblages in earlier periods warrants limiting time-vectored analyses to pre-Magdalenian and Magdalenian comparisons, as presented in Table 12.4. Here, the prevalence of narrative depictions in all parietal contexts is clearer, as is notable consistency in the average number of scenes on shelter and cave walls through time. Scenes appear to be at least twice as common in the decorated caves and certainly are not rare, as often claimed. The high TABLE 12.4. Number of scenes identified on all frequency is somewhat unexpected; however, if the pre-Magdalenian and Magdalenian portable and number of identified scenes were considered relative parietal contexts with mean number of scenes per to the number of images in all cave assemblages in image assemblage (x¯). a given sample—a measure not captured here—the Portable Shelter Cave apparent abundance would be significantly lower. A Time Period Objects Walls Walls similar decrease in narrative depictions on portable Magdalenian 37 (3.7) 4 (.67) 18 (1.38) objects and shelter walls would not apply given that most portable objects have only one or two images Pre0 3 (.6) 12 (1.5) or compositions and that rock shelters typically have Magdalenian far fewer images than caves. The general subject matter of each scene was also recorded in order to identify potential relationships between content and context of narrative depiction. More specifically, event participants were described as “animals,” “humans,” or “ritualized humans.” Agents in scenes depicting one or more animals independently or mutually engaged in an activity were classified as “animals.” In cases where participants include both animals and humans, classification was based on the human agent or agents. Male and female humans were found in two classes: those that were ritualized and those that were not. Those that do not appear ritualized in any way were classified simply as “humans.” Ritualized humans include agents with a mix of human and animal features, such as the “bird man” in the shaft scene at Lascaux (Figure 12.1) and the “sorcerers” at Les Trois-Frères (Figure 12.3) that are consistent with the ritual use of masks and transformational capacities assigned to ritual practitioners in many small-scale cultures (as described, for example, by Clottes and Lewis-Williams 1998). The “wounded man” figures with a series of straight lines projecting from the torso that are often interpreted as spears and most famously represented at Cougnac are also classified as ritualized humans. Both the sorcerer and wounded man are recurring motifs frequently depicted with bent postures as seen in ritual dance and with erect penises. They may or may not be associated with one or more animals and vary in extent of detailing. Those identified as images of sorcerers tend to be more detailed and, while very similar to each other, nevertheless idiosyncratic. Clottes and Lewis-Williams (1998) and Clottes (2008) cite these figures among evidence for shamanic ritual as the ideological context for image making in cave walls; however, iterations are found on portable objects from La Madeleine (Raux 2013). The class “ritualized human” is used here in an effort to avoid imposing a specific interpretation of who or what the figures represent while still emphasizing their consistency with known ritual behaviors and specifically their status outside the mundane.

A COMPARISON OF “SCENES” IN PARIETAL AND NON-PARIETAL UPPER PALEOLITHIC IMAGERY

189

Table 12.5 displays the percentage of scenes from each pre-Magdalenian and Magdalenian archaeological context with animal, human, and ritualized human agents. NARRATIVE EVENTS were not depicted on portable objects during the early periods of the Upper Paleolithic of France, but during the Magdalenian scenes most often include animals and, to a lesser extent, humans. Scenes with ritualized humans are fairly unusual, even when iconic depictions, and specifically narrative depictions, are common. Scenes made on rock shelter walls prior to the Magdalenian are dominated by events with human participants, but toward the end of the Ice Age they include only animals. Ritualized humans were not represented in narrative depictions on rock shelter walls. In sharp contrast, while scenes that were made on cave walls in early periods most frequently depict animals engaged in various activities, ritualized human actors are also represented with some frequency. Moreover, the proportion of narrative events with ritualized humans increased through time. Other human actors were not depicted on cave walls until the Magdalenian, and then only rarely. It is posited here that portable and parietal imagery constitute different types of information exchange that, in turn, account for the varying distribution of narrative events and event participants. Images on weaponry and other portable objects are necessarily small and visually accessible only to group members in close proximity. The information that is implicit in the artistic representations is exchanged between individuals (Culley FIGURE 12.3. The “sorcerer” at 2016; Davidson 1989; Wiessner 1983) and consumed passively as deco- Les Trois-Frères consistent with rated objects are used during every day activities. the ritual use of masks and transPortable objects are the ideal medium for narrative depiction that uses formational capacities assigned to smallness of elements to create a sense of motion and distance and hinges ritual practitioners. Wellcome Lion the separation of events and perceiving agents to create imaginational brary, London, Wikimedia Comspace. Moreover, the ability to create and manipulate imaginational space mons, CC BY-SA 4.0. through detailing increases the efficacy of portable art as a primary mechanism through which information about individuals’ histories, social roles, and affiliations is conveyed. Portable objects and narrative depictions reciprocally manifest their respective semiotic potentials. However, portable art is a poor mechanism TABLE 12.5. Percentage of Scenes Depicting Ritualized Humans, for exchanging group-level information Other Humans, or Only Animals. such as shared beliefs that are negotiated through group-level interactions, parMagdalenian Ritualized ticularly those involving ritual practices. Context Animals Humans Humans The large number of scenes depicting huportable objects mans and/or animals engaged in events (n=37) 65% 27% 8% of daily life and near absence of ritualized rock-shelter walls human event participants that are appar(n=4) 100% 0 0 ent in the sample assemblages should, cave walls then, not be a surprise. (n=18) 67% 6% 28% Parietal imagery affects inter- and intragroup-level information exchange due Pre-Magdalenian Ritualized to the physical independence from speContext Animals Humans Humans cific individuals and visual accessibility portable objects to all group members (Culley 2016; Wi(n=0) – – – essner 1983). Under this model, narrarock-shelter walls tive depiction should be far less common (n=3) 33% 67% 0 than on portable media, as seen in both cave walls the rock shelter and cave assemblages. (n=12) 83% 0 17% Where scenes are present on rock shelter

190

ELISABETH CULLEY

walls, their association with habitation sites and thus with mundane activities and social roles make the representation of ritualized human event participants unlikely. In addition to the group level dynamics of parietal image consumption that should limit narrative depiction on cave walls, scenes are inconsistent with apparent cave ontology as suggested by other attributes of cave art. Indeed, cave art is renowned for imagery that violates the two-dimensionality required of imaginational space. Many images create the dimensionality of animal bodies by incorporating natural features of the caves, including stalagmites, hollows, and bosses (Bahn and Vertut 1997: 105–6). The sense of relief would have been enhanced by flickering torch light, with the effect of animals extending from the walls into observational space (Clottes 2009). In other cases, animals were created as if emerging from or entering into crevices in the cave walls. Although the figures are almost always depicted in profile, their orientation out from or into wall features creates the same effect as looming figures that violate the boundaries of imaginative space and engage observers. Moreover, many decorated caves have handprints of the men, women, and children who entered the galleries and/or small bones and stones stuck into crevices (Clottes 2009), effectively merging artist, audience, images, artifacts, and caves. Each of these representational trends indicates that caves and cave art were not observed passively, but engaged with physically. The imagery is experiential and likely transformative. Within the context of caves as numinous points of engagement with the subjects represented, the images are the antithesis of narrative imaginative space viewed from the outside looking in. Certainly depictions of ordinary humans engaged in daily activities are unexpected there, while the depiction of ritualized humans may reify cave ontology.

Concluding Discussion In defining scenes as a universal perceptual category and depictive option, as well as universal markers that effect the recognition of narrative depiction, Dobrez has made a significant contribution to image classification efforts. This review and application of Dobrez’s perspective relative to Upper Paleolithic imagery points to some limitations in the use of perceptual markers to identify scenes. Most notably, there remains some question as to how many and which markers are necessary and sufficient to identify scenes with certainty, as well as how objective researchers can be in determining when visual elements constitute scene markers. Moreover, using only universal perceptual markers to identify scenes can exclude narrative depictions that would be identified using other information. The preliminary study also underscores the difficulties in establishing chronological control over Upper Paleolithic imagery. More specifically, it shows the lack of such control has masked variation in narrative depiction in Ice Age Europe. The analysis also highlights the importance of using the appropriate measures of scene frequencies. Nevertheless, even this limited analysis has isolated patterning in narrative depiction prior to and during the Magdalenian, and generated testable hypotheses regarding the semiotic potential of image-making contexts and the consequent distribution of narrative events and event participants. Ultimately, the study highlights significant potential for better understanding narrative depiction in the Upper Paleolithic and, more generally, for Dobrez’s perspective to make significant contributions to the study of art worldwide. Elisabeth Culley is a consultant with Matrix Design Group in Phoenix, Arizona (US). She is a specialist in the evolution of human cognition and the origins of art, symboling, and ideation and has worked in France, Spain, South Africa, the United States, and Guam. Her publications include Examining Metaphorical Reasoning in Rock Art Production (2003), Making Marks: Graduate Studies in Rock Art Research, ed. (2005), Supernatural Metaphors and Belief in the Past: Defining an Archaeology of Religion (2008), and the forthcoming “Art, Sign and Representation” with Iain Davidson in The Oxford Handbook of Human Symbolic Evolution, Second Edition.

A COMPARISON OF “SCENES” IN PARIETAL AND NON-PARIETAL UPPER PALEOLITHIC IMAGERY

191

Notes 1. Perceptual categories are represented in capital letters following established convention in cognitive linguistics in order to emphasize the structural nature of perceptual schema and to distinguish them from individual expressions of them. In this context, NARRATIVE EVENT is a perceptual category that effects the perception of specific narrative events, such as two animals crossing a stream. 2. To view these images see http://www.edwardhopper.net/images/paintings/girl-at-sewing-machine.jpg http://www.jwwaterhouse.net/author/waterhouse/page/2/ 3. For a more recent review of dating methods see the section on Numerical Dating Methods in I. Davidson (2017).

References Aubert, M., A. Brumm, M. Ramli, T. Sutikna, E. W. Saptomo, B. Hakim, M. J. Morwood, G. D. van den Bergh, L. Kinsley, and A. Dosseto. 2014. “Pleistocene Cave Art from Sulawesi, Indonesia.” Nature 514 (7521): 223–27. Aubert M, R. Lebe, A. A. Oktaviana, M. Tang, B. Burhan, J. A. Hamrullah, H. B. Abdullah, J. X. Zhao, I. M Geria, P. H. Sulistyarto, R. Sardi, A. Brumm A 2019. “Earliest Hunting Scene in Prehistoric Art.” Nature 576 (7787): 442–45. Aubert, M., P. Setiawan, A. A. Oktaviana, A. Brumm, P. H. Sulistyarto, E. W. Saptomo, B. Istiawan, T. A. Ma’rifat, V. N. Wahyuono, F. T. Atmoko, J. X. Zhao, J. Huntley, P. S. C. Taçon, D. L. Howard, and H. E. A. Brand. 2018. “Palaeolithic Cave Art in Borneo.” Nature 564 (7735): 254–57. Azéma, M. 1992. ”La représentation du mouvement dans l’art animalier paléolithique des Pyrénées.” Bulletin de la Société Préhistorique Ariège-Pyrénées 47: 19–76. ———. 2005. “Breaking down Movement in Paleolithic Art.” International Newsletter on Rock Art 43: 14–21. ———. 2006. “La représentation du mouvement au Paléolithique supérieur. Apport du comparatisme éthographique à l’interprétation de l’art pariétal.” Bulletin de la Société Préhistorique Française 103(3): 479–505. ———. 2008. “Representation of Movement in Palaeolithic Parietal Art: An ethographical approach.” Anthropozoologica 43(1): 117–54. ———. 2011. Préhistoire du cinéma. Origines de la narration graphique et du cinématographe. Paris: Errance. Azéma, M., and F. Rivière 2012. “Animation in Palaeolithic Art: A Pre-echo of Cinema.” Antiquity 86 (332): 316–24. Bahn, P. 1998. The Cambridge Illustrated History of Prehistoric Art. Cambridge: Cambridge University Press. Bahn, P., and J. Vertut. 1997. Journey Through the Ice Age. Berkeley: University of California Press. Bednarik, R. 1995. “The Coa Petroglyphs: An Obituary to the Stylistic Dating of Palaeolithic Rock-Art.” Antiquity 69(266): 877–84. Cantlon, J., P. Pinel, S. Dehaene, and K. Pelphrey. 2011. “Cortical Representations of Symbols, Objects, and Faces Are Pruned Back during Early Childhood.” Cerebral Cortex 21(1): 191–99. Chauvet, J.-M., E. B. Deschamps, and C. Hillaire. 1995. La Grotte Chauvet à Vallon-Pont-d’Arc. Paris: Seuil. Clark, G., and J. Riel-Salvatore. 2009. “What’s in a Name? Observations on the Compositional Integrity of the Aurignacian.” In The Mediterranean from 50,000 to 25,000 BP. Turning Points and New Directions, ed. M. Camps and C. Szmidt, 323–38. Oxford: Oxbow Books. Clottes, J. 1993. “Post-Stylistic?” In Rock Art Studies: The Post-Stylistic Era or Where Do We Go From Here? ed. P. Bahn and M. Lorblanchet, 19–25. Oxford: Oxbow. ———. 2001. “Paleolithic Europe.” In Handbook of Rock Art Research, ed. D.S. Whitley, 459–81. Walnut Creek, CA: AltaMira Press. ———. 2008. Cave Art. London: Phaidon. Clottes, J. 2009. “Sticking Bones into Cracks in the Upper Palaeolithic.” In Becoming Human: Innovation in Prehistoric Material and Spiritual Culture, ed C. Renfrew and I. Morley, 195–211. Cambridge: Cambridge University Press. Clottes, J., J.-M. Chauvet, E. Brunel-Deschamps, C. Hillaire, J.-P. Daugas, M. Arnold, H. Cachier, J. Evin, P. Fortin, C. Oberlin, N. Tisnerat, and H. Valladas. 1995. “Radiocarbon Dates for the Chauvet-Pont-d’Arc Cave.” International Newsletter on Rock Art 11: 1–2. Clottes, J., M. Garner, and G. Maury. 1994. “Magdalenian Bison in the Caves of the Ariège.” Rock Art Research 11(1): 58–70. Clottes, J., and D. Lewis-Williams. 1998. The Shamans of Prehistory: Trance and Magic in the Painted Caves. New York: Harry N. Abrams. Conard, N., M. Malina, and S. Münzel. 2009. “New Flutes Document the Earliest Musical Tradition in Southwestern Germany.” Nature 460: 737–40. Culley, E. 2006. “The Meaning of Metaphor: A Cognitive Approach to Prehistoric Ideation.” M.A. thesis. Flagstaff, AZ: Northern Arizona University.

192

ELISABETH CULLEY

———. 2008. “Supernatural Metaphors and Belief in the Past.” In Belief in the Past: Theoretical Approaches to the Archaeology of Religion, ed. D. S. Whitley and K. Hays-Gilpin, 67–84. Walnut Creek, CA: Left Coast Press, Inc. ———. 2016. “A Semiotic Approach to the Evolution of Symboling Capacities During the Late Pleistocene with Implications for Claims of Modernity in Early Human Groups.” Ph.D. dissertation. Tempe, AZ: Arizona State University. Davenport, D., and M. Jochim. 1988. “The Scene in the Shaft at Lascaux.” Antiquity 62 (236): 558–62. Davidson, I. 1989. “Freedom of Information: Aspects of Art and Society in Western Europe during the Last Ice Age.” In Animals into art, ed. H. Morphy, 440–56. London: Unwin Hyman. ———. 2017. “Paleolithic Art.” In Oxford Bibliographies in Anthropology, ed. J. Jackson. New York: Oxford University Press. doi: 10.1093/obo/9780199766567-0173. Deacon, T. 2012. “The Symbol Concept.” In The Oxford Handbook of Language Evolution, ed. M. M. Tallerman and K. R. K. Gibson, 393–405. Oxford: Oxford University Press. Delporte, H. 1984. “L’art Mobilier et ses Rapports avec La Faune Paléolithique.” In La Contribution de la Zoologie et de L’ethologie à L’interprétation de L’art des Peuples Chasseurs Préhistoriques. Numero 3 Colloque de la Societe Suisse des Sciences Humaines, ed. H. G. Bandi, W. W. Huber, M.-R. Sauter, and B. Sitter, 111–42. Fribourg: Editions Universitaires Fribourg Suisse. Dobrez, L. 2011. “Looking at Our Looking: A New Approach to the Definition of a Rock Art Scene.” American Indian Rock Art 37: 251–64. ———. 2012. “Towards a More Rigorous Definition of Terms: Are There Scenes in European Palaeolithic Art?” In L’Art Pléistocène dans le Monde/ Pleistocene Art of the World / Arte Pleistoceno en el Mundo, Actes du Congrès IFRAO, Tarascon-sur-Ariège, September 2010. Numero Spécial de Prehistoire, Art et Sociétés, Bulletin se la Société Préhistorique Ariège-Pyrénées, LXV–LXVI, 2010–2011, ed. J. Clottes, CD-1837–51. ———. 2013. “The Perception of Depicted Motion.” Arts 2: 383–446. ———. 2015. “Making Sense of Pictures.” Bollettino del Centro Camuno di Studi Preistorici 40(7): 31–47. Dobrez, L., and P. Dobrez. 2013a. “Canonical Form and the Identification of Rock Art Figures.” American Indian Rock Art 39: 115–29. ———. 2013b. “Rock Art Animals in Profile: Visual Recognition and the Principles of Canonical Form.” Rock Art Research 30: 75–90. ———. 2014. “Canonical Figures and the Recognition of Animals in Art and Life.” Boletín del Museo Chileno de Arte Precolombino 19(1): 9–22. Douka, K., V. Slon, Z. Jacobs, C. Bronk Ramsey, M. Shunkov, A. Derevianko, F. Mafessoni, M. Kozlikin, B. Li, R. Grün, D. Comeskey, T. Devièse, S. Brown, B. Viola, L. Kinsley, M. Buckley, M. Meyer, R. Roberts, S. Pääbo, J. Kelso, and T. Higham. 2019. “Age Estimates for Hominin Fossils and the Onset of the Upper Palaeolithic at Denisova Cave.” Nature 565 (7741): 640–44. Goldstone, R. 1994. “Influences of Categorization on Perceptual Discrimination.” Journal of Experimental Psychology 123(2): 178–200. Guthrie, D. 1984. “Ethological Observations from Palaeolithic Art.” In La Contribution de la Zoologie et de L’ethologie à L’interprétation de L’art des Peuples Chasseurs Préhistoriques. Numero 3 Colloque de la Societe Suisse des Sciences Humaines, ed. H. G. Bandi, W. W. Huber, M.-R. Sauter, and B. Sitter, 35–74. Fribourg: Editions Universitaires Fribourg Suisse. Halverson, J. 1992. “Paleolithic Art and Cognition.” The Journal of Psychology 126(3): 221–36. Iglesias, J., J. Eriksson, F. Grize, M. Tomassini, and A. Villa. 2005. “Dynamics of Pruning in Simulated Large-Scale Spiking Neural Networks.” BioSystems 79(9): 11–20. Jewitt, C., and R. Oyama. 2001. “Visual Meaning: A Social Semiotic Approach.” In Handbook of Visual Analysis, ed. T. van Leeuwen and C. Jewitt, 134–56. London: Sage. Kay, P., B. Berlin, and W. Merrifield. 1995. “Biocultural Implications of Systems of Color Naming.” In Language, Culture, and Society: A Book of Readings, ed. B. G. Blount, 456–69. Prospect Heights, IL: Waveland Press, Inc. Keyser, J. 2001. “Relative Dating Methods.” In Handbook of Rock Art Research, ed. D. S. Whitley, 116–38. Walnut Creek, CA: AltaMira Press. Lakoff, G. 1987. Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press. Lakoff, G., and M. Johnson. 1999. Philosophy in the Flesh. New York: Basic Books. Leason, P.A. 1939. “A New View of the Western European Group of Quaternary Cave Art.” Proceedings of the Prehistoric Society 5(1): 51–60. Lenssen-Erz, T. 1992. “Coherence—A Constituent of Scenes in Rock Art: The Transformation of Linguistic Analytical Models for the Study of Rock Paintings in Namibia (with Commentaries and Reply). Rock Art Research 9: 87–105. Levine, M. 1957. “Prehistoric Art and Ideology.” American Anthropologist 59(6): 949–64. Mervis, C., and E. Rosch. 1981. “Categorization of Natural Objects.” Annual Review of Psychology 32(1): 89–115. Nowell, A. 2015. “Learning to See and Seeing to Learn: Children, Communities of Practice and Pleistocene Visual Cultures.” Cambridge Archaeology Journal 25(4): 889–99. Nöth, W. 1990. Handbook of Semiotics. Bloomington: Indiana University Press. Peirce, C. (1931) 1960. The Collected Papers of Charles Sanders Peirce: Volumes I & II, ed. C. Hartshorne and P. Weiss. Cambridge, MA: The Belknap Press.

A COMPARISON OF “SCENES” IN PARIETAL AND NON-PARIETAL UPPER PALEOLITHIC IMAGERY

193

Pike, A. W., D. L. Hoffmann, M. García-Diez, P. B. Pettitt, J. Alcolea, R. De Balbin, C. González-Sainz, C. de las Heras, J. A. Lasheras, R. Montes, and J. Zilhao. 2012. “U-Series Dating of Paleolithic Art in 11 Caves in Spain.” Science 336(6087): 1409–13. Prather, J., S. Nowicki, R. Anderson, S. Peters, and R. Mooney. 2009. “Neural Correlates of Categorical Perception in Learned Vocal Communication.” Nature Neuroscience 12(2): 221–28. Quiles, A., H. Valladas, H. Bocherens, E. Delqué-Kolicˇ, E. Kaltnecker, J. van der Plicht, J.-J. Delannoy, V. Feruglio, C. Fritz, J. Monney, M. Philippe, G. Tosello, J. Clottes, and J.-M. Geneste. 2016. “A High-Precision Chronological Model for the Decorated Upper Paleolithic Cave of Chauvet-Pont d’Arc, Ardèche, France.” Proceedings of the National Academy of Sciences 113(17): 4670–75. Raux, P. 2013. “Les Sorciers de la Préhistoire dans l’Art Paléolithique.” Bulletin de la Société d’Études et de Recherches Préhistoriques des Eyzies 62: 121–31. Roberson, D., I. Davies, and J. Davidoff. 2000. “Color Categories are Not Universal: Replications and New Evidence from a Stone-Age Culture.” Journal of Experimental Psychology 129(3): 369–98. Rosch, E. 1975. “Universals and Cultural Specifics in Human Categorization.” In Cross-Cultural Perspectives on Learning, ed. R. W. Brislin, S. Bochner, and W. J. Lonner, vol. 1, 177–206. Hoboken, NJ: John Wiley & Sons. ———. 1977. “Human Categorization.” In Studies in Cross-Cultural Psychology, ed. N. Warren, 1–49. London: Academic Press. ———. 1978. “Principles of Categorization.” In Cognition and Categorization, ed. E. Rosch and B. B. Lloyd, 27–48. Hillsdale, NJ: Lawrence Erlbaum Associates. Rosch, E., and B. Lloyd. 1978. Cognition and Categorization. Hillsdale, NJ: Lawrence Erlbaum Associates. Rosch, E., C. Mervis, W. Gray, D. Johnson, and P. Boyes-Braem. 1976. “Basic Objects in Natural Categories.” Cognitive Psychology 8: 382–439. Rowe, W. 2001. “AMS Radiocarbon Analysis.” In Handbook of Rock Art Research, ed. D. S. Whitley, 139–66. Walnut Creek, CA: AltaMira Press. Salazar-Sutil, N., and S. Sebastián Melo. 2016. “Exposed to Time: Cross-histories of Human Motion Visualization from Chrono-to Dynamophotography.” In The Oxford Handbook of Screendance Studies, ed. D. Rosenberg, 143–67. Oxford: Oxford University Press. Sloutsky, V. 2010. “From Perceptual Categories to Concepts: What Develops?” Cognitive Science 34(7): 1244–86. Ucko, P., and A. Rosenfeld. 1967. Palaeolithic Cave Art. New York: McGraw-Hill. Valladas, H., N. Tisnérat-Laborde, H. Cachier, M. Arnold, F. Bernaldo de Quirós, V. Cabrera-Valdés, J. Clottes, J. Courtin, J. J. Fortea-Pérez, C. Gonzáles-Sainz, and A. Moure-Romanillo. 2001. “Radiocarbon AMS Dates for Paleolithic Cave Paintings.” Radiocarbon 43(2B): 977–86. Von Petzinger, G., and A. Nowell. 2011. “Question of Style: Reconsidering the Stylistic Approach to Dating Paleolithic Parietal Art in France.” Antiquity 85(330): 1165–83. Von Petzinger, G., and A. Nowell. 2014. “Non-Figurative Signs and the Antiquity of Chauvet Cave.” Journal of Human Evolution 74: 37–54. White, R. 2003. Prehistoric Art: The Symbolic Journey of Humankind. New York: Harry N. Abrams. Wiessner, Polly. 1983. “Style and Social Information in Kalahari San Projectile Points.” American Antiquity 48(2): 253–76. Zentall T., E. Wasserman, O. Lazareva, R. Thompson, and M. J. Rattermann. 2008. “Concept Learning in Animals.” Comparative Cognition & Behavior Reviews 3: 13–45.

==== •

•

•

13. SCENE MAKERS Finger Fluters in Rouffignac Cave (France) Leslie Van Gelder and April Nowell

Introduction Finger flutings are lines drawn by peoples with their fingers in the soft surfaces of cave walls, ceilings and floors. To “flute” means to create a surface that is grooved, channeled or ribbed and thus the term “finger fluting” captures the three-dimensional morphology of these lines. Finger flutings are found in caves in modern day France, Spain, and Australia. Though no finger flutings have been directly dated themselves, in Europe they are associated with art dating to the Late Pleistocene in caves such El Castillo (ca. 40,800 BP), Chauvet (ca. 36,000 BP), Gargas (ca. 27,000 BP), Pech Merle (ca. 25,000 BP) and Altamira (ca. 18,500 BP), as well as younger caves of the Magdalenian (ca. 18,000–12,000 BP) such as Las Chimeneas and Rouffignac.1 Although some finger flutings depict figurative images of bison, mammoth, rhinoceros, and deer, and others depict non-figurative signs, they are mostly non-representational lines. Finger flutings have been variously interpreted as doodling (Breuil 1912), serpents (Barrière 1982; Nougier and Robert 1958), or water images (Marshack 1977), the residue of surface preparation for making art (Lorblanchet 1992), and evidence of the moment when a shaman touches the “skin” of the otherworld (Lewis-Williams 2002). Though they make up a larger percentage of cave art than any other form, finger flutings were largely ignored during the first century of the study of Paleolithic art, as they are enigmatic and not always visually appealing. Breuil (1912) famously referred to them as “traits parasites” (parasite lines) and deleted them from his re-drawings of cave images, believing they detracted from the figurative art. The nonfigurative nature of the vast majority of finger-fluted lines poses unique challenges for a discussion of scenes in Paleolithic rock art. Is it possible to identify certain panels of finger flutings as deliberately created “scenes” by the artists who made them? Alternatively, are we at risk of ascribing our own modern visual values to finger-fluted panels because the visual literacy of our time recognizes an abstract set of lines bounded by white space as being similar to a modern-day artist’s tableau or canvas? Similarly, are we more likely to give greater weight to the interpretation of seemingly orderly flutings versus those that do not have the same linearity or other apparent patterning (Figure 13.1a, 1b)? Drawing on examples from Rouffignac Cave in southern France, this chapter considers whether scenes can be identified among the finger flutings by exploring what is meant by “scenes” in relation to unique aspects of finger fluting data.

Are There Scenes Among the Finger Flutings? What is a Scene? At its most basic, a scene in rock art can be defined as the intentional placement of images (figurative or non-figurative) in close proximity to each other with some distance from other such images where some interaction or relationship between the images is indicated (e.g., Lenssen-Erz 1992). While examples of this kind of scene will be presented below, finger flutings may be able to contribute more deeply to discussions about scenes in rock art if we are willing to reframe the conversation—to shift our perspective somewhat.

SCENE MAKERS

195

FIGURE 13.1a–b. (a) Flutings from the ceiling of Chamber E, Rouffignac Cave, southern France, appear to be organized into specific units or images with clear spaces between them. (b) flutings from Chamber A1 of the same cave show meandering lines that some researchers have referred to as snake images (Nougier 1958; Barrière 1982: 205), water images (1977) or male symbols (Leroi-Gourhan 1958). Copyright Leslie Van Gelder.

Unlike many forms of rock art that leave behind images but little forensic information about the artists themselves, finger flutings afford researchers the opportunity to identify unique individuals within the cave setting, and in many instances we are able to determine “who’s who” in the caves. Focusing on the actions of individual artists allows us to turn the visual tables, so that instead of focusing our attention solely on the outcome or product of their creations as scenes, the actions of the fluters help us define what may be a scene in terms of what was meaningful to them. The “scene” we see on the cave walls in the form of their finger flutings becomes a moment frozen in time, capturing the relationships among the people who came to the cave to mark its ceilings and walls. We refer to this as an “archaeology of intimacy” or an “archaeology of the small.”

Communities of Practice When we are trying to reconstruct relationships between people and materials, particularly in the context of craft production, it is often useful to approach them within the context of “communities of practice.” A community of practice is defined by a “network of relations among people and objects mediated by [the] actions they conduct” (Joyce 2012: 150). In Paleolithic art, Conkey (2009) argues that the “images and forms were generated within and by communities of practice . . . in that sense [they are] not art images, rather they are [the] artful integration of many entangled material and social factors.” Community of practice research employs a “biography of things” perspective whereby objects have histories in their own right and in which their relationships, importance, and roles change over time (Gosden and Marshall 1999). Importantly for our purposes here, it also focuses on the role that learning and apprenticeship play in the cycle of the social reproduction of knowledge and practice (Nowell 2015, 2017). The approach provides the necessary theoretical tools and vocabulary to explore social decision making and social participation, and how these might have changed over time and space (Nowell 2017). By identifying who entered a cave, who fluted with whom, what they chose to draw with their fingers (be it figurative or non-figurative), and what the totality of their interactions were, we can begin to document what were meaningful bounded units or (broadly defined) “scenes” to these people. To illustrate this approach, we discuss finger fluting panels associated with the two of the eight fluters identified in Rouffignac Cave who are among its most prolific fluters and participated in the creation of multiple panels in diverse locations within the cave. The six examples we present document the diversity of scenes in which these two individuals engaged with each other and others in the cave environment while they were leaving their marks upon the walls.

196

LESLIE VAN GELDER AND APRIL NOWELL

Rouffignac Cave Rouffignac Cave is located 4 km from the village of Rouffignac in the French department of the Dordogne. Rouffignac contains over 500 m2 of finger flutings (Plassard 1999: 62). The art in Rouffignac has never been formally dated, but based on stylistic considerations it is considered to be 13–14,000 years old. However, Sharpe and Van Gelder (2006c) have suggested it could be more than 20,000 years old because there are cave bear scratches on top of two of the drawings and some of the flutings, and cave bears are thought to have died out in this region some 20,000 years ago (Pacher and Stuart 2009; Sharpe and Van Gelder 2006c). Finger flutings are found in eight chambers of the cave. Figurative drawings made with manganese dioxide, as well as engravings made with sharp tools, are found largely in the chambers known as the Salon Rouge (Chamber B), the Galerie Henri Breuil (Chamber G), and the Voie Sacrée (Chamber G), with the highest concentration of images found at the Grand Platfond. The farthest decorated section of the cave, the Grande Fosse (Chamber F), which contains finger flutings, is located nearly 0.94 km from the entrance. The original cave floor had a large number of bear pits, as bears regularly wintered in the cave. The cave has since been modified for tourist visits involving a small train, and the floor along the train route has been compromised; however, the floors along the edges of the chambers and the chambers in which the train does not run have not been impacted.

Method for Identifying Individual Fluters Early interpretations of finger flutings were proposed without supporting evidence, as no methodology by which to study flutings existed until the beginning of the twenty-first century. In 2001–2008, while working at the French cave sites of Rouffignac and Gargas, the late Kevin Sharpe and Leslie Van Gelder developed, tested, and applied a method for systematically studying flutings. This endeavor included the development of a system of nomenclature (Sharpe and Van Gelder 2006a, 2006b, 2010a), replicable data collection protocols (Sharpe and Van Gelder 2006d, 2007, 2010a, 2010b), a comparative database to identify the age of fluters (Sharpe and Van Gelder 2004, 2006a, 2006c), means to identify the sex of fluters (Sharpe and Van Gelder 2009), and the application of Zipf ’s Law (communications theory) to determine which panels might represent meaningful communication (Sharpe and Van Gelder 2009; Van Gelder 2015d). In 2009, the research program was expanded to an additional twelve caves in France and Cantabrian Spain (El Castillo, Las Chimeneas, El Cudón, Hornos de la Peña, La Clotilde, La Estación, El Calero II, Las Brujas, El Juyo, El Salitre, La Flecha, and Castro Urdiales) and eventually to one cave in Australia (Koonalda). These new studies led to the development of methods for studying figurative animals drawn with one finger (Van Gelder 2010, the role of children in the production of cave art (Van Gelder 2012, 2015a, 2015b, 2015e), group sizes of parties entering caves (Van Gelder 2015d), and collaborations between fluters in the production of fluted panels (Van Gelder 2015c). Our method for analyzing finger flutings utilizes the width of lines drawn by the middle three fingers of the hand as the source for identifying unique individuals (Sharpe and Van Gelder 2005, 2006a; 2006b; 2006c; 2006d; Van Gelder 2010). The middle three fingers are used because the fluters used these most consistently and one need not determine right/left hand to study their width, as they remain the same three fingers regardless. Lines are measured at the narrowest point of a three-fingered unit in a fluted stream so as to avoid debris and to gain the most accurate measure of the fingers themselves. We use a margin of +/- 2 mm in the determination of an individual, since the same person may flute in the same panel with fingers closer together and wider apart. This also accounts for bodily asymmetry between right and left hands, as we have found in recent studies that there may be up to 2 mm difference between the three-finger widths of the left and right hands of the same individual (Van Gelder 2015a). Overlays and underlays are recorded to determine superposition of lines. Whenever possible, lines that would have been drawn contemporaneously are noted (i.e., when lines cross both over and under each other) as this indicates two individuals fluting in concert with each other. Hand studies have allowed us to determine that a three-finger width of 30 mm or smaller is found only in children under the age of five, while widths of 33 mm or smaller are generally found to predominate in children aged seven years or younger or young adolescents (Sharpe and Van Gelder 2006a: 943; Van Gelder

SCENE MAKERS

197

2015). In our studies of children’s hands we found that children as young as five can have a three-fingered width as wide as 40 mm and an adolescent of twelve can have a three-fingered width as wide as 51 mm, which parallels widths found among adult males. This suggests that although we can establish that only children could have made the very small-width flutings, we recognize that young children could also could be responsible for other flutings that we cannot at present distinguish from those made by non-children. When fingertips are visible and the fluting hand (right or left) can be established, we use this data to determine the sex of individual fluters using the 2D:4D ratio (Sharpe and Van Gelder 2005; Van Gelder and Sharpe 2009). This ratio refers to the relationship between the length of a person’s second digit (or index finger) to his or her fourth digit (or ring finger). The 2D:4D ratio is predetermined in utero through exposure to estrogen and testosterone (Koehler, Simmons, and Rhodes 2004). Ratios < 1.0 are typical of males while ratios of >1.0 are typical of females (Manning et al. 1998). This ratio has also been applied more recently to handprints in rock art (Snow 2013). Distances between thumb and index finger are measured to determine hand size when possible. Depth of fluting, height from the floor, speed of fluting (faster fluting creates more buildup of debris), length of fluting, and hand used to create the fluting are also noted, as well as idiosyncratic choices such as the making of arcs, diagonals, or zigzags. We use these variables in addition to the three-fingered width data to develop unique profiles of individuals. If we were to use only the width data it might be possible to confound individuals, and it remains possible that we are identifying the minimum number of individuals responsible for fluting in a cave, but the fact that there are regularities/patterns of choices and preferences associated with specific widths strongly suggests that we are capturing the products of unique individuals. Chambers and panels are analyzed for the number of individuals within them and their physical relationships to each other within the cave context. We are aware that the height from the floor can be a false measure of the height of an individual, so we also note ways in which an individual could reach certain heights if no specific physical structure in the cave could provide access.

Two Individuals Though eight individuals (Table 1) have been identified as fluters in Rouffignac Cave (Van Gelder 2016b), this chapter will focus on scenes in which two of the individuals have participated to highlight the diversity of scenes that they created and the circumstances in which it is possible to envision them. The first of the two fluters has a three-fingered marking width of 38 mm. Based on the profiles of this person’s hand, we believe the fluter to be male as his 2D is lower than his 4D relative to his middle finger (Van Gelder and Sharpe 2009). Acknowledging the controversial nature of using the 2D:4D ratio in determining the sex of individuals based on their hands (e.g., Barrett and Case 2014), we do not consider TABLE 13.1. List of individual finger fluters documented at this an absolute, but for the purposes of this Rouffignac Cave. chapter we will refer to this individual as a Measure of Number of male. He is among the most prolific of the 3-fingered width of individuals fluters in the cave, as his flutings are found Panel or Location individuals in mm creating panel in Chambers A1, A2, E, F, G, Voie Sacrée, 38, 48 2 H, and H1, and he produced seven of the Chamber G Mammoths of fourteen identified tectiforms in the cave (Van Gelder 2016a). A tectiform is one of Discovery Panel a limited number of non-figurative signs in Paleolithic art. It has the rough shape of an upward pointing arrow with a bar underneath (Figure 13.2). The name comes from the Latin meaning “roof ” as early researchers believed they might be iconic representations of dwelling structures (Bahn 2016; von Petzinger 2016), but currently their meaning remains enigmatic.

Chamber G Galerie Henri Breuil Patriarch Panel

34, 38

2

28, 34, 38 , 41, 44, 48

6

Chamber G4 Saiga Panel

31, 34, 38, 44, 48

5

Grand Fosse

31, 34, 38

3

Chamber G Voie Sacre Rhinoceros 30 Panel

198

LESLIE VAN GELDER AND APRIL NOWELL

FIGURE 13.2. Tectiforms 35 and 36, Voie Sacrée, Rouffignac Cave. Copyright Leslie Van Gelder.

The second individual on whom we will focus has a 34-mm three-fingered marking width. Based on the profile of this fluter’s hand, we believe her to be a female (Van Gelder and Sharpe 2009). Though a hand width of 33 mm is considered the upper limit for a child (Sharpe and Van Gelder 2006a), we note that other aspects of her flutings suggest that she is a youth. This is based on the height of her flutings, and her engagement with the other children’s flutings that are oftentimes lower than those of Fluter 38 mm when they are in concert with each other. While we cannot say this for certain, it is likely that she was young, perhaps seven or eight years old. She is prolific in her flutings, especially in Chambers A1, A2, G, H, and F. She, more than any of the other finger fluters in the cave, is prone to drawing curved and arc-like lines (Cooney 2013: 177).

Case Studies from Rouffignac Cave Tectiform Pair In the Voie Sacrée, on the right-side wall approximately 504 m from the cave entrance, there are two tectiforms (Figure 13.2) (named tectiform 35 and 36 by Barrière 1982: 30). They show a scene of the relationship between Fluter 34 mm and Fluter 38 mm. The tectiform to the left was drawn by Fluter 34 mm and on the right by Fluter 38 mm. The tectiform drawn by Fluter 34 mm is 50 cm in height (1.20 m from the floor at its highest point), and the tectiform drawn by Fluter 38 mm is 70 cm (1.42 m from the floor at its highest point). If these were drawn at the same time, they likely show the height difference between the two fluters. The top right-side lines of Fluter 34 mm’s tectiform cross over the bottom left lines of Fluter 38 mm’s tectiform suggesting that her tectiform was drawn after Fluter 38 mm’s. Whether drawn in imitation of Fluter 38 mm’s more complex tectiform, we do not know, but this likely shows the close proximity of the two fluters.

SCENE MAKERS

199

FIGURE 13.3. Chamber G4 Panel, Rouffignac Cave. Copyright Leslie Van Gelder.

Chamber G4 Panel In Chamber G4, located past two rhinoceros drawn on the ceiling of Chamber G4, approximately 455 m from the cave entrance (Barrière 1982: 19), there is a panel 1.62 m high and 0.82 m wide that is made up of 19 thick lines drawn with a tool (most likely a bone tool based on the thickness of the line and the lack of striations which a stick would leave behind). Seven streams of finger flutings were drawn over the lines (Figure 13.3). Animals scratched over parts of the bottom of the panel after the lines were made. The streams of finger flutings are all visually similar in that they are in each case a single swipe of a hand made with three or four fingers, all but one drawn at a similar angle and spaced closely. At first glance, it would be easy to imagine this to be the work of a single artist, especially because of the similarity of the angle of the flutings. However, the measurements of the fluted lines offer a different story and create the possibility of envisioning a scene where the lines were drawn by a group of five different individuals whose three-fingered hand measures are 48 mm, 44 mm, 38 mm, 34 mm, and 31 mm, respectively. All of these individuals are found in other parts of the cave, and previous studies on sex have suggested that Fluter 48 mm, Fluter 44 mm, and Fluter 34 mm are female while Fluter 38 mm is likely male and Fluter 31 mm is indeterminate (Van Gelder and Sharpe 2009). Fluter 38 mm has drawn three of the streams of flutings. Each of the others has only fluted once. One of the fluters, Fluter 31 mm, is a child, and yet his/her marks are nearly at the highest point of the panel (1.62 m) suggesting that perhaps s/he was held up on an adult’s hip or lifted to reach that height.

Rhinoceros Panel Fewer than 200 meters away from the G4 panel, 488 m from the entrance of the cave (Barrière 1982: 27) in the Voie Sacrée, is the Rhinoceros Panel (named for a partially drawn rhinoceros to the right of the panel)

200

LESLIE VAN GELDER AND APRIL NOWELL

FIGURE 13.4. Rhinoceros Panel, Rouffignac Cave. Copyright Leslie Van Gelder.

(Figure 13.4). In this panel, five individuals have fluted. The fifth fluter is not visible in this image but has fluted just to the right of this image with a low 48 mm fluting. Of those fluting, Fluter 38 mm is once again the most prolific on the panel, having drawn five of the twelve measurable streams. The flutings of Fluter 34 mm appear next to Fluter 38 mm’s lines. To the left of Fluter 34 mm are lines drawn by one of the other children, Fluter 28 mm. We believe that both Fluter 28 mm and Fluter 34 mm were likely held up to reach the height of 1.65 m on this panel. Here, as in the G4 panel, our two fluters are in the company of other members of the group, all of whom left at least one stream of flutings and together create a visual scene with their marks.

Saiga Panel Three of the individuals (Fluter 31 mm, Fluter 34 mm, and Fluter 38 mm) who appeared in the previous two panels are also found at the Saiga Panel, located at the terminus of the Grande Fosse, 0.94 km from the entrance of the cave, at the furthest point in which finger flutings are found in the cave (Figure 13.5). A person must cross many bear pits to reach this section of the cave. For an adult, reaching this location would require crawling for long sections, though perhaps a smaller child would be able to walk or crouch. Our research suggests that at least one if not two of the individuals in this group were likely children under the age of seven (Sharpe and Van Gelder 2006a; Van Gelder 2015a). The flutings are drawn approximately 1.2 m from the floor of the chamber.

SCENE MAKERS

201

FIGURE 13.5. Saiga Panel, Grande Fosse, Rouffignac Cave. Copyright Leslie Van Gelder.

Barrière has described these flutings as an image of a saiga antelope drawn with fingers (Barrière 1982: 87). If the interpretation of this drawing as an antelope is correct, then it represents an image that was co-created by Fluter 34 mm and Fluter 38 mm. The third line of the drawing (not shown here), which is sometimes included by those representing the panel, was drawn by Fluter 31 mm. Here, then, is a scene of two children co-creating their work with a third, older member of the group.

Frieze of Five Mammoths One of the more surprising aspects of finger flutings is the general absence of navigational lines. One would expect people unfamiliar with a cave, especially in the dark, to drag their hands along the cave walls as they walked, or to pause and leave handprints where they leaned into the wall (the walls are still soft enough to record these marks). The only instance of hand dragging is in a very small section of the Voie Sacrée in Rouffignac where Fluter 38 mm, Fluter 34 mm, and Fluter 28 mm, at different heights, dragged their hands horizontally across the same 3-m section of wall near a complex engraved frieze of five mammoths (Figure 13.6). The navigational lines drawn along this wall may also capture their relative heights, and it is possible to envision the scene of the two younger children and the older 38 mm all walking along this wall dragging their hands as they walked.

Patriarch Panel Another panel is located in Galerie H. Breuil across from the Frise de Onze Mammouths approximately 488 m into the cave (Figure 13.7). The mammoth on the left side, drawn by Fluter 38 mm, is 110 cm long. The mammoth on the right side, drawn by Fluter 34 mm, is approximately 73 cm long. Though the panel

202

LESLIE VAN GELDER AND APRIL NOWELL

FIGURE 13.6. Navigational lines, Voie Sacrée, Rouffignac Cave. Copyright Leslie Van Gelder.

FIGURE 13.7. Patriarch Panel, Galerie H. Breuil, Rouffignac Cave. Copyright Leslie Van Gelder.

SCENE MAKERS

203

continues on to the right, where lines have also been drawn by Fluter 28 mm, in this scene we see the relationship between Fluter 38 mm and Fluter 34 mm as they drew two mammoths facing each other. In three instances in this section of the panel, Fluter 38 mm’s lines overlap those drawn by Fluter 34 mm.

Discussion and Conclusion As noted above, at its most basic, a scene in rock art can be defined as the intentional placement of images (figurative or non-figurative) in close proximity to each other at some distance from other such images, where some interaction or relationship between the images is indicated. Repetition of figures or markings is often cited as an example of this kind of relationship (e.g., Lenssen-Erz et al. this volume; Ross this volume; Villaverde this volume). In the panels described above, repetition is a device that is often employed by the fluters. For example, along the Voie Sacrée, two tectiforms are placed side by side (Figure 13.2) while in Chamber G4, six carefully fluted arcs are aligned in a row, all angled in the same direction and diminishing in size as one moves toward the ceiling (Figure 13.3). Another means of inferring a relationship between figures is when there is some interaction between them (L. Dobrez 2013, this volume; Brusgaard and Akkermans this volume). This kind of scene is often described as rare in Paleolithic parietal art (Bahn 2016; but see Azéma and Rivière 2012; Cully this volume). Though only minimally rendered, the two fluted mammoths on the Patriarch Panel, pictured in Figure 13.7, are facing each other, locked in a shared gaze. This scene is reminiscent of two manganese-drawn mammoths in Chamber G3 in Rouffignac who are facing each other and are often described as being in combat (Barrière 1982: 113). Other ways of identifying a scene include narrativity, a sequence of actions (e.g., the lioness hunt at Chauvet), or a breaking down of the “fourth wall”—in other words, an engagement with viewers not involved in mark making. None of these elements can be said to characterize the fluted panels at Rouffignac, but by applying a communities of practice approach to an analysis of the flutings, we can make additional arguments about relationships between images and thus for the existence of scenes. The eleven flutings on the Rhinoceros Panel (Figure 13.4) and the co-created saiga (Figure 13.5) are two good examples. The eleven fluting streams on the Rhinoceros Panel lack the internal organization of the arced flutings in Chamber G4 (Figure 13.4), so in this case the repetition is more likely a function of the medium itself than any obvious attempt at duplicating a particular form. However, the intentional relationship between the fluting streams becomes clear when we realize that they were made by different hands. Dobrez (2016: 8) defines the composition of a scene as “formal arrangement in the eye of an observer, whether or not it was also in the eye of the maker.” In our research, intentionality and cohesiveness of the scene can be ascertained by documenting the actions of the fluters—we do not have to guess what they intended. Obviously, any human-produced marks are “intentional” on some level, but the fact that different hands were creating marks in a similar way in close proximity to each other suggests true intentionality. Similarly, the saiga in the Grande Fosse (Figure 13.5) was co-created by two or possibly three individuals fluting in concert. We would normally not consider a single figure a “scene” (see L. Dobrez 2013, 2016, this volume) unless there was an indication of action or unless, knowing the cultural context in which the figure was produced, we could confidently assert that a figure was referencing a larger narrative (see Dobrez 2015, 2016). In this case, however, the scene is a result of the bodily interactions between fluters as they engaged in the production of this figure. The saiga and even the flutings on the Rhinoceros Panel bring to mind the Panel of Dots at Chauvet Cave, France, dating to approximately 34,000 years ago (Figure 13.8). The figure has been described variously as a bison, a rhinoceros (Mohen 2002), or a mammoth (Clottes 2010). The image was created by several people pressing their palms into pigment and then pressing their hands onto the cave wall to form a coherent image. On a superficial level, there is no narrative quality to the Chauvet figure per se, but when we delve more deeply into its production, its role in a larger narrative and its “scene-ness” become apparent. This figure, like the Rouffignac flutings, becomes a moment fixed in time that captures the relationships among the people who came to the cave and intentionally created these visual units bounded in space and time.

204

LESLIE VAN GELDER AND APRIL NOWELL

FIGURE 13.8. Panel of dots at Chauvet Cave, France, dating to approximately 34,000 years ago, described variously as a bison, a rhinoceros (Mohen 2002), or a mammoth (Clottes 2010). The image was created by several people dipping their palms into pigment and then pressing their hands onto the cave wall to form a coherent image. Copyright Professor Jean Clottes.

In a similar fashion, the communities of practice approach lends an added dimension to our understanding of other fluted panels such as the two tectiforms along the Voie Sacrée. The fact that the tectiform created by 38 mm overlaps the one rendered by 34 mm gives us at a minimum the order of production but also suggests that they were made very closely in time—likely at the same time—given the evidence for other interactions between these two throughout the cave. Interpreting the navigational lines (Figure 13.6) as a scene in a traditional sense is less straightforward, as the lines appear to have been made haphazardly—an unintended consequence of navigating uneven (and unfamiliar?) terrain in low light. What undermines that interpretation is the fact that, to our knowledge, there are no other stray lines like this in the caves that we have studied. In conclusion, scenes in Paleolithic parietal art are thought to be rare, but by adding finger-fluted data, we can expand the number of examples of scenes in this time period. Not only is the content of many finger-fluted panels consistent with the definition of scenes, but an analysis of the forensic data afforded by the very nature of the fluting streams provides a window onto the intentionality and cohesiveness of scenes not available for other media. In the six examples from Rouffignac Cave provided here, we have documented moments frozen in time that capture the relationships among the people who came to the cave and marked its ceiling and walls. In some sense, that is the scene we are ultimately able to see through the residue they left behind: their story. Leslie Van Gelder PhD is a member of the PhD Faculty of the Walden University’s Richard W. Riley College of Education and Leadership. Her archaeological research focuses on finger flutings in the Upper Paleolithic and Pleistocene and also on themes of embodiment, children, gender, and light. She is the Director of the Whakatipu Wildlife Trust and lives in Glenorchy, New Zealand.

SCENE MAKERS

205

April Nowell is a Paleolithic archaeologist and Professor and Chair of Anthropology at the University of Victoria. She directs an international team of researchers in the study of Lower and Middle Paleolithic sites in Jordan and collaborates with colleagues on the study of Pleistocene rock art in Australia and France. She is known for her publications on cognitive archaeology, Paleolithic art, the archaeology of children, and the relationship between science, pop culture, and the media. She is the author of Growing Up in the Ice Age: Fossil and Archaeological Evidence of the Lived Lives of Plio-Pleistocene Children (2021) and coeditor of Stone Tools and the Evolution of Human Cognition (2010) with Prof. Iain Davidson, Archaeology of Night: Life After Dark in the Ancient World (2018) with Dr. Nancy Gonlin, and the forthcoming Culturing the Body: Prehistoric Perspectives on Identity and Sociality with Dr. Benjamin Collins.

Note 1. Please note that the dates given are the oldest dates for these sites that often have multiple periods of activity.

References Azéma, M., and F. Rivière. 2012. “Animation in Palaeolithic Art: A Pre-echo of Cinema.” Antiquity 86: 316–24. Bahn, P. 2016. Images of the Ice Age. Oxford: Oxford University Press. Barrett, C.K., and D. T. Case. 2014. “Use of 2D:4D Digit Ratios to Determine Sex.” Journal of Forensic Sciences 59(5): 1315–20. Barrière, C. 1982. L’Art Parietal de Rouffignac: La Grotte aux Cent Mammouths. Paris: Picard. Breuil, H. 1912. “L’age des caverns et roches ornées de France et d’Espagne.” Revue Arch 19: 193–234. Clottes, J. 2010. Cave Art. London: Phaidon. Conkey, M. 2009. “Materiality and Meaning-Making in the Understanding of the Palaeolithic Arts.” In Becoming Human: Innovation in Prehistoric Material and Spiritual Culture, ed. C. Morley and C. Renfrew. Cooney, J. 2013. “The Child in the Cave: The Contribution of Non-Adults to the Creation of Cave Art and Community in the Upper Palaeolithic.” PhD dissertation. Cambridge: Cambridge University. Dobrez, L. 2013. “The Perception of Depicted Motion.” Arts 2: 383–446. ———. 2015. “Depicted Motion, Interaction and Causality in Rock Art.” American Indian Rock Art 41: 57–68. ———. 2016. “Theoretical Approaches to Rock Art, with Comments and Reply.” Rock Art Research 33(2): 1–24. Gosden, C., and Y. Marshall. 1999. “The Cultural Biography of Objects.” World Archaeology 31(2): 169–78. Joyce, R. 2012. “Thinking about Pottery Production as Community Practice.” In Potters and Communities of Practice: Glaze Paint and Polychrome Pottery in the American Southwest, A.D. 1250–1700, ed. L. S. Cordell and J. Habicht-Mauche, 149–54. Tucson: Anthropological Papers of the University of Arizona. Koehler, N., L. Simmons, and G. Rhodes. 2004. “How Well Does Second-to-Fourth Digit Ratio in Hands Correlate with Other Indications of Masculinity in Males?” Proceedings of the Royal Society of London. B (Suppl) 271: S296–8. Lenssen-Erz, T. 1992. “Coherence—a Constituent of ‘Scenes’ in Rock Art: The Transformation of Linguistic Analytical Models for the Study of Rock Paintings in Namibia, with Comments and Reply.” Rock Art Research 9(2): 87–105. Leroi-Gourhan. A. 1958. “I,a Fonction des Signes dans les Sanctuaires Paleolithiques.” Bulletin de la Societe Prehistoriques Francaise 55: 307–21. Lewis-Williams, D. 2002. The Mind in the Cave: Consciousness and the Origins of Art. London: Thames and Hudson. Lorblanchet, M. 1992. “Finger Markings in Peche Merle and Their Place in Prehistoric Art.” In Rock Art in the Old World, ed. M. Lorblanchet, 451–90. New Delhi: Indira Gandhi National Centre for the Arts. Manning, J. T., D. Scutt, J. Wilson, and D. Lewis-Jones. 1998. “The Ratio of 2nd to 4th Digit Length: A Predictor of Sperm Numbers and Concentrations of Testosterone, Luteinizing Hormone and Estrogen.” Human Reproduction 13(11): 3000–4. Marshack, A. 1977. “The Meander as a System: The Analysis and Recognition of Iconographic Units in Upper Palaeolithic Compositions.” In Form in Indigenous Art: Schematization in the Art of Aboriginal Australia and Prehistoric Europe, Prehistory and Material Culture Series, ed. P. J. Ucko, vol. 13, 286–317. Canberra: Australian Institute of Aboriginal Studies. Mohen, J.-P. 2002. Prehistoric Art. Paris: Éditions Pierre Terrail. Nougier, L.-R. and R. Robert. 1958. The Cave of Rouffignac. Trans. D. Scott. London: George Newnes. Nowell, A. 2015. “Learning to See and Seeing to Learn: Children, Communities of Practice and Pleistocene Visual Cultures.” Cambridge Archaeology Journal 25(4): 889–99. ———. 2017. “Art as Social Practice in the Paleolithic. Contribution to Special issue: Art: Art, Material Culture, Visual Culture, or Something Else.” Cambridge Archaeological Journal 27(4): 599–606. Pacher, M., and A. J. Stuart. 2009. “Extinction Chronology and Palaeobiology of the Cave Bear (Ursus spelaeus).” Boreas 38(2): 189–206.

206

LESLIE VAN GELDER AND APRIL NOWELL

Plassard, J. 1999. Rouffignac: Le Sanctuaire des Mammouths. Paris: Seuil. Sharpe, K,. and L. Van Gelder. 2004. “Children and Paleolithic ‘Art’: Indications From Rouffignac Cave, France.” International Newsletter on Rock Art 38: 9–17. ———. 2005. “Techniques for Studying Finger Flutings.” Society of Primitive Technology Bulletin 30: 68–74. ———. 2006a. “Evidence of Cave Marking by Paleolithic Children.” Antiquity 80(310): 937–47. ———. 2006b. “The Study of Finger Flutings.” Cambridge Archaeological Journal 16(3): 281–95. ———. 2006c. “Finger Flutings in Chamber A1 of Rouffignac Cave, France.” Rock Art Research 23(2): 179–98. ———. 2006d. “A Method for Studying Finger Flutings.” In Exploring the Mind of Ancient Man: Festschrift to Robert G. Bednarik, ed. P. Chenna Reddy. New Delhi, India: Research India Press. ———. 2007. “More about ‘More about Finger Flutings.’” Rock Art Research 24(1): 133–35. ———. 2009. “Paleolithic Finger Flutings as Efficient Communication: Applying Zipf ’s Law to Two Panels in Rouffignac Cave, France.” Semiotica 177: 171–90. ———. 2010a. “Four Forms of Finger Flutings as Seen in Rouffignac Cave, France.” In An Enquiring Mind: Studies in Honor of Alexander Marshack, ed. P. Bahn, 269–85. Oxford: Oxbow Books. ———. 2010b. “Fluted Animals in the Zone of Crevices, Gargas Cave, France.” Proceedings of the Eleventh Congress of the International Federation of Rock Art Organizations, Lisbon, Portugal, 4–9 September 2006. Snow, D. 2013. “Sexual Dimorphism in European Upper Paleolithic Cave Art.” American Antiquity 4: 746–61. Van Gelder, L. 2010. “New Methods and Approaches in the Study of Finger Flutings.” Proceedings from the IFRAO Conference: Pleistocene Art of the World, Tarascon-sur-Ariège and Foix, France, 6–11 September 2010. ———. 2012. “Rouffignac Rock Art: Finger Fluting and the Cave Children of France.” Current World Archaeology 5(2): 28–30. ———. 2015a. “Counting the Children: The Role of Children in the Production of Fingerflutings in Four Upper Paleolithic Caves.” Oxford Journal of Archaeology 34(2): 120–31. ———. 2015b. “Finger Flutings in Koonalda Cave: New Explorations, Discoveries, and Methodological Challenges.” Nineteenth Conference of the International Federation of Rock Art Researchers, Caceres, Spain. 31 August–4 September 2015. ———. 2015c. “Using Finger Flutings as a Method for Examining Interpersonal Relationships among Upper Paleolithic Cave Artists.” Nineteenth Conference of the International Federation of Rock Art Researchers, Caceres, Spain. 31 August–4 September 2015. ———. 2015d. “Group Sizes of Upper Paleolithic Cave Artists.” Eleventh Conference on Hunting and Gathering Societies, Vienna, Austria. 7–11 September 2015. ———. 2015e. “The Role of Children in the Creation of Finger Flutings in Koonalda Cave, South Australia.” Childhood in the Past: An International Journal 8(2): 149–60. ———. 2016a. “Finger Flutings, Tectiforms, and the Audacity of Hope.” Expression 13: 78–86. ———. 2016b. “Evidence of Collaboration among Art Makers in Twelve Upper Paleolithic Caves.” In Paleoart and Materiality: The Scientific Study of Rock Art, ed. R. Bednarik, D. Fiore, and Basile, 195–203. Oxford: Archeopress. Van Gelder, L., and K. Sharpe. 2009. “Women and Girls as Upper Paleolithic Cave ‘Artists:’ Deciphering the Sexes of Finger Fluters in Rouffignac Cave.” Oxford Journal of Archaeology 28(4): 323–333. Von Petzinger, G. 2016. The First Signs. New York: Atria Books

==== •

•

•

14. MAPS IN PREHISTORIC ART Pilar Utrilla, Carlos Mazo, Rafael Domingo, and Manuel Bea

Introduction When Iain Davidson proposed that we should write a chapter on “Maps in Prehistory,” we spent a long time thinking about it. Indeed, it is not an easy topic to address. First of all, we have to try to define what a prehistoric map is, as well as try to decipher and identify prehistoric representations that may be interpreted as maps, according to current criteria. We have established some starting points by noting the possibilities that (1) some “maps” cannot in fact be interpreted as maps according to our modern concept of maps, (2) the examples we describe in this chapter may not have had the same (or similar) attributions and significance that we accord to maps nowadays, and last but not least (3) the representations we define as “maps” may not have been maps originally. Hence, in this chapter we consider three categories, all of them theoretical, in order to classify the reliability of prehistoric maps. We will consider some of the most significant examples. “Maps are graphic representations that facilitate a spatial understanding of things, concepts, conditions, processes, or events in the human world” (Woodward and Lewis 1987). In traditional societies, maps usually lack elements inherent to modern cartography. Their geometry is based on topology, where concepts such as linearity, center and periphery, contiguity, and connectedness are far more important than accuracy based on coordinated positions in an abstract space. When describing a river, a hunter-gatherer would give more importance to the direction of its flow (downstream, upstream) and what derives from this (higher lands upstream, lower terrain downstream) than to its length, which greatly depends on the intricacy of its meanders. Unlike other types of data such as temporally structured information (for example, speech or music), spatial information cannot be transmitted in a sequential mode. Speech and music are ephemeral, but when graphic forms emerged as a common way of communicating information (at least forty thousand years ago), they had the advantage of being more permanent. For us there are clear advantages in this type of systematic transmission of information, but for many millennia mapmaking must have been an almost unconscious form of graphic expression. Thus, ancient non-formal maps are barely recognizable to modern researchers but must have been very obvious to those who made them and to the group or groups that shared the same conceptual structures concerning spatial representations. Many researchers have addressed the possibility that there were maps in prehistory. In Eastern Europe there are documented cases of topographic representations of the immediate surroundings of several sites (Klíma 1991; Marshack 1979; Svoboda 1997, 2007, 2017; Züchner 1988, 1996). Kozlowski (1992) lists possible maps from Kiev-Kirillovskaya and Mezhirich in Ukraine, as well as those from Pavlov in Moravia, showing mountains and the meanders of a river, and from Dolni Vestonice, with engraved small arches suggesting hut representations. In contrast, some of the cases mentioned in Western Europe, such as the supposed scenery of marshlands reflected in a piece from El Pendo Cave (Cantabria, Spain), were severely criticized by Barandiarán (1993). An engraved rib from the Middle Magdalenian epoch found in the Llonín Cave (Asturias, Spain) (Fortea, de la Rasilla, and Rodríguez 1992) could represent another example of landscape features in portable art. According to Züchner (1996), the engraving shows a mountain crest surrounded by a river and animals living in that area. Romanelli (Southern Italy) is also especially informative. Besides animals, there are long meandering lines that Marshack (1977) interpreted as watercourses or as life-giving water and also provided more examples from Magdalenian layers found in Parpalló. Züchner studied several examples of “Occidental” and Megalithic art where personal properties, fields, and villages are shown with

208

PILAR UTRILLA, CARLOS MAZO, RAFAEL DOMINGO, AND MANUEL BEA

some degree of complexity at sites such as Mont Bego, Valcamonica, or Antelas. Fields, watercourses, valleys, hills, and a large building seem to be recognizable (Delano-Smith 1987; Züchner 1996). Engravings in Copt Howe (Cumbria, England) have been interpreted as a scene reproducing a real sunset landscape with the mountain and the sun during the solstice (Sharpe 2009: 167–68, Figures 14 and 15). Despite this impressive list, not all examples have the same degree of credibility, as we will see below.

Case Studies Quite a large number of dwellings, huts, tents, and fenced areas, or livestock enclosure representations have been found in different rock art regions and chrono-cultural contexts: Cedarberg (South Africa), shelter IIE21 on Bhonrawali Hill (India), Puntal del Tío Garrillas (Spain), Rujum Hani (Jordan), I-N-Eten in Tassili n’Ajjer (Algeria). Nevertheless, they are not always related to other representations of the environment that could allow us to consider them real representations of a landscape or cartographic figurations of reality. In light of this, this chapter focuses only on those prehistoric examples in which a natural or humanized landscape may have been represented. We first considered what criteria should be used to define what a map is or can be. We envisaged a classification in three categories, depending on the degree of credibility of its reading as a map: 1. High credibility: the depicted motifs can be recognized in a current, real landscape in an environment more or less nearby. This is the case of Block 1 of Abauntz, where the surrounding landscape is reproduced (Utrilla et al. 2009), or the representation of the Neolithic village of Çatalhöyük, where the Hasan Dag volcano was painted during an eruption (Mellaart 1967). 2. Medium credibility: one or more geographical elements concur, but not the entire representation. This is the case of Gargas (Barrière 1976; Utrilla et al. 2007–2008), where a meander of the Garonne River and part of the course of the Neste River were represented close to their confluence, together with a varied group of ungulates. Also, Panel XV of Fuente del Trucho (Utrilla and Bea 2015) might represent the orography of the central paths of the Pyrenees, together with other features (hand stencils, horses, and trefoil signs). 3. Speculative claims: the landscape is not clearly recognizable and the possible map requires a high degree of interpretation. This is the case of the diaphysis from La Rochette (Delporte 1962) that might represent the theoretical plan of a cave much like the “feline gallery” of the nearby cave of Lascaux; or the mammoth tusk from Pavlov (Svoboda 2017), where it seems possible to identify rivers and mountains despite there being no clear similarity with those in its nearby surroundings; or the group of motifs that can be identified as tents and huts represented in Mezhirich (Kozlowski 1992). Much speculation is also needed to interpret two Venus figurines from Prˇedmostí as mammoth hunting scenes or the coastline of the “map” of Tito Bustillo indicating inhabited caves. Interpretations referred to in this category should be regarded as probable cognitive distortions produced by an abstract selection of details (Bednarik 2016). Such readings are deeply influenced by subjectivity: a different group of researchers can “see” different motifs. We also thought about the presumed reasons that the prehistoric people could have had to undertake the abstraction effort necessary to represent the real landscape in a simplified image: 1. To mark the possession of territories for hunting or raw materials (Abauntz, Gargas, Kesslerloch). Ethnology provides us with rich evidence of how recent hunter-gatherers produce simple but practical “maps” and give names to geographical features in the landscape (Svoboda 2007). 2. To establish mobility routes such as the roads in Romanelli, Kesslerloch, or Fuente del Trucho, or access to agricultural plots at Hayonin, Valcamonica, or Mont Bego. 3. To record a special event: for example, an eruption of the Çatalhöyük volcano, the freezing of fingers in the central passes of the Pyrenees (Fuente del Trucho, Gargas), or the great mammoth hunt of Predmost. 4. To document a scale map of the surrounding landscape, either the plan of a cave or a settlement (La Rochette, Mezirich, Pavlov) or even a wider territory (Abauntz, Tito Bustillo, La Laja Alta).

MAPS IN PREHISTORIC ART

FIGURE 14.1. Map from Block 1of the cave of Abauntz and surrounding landscapes.

209

210

PILAR UTRILLA, CARLOS MAZO, RAFAEL DOMINGO, AND MANUEL BEA

Some of the maps could respond to two or more motivations, but we will prefer the one that seems most important with respect to the ensemble of images represented. According to this, we will present the following cases.

Possession Abauntz Cave The cave of Abauntz is located in Arraitz (Navarra, Spain), on the southern side of the Pyrenees, not far from the numerous Magdalenian sites of the Cantabrian coast and the Isturitz Cave. A Late Magdalenian level (dated to ca. 13,700 cal BP) yielded three figuratively engraved portable art blocks (Utrilla et al. 2004). Block 1 has been recognized as a map (Utrilla et al. 2007–2008, 2009), and new, complementary readings have recently added the identification of possible tents and huts within the landscape (Utrilla and Mazo 2011). The occupation features (scarce toolkit, supposed small band of hunters) could be compared to Binford’s proposal for Nunamiut hunters at the Mask Site, who spent time monitoring the movement of game and engaged in activities such as tool making to reduce boredom (Binford 1980). A river passing by a narrow path through the mountains is engraved on the Block 1 surface (Figure 14.1). The engraved lines seem to reproduce the twisting course of the Zaldazaín River, crossing the upper part of Face A of the block and also representing the tributaries along its route. The tangled engraved lines have been interpreted as images of ponds or puddles. The contemporary landscape is surprisingly similar, with the mountain in front of the cave, the Zaldazaín River running at its feet in a narrow canyon, and flat flooded areas to the south. Different animal species were represented in the different environments: ibex on the mountains (becoming smaller and more schematic the farther they recede from the foreground), deer on the flatlands, and reindeer or bovids in the meadows. Some other signs (short parallel traces, pointed and semicircular signs) could be interpreted as paths, fords, or even bridges across the river; six ellipsoidal signs in spiral form, some of them filling hollows, complete the representation. Could they mark the location of shelters, caves, springs, or raw materials? (Utrilla et al. 2007–2008). In sum, face A of Block 1 can be seen as a sketch that synthesizes most of the natural milestones of the environment on a small surface. Simplification, optimization, and standardization of the elements enable the provision of quite a lot of information. A living space was categorized and constructed by combining a “linear code” and the relief of the block itself. The following discussion is based on the identification of the art outlined in the publications cited above. Was Block 1 a sort of graphic story occurring in a real natural landscape? Perhaps it is a real map that could be used as a reminder for another visit to the area. The idea that this is a story seems plausible. After the daily hunting activities, a small group of people would meet in the cave to relate the adventures of the day, which the artist would capture on a block of stone, later abandoned in situ. Was it left on purpose so that later visitors would “read” it and learn about the “feat” (the hunting success)? None of the animal motifs appear to be injured by arrows or wounded. In this context, as has previously been pointed out, a physical landscape becomes a human landscape by the addition of storied, mythological, and utilized environments. The possibility that Block 1 functioned as a topographical sketch providing information on biotic and/or abiotic resources becomes more plausible if we give greater weight to certain “minor” elements such as dotted lines, ogives or “scaleriforms.” It could thus be seen as an outline or map with interesting areas marked by the author(s), that is, as a graphical representation of a real landscape that could have served as a reminder of previous visits to the area, according to its schematic representation of the territory, without strict accuracy or selective details (Utrilla et al. 2007–2008: 243).

Gargas In the famous cave of Gargas (Aventignan, France) (Barrière 1976; Foucher et al. 2012), located near the confluence of the Neste and Garonne rivers, the ceiling of the section called the “Chambre du Camarin” (an important sanctuary during the Gravettian, ca. 27,000 cal BP) is entirely decorated with animals. Together with the animal motifs (especially horses, bison, aurochs, ibex, mammoth, and some other minor species such as deer, a possible duck, and even mixed or hybrid animals) there are some carefully engraved curvi-

MAPS IN PREHISTORIC ART

211

linear multiple traces very similar to those defined as rivers, already referred to in connection with Block 1 of Abauntz. According to Marshack (1977), it seems probable that these traces represent watercourses or paths, but they may not necessarily depict real landscape features, possibly instead showing mythological rivers or routes (e.g., the route by which a shaman takes people into another world, the origin of a person or a tribe arriving in the territory, the river of the underworld, the birth-giving nature of water itself). Nevertheless, two sections of one of the snake-like traces describe a well-defined meander very similar to that of the Garonne River itself, as well as part of the course of the Neste (Figure 14.2), whose confluence is near the cave. Thus it may be that the panel of engravings at least partially represents a natural landscape arranged around the confluence of the rivers Garonne and Neste, with animals occupying their natural ecosystem. a

b

c

FIGURE 14.2a–c. (a) Tracing of the palimpsest of the cave of Gargas (after Barrière 1976). (b) Selected details of the Gargas palimpsest. (c) Map of the real course of the Garonne and Neste next to the location of the cave (OpenStreetMaps).

212

PILAR UTRILLA, CARLOS MAZO, RAFAEL DOMINGO, AND MANUEL BEA

Meander-like shapes and multiple tracings are perfect conventions for representing watercourses. The tracings’ (flutings’?) similarity to the actual outline of the rivers near the cave supports the view that the panel is a representation of the environment surrounding the cave. A similar case is the figure represented in the two decorated bone plaques from the Kesslerloch cave (Switzerland), interpreted as a map of the surrounding district (Rödiger 1891). Neither of the items in question was among those later discovered to have been faked, as was the case for some other objects from Kesslerloch now considered forgeries (Delano-Smith 1987).

Mobility Routes Fuente del Trucho The Cave of Fuente del Trucho in the Central Pre-Pyrenees has been defined as a major site of Paleolithic art, with two main decorated areas—engravings at the cave entrance and paintings inside the cave—featuring diverse subject matter, including some animals (horses, ibex), signs, series of dots, and hand stencils that are ubiquitous throughout the cave. Most of these images have been described recently (Utrilla et al. 2013; Utrilla and Bea 2015). U/Th dates have been obtained (Hoffmann et al. 2017) for calcite crusts covering the paintings: those considered here date to the Gravettian period, roughly contemporary with the paintings discovered in the Gargas cave (ca. 27,000 BP). Panel XV, with its complex series of dots, appears to be associated with three black and two red hand stencils that seem to form radial and curvilinear geometric patterns. All of them are clearly visible in the center of the ceiling. The dots have been interpreted as a starry sky (Beltrán 1993) as well as a sort of guide marking routes through the Pyrenean passes (Utrilla 2005). There are in fact surprising similarities between the passage seen on a real map of the Pyrenees and the particular dispersion of curvilinear and radial dot series (Panel XV) depicted in the cave (Figure 14.3). If we accept this interpretation, Paleolithic artists may have represented a schematic map in which the departure and arrival points on both sides of the Pyrenees are represented by the hand stencils (the caves of Gargas and Fuente del Trucho are 110 km apart as the crow flies), and the complex dot series could be seen as different geographical features as well, perhaps the routes they followed. The peculiar hand stencils with some fingers missing painted in Gargas and Fuente del Trucho could have been a reminder of a traumatic event: perhaps some members of the group (younger ones, since some of the small hand stencils in Fuente del Trucho seem to belong to children around four years old) lost some of their fingers due to frostbite when passing the glaciated mountain passes of the Pyrenees (Utrilla and Bea 2015). Interestingly, some of the chert tools found at Fuente del Trucho were made of Northern Pyrenean lithic raw materials (Sánchez de la Torre, pers. com.)

Hayonim In 1997, a limestone slab from a Natufian occupation was found at Locus 11 in Hayonim Cave in Israel. Like other findings from the same site and other neighboring spots (i.e., Kebara Cave), it was decorated with a complex series of engraved lines that formed irregular geometric patterns (Bar-Yosef and BelferCohen 1999). The motifs were rectangular areas comprising clusters of lines arranged at both sides of a ladder-like design running along the center of the slab. The left ensemble of clusters included horizontal and vertical lines separated by horizontal engravings, while the right one comprised only vertical lines separated by a meandering pattern. The motifs were interpreted as a schematized representation of space, the different clusters being “fields” similar to those viewed on the mammoth tusk found at Pavlov (Svoboda 1997). Bar-Yosef and Belfer-Cohen (1999) stressed the existence of a notion of territoriality among the Natufians and the fact that they were the first to cultivate fields near the inhabited sites. The engraved slab could thus be interpreted as a map of particular fields located not far from the site. The authors recognize the impossibility of locating the outlines of such ancient fields after ten millennia of erosion and land use. The same argument would apply to the examples from Mont Bego or Valcamonica, specifically the “maps” from Seradina-Bedolina interpreted as “topographical type geometric compositions” (Marretta 2010: 152 and 157; Züchner 1996, Figure 8).

MAPS IN PREHISTORIC ART

213

FIGURE 14.3a–b. (a) Satellite image of the Pyrenees and location of the caves (Gargas and Fuente del Trucho) (Wikimedia Commons). (b) Tracing of Panel XV. Hand stencils could mark locations of caves, while dot series would represent passages or routes through the Pyrenees.

214

PILAR UTRILLA, CARLOS MAZO, RAFAEL DOMINGO, AND MANUEL BEA

Special Events Çatalhöyük Çatalhöyük is a well-known Neolithic village in Anatolia whose occupation spanned from ca. 9400 to 8600 cal BP. Since the late 1960s, the famous painting on the wall of one of the domestic shrines of Çatalhöyük has been interpreted as the oldest map in the world (Delano-Smith 1987). Such claims can still be found in the literature (Schmitt et al. 2014), despite the well-known publication of the map of Abauntz (Utrilla et al. 2009), which predates the Anatolian example by six millennia (Clarke 2013). The celebrated Anatolian picture is also of interest to geologists, as it has been said to represent the volcano Hasan Dag during an eruption (Mellaart 1967), an interpretation that could be true, according to some scholars (Forest 2003: 50). Nevertheless, the original excavator, Mellaart, thought at first that the image represented a leopard skin and a dense pattern of quadrangular elements, also depicted elsewhere in the urban complex (Meece et al. 2006). Not until some years after its discovery did he interpret the painting as an aerial view of the village of Çatalhöyük (the geometric patterns) with a distant view of the Hasan Dag volcano (which is actually 130 km northeast of the site) during an eruption. The leopard skin interpretation, actively supported by Meece et al. (2006), is based on frequent representations of leopards (in wall paintings as well as large and portable sculptures) not only in Çatalhöyük but also elsewhere in Neolithic Anatolia. These authors also suggest that the interpretation of the geometric figures as a plan of the city could only come from an archaeologist, as it is strikingly similar to a bird’s eye, two-dimensional, roof-removed representation of an urban archaeological site; however, this interpretation is very different from what a Neolithic inhabitant of Çatalhöyük would have depicted—most likely a groundlevel view. Meece et al. (2006) also criticize the volcano figure, explaining that the Hasan Dag indeed has two peaks as shown in the wall painting, but when seen from the direction of Çatalhöyük they are reversed: the higher being on the left and the smaller on the right. Other nearby volcanoes do not fit the depicted outline. It had been suggested that the obsidian found at Çatalhöyük was collected in the Hasan Dag area, but more recently this has been refuted (Carter et al. 2006). Nevertheless, a recent paper (Schmitt et al. 2014) has revived the interpretation of the painting as a map. An eruption has been dated by U-Th/He zircon geochronology to 8970±640 BP, and the wall painting belongs to an archaeological level roughly dated to 8600 cal BP. The authors propose that such a violent episode would have made a deep impression on the inhabitants of nearby regions, meriting its depiction in one of the shrines of Çatalhöyük. A possible compromise could conjoin the two interpretations (volcano versus leopard) by combining their natural attributes: the force and the noise of the volcano during its eruption and the strength and roar of the leopard. This could unify two important symbols in one.

Mezhirich Eastern Europe offers some examples of graphic representations traditionally defined as maps. Roughly twenty Paleolithic open-air sites have been recorded in the Dnieper and Desna basins. The most renowned of these sites is Mezhirich (Ukraine), located on a promontory controlling the confluence of the Ros and Rosava rivers. Since 1966, several archaeological excavations have provided some exceptional remains, among them four groupings of mammoth bones sometimes interpreted as huts (Soffer et al. 1997). A recent series of AMS dates on bones from the huts (Marquer et al. 2012) shows that the site was occupied between 14,850 and 14,315 cal BP. Some representations found at this site have been explained as topographic motifs describing the surroundings. This is the case with some of the portable engravings seen on mammoth tusks and ivory plaques found in Kiev-Kirillovskaya, especially those discovered in Mezhirich (Kozlowski 1992). Pidoplichcko (1969) described the plaque of Structure 1 in Mezhirich as a map of a settlement with a group of huts beside a riverbank in the mountains. Iakovleva (1985) referred to four striped motifs regarded as schematic huts in order to support her ideas about the exterior decoration of the Mezhirich huts. Thus, the distribution of the different structural elements would have been made not only for technical or functional purposes, but also for aesthetic reasons comprising two decorative motifs: zigzags and parallel lines (cited in Kozlowski 1992).

MAPS IN PREHISTORIC ART

215

Marshack (1979) revised the ensemble of ivory plaques from the Russian Plateau and posed several questions. He argued that at their correct orientation, with a 180º vertical rotation, the shapes described by some as huts seem to represent twin peaks with an astronomical body between them, despite the fact that such features are not present in a territory characterized by soft hills and valleys.

La Rochette In Western Europe, Hauser (1911) discovered in La Rochette rock shelter (Saint-Léon-sur-Vézère, France), a site with a long stratigraphy (Delporte 1962) containing a decorated bone whose provenience was thought to be at an (undated) Upper Aurignacian level, or—because the Perigordian was then unknown—a transitional level (Upper Aurignacian-Solutrean), as Hauser mentions (Beinhauer 1986). The engraved surface shows five successive circular lobes separated by four narrowing passes preceded by a sort of access corridor. The inner part of the shape is filled with eight longitudinal lines. Each lobe has a different fill-pattern, the first partially longitudinal and the third and fourth partially transversal. The fill of the second lobe is lost. A wavy line encloses the piece on its left side, while on the right there is a rectilinear line with fringes. Beinhauer (1986) interprets the motif as a map of a cave with a watercourse running nearby to the left (or maybe a snake) and a cliff on the right side. Such a map would not represent the La Rochette rock shelter itself, but nearby there are several caves with a morphology that is consistent with the engraved lines. The cave of Lascaux (8 km from La Rochette) has a similar topography in the Gallery of Mondmilch and its continuation in the Diverticule des félins (Figure 14.4). Indeed, the plan of Lascaux, in its final “diverticule,” is divided into five circular lobes separated by four narrowing passes like those represented on the decorated bone from La Rochette. Beinhauer had already pointed out this similarity, emphasizing the proximity of the two caves. This is a very suggestive hypothesis that we have referred to previously (Utrilla et al. 2007–2008). It is not possible to establish a link between the time when the La Rochette piece was engraved (Aurignacian) and the age when Lascaux was decorated (Solutrean, Lower Magdalenian), but of course the cave already existed during the Aurignacian. The bestiary represented in that area of Lascaux (felines) is very suggestive, this species being closely related to Aurignacian art (e.g., Chauvet, Vogelherd, or Hohlenstein-Stadel), although the feline motif is also present in Magdalenian caves in the same area (e.g., Combarelles I).

Tito Bustillo The absence of identifiable details on a decorated panel makes a non-speculative interpretation more difficult. This is the case with a possible cartographic representation found in the cave of Tito Bustillo (Ribadesella, Spain) (Ruiz 2007). According to this interpretation, the map is of the Sella River basin: the signs (a double parallel line and groups of fingertips or short lines in groups of two or three), depicted on a boulder, could correspond to the real location of the Sella basin itself (long parallel lines) and various caves in their correct location (Cuevona, Cova Rosa, San Antonio, Pedral de Arra, Cobayu, Tinganón). Furthermore, the upper border of the boulder is reminiscent of the coastal line of the territory. Balbín (2006) disputed this interpretation by pointing out that the signs on the stone do not reflect a map of the exterior geography of the cave.1 Meanwhile, an important hole in this argument concerns the real coastline during the Upper Paleolithic period, which was no less than 6 km away from the current one. Taking this into account and accepting the argument that the image was a map, some of the signs (defined as caves or sites) could correspond to sites nowadays submerged. The current sinuous progression of the Sella basin would not have been well traced on the decorated panel (Utrilla et al. 2007–2008). Although we know that the the Pleniglacial delimitation of the coastline was relatively far from the current one (see www.azti.es) and the Sella River nowadays presents a more sinuous development (due to modification of the slope), it is still surprising that the artists in Tito Bustillo would have chosen to place such a large freestanding block in a prominent place at the end of the gallery, depicting dots and lines in places that seem to coincide “grosso modo” with the location of the main decorated caves. In this case, its definition as a probable map (Ruiz 2007) could be taken into account (Figure 14.5).

216

PILAR UTRILLA, CARLOS MAZO, RAFAEL DOMINGO, AND MANUEL BEA

FIGURE 14.4a–c. (a) General plan of the Lascaux Cave (from Leroi-Gourhan 1984: 181); (b) Detailed plan of the “Cabinet des felides” of Lascaux Cave (modified after Leroi-Gourhan and Allain 1979: 61); (c) Tracing of the decorated bone of La Rochette (after Beinhauer, 1986: 142).

La Laja Alta The post-Palaeolithic group of La Laja Alta (La Janda, Cádiz, Spain) could be similarly interpreted as a map. La Laja Alta is an open-air rock art shelter in which several boat motifs are depicted, along with other abstract or schematic elements. Besides the chrono-cultural filiation and interpretation of the boats (Barroso 1980; Giles and Sáenz 1978), a new reading of this imagery and its distribution on the panel can be suggested. A possible interpretation as a harbor was initially pointed out (maybe Carteia) (Giles and Sáenz 1978), but it is possible to go further. By superimposing the rock art motifs on a map of the coastline, it is possible to see a reasonably coherent distribution of the motifs (and their identification) on the map, showing the boats in the bottom part of the scene (on the sea) while the rest of the figures are arranged on the coast line, as if representing landmarks like villages, towns, or ports.

MAPS IN PREHISTORIC ART

217

FIGURE 14.5a–b. (a) (top). Location of the decorated block in Tito Bustillo. (b) (bottom). Detail of the block and dots and lines depicted on it. Their location on the block’s surface seems to have a distribution pattern quite similar to the one observed for major decorated caves in the territory (after Ruiz 2007).

218

PILAR UTRILLA, CARLOS MAZO, RAFAEL DOMINGO, AND MANUEL BEA

Discussion Geographical information is communicated by animals, for example, by the scent marking of territories. The information requires that both the animal that delimits the space and the potential receivers of information be in the same territory, at least within a short time interval. Humans have marked their claims to terrains in a similar way. Boundaries between fields, properties, or countries have traditionally been signaled with stones, signs, or barriers. These markers are more permanent than mere scent, but their significance is very similar. The hunter-gatherer way of life also involved spatial frames related to processes, changes, and events that were essential to those people’s survival or potential sources of danger . The cooperative strategies followed by hunter-gatherers improved the development of language, and sharing spatial information was undoubtedly a basic element of their daily interrelations (Lewis 1987). Apart from a strictly geographical point of view, prehistoric territory can also be regarded as being integrated (and understood) in a symbolic landscape. As Bahn (2014) wrote, “we can never really know the motivations behind any prehistoric art.” In any case, when following geographical approaches relating to human groups, it may be valuable to differentiate between “environment” and “landscape,” so it would be interesting to try using different methods to establish how a territory becomes an appropriated landscape for humans. As already pointed out, features of a landscape or site were associated with particular myths, legends, or events; traditional or tribal territories; sacred or holy areas; taboos; and so on (Bahn 2014), with maps being an explanation of some rock art sites (Mountford 1956, 1961). It could be that mapping, such as that identified at Abauntz, involved the transformation of a territory—representing an effective use and control of lands and resources or places where different activities would take place (Eastham and Eastham 1991)—into a landscape, that is, a place rich in meaning and experience (Zedeño 2009: 215). As the Old World prehistoric past is far too remote for us to infer data from traditions or legends, ethnographic parallels could be a good source of information. First, graphic manifestations produced in ethnographic contexts are closely linked to a precise social context, and as such should be studied from a holistic perspective rather than in isolation (this being very common in prehistoric rock art revisions). Common repeated features should be the basis of our studies, in order to extract the most important concepts and deduce how they have been translated into images. Secondly, ethnographic rock art rarely seems to be devoted to daily life: its main purpose is not obtaining food by means of sympathetic magic. Besides, most huntergatherer groups do not need a map to help them find their way: only societies living in monotonous environments such as the open ocean or wide plains require directional aids. The use of symbols in early maps poses another problem for modern scholars: in indigenous groups, the limited variety of symbols is easily learned, as they arise from shared experiences. Modern maps also employ symbols, but they are frequently explained in written descriptions, which is not the case with prehistoric maps or in oral societies. For this reason, the recognition of true prehistoric maps is no easy task (in fact, it may almost be impossible). They are often identified as such by spontaneous recognition (“it looks like a map”) (Delano-Smith 1987), but this is based on the researcher’s experience with modern maps. This compels us to ask what exactly a map is, and moreover, what prehistoric maps were intended for. Spontaneous recognition implies that the researcher recognizes figures as a whole, assuming they are contemporary, but cumulative scenes in prehistoric arts are scarce and usually difficult to interpret. Figures appear superimposed, in different angles and sizes, and in Paleolithic times they almost always lack a frame other than the limits of the object or the panel on which they are represented. Palimpsests may have been created on purpose, in a single art session, or may be the result of many different unrelated interventions. Similarities in the technique employed, the ductus, or the style could help to isolate groups of figures and look for cartographic meanings in their spatial representation. In order to read palimpsests as cartographic images, we must look for some recognizable naturalistic symbols, like animals or dwelling structures, that may have spatial significance. Besides, repeated figures could occur, as in modern maps where some symbolic images appear many times. Following these criteria, scholars have, with difficulty, identified some supposed prehistoric maps that represent the landscape from above. They always represented small areas, perhaps due to a minor degree of abstraction in their makers’ spatial thinking. Their intended purpose is difficult to decipher, but researchers (Delano-Smith 1987) agree that they were not mere cartographic depictions, as we have highlighted above. In an ethnographic context, maps have a deeper, frequently symbolic meaning.

MAPS IN PREHISTORIC ART

219

Conclusions As Eastham and Eastham (1991) wrote, “there has been little systematic application of hypothesis to Paleolithic images”—an observation that can be applied to the whole of prehistoric art. In any case, significance can vary from one case to another. Representation of the environment (more or less close to reality) could have had different purposes in different moments, regions, and cultures. Prehistoric groups’ concepts of space and territory would very likely have been different from ours. That is why only those compositions where we can discern some geographical elements coinciding with current features can be defined, tentatively, as “maps.” We will never be able to know whether one decorated panel or another may also have been intended to represent a landscape or a real environment addressing features relevant to the prehistoric artist (which would not be the same for us). Narratives, mythical events, and historical facts may also be represented within organized spaces on decorated caves or objects, all of them with specific meanings, not only cartographical or ecological. Places of memory (whether social and/or symbolic territories) should also be taken into account (Van Dyke 2008). Whatever their significance may have been, “maps” provide us with additional information we can use to approach certain social aspects of society. The graphic representation of a society’s surroundings and natural environmen humanizes a space. By engraving our cultural concepts, we can assert our territorial or property rights over a territory that we are finally able to comprehend.

Acknowledgements This study has been funded by the project “Gaps and sites. Gaps and occupations in the Prehistory of the Ebro Basin” (HAR2017-85093-P) and the Research Group “P3A. First settlers and Archaeological Heritage of the Ebro Basin” (H14_20R, Government of Aragón and European Social Fund). The authors belong to the University Institute of Environmental Sciences (IUCA) (C. Mazo) and to the University Institute of Heritage and Humanities (IPH) (P. Utrilla, R. Domingo and M. Bea), University of Zaragoza (Spain). R. Domingo has been a Ramón y Cajal Research Fellow (ref: RyC2013-12613, Spanish Ministry of Economy and Innovation). Pilar Utrilla, full professor at the University of Zaragoza (Spain), focuses her line of research on Cantabrian Magdalenian hunter-gatherers of the Ebro Basin, Mesolithic-Neolithic transition, and rock art (Paeolithic and post-Paleolithic). She has led nine National Research Projects (Ministry of Science and Innovation) as well as the PPVE Research Group, published more than 250 scientific papers and books, and supervised 15 PhDs. She is a member of many scientific journal editorial boards, an advisor for the CNEAI (responsible for the History and Art commission), and chairwoman of the UISPP’s 8th Commission. Her high work capacity has allowed her to lead more than forty archaeological campaigns from the Middle Palaeolithic to the Neolithic. Carlos Mazo is full professor at the University of Zaragoza (Spain). His PhD (1989) was focused on the functionality of lithic tools from the Magdalenian levels of the cave of Abauntz. He has directed many archaeological campaigns (prospection and excavation) linked to the Paleolithic and Epipaleolithic-Neolithic transition in the Middle Ebro basin and has participated in more than 150 scientific publications. Currently, his main line of research is focused on the final presence of Neanderthals and Upper Paleolithic transition in the Iberian System. Rafael Domingo obtained his PhD in Prehistory in 2003. He was a postdoctoral fellow in the CNRS (Toulouse, France) from 2003 to 2005. He has been a senior researcher “Ramón y Cajal” at the University of Zaragoza from 2014 to 2019, where since then he is a Lecturer. His main research is devoted to the reconstruction (by means of archaeological fieldwork and functional analyses of lithic tools) of the prehistoric societies in northeastern Iberia from the Middle Paleolithic to the Neolithic. He has published more than a hundred papers, book chapters, and books, and has contributed to dozens of international and national congresses.

220

PILAR UTRILLA, CARLOS MAZO, RAFAEL DOMINGO, AND MANUEL BEA

Manuel Bea earned a doctorate in History (Prehistory) focused on Post-Paleolithic rock art from the University of Zaragoza (2005). He has been postdoctoral research fellow at the University of Zaragoza (Spain), being curretly recognised researcher (ERC-CoGLArcHer project) at the University Jaume I (Castellón) and member of the National Spanish Committee of ICOMOS. His interest in the scientific study of rock art and documentation by means of new technologies is combined with an interest in evaluating the social and economic values of prehistoric societies. His research has been published in numerous national and international papers, book chapters, and books. He has been the scientific director of many archaeological projects as well as collaborator in international projects in Europe, the United States, and the Middle East.

Note 1. After Balbín’s statement to the journal “El Comercio” (December 16, 2006, page 76).

References Bahn, P. G. 2014. Prehistoric Rock Art: Polemics and Progress. Cambridge: University of Cambridge Press. Barandiarán, I. 1993. “El lobo feroz: la vacuidad de un cuento magdaleniense.” Veleia 10: 7–37. Barrière, C. 1976. L’art parietal de la Grotte de Gargas. Oxford: BAR Supplementary Series, 14. Barroso, C. 1980. “Nuevas pinturas rupestres en Jimena de la Frontera (Cádiz: Abrigo de Laja Alta).” Zephyrus 30–31: 23–42. Bar-Yosef, O., and A. Belfer-Cohen. 1999. “Encoding Information: Unique Natufian Objects from Hayonim Cave, Western Galilee, Israel.” Antiquity 73: 402–10. Bednarik, R. G. 2016. “Rock art and Pareidolia.” Rock Art Research 33(2): 167–81. Beinhauer, K. W. 1986. “Eine aussergewöhnliche Knochen-Gravur des Jungpaläolithikums aus La Rochette (Dép. Dordogne) in den Archäologischen Sammlungen des Städtischen Reiss-Museums Mannheim.” Mitteilungen der Anthropologischen Gesellschaft in Wien 116: 141–53. Beltrán, A. 1993. Arte prehistórico en Aragón. Zaragoza, Spain: Ibercaja. Binford, L. 1980. “Willow Smoke and Dog’s Tails: Hunter-Gatherer Settlement Systems and Archaeological Site Formation.” American Antiquity 45(1): 4–20. Carter, T., G. Poupeau, C. Bressy, and N. J. G. Pearce. 2006. “A New Programme of Obsidian Characterization at Çatalhöyük, Turkey.” Journal of Archaeological Sciences 33: 893–909. Clarke, K. C. 2013. “What Is the World’s Oldest Map?” Cartographic Journal 50: 136–43. Delano-Smith, C. D. 1987. “Cartography in the Prehistoric Period in the Old World: Europe, the Middle East, and North Africa.” In The History of Cartography, vol. 1, 54–101. Chicago: University of Chicago Press. Delporte, H. 1962. “Le gisement paléolithique de La Rochette (Commune de Saint-Léon-sur-Vézere, Dordogne).” Gallia Préhistoire 5(1): 1–22. Eastham, M., and A. Eastham. 1991. “Palaeolithic Parietal Art and Its Topographical Context.” Proceedings of the Prehistoric Society 57(1): 115–28. Forest, J. D. 2003. “Çatal Huyuk et son décor. Pour le déchiffrement d’un code symbolique.” In Arts et symboles du Néolithique à la Protohistoire, ed. J. Guilaine, 41–58. Paris, France: Éditions Errance. Fortea, J., M. de la Rasilla, and V. Rodríguez. 1992. “La cueva de Llonín (Llonín, Peñamellera Alta).” In Campañas de 1987 a 1990. Excavaciones Arqueológicas en Asturias 1987-1990, 9–18. Oviedo, Spain: Servicio de Publicaciones del Principado de Asturias. Foucher, P., C. San Juan-Foucher, C. Vercoutère, and C. Ferrier. 2012. “La grotte de Gargas (Hautes-Pyrénées, France): l’apport du contexte archéologique à l’interprétation de l’art pariétal.” Bulletin de la Société Préhistorique Ariège-Pyrénées 65/66: 52–53 and CD 209–25. Giles, F., and A. Sáenz. 1978. “Prehistoria de la Laguna de la Janda: nuevas aportaciones.” Boletín del Museo de Cádiz 1: 7–35. Hauser, O. 1911. Le Périgord préhistorique. Périgueux: Réjou impr. Hoffmann, D., P. Utrilla, M. Bea, A. W. G. Pike, M. Garcíaz-Díez, J. Zilhao, and R. Domingo. 2017. “U-Series Dating Palaeolithic Rock Art at Fuente del Trucho (Aragón, Spain).” Quaternary International 432(part B): 50–58. Iakovleva, L. A. 1985. “Décoration architecturale des cabanes paléolithiques. Construction et architecture.” Budivelni 3: 25–28. Klíma, B. 1991. Die jungpaläolithischen Mammtjäger-Siedlungen Dolní Vestonice und Pavlov in Südmähren. Liestal, Switzerland: CSFR, Archáologie und Museum. Kozlowski, J. 1992. L’art de la préhistoire en Europe Orientale. Paris, France: CNRS.

MAPS IN PREHISTORIC ART

221

Leroi-Gourhan, A. 1979. La stratigraphie et les fouilles de la grotte de Lascaux. Lascaux inconnu XII, supplement à Gallia Préhistoire. Oqaris, France: CNRS. ———. 1984. “Grotte de Lascaux.” In L’art des cavernes. Atlas des grottes ornées paleolithiques françaises, ed. A. Leroi-Gourhan. Paris, France: Ministère de la Culture. Leroi-Gourhan, A. and J. Allain. 1979. Lascaux inconnu. Suppléments à Gallia Préhistoire, 180–200. Paris:. Éditions du CNRS. Lewis, G. M. 1987. “The Origins of Cartography.” In The History of Cartography, ed. J. B. Harley and D. Woodward, 50–53. Chicago: University of Chicago Press. Marquer L., V. Lebreton, T. Otto, H. Valladas, P. Haesaerts, E. Messager, D. Nuzhnyi, and S. Péan. 2012. “Charcoal Scarcity in Epigravettian Settlements with Mammoth Bone Dwellings: The Taphonomic Evidence from Mezhyrich (Ukraine).” Journal of Archaeological Science 39: 109–20. Marretta, A. 2010. “Il parco archeologico comunale di Seradina-Bedolina.” In I parchi con arte rupestre della Valcamonica, eds. Marreta, A. and T. Cittadini, 138–59. Capo di Ponte: CCSP. Marshack, A. 1977. “The Meander as a System: The Analysis and Recognition of Iconographic Units in Upper Palaeolithic Compositions.” In Form in Indigenous Art: Schematisation in the Art of Aboriginal Australia and Prehistoric Europe, ed. P. J. Ucko, 286–317. Canberra, Australia: Australian Institute for Aboriginal Studies. ———. 1979. “Upper Palaeolithic Symbol Systems of the Russian Plain: Cognitive and Comparative Analysis.” Current Anthropology 20(2): 271–305. Meece, S. 2006. “A Bird’s Eye View of a Leopard’s Spots: The Çatalhöyük ‘Map’ and the Development of Cartographic Representation in Prehistory.” Anatol. Stud. 56: 1–16. Mellaart, J., and M. Wheeler. 1967. Çatal Hüyük, a Neolithic Town in Anatolia. London,: Thames and Hudson. Mountford, C. P. 1956. Art, Myth and Symbolism. Cambridge: Cambridge University Press. ———. 1961. “The Artist and His Art in Australian Aboriginal Culture.” In The Artist in Tribal Society, ed. M. Smith, 1–14. London: Routledge and Kegan Paul. Pidoplichcko, I. G. 1969. Late Paleolithic Dwellings of Mammoth Bone in the Ukraine. Kiev, Ukraine: Institute of Zoology, Ukrainian Academy of Sciences. Rödiger, F. 1891. “Vorgeschichtliche Kartenzeichnungen inder Schweiz.” Zeitschrift für Ethnologie 23: 237–42. Ruiz, M. 2007. “El mapa rupestre de Ribadesella (Asturis), uno de los primeros en la cronología cartográfica.” Mapping 119: 6–9. Schmitt, A. K., M. Danišík, E. Aydar, E. S¸en, I. Ulusoy, and O. M. Lovera. 2014. “Identifying the Volcanic Eruption Depicted in a Neolithic Painting at Çatalhöyük, Central Anatolia, Turkey.” PLoS One 9: e84711. Sharpe, K. E. 2009. “Rock-art and Rough Outs: Exploring the Sacred and Social Dimensions of Prehistoric Carvings at Cop Howe, Cumbria.” In Art as Metaphor: The Prehistoric Rock-Art of Britain, ed. A. Mazel, G. Nash, and C. Waddington, 151– 73. Oxford: Archaeopress. Soffer, O., J. M. Adovasio, N. L. Kornietz, A. A. Velichko, Y. N. Gribchenko, B. R. Lenz, and V. Y. Suntsov. 1997. “Cultural Stratigraphy at Mezhirich, an Upper Palaeolithic Site in Ukraine with Multiple Occupations.” Antiquity 71: 48–62. Svoboda, J. 1997. “Symbolisme gravettien en Moravie.” Bulletin de la Société Préhistorique de l’Ariège 52: 87–103. ———. 2007. “Spatial Representations in the Upper Paleolithic: The Cases from Pavlov and Prˇedmost.” In Anati, E., and Mohen, J.P. (eds.) Les expressions intelectuelles et spirituelles des peuples sans écriture 123–26. Capo di Monte, Italy: UISPP/ CISENP. ———. 2017. “On Landscapes, Maps and Upper Paleolithic Lifestyles in the Central European Corridor: The Images of Pavlov and Prˇedmost.” Veleia 34: 67–74. Utrilla, P. 2005. “El Arte Rupestre en Aragón. 100 Años después de Calapatá.” In Arte Rupestre en la España Mediterránea, ed. M. S. Hernández and J. Soler, 341–77. Alicante, Spain: MARQ. Utrilla, P., and M. Bea. 2015. “Fuente del Trucho, Huesca (Spain): Reading Interaction in Palaeolithic Art.” In Prehistoric Art as Prehistoric Culture: Studies in Honour of Professor Rodrigo de Balbín Behrmann, ed. P. Bueno and P. Bahn, 69–78. Oxford: Archaeopress. Utrilla, P. and C. Mazo. 2011. “Los cantos pintados de la cueva de Abauntz y algunas nuevas lecturas del bloque 1.” Príncipe de Viana 253: 23–42. Utrilla, P., V. Baldellou, M. Bea, and R. Viñas. 2013. “La cueva de la Fuente del Trucho (Asque-Colungo, Huesca): Una cueva mayor del arte gravetiense.” In Congreso Internacional: El Gravetiense Cantábrico, estado de la cuestión, 526–37. Madrid, Spain: Museo Nacional de Altamira. Utrilla, P., C. Mazo, M. C. Sopena, R. Domingo, and O. Nagore. 2004. “L’art mobilier sur pierre du versant sud des Pyrénées: les blocs gravés de la grotte d’Abauntz.” In L’art du Paléolithique supérieur, ed. M. Lejeune and A. C. Welté, 199–218. Liège, Belgium: Presses Universitaires de Liège. Utrilla, P., C. Mazo, M. C. Sopena, M. Martínez-Bea, and R. Domingo. 2009. “A Palaeolithic Map from 13660 calBP: Block 1 of the Cave of Abauntz (Spain).” Journal of Human Evolution 57: 99–111. Utrilla, P., C. Mazo, M. C. Sopena, R. Domingo, and M. Bea. 2007–2008. “Ríos, montañas y charcas: una representación de paisaje en el bloque 1 de la Cueva de Abauntz.” Veleia, 24/25: 229–60.

222

PILAR UTRILLA, CARLOS MAZO, RAFAEL DOMINGO, AND MANUEL BEA

Van Dyke, R. M. 2008. “Memory, Place, and the Memorialization of Landscape.” In Handbook of Landscape Archaeology, ed. B. David and J. Thomas, 277–84. Walnut Creek, CA: Left Coast Press. Woodward, D., and G. M. Lewis. 1987. “Concluding Remarks.” In The History of Cartography, vol. 2, ed. J. B. Harley and D. Woodward, 537–41. Chicago: University of Chicago Press. Zedeño, M. N. 2009. “The Archaeology of Territory and Territoriality.” In Handbook of Landscape Archaeology, ed. B. David and J. Thomas, 210–17. Walnut Creek, CA: Left Coast Press. Züchner, C. 1988. “Kartenbilder des 3. und 2 Jahrtausends v. Chr.” Mitteilungen der Fränkischen Geographischen Gesellschaft 33/34: 321–44. ———. 1996. “The Scaliform Sign of Altamira and the Origin of Maps in Prehistoric Europe.” In El Hombre Fósil 80 años después. Homenaje a Hugo Obermaier, ed. A. Moure, 325–43. Santander, Spain: Universidad de Cantabria.

==== •

•

•

15. SCENES IN THE PALEOLITHIC AND LEVANTINE ART OF EASTERN SPAIN Valentín Villaverde

Introduction The aim of this chapter is to assess and compare the scenes found on the portable art assemblage at Cova del Parpalló (Figure 15.1) (Villaverde 1994) and those represented in the Levantine art of the Mediterranean region of Spain (e.g., Beltran 1982). The portable art assemblage at Parpalló comprises 5,612 pieces, with a total of 766 zoomorphic representations, 11 anthropomorphs, and 4,279 defined symbols or ideomorphs, of which 202 are painted and 4,077 engraved (Villaverde 1994). Some of the pieces have traces of coloring or engravings, the symbolic interpretation of which is doubtful (Roldán García et al. 2016). The artistic sequence spans from the Gravettian (26,000–21,000 BP) all the way to the Upper Magdalenian, therefore representing an exceptional case on the European Paleolithic portable art scene. From a thematic and stylistic point of view, this portable art assemblage is fully consistent with Paleolithic art assemblages found throughout Western Europe. On the other hand, in the Iberian Mediterranean region a large number of rock shelters are painted with figurative themes, referred to as Levantine Art since they were first discovered (for history and distribution see Beltrán 1982). Only very rarely was engraving used as a technique in this kind of art (Utrilla and Villaverde 2004), which from a technical point of view sets it apart from the regional Paleolithic cave art in eastern Iberia, where painting and engraving coexist. Levantine art is characterized by the depiction of human and animal representations, the latter comprising only a small number of species, mainly deer and ibex or goats, and to a lesser extent bulls, horses and wild boar. Additional characteristics separating Levantine art from the regional style of Paleolithic art are the frequent findings of figures within scenes and the decorated rocky walls in shallow rock shelters preferentially located in interior mountain areas. The post-Paleolithic chronology of Levantine art is the focus of much debate (García-Robles, García Puchol, and Molina Balaguer 2004; Hernández Pérez 2012; Utrilla, Baldellou, and Bea 2012; Villaverde et al. 2012), but is more generally accepted (see also the discussion in Domingo, Chapter 16). Comparing two assemblages so different in type and so well separated stylistically, whatever the ultimate decisions about the chronology, not only shows the changes in the frequency of scenes represented, but also allows a consideration of the differences in the kinds of scenes and the differing treatment of time in the narratives that can be inferred from a number of them. The different themes and treatments of time need to be regarded as distinctive features of the relevant cultures. The treatment of time in the scenes on the plaquettes at Parpalló does not differ from that which is common to European Paleolithic art generally. On the other hand, Levantine Art represents distinctive treatments of time (Villaverde 2005a) using a type of scene that usually is found in artistic contexts related to agricultural and pastoral societies. In this analysis, there are significant differences between the two kinds of art in both the thematic variations and the treatment of time in scenic narratives. The differences are due to the fact that these two stylistic productions are linked to two different economic and social systems: the hunter-gatherer societies of the Upper Paleolithic, and the agricultural and pastoral societies of the Neolithic. These differences in the use of time will be demonstrated in the remainder of this chapter.

224

VALENTÍN VILLAVERDE

The Concept of the Scene in Prehistoric Art Identifying a scene in prehistoric art entails a series of requirements, sometimes subject to discussion. Delporte (1981), in his analysis of figure associations in Paleolithic art (mostly mobile art; see Chapter 12 by Culley for a discussion of the different treatments of scenes in Upper Paleolithic mobile and parietal art), distinguishes up to five different variants: (1) chance or accidental association, which is more common in cave art assemblages and on portable art pieces and entails an accumulation of different animal species within the same graphic space without obvious stylistic or thematic connection; (2) repetitive superimposition, where various animals of the same species share a graphic space but cannot be interpreted with certainty as representing a herd or group, that is to say, they cannot be identified as a scene; (3) narrative association, or what could be termed the configuration of a scene, given that the individuals represented are taking part in a shared activity such as in scenes of hunting or mating; (4) geometric association, typically involving friezes where animals are often shown in line, or confrontations where the same or different species are shown opposed to each other, both of which might be interpreted as scenes—the friezes as migrations of animals, and the confrontations possibly as typical fights between rutting males of certain species); and (5) thematic association (rare examples in which the action cannot be specified, even when the intentionality of the artist is clear). Aware of the difficulties around the identification of some of these scenes, or the subjectivity that determines whether they can be identified as scenes, Delporte notes that “in order to talk about a scene it is not enough for a number of characters to coexist, but an action must be discernible, which, engages at the same time as it confirms the nature of the characters; and it is precisely the definition of the action that is the most difficult thing to demonstrate unambiguously.” Given the frequency of the repetitive and random associations found in most European Paleolithic portable and cave art, it seems appropriate, when defining a scene, to concentrate on the action that the individuals need to take part in, whether they are animals, humans, or a combination of both. Without discernible and identifiable action, it does not seem possible to talk about the creation of a scene. Another matter, as we shall see later, is whether several characters—animals, humans, or a combination of both—need to take part in the action, or whether, on the contrary, the representation of a single character in which the action is clearly discernible is sufficient to identify a scene. On the other hand, the distinction between a “narrative association,” or scene, and a “thematic association,” in which the action cannot be specified but the represented elements possess sufficient features to rule out being interpreted as a random or “accidental association,” is clearly necessary although it is not always easy to establish. The most paradigmatic case representing the latter would be, in my opinion, the wrongly termed “shaft scene of Lascaux” or the “Woman under the reindeer scene” from Laugerie-Basse. Whereas the content of the scenes can be inferred from the subjects represented, and from their actions, in the thematic associations an interpretation can only be put forward in general terms, making reference to the world of ideas or myths (D’Huy 2016; Le Quellec 2014), or, in many cases, making use of ethnographic references far removed in time and space from those depicted (Bahn 2016). Although ethnographic comparisons can be useful when it comes to “reading” certain content, most of the time what they really inform us of is the difficulty of translating the meanings of certain compositions when the cultural context under which they were produced is unknown to us (Davidson 2017; Ingold 2000). Not only is it about being aware of the difficulty of transferring the thought schemes implicit in our Western concept of art to the production of works by groups in the past, but it also questions the possibility of establishing or applying general ontologies based on images or their associations, when it comes to interpreting the spatial associations of the figures found within these works. The animal scenes portray behaviors that can be easily evaluated in ethological terms, but the scenes in which humans are present are not mere “photographs” of past actions given that they also convey ideas about the social and cultural backgrounds of their authors, and refer to ideas and values that are beyond the actions themselves. Conversely, given that we are here going to be dealing with scenes from different time periods and social and economic contexts, it is also worth noting that in prehistoric art scenes there can be intervening actions that not only imply interactions between humans, between animals, or between humans and animals, but that in some cases the action can be deduced from the objects carried or used by the humans. Even the interpretation of the actions that arise from the handling or carrying of certain tools presents some difficulties because their identification as a scene is always at the verge of a type of distinction that is important when

SCENES IN THE PALEOLITHIC AND LEVANTINE ART OF EASTERN SPAIN

225

dealing with scenes such as these. When referring to animals, one factor is the pictographic component associated with animation or the detail that derives from the expression of the coat or sex, or from an action such as jumping, galloping or alert signaling, which does not necessarily imply a scenic component. A second consideration is the detail in the arrangement or the material objects carried by the humans represented, given that they in themselves can evoke the narrative or scenic character of the representation within which they are integrated. In relation to this problem, it is common to refer only to scenes where several individuals participate in the action, and to name all aspects of age, season, individual behavior or movement as characteristic of the “pictography.” I will return to these distinction terms later in Section 4 as they are also useful when discussing Levantine art scenes. At present, it will suffice to say that “pictographic,” as discussed by Leroi-Gourhan (1983), is everything that individualizes the subject being represented. In my opinion, the pictographic elements capable of individualizing attitudes or indicating the different seasons of the year, such as the fur on animals, antlers, specific behaviors, or the ages of certain individuals, are much more common in Paleolithic art than Levantine art. These elements can help with the identification of the actions of the scenes in which they are integrated, but are not by themselves, sufficient to define scenes. The analysis of animation, or of movement, in Paleolithic or Levantine art, clearly illustrates this point. In Paleolithic art inanimate representations coexist (what Leroi-Gourhan (1983) termed “null animation”) with segmenting animations, centered around a body part, or coordinated animation, in which movement reaches the various parts of the body being represented. In Levantine art, the human figure predominantly responds to a pattern of animation that in most cases covers different parts of the body. Null animation is uncommon, especially in terms of human figure representations. In Paleolithic art some animal arrangements, given their animation, give the impression of ethologically identifiable actions, but their interpretation as scenes is much more problematic. The same is true in Levantine art. Various examples from the plaquettes of Parpalló might be useful in this regard. In order to establish or interpret a scene, is it sufficient to note the presence of an animal jumping, moving a leg, or lowering its head down to the ground either because it is eating or because it is puffing in a threatening manner? In other words, if in order to talk about a scene there must be several individuals participating in a common action that unites them, in the sense in which Delporte uses the term, it is clear that we would not be identifying true scenes, but instead pictographic elements that account for behaviors that may have a high degree of ethological significance. The animation and pictography of Paleolithic art have been the subject of numerous works in recent years, and the number of figures that can be included in these categories has increased considerably (Azéma 2009, 2010), but the scenes continue to represent a small number of the total portable art, and most are from the Magdalenian (17,000–12,500 BP) (Barandiarán 2003).

Scenes on the Paleolithic Portable Art at Parpalló As at other sites of Paleolithic portable art, scenes are not abundant in the Parpalló assemblage, but despite this it is worth noting that they are present from almost the beginning of the sequence and have reliably been dated to the Solutrean period (21,000–17,000 BP), even in its oldest stages, that is to say, those corresponding to the lower and middle Solutrean (21,000–20,000 BP). Other pieces from the Evolved Solutrean (20,000–17,000 BP) and the Magdalenian (17,000–12,000 BP) have in previous works been classified as scenes (Villaverde 1990, 1994), but at present I believe that most of these cases do not meet the necessary requirements for classification beyond simple repetitive associations of figures of the same species. Meanwhile, if we focus our analysis on discussing unproblematically identifiable scenes and comparing the scenes from Parpalló with the assemblage of scenes documented in Paleolithic portable art, we can observe a clear difference. Several scenes were produced during the Solutrean at Parpalló, yet scenes are very rare in the rest of the Paleolithic portable art of this chronology. On the other hand, while the largest numbers of scenes in the Cantabrian or the French southern regions are dated to the Magdalenian, at these dates in Parpalló their numbers are few. Without a doubt, this evident peculiarity shows the importance of the Solutrean occupation in the Mediterranean region of Iberia and the cultural individualization of the Iberian Mediterranean Magdalenian.

226

VALENTÍN VILLAVERDE

Four pieces from the Parpalló assemblage clearly meet the common action requirements necessary to identify a scene, whereas two others are not as clear. We will focus on the contents of these six pieces, ordering them chronologically from the oldest to the most recent. On the first piece, from the Lower Solutrean (21,000–20,000 BP), it is possible to note a stag and a doe (Figure 15.1a, the specimen number in Villaverde

b a

c

FIGURE 15.1a–c. (a) Parpalló, tracing of Plaquette 16079A. Lower Solutrean. The anatomical details and the disposition of the two animals, a stag and a doe, lead to the identification of a pre-coupling scene. (b) Parpalló, tracing of Plaquette 16182. Middle Solutrean. Maternal scene of a doe suckling her fawn. The internal filling of the fawn by means of paired traces showcases the whitish spots found on young deer. (c) Detail of the previous piece. Double representation of the head of the fawn: one complete but not reaching the maternal abdomen and the other partial but reaching the mother’s body. Copyright Valentín Villaverde.

SCENES IN THE PALEOLITHIC AND LEVANTINE ART OF EASTERN SPAIN

227

1994 is #16079A). The anatomical details and the disposition of the two animals lead to the possible identification of a thematic association between a male and a female that evokes a typical pre-coupling scene. Not only does the female figure occupy the graphic space available in front of the partial representation of the male deer, but the male, with details on its antlers, is also depicted with its mouth open and associated traces that could well represent the bellows typical of the mating season. Additionally, the female appears to have her tail up, indicating that she is ready to be mounted. This common type of scene in Paleolithic art incorporates the pre-copulation behavior of various species (reindeer, aurochs, ibex, etc.) which, in this case, departs from the following and sniffing variant and incorporates the rutting and approach. Two other pieces are from the lower half of the Middle Solutrean (20,000–19,000 BP). The first (Figure 15.1b and c, #16182) is one of the best known plaquettes from Parpalló, given its high degree of realism and the detail with which it resolves into a scene of a doe nursing her fawn (see Chapter 1 by Davidson). The disposition of the fawn, with its body infilled by means of paired traces that showcase the whitish spots found on young deer and its head looking toward the mammary glands of its mother, constitutes key elements in the precise identification of the scene. The head of the fawn appears to be represented twice: one is represented whole but not reaching as far as the doe’s abdomen, whereas the second clearly reaches it. This has been interpreted as a decomposition of movement in two successive phases (Azéma 2008; Villaverde 1990). Lastly, the sizes of the female and her young one and the way the latter gets close to her are very similar to what this would look like in real life. The position of the female, resting, is also correct given this particular circumstance. Whereas maternal scenes are relatively well documented in Paleolithic art, the suckling detail has only been noted at Parpalló. The association of a female reindeer and her young at La Bigourdane, with the latter approaching her abdomen, does not illustrate the suckling but simply suggests it is about to take place (Lorblanchet 2010: 415). The other scene from this same period, found on a large block, consists of a lynx jumping toward the neck of an ibex (Figure 15.2a, #16341). Again, the representational details of the lynx enable the identification of the species: a short, raised tail, a triangular head with well-profiled ears, and the filling of the body by means of short, tight traces that resemble the animal’s fur. The hunting of an ibex by a lynx is currently documented at Doñana Spanish National Park (Rodríguez and Delibes 1990), so it is possible to interpret the plaquette of Parpalló as one of the few recorded cases in Paleolithic art where a scene represents the hunting activity of a carnivore. Another noteworthy case is that of a lion jumping on top of a horse at Le Tucd’Audoubert (Azéma 2010: 345), about which some scholars have expressed doubts (Testart 2016). Other similar scenes documented in Paleolithic portable art are limited to a simple association between the carnivore and the herbivore, without portrayal of details related to the attack (Barandiarán 1993). The fourth piece (Figure 15.2b, #17757 A) is from the Evolved Solutrean (19,000–17,000 BP), and like some of the other pieces, the scene represented—a doe walking with her two young—is placed below the representation of an unrelated equid. In the case of the suckling scene, an ibex is located below the mother and young, and on the plaquette of the lynx and the ibex, there is an image of a large equid. In the three cases, the engravings of the animals taking part in the scenes are different from the engravings that are randomly associated with them, indicating that the scenes and random associations coexist on a single piece. Going back to the last maternal scene cited (#17757A), its unequivocal identification stems from various factors. First, the doe and her fawns rest on the same imaginary ground, that is, the lower limit of the plaquette. Second, the three are engraved using the same technique, that of multiple tracing. Finally, the size of the two fawns and their position between the legs of their forward-facing mother are consistent with the ages of the young and clearly correlates with well-documented doe behavior observed in nature. These four pieces enable us to note a fact that we believe to be relevant, namely that the scenic content is resolved by paying special attention to details, whether it be the subjects’ arrangement, relative sizes, fur, or posture, leading to a clear identification of the action being represented. The pictographic reinforcement, as I pointed out earlier, is capable of removing uncertainty regarding the identification or the age of the animals taking part in the scene, which is consistent with an intention to construct a narrative. The four scenes also depict animal behavior and include common themes relating to pre-coupling, looking after young, and carnivores’ predation on herbivores. Their parallels can be seen in later chronological contexts more typical of the Magdalenian. However, another plaquette (#18946 A) from the beginning of the Magdalenian (17,000–15,000 BP) could be included among the pre-coupling scenes, in this case involving a pair of ibex. A female occupies the available space behind the representation of a male, who is larger in size and has more

228

VALENTÍN VILLAVERDE

FIGURE 15.2a–b. (a) Parpalló, plaquette 16341. Middle Solutrean. Lynx jumping toward the neck of a goat. (b) Parpalló, plaquette 17757A. Evolved Solutrean. A doe walking with her two young. The three animals rest on the same imaginary ground and are engraved using the same technique of multiple tracing. Copyright © Valentín Villaverde.

229

SCENES IN THE PALEOLITHIC AND LEVANTINE ART OF EASTERN SPAIN

developed horns. Only the fact of her elevated tail would suggest the female’s receptive attitude and therefore consideration of the plaquette as a scene. Another example (#20777), found in the part of the cave known as Galerías, offers no secure chronological data, even though this part of the cave was probably accessible only in the Magdalenian. It does, however, present an association of three ibex, the interpretation of which suggests two possibilities: the male sniffing the female, or a maternal-type association by which the female and her young would be associated with a male ibex. In this case, too, it is hard to distinguish between a repetitive thematic association and a true scene. Although similar associations have been identified as scenes, we believe that greater rigor should be maintained in this type of identification. The differences in size are ambiguous, and numerous examples in cave and portable art make us want to be prudent regarding this criterion, if no other elements intervene in the identification of the scene. Several things to consider emerge from the discussion of the pieces above. Firstly, this is a very small number of plaquettes and of animals integrated into scenes. All the examples from Parpalló are exclusively of animals. None of the eleven possible anthropormophic representations at the site is part of a scene, though repetitive thematic associations are present in three cases. Secondly, if the association tendency of the animals represented on the pieces at Parpalló is taken as the unit of analysis, using both sides of the plaquette when both are decorated, it is clear that surfaces with zoomorphic representations very commonly display thematic associations that are repetitive (same species) or random (different species). From a total of 766 animals represented, 358 appear integrated in association with other animals. Taking all of the species globally, it is the aurochs and horses that present the greatest association percentages, followed by ibex and deer. Noticeably, these percentages nonetheless present variations throughout the Parpalló sequence (Table 15.1). These data are sufficient to suggest that the graphic behavior observed at Parpalló is similar to the majority of Paleolithic portable and cave art known in Europe (Delporte 1990; Sauvet and Wlodarczyk 1995), the most common scenes are those that focus on animal behavior, and the number of scenic themes is rather small. Furthermore, all of the scenes refer to an immediate moment in time, as they document actions that are resolved in a short time span. The doe nurses; the lynx jumps on the neck of the ibex; the doe moves, accompanied by her two fawns; the male deer bellows, and the act of mounting the female appears to be imminent. These are like snapshots of a well-known ethological reality, but they do not imply either past times or a relatively distant future. The narrative ends in itself. This circumstance is not exclusive to Parpalló. It is well known that the associative tendency of the animals and symbols defines Paleolithic cave and portable art, but scenes are not dominant. From this perspective, the portable art assemblage at Parpalló is quite classic and coincides with both the thematic and the associative content found in Cantabrian and French art, as well as that from the Mediterranean Iberian region. In the same Iberian central Mediterranean region, several Paleolithic parietal assemblages have been found with the same patterns as those observed in the rest of southwestern Europe—repetitive associations that have given rise to narrations or scenes, though only on a few occasions. In any case, no significant quantity of scenes like those documented on French or Cantabrian Magdalenian portable art has been noted in the Iberian central Mediterranean region, indicating that a process of regionalization, already noted on other occasions (Villaverde 1994, 2005b), was under way on those dates. However, there is no shortage of elements that repeatedly remind us of supra-regional contacts, an observation that echoes the consistency in culture change processes in the lithic and bone assemblages (Villaverde, Borao Álvarez, and Cardona 2015). TABLE 15.1. Parpalló animal associations. PM: Pre-Magdalenian levels (25,000–17,000 BP), M: Magdalenian levels (17,000–12,000 BP), AA: Number of associated animals, AR: Number of represented animals, %: Percentage of AA in relation of the number in AR. Pre-Magdalenian

Magdalenian

Total

TAA

TAR

%

TAA

TAR

%

TAA

TAR

%

9

22

40.9

14

19

73.6

23

41

56.1

Horses

34

77

44.1

15

36

41.6

49

113

43.3

Does

25

55

45.5

6

32

18.7

31

87

36.6

Ibex

34

96

35.4

16

35

45.7

50

131

38.1

Aurochs

230

VALENTÍN VILLAVERDE

We can say that images and scenes in Paleolithic art refer to an “eternal present” and lack any reference to past or future temporal perspectives, even if the stories accompanying these images possessed that temporal dimension. Not even in animation, analyzed in depth in the case of cave and portable art, does movement imply a very different temporal dimension. It is about accounting for a behavior that, although it implies a certain temporal duration, never goes beyond a very limited span: the movement of the limbs by means of various representations of each or the changing arrangement of the head. Even in the case of the Penascosa block 4 at Cõa (Baptista 1999)—a scene of a male horse mounting a mare, with various arrangements of the male’s head—the expression of the movement or the action continues to be short in duration, almost instantaneous. The only exceptions appear to be the well-known spear-thrower from Mas d’Azil, on which four horse heads appear sculpted and engraved, evoking different phases of the life of the animal, including a detail of the defleshed head of the animal following its death (Delporte 1990); some animal representations from Lascaux interpreted as the regeneration of time through the succession of the different depicted species (Aujoulat 2004); and four horse heads from Chauvet, interpreted as a unique animal in four different behaviors or life phases (White 2003: 79).

Levantine Art Scenes Levantine art, which can be found over a large geographical area in the Iberian Mediterranean area, is characterized by the importance of scenic compositions that include human and animal figures, either separately or in an integrated way (Beltrán 1982). This art’s chronology has been a matter of debate since its discovery. It was first thought to be Paleolithic, then Mesolithic, and finally Neolithic. The art’s temporal assignment is based fundamentally on criteria related to the subject matter, spatial studies centered on the correlation between the distribution of the rock shelter art assemblages and the archaeological evidence, and the territory occupation, all of which suggest a strong correlation with the Neolithic (Fairén 2006; Martínez i Rubio and Martorell 2012; Villaverde et al. 2016). It is important to note a certain degree of variability in the way Levantine art represents animals and humans. Especially in the case of the latter, the majority of regional studies have established the existence of different styles or so-called graphic horizons, with significant variations in the degree of naturalism of the representations, the body proportions, and/or the aptitudes for and elements of adornment, weaponry, or clothing (Domingo 2008; Mateo Saura 2006; Utrilla and Martínez-Bea 2007). The variety of scenes and their quantitative importance constitute not only the two characteristic features of Levantine art, but also their main differences from Paleolithic art. There are also strong stylistic differences in the ways the artists represented the same subjects. However, it is not always easy to determine the presence of scenes in either category of art because repetitive association is also present throughout the different graphic horizons found in Levantine art. In addition, scenes involving only animals are unusual—a substantial change from what is typical of Paleolithic art. An additional problem in the analysis of scenes is the addition of figures in different styles (Sebastián 1987), as at times the additions have ended up modifying the content of the scenes they are joining. This circumstance is especially important and must be analyzed carefully when interpreting scenes or specifying the action of which the motifs are a part, whether they be human or animal. We will begin by looking at the variety of scenic content in order to carefully assess this second issue, using as an example one of the best known and most reproduced scenes in Levantine art: the hunting of a herd of deer found at Cova dels Cavalls. One of the few global approaches that have so far been proposed for the analysis of scenes in Levantine art was that by López-Montalvo (2005). Her work is restricted by the scarcity of systematic treatment of this topic in a significant number of publications on rock art sites, due mostly to the application of different criteria to identify the scenes and the uneven quality of the available graphic information. In any case, there are two predominating scene themes in the Levantine art complex: hunting scenes, with many variants; and war or fight scenes, also in numerous variations. In the area of Valltorta-Gassulla (Castellón)—one of the most characteristic concentrations of Levantine art—a systematic study of hunting scenes carried out by López-Montalvo (2005) allowed a distinction to be established between single and collective scenes and the following variations of each: (a) the interception or direct hunter-prey confrontation, and (b) the stalking, tracking, or pursuit. In the assemblages studied by

SCENES IN THE PALEOLITHIC AND LEVANTINE ART OF EASTERN SPAIN

231

López-Montalvo, she observed a balance between the number of scenes with a single hunter and those of a collective type, as well as a certain predilection for some species, such as deer and wild boar. The most interesting aspect of these scenes, however, is the difference in the way time is treated, compared to its treatment in in Paleolithic art. The scenic content ceases to represent just a single instance or short space of time, and instead incorporates processes that entail a long temporal duration. In the pursuit or tracking categories, we can frequently note an intention to break the hunting process down into at least two separate stages: the sighting of the prey, and the following of its tracks, which in some cases cover a relatively great distance (in Cingle de la Mola Remigia, for example, the tracks are a meter long). Similarly, a scene may include first encountering the prey and then bringing it down. In some cases the encounter phase is omitted or perhaps has disappeared due to conservation problems. One of the most interesting examples of this type of temporal and perhaps spatial narrative is from the third panel in de La Palla (Figure 15.3a) (Hernández a

b

c

d

FIGURE 15.3a–d. (a) An archer integrated into the hunting scene of de La Palla rock shelter. Copyright Archivo del IVCR+i. (b) Tracing of the complete scene of de La Palla rock shelter (the temporal narrative must be read from bottom to top; for more details see the text) (Hernández Pérez et al. 1988). (c) Rock shelter VIII of Cingle de la Mola Remigia. DStrech enhancement (ybk filter) of photograph in 15D, with a scene in which a fallen individual is carried in the arms of his colleague (d) Photograph of the previous image. Copyright Valentín Villaverde.

232

VALENTÍN VILLAVERDE

Pérez, Ferrer, and Català 1988; Villaverde 2005a), where up to four stages are depicted: the sighting of prey and launching of a weapon; the following of the trail or tracking of the wounded prey; the final shot to the injured animal; and its ultimate collapse, including the vomiting of blood and displays of agony. Space and time are integrated into a scene that underpins the story of an action of certain duration. Another example of what we have just noted is found at La Araña rock shelter (Figure 15.4), in a wellknown hunting scene on the left-hand side of the second cavity. The reconstruction of the scene by Hernández Pacheco (1924) (Figure 15.4a), which fills in the deteriorated figures to resemble something close to their original appearance, allows us to see that this scene, when taken as a whole, is impossible or contradictory to analyze in immediate temporal narrative terms. A herd of ibex in a line appear to cut across a narrow passage. Next to them are scattered animals, some brought down, others running. We see two phases of the collective hunting process in a single image—the stalking and waiting for the animals to pass by, and the shooting and bringing down the prey once it reaches the ambush point from where arrows would be shot, according to plan. All the animals, even those that appear to know what is about to happen, appear to have been shot. Twelve archers and nine ibex are part of the scene. Their sizes, methods of representation, and details of the bows and arrows are portrayed in the same style. This suggests that the scene was conceived as a single narrative composition with some slight variation, and may have been produced by a single hand. Something similar can be noted in some of the wild boar hunting scenes from Cova Remigia (Sarrià 1989). In a scene from cavity V, an adult boar running to escape while being chased from behind and on both sides by at least four archers is associated with a young boar already brought down with its legs up in the air in a way that is repeated quite often in scenes where these and other animals are represented as hunted prey. Arrows have already been fired and indeed appear to be stuck in the body of the mature wild boar; now it is the pursuit of the wounded animal trying to escape that is being represented. In Levantine art, this continuous suggestion of past or future events, including both the background and the expected resolution of the actions, is repeated with sufficient consistency to make us think that this temporal narrative concept is part of a standard means of graphic expression on the part of their creators. The vividness of some of the scenes comes precisely from their representation of the passage of time: such scenes are not static representations of a moment, but actions that develop through time and contain within their structure the keys to reading them. In all three cases noted here, there is a formal unity that allows us to conceive of the scenes as unitary. The final case we want to discuss is the scene at Cova dels Cavalls (Figure 15.5), narration of which is complicated by the successive addition of figures to the original core of the scene, generating a complexity that must now be interpreted as resulting from the addition process itself. Careful study of this composition has enabled us to reconstruct the process by which the depicted figures were executed, using the stylistic seriation established at the regional level, the repainting and rectification of some of the figures, and the positions of the different themes within the composition (Villaverde and Martínez-Valle 2002). Provision of a detailed description of the scene is beyond the scope of the present chapter, but it is sufficient it to say that although the animals represent a central nucleus and conform to the same formal and stylistic criteria, the same cannot be said of the archers that surround them. Not even the assemblage of four archers waiting for the running herd to arrive corresponds to the same stylistic phase. What appears to be a representation of a strategically planned hunt, in which several archers drive animals into an area where other archers lie in wait to kill them, can only be read as a scene in this way because of the progressive accumulation of archers on the flanks of the herd. The important thing is to assess how each new addition changed the exact sense of the scene in a process entailing a progressive appropriation of the narration by those who added to and modified the scene. Thorough study of this composition has enabled the reconstruction of the process of the figures’ execution by way of a stylistic seriation established at the regional level, the repainting and fixing of some of the figures, and the general positions of the different themes on the whole. The unitary character of a herd of deer comprising two males, two young, and at least seven females emerges in part from the uniformity and consistency of the relative sizes of the figures, the arrangements of the heads, necks, and limbs with their identical movement conventions, their orientation, and their slight downward tilt, but also stems from their position of equidistance and respect for the available spaces. Meanwhile, diversity is evident in the archer figures accompanying or waiting for them, with at least four different styles documented around the scene. Here we can see an addition process that preserved the theme of the scene but transformed its specific content, generating a final vision of the planned process in which many artists were involved.

SCENES IN THE PALEOLITHIC AND LEVANTINE ART OF EASTERN SPAIN

233

FIGURE 15.4a–b. (a) La Araña, second cavity. Tracing of the hunting scene of a herd of ibex (Hernández Pacheco 1924). (b) Photographic detail of the previous scene. Copyright Archivo del IVCR+i.

234

VALENTÍN VILLAVERDE

a b

c

FIGURE 15.5a–c. (a) Cova dels Cavalls. Tracing of the complete hunting scene of a herd of deer. Human archers were added at different times, transforming the character of the scene (Martínez Valle and Villaverde 2002). (b) Archer 25 of the previous tracing. This is one of the archers situated at the left of the scene, awaiting the arrival of the herd. (c) Detail of the deer herd. Two fawns with typical whitish spots are represented in the center of the herd, surrounded by several females. Copyright © Valentín Villaverde.

SCENES IN THE PALEOLITHIC AND LEVANTINE ART OF EASTERN SPAIN

235

The other predominant kind of scene in Levantine art, that of conflict/warfare, also appears in numerous variants that range from organized conflict to representations of very specific aspects relating to warfare. This issue has been thoroughly discussed by López-Montalvo (2011, 2015), who assessed both the variations noted in the chronological sequence and spatial variations. We can say that violence between humans is present throughout the whole Levantine artistic sequence, and that—as is the case in the hunting scenes—the treatment of these scenes also incorporates a wide temporal perspective. The scenes included in this section range from organized fighting between archers (e.g., rock shelter II at Civil; Les Dogues; rock shelter IX at Cingle de Mola Remigia; de la Fuente del Sabuco I rock shelter; and Torcal de las Bojadillas), to executions carried out by groups of archers who brandish their arrows in front of a defeated, shot person who is placed in front of them (rock shelter V at Remigia). There are also images of people who have been shot and are clearly wounded, even if they have not been brought down or are on the point of collapse (rock shelter III at Remigia or rock shelter VII at Saltadora), as well as the explicit representation of a collapsed individual carried in the arms of his colleague (rock shelter VIII of Cingle de la Mola Remigia) (Figure 15.3b). A special kind of composition, one that is difficult to distinguish from repetitive associations, is that comprising groups in movement. This kind of composition may adopt a simple succession of armed characters (Abric del Voro or rock shelter IX at Cingle de la Mola Remigia, to name a couple of examples that integrate four and five archers respectively), or the evocation of a group comprising men, women, and children clearly arranged as if moving across the landscape. The clearest example of this kind of scene is found in Abrigo de Centelles (Villaverde, Guillem, and Martínez 2006). The composition fulfills all of the requirements for being considered a scene: one archer carries two arrows, whereas another, weaponless archer carries bags or bundles of a certain size; two inner figures who are dressed in a different way and have small breasts marking them as female carry two big bundles on which the figures of young children are represented, sheltered from the sun by parasols; another unarmed figure that could well be a female carries a bag on her back, while another, smaller in size, appears to represent a child. Bundles, objects, the identification of the different sexes, the spatial arrangement and articulation of the characters are all aspects indicating that this represents a traveling group carrying various things. The impression is one of a scene focused on the representation of movement. It is even possible that one of the archer figures seen in the walking line was an addition, given the differences it presents with regard to the other figures. In any case, its presence does not in any way modify the narrative structure of the composition. On certain graphic horizons in Levantine art, one can observe a tendency toward compositions that represent numerous human figures predominantly facing the same direction. It is very difficult to classify these kinds of repetitive associations as a scene, because aside from the walking posture, nothing indicates what type of action they are involved in. In Levantine art, one fairly often finds groupings made up of a successive repetition of two to four figures and ending with very numerous configured groups that are difficult to assess. Are these associations meant to evoke the idea of a territorial domain by these groups, or are they simply a gesture of addition that, due to successive repetition, keeps the associated symbolism of certain panels alive, with the final reading not being more than the sum of the additions? With regard to the last question, it might be appropriate to remember that repetition is an essential element of ritual, as Ross and Davidson (2006) have pointed out in relation to Central Australian rock art. Only when the action can be precisely identified can we talk of a scene, and this requirement, which also applies when dealing with Paleolithic art, is not necessarily easy to figure out in Levantine art and even less so in this kind of thematic repetition. A grouping of animals of the same species does not necessarily imply the representation of a scene, and the same holds true for a group of human figures. Following these assessments, it is clearly no easy task to determine scenes of a social nature. With regard to this subject, a number of potentially useful clarifications can be established. Although the action cannot be determined or else is limited to a simple interaction due to proximity, the groups of two or more figures that can clearly be stylistically or spatially differentiated from the rest of the figures in a panel could be regarded as the clearest example of a scene that evokes patterns of social behavior: individuals talk while walking; women or both women and men participate in these same potential conversations, which are discernible due to the arrangement of these participants in groups or face to face; or other individuals carrying out complicated, unidentifiable actions. In this case, it is the individualization of the represented characters, based on the details, attitudes, and/or the rarity of the arrangement or objects, that enables us to understand

236

VALENTÍN VILLAVERDE

that we are looking at true scenes. The interpretation of the objects with which they interact can modify the reading of the action(s). Interpretative ambiguity is hardly avoidable, considering the difficulty and range of possible alternatives that emerge from some of these compositions. For example, the association between the human figures 30 and 31 from Panel 1 at Abrigo II de la Sarga is beyond doubt, given the way they are represented: legs overlapping, arms positioned one over the other, suggesting that they are touching or holding each other, the face to face position of the two characters, and their identical proportions and style. Although the two figures integrate clearly into a common action, the interpretation of the scene is still unclear. It is difficult to establish whether a fight is being depicted, in which the action consist of one of the two characters is trying to stab the other in the chest, as was suggested by the authors of the publication in which these figures were first presented (Hernández Pérez et al. 2007), or whether it is another kind of action of a social or symbolic nature. The proximity of the figures whose actions appear not to have anything to do with this part of the composition does nothing to help clarify the scene. Another set of compositions, in this case not necessarily involving pairs or groups with more individuals, is associated with scenes that depict activities that concern neither hunting nor warfare. This is so in plant-collecting scenes or scenes linked to certain female characters—or dancing scenes, usually staged by a group of female characters wearing skirts and showing their breasts. The elements of dance are not always explicit, but the deliberate association seems to indicate the artist’s willingness to account for the joint action. Plant-collecting scenes and scenes representing human figures climbing up a rope, generally interpreted as honey collecting, suggest that in certain circumstances, a single human character can constitute a scene. It is clear that the scenic component in these cases derives from the incorporation of other objects or, as in the honey collection scenes, the explicit representation of the insects flying around the honeycomb. Compositions that can be grouped under this theme are few and have a very restricted geographical distribution to boot (Martínez i Rubio 2009). At Cingle de l’Ermità, various figures are represented climbing ropes, sometimes using cordages, in order to gain access to the honeycomb. At El Mas d’En Salvador only one figure is represented; the same is true for as at Chorradores and Arpán L, although in this case the peculiarity stems from the fact that the character going to the honeycomb is descending, whereas and at La Araña and Abric de la Penya the honey gatherers are in pairs, even when a different way of representing the two figures suggests the possibility that here we may be dealing with two cases of additions, unless the different treatments were intended to signify distance between the figures or their roles in the action. Given this kind of variation and the requirements linked to the identification of a scene, we are interested in scenes in which only a human figure is represented, because in this case it is the association with the flying insect which guarantees their interpretation. A different case on the verge of being identified as a scene comes from the Cinto de las Letras (Jordá and Alcácer 1951; Martínez i Rubio 2006), where a woman who appears to be carrying a bag or bundle slung across her back points her hands in the direction of what appears to be a representation of a plant. In this case, the difficulty of identification arises from the plant motif. We have also run into difficulties in interpreting a good number of (so-called) scenes of climbers, whose action consists exclusively of climbing up what appear to be rather large tree trunks. In some cases the presence of bags reinforces the scenic character of the representation, even if the reason behind their climbing, such as picking fruit from trees, is not visible for identification. The common action suggests something that has not so far been identified in European Paleolithic art—that is, interaction between humans and the plant world.

Final Thoughts We can conclude that the analysis of scenes requires much methodological rigor and careful characterization of what we mean by “scenes.” Both the Levantine and the Paleolithic art of the Iberian Mediterranean region offer clear examples of scenes that fit Delporte’s (1981) definition of a scene and meet his requirements for establishing a rock art scene as such. Most researchers more or less explicitly accept that a scene involves different characters taking part in a common action; however, an action in relation to plants or insects might involve only a single character. Overall, this is not common in Levantine art, but it does not

SCENES IN THE PALEOLITHIC AND LEVANTINE ART OF EASTERN SPAIN

237

differ in concept or in the number of elements involved in the action of a hunting scene that involves only a single person and a single animal. Pictographic elements reinforce the identification of the actions and therefore favor the narrative character of the scenes. The fact that these elements appear early in the portable art at Parpalló indicates that the authors of these scenes wanted to differentiate them from the more general trend observed in European Paleolithic art, that is, considerable numbers of repetitive associations lacking scenic significance. Levantine and Paleolithic art differ substantially in the number of scenes documented, showing that these two types of art praxes represent different concepts of graphic expression. The narrative character of Levantine art fostered a predominance of hunting and war themes, as well as a comparatively greater presence of human representations. Hunting might not necessarily have been associated with economic activity so much as it was associated with prestige. War is a clear indicator of social conflict in a Neolithic context. In addition, these themes incorporate a narrative procedure in which different phases of an action appear integrated. In Paleolithic art, these kinds of references to different times are hardly ever documented, but they are documented on numerous Levantine art assemblages, and with remarkable variety in terms of the approaches and results. It would be of interest to delve into this further in order to overcome the obstacles posed by mere descriptions and assess readings of the scenes in Levantine art with greater rigor. This would mean accepting that some themes need to be broken down into their different phases if they are to be correctly interpreted. This kind of analysis, however, is hindered by the cumulative nature of some of these compositions, in which additions clearly take over previous content and at times even end up transforming the scenes.

Acknowledgments Research was funded by the Spanish Ministerio de la Ciencia e Innovación (HAR2014-52671-P and by the Generalitat Valènciana (PROMETEO/2017/060). I wish to thank Iain Davidson and April Nowell for their invitation to participate in this volume and the revision of the English text. Valentín Villaverde is Professor of Prehistory at the University of Valencia (Spain) and Director of PREMEDOC (Prehistory of the Western Mediterranean) Research Group. In relation to the study of Paleolithic Art, he is author of Arte Paleolítico de la Cova del Parpalló, edited by Servicio de Investigación Prehistórica de Valencia, and co-author of three books dedicated to the study of Levantine Art: La Cova dels Cavalls en el Barranc de la Valltorta, Los Abrigos VII, VIII y IX de les Coves de la Saltadora y Arte Rupestre en el Riu de les Coves (Castellón), all of them published by Generalitat Valenciana.

References Aujoulat, N. 2004. Lascaux. Le geste, l’espace et le temps. Luçon: Éditions du Seuil. Azéma, M. 2008. “Representation of Movement in the Upper Paleolithic: An ethological Approach to the Interpretation of Parietal Art.” Anthropozoologica 43(1): 117–54. ———. 2009. L’Art des cavernes en action. Tome 1: Les animaux modèles. Aspect, locomotion, comportement. Paris: Éditions Errance. ———. 2010. L’Art des cavernes en action. Les animaux figurés. Animation et mouvement, l’illusion de la vie. Paris: Editions Errance. Bahn, P. G. 2016. Images of the Ice Age. Oxford: Oxford University Press. Baptista, A. 1999. No tempo sem tempo. A arte dos caçadores paleolíticos do Vale do Côa. Vila Nova de Foz Cõa: Instituto Portugués de Arqueologia. Barandiarán, I. 1993. “El lobo feroz: la vacuidad de un cuento magdaleniense.” Veleia 10: 7–37. ———. 2003. Grupos homoespecíficos en el imaginario mobiliar magdaleniense. Retratos de familia y cuadros de género, Serie Minor, 21. Vitoria: Universidad del País Vasco. Anejos de Veleia. Beltrán, A. 1982. Rock Art of the Spanish Levant: The Imprint of man. Cambridge: Cambridge University Press. Davidson, I. 2017. “Images of Animals in Rock Art: Not Just Good to Think.” In The Oxford Handbook of the Archaeology and Anthropology of Rock Art, ed. B. David and I. McNiven, 435–68. Oxford: Oxford University Press. Delporte, H. 1981. L’objet d’art préhistorique. Paris: Réunion des Musées Nationaux. ———. 1990. L’image des animaux dans l’art préhistorique. Cahors: Picard.

238

VALENTÍN VILLAVERDE

D’Huy, J. 2016. “Peut-on tester scientifiquement la phylogenetique des mythes? A propos du motif de la chasse cosmique.” Mythologie Française 263: 20–24. Domingo, I. 2008. “From the Form to the Artists: Changing Identities in Levantine Rock Art (Spain).” In Archaeologies of Art: Time, Place, Identity, ed. I. Domingo Sanz, D. Fiore, and S. K. May, 99–129. Walnut Creek, CA: Left Coast Press. Fairén, S. 2006. El paisaje de la neolitización. Arte rupestre, poblamiento y mundo funerario en las comarcas centro-meridionales valencianas. Murcia: Universidad de Alicante. García-Robles, R. M., O. García Puchol, and L. Molina Balaguer. 2004. “El Arte Levantino y el proceso de neolitización en el arco mediterráneo peninsular: el contexto arqueológico y su significado.” Archivo de prehistoria levantina 25: 61–90. Hernández Pacheco, F. 1924. Las pinturas prehistóricas de las Cuevas de La Araña (Valencia): Evolución del arte rupestre en España. Madrid: Comisión de Investigaciones Paleontológicas y Prehistóricas. Hernández Pérez, M. S. 2012. “Defining Neolithic Art: Levantine, Macroschematic and Schematic art in the Mediterranean Arc of the Iberian Peninsula.” In The Levantine Question: Post-Paleolithic Rock Art in the Iberian Peninsula, ed. J. García Arranza, H. Collado Giraldo, and G. Nash, 145–65. Budapest: Archaeolingua. Hernández Pérez, M., P. Ferrer, and E. Català. 1988. Arte rupestre en Alicante. Alicante: Fundación del Banco Exterior del Banco de Alicante. ———. 2007. “La Sarga (Alcoi, Alicante). Nuevas imágenes, nuevas interpretaciones.” Recerques del Museu d’Alicoi 16: 35–60. Ingold, T. 2000. The Perception of the Environment: Essays on Livelihood, Dwelling and Skill. London: Routledge. Jordá, F., and J. Alcácer. 1951. Las pinturas rupestres de Dos Aguas. Valencia: Trabajos Varios del SIP. Le Quellec, J. L. 2014. “Une chrono-stratigraphie des mythes de création.” In Mémoire culturelle et transmission des légendes, ed. Y. Vade, 51–72. Paris: Editions L’Harrmattan. Leroi-Gourhan, A. 1983. Los primeros artistas de Europa. Introducción al arte parietal paleolítico. Madrid: Ediciones Encuentro. López Montalvo, M. E. 2005. Análisis interno del arte levantino: la composición y el espacio a partir de la sistematización del núcleo Valltorta-Gasulla. València: Universitat de València. ———. 2011. “Violence et mort dans l’art rupestre du Levant: groupes humaines et territoires.” In L’armament et l’image du guerrier dans les sociétés anciennes: de l’objet à la tombe, ed. L. Baray, M. Honegger, and Dias-Meirinho, 19–42. Dijon: Éditions Universitaires de Dijon. ———. 2015. “Violence in Neolithic Iberia: New Readings of Levantine Rock Art.” Antiquity 89: 309–27. Lorblanchet, M. 2010. Art Pariétal. Grottes ornées du Quercy. Rodez: Éditions de Rouergue. Martínez i Rubio, T. 2006. “El Cinto de las Letras (Dos Aguas), revisió del panell 55 anys després de la seua publicació.” Saguntum PLAV 38: 61–91. ———. 2009. “El Abrigo de los Chorradores (Millares, Valencia). Una nueva representación de recolección de la miel a orillas del Río Jucar.” In El Arte Rupestre del Arco Mediterráneo de la Península Ibérica, ed. J. A. López Mira, R. Martínez Valle, and C. Matamoros de Villa, 95–104. València: Generalitat Valenciana. Martínez i Rubio, T., and X. Martorell. 2012. “La senda heredada: contribución al estudio de la red de caminos óptimos entre yacimientos de hábitat y de arte rupestre neolíticos en el Macizo del Caroig (Valencia).” Zephyrus 70: 69–84. Martínez-Valle, R., and Villaverde, V. 2002. La Cova dels Cavalls en el Barranc de la Valltorta. Monografías del Instituto de Arte Rupestre, 1. València. Generalitat Valenciana. Mateo Saura, M. 2006. “Aproximación al estudio de la figura humana en el arte rupestre levantino del Alto Segura.” Cuadernos de Arte Rupestre 3: 125–60. Rodríguez, A., and M. Delibes. 1990. El lince ibérico (Lynx pardina) en España. Distribución y problemas de conservación. Colécción Técnica: Ministerio de Agricultura, pesca y alimentación. Roldán García, C., V. Villaverde Bonilla, I. Ródenas Marín, and S. Murcia Mascarós. 2016. “A Unique Collection of Paleolithic Painted Portable Art: Characterization of Red and Yellow Pigments from the Parpalló Cave (Spain).” PLoS ONE 11(10): e0163565. Ross, J., and I. Davidson. 2006. “Rock Art Ritual: An Archaeological Analysis of Rock Art in Arid Central Australia.” Journal of Archaeological Method and Theory 13(4): 305–41. Sarrià, E. 1989. “Las pinturas rupestres de Cova Remigia (Ares del Maestre, Castellón).” Lucentum 7/8: 7–33. Sauvet, G., and A. Wlodarczyk. 1995. “Eléments d’une grammaire formelle de l’Art pariétal paléolithique.” L’Anthropologie 99: 193–211. Sebastián, A. 1987. “Escenas acumulativas en el arte rupestre levantino.” Bajo Aragón, Prehistoria 7/8: 377–97. Testart, A. 2016. Art et religion. De Chauvet à Lascaux. Luçon: Éditions Gallimard. Utrilla, P., V. Baldellou, and M. Bea. 2012. “Aragon and Spanish Levantine Art: A Territorial Study.” In The Levantine Question: Post-paleolithic Rock Art in the Iberian Peninsula, ed. J. García Arranza, H. Collado Giraldo, and G. Nash, 263–82. Budapest: Archaeolingua. Utrilla, P., and M. Martínez-Bea. 2007. “La figura humana en el arte levanitno aragonés.” Cuadernos de Arte Rupestre 4: 163–205. Utrilla, P., and V. Villaverde. 2004. Los grabados levantinos del Barranco Hondo, Castellote (Teruel), Zaragoza: Diputación General de Aragón.

SCENES IN THE PALEOLITHIC AND LEVANTINE ART OF EASTERN SPAIN

239

Villaverde, V. 1990. “Animation et scènes sur les plaquettes du Parpalló (Gandia, Espagne): quelques considérations sur la pictographie dans l’art mobilier.” In L’art des objets au Paleolithique. Tome 2: Les voies de la recherche, ed. J. Clottes, 227–41. Clamecy: Ministère Culture, Direction de Patrimoine. ———. 1994. Arte paleolítico de la Cova del Parpalló (estudio de la colección de plaquetas y cantos grabados y pintados). València: Museu de Prehistòria de València. ———. 2005a. “Arte Levantino: entre la narración y el simbolismo.” In Arte rupestre en la Comunidad Valenciana, ed. R. Martínez Valle, 197–226. València: Generalitat Valenciana. ———. 2005b. “Arte Paleolítico de la región mediterránea de la Peínsula Ibérica: de la Cueva de la Pileta a la Cova de les Meravelles.” In Congreso de Arte Rupestre en la España Mediterránea, ed. M. Hernández and J. A. Soler, 17–43. Alicante: Instituto Alicantino de Cultura Juan Gil-Albert. Villaverde, V., M. Borao Álvarez, and J. Cardona. 2015. “Dos piezas del Paleolítico superior del Mediterráneo ibérico con paralelos extra-meditrerrnáneos.” Cuadernos de Prehistoria y Arqueología 41: 147–63. Villaverde, V., P. M. Guillem, and R. Martínez. 2006. “El horizonte gráfico Centelles y su posición en la secuencia del Arte Levantino del Maestrazgo.” Zephyrus: Revista de prehistoria y arqueología 59: 181–98. Villaverde, V., T. Martínez i Rubio, P. M. Guillem Calatayud, R. Martínez Valle, and J. A. Martínez Álvarez. 2016. “Arte rupestre y hábitat en la prehistoria del Riu de les Coves. Aproximación a la cronología del Arte Levantino a través de la red de caminos óptimos.” In Del neolític a l’edat del bronze en el Mediterrani occidental, coord. J. Juan Cabanilles, 501–20. València: SIP. Villaverde, V., and R. Martínez-Valle. 2002. La Cova dels Cavalls en el Barranc de la Valltorta. Tirig: Monografías del Instituto de Arte Rupestre, Museu de la Vallatorta. Villaverde, V., R. Martínez Valle, P. M. Guillem Calatayud, E. López-Montalvo, and I. Domingo Sanz. 2012. “What Do We Mean by Levantine Rock Art?” In The Levantine Question: Post-Paleolithic Rock Art in the Iberian Peninsula, ed. J. García Arranza, H. Collado Giraldo, and G. Nash, 81–106. Budapest: Archaeolingua. White, R. 2003. Prehistoric Art: The symbolic Journey of Humankind. New York: Harry N. Abrams, Inc.

==== •

•

•

16. NEW INSIGHTS INTO THE ANALYSIS OF LEVANTINE ROCK ART SCENES INFORMED BY OBSERVATIONS ON WESTERN ARNHEM LAND ROCK ART Inés Domingo Sanz

Introduction Rock art is a uniquely human visual form of communication and information exchange, driven by a human need to share information on the natural, the cultural, or symbolic world. This cultural form of communication has occurred all over the world, at different points in time. Throughout history, humans have used visual media to call their peers’ attention to particular aspects of the material and symbolic world, as was also the case in prehistoric times. Exploration of continuities and shifts in these unique cultural products through analysis of motifs, techniques, content, patterns of composition, distribution, and so forth has the potential to shed light on the evolution of human creativity, human cognition, and cultural practices through space and time. Of special interest to this volume is the appearance of scenes in the history of prehistoric art, which brought humans one step further in their evolution of creative thinking and the human capacity to communicate through images. In Europe, this further step is well illustrated by Levantine rock art (LRA), a rock art tradition featuring substantial changes in subject matter and patterns of composition when compared to previous art, with narrative scenes extensively used for the first time as innovative visual storytelling conventions. Levantine rock art, a rock art tradition unique to Mediterranean Iberia, represents a narrative or scenic turn in the history of prehistoric art in this part of the world. While previous European Paleolithic art generally emphasized aspects of the animal world, with naturalistic animals dominating the panels mainly in non-scenic compositions, LRA moved to a new, explicit focus on humans, their material culture, and their cultural practices through narrative and dynamic scenes. With LRA we face the first literal representation of humans fighting, hunting, or developing other sorts of activities in these territories. However, the origin, timing and factors behind such a change (whether resulting from endogenous or exogenous influences) are still a matter of debate. Was this art produced by the last local hunter-gatherers at some point after the Pleistocene-Holocene transition, either as a result of internal sociocultural changes or in response to external pressures (such as Neolithic populations’ arrival in these lands around 7,500 years ago)? Or was it introduced with agriculture, despite being an art that is not about agriculture? Analyses of the content of the scenes depicted in LRA and the relation between this art and other rock art traditions identified in the regional sequence have shed some light on these discussions, which I shall summarize below. Still, the exact chronology of this particular change and the modes of subsistence of the producers remain uncertain. To address this situation, alternative sources of information are needed to further inform existing debates and to clarify whether this shift toward a new scenic art emphasizing things related to humans and culture was led by a hunter-gatherer or a farming society. Interestingly, the change from a mostly non-scenic art focused on the animal world to a narrative art focused on humans and their practices has also been identified in other parts of the world at different points in time and in different sociocultural contexts. Exploration of similar developments in other territories may provide further evidence to advance our understanding of this tradition. Based on long-term experience exploring rock art in Arnhem Land, where both endogenous and exogenous variables caused changes in the art of Aboriginal people over time, I critically reflect on some aspects of the debate about the producers of

NEW INSIGHTS INTO THE ANALYSIS OF LEVANTINE ROCK ART SCENES

241

LRA, and thus about the emergence of scenes in these lands. Acknowledging that these initial reflections do not end the discussions, I aim to offer a starting point for future comparative analysis of major shifts in the history of prehistoric art to explore global questions about the evolution of human cognition and human visual communication, as well as the potential mechanisms driving major cultural and cognitive changes such as the one treated here.

Levantine Rock Art: An Art of Narrative Scenes. With more than a thousand sites distributed throughout Mediterranean Iberia, LRA is one of the largest bodies of rock art in all Europe, offering exceptional visual records of past humans’ interactions with nature and with each other. It is located in open-air rock shelters overlooking the extensive network of rivers and tributaries flowing toward the Mediterranean Sea within the lands of Aragon, Catalonia, Castilla la Mancha, Valencian Community, the Region of Murcia, and Andalusia (Figure 16.1). This large body of figurative

FIGURE 16.1. Geographic distribution of (1) Macroschematic Art (MA), (2) Levantine Rock Art (LRA), and (3) Schematic Art (SA) in Iberia. LRA is exclusive to the eastern side of the Iberian Peninsula, MS is limited to Alicante province, and SA appears throughout Iberia. Base map created by NASA, MODIS / LANCE. Public domain.

242

INÉS DOMINGO SANZ

art includes brush-painted and, rarely, engraved (Utrilla and Villaverde 2004) figures, in which attention is placed on silhouettes without internal details. Nevertheless, contour details are explicit enough to differentiate between animal species and sometimes even between sexes (at least for humans and some animal species, such as deer, at some points in time). The level of detail proves great knowledge of the anatomy of the wild fauna depicted, mainly deer, wild goats, boars, bulls, horses, and less often, carnivores and insects. In some species, differences between adults and young are made explicit by characteristic hair coloration (e.g., dots on fawns and parallel lines on young wild boars). Images of animal-headed or composite humans (mostly antler-headed humans) also exist, but they are rare. Vegetation is incidental, with just a few wild trees (la Sarga in Alcoi) and herbaceous plants (Torcal de las Bojadillas site in Nerpio, Santa Maira in Castell de Castells or Civil site in Albocàsser). But the principal foci of LRA are almost certainly humans, with explicit differentiation between men, women and occasionally, in the early stages, children (for a synthesis on potential representations of children see Bea 2012). LRA shows humans engaged in dynamic scenes reproducing all sorts of anthropogenic activities and behaviors previously unseen in the rock art. These depictions portray very specific attributes of human anatomy (distinctive facial and other physical features such as sexual attributes, beards, and different sorts of hairstyles) as well as all sorts of adornments (headdresses, bracelets, belts, straps), clothing (long and short pants and skirts) and equipment (arrows, bows, quivers, boomerang-shaped tools, bags, baskets, and bundles). The body shapes range from quite naturalistic to idealized forms, but they are unquestionably representations of anthropomorphic figures, whether or not they represent real or imaginary beings. Thus they differ clearly from the preceding ambiguous Paleolithic human representations, which are consistently non-realistic, extremely simplified, and often even questionable. Lorblanchet (1989) interprets this ambiguity among Paleolithic humans as a deliberate means of visually expressing the bonds (i.e., the complementary dimension, undifferentiation, or permeability of boundaries) between humans and animals. On the contrary, LRA speaks of humans, who now play the dominant part, marking the birth of visual anthropocentrism in this part of the world. The primacy of humans is achieved through large-scale visual narratives emphasizing human actions. But the question is, why did this dramatic change come about in the content of the art? Does it necessarily reflect different cultural backgrounds between Paleolithic and Levantine art producers? Or could it result from endogenous factors? I will come back to this matter later in the chapter. A focus on the compositional patterns reveals that LRA sites are dominated by narrative scenes involving new ways of arranging motifs within compositions to create a sense of motion or action. According to Lenssen-Erz (1992), action is the defining element of a scene in prehistoric art. In the same vein, Dobrez (2012: 1839) suggests that a scene is best described as a visual narrative in which the motifs in the composition are involved in doing something, as implied by multiple visual markers in both the motifs and their patterns of association. A scene could then be defined as “a visual description of an activity.” Stylistic analysis of motifs interacting may reveal single or several events in the construction of a scene, which can result either from being represented as such originally, or from adding figures to a previous creative event (Domingo 2012b; May and Domingo 2010; Sebastian 1993). Direction of motion and individual and group animation also play crucial roles in the identification of scenes (Domingo 2005: 126). It could be argued that in Europe, the appearance of scenic compositions predates LRA. In fact, a subtle increase in the number of scenic associations has already been detected at the end of the Upper Paleolithic, during the Magdalenian period (16500–11800 BP / 20000–13500 cal BP). However, these early scenes mostly reproduce the ecological behavior of certain animal species. Only a very small number of compositions (most of them on pieces of portable art) begin to introduce scenic associations of both anthropomorphic figures and animals, like those in the scenes from Bruniquel, Château des Eyzies, Laugerie-Basse, Mas d’Azil, Péchialet, Raymonden, Roc-de-Sers, La Vache (Duhard 1996), and Hort de la Boquera (Domingo et al. 2019), to name a few. However, while Paleolithic subjects are mostly in a rigid pose or display only partial movement (Dobrez 2012), LRA portrays humans and animals in dynamic action. Now many elements of the motifs and their spatial relationships (as suggested by their position, direction, and planes of representation) imply motion, resulting in the creation of a sense of movement and sometimes even speed (Figure 16.2). To convey a feeling of movement, Levantine artists introduced new patterns, such as linear, parallel, staggered, and radial

NEW INSIGHTS INTO THE ANALYSIS OF LEVANTINE ROCK ART SCENES

243

FIGURE 16.2. Archer from la Covatina site (Vilafranca). Motion is suggested by the exaggerated length of the figure’s stride, the sense of wind blowing as suggested by the figure’s clothing and the diagonal plane of representation. Copyright Inés Domingo Sanz.

distributions in consecutive, confronted, or opposed arrangements, on a combination of horizontal, diagonal or even vertical planes. These patterns are also used to illustrate the dimensions of the space where the action depicted takes place. Floor lines are rarely depicted, but they are implied by the spatial distribution of the motifs. Occasionally, scenes include animal tracks or blood trails that explicitly show the pathways of movement. Based on these general features and in contrast to Paleolithic Art, LRA scenes can be described as action paintings or snapshots of actions. In fact, action and motion became so important that some LRA scenes even combine snapshots of the same action to illustrate different events from the same story, suggesting sequential moments in time. For example, a hunting scene at Abric del Barranc de la Palla (Tormos, Alacant) depicts a hunter following an animal blood track and shooting the animal, and the wounded animal bleeding, vomiting, and falling at the same time (Villaverde 2005; Villaverde et al. 2012; also see Villaverde this volume). Similarly, a gamedrive hunting scene at Cova dels Cavalls depicts hunters driving the game toward other waiting hunters; the waiting hunters shooting arrows at the prey; and then the wounded prey—much as in a cinematographic sequence (see Figure 16.5d later in the chapter). This sort of record of the path or sequence of the completion of an action seems to deal with the temporal dimension of narrative, which is crucial to telling a story and visually illustrating the effect of time in a specific action, creating an illusion of time passing. To an uninformed outsider, these scenes are so explicit that they look like the sort of storyboarding used to prepare for an activity or recall it once it is done. But although the literalism of the actions depicted is unquestionable, it is hard to tell whether they are illustrations of real or symbolic worlds. The central role of humans in LRA, whether as a single individual or as a group, is present throughout the sequence in a variety of scenes that are economic, social, or cultural in nature. The panels are dominated

244

INÉS DOMINGO SANZ

by narratives of hunting (mainly for big game, including deer, aurochs, wild boars, and wild goats), violence and warfare (single individuals with arrows embedded in their bodies, warrior parades, and well-structured battles), marches and dances, honey hunting, maternity, and death, with humans exhibiting multifarious anthropogenic items (weaponry, bags and baskets, dresses, ornaments) and behaviors not previously seen in prehistoric art in this part of the world (Beltran 1982; Domingo 2006; García, Collado, and Nash 2012; Nash 2005; Villaverde et al. 2012; etc.). More interestingly, when scenes, whether ritual or genre, are analyzed diachronically, these visual narratives talk about changes in social dynamics and cultural practices, providing unique insights into past human behaviors (Domingo 2006, 2008, 2012a, 2012b; MartínezRubio 2011; Mateo Saura 2006; Utrilla and Bea 2015; Utrilla and Martínez Bea 2007; Villaverde, Guillem, and Martínez 2006). Temporal and spatial variations affected not only the way LRA depicted humans (with variations in size, proportions, techniques, etc.) as gradually evolving toward less naturalistic, proportioned, and smaller human depictions, but also the activities of the humans engaged in the scenes. Systematic analysis of superimpositions has enabled identification of a minimum of five well-defined artistic phases (Domingo 2005, 2006; Mateo Saura 2006; Utrilla and Martínez Bea 2007) (Figure 16.3). For example, in the ValltortaGasulla region, the first stylistic episode (or Centelles horizon) includes large, naturalistic human figures (with average height between 20 and 40 cm) with proportioned bodies, all sorts of adornments, and sex and age differentiations (men, women, and children depicted). They appear in scenes focusing on different sorts of social activities (humans marching, maternity scenes, wounded individuals, death scenes) (Domingo 2006, 2008). In this first stage, however, hunting scenes, which are usually considered the central subject matter of this tradition, are absent from the panels. Hunting is depicted from the second (or Civil) phase onward; the human figures are similar in size, but their bodies are more elongated than in the previous phase (for details on the main characteristics of the different phases see Domingo 2008). Once hunting becomes part of the Levantine art repertoire, the rock art shows an evolution from an initial focus on deer hunting (during the

FIGURE 16.3. Las Monteses site (Jalance). Is the figure on the left holding a hafted stone axe? Copyright Inés Domingo Sanz.

NEW INSIGHTS INTO THE ANALYSIS OF LEVANTINE ROCK ART SCENES

245

Civil phase) to a later diversification of target species (deer, wild boar, and wild goat) in subsequent phases, characterized by a downsizing of motif dimensions and volume and a new trend toward non-naturalistic proportions resulting in elongated torsos (Mas d’en Josep, Cingle, and Linear phases), at least in this region. Explicit representations of women and children, seen during the Centelles and Civil phases, disappear as the size and naturalism of the human figures decreases. An increase in violent scenes is observed at the end of the sequence, with a proliferation of explicit battles waged between two distinct warring factions. In the past few decades, the increase in regional approaches to LRA has led to a more global understanding of the content of this art. Now it is widely accepted that LRA is more than a global tradition focused primarily on hunting. Rather, it is a network of artistic territories involved in continuous processes of change and adaptation, with hunting missing in the initial stage but appearing in panels, along with a changing variety of subject matter, in the later phases. While some shared features are seen throughout the Levantine territory, reflecting long-distance social interactions, others are specific to particular regions, reflecting more regional identities (Domingo 2012a, 2012b; Utrilla and Martínez Bea 2007; Villaverde et al. 2012). Despite progress in the definition of LRA, debates on the chronology and life ways of LRA producers are still open.

Chronology and Sequence Mediterranean Iberia is home to several bodies of prehistoric rock art assemblages containing masterpieces of many generations of artists. Also preserved alongside LRA are examples of Paleolithic art (PA) and two other post-paleolithic (or Holocene) rock art traditions known as Macroschematic art (MA) and Schematic art (SA) (Figure 16.1). Spread throughout Europe and focused primarily on the animal world in non-scenic compositions, PA involves a combination of figurative art (naturalistic animals and few, ambiguous representations of humans) and symbols dating to the European Upper Paleolithic and lasting up until 10,000 BP (11,500 cal BP) (García-Díez and Vaquero 2006; Román et al. 2016; Román and Domingo 2017). As for LRA, MA, and SA, they share landscapes and even open-air rock shelters, but they differ in content and distribution. Macroschematic art focuses on large, schematic anthropomorphic figures (up to 1 m high) and geometric forms (meander-like forms, zigzag patterns, etc.) arranged on panels in non-scenic compositions. The images in this tradition differ in size and content from those in Schematic art, and the tradition itself is only found in the central Mediterranean region, with an original focus in Alicante. Parallels between images in the rock art and others on cardial ware pottery date it to the early Neolithic (between 6700 BP or 5600 cal. BC and 5500 BP or 4358–4332 cal. BC) (Hernández 2012: 149). Immediately thereafter there emerged a phase of expansion of Macroschematic art regionally known as Ancient Schematic art (ASA), which was focused on large zigzag forms. Examples of this art form are spread over the entire Mediterranean region of Iberia (Hernández 2012, 2013; Utrilla 2013; Villaverde 2012), including, from north to south, the sites of Labarta (Huesca), los Estrechos and los Chaparros (Teruel); Cova del Civil and Cova dels Cavalls (Castelló); Marmallo IV and Cueva del Tio Modesto (Cuenca); Cueva de la Araña, Los Gineses, Abric de Roser, Barranc del Bosquet and Tortosilla (Valencia); Cueva de la Vieja and Cueva del Queso (Albacete); Cantos de la Visera (Murcia) and Tabla de Pochico (Jaen). Interestingly, together with some MA motifs, some ASA zig-zag forms have been identified under some LRA motifs at a number of sites, providing some relative chronology for some LRA phases, as I will discuss below. The sequence ends with Schematic art, which includes very simplified depictions of humans, animals, and geometric forms, only occasionally arranged in scenes. There are wide geographic and temporal variations across Iberia and Europe in Schematic Art, which in this area spans from the early Neolithic to the Bronze Age (Hernández 2003). Placing the focus on the socioeconomic background of the producers of these traditions, it appears that while Paleolithic art was produced by local hunter-gatherer societies, Macroschematic art and Schematic art were more likely to be linked to the Neolithic expansion or the transition to farming. During this transition, agriculture, animal husbandry, pottery, and other new items and technologies were introduced to Europe from the Near East (for an updated overview on the Neolithic transition in the Mediterranean see Manen, Perrin, and Guilaine 2014). Finally, it is still under discussion whether some of the Ancient Schematic art

246

INÉS DOMINGO SANZ

zigzag forms are linked to Macroschematic art (early Neolithic art) or to some geometric art produced by the last hunter-gatherers during the Epipaleolithic (the final period of the hunter-gatherer cultures, which in this region spans roughly from 13500 to 7700 cal BP) (for more details on these complex debates see Hernández and Martí 2002 or Bueno and Balbín 2016). So where in this more or less well-defined prehistoric artistic sequence is Levantine art located? Does it predate, coexist with, or follow Macroschematic art? The answer to this question is key to this chapter’s aim to establish whether the emergence of scenes in the art of this region is related to hunter-gatherer or farming communities. Originally, Levantine art was dated to the Paleolithic and described as a regional variation of FrancoCantabrian rock art (Breuil 1908; Obermaier 1916), but today it is generally accepted that differences between Paleolithic and Levantine art are due not to territorial but to temporal and sociocultural differences. However, discussions of the chronology and the sources of the producers’ livelihood are still open. These discussions are especially constrained by the problems of applying reliable chronometric dating methods to this tradition (for a full approach to the complexities of dating this art, see Fernández 2014; Martí 2003; Villaverde et al. 2012). Due to this lack of chronometric dating, debates have been addressed by way of analysis of the content— either the material culture (Cabré 1915; Fernández 2006; Galiana 1986; Jordá 1971a, 1971b, 1974, 1975a, 1975b, etc.), the subject matter of the actions depicted, or the context (Fernández 2014; Martí 2003)—or else through landscape archaeology (Cruz and Vicent 2007; Torregrosa and Galiana 2001; Villaverde et al. 2016; among others). All of these perspectives have advanced the discussion and brought focus to some of the problems, but the debate is still ongoing because all of the resulting hypotheses include some bias and can be questioned to some extent. The identity of the artists and their life ways are still being discussed, but regardless of whether they were the last local hunter-gatherers or the new Neolithic populations arriving from the Middle East, some kind of link to the Neolithic is widely accepted (Bernabéu 2002; Cruz and Vicent 2007; Fortea and Aura 1987; Hernández 2003, 2012; Martí 2003; Martí and Cabanilles 2002; Martínez and Villaverde 2002; McClure, Molina, and Bernabéu 2008; Molina, García, and García 2003; Utrilla 2000; Utrilla and Bea 2015; Utrilla and Martínez Bea 2007, among others). As previously mentioned, parallels in Early Neolithic ceramic decorations date MA to this period (Hernández 2003; Martí 2003). Superimpositions of some LRA motifs on top of MA and ASA rock art at some sites provide a minimum date for LRA, or at least for those Levantine rock art phases covering MA and ASA (Villaverde et al. 2012). On these bases, now it is broadly accepted that Levantine rock art was (at least partially) produced during the Neolithic. Keeping this in mind, current discussions focus on who the producers were and on the precise inception of this art: • Did it start during the Mesolithic (a term used in this area to refer to the last hunter-gatherer phases of the Epipaleolithic, beginning roughly about 11000 cal BP and ending with the introduction of farming), as a continuation of a rock art tradition initiated by local hunter-gatherers and continued long after the Neolithic expansion, as suggested by the superimposition of some Levantine motifs on top of Macroschematic and Ancient Schematic paintings? • Does it result instead from a situation of contact and territorial conflict between local hunter-gatherers and foreign Neolithic producers during the spread of the Neolithic culture to this area, with the local hunter-gatherers introducing LRA and the newcomers introducing MA, ASA, and SA? • Or are the authors fully Neolithic populations that, once settled in this area, were exploring and marking their new landscape with LRA to secure access to new resources? If we focus on the content, we see that no domestic animals and plants or clearly Neolithic items (such as pottery vessels, ground stone axes, and adzes or even sickles) are present in the art, except for a few controversial exceptions. On three occasions only, that is, in two hunting scenes and one enigmatic one, an oval-shaped object held in the hand of an individual has been interpreted as a potential hafted stone axe (Domingo et al. 2013; Hernández, Ferrer, and Catalá 1988; Martínez-Rubio 2006), but other interpretations are also possible, since the objects are too small to include conclusive details (Figure 16.4). But if weaponry was so important in the Levantine panels, with bows and arrows present in almost every scene, how can we explain the absence of such an innovation (ground tools) in the scenes?

NEW INSIGHTS INTO THE ANALYSIS OF LEVANTINE ROCK ART SCENES

FIGURE 16.4. Ancient hunting tactics as depicted in LRA. Copyright Inés Domingo Sanz.

247

248

INÉS DOMINGO SANZ

Moving from the individual objects to the content of the scenes, LRA calls attention to a wide range of social and economic activities. Animals are sometimes represented alone or in groups, but scenes with humans predominate. For a substantial part of the last century, the role of hunting in the Levantine panels has been stressed as key to deducing chronology. The “assumed” emphasis on wild fauna and the lack of domestic activities have been interpreted as indicators of the hunter-gatherer life ways of the producers (Alonso and Grimal 1994; Mateo Saura 2012). This is not surprising, as LRA provides a unique visual record of ancient hunting tactics deserving special attention. Evidence of stand hunting (with the hunter waiting for the prey at a concealed or elevated point), game-drive hunting (with hunters driving game toward other, waiting hunters), tracking (following animal tracks or the blood trail left by wounded game), stalking (with care to avoid being detected), chasing, or prey interception illustrate well the attention given by the authors to this activity, and their level of expertise (Figure 16.5). Both individual and cooperative hunting strategies are depicted in such detail that they look more like a fully illustrated handbook for successful hunting than a collection of territorial markers. However, as explained above, hunting activities are not the only subjects of scenes. They share the panels with a variety of subject matters illustrating other social and economic activities, including violence, warfare (with spear-wounded individuals falling or tumbling; warrior parades, sometimes with raised bows as in a dance or celebration; and other, more complex battle-scenes) and death, dancing, alignments of figures marching, honey harvesting, climbers, maternity scenes, and so on. Often humans are also involved in other unidentified actions. The challenge to identify the actions of some scenes is a further example of the com-

FIGURE 16.5. Evolution of styles and subject matters in LRA. Copyright Inés Domingo Sanz.

NEW INSIGHTS INTO THE ANALYSIS OF LEVANTINE ROCK ART SCENES

249

plexities of interpreting the art of other cultures without informants. The panels provide no clear, explicit evidence of agricultural practices or pastoralism. As previously mentioned, time and territorial behavior of the prehistoric human groups caused gradual changes and discontinuities in LRA that are especially evident in the altered style of the human figures and the changing themes in which they are involved. By exploring LRA sequentially and considering variations in the themes represented, a more complete picture of this art and the changing practices of the authors is achieved, with interesting implications for debates on the chronology and the authors of LRA. When scenes are analyzed diachronically, in light of the new diagnostic sequences, hunting activities are non-existent during the first stage, so the dominance of this activity is reduced, and the argument giving chronological value to a “dominant” theme of hunting is consequently less powerful (Villaverde et al. 2012). In other words, the dominant role of hunting scenes can no longer be the main empirical data supporting the claim that this art was pioneered by hunter-gatherers, as some researchers have suggested (Alonso and Grimal 1994). Moreover, the presence of wild resources in the archaeological levels of some pastoralist sites, such as Cinto Mariano (Juan Cabanilles et al. 2005), Fuente Flores (Juan Cabanilles and Martínez 1988), Ereta del Pedregal (Pérez-Ripoll 1990) or Cueva del Tollo (Domingo et al. 2013), dating to more recent Neolithic periods, show that game meat was still important in the economy of Neolithic groups (Hernández 2012; Villaverde 2012). Thus, the role of hunting in Neolithic societies should not be underestimated, whether as a source of food or as a prestige activity (Marti 2003). This means that hunting scenes could be depicted by either Mesolithic or Neolithic societies, leaving the discussion open. The same goes for a focus on images of violence and war. Violence is implicit in LRA from the early stages. In the first phase, or Centelles style, images of speared individuals are depicted. Determining whether these early images of isolated speared individuals refer to symbolic stories or real narratives of violence and warfare is not an easy matter. These images could also represent individual punishments (Domingo et al. 2007: 121). In Australian rock art in Arnhem Land, for example, knowledgeable informants have said that equivalent scenes are used to teach “moral lessons” and influence social behavior (Domingo and May 2008: 86). As such, these images could be used to regulate and memorialize appropriate behavior, in contrast to other (later) situations where violence and warfare are more explicitly shown in Levantine rock art, as discussed below. While I do not share this interpretation, other scenes previously identified as maternity scenes or as groups of individuals (Domingo 2005; Domingo et al. 2007) have recently been reinterpreted as scenes with possibly abducted women (Bea 2020; Utrilla and Bea 2015), and as such could be seen as further illustrations of violence at this early stage of the LRA sequence. Violence becomes even more explicit later on in the sequence, when the number of unequivocal representations of complex battles increases, especially during the last two stylistic phases (Cingle and Linear styles). Along with these combats, new scenes emerge in which a single individual is speared by a group of archers. This pattern reflects a shift in social dynamics from images mainly emphasizing social cooperation, as exemplified in both social activities and group hunting throughout the sequence, to new images also illustrating increasing competition and conflict at the end of the sequence, even though cooperation is still visible within the confronted groups. The presence of violent scenes in LRA has been analyzed in the context of both Mesolithic (Bea 2020; Mateo Saura 2000; Nash 2005; Utrilla and Bea 2015) and Neolithic societies (López 2015). What these scenes surely demonstrate is that the producers were familiar with violence and conflict. But to ignore the exact chronology and way of life of the producers is clearly to preclude analysis of this art in context and deeper study of the question of the meaning of these and other scenes. Interpretations of these scenes as literal representations of the authors’ daily lives overlook rock art’s potential to encode symbolic meanings. Are they purely representing a daily activity, a prestige activity, a historic account, a legend, or even something entirely unrelated to the activity or the motifs represented? If we cannot answer these questions, then the themes identified cannot be used to literally interpret the art and assign it to specific time periods and specific ways of life—at least, not until we have more robust evidence of the chronology of this art. Thus, alternative approaches should be explored to further advance knowledge about this unique heritage of Mediterranean Iberia. Using information deriving from another body of rock art (this one in Western Arnhem Land) in which both exogenous and endogenous variables caused changes in the content of the art through time, I aim to bring new data into existing debates in LRA studies to gain new understandings of both this millenary legacy and the appearance of scenes in prehistoric art.

250

INÉS DOMINGO SANZ

The Emergence of Scenes in Mediterranean Iberia: Exploring the Sequence of Western Arnhem Land Rock Art (Australia) as an Indirect Source for Reflection The synthesis of LRA offered so far identifies some key questions, whose answers may contribute to current debates on the authors and chronology of LRA, and thus on the timing of the emergence of scenes in prehistoric art in this part of the world: • Can a radical change from an art focused on the animal world to an art focused on humans and their cultural practices by means of narrative scenes take place within the same human group in cultural transition? • Or should we expect such a change to occur in a context of cross-cultural encounters between cultures with conflicting economic and social systems? • What happens in other traditions that emerged in a context of cross-cultural encounters between cultures with different economic and technologic backgrounds? Does the art reflect any of the innovations introduced by newcomers? The panorama of uncertainties that we still face in Mediterranean Iberia led me to seek alternative sources of information to explore the contexts of, and reasons for, humans’ shift toward an art focusing on humans and scenes in other territories; and to assess the impact of a situation of contact between significantly different cultures with significant differences in technologies and subsistence practices in the content of art. Western Arnhem Land, in Australia, provides an interesting case study by which to critically reflect on some of these questions. I am aware of the spatial and temporal distance between the two cultures and regions under study, and my intention is not to use this research for analogical reasoning, but to provide a new framework for reflecting on the different currently debated hypotheses about the producers of LRA. Western Arnhem Land boasts an enduring rock art tradition that began at least 28000 years ago (David et al. 2013) and lasted until the end of the last century. Aboriginal communities inhabited these lands throughout this period, persisting in their hunter-gatherer lifestyle until contact with Europeans. Changes in the styles and subjects of rock art as well as in material culture show both an internal evolution and regional variations of these societies over time. Interestingly, Indigenous societies in Western Arnhem Land also experienced a change in the content of art around the Pleistocene-Holocene transition, marked by rising sea levels and changed hydrological conditions (Chippindale and Taçon 1998; May et al. 2018). Here, an art of large naturalistic animals (particularly macropods), stencils, abstract designs, and non-scenic associations (Chaloupka 1993; Chippindale and Taçon 1998; Taçon and Chippindale 1994) shifted to an art focused on humans and their cultural practices. This new human-based art was not a single style; rather, it included several styles (Dynamic Figure, Post-dynamic Figure, Simple Figure and Northern Running Figure) (Chippindale and Taçon 1998: 102). According to these researchers, these art styles are related to a distinct period of time ranging roughly between 10,000 (?) to 6,000 years. They were not contemporary, and they have distinctive stylistic traits, but all were dominated by humans, motion, and narrative scenes, with a trend toward simplification of images of the human body over time. When scenes are identified, figures are arranged in domestic, ceremonial, hunting, and chasing or fighting scenes, large fighting scenes becoming more frequent in the more recent art (Taçon and Chippindale 1994). Apart from this period, and in the context of this research, a second point of interest in the rock art sequence of Western Arnhem Land is the changes observed at the end of the sequence. The most recent period in Arnhem Land rock art is known as “X-ray” art, with an estimated time frame of 3,000 years ago to the present. X-ray art includes different sorts of human and animal figures in depictions featuring their internal anatomy, as well as other stick figures, stencils, prints, and beeswax rock art (Taçon and Chippindale 1994). Together with this style, the last 400 years have also included representations of contact art, a new subject matter derived from intercultural encounters between local communities and diverse foreign populations (initially from Southeast Asia and later from Europe) on the northern coast of Australia. On panels created during this period, traditional motifs in traditional styles alternate with panels showing the imagery of contact (depictions of Macassans, Malays, Europeans, ships, guns, pipes, introduced animals, etc.) (May, Domingo, and Taçon 2015; Taçon et al. 2012). Thus, the situation of contact between cultures with different technologies not only ushered in major change in rock art’s

NEW INSIGHTS INTO THE ANALYSIS OF LEVANTINE ROCK ART SCENES

251

content and manner, but additionally brought with it new imagery representing a sample of the new items and innovations introduced by the newcomers, which then coexisted with (rather than replacing) traditional techniques and subject matters. Within this relative sequence of styles of Western Arnhem Land, the change toward an art dominated by humans and the different sub-styles developed recall the description and sequence of LRA provided in previous sections. Interestingly, and very broadly speaking, the new human-based art styles from Western Arnhem Land show an evolution from more naturalistic human figures to stick-like humans, with motion and narrative as a constant through the sequence. The most interesting point is that in Western Arnhem Land, the change toward a narrative art focusing on humans and their cultural practices takes place in the heart of a hunter-gatherer society adapting to new environmental conditions. Considering this example, it could be argued that maybe LRA could eventually have resulted also from an internal evolution of a long-lasting rock art tradition started and continued by hunter-gatherer societies in Iberia, instead of resulting from a situation of intercultural contact between hunter-gatherers and farmers during the transition to agriculture. The hypothesis of LRA being some sort of continuation of a rock art practice started by Pleistocene hunter-gatherers and continuing during the Holocene is not new. This option is still supported today by some scholars (Bueno and Balbín 2016; Mateo Saura 2012; Viñas 2012, among others). Some of them think they see some sort of continuity between the Final Paleolithic rock art engravings from Castellón (published by Guillem, Martínez, and Melià 2001; Guillem and Martínez 2009; Martínez, Guillem, and Villaverde 2009; Villaverde, Guillem, and Martínez 2006) and the Levantine engravings from Barranco Hondo (Utrilla and Villaverde 2004) and Llavería P-IV (Viñas 2012; Viñas and Sarriá 2011). In both cases, the motifs are very finely engraved but show some fundamental differences in the anatomic proportions of the animals and in the presence of humans in those considered to be of Levantine style (Barranco Hondo). Other researchers even see a potential Paleolithic reminiscence in Macroschematic art (Bueno and Balbín 2016). Other scholars, myself included, have argued that the disappearance of figurative art from the portable art record around 10000 BP (11,500 cal BP) in the region, replaced by purely geometric art, does not seem to support the suggested continuity between Paleolithic and Levantine rock art advocated by these researchers (Fullola et al. 2015; Villaverde 2012). However, analysis of the changes observed in the rock art sequence of Western Arnhem Land and the parallelism with that of Mediterranean Iberia leads us to revisit some of the arguments we have previously used to reject this hypothesis. The lack of evidence in the portable art record does not necessarily mean that LRA could not have been produced during the Mesolithic, as not all rock art traditions necessarily have parallels in other media. Moreover, it has been suggested that the production of rock art throughout the history of Arnhem Land is neither continuous nor steady, despite the continuity of hunter-gatherer life ways. The sequence of Arnhem Land rock art shows changes over time in its subjects and its manner (Chippindale and Taçon 1998: 106), and these changes do not occur at a constant rate over time; rather, some periods are more prolific than others (Taçon 1993: 113). Even accepting that Eastern Iberian artists could have had a period of minor artistic production, as the portable art record suggests, there is no reason to reject the idea that they could also have started a new figurative rock art tradition with a new focus on humans and scenes at some point during the Mesolithic, as suggested by Utrilla and Bea (2015). Thus, until we find arguments to deny the possibility of LRA being initiated by local hunter-gatherer societies that are more persuasive than the lack of parallels in the portable art record, this option remains open. Even more suggestive of such a possibility are some general similarities in subject matter between the new narratives from Western Arnhem Land and those from LRA. As described by Taçon and Chippindale, this new group of styles focused on humans and scenes draws attention to similar activities, including domestic, ceremonial, hunting, chasing, or fighting scenes (Taçon and Chippindale 1994: 219). This list of subject matters, as well as the manner in which some humans are represented or the role of motion and action in the panels, is comparable to some of the main features and themes characteristic of LRA. As such, it provides an excellent indirect source to assess the validity of interpretations offered for some of the items, motifs, and scenes identified in LRA. For example, like the Dynamic Figures art of Arnhem Land, the earliest human style in LRA (Centelles style) includes naturalistic humans with all sorts of body decorations. These include elaborated bracelets that, based on their shape and apparent solid structure, have been interpreted as Neolithic (Barciela 2005). LRA, however, is an art of flat images, with solid infill as the main technique. As such, the only way to feature the presence of a bracelet in the silhouette is by depicting a protruding shape, which

252

INÉS DOMINGO SANZ

does not necessarily imply that the bracelets were made of hard raw materials, such as stone. The volume of the protrusion signifying a bracelet in both sets of human figures, Levantine and Dynamic, is similar, so both the bracelets depicted respectively in Dynamic Figure style and in LRA could have been made of perishable materials. The parallelism in the way human figures are displayed in the panels in both artistic traditions is also surprising, with the majority of them “shown running or moving across the rock face, without further context to define the action” (May et al. 2018: 6). Also comparable is the presence in both traditions of images of violence and the increase in battle scenes in the final stages. In both arts the battlefield is recreated by placing humans in two confrontational and well-structured groups. Are these similarities enough to suggest that both traditions must have been produced by analogous societies? In other words, do these similarities necessarily reflect patterns of behavioral change within huntergatherer societies in both countries? Or could they just be universal human reactions or adaptations to external pressures, whether caused by complex environmental changes (such as those of the PleistoceneHolocene transitions) or by complex intercultural encounters (such as those taking place during the Neolithic transition in Mediterranean Iberia)? Then again, if this change was produced by two completely different societies with different subsistence practices, how can we explain the similarities between the activities depicted, especially considering that no obvious scenes, animals, or items related to agriculture or pastoralism are included in LRA? Further reflection on this matter would require extra research, including a detailed breakdown of the characteristics of each artistic style and an in-depth redefinition of the chronology and sequence of the art in each territory, but such an approach is beyond the scope of this chapter. Nevertheless, in the future it will certainly be interesting to explore whether this shift—from rock art traditions focused on the animal world to others focused on humans and their cultural practices through narrative scenes— responds to a more global phenomenon connected to changes in human cognitive evolution, behavior, and perception of the self, or to other global exogenous pressures (i.e., environmental changes). However, if we ignore the question of whether rock art is an account of daily life or a metaphor of a symbolic world, and whether the materials represented are depictions of real or imaginary items, then our interpretations will be constrained. While I certainly see parallelisms between the two bodies of art, the truth is that the complexities of the Holocene artistic sequence in Mediterranean Iberia complicate the quest for answers to this question. As I have already mentioned, Levantine motifs are depicted on top of Macroschematic art at a number of sites. These patterns suggest that some (if not all) of the Levantine phases should be dated to the Neolithic. Thus, either LRA and MA coexisted as expressions of two different cultures (hunter-gatherers and farmers) or, in contrast to Western Arnhem Land, the transition to a scenic art in Mediterranean Iberia was pioneered by fully Neolithic societies once they settled in the new territories (Hernandez 2003; Villaverde et al. 2016; among many others). Considering the inventory of similarities between LRA and Holocene art in Western Arnhem Land, could we accept that two societies (hunter-gatherers and farmers) with very different social contexts, technologies and subsistence strategies could have produced such similar graphic expressions in two different parts of the world? If they did, what driving forces might have been behind these two distant societies’ sudden need for self-affirmation through a scenic art emphasizing specific human actions? If this was the case, these similarities can only be understood as one more step in the universal needs that drive human behavior, outside their sociocultural and economic context. Comparisons between the Iberian Mediterranean and the Arnhem Land sequences also provide food for thought on the second hypothesis stated at the beginning of this section: the possibility that LRA and MA resulted from a situation of contact between two cultures with conflicting interests. In Mediterranean Iberia, the analysis of the archaeological context does not well reflect the initial contact and later continuity of the last hunter-gatherers during the Neolithic, as summarized by Hernandez and Martí (2002) or Fernández (2014), among others. The archaeological record is still scant, and further research is needed to understand and clarify what happened between the last hunter-gatherers and the early farmers in this region, and especially to gauge the strength of the last hunter-gatherers’ influence in this cultural interaction. The initial confinement of Macroschematic art to the region of Alicante, matching with the Early Neolithic occupations in the area, and of LRA more inland, was very much in keeping with the idea of these two rock art traditions coexisting in time and being produced by two different societies. This hypothesis would also match

NEW INSIGHTS INTO THE ANALYSIS OF LEVANTINE ROCK ART SCENES

253

the idea of LRA being produced in the context of a changing hunter-gatherer society under changing external circumstances (environmental change in Arnhem Land vs. a situation of contact in Mediterranean Iberia) considering the similarities between LRA and Holocene art from Arnhem Land. However, if this was the case, then at some point in time Iberian Mediterranean hunter-gatherers suddenly came into contact with foreign human groups bringing in new items and technologies not previously seen in this part of the world (domestic animal and plant species, pottery, ground stone axes, sickles, wooden spoons, etc.). Shouldn’t we expect some of these innovations previously unknown in the region to be somehow reproduced in the art? In Arnhem Land, in a context of Macassans and Europeans interacting with Indigenous people in an area without rapid replacement of the local population but skilled in coexistence and interaction, the impact of these interactions is reflected in both the archaeological and the artistic record. Rock art produced during and after the contact period includes both traditional paintings (which, as evidence of cultural continuity, were used to reinforce the Indigenous identity against others) and depictions of introduced material culture (Macassan steel axes and knives, tobacco pipes, pottery, boats, smokehouses for curing trepang, European boats, handguns and rifles, airplanes, bicycles, tobacco tins, etc.), animal species (Macassan monkeys, horses, sometimes horses with riders, buffalo, sheep, goats, etc.) and even representations of the others, be they Macassan visitors or European invaders with their particular clothing (Chaloupka 1993; May, Domingo, and Taçon 2015; Taçon et al. 2012; etc.). This makes me wonder if, given a similar situation in the past, we should expect depictions of traditional subject matters (reinforcing people’s own cultural identity), together with some depictions of others as a form of appropriation of exogenous cultural elements. In any case, no explicit or uncontested depictions of introduced objects, animals, plants or activities (including farming) are obvious in LRA. Hence, the data currently available in the Levantine panels do not support a situation of contact between two very different cultures, as observed in western Arnhem Land or other places in South Africa, North America, or South America (Arenas and Martínez 2009; Berrojadil 2015; Campbell 1986; Molyneaux 1989; Ouzman 2003; Ouzman and Smith 2004; Strecker and Taboada 2004). Given the lack of introduced items in the art, and according to the data available today, it is more likely that the change to a scenic art in Iberia was introduced by either fully Mesolithic or fully Neolithic groups, rather than somewhere in between. Considering the parallelisms with the artistic sequence from Arnhem Land (a similar focus on humans, similar subject matters of the scenes, and the absence of Neolithic items) incline me to support the first hypothesis, that it was produced by a changing hunter-gatherer society. But in fact, the superimpositions of LRA over MA tip the balance even further to the second hypothesis, that LRA was produced not just during the Neolithic but by a changing Neolithic society—so long as we can further support the potential parallelism between some motifs in MA and in early Neolithic ceramics, and no other new dates or evidence prove the opposite.

Conclusions The Levantine rock art of Mediterranean Iberia is a unique rock art tradition that marks a singular change in the subject matter of prehistoric art: it was the first tradition to fix the audience’s attention on humans, human-made items, and human activities and behaviors through narrative scenes. European Paleolithic rock art was mainly an art of animals in stiff poses or partial movement, geometric signs, and very few, unrealistic humans mainly arranged in non-scenic compositions. LRA, on the contrary, was an art of humans and their cultural practices. Motion and scenic compositions dominate the panels, where new artistic patterns were introduced to deal with the temporal dimension of narratives. It has previously been argued that although the iconography of LRA neither proves nor denies the hunter-gatherer origin of this art, further explanations are needed to understand the potential causes driving such a change in the art, which as Hernandez and Martí (2002: 257) suggest, does not seem to be directly related to previous traditions, derived from external pressures, or responding to sudden endogenous changes. In any case, as mentioned before, an incipient interest in humans and scenes is already latent in Magdalenian rock art, but the gap between the end of Paleolithic art and LRA has been difficult to explain given the data available in this region, unless LRA is linked to the Neolithic transition. Interestingly, a change in prehistoric rock art that is nearly unique within Europe (i.e., the shift from Paleolithic to Levantine rock art), is not unique to

254

INÉS DOMINGO SANZ

Europe. For this reason, and being familiar with the evolution of Arnhem Land rock art through internal changes resulting from both endogenous and exogenous influences but always produced by hunter-gatherer communities, I considered it interesting to explore the evolution of rock art in this territory, to further knowledge and bring new insights to debates on the shift toward an art of scenes in prehistoric times, as well as on the chronology and the potential producers of LRA. This new pairing of human and scene, represented in Europe by LRA, marks a major shift in the evolution of visual storytelling and in humans’ perception of the self, or at least in the way humans decided to project their own images, behaviors and practices in the visual arts. Whether depicting allegories, history paintings, or genre art, this new art of scenes has the capacity to tell an entire story in a single glance. This radical change in the content of the art reveals new values and thoughts, and a new perception of the world in which humans become the center of attention. In Mediterranean Iberia, the question of whether this change follows an earlier tradition of prehistoric art focusing on the animal world through naturalistic depictions (Paleolithic art) or a Macroschematic tradition focusing on geometric forms and some schematic humans is still under discussion. Could this new way of depicting human/animal relationships produce a new exploiter/resources relationship that can only be explained by a major transition to a food-producing society? The emergence of a similar art produced by hunter-gatherer communities in Western Arnhem Land shows that a change in subsistence practices is not necessary to bring about such a change as this in art. Thus, it is worth investigating why this sudden emphasis on humans, and on scenes externalizing social facets and cultural practices, developed independently in different parts of the world at different points in time and potentially in relation to different subsistence strategies. Such a shift in the history of art may be seen as signaling a new global change—in the human mind, in human visual communication patterns, and in the way humans conceptualize themselves in art, regardless of the art producers’ subsistence practices—that needs further attention. Furthermore, the advent of a new narrative art in Mediterranean Iberia promises interesting evidence by which to explore changes in prehistoric social dynamics. A preliminary analysis suggests that LRA shows more evidence of human cooperation in the early stages and an increase in competition later on in the sequence. However, more detailed analysis must be conducted in the future if we are to expand our knowledge of prehistoric social behavior.

Acknowledgments Most of the reflections and perspectives collected in this chapter are outcomes of the project HAR201125440: “Contact Rock Art? Contributing to the Debate about the Authors of Postpaleolithic Rock Art of Mediterranean Spain from an Ethnoarcheological Perspective, Funded by the Spanish Ministry of Economy and Competitiveness (2011–2014). Inés Domingo Sanz, who has been ICREA (Institució Catalana de Reserca I Estudis Avançats) Research Professor at the Universitat de Barcelona (Dpt. of History and Archaeology) since 2010, is Vice-president of the World Archaeological Congress (WAC) (2017-2022). She was recently awarded an ERC (European Research Council) Consolidator grant (grant agreement No 819404) for her research in archaeology and rock art and has also earned various academic awards and posts, such as the Blaze O’Connor Memorial Award (WAC, Jordan, 2013), Honorary Associate Researcher at the Dep. of Archaeology, Flinders University, Australia (since 2009), and Honorary Guest Professor at HeTao University, Inner Mongolia, China (2010). She has authored 38 papers in refereed journals, 43 book chapters (19 in international volumes), 6 edited volumes, and 5 books.

References Alonso, A., and A. Grimal. 1994. “El arte levantino o el “trasiego” cronológico de un arte prehistórico.” Pyrenae 25: 51–70. Arenas, M. A., and J. L. Martínez. 2009. “Construyendo nuevas imágenes sobre los Otros en el arte rupestre andino colonial.” Revista Chilena de Antropología Visual 13: 17–36.

NEW INSIGHTS INTO THE ANALYSIS OF LEVANTINE ROCK ART SCENES

255

Barciela, V. 2005. “Vestimenta y adornos.” In Arte rupestre en la Comunidad Valenciana, ed. R. Martínez, 253–64. Valencia: Generalitat de València. Bea, M. 2012. “Representaciones infantiles en el arte Levantino.” In Estudios sobre infancia en el Mediterráneo Antiguo. Zaragoza: Prensas Universitarias de Zaragoza. ———. 2020. “When Not Everything Is as Nice as It Looks: Social Veiled Conflicts in Levantine Rock Art.” Quaternary International, 544: 12–22. Beltran, A. 1982. Rock Art of the Spanish Levant. Cambridge: Cambridge University Press. Bernabéu , J. 2002. “The Social and Symbolic Context of Neolithization: El paisaje en el Neolítico Meditarráneo.” SaguntumExtra 5: 209–33. Berrojadil, F. 2015. La vitalidad de las voces indígenas: arte rupestre del contacto y en sociedades coloniales. Mexico City: Universidad Nacional Autónoma de México. Breuil, H. 1908. “Les pintures quaternàries de la Roca del Cogul.” Butlletí del Centre Excursionista de Lleyda 10: 10–14. Bueno, P., and R. Balbín. 2016. “De cazadores a productores. Transiciones y tradiciones.” In Del neolític a l’edat del Bronze en el Mediterrani occidental. Estudis en homenatge a Bernat Martí Oliver, 465–80. València: Museu de Prehistòria de València. Cabré, J. 1915. El Arte Rupestre en España. Madrid: Comisión de Investigaciones Paleontológicas y Prehistóricas. Campbell, C. 1986. “Images of War: A Problem in San Rock Art Research.” World Archaeology 18(2): 255–67. Chaloupka, G. 1993. Journey in Time: The World’s Longest Continuing Art Tradition. Chatswood, NSW: Reed. Chippindale, C., and P. Taçon. 1998. “The Many Ways of Dating Arnhem Land Rock Art.” In The Archaeology of Rock Art, ed. C. Chippindale and P. Taçon, 90–111. Cambridge: Cambridge University Press. Cruz, M., and J. Vicent. 2007. “Rock Art as an Archaeological and Social Indicator: The Neolithisation of the Iberian Peninsula.” Journal of Anthropological Archaeology 26: 676–97. David, B., B. Barker, F. Petchey, J. J. Delannoy, J. M. Geneste, C. Rowe, M. Eccleston, L. Lamb, and R. Whear. 2013. “A 28,000Year Old Excavated Painted Rock from Nawarla Gabarnmang, Northern Australia.” Journal of Archaeological Science 40: 2493–501. Dobrez, L. 2012. “Towards a More Rigorous Definition of Terms: Are There Scenes in European Paleolithic art?” In Pleistocene art of the World. Proceedings of the IFRAO Congress, Tarascon-sur-Ariège, September 2010, ed. J. Clottes, 1837–51. Domingo, I. 2005. Técnica y ejecución de la figura en el Arte Rupestre Levantino. Hacia una definición actualizada del concepto de estilo: validez y limitaciones. València: Servei de Publicacions de la Universitat de València. ———. 2006. “La figura humana, paradigma de continuidad y cambio en el Arte Rupestre Levantino.” Archivo de Prehistoria Levantina 26: 161–91. ———. 2008. “From the Form to the Artists: Changing Identities in Levantine Rock Art (Spain).” In Archaeologies of Art: Time, Place and Identity, ed. I. Domingo Sanz, D. Fiore, and S. May, 99–129. Walnut Creek, CA: Left Coast Press. ———. 2012a. “Human Figures, Techniques and Territories: Towards a Technical Redefinition of Levantine Rock Art.” In The Levantine Question: Post-Palaeolithic Rock Art in the Iberian Peninsula, ed. J. L. García, H. Collado, and G. Nash, 117–44. Budapest: Archaeolingua. ———. 2012b. “A Theoretical Approach to Style in Levantine Rock Art.” In A Companion to Rock Art, ed. J. McDonald and P. Veth, 306–22. Boston: Blackwell. Domingo, I., E. López-Montalvo, V. Villaverde, and R. Martínez. 2007. Los Abrigos VII, VIII y IX de les Coves de la Saltadora. Monografías del Instituto de Arte Rupestre, 2. Valencia: Generalitat Valenciana. Domingo, I., P. García-Argüelles, J. Nadal, J. M. Fullola, J. L. Lerma, and M. Cabrelles. 2019. “Humanizing European Palaeolithic Art: New Visual Evidence of Human/Bird Interactions at L’Hort de la Boquera Site (Margalef de Montsant, Tarragona, Spain).” L’Anthropologie, 123: 1–18. Domingo, I., and S. K. May. 2008. “La pintura y su simbología en las comunidades de cazadores-recolectores de la Tierra de Arnhem (Territorio del Norte, Australia).” In Mundos Tribales. Una visión etnoarqueológica, ed. J. Salazar, I. Domingo, J. Azkárraga, and H. Bonet, 78–91. Valencia: Museu de Prehistoria. Domingo, I., B. Rives, D. Román, and R. Rubio. 2013. Imágenes en la piedra. Arte rupestre en el abrigo de las Monteses y su entorno (Jalance). Jalance: Ministerio de Cultura / Ayuntamiento de Jalance. Duhard, J. P. 1996. Réalisme de l’image masculine paleolithique. Grenoble: Jérôme Millon. Fernández, J. 2006. “Las flechas en el arte levantino: aportaciones desde el análisis de los proyectiles del registro arqueológico del riu de les Coves (Alt MAestrat, Castelló).” A.P.L. 26: 101–59. ———. 2014. “Traditions artistiques, interactions culturelles et contextes symboliques de la transition néolithique dans la region méditerranéenne espagnole.” In La transition néolithique en Méditerranée: The Neolithic Transition in the Mediterranean, ed. C. Manen, T. Perrin, and J. Guilaine, 371–403. Errance: AEP. Fortea, F. J., and E. Aura. 1987. “Una escena de vareo en la Sarga (Alcoi). Aportaciones a los problemas del arte levantino.” A.P.L. 17: 97–120. Fullola, J. M, I. Domingo, D. Román, P. Garcia-Argüelles, M. García-Diez, and J. Nadal. 2015. “Small Seeds for Big Debates: Past and Present Contributions to Paleoart Studies from North-eastern Iberia.” In Prehistoric Art as Prehistoric Culture: Studies in Honour of Professor Rodrigo de Balbín-Behrmann, ed. P. Bueno-Ramírez and P. Bahn, 157-69. Oxford: Archaeopress Archaeology.

256

INÉS DOMINGO SANZ

Galiana, M. F. 1986. “Contribución al Arte Rupestre Levantino: análisis etnográfico de las figuras antropomorfas.” Lucentum 9: 55–87. García, J. L., H. Collado, and G. Nash. 2012. The Levantine Question: Post-Palaeolithic Rock Art in the Iberian Peninsula. Budapest: Archaeolingua. García-Diez, M. and M. Vaquero. 2006. “La variabilité graphique du Molí del Salt (Vimbodí, Catalogne, Espagne) et l’art mobilier de la fin du Paléolithique supérieur à l’est de la Péninsule Ibérique.” L’Anthropologie 110: 453–81. Guillem, P., and R. Martínez. 2009. “Arte rupestre en el Cingle del Barranc de l’Espigolar (La Serratella, Castelló).” In El arte rupestre del Arco Mediterráneo de la Península Ibérica: 10 años en la lista del Patrimonio Mundial de la UNESCO, ed. J. A. López, R. Martínez, and C. Matamoros, 35–48. València: Generalitat Valenciana. Guillem, P., R. Martínez, and F. Melià. 2001. “Hallazgo de grabados rupestres de estilo paleolítico en el norte de la provincia de Castellón: el Abric d’en Melià (Serra d’en Galceran).” Saguntum-PLAV 33: 133–140. Hernández, M. S. 2003. “Las imágenes del arte Macroesquemático.” In Arqueología e iconografía: indagar en las imágenes, ed. T. Tortosa and J. A. Santos, 41–58. Rome: L’Erma di Bretschneider. ———. 2012. “Defining Neolithic Art: Levantine, Macroschematic and Schematic Art in the Mediterranean Arc of the Iberian Peninsula.” In The Levantine Question: Post-Palaeolithic Rock Art in the Iberian Peninsula, ed. J. L. García, H. Collado, and G. Nash, 145–65. Budapest: Archaeolingua. ———. 2013. “Reflexiones sobre los artes esquemáticos entre las cuencas de los ríos Segura y Júcar.” In Arte Rupestre Esquemático en la Península Ibérica. Comarca de los Vélez, 2004, ed. J. Martínez and M. Hernández, 141–51. Almería: Ayuntamiento de Velez Blanco. Hernández, M., P. Ferrer, and E. Catalá. 1988. Arte Rupestre en Alicante. Alicante: Centre d’Estudis Contestans. Hernández, M. S., and B. Martí. 2002. “El arte rupestre de la fachada Mediterránea. Entre la tradición Epipaleolítica y la expansión Neolítica.” Zephyrus 53/54: 241–65. Jordá, F. 1971a. “Bastones de cavar, layas y arado en el Arte Rupestre Levantino.” Munibe 23: 241–48. ———. 1971b. “Los tocados de plumas en el arte rupestre Levantino.” Zephyrus 21/22: 5–75. ———. 1974 “Formas de vida económica en el Arte Rupestre Levantino.” Zephyrus 25: 209–23. ———. 1975a. “La sociedad en el Arte Rupestre Levantino.” Papeles del Laboratorio de Arqueología de Valencia 11: 159–84. ———. 1975b. “Las puntas de flecha en el Arte Levantino.” XIII Congreso Nacional de Arqueología, 219–26. Juan Cabanilles, J., and R. Martínez. 1988. “Fuente Flores (Requena, Valencia). Nuevos datos sobre el poblamiento y la economía del Neolítico valenciano.” Archivo de Prehistoria Levantina 18: 181–233. Juan Cabanilles, J., R. Martínez, T. Badal, T. Orozco, and C. Verdasco. 2005. “Un aprisco bajo abrigo en el yacimiento eneolítico de “El Cinto Mariano” (Requena, Valencia).” In III Congreso de Neolítico en la Península Ibérica: Santander, 5 a 8 de octubre de 2003, ed. P. Arias, R. Ontañon, and C. García-Mancó, 165–74. Cantabria: Universidad de Cantabria. Lenssen-Erz, T. 1992. “Coherence—A Constituent of ‘Scenes’ in Rock Art: The Transformation of Linguistic Analytical Models for the Study of Rock Paintings in Namibia.” Rock Art Research 9(2): 87–105. López, E. 2015. “Violence in Neolithic Iberia: New Readings of Levantine Rock Art.” Antiquity 89(344): 309–27. Lorblanchet, M. 1989. “From Man, to Animal and Sign in Palaeolithic Rock Art.” In Animals into Art, ed. F. Morphy, 109–43. London: Unwin Hyman. Manen, C., T. Perrin, and J. Guilaine. 2014. La transition néolithique en Méditerranée. The Neolithic Transition in the Mediterranean. Arles: Éditions Errance AEP. Marti, B. 2003. “El arte rupestre levantino y la imagen del modo de vida cazador: entre lo narrativo y los simbólico.” In Arqueología e iconografía, ed. T. Tortosa and J. A. Santos, 59–78. Italy: L’Erma di Bretschneider. Martí, B., and J. Juan-Cabanilles. 2002. “La decoració de les ceràmiques neolítiques i la seua relació amb les pintures rupestres dels abrics de la Sarga.” In La Sarga. Arte Rupestre y Territorio, ed. M. S. Hernández and J. M. Segura, 147–70. Alcoi: Ajuntament, Caja de Ahorros del Mediterráneo. Martínez, R., P. M. Guillem, and V. Villaverde. 2009. “Grabados rupestres de estilo paleolítico en el norte de Castellón.” In Arte Prehistórico al aire libre en el Sur de Europa, ed. R. de Balbín, 225–36. Documentos PAHIS, Castilla y León: Junta de Castilla y León. Martínez, R. and V. Villaverde. 2002. La Cova dels Cavalls en el Barranc de la Valltorta. Monografías del Instituto de Arte Rupestre, 1. Martínez-Rubio, T. 2006. “El cinto de las Letras (Dos Aguas). Revisió dels Panells 55 anys després de la seua publicació.” Saguntum 38: 61–91. ———. 2011. Evolució i pautes de localització de l’art rupestre postpaleolític en Millares (València) i el seu entorn geogràfic comarcal. Aproximació al territori des de l’art. Valencia: Universitat de València. Mateo Saura, M. A. 2000. “La Guerra en la vida de las comunidades Epipaleolíticas del Mediterráneo peninsular.” Era 2: 111–27. ———. 2006. “Aproximación al estudio de la figura humana en el arte rupestre levantino del Alto Segura.” Cuadernos de Arte Rupestre 3: 125–60 ———. 2012. “From Palaeolithic Art to Spanish Levantine Art: Continuity or Rupture.” In The Levantine Question: PostPalaeolithic Rock Art in the Iberian Peninsula, ed. J. L. García, H. Collado, and G. Nash, 167–86. Budapest: Archaeolingua.

NEW INSIGHTS INTO THE ANALYSIS OF LEVANTINE ROCK ART SCENES

257

May, S. K. and I. Domingo. 2010. “Making Sense of Scenes.” Rock Art Research 27(1): 35–42. May, S. K., I. Domingo, and P. Taçon. 2015. “Arte rupestre de contacto: la versión Indígena de los encuentros interculturales en el noroeste de la Tierra de Arnhem (Australia).” In La vitalidad de las voces indígenas: Arte rupestre del contacto y en sociedades coloniales, ed. F. Berrojadil, 83–106. Mexico City: Instituto de Investigaciones Estéticas de la Universidad Nacional Autónoma de México. May S. K, I. Johnston, P. Taçon, I. Domingo, and J. Goldhahn. 2018. “Early Australian Anthropomorphs: Jabiluka’s Dynamic Figure Rock Paintings.” Cambridge Archaeological Journal 28(1): 67–83. McClure, S., L. Molina, and J. Bernabéu . 2008. “Neolithic Rock Art in Context: Landscape History and the Transition to Agriculture in Mediterranean Spain.” Journal of Anthropological Archaeology 27: 326–37. Molina, L., O. García, and M. R. García. 2003. “Apuntes al marco crono-cultural del arte levantino: Neolítico vs neolitización.” Saguntum 35: 51–67. Molyneaux, B. 1989. “Post-Contact Micmac Rock Art.” In Animals into Art, ed. H. Morphy, 193–214. London: Unwin Hyman. Nash, G. 2005. “Assessing Rank and Warfare-Strategy in Prehistoric Hunter-Gatherer Society: A Study of Representational Warrior Figures in Rock Art from the Spanish Levant, Southeastern Spain.” In Warfare, Violence and Slavery in Prehistory, ed. P. Parker and I. J. N. Thorpe, 75–86. Oxford: Archaeopress. Obermaier, H. 1916. El hombre fósil. Mem. núm. 9 de la Com. De Inv. Paleontológicas y Prehistóricas. Ouzman, S. 2003. “Indigenous Images of a Colonial Exotic: Imaginings from Bushmen Southern Africa.” Before Farming: The Archaeology and Anthropology of Hunter-Gatherers 1(6): 239–56. Ouzman, S., and B. Smith. 2004. “Taking Stock: Identifying Khoekhoen Herder Rock Art in Southern Africa.” Current Anthropology 45(4): 499–526. Pérez-Ripoll, M. 1990. “La ganadería y la caza en la Ereta del Pedregal (Navarrés, Valencia).” Archivo de Prehistoria Levantina 10: 223–39. Román, D., and I. Domingo. 2017. “The End of the Upper Palaeolithic in Castellón: A Key Territory to Understand the End of the Pleistocene in Mediterranean Iberia.” Pyrenae 48(1): 47–70. Román, D., J. Nadal, I. Domingo, P. García-Argüelles, L. Lloberas, and J. M. Fullola. 2016. “La fin du Paléolithique dans la Catalogne méridionale ibérique revisitée: nouvelles réponses pour anciennes questions.” L’Anthropologie 120: 610–28. Sebastián, A. 1993. “Estudio sobre la composición en el Arte Levantino.” PhD Dissertation. València: Universitat de València. Strecker, M., and F. Taboada. 2004. “Aymara Rock Art of Lake Titicaca.” Rock Art Research 21(2): 111–25. Taçon, P. 1993. “Regionalism in the Rock Art of Western Arnhem Land, Northern Territory.” Archaeology in Oceania 28: 112–20. Taçon, P., and C. Chippindale. 1994. “Australia’s Ancient Warriors: Changing Depictions of Fighting in the Rock Art of Arnhem Land, N.T.” Cambridge Archaeological Journal 4(2): 211–48. Taçon, P., J. Ross, A. Paterson, and S. May. 2012. “Picturing Change and Changing Pictures: Contact Period Rock Art of Australia.” In A Companion to Rock Art, ed. J. McDonald and P. Veth, 420–36. Boston: Blackwell. Torregrosa, P., and M. F. Galiana. 2001. “El arte esquemático del levante peninsular: Una aproximación a su dimensión temporal.” Millars 14: 153–98. Utrilla, P. 2000. El Arte Rupestre en Aragón. Zaragoza: CAI00. ———. 2013. “Arte Esquemático en la cuenca del Ebro 2: Extensión, paralelos miebles y yacimientos asociados.” In Arte Rupestre Esquemático en la Península Ibérica. Comarca de los Vélez, 2004, ed. J. Martínez and M. Hernández, 223–41. Almería: Ayuntamiento de Velez Blanco. Utrilla, P., and M. Bea. 2015. “Los ‘paquípodos’: su difícil encaje en la cronología del arte Levantino.” Cuadernos de Prehistoria y Arqueología UAM 41: 127–46. Utrilla, P., and M. Martínez Bea. 2007. “La figura humana en el arte Levantino aragonés.” Cuadernos de arte rupestre 4: 163–205. Utrilla, P., and V. Villaverde. 2004. Los grabados levantinos del Barranco Hondo (Castellote, Teruel). Zaragoza: Gobierno de Aragón. Villaverde, V. 2005. “Entre la narración y el simbolismo.” In Arte rupestre en la Comunidad Valenciana, ed. R. Martinez, 197–226. Valencia: Generalitat Valenciana. ———. 2012. “La cronología del arte Levantino.” In Actas de las Jornadas. Abrigo de Tortosilla. 100 aniversario de su descubrimiento, ed. I. Domingo, R. Rubio, and B. Rives, 39–45. Valencia: Ayuntamiento de Ayora-Diputación de Valencia. ———. 2015. “Palaeolithic Art in the Iberian Mediterranean Region: Characteristics and Territorial Variation.” In Prehistoric Art as Prehistoric Culture: Studies in Honour of Professor Rodrigo de Balbín-Behrmann, ed. P. Bueno and P. Bahn, 145–55. Oxford: Archaeopress. Villaverde, V., P. M. Guillem, and R. Martínez. 2006. “El horizonte gráfico Centelles y su posición en la secuencia del arte Levantino del Maestrazgo.” Zephyrus 59: 181–98. Villaverde, V., R. Martínez, P. M. Guillem, E. López, and I. Domingo. 2012. “What Do We Mean by Levantine Rock Art?” In The Levantine Question: Post-Palaeolithic Rock Art in the Iberian Peninsula, ed. J. L. García, H. Collado, and G. Nash, 81–115. Budapest: Archaeolingua. Villaverde, V., T. Martínez, P. M. Guillem, R. Martínez, and J. A. Martínez. 2016. “Arte rupestre y hábitat en la prehistoria del Riu de les Coves. Aproximación a la cronología del Arte Levantino a través de la red de caminos óptimos.” TV SIP 119: 501–20.

258

INÉS DOMINGO SANZ

Viñas, R. 2012. “Superimpositions in Spanish Levantine Rock Art: Previous Proposals and New Evidence for a Reassessment.” In The Levantine Question: Post-Palaeolithic Rock Art in the Iberian Peninsula, ed. J. L. García, H. Collado, and G. Nash, 55–80. Budapest: Archaeolingua. Viñas, R., and E. Sarrià. 2011. “Documentació dels nous conjunts d’art rupestre del Priorat (Tarragona).” Tribuna d’Arqueologia 25(2009/2010): 53–84.

==== •

•

•

17. RULES OF ORDERING AND GROUPING IN THE PITOTI, THE LATER PREHISTORIC ROCK ENGRAVINGS OF VALCAMONICA (BS), ITALY From Solitary Figures through Clusters, Graphic Groups, and Scenes to Narrative Craig Alexander, Alberto Marretta, Thomas Huet, and Christopher Chippindale

Introduction: Recognizing Scenes in Rock Art, in Valcamonica or Elsewhere Understanding when several individual figures in rock art actually should be taken together as depicting a scene seems an obvious aspect of ancient pictures to explore, yet it seems to happen only rarely. In this chapter we study the particulars of Valcamonica, a valley near Brescia in northern Italy that holds some of the largest and best known later prehistoric rock art groupings in Europe, but the issues we address are general and universal in rock art studies. The major specifics of the pitoti rock art of Valcamonica are well known and published in an expansive literature that includes monographs on individual surfaces or groups of surfaces (e.g., Naquane [Anati 1960b] and Seradina 12 [Marretta 2018]), and syntheses of the whole entity in its European context by Emmanuel Anati (1960a, 1960b, 1963, 1964, 1982, 1990, 2004), as well as treatments of many individual aspects, especially chronology (e.g., Anati 1963; Pause 1992), and opinions on just what distinct motifs depict or may stand for (e.g., Sansoni 1977; Martinotti 2009). Meanwhile, recognizing what constitutes a “scene” in rock art has seldom been addressed explicitly, either with regard to specific examples (e.g., Davenport and Jochim 1988) or in general (e.g., Lenssen-Erz 1992)—hence the need for the present volume. It does not much directly address scenes in Valcamonica, even though the figures on some surfaces like the statue-stele (Figure 17.4 below) are evidently ordered by some kind of guiding grammar of graphic rules. We are still in a land of “Gaze—or gawp!—and guess.” Variously influenced, including by Hagen’s notion of the “picture problem” (Hagen 1986), and the shape and graphic grammar formalisms first developed by Stiny (1975), we have tried to build afresh. Although we see our study as our own and original, not consciously applying or translating others’ frames of thinking, it is sure to have precedents and analogues in approaches to rock art or related graphic issues.

Issues in Defining Terms Useful for Valcamonica: Figures, Surfaces, Pictures, Points of View, Manner in Which Figures Are Placed We refer to the figures as pitoti, their name in the valley’s historic dialect, rather than “rock engravings” or “petroglyphs” (US English), incisioni rupestri (Italian), or gravures rupestres (French). A typical Valcamonica surface carrying pitoti is an open-air slab of rock nestled in a hillside, sloping or undulating, more or less smooth where its hard surface has been scraped and polished by ancient ice sheets. The dimensions of such a slab are generally a matter of meters. It holds a set of markings cut into the surface to a depth of a few or many millimeters, usually by hammering into the rock surface, leaving “peck-marks”: distinct areas of dots, lines, or shapes whose rough texture makes them visibly different. The dimensions of such figures typically

260

CRAIG ALEXANDER, ALBERTO MARRETTA, THOMAS HUET, AND CHRISTOPHER CHIPPINDALE

range from a few centimeters to a few tens of centimeters. Usually these markings are either separate or, where they overlap, are of sufficiently different surface texture that we can recognize separate figures—as many as several hundred on the most decorated rocks. The majority of the figures are of prehistoric date, from the Neolithic through the Chalcolithic (Copper Age) and into the pre-Roman Iron Age. Overall, the bulk of the figures come from the very end of this extended period—from, say 4500 BC up to the very last centuries BC. A smaller set of figures belong to the medieval period, when the custom of marking the rocks was revived. We can define a “figure” as simply a confined zone of lines or an area cut into the rock that is separated from other figures by untouched rock1 (we do not follow the very separate definition of a figure as being an image, defined here as a shape distinctive of, say, an animal or a human). Generally, a figure is continuous, and the individual cuts or peck-marks that create it are of a consistent depth (for drawings of some surfaces, see Figures 17.1–17.5). Where elements run into each other, we can often observe similarity or difference in the depth and in the pecking that helps us decide if both parts belong to a single figure. If not, they may be separate figures running into each other, or may represent the superposition of one figure over another at different times. Sometimes the nature of the subject reveals a difference in date: for example, an image of a halberd or an archaic metal dagger known to be of an early date in the Chalcolithic period (Copper Age) may be found near an image that is representative of a later date, such as a human figure with spear, shield, and helmet reminiscent of forms associated with the Iron Age, more than twenty centuries later. We can define a pitoti surface as an area of rock with pitoti, defined by a distinct edge where the rock ends or enclosed by other areas of rock that do not have pitoti. Sometimes, on a Valcamonica surface crowded with many figures, close inspection shows there are fewer superpositions and overlaps between figures than one would expect: it looks as though the figures’ form, or their placement, or both, were chosen so as to avoid overlaps with existing figures. We call this a “weak avoidance rule”—avoidance because figures are evidently avoiding each other, but weak because the rule is not absolute—figures do sometimes overlap. Other, “stronger” rules like the conventional orientation of figures or their relative scale may be more significant. A simple modeling of this apparent “avoidance” habit was carried out many years ago for figured surfaces in Mont Bego, another major Alpine site, near Tende in the French Maritime Alps (Chippindale 1988). The rock art there has a limited range of motifs that mostly also occur in Valcamonica, but the majority of those in Valcamonica are absent from Bego. This small study both confirmed this effect is in play and began to define the “weakness” or “strength” of the avoidance convention (Chippindale 1988). A fuller study with more advanced mathematical modeling is required for Valcamonica. In Valcamonica, we can confidently recognize many figures as pictures, that is, as images of actual real-world objects: daggers, axes, deer, dogs, horses, people. A moment’s thought will tell us this is because the figures are the same shape as the objects they depict, despite substantial differences in size, color, materials, and the extent of their flatness or three-dimensionality. Thus the most essential of these several variables is shape. A trivial but relevant analogue from our own time is tiny, usually pink plastic elephant forms in which we freeze water, creating ice shapes for use as novelties in drinks. As long as they are the same shape as elephants (or hippopotamuses or kangaroos or whatever), they are immediately recognizable. Further, certain points of view are preferred in the interests of a clear image: only one side of an animal, for instance, is depicted because the view from the side makes all four limbs clear, in contrast to, say, the view from the front or from above. Perspective—the dominant Western system of representation in recent centuries—seeks to present elements of a 3-dimensional subject from a single viewpoint. Valcamonica art, and other depictive systems in later prehistory and protohistory generally, prefer a “flat” system of representation in which different points of view are added up to create the figure. This is usefully called “aspective,” in opposition to “perspective” (Huth 2011). Here we use the word aspective and consider it fitting, where different aspects or points of view exist within a single figure. We are not aware of use of aspective in the context of prehistoric Europe; however, Egyptologists, noticing the same habit in classical Egyptian art, have been using it for many years (Brunner-Traut 1974). Moreover—and again, as is common in regional rock art traditions—distinctive features identify each kind of animal. In Valcamonica, the dog’s mouth is open and its teeth are shown; also, the dog’s tail is raised and curving forward over the animal’s back. Indeed, the tail of a real dog does just that (in dogs whose tail forms are not modified by selective breeding). Unmodified dogs’ tails are shown raised and curved forward in,

RULES OF ORDERING AND GROUPING IN THE PITOTI

261

for instance, Classical images dated to times close to those of the Iron Age figures in Valcamonica. Should the tail instead curve down and back, the dog may be identified as a wolf or fox. Finally, in this regard, the habits that we can call “picture conventions” do not always closely follow the real-world form. We call this conventional relation between the representation and the represented concept conventional iconicity: it is not necessarily mimetic, that is, closely modeling the real-world form. The antlers of Valcamonica deer are often shown as two parallel lines, each with a series of straight, short, diagonal parallel lines diverging from that main “stem” so that clearly, neither the actual shape of the main antler nor the way that deer antlers divide into side tines is mimetic. Instead, a Valcamonica convention determines how antlers are shown. Some figures take a distinct form, often a repeated form, that we do not recognize with confidence as having the distinctive form of a real-world object. If they are pictures, we do not recognize the real-world objects they depict. Chief amongst these, ever since the first of them was reported at Bedolina (Turconi 1998), are complex figures that resemble maps (see Pilar Utrilla, this volume) of our own era, including lines, areas completely pecked, and other areas with regularly spaced dots. One can read these features as boundaries, paths, streams, and fields (Arca 2004); similarly, areas with dots of different shapes may correspond to different uses of fields in a farming landscape. We do not address these in this essay. Many physical objects are depicted with such clarity that distinctive forms already known in documented and archaeologically dated objects can be recognized in the art, notably in the Copper Age and Iron Age: that is, they have naturalistic iconicity. Other subjects in the art, such as deer, dogs, and human beings (if the humans are not shown with objects of distinctive type and date, such as weapons or helmets), were surely present in the valley over long periods, or even over the whole period in which the ancient art was made. As a strong basis for dating is lacking, but typo-chronology and superimpositions may help to draw a chronological frame. Some years ago, Paul Taçon and Christopher Chippindale (1998) found it useful to avoid the usual but confusing term “style,” a once-clear word made obscure by having acquired too many different meanings. Instead of referring to “style,” they defined a figure or picture in terms of its subject, the particular manner in which it is depicted, and the technique used to make it. These distinctions are useful in Valcamonica, where human figures, for example, are depicted in many different manners. Valcamonica rock art research has focused mainly on typologies of subject—daggers, deer, etc. In this chapter, we concentrate instead on typologies of placement (situation or position). The two approaches are complementary, so here we explore groupings of figures in terms of individual figures and their “spatial order,” that is, the particular manner in which figures are placed on the figured surface in relation to each other. One of these is a scene (see below for our definition), but we can recognize other kinds of order in where figures are placed, and we here attempt to resolve these into distinct classes.

Working Definitions for This Chapter This chapter addresses how figures are related to each other, so it is not the place to explore in detail what a “figure” is. Still, some working definition is essential here. In the case of pictures of creatures, we can define a figure as the animal we recognize as depicted: a deer, a dog, a horse, a bird, an unidentified quadruped; or a human; or an object such as a halberd or dagger. We call these “simple figures” when they refer to the being or the object alone. Something equivalent holds for other subjects or geometric shapes. A picture of a human figure may incorporate objects they are holding or wearing: a sword, a shield, articles of clothing, a helmet. Since these can be separated from the human, the figure is a composite; thus we call these separable items the “elements” of a “composite figure.” Any element of any figure, simple or complex, can be suitably divided into components for analysis: the antlers, head, body, and legs of a deer, for example; the head (presumably iron) and shaft (presumably wood) of a spear as a single object that was made by uniting two parts in different materials; the different parts of a human body, subdivided in as many ways as are useful: maybe the head, the body, the arms, the hands, the phallus, the legs, the feet; or into other or smaller units. The degree of detail and precision in the description

262

CRAIG ALEXANDER, ALBERTO MARRETTA, THOMAS HUET, AND CHRISTOPHER CHIPPINDALE

is a priori unlimited, so ever smaller elements of the analysis can be brought into the typology: not just a blade but also the shape of the blade; not just a helmet but also the form of the helmet. We use group as a broad term for figures that occur together, whether or not they fall into one of the classes defined in this chapter. We do not use the broad word “composition,” as we find it too vague to be helpful here.

Six Kinds of Spatial Order in Valcamonica Rock Art Study of placement, position, and situation is generally seen as the subject of geography. Accordingly, there are fitting mathematical methods to define and measure these variables. Commonly, feature distributions— modeled as point distributions—are studied through first-order neighborhood analysis (e.g., geometric center, Average Nearest Neighbors, etc.) and second-order neighborhood analysis (e.g., Ripley’s K, L-Besag), and then classified into three point pattern classes: clustered, random, and regular. Non-mathematically minded readers may already have recognized these three classes: those where figures are grouped together into clusters with large empty spaces between; those where they seem scattered in a haphazard or random way, sometimes clustered, sometimes not; and those where they seem to be quite regular or tend to be evenly spaced. A precursor to this approach is Alexander (2011), which modeled and measured the pitoti on a larger socio-spatial scale. That study explored and analyzed the distribution of figured surfaces across the broader landscape, on scales of hundreds of meters or kilometers; whereas the present essay explores and analyzes the distribution of figures across a single surface measured in centimeters and meters. John Clegg (1995) used the word “mathesis” for an analytical method that measures things not normally measured. Before we can know just what to measure, we must understand what the measurement intends to report, that is, exactly which are the vital variables, the ones whose measurement will actually be informative. This study is therefore a pre-mathesis—a search for those variables. Looking ahead, therefore, for each class we propose, provisionally, some measures we hope will result in a good mathesis. The next step is to use them as a means to determine what the essential variables are, and how they can usefully be measured. When we divide groups of figures into distinct classes according to placement, we recognize six kinds of Valcamonica rock art. This chapter has limited space for illustrations, so most of the six kinds are illustrated by just one example of each. We have chosen to take most from Seradina 12, which offers the most recent records and the highest standard of conventional 2-D tracing of Valcamonica’s major surfaces (Marretta 2018). Seradina 12 is both unusual and outstanding amongst the Valcamonica surfaces in several ways: it is very large and well preserved; it has a very large number of figures, including some interesting ones; and it has some very striking placings of figures in relation to each other. We can assure the reader that our classes (except for “narrative”) arise from field knowledge of thousands of figures on hundreds of surfaces, not the single instance here shown. Many other examples of all the others can be found in the numerous drawings and clear photographs of Valcamonica surfaces published in, for example, Anati (1960b, 1982), Sansoni (1987), Sansoni and Gavaldo (1995, 2009), and Marretta, Cittadini, and Fossati (2007). The six kinds of Valcomonica rock art are: (1) Solitary figure: a figure that is alone on the surface (2) Figures in clusters: groupings of figures close together but apparently unrelated Beyond clusters, we recognize four modes in which figures are placed together with apparent intent: (3) Figures in arrays. (4) Figures in graphic designs. (5) Figures together making scenes, divided between contiguous-figure scenes and adjacent-figure scenes. (6) A narrative. We think it helpful to use a different name for each type of placement in this informal treatment of the subject: solitary figure, cluster, array, graphic design, scene, and narrative. Some of these terms have more precise meanings in mathematics, so, if this study were later to advance into a more formal mathematical analysis, different terminology might be needed.

RULES OF ORDERING AND GROUPING IN THE PITOTI

263

Solitary Figures Occasionally, a figure appears alone (Figure 17.1). It may be absolutely alone, as the only marking on the currently exposed surface. Solitary figures follow a consistent rule of Valcamonica art, which those grouped in any way follow too: where a surface has a slope, as most figured surfaces do, the figure is placed “upright” on the rock. Thus humans and animals are placed with their feet below and their bodies above. Inanimate objects, such as daggers and halberds, do not have this kind of “natural position,” but regularities are seen: daggers are generally placed horizontal relative to the slope; halberds are placed with the shaft vertical and the right-angled blade horizontal and at the top. This order for halberds is consistent with the few images in both Valcamonica and its ”twin” mountain site, Mont Bego in the French Maritime Alps, that show halberds held by men in this way—vertically, with the shaft held high and the blade above the holder’s head. This “upright rule” applies equally to figures in the varied kinds of groups: see FIGURE 17.1. A solitary figure. Detail of figured surface, Seradina 12. Copyright Alberto Marretta. The motif is a human, (Figures 17.1-17.6). with the manner of depiction of a common variant. Notice the In a fieldwork convention generally folfaint marks above the figure, which are perhaps eroded traces lowed in the valley, a search for further art on of pecking. It may be that this was not a solitary figure but aca surface generally stops when no marks are tually one with a companion, now weathered away, leaving the seen within one meter of a figure. Given that appearance of a solitary figure. That there are so very few solithe maximum dimensions of Valcamonica tary figures on this vast surface with its many hundred figures figures are commonly a few centimeters or underlines how rarely figures are solitary. tens of centimeters (though sometimes far larger), we propose that any kind of interaction between figures, and certainly their constituting a scene, is unlikely once figures are separated by the fieldwork convention. Therefore we propose that figures separated by one meter of unmarked rock also be seen as solitary, alongside those that occupy a surface alone. To summarize, we can define a solitary figure as a figure that is either in complete isolation, as the only figure on a rock surface, or in the comparative isolation of not being close to any other figure. Measures for solitary figures We propose the following measurements. When the figure is the only one: • Fraction of a surface covered by a solitary figure. • Distance to the center of the surface–centrality. • Distance to the border of the surface (eccentricity). When the figure is not the only one: • Distance from the next figure (always more than one meter). This could be adjusted to a multiple of figure dimensions, rather than absolute size. • Area around figure which is closer to it than any other figure. This defined zone can be used to create Thyssen polygons, and a measure of the polygons’ relative size.

264

CRAIG ALEXANDER, ALBERTO MARRETTA, THOMAS HUET, AND CHRISTOPHER CHIPPINDALE

Figure Clusters On Valcamonica surfaces, many or most figures are often grouped together; that is, many of them are concentrated in what can fairly be called clusters (Figure 17.2). Between the clusters are areas without figures. Sometimes the geological structure of the surface varies within an individual rock, being more or less smooth and regular; one thinks of those rougher zones as less attractive, less inviting, less “good” for figures. No study has been made to explore how many of the figures recorded in Valcamonica are solitary, either in the stricter definition of being the only figure on a surface, or in the looser definition of being more than 1 meter from neighboring figures. Our sense, derived from many years’ field experience in Valcamonica (largely by one of the present authors, AM), is that figures do tend strongly to cluster, rather than being often solitary. In discussions between the four of us, we were hard pressed to mention many individual surfaces with solitary figures, despite knowing of many hundreds of clusters both large and small. Our sense from experience in several other rock art regions across the world (largely of another author, CC) (Chippindale 2004; Taçon and Chippindale 1994) is that clustering is common in rock art. Two straightforward reasons for clustering in general could be, first, that several figures were made at much the same time in a single archaeological event; and second, that the presence of an existing figure prompted the artist to choose the same surface, so there were several archaeological events. In Valcamonica, one can reasonably imagine a slow colonization of bare rock by lichens, moss, leaf-fall, litter, soil, vegetation, and finally woodland that swallows up and completely conceals the surface. The zones where clusters accumulate will be those still consisting of bare rock, where there is no covering material and where, because there is no cover, a new figure is more likely to survive in the open than to be rapidly buried. Often there is evidence, either in the subject, or in the manner of depiction, or in both, that figures in a cluster are of broadly the same date. The stylistic and spatial proximities between clustered engravings are surely related to Waldo Tobler’s First Law of Geography, “everything is related to everything else, but near things are more related than distant things” (Miller 2004: 284; after Tobler 1970: 236). To summarize, we can define a figure cluster as a number of figures that are close together but seem to be independent of each other; that is, they do not show the more structured relationship evident in an array, a graphic design, a scene, or a narrative (as discussed below). They differ from solitary figures only in their placement together. Whatever the reasons for clustering, the basis in field observation is robust: the exact placement and separation of figures can be measured. Consequently, spatial statistical analysis of the degree to which figures are clustered or distributed is feasible. The spatial configuration of the rock surface, when it is alternatively suitable and not suitable for engraving, creates an inhomogeneous space. On such a surface, a Poisson point process distribution of engravings, equivalent to the normal law (or Gaussian), will produce clusters, but an “attractor rule” (see below) will also produce clusters. Since at least two different factors (Poisson law, attractor rule) can produce the same result (clusters), without any further arguments this simple explanatory model can be said to be “underdetermined.” Notice in Figure 17.2 how the cluster is surrounded by an extensive area without figures; that is, the area of the cluster is defined. There are three similar human figures within the cluster. Each figure is depicted in a similar fashion holding a spear in one outstretched arm and a round shield in the other. As discussed below, an array is defined as a group of figures at a reasonably regular spacing and thus it is possible that these three figures constitute an array. The authors are much aware that their analytical classes are still neither certain nor proven. They will need to be used with care. Their use will probably result in some adjustment of their definitions. Measures for figure clusters We propose these measurements: • • • • •

Number of figures. Number of recognized motifs. A measure of the relation of those numbers, which will be a measure of diversity within the cluster. A measure of spatial order/randomness with a technique such as Ripley’s K statistic (Ripley 1976). The fraction of the surface occupied by the convex hull of the cluster (a convex hull is the smallest convex shape containing all the figures).

RULES OF ORDERING AND GROUPING IN THE PITOTI

265

FIGURE 17.2. A figure cluster. Detail of figured surface, Seradina 12. Copyright Alberto Marretta. Notice how the cluster is surrounded by an extensive area without figures, that is, the area of the cluster is defined. Notice that within the cluster are three quite similar human figures, each depicted in much the same way, holding in one outstretched arm a spear with its blade uppermost, and in the other a round shield depicted as a semicircle. Notice (below) that an array is defined as a group of similar figures spaced reasonably regularly. Perhaps the three figures in the cluster make an array. The authors are very much aware that care will be required in the use of their analytical classes, and that some adjustment of their definitions will likely be necessary.

266

CRAIG ALEXANDER, ALBERTO MARRETTA, THOMAS HUET, AND CHRISTOPHER CHIPPINDALE

• Shape indexes, suitably chosen from the many available, such as the largest axis length (Feret diameter), the roundness, etc.

Figures in Arrays It is often noticeable that several or many examples of a figure type are concentrated in a limited area (Figure 17.3). A surface bears not a single dagger but several daggers; not one water bird but several; not a few paired warriors but many; not single footprints or sandals, but a pair matched left and right, or several pairs close together. We also notice quite a large number of figures that are consistent in both their subject, a human, and their manner of depiction: a distinct round head; a long neck; broad shoulders; the arms held out and curving down and well away from the body; the body small, round, or squarish; the legs well separated; the feet pointing out to left and right. The figures therefore appear to be standing, facing us square on—not with their feet facing in the same direction, as would suggest walking or running. Some have phalluses. One figure, bigger than the others, with large fingers and very large toes, belongs to a recognized class of Valcamonica anthropomorphs called grandi mani (“big hands,” in Italian) figures. The figures in the array are strikingly well spaced: no two touch or overlap. The few that touch other figures mostly touch a different kind of figure, not another of the type making the array. If they were placed at random, there would be both overlaps and larger areas without a member of the array, so it is evident by inspection that they are placed in considered positions, each relative to the other—a positioning defined by an avoidance rule (above). One could create some kind of an agency model: if one thinks of figures as autonomous beings—as actual actors—then one can say of them, “Figures prefer to be placed so that they do not overlap.” Other examples of arrays at well-known Valcamonica sites include the several looms on Naquane rock 1; the many human figures of the distinctive type known as oranti on Naquane rock 50; and the several water birds on a Foppe di Nadro rock. These arrays are consistently of the same motif and, in the case of humans, a particular manner of depicting a human. We call these groupings of figures “arrays.” We choose that word over, say, “groups,” because it evokes the everyday meaning of array as a set of well-spaced objects with an evident order. To summarize, we can define a figure array as a number of figures, of the same or related motifs, that are adjacent, overlapping, or otherwise in a close spatial relationship but are not clearly placed each in relation to others (as is the case for the graphic designs and scenes defined in the next sections). Whatever the reasons for an array, the basis in field observation is robust: instances where the similarity in subject and form is evident. As with a cluster, the exact separation of figures can be reliably studied and measured. Consequently, it is possible to statistically analyze both the degree to which the figures in the array are similar, and the degree of separation. Measures for arrays We propose these measurements: • Ratio of the number of the repeated image to the number of other motifs within convex hull of repeated images making the array. This is a kind of measure of its “purity,” that is, of how much the cluster is composed of that dominant motif, and of how much that dominance is “diluted” by other motifs. • The same measures as were used for a cluster (see above), as an array is a cluster that is distinguished by consisting largely of a single motif and may be more structured, given the more even spacing of the examples of that motif. • Number of figures. • Number of recognized motifs. • A measure of the relation of those numbers, which will be a measure of diversity within the cluster. • A measure of spatial order/randomness with a technique such as Ripley’s K statistic (Ripley 1976). • The fraction of the surface occupied by the convex hull of the cluster. • Shape indexes, suitably chosen from the many available, such as the largest axis length (Feret diameter), the roundness, etc.

RULES OF ORDERING AND GROUPING IN THE PITOTI

267

FIGURE 17.3. An array of figures. Detail of figured surface Seradina 12. Copyright Alberto Marretta. Amongst the figures are many images of humans depicted in the same manner: arms stretched out and curving down, legs well apart with feet turned out. These form an array. Other subjects present—a human on horseback, dogs, a spear, humans holding a spear, two opposed warriors—are not part of the array but are within the same part of the surface.

268

CRAIG ALEXANDER, ALBERTO MARRETTA, THOMAS HUET, AND CHRISTOPHER CHIPPINDALE

Figures in Graphic Designs Our next class recognizes those surfaces where the figures are not just in clusters or arrays but also adhere to a more exacting spatial order in which the positioning of figures on the surface follows clear, consistent patterns that exhibit more structure than do arrays and typically involve more than one type of rock art figure (Figure 17.4). Here we can see more controlling “rules of placement.” Their characteristic spatial order consists of repeated layouts seen on many surfaces, such as the daggers placed horizontally and closely, one above the other (seen in both examples illustrated in Figure 17.4) and the halberds placed vertically side by side—yet, every instance is in some respect different. One reason we call these “graphic designs” is that this combining of repeated conventions (or “rules”) with variation in each instance or single design rather echoes the nature of graphic design in our own culture today, as well as the practice of the graphic designers who shape them. As one of the volume editors remarked to us, modern researchers studying ancient rock art sometimes forget the attractive repetition of pattern that is so common in graphic designs of all historic periods, including our own. The aesthetic of repetition is yet another plausible base for striking aspects of ancient art.

a

b

FIGURE 17.4a–b. (a) Graphic design: statue-stele, Bagnolo 2. (After E. Anati, 1990, Fig. 224.). (b) Graphic design: rock art on bedrock, “monumental composition.” Plas Roccia 1: Capitello dei Due Pini, (After S. Casini, ed., 1994). Figures in graphic designs, examples of the two variants as distinguished in the research literature: the statue-stele on a small, rounded, portable boulder, often of broadly cylindrical shape; and the “monumental composition,” a comparable design differing only in that it is placed on bedrock, or on a large detached boulder that offers a large surface comparable to a bedrock panel.

RULES OF ORDERING AND GROUPING IN THE PITOTI

269

These conventions are vividly present in the statue-stele and on a small number of other surfaces. Figure 17.4a, a statue-stele, illustrates some of these rules. At the top, a simple geometric form—a round motif of a line with small lines projecting. To the right are several animals of similar form facing the same way and set in a vertical line. To the left is another animal, horizontal and facing the opposite way. Also on the right, two hafted objects—halberds or variants thereof—face the same way, the upper one nested in the lower one below and to one side of its partner. At center there is a more complex geometric form with concentric curved lines, projecting lines, and immediately below, two linked concentric sets of circles. Below to the right, two daggers, horizontal, both point to the left. At the bottom, a plough is drawn by two oxen facing right. We can recognize several organizing principles or rules that constitute these graphic designs: • Repetition: the same motif repeated—many examples. • Consistency in manner of depiction: the motif is repeated in very nearly the same form—many examples. • Consistency in size: the motif is repeated at nearly the same size: each type of engraving (e.g., daggers, deer, etc.) has a constant relative size (scale) with respect to the other engravings—many examples. • Verticality: the same motif is repeated as if in a vertical stack, as in the case of the animals and daggers. • Horizontality: the same motif is repeated across a horizontal line, as in the human figures. • Diagonality: the same motif is repeated across a diagonal, as in the case of the halberds. (These last three—verticality, horizontality, and diagonality–are all special cases of translation, ones that we think may be useful to keep distinct when exploring that aspect.) • Facingness: the same motif facing the same way, as in the vertical stacks of animals and daggers, and the diagonal stacks of halberds. • Symmetry, left to right: as in opposed sets of stacked daggers. • Top-to-bottom-ness: as in the case of the oval line that is always at the top of the figured area, and the case of the stacks of animals, which usually start at or near the upper edge of the surface. • Inconsistency of scale between the different motifs: as in the relative sizes of daggers (known from archaeological examples) to humans and large animals. The multi-figure groups in Casini (1994) and the more complete illustrated listing in Marretta (n.d.; unpublished corpus of Valcamonica statue-stele) show that these principles are seen repeatedly. It is possible that a closer study would identify more principles than these ten, which are so strong and standard that they are evident in an inspection of only a few tens of these graphic designs. Some but not all of these design rules are visible in the two illustrated here. The distinctive aspect of these groupings is they are ordered by a controlling graphic logic that governs what is placed in relation to what, as partly stated in those principles. None of these groupings looks at all like an image capturing how objects, animals, and people commonly place themselves in relation to each other in the real world. We must and do admit that this is a theoretical possibility. To summarize, we can define a graphic design as a number of figures of the same or related motifs, which are adjacent or overlapping, and appear in a close spatial relationship defined by geometric placing. We see that spatial order, often including some pattern of repetition, as unlikely to be expressing an actual spatial relationship in the real world. Measures for graphic designs We propose these measurements: • • • •

Number of figures. Number of recognized motifs. A measure of the relation of those numbers, which will be a measure of diversity within the cluster. Portion of surface occupied by convex hull. We expect this to be high for statue-stele; it may be lower for monumental compositions. Convex hull is clearly suitable, but there is more than one shape index that could be used (e.g., factor compactness).

One could also: • Define in suitable terms a “standard design” regarding presence/absence (not number of instances) and so create an “agreement index” for each example of a graphic design.

270

CRAIG ALEXANDER, ALBERTO MARRETTA, THOMAS HUET, AND CHRISTOPHER CHIPPINDALE

Scenes We recognize two kinds of scenes (Figure 17.5). Our defining names for them, “contiguous-figure scenes” and “adjacent-figure scenes,” are chosen to state the issue that divides scenes into two kinds. Contiguous-figure scenes Whereas the figure groups illustrated in Figure 17.4 are spatially organized by a graphic logic, those illustrated in Figure 17.5 are instead structured by their real-world relationship. In each, the subject is clear: two animals, connected by a yoke at the shoulders, pull a drawbar connected to a plough guided by a human being. We call these scenes, defining them as a group of figures whose placement on the rock we reasonably think of as closely reflecting the real-world spatial order of their component elements. We notice that, in a convention already observed in other figures, the component elements are placed in such a way that their form is clear, although in actuality, when seen from the side, the nearer of the two animals would largely obscure its twin alongside. This convention is an effective means of showing both the form of the creature and the two animals’ spatial relationship to each other, to the yoke, and to the plough’s draw-shaft. To our modern eye, the plough team appears to be seen diagonally from above. Indeed, Clarence Bicknell, a pioneer student of Alpine rock art and the first to grasp that figures of this geometry were images

FIGURE 17.5. Scenes. Detail of figured surface Seradina 12. Copyright Alberto Marretta. In this small area are several separate scenes worth describing and defining in a necessarily lengthy caption. The two scenes of ploughing (center-right and below) are contiguous-figure scenes, in which the whole scene with its leading human with mattock, the two beasts they lead, the yoke and shaft of the plough, the digging arm of the plough, and the ploughing human behind the plough are all depicted with a single continuous line. At top left are two pairs of opposed warriors, adjacent-figure scenes in which the figures are in a symmetrical opposition. At top right is a scene both symmetrical and sequential: two dogs on the left confront a dog on the right. Here, on a few square centimeters of a single surface, one sees the complexity in scenes, which is both our duty and joy as researchers to understand in a fitting rationality.

RULES OF ORDERING AND GROUPING IN THE PITOTI

271

of plough teams, himself came to this insight when he looked diagonally down from a high slope onto a team ploughing in a valley below (Bicknell 1911, 1913). Note also that the ploughmen are shown face on, so their legs and phallus are clear. Were the ploughman to be drawn from the side, in their actual position in relation to the plough and its pulling beasts, they would obscure each other. We notice in all these figures that the pecking that creates the figures is continuous in line and area, save for an occasional minute gap or a zone where weathering has removed part of the surface and figure. That is, these figures are contiguous. We call them contiguous-figure scenes. Adjacent-figure scenes A common pattern in the figures (Figure 17.5) shows a pair of human figures standing face to face, a short distance apart. The placing of the legs and feet shows they are indeed facing each other, rather than facing in the same direction as if one were following the other, or facing away from each other as if walking away from each other or standing back to back. Usually the figures carry objects in each arm; that is, each is itself a composite figure. Commonly one arm holds an object of rectangular, round, or irregular form recognizable as a shield; usually it is held out towards the body of the opposing figure. The other arm is typically raised above, or above and behind, the head, and holds a linear object held at one end, recognized as a sword, or a longer and thinner linear object grasped in the middle, recognized as a spear or lance. Reasonably, these combinations of paired figures bearing arms against each other, of which several hundred are known, are seen as fighters. Sometimes another figure, usually of rather larger size, stands behind each fighter, making a group of four ordered in two opposing pairs. Other scenes of similar clarity appear to show the hunting of deer (further discussed in the next section) in a common combination, repeated on many surfaces. In these scenes, a stag, identified by its antlers, is followed by a dog, which in turn is followed by a rider on horseback. Usually they all face the same direction, so that the spatial order is respected: hunter on horseback follows dog, which pursues deer. Other details are closely consistent. For instance, the relative heights of the actors, without matching their real-world proportions exactly, do reflect those proportions: the dog is shorter than the deer, as it is in life; the mounted rider is taller than the deer, as a rider is in life. The feet of the three actors stand on much the same baseline, consistent with them crossing the same ground surface. Never is one placed “up in the air” or “down into the ground” in relation to the other two. Again, the placing of the actors–the deer, the dog, the rider on horseback–is consistent with a real-world spatial relationship as seen from a single viewpoint, with the viewer watching the action from the side. One can also notice a refinement, or perhaps a necessity. There is only the slightest separation between rider on horseback and dog, and between dog and deer. So does this show the moment of capture, when the dog seizes the deer by its rear leg or haunches, with the human hunter right behind? We think not, for although the dog is close to the deer and has its jaws open, it rarely bites or comes into actual physical contact with the deer. Also, surely, in actual hunting the human hunter will not always be immediately behind the dog and deer. Suppose instead that the dog is several deer-lengths behind its prey, and the human hunter yet again several deer- or horse-lengths behind. If they are so separated, especially on a surface with many other figures, then the viewer might not link the deer, the dog, and the human on horseback into a scene. That is, the actual spatial relationships are modified in the depiction so as to clarify and emphasize the subject—the hunt. To summarize, we can define a figure scene as two or more figures of the same or related motifs that are contiguous, adjacent, or overlapping, and in close placement together are recognizable as reflecting a real-world spatial relationship. It is seen from either a single viewpoint or from multiple viewpoints, if components of the scene cannot be plainly shown from one viewpoint alone. The viewpoint or viewpoints chosen are such that the different elements of the scene, that is, its actors, are depicted most clearly and most distinctly one from another. These scenes can be divided into contiguous-figure and adjacent-figure scenes. Measures for scenes We propose the following measurements, which need to be slightly different for the two different kinds of scenes. Measures for contiguous-figure scenes: Given the precision/clarity of the definition, we do not immediately see quantitative measurements that might further strengthen the classification of an image into this class.

272

CRAIG ALEXANDER, ALBERTO MARRETTA, THOMAS HUET, AND CHRISTOPHER CHIPPINDALE

Measures for adjacent-figure scenes, either paired/opposed or sequential: • Index of formal similarity, based on co-presence of attributes. This should be an especially effective measure for paired warriors. • Dimensions of each figure in the scene. • Distance between figure centroids as a fraction of overall scene size. • A measure of symmetry, applying to paired/opposed figures and surely especially useful for paired warriors: the extent to which figures are reflections and rotations of one another. For example, a GPA (Generalized Procrustes Analysis) might explore such aspects as scale, rotation, and translation. In this way, it will be possible to capture and measure how much each warrior is indeed a “mirror image” of his opponent. Still, it is not yet clear which of several measures of symmetry is most suitable. • For sequential scenes, the ratio between maximum dimensions of length and height of figures in the scene (e.g., deer, dog, or mounted human in hunting scenes), in relation to their relative proportions and ratio in the real-world subjects depicted. Narrative scenes One common scene, repeated in many places and reported above, depicts a deer hunt (Figure 17.6): a stag, followed by a dog, followed by a rider on horseback. At Seradina 12 (see above) we have a variant and elaboration of the scene (Figure 17.5, upper right). It seems to be unique to this one surface. But, we remind ourselves, something similar may exist on other known surfaces without having been seen or thought of in this way, or on the surely very many surfaces largely or wholly now buried in Valcamonica and not yet known to fieldworkers. In one group of figures, a line of several deer is followed by a line of several dogs, which is followed by a line of several riders with spears on horseback. The four deer are very similar in form. After them is a gap, then four dogs, again closely similar in form and close together. The gap between the deer and the dogs reveals only the left-hand, rear part of an animal, and the front part is missing—it was not completed or was lost where the rock flaked and weathered away. Since the rear bodies, legs, and tails of deer and dogs are shown in similar ways, it is unclear whether this is another deer or another dog. We think it more likely to have been a dog because it is close to the other dogs and decidedly separated from the deer. Altogether, then, there are fourteen actors that together make a single line that runs diagonally across and up the surface.

FIGURE 17.6. Narrative? Detail of figured surface Seradina 12. Copyright Alberto Marretta. Two adjacent-figure scenes: deer-hunting narratives. Also other figures variously placed, some in configurations that our other five classes may usefully define.

RULES OF ORDERING AND GROUPING IN THE PITOTI

273

Below and to the left, we see much the same combination, again slightly weathered so the exact number of animals is not fully clear. We see three deer, then three dogs. Next is the front legs and head of an unclear animal that we think, judging from its form and spacing, is another dog, making four dogs altogether. After that come the rear body and back legs with scratched lines above of another unclear animal that we think, based on its form and spacing, is a rider with a spear on horseback. Finally, there are two clear riders with spears on horseback—three riders altogether—bringing the total number of actors to ten. All are in the same order, with a larger gap between each animal group. All are on the same ground-line running diagonally across and up. Further below is another, less clear, repetition of the group. In it, again, are several deer, dogs, and riders with spears on horseback. Again, all are in the same order. There is a larger gap between each animal group than there is between animals in the same group. All are on the same ground-line, diagonally across and up. Read as a single scene, each shows a larger hunt. Several mounted hunters, each with about one dog, have found and are pursuing about the same number of deer. But they are on the same ground-line, and each neatly follows behind the other, as if walking on a modern tarmac path. Surely, an actual ancient hunt through woodland or across meadow would be a more scattered and disorderly affair, not a neat line of seemingly domesticated animals marching as if in a circus procession! Thus, we propose and prefer a different explanation of this singular occurrence: that it is a narrative, depicting a scene at successive moments in time: one deer is tracked by one dog, which is followed by a mounted hunter, and the figure depicts their motions across the landscape as they would be seen in several successive moments. Another striking aspect of this configuration of up to fourteen figures is that it is shown three times on this same surface. The sequence of figures is always the same; the relative proportions and consistent detail is much the same; even the overall size and length of the grouping is quite similar. This, we believe, is the first time a narrative has been proposed as a persuasive account of why a number of Valcamonica figures considered together take a particular form. Of the several words that could be used, we prefer “narrative,” given its focus on how a sequence of events develops in time. Over time, the hunt moves across the landscape and is duly depicted moving across the rock surface. To be sure, it is a simple narrative of the middle period of the hunt, that is, the pursuit phase in which not much changes, as opposed to the start, when the deer is flushed from cover and the pursuit begins, or the end, when the deer is caught and killed or else escapes. Our late colleague Frederick Baker, himself interested in precursors to the invention of cinema near the turn of the twentieth century, liked to call this “proto-cinema,” an attractive, vivid term we prefer not to use. We see the hunting narrative’s similarity to the sequential projection of frames in a film as a matter of simple coincidence. The better modern analogy is the cartoon strip, in which each frame shows the story at the next stage in its progress. In 2012, the innovative • P • I • T • O • T • I • project (Chippindale and Baker 2012) explored aspects of rock art from prehistoric Europe, using techniques from digital media and even treating pitoti as if they were modern digital images. It used Valcamonica for all its working examples. One aspect it explored was to treat images like stills from an animated cartoon, in a way that somewhat echoes Marc Azéma’s (2008, 2009, 2010; see also Nowell 2015) exploration likening Palaeolithic art to cartoons (“bande dessinée” in French); however, the • P • I • T • O • T • I • team developed its approach independently of Azéma. Despite agreeing with both Baker and Azéma that there are intriguing graphic parallels, we are unaware of any basis for Azéma’s claim that Palaeolithic art is either the “prehistory of cinema” or the “origin of cinematography” (Azéma 2011). Rather, we see Palaeolithic art, post-glacial art (like that found in Valcamonica), and some kinds of cinema as similar solutions arrived at independently in response to similar picture problems, to use Hagen’s way of thinking. This specific picture problem, then, is how the passage of time can be represented using images that are themselves static and unchanging through time. One answer is to represent the passage of time with a set of images, each static and unchanging through time, but in a sequence that captures movement and change. In this respect, we notice that scenes of the hunt are invariably of the pursuit. Not one scene is known that shows the beginning or the end of the hunt. We also notice that the same choice of the middle phase likewise applies in the many hundreds of scenes of fighters. They are never shown limbering up at a distance,

274

CRAIG ALEXANDER, ALBERTO MARRETTA, THOMAS HUET, AND CHRISTOPHER CHIPPINDALE

or not yet holding their weapons, and there is only a single known example that seems to show the end of the combat, with the loser prostrate and run though by a spear. Instead, the fight is shown at the middle phase: the two fighters are face to face and very close, their shields up, and their swords or spears raised overhead, ready to strike. In his chapter in this book, Villaverde presents a similar idea developed independently from ourselves: that what can be read as single scenes may actually represent successions of events—the same notion that the figures depict a narrative. Measures for narratives We propose these measurements: • Number of image elements in each “narrative phase”: is it consistent, or does it vary? • Dimensions of each figure (as for scenes); consistency within a “narrative” when the image repeats, and consistency across narrative phases. • As in the case of scenes, use GPA to compare instances of a given figure type.

Concluding Discussion In this chapter, which is necessarily quite short in length, we have developed and presented a system of analytical classifications for understanding how figures are placed in spatial relation to each other on Valcamonica rock art surfaces. A typology, especially one with as few classes or types as this one has, is a meager basis for useful analysis unless there also exists some finer gradation of quantitative information. Thus, we have proposed some measures that can be developed in various working examples. At this early stage, though, trial and error, followed by adjustment, are needed to find out just what should be measured and how a collective pattern should be analyzed. We think this approach—based on just a few classes, each of which is further explored through simple mathematical measures—is more promising than a typology with many more types, or many sub-types recognized within the overall types. Here we regard Valcamonica as a case study and source of worked examples, rather than a subject to be covered in a self-contained way. It is clear that many aspects of rock engravings must be common across the world. All such aspects that have to do with the physical making of the art relate to, and derive from, the way rock behaves when it is assaulted with a stone or metal instrument and thereby scratched, incised, or pecked. The needed analysis should rely primarily on geology, material science, and the physical aspects of human behavior and use of the hands. Equally, we are certain the “picture problem” issues of depicting complex 3-dimensional shapes in simple 2-dimensional figures are related, even nearly identical across rock-art traditions. Hagen (1986) we see as a good starting point (see discussion above). As regards scenes, rock art researchers everywhere are still in a primeval stage: ‘It looks very much to me like a scene to me, but I cannot clearly state just why I know that—it’s just sort of obvious!’ Even had we chosen to report and reference in detail the collected Valcamonica literature touching on this matter, there would not have been much to say. As for the bigger picture of the systematics of rock art scenes within rock art in general, we note that the present volume is the first coherent and comparative collection ever to address this.

Acknowledgments We are grateful to colleagues for our discussions about these universal issues in rock art research, especially those in the • P • I • T • O • T • I • and 3D-PITOTI projects, with their focus on Valcamonica specifics. We thank our colleague, Kate Sullivan, and the volume editors for their help in moving our draft forward into this much-improved final form. Nearly all the images came from Seradina 12 and are are found in Marretta (2018). The others were redrawn by ourselves.

RULES OF ORDERING AND GROUPING IN THE PITOTI

275

Craig Alexander has worked on the Iron Age rock art of Valcamonica (northern Italy), the sourcing and exchange of ceramics in the Apulian Neolithic (southern Italy), the dating of the Neolithization of Apulia, the dating of the Neolithization of Italy as a whole, and Bayesian modeling of the chronology of a large medieval cemetery in Cambridge (UK). He has also taken part in various contract archaeology projects, typically in a GIS role. Craig is particularly interested in GIS and statistical/methodological issues and has published several archaeological articles, having worked in the business world for many years before studying for a Masters and PhD in archaeology. He has also done work for the University of Cambridge as an independent scholar. Alberto Marretta is the Scientific Director of the Parco Comunale di Seradina-Bedolina in Capo di Ponte, Valcamonica, Italy. The park contains many engraved rocks, the most notable of which include the Bedolina map and the enormous rock 12 with its thousand plus images. Alberto has published extensive works on the rock art of Valcamonica, most recently a volume dedicated to rock 12 at Seradina. Thomas Huet has worked mainly on the Neolithic and Protohistoric system of imagery (rock art, steles, pottery decorations, etc.) in the northwest Mediterranean. His approach is mainly computer-based and datadriven. He developed routines of analyses that applied programming languages to large and heterogeneous datasets. Formerly he was contracted as IT for universities and research institutes working on different topics, such as the spread of farming during the Early Neolithic, or the networks of personal goods and importations that arose during the Iron Age). He has published numerous scientific articles and book chapters on various topics and is a member of several different reviewing committees linked to computer applications in archaeology, a post that also implies some teaching at the university. Christopher Chippindale studies the prehistoric rock art of Mont Bego and Valcamonica in southern Europe and the rock art of Aboriginal peoples in Australia. Christopher has long been interested in the possibilities opened up by more formal analyses of rock art imagery. For many years he was a curator at the Museum of Archaeology and Anthropology of the University of Cambridge and is now affiliated with the McDonald Institute for Archaeological Research in Cambridge.

Note 1. Editorial note: this usage creates a problem for the conventional labeling of published pictures from the site. We have chosen not to follow the authors in labeling these “Image 1-5”; instead, we will stick to the usage followed by all other authors in the volume and label them, e.g., Figure 1.5,” meaning the fifth figure (i.e., image, usually a photograph or artwork) in Chapter 1.

References Alexander, C. 2011. “Valley of Pitòti: GIS-based Socio-spatial Analysis of Iron Age Rock Art in Valcamonica (BS), Lombardy, Italy.” Ph.D dissertation. Cambridge: University of Cambridge. Anati, E. 1960a. La civilization du Val Camonica. Paris: Editions Arthaud. ———. 1960b. La Grande roche de Naquane. Paris: Masson. ———. 1963. La datazione dell’arte preistorica camuna. Breno (BS): Tipografia Camuna. ———. 1964. Camonica Valley: a Depiction of Village Life in the Alps from Neolithic Times to the Birth of Christ, as Revealed by Thousands of Newly Found Rock Carvings. Trans. L. Asher. London: Jonathan Cape. ———. 1982. Luine: Collina Sacra. Capo di Ponte: Edizion del Centro. ———. 1990. “L’altopiano di Ossimo-Borno nella preistoria: le statue-menhir—relazione preliminare.” Bollettino del Centro Camuno di Studi Preistorici, 25/26: 269–358. ———. 2004. La civiltà delle pietre: Valcamonica una storia per l’Europa. Capo di Ponte (BS): Edizion del Centro. Arca, A. 2004. “The Topographic Engravings of Alpine Rock Art: Fields, Settlements and Agricultural Landscapes.” In The Figured Landscape of Rock-Art: Looking at Pictures in Place, ed. C. Chippindale and G. Nash, 318–50. Cambridge: Cambridge University Press.

276

CRAIG ALEXANDER, ALBERTO MARRETTA, THOMAS HUET, AND CHRISTOPHER CHIPPINDALE

Azéma, M. 2008. Préhistoire de la bande dessinée et du dessin-animé. Tarascon: Parc de Préhistoire de Tarascon sur Ariège. ———. 2009. L’Art des cavernes en action 1. Paris: Errance. ———. 2010. L’Art des cavernes en action 2. Paris: Errance. ———. 2011. La Préhistoire du cinéma: origines paléolithiques de la narration graphique et du cinématographe. Paris: Errance. Bicknell, C. 1911. The Prehistoric Rock Engravings in the Italian Maritime Alps, 2nd ed. Bordighera: Pietro Gibelli. ———. 1913. A Guide to the Prehistoric Rock-Engravings in the Italian Maritime Alps. Bordighera: Giuseppe Bessone. Brunner-Traut, E. 1974. “Epilogue; Aspective, in Heinrich Schäfer.” In Principles of Egyptian Art, ed. and trans. J. Baines, 424– 43. Oxford: Clarendon Press. Casini, S. 1994. Le pietre degli dei: menhir e stele dell’Età del Rame in Valcamonica e Valtellina. Bergamo: Centro Culturale Nicolò Rezzara. Chippindale, C. 1988. “The Later Prehistoric Rock-Engravings of Val Fontanalbe, Mont Bego, Tende, Alpes-Maritimes, France.” Ph.D dissertation. Cambridge: University of Cambridge. ———. 2004. “From Millimetre up to Kilometre: A Framework of Space and of Scale for Reporting and Studying Rock Art in Its Landscape.” In The Figured Landscapes of Rock Art: Looking at Pictures in Place, ed. C. Chippindale and G. Nash, 102–17. Cambridge: Cambridge University Press. Chippindale, C., and F. Baker. 2012. • P • I • T • O • T • I • : Digital Rock Art from Prehistoric Europe: Heritage, Film, Archaeology / • P • I • T • O • T • I • : arte rupestre digitale dell’Europa preistorica: beni culturali, film, archeologia. Milan: Skira. Clegg, J. 1995. Notes towards Mathesis Art, 3rd ed. Balmain (NSW): Clegg Calendars. Davenport, D., and M. Jochim. 1988. “The Scene in the Shaft at Lascaux.” Antiquity 62: 558–62. Hagen, M. 1986. Varieties of Realism: Geometries of Representational Art. Cambridge: Cambridge University Press. Huth, C. 2011. “Contributions in: Qu’est-ce que l’art protohistorique? Points de vue d’Henri-Paul Francfort, Christoph Huth, Ricardo Olmos. Miklós Szabó, & Stéphane Verger.” Perspective: La revue de l’INHA 2010/11(2): 195–214. Lenssen-Erz, T. 1992. “Coherence—a Constituent of ‘Scenes’ in Rock Art.” Rock Art Research 9: 87–105. Marretta, A. 2018. La roccia 12 di Seradina 1: documentazione,analisi e interpretazionedi un capolivro dell’arte rupestre alpina. Capo di Ponte (BS): Edizioni del Parco di Seradina-Bedolina. ———. n.d. Listing of statue-stele in Valcamonica. Unpublished. Marretta, A., T. Cittadini, and A. Fossati. 2007. La Riserva Naturale Incisioni Rupestri di Ceto, Cimbergo, Paspardo: guida al percorsi di visita. Capo di Ponte: Centro Camuno di Studi Preistorici. Martinotti, A. 2009. “Instrumentum rituale: simbologia ed idelogia della ‘paletta’ nell’Italia protohistorica tra archeologia ed arte rupestre.” In Valcamonica Symposium 2009: fare storia dalla preistoria: Capo di Ponte 28 ottobre—2 novembre 2009: pre-atti del convegno, ed. E. Anati, 243–54. Capo di Ponte: Centro Camuno di Studi Preistorici. Miller, H. 2004. “Tobler’s First Law and Spatial Analysis.” Annals of the Association of American Geographers 94(2): 284–89. Nowell, A. 2015. “Learning to see and seeing to learn: Children, communities of practice and Pleistocene visual cultures.” Cambridge Archaeology Journal 25(4): 889–899. Pause, K. 1992. “Le scene di duello nell’arte halstattiana e nell’arte rupestre camuna: confronti ed elementi di datazione.” Appunti 19: 36–40. Ripley, K. 1976. “The Second-Order Analysis of Stationary Point Processes.” Journal of Applied Probability 13: 255–66. Sansoni, U. 1977. “Le figurazioni umane a grandi mani nella problematica magico-religiosa delle incisioni rupestri della Valcamonica.” Bologna: Tesi di Laurea, Universita degli Studi di Bologna. ———. 1987. L’arte rupestre di Sellero. Capo di Ponte: Edizione del Centro. Sansoni, U., and S. Gavaldo. 1995. L’arte rupestre del Pia d’Ort: la vicenda di un santuario preistorico alpino. Capo di Ponte: Edizione del Centro. ———. 2009. Lucus rupestris: sei millenni d’arte rupestre a Campanine di Cimbergo. Capo di Ponte: Edizione del Centro. Stiny, G. 1975. Pictorial and Formal Aspects of Shape and Shapegrammars. Basel: Birkhäuser. Taçon, P., and C. Chippindale. 1994. “Australia’s Ancient Warriors: Changing Depictions of fighting in the Rock Art of Arnhem Land, N.T.” Cambridge Archaeological Journal 4(2): 1–20 ———. 1998. “Introduction: An Archaeology of Rock Art through Informed Methods and Formal Methods.” In The Archaeology of rock art, ed. C. Chippindale and P. Taçon, 1–10. Cambridge: Cambridge University Press. Tobler, W. 1970. “A Computer movie simulating urban growth in the Detroit Region.” Economic Geography 46(S): 234–40. Turconi, C. 1998. La mappa di Bedolina nel quadro dell’arte rupestre della Valcamonica, Notizie Archeologiche Bergomensi 5: 85–113.

==== •

•

•

18. FINDING ORDER OUT OF CHAOS A Statistical Analysis of Nine Mile Canyon Rock Art Jerry D. Spangler and Iain Davidson

Introduction: The Nine Mile Paradigm Nine Mile Creek (which is actually about 144 km long) is a perennial west-to-east tributary of the Green River, itself the largest tributary of the Colorado River and part of the greater Colorado Plateau, the largest drainage system in the western United States. It is a high-desert region and generally arid, with an agricultural zone in the valley bottom that receives less than thirty centimeters of rain annually—an amount insufficient for agriculture without irrigation. Nine Mile Creek is a small stream, generally two to four meters wide and no more than twenty to thirty centimeters deep, that is fed by high plateau spring snow melt and small springs along its entire course. One of only four flowing water sources in the West Tavaputs Plateau, the creek is a predictable and consistent source of water even during the worst droughts. Nine Mile Canyon, through which the Creek flows, probably has more rock art than any other place in the continental United States. Most of those who choose to work in Nine Mile Canyon in eastern Utah are awestruck by the sheer quantity of images, and in some cases by the remarkable skills and fearlessness of the ancient artists, given that images can be found along the narrowest of cliff ledges where one slip meant certain death. To date, archaeologists have thoroughly inventoried about 10 percent of the canyon, where they have documented some 1,200 prehistoric sites of all types, mostly rock art sites and agricultural residences and hamlets typically assigned to the Fremont Complex (AD 500–1300) (e.g., Simms 2010). It is estimated that a complete inventory of the canyon would result in the identification of about ten thousand prehistoric sites, most of which would have rock art features, perhaps with as many as 75,000 to 100,000 images in total. About half of the rock art clusters are found near the canyon bottom, and because of an access road in the bottom of the canyon these images are well known to the public and rock art researchers and touted by tourism boosters as “The World’s Longest Art Gallery.” The creek has eroded the soft Tertiary-age sandstone of the Green River Formation, creating a stair-step series of cliffs rising 300–400 meters above the valley floor. These cliff faces are often quite smooth and patinated with accretions of red-to-black iron oxide, popularly referred to as “desert varnish,” that have proved to be an attractive canvas for prehistoric rock artists. The rock art is found along the lower 72 kilometers of the canyon, starting at an elevation of about 2,040 meters and continuing uninterrupted to the creek’s confluence with the Green River at 1,400 meters above sea level. Since 2006, the Colorado Plateau Archaeological Alliance (CPAA), a nonprofit working to protect cultural sites on public lands in the American West, has conducted systematic inventories of the canyon, establishing uniform methods for documenting, organizing, and describing prehistoric rock art consisting of petroglyphs (by far the most common rock art) and pictographs (few in number and preserved only in sheltered settings).

Research Questions The analysis and description that follow review some of the results of a general survey of the rock art of the canyon in order to consider some questions about the cultural origins of the imagery. This work began with Polly Schaafsma’s work in the 1970s, which brought some semblance of order to rock art studies north and

278

JERRY D. SPANGLER AND IAIN DAVIDSON

west of the Colorado River. Using photographic collections at the Peabody Museum at Harvard University, she defined regional variations in rock art styles and compositional elements (Schaafsma 1971). More recently, the spatial dimension of our survey permitted further analysis in relation to the association between the art and the other archaeology, and then in terms of the presences and absences of clustered images that might be regarded as scenes. The sorts of questions we asked concerned whether spatial analysis of individual image classes or clusters of images could provide answers about the nature and function of the art. Nine Mile Canyon art is famous for some of its obvious scenes, but within a tradition that includes the production of scenes, are there are other categories of art production that cannot be construed as scenes? If there are, can analysis of the spatial distribution of such sites help us move toward new understandings about the process of the production of the art? Finally, does the pattern of combinations of images in clusters of images permit an interpretation in terms of ritual or of casual production?

Methods For the purposes of this chapter, CPAA compiled data collected on the five recent block inventories in the canyon—five noncontiguous areas encompassing about 1,100 hectares (see Spangler 2011a, 2011b, 2011c, 2012, 2014). These included one block in the upper portion of the canyon where agriculture would have been difficult due to shorter growing seasons, one each in the upper middle and lower middle portions where site densities were known to be exceptionally high, one in the middle portion but focused on a side drainage on the south side of the main canyon, and one in the lower portion of the canyon where natural vegetation is scant but evidence of prehistoric architecture is nonetheless abundant. Our intent was to define the spatial patterning of rock art images, both horizontally along the length of the west-to-east canyon bottom and vertically from the canyon bottom to the canyon rim more than 300 meters above. Using geographic information system software (ARCGIS), we plotted the distribution of rock art localities by relative complexity, general compositional qualities, and their spatial relationships with residential and storage sites believed to be associated with prehistoric agriculture. Our intent was to establish a valid sample of the canyon’s rock art to test certain long-held assumptions. To test the assumption that most of the rock art was attributable to the Fremont, we posited several working hypotheses: (1) If rock art clusters are associated with agricultural activities, then they should be found at those elevations where agriculture would have been possible, and conversely they would be rare in settings where agriculture is not possible; (2) The density of rock art clusters would increase or decrease in association with the agricultural potential of the adjacent floodplain in the same area; (3) Optimal farming locations would have fostered larger populations, as evidenced by greater numbers of residences, which would correspond to greater numbers and density of rock art clusters, and (4) rock art clusters would be found primarily in close proximity to the residences, or within 30 meters’ vertical distance above the floodplain. Our analysis supported some but not all of our working hypotheses. By standard practice on public lands in the United States, an archaeological site is any location where (a) there are artifacts and/or features in close proximity to one another, and (b) these elements are separated from another group of artifacts or features by at least 30 meters. In other words, an archaeological site in Nine Mile Canyon might be several thousand square meters, with dozens of individual features ranging from pit houses and granaries to clusters of rock art, so long as the features are within 30 m of another feature. If there is a separation of more than 30 meters, a different site number is assigned. This rather arbitrary 30-meter separation between sites ignores the likelihood that a rock art locality 40 meters away from a pit house is probably related to the pit house and the individuals who lived there. In Nine Mile Canyon, this practice has resulted in an astonishingly high site density (more than 60 sites per km2) that most certainly masks the interrelated nature of the features being documented. In principle, data of this sort should allow a distinction between art made close to living sites that was perhaps casual, and art made at a distance that required a special-purpose trip. That distinction is not possible at the present stage of the analysis. During the CPAA inventories on Warrior Ridge, Utah State Archaeologist Kevin Jones encouraged us to ignore the 30-meter standard and instead document all ninety features along the 1.6-kilometer

FINDING ORDER OUT OF CHAOS

279

ridgeline as a single site (42Dc1). This facilitated a more thorough discussion of the interrelatedness of rock art to residential and storage features but also revealed the disadvantage that the Warrior Ridge site data are not directly comparable to the site data from the other four survey blocks where the 30-meter standard was applied. Another problem we encountered is the common use of the term “rock art panel” to describe different groups of rock art images at a single archaeological site. Currently, there is no standard as to what constitutes a “panel” or what the spatial separation should be to distinguish one panel from the next. Some archaeologists have assigned different panel numbers when there was no spatial separation whatsoever. Hence, the term as traditionally used in Nine Mile Canyon is meaningless in terms of statistical analysis. To establish consistency in our own analyses, we avoided using “panel” and instead organized the data into “clusters” to denote a group of images separated from another group of images by at least 2 meters. This is very similar to what Alexander, Marretta, Huet, and Chippindale (this volume) defined as a cluster in their work at Valcamonica, but without any assumptions about whether the images were related to one another or if they reflected scenes or narratives. We agree with their reasoning that a cluster, in most instances, probably reflects images made at about the same time, and/or that the presence of a single image prompted the creation of additional images on the same surface. Our inventories were conducted using standard archaeological survey techniques (e.g., 15-meter parallel transects whenever possible), but with modifications to accommodate the rugged canyon topography. Slopes of more than 30 degrees were omitted, and instead of uniform transects on the steep slopes, crew members were assigned to walk along the base and top of each ascending cliff level and associated ledges, with special attention directed at isolated outcrops, crevices, monoliths, buttes, and other natural features where archaeological features would be expected. Crews worked east-west and parallel to the canyon bottom within the inventory boundaries defined by legal sections found on U.S. topographical maps. All archaeological sites were documented on standardized forms required by the state of Utah, and all rock art clusters were photographed and sketched. Survey data were then organized on Excel spreadsheets by GIS data, topography, elevation, size, complexity, compositional qualities, and selected representational elements. The five survey blocks discussed here were chosen because they represent different portions of the canyon with different environmental variables, and because the data were collected and organized in exactly the same manner. The identification of animals by species proved quite frustrating and rather subjective. For our purposes, an animal was labeled a bighorn sheep only if it depicted long horns that curled back toward the hind quarters; of course, this eliminated what were probably depictions of bighorn ewes and lambs without horns. Images were labeled as deer or elk if antlers with multiple tines were clearly depicted; this also eliminated does, cow elk, and juveniles without antlers. Animals were labeled as bison if they had a distinctive hump on the shoulders and short horns, but there were only three examples of this, all in the Owl Canyon block (numerous other examples in survey areas are not considered in this chapter). We suspected that some images (with short curled tails and open mouths as if barking) depicted canines, but in the absence of more defining characteristics they were not cataloged as such. Animals without clearly identifying traits were simply labeled as unknown. We also wanted to test Schaafsma’s assertion that the trapezoidal anthropomorph was the typological norm in this region. In fact, it is a long-standing truism that the trapezoidal anthropomorph is a defining characteristic of the Fremont Complex. The 969 anthropomorphic images were initially organized by torso shape into eight categories: trapezoidal, round, rectangular, oval, cigar-shaped, stick figures, backpackers, and “other,” which encompassed abstract bipedal representations, multiple torso shapes, and unique stylized anthropomorphs. Distinguishing between oval and cigar-shaped images proved meaningless, so these were subsequently combined.

Definitions In order to define the contexts of the analysis, we need to describe the rock art of the canyon and the region. Schaafsma (1971) assigned the rock art of Nine Mile Canyon to what she called “San Rafael Fremont: North-

280

JERRY D. SPANGLER AND IAIN DAVIDSON

ern Zone.” She suggested these “petroglyphs and rock paintings . . . exhibit a stylistic phase of Fremont rock art which is internally consistent and distinct from that of the Uinta region and which can be differentiated from that of the southern San Rafael zone” (1971: 28–29). In particular, the Northern Zone rock art prevalent in Nine Mile Canyon lacked the detailed ornamentation commonly found in the Vernal district north of Nine Mile Canyon (see for example Simms 2010: 95–97). Instead, the panels were crowded with small, solidly pecked figures, often carelessly executed and ill-defined. Schaafsma’s analysis found Northern Zone anthropomorphs exhibited a typological norm of a trapezoidal body and bucket-shaped head. The body form might be flared at the base to suggest a kilt, and some anthropomorphs are portrayed with long rake-like horns or antlers. These figures had arms, which were commonly bent at the elbow, and hands with spread fingers. Legs were usually short and straight. Some anthropomorphs displayed rectangular or triangular body shapes instead of the usual trapezoidal configuration. The majority of anthropomorphs were painted or pecked in solid infill (1971: 29–31; see Figure 18.1), not in outline as in some other regional art (for examples of the variation, see Simms 2010: 80–108). A general consensus among archaeologists working in the region asserts that maize agriculturalists occupied the northern Colorado Plateau by AD 200 and Nine Mile Canyon by AD 500, and that agriculture persisted until AD 1300 and perhaps two centuries later in isolated niches. Circular residences built in stone are very widespread, situated on gravel benches 20 to 30 meters above the floodplain, and stone-and-mud granaries are tucked onto cliff faces. Both types of structure can be attributed to these earliest farmers. Scores of residences have been documented, mostly occurring as a single residence or in clusters of two or three. Occasionally there are small hamlets or villages of seven to ten residences. Dozens upon dozens of granaries have been documented, usually in close proximity to the residences and reflecting a variety of storage strategies. Because evidence of earlier (Archaic) or later (Ancestral Ute) occupations is actually quite rare, it has often been assumed that many if not most of the rock art depictions in the canyon are also attributable to these Fremont farmers. A predictable weakness in Schaafsma’s research derives from the limitations of the raw data available to her, which was inherently biased toward FIGURE 18.1. The typological norm for the Fremont Complex in large or aesthetically pleasing sites that Nine Mile Canyon, as suggested by Schaafsma, is a trapezoidal or drew the attention of photographers rectangular anthropomorph. These actually represent a minority over the years. The photographers had of the anthropomorphs here. Photo courtesy of the Colorado Pla- made no attempt to catalog all rock art teau Archaeological Alliance. sites in any given area, and she had no

FINDING ORDER OUT OF CHAOS

281

idea whether these large panels were a representative sample of the thousands of rock art localities that are present in Nine Mile Canyon. We adopted Davidson’s definition of a scene (Chapter 1 in this volume) as “a set of images in spatial proximity to each other from which, without any knowledge other than the images themselves, an observer can infer actions taking place among the actors represented in the image.”

Results CPAA’s large-scale surveys of Nine Mile Canyon rock art have demonstrated that the classic trapezoidal Fremont anthropomorph–the human figure with broad shoulders tapering to a narrower waist–is comparatively rare. In fact, it represents a minor portion (17 percent) of all anthropomorphs depicted in the Canyon. Nine Mile Canyon anthropomorphs come in every imaginable shape: round, oval, cigar-shaped, square, rectangular, stick-figure, abstract, trapezoidal, and even combinations of shapes. Some are imposingly large (greater than 1 meter in height) and others are tiny (less than 5 centimeters high). In short, when the entire catalog of images is considered, it is unclear whether Nine Mile Canyon rock art conforms to any typological norm for Fremont rock art, at least as far as Schaafsma defined it. Certain elements Schaafsma dismissed as not present are in fact present in significant numbers. For example, she claimed that the humpbacked flute player motif was not present in the Nine Mile Canyon region, yet this motif has been observed throughout Nine Mile Canyon (Gillin 1938: 30; Reagan 1933: 62–63; Spangler 2011a; Strevell and Pulver 1935: 17). In another example, Schaafsma (1971: 36) noted the rarity of serpents with horns or plumes. Since 1989, intensive surveys in middle and lower Nine Mile Canyon have revealed more than two hundred such images, which were rare in the photographs Schaafsma studied. Despite the broad variability evident in the Nine Mile Canyon rock art, several patterns emerged from our combined data set: (1) individual clusters of rock art are usually small, ranging from 3 to 8.5 images per locality; (2) complex sites with at least a hundred images and obvious compositional qualities are rare, but there is at least one in each of the survey blocks, all of which are in close proximity to a small village; (3) rock art clusters are found at every level from the canyon bottom to more than 300 meters above, but cluster size and cluster density decrease substantially with increased distance above the valley floor; and (4) the identifiable images are almost equally divided between zoomorphic representations, bipedal or anthropomorphic representations, and geometric patterns that exhibit repetitious uniformity throughout the canyon (see Table 18.1 for the data set used in the following discussion). TABLE 18.1. Gross rock art data based on upper, middle and lower survey blocks and tabulated by images per site and images per cluster (panel).

Location

Survey Name

Total Sites With Rock Art

Individual Rock Art Clusters

Total Images

Median Images/ Site

Mean Images/ Site

Median Images/ Cluster

Mean Images/ Cluster

Upper Canyon

First Canyon

28

61

361

3

12.9

3

5.9

Middle Canyon

Owl Canyon

58

148

1,496

14

25.8

5.5

10.1

Middle Canyon

Warrior Ridge

n/a

54

1,000

n/a

n/a

8.5

18.5

Middle Canyon

Petes Canyon

57

164

1,057

7

18.5

3

6.4

Lower Canyon

Devils Canyon

24

49

399

10

16.6

4

8

282

JERRY D. SPANGLER AND IAIN DAVIDSON

Our attempts to organize rock art clusters by compositional qualities revealed that Nine Mile Canyon rock art does not conform to a single definition or typological norm. There are many large clusters where images were clearly added over a long period of time, as evidenced by different levels of re-patination and superimposition (Figure 18.2). Yet in many other large clusters there is no evidence of this whatsoever, suggesting the images were created over a short period of time and are ideal examples of what Alexander, Marretta, and Chippindale (this volume) refer to as a “weak avoidance rule” in that the placement of figures was intentionally chosen so as to avoid overlap with existing figures, although there might be exceptions (Figure 18.3). This is especially evident in the small clusters (100 m

Upper Canyon

First Site

38

16

3

4

Middle Canyon

Owl Canyon

73

42

26

7

Middle Canyon

Petes Canyon

67

48

20

24

178

106

49

35

Total

The Warrior Ridge data appears anomalous in that a high density of rock art images (n=1,000) are clearly associated with residences and storage features, but none of these are in close proximity to potential fields on the floodplain or to a permanent water source. Access to the ridgetop is steep, and the energy costs of transporting water to a large population aggregated there would have been significant. In the winter months, though, snow accumulated on the north face of the ridge could have been melted, reducing the cost of retrieving water. The rock art images found on Warrior Ridge might therefore have been created in the dreary winter months while farmers awaited their annual return to spring-summer residences near their fields on the valley floor. In summary, our spatial analysis supports the hypothesis that these rock art clusters are generally associated with residential sites oriented to maize agriculture (field preparation, planting, weeding, etc.), and that the size and density of the clusters decreases at elevations above 1,829 meters (6,000 feet), where agriculture would have been increasingly risky. About half of all rock art clusters are located within 30 meters’ vertical distance above the floodplain, but almost three out of every four clusters are located within 60 meters of elevation over the floodplain. This suggests that the images’ creators chose suitable cliff surfaces in relatively close proximity to residences, and that the activity of creating rock art would have been peripheral to maize farming. The exception to the pattern of close proximity to fields is Warrior Ridge, at 100 to 300 meters above the floodplain. As discussed above, the high density of rock art images there is indeed associated with residences (n=10). In this area, however, the massive quantities of fire-cracked rock and the optimal solar aspect of this area’s residential features suggest it was occupied during colder months, when no agricultural activities would have occurred. In this sense, rock art clusters are associated with farmers rather than farming. And the creation of rock art images, at least in the case of Warrior Ridge, might have been a seasonal activity pursued when residents were not distracted by food production requirements. When the data sets from the other survey blocks are considered collectively, the fact that one out of every four rock art clusters is not in close proximity to a residence suggests that other, unidentified factors also influenced decisions about where to create some images, and these images might reflect activities unrelated to farming (e.g., hunting forays, overland travel and trade, and ritual). With the exception of Warrior Ridge, rock art clusters within the survey blocks at higher elevations over the floodplain and/or residences tend to be small (fewer than 5 images) and widely dispersed, but otherwise are stylistically indistinguishable from those found at lower elevations. These small clusters typically feature geometric figures, often with a single anthropomorph or quadruped, but only rarely do they exhibit multiple image types. Typical of these clusters would be a single, small anthropomorph with an undulating line and a circle, or a quadruped with a meandering line.

Relative Size and Complexity Further exploration of issues about relations between the rock art and the settlements was beyond the scope of our baseline inventories. Nonetheless, we were able to organize the data to address the relative size and relative complexity of rock art clusters in the canyon. We classified sites based on the conservative estimate that one individual could, on average, produce one small petroglyph per day, and that the number of images at any given location would reflect a time commitment (cost) in actual days. Using this estimate, a rock art cluster with five or fewer images probably represents a single event lasting only a few days. A cluster with up

286

JERRY D. SPANGLER AND IAIN DAVIDSON

to ten images might have required a week or a little longer to create, but also might have been a single event. We used this distinction as a basis for classifying rock art clusters. Based on this approach, 58 percent of all rock art clusters have five or fewer images, and 78 percent have ten or fewer (see Table 18.3). In other words, the vast majority of all rock art clusters might have been the outcome of a single, short-term event that would not have inhibited participation in food production (agriculture) or procurement (hunting and gathering) activities. The creation of these small clusters would have required a time commitment of only about 1 to 3 percent during a given year (365 days divided by 5 days and 10 days, respectively). Three possibilities are suggested: (1) Stylistic uniformity within individual clusters can be attributed to their being created by a single individual in a short period of time, and therefore they reflect a narrative scene of significance to that individual; (2) Creating rock art was for the most part an activity in which many or most families participated, and expressions of common iconography varied only subtly from one group to another; and (3) the stylistic uniformity repeated throughout the canyon reflects a conformity to values shared by other social groups in the same area, including the element of repetition, an essential element of ritual (Ross and Davidson 2006). Nine Mile Canyon is famous for its large and occasionally complex rock art sites, replicas of which now grace book covers, T-shirts, and other tourist souvenirs, but these sites represent a decided minority of rock art sites in the canyon. Only 4 percent of rock art clusters have 21–30 images, 2 percent have 31–40 images, and 3 percent have more than 40 images. These larger clusters are typically located in close proximity to a small agricultural village or cluster of residences, and they might be relevant to larger group activities, even though they are stylistically identical to the smaller clusters. Most of these larger sites exhibit greater attention to compositional detail, portraying an abundance of large, interacting figures. Some appear to depict in elaborate detail a single event, such as the Great Hunt site in Cottonwood Canyon and the combat scenes at Warrior Ridge. Others can and often do exhibit evidence of modifications and augmentation through time, sometimes creating a bewildering maze of older and younger images. These serve as a reminder that some of the larger clusters might actually be the accumulation of many single-event activities. Separating “scenes” from accidental accumulations is a fundamental task for those who want to talk about the importance of scenes. One of us (see Davidson this volume) suggested a hierarchy of image combinations with the intention of teasing out aspects of the way clusters of images were produced. Looking only at Davidson’s Category A— single images on a panel—we examined the database for Nine Mile Canyon and found 99 rock art clusters (21 percent) in this category. Of those, 21 were abstract or indeterminable, 19 were anthropomorphs (all small), and 42 were geometric or repeating images (mostly meandering and undulating lines). Only 17 clusters were instances where a quadruped appears by itself (17 percent). Of these, only three were confidently identified as bighorn sheep and one as an elk (at Warrior Ridge, Figure 18.4). All others were listed as unknown quadrupeds. All except the elk are rather carelessly executed, or at least are not bound to precision and detail. All except the elk are quite small. In other words, the elk figure is an outlier, something that does not occur anywhere else in the five inventory blocks. TABLE 18.3. Relative rock art complexity per rock art panel for each of the five survey blocks. Total Clusters

1–5 Images

6–10 Images

11–20 Images

21–30 Images

31–40 Images

>40 Images

First Canyon

61

43

8

5

3

2

0

Owl Canyon

146

74

38

21

6

0

7

Warrior Ridge

54

17

12

11

5

4

5

Petes Canyon

164

106

31

17

5

3

2

Devils Canyon

49

34

6

6

1

1

1

Total

474

274

95

60

20

10

15

58

20

13

4

2

3

Percentage

FINDING ORDER OUT OF CHAOS

287

FIGURE 18.4. The unusually large, detailed elk figure at Warrior Ridge is unique in that no similar animal representations have yet been identified in the absence of anthropomorphic or geometric representations. Jerry Spangler shows how large the image is. Also shown is a panel above the elk that is probably not part of the same act of making. Copyright Iain Davidson.

288

JERRY D. SPANGLER AND IAIN DAVIDSON

The elk becomes even more of an outlier when the clusters with only a single anthropomorph are considered. All anthropomorphs are rather small, and most lack attention to precise details. There are no significant differences in relation to the shapes of torsos: the numbers are divided roughly equally among rectilinear, curvilinear, and other. Although the sample size is small, two out of every three clusters with a single anthropomorph are located between 130 and 365 meters above the floodplain—in other words, nowhere near a suspected residence. Instead, anthropomorphs are found on high outcrops, where those who had expended significant energy to climb the near-vertical slopes would have been rewarded with commanding views up and down the main canyon. Far from revealing the importance of single images (as we might have expected from the outstanding nature of the elk at Warrior Ridge), this inventory suggests something more mundane: that there are at least four subdivisions in Category A: A1, Finely executed large single figures; A2, small anthropomorphs; A3, carelessly executed animal images that might have formed groups via the addition of similar images; and A4, geometric or nonrepresentational marks. The location of the elk and of the single anthropomorphs might suggest that the presence of single figures was not principally concerned with narratives in the way that scenes were. The larger clusters’ location close to residential concentrations, along with the fact that several of them show attention to compositional detail and thus might be called scenes, suggests that scenes may have been related to domestic or agricultural activity rather than to a context where special functions were performed. On the other hand, the presence of isolated anthropomorphs at heights above the residences and floodplains suggests an entirely different context.

Representational Qualities It is extremely risky to interpret what is being depicted in the rock art images, especially when there are no modern analogues, nor any people who can demonstrate historical continuity with the Fremont Complex. What appears to us to be an animal might, in the artist’s eye, have been something entirely different, as has been documented in Australia (Macintosh 1977). However, that has not deterred many from speculating about whether or not images reflect actual events or abstract concepts. Such conclusions, based on a single site or a handful of sites, are tenuous and remain untested. The CPAA baseline study was instead focused on identifying patterns in spatial distributions at different scales without necessarily attributing narrative meaning to those patterns and whether or not those patterns are spatially consistent through the canyon. The review of more than 1,000 field sketches found considerable variability in styles, particularly in the depiction of humans and animals. Some were rendered rather carelessly and others exhibited precise attention to detail that allowed for finer-scale identification. Rather than concentrate on such differences, we chose to organize the data into three general categories: Quadrupeds, anthropomorphs, and geometric or repetitive images. A fourth category of “unknown” emerged as a catch-all for images that were undefinable, abstract, or had eroded to the extent that the original image could not be determined (see Table 18.4). TABLE 18.4. Tabulations of image types by broad categories (quadrupeds, anthropomorphs, and abstract unknown) for each of the five survey blocks, Survey Name

Quadrupeds

Anthropomorphs

Geometric

Abstract/Unknown

First Site

85

62

121

91

Owl Canyon

279

239

422

532

Warrior Ridge

223

358

111

307

Petes Canyon

148

247

259

403

Devils Canyon

113

56

71

159

Total

848

962

984

1,492

FINDING ORDER OUT OF CHAOS

289

It is a common perception that animal images predominate at Nine Mile Canyon (see Matheny et al. 2004; Matheny, Smith, and Matheny 1997). That is certainly true in the Devils Canyon Survey block, but it is not the case when the images are considered in aggregate. In fact, the proportions of quadrupeds (20 percent), anthropomorphs (22 percent), and geometric/repetitive images (23 percent) differ only minimally from each other, although there are certainly differences from one survey block to the next. The distribution of quadrupeds by image class, according to the criteria described in the Methods section, is shown in Table 18.5. Most of the quadrupeds (79 percent) were subsequently categorized as unknown. More surprising was just how rare deer and elk depictions are (n=15), even though Nine Mile Canyon today is prime deer and elk habitat, as it probably was in prehistoric times as well. By comparison, 156 images (18 percent) were identified as bighorn sheep. Bighorn sheep, once extirpated by Euro-Americans, have been reintroduced to the region and are thriving in more remote, rugged canyon terrain, but they are rare in Nine Mile Canyon itself. Most of these images are located in close proximity to residences that are assumed to have been agricultural in purpose, even if the scenes themselves depicted hunting activities. Depictions of farming are extremely rare or have not been identified as such. According to the tabulated data, the trapezoidal anthropomorph, so important to Schaafsma’s (1971: 6) definition of Fremont, is not all that common (17 percent of all anthropomorphs), and it cannot be considered a typological norm. In fact, there does not appear to be a typological norm among anthropomorphs in Nine Mile Canyon. Oval and cigar-shaped anthropomorphs are more common (36 percent), but the data set is certainly biased by the unusually high number of such figures at Warrior Ridge, where they are engaged in scenes best interpreted as combat (Figure 18.5). There are more of this type at Warrior Ridge than in the other four survey blocks combined. Further analysis will look at the frequency with which cigar/oval figures are associated only with scenes of interaction with anthropomorphs, as opposed to interaction with animals or juxtaposition with them. We then examined this data set from a different perspective to better understand whether the Nine Mile Canyon anthropomorphs could be encompassed within Schaafsma’s definitions for Fremont rock art. She acknowledged that some Fremont anthropomorphs were rectangular and triangular. The data were then organized into three broad categories: rectilinear (a torso defined by straight TABLE 18.5. Tabulations of quadruped images by identifiable animal lines in the shape of a trapezoid, trian- for each of the five survey blocks. gle, or rectangle), curvilinear (a torso Bighorn defined by curving lines in circular and Survey Name Sheep Deer/Elk Bison Unknown oval shapes), and other. Under this orFirst Canyon 18 1 0 66 ganizational scheme, rectilinear images Owl Canyon 45 3 3 228 were considered to be Fremont anthropomorphs as defined by Schaafsma (see Warrior Ridge 33 2 0 188 Table 18.7). Petes Canyon 32 7 0 109 The results were not convincing. Devils Canyon 28 2 0 83 Only 22 percent of all anthropomorphs were rectilinear and therefore possibly Total 156 15 3 674 TABLE 18.6. Tabulations of anthropomorphic images by image shape for each of the five survey blocks. Survey Name

Trapezoidal

Round

Rectangular

Cigar/Oval

Stick

Backpacker

Other

First Canyon

11

7

0

28

5

2

11

Owl Canyon

61

23

21

74

25

16

19

Warrior Ridge

26

18

11

184

50

13

56

Petes Canyon

49

9

13

57

36

71

12

Devils Canyon

18

7

4

7

16

4

5

Total

165

64

49

350

132

106

103

290

JERRY D. SPANGLER AND IAIN DAVIDSON

FIGURE 18.5. This is one of more than a dozen different scenes depicting conflict at Warrior Ridge, especially above the large anthropomorph. Note that despite the high number of interacting images, there appears to be little overlap of images. The image also shows patterns of dots in rows, squares, rectangles, or polygons that are extremely common in Nine Mile Canyon but rare outside of it. Image courtesy of the Colorado Plateau Archaeological Alliance.

Fremont. Almost twice as many (43 percent) were curvilinear and therefore outside the standard definition for Fremont rock art. In summary, some anthropomorphs fit within the norm for Fremont rock art, but far more are inconsistent with Schaafsma’s traditional definition. The absence of Fremont figures in scenes with animals or as agents in scenes with other such figures suggests that the function of the Fremont figures was very different from the function of the cigar/oval figures. Among the anthropomorphs called “backpackers” or “burden bearers,” 106 are found in the five survey blocks, and several times that many have been documented elsewhere in the canyon. They typically depict a stick figure-like human in a walking motion (front TABLE 18.7. Tabulation of abstract/unknown images by broad foot raised) and carrying a large, oval pack on its back. They often occur in rows of ten categories (rectilinear, curvilinear, other). or more, which might be considered a scene Survey Name Rectilinear Curvilinear Other (Figure 18.6a). Certainly that interpretation First Canyon 11 35 18 would be consistent with Davidson’s (this volume) definition of a scene. The “backpacker” Owl Canyon 82 97 60 images are frequently found in association Warrior Ridge 37 202 119 with Fremont trapezoidal anthropomorphs Petes Canyon 62 66 119 and, based on identical re-patination of the images, are considered temporally consistent Devils Canyon 22 14 25 with the Fremont anthropomorphs. These Total 214 414 341 backpackers are extremely rare outside of Percentage 22 43 35 Nine Mile Canyon (a few have been identified

FINDING ORDER OUT OF CHAOS

291

FIGURE 18.6a–b. (a) Backpacker figures are extremely rare outside of Nine Mile Canyon and might be considered an identifier unique to the canyon. They are found near Desbrough Canyon in direct association with several trapezoidal anthropomorphs. Photo courtesy of the Colorado Plateau Archaeological Alliance. (b) There are more than two hundred figures in Nine Mile Canyon where undulating lines and spirals terminate with a head featuring what appear to be horns (and occasionally antlers). The horned snake motif is found throughout North America but is especially common among Puebloan peoples, who associate it with the Water Clan. Photo courtesy of the Colorado Plateau Archaeological Alliance.

292

JERRY D. SPANGLER AND IAIN DAVIDSON

in nearby Range Creek and Desolation Canyon, and some are found in the San Juan River Basin but are stylistically different), yet they are found in abundance the entire length of Nine Mile Canyon. Longtime Nine Mile Canyon scholar Ray Matheny suggested that the backpacker images, when considered in the context of the large numbers of bighorn sheep depicted in the rock art, reflect a winter-based hunting economy in which game animals were hunted, processed, and traded with distant groups for other resources. The clustering of rock art around the mouths of side canyons was consistent with trade routes in and out of the canyon, as well as access points to the high plateaus where game was procured (Matheny et. al. 2004). There are two problems with this conclusion: (1) There is little evidence of the significant stone tool production and maintenance that would be expected in a hunting economy, and (2) the backpacker images only rarely occur in association with hunting scenes or quadrupeds. It seems just as likely that the backpacker images reflect transportation of harvested maize between seasonal residences. Other images also appear to be unique to Nine Mile Canyon, or at least are certainly much more common there. The depiction of long-legged water birds (perhaps herons or cranes), sometimes with one leg raised, are especially common in middle Nine Mile Canyon, and there is one rock art cluster that depicts owls. Undulating lines and spirals depicting heads with horns, referred to as horned snakes, are ubiquitous throughout the entire length of the canyon (44 of them are located within the five study blocks), but they are rather rare in Fremont contexts outside of Nine Mile Canyon (Figure 18.6b). Likewise, dot matrixes (rows of dots in square, rectangular, or polygonal patterns) are found in abundance here but are rare outside the canyon (Figure 18.5). These can all occur as isolated images, but they are more commonly found in association with Fremont anthropomorphs (rectilinear), backpackers, and horned snakes. It also became quite obvious that certain geometric images are repeated with remarkable consistency throughout the canyon (see Table 18.8). The repetitive nature of these images suggests they had shared meanings understood by everyone, regardless of individual group identity. Munn (1973) has argued that geometric figures are deliberately polysemic, or that they have different meanings in different contexts. But the remarkable uniformity of certain geometric and repeating images in Nine Mile Canyon would appear more consistent with the principal characteristics of ritual behavior, in particular invariance, repetition, and form (see Rappaport 1999; Ross and Davidson 2006). The implications of a shared iconography certainly warrant additional study. Evidence of social interaction between humans is equivocal. When multiple anthropomorphs are depicted, they are usually represented as a row of individuals standing side by side, as is the case at the Family TABLE 18.8. Tabulation of non-anthropomorph and non-zoomorphic images that are repetitive throughout the canyon organized by survey block. First Canyon

Owl Canyon

Warrior Ridge Petes Canyon Devils Canyon

Totals

Dot Matrixes

10

28

3

11

1

53

Dot Rows

6

35

27

22

5

95

Concentric Circles/ Mazes

1

12

1

5

2

21

Spirals

1

4

4

14

4

27

Circles/Squares

53

71

18

41

14

197

Undulating Line

20

112

39

54

21

246

Horned Snakes

3

11

5

18

7

44

Meandering Line

9

52

4

18

11

94

Zigzags

0

6

1

2

0

9

Rakes/Ladders

5

15

0

5

6

31

Cross Hatches/Grids

1

7

7

15

0

30

Straight Lines

12

69

2

54

0

137

FINDING ORDER OUT OF CHAOS

293

Panel, which looks much like a family portrait. In this sense, the anthropomorphs seem to be interacting with the viewer, not each other. In some instances the anthropomorphs are holding hands. Scenes depicting dance or ritual have not been identified as such, although Rasmussen Cave has what appears to be dancing bipedal animals, which might be costumed participants in ritual dances. Scenes depicting human sexual activities are rare throughout the canyon and are absent from the five inventory blocks considered here. Scenes of humans interacting with animals are much more common but largely limited to scenes of a hunter aiming a bow and arrow at a quadruped, an anthropomorph standing amid a herd of quadrupeds, or an anthropomorph connected to a quadruped by an undulating or meandering line. In instances where horned snakes are depicted, the snake figure sometimes has an open mouth as if talking to an adjacent anthropomorph, lending support to the idea the horned snake is symbolic of ritual behavior.

Summary Most American archaeologists are notoriously reticent about studying indigenous rock art within the constructs of scientific methods. A fundamental goal of archaeology is to explain human behavior, and therein lies the problem: in the twenty-first century, how do you construct a sound, testable scientific hypothesis related to rock art when there are no other clues as to the ancients’ world view? Religion, cosmology, iconography, belief systems—all are abstract concepts that archaeologists find difficult to address within the constructs of scientific methods. In many respects, archaeologists working in the American West view rock art as background noise, a distraction to the broader behavioral questions like human responses to changing climates. In our view, however, thorough analysis of rock art in the context of other archaeological studies can enhance the understanding of the behavior of the people who made it. The irony is that rock art represents visible, tangible evidence of what is otherwise intangible. As such, Nine Mile Canyon presents an ideal outdoor laboratory for the study of rock art within the context of sound scientific methods, including other aspects of archaeology. A key element of our own rock art studies is statistical analysis of spatial distribution of rock art sites and its diagnostic images, specifically their relationship to topographic features and environmental variables, as well as to other archaeological sites. The rich tapestry of images in this region first came to the attention of antiquarians in the 1890s. Later, in the 1930s, it received international acclaim in dozens of short reports in natural history journals (a detailed history of investigations of the archaeology of Nine Mile Canyon is given in Spangler 2013). These reports offered up fantastical speculations on the meanings of the images. Many researchers of that day interpreted the rock art scenes as mass migrations of Toltecs and Aztecs from Mexico, rampant internecine warfare, and militaristic annihilations by invading groups (see Montgomery 1894 and Reagan 1934 as two examples). Many of these notions persist today through continued use of terms like “headhunters” and “fortresses” in the popular media (see Alberty 2017 for the most recent example). This initial review of the spatial distribution of rock art in Nine Mile Canyon as individual image classes or as clusters of images has not produced dramatic narratives in the manner of those early interpretations, and it has brought up more questions than answers about the nature and function of the art. Most importantly, it has shown that the analysis of a database such as that collected for the CPAA project can find that within a single cultural tradition there may have been several different reasons for producing the images. It also shows that within a tradition that includes the production of scenes, there are other categories of art production that cannot be construed as scenes. The evidence presented here indicates there is a good case to argue that some images were produced casually while others might have been art produced for special purposes, including ritual in some instances. Finally, the spatial analysis presented here suggests that scenes may often have been produced at a short distance from settlements and fields, whereas both formal and informal individual figures were often produced at some distance above settlements, suggesting the possibility of a different function. Jerry Spangler is a professional archaeologist and executive director of the Colorado Plateau Archaeological Alliance, a Utah-based nonprofit that works collaboratively with governments, private entities, and conservation organizations to develop strategies that will further the protection and preservation of historic and

294

JERRY D. SPANGLER AND IAIN DAVIDSON

archaeological sites on public lands in the western United States. He is a recognized expert on prehistoric Fremont and Ancestral Puebloan peoples of the northern Colorado Plateau, and has published widely in peer-reviewed journals, monographs, and scholarly books, including four books related to Nine Mile Canyon. He is currently involved in collaborative partnerships with federal and state agencies, as well as various universities. Iain Davidson has worked on Spanish Upper Palaeolithic Art, archaeology and ethnography of Northwest Queensland, Australian rock art, colonization of Sahul, language origins, and cognitive evolution. He has taught in England, Australia, and the United States, and has also worked with ten different Aboriginal groups and undertaken major archaeological consultancy for many of Australia’s leading industries. His publications include four books, seven edited volumes, more than sixy chapters and eighty articles in journals. He is a Fellow of the Australian Academy of Humanities and was awarded the Rhys Jones Medal of the Australian Archaeological Association 2010.

References Alberty, E. 2017. “Severed Heads Tell a Violent Mystery at Utah’s Dry Fork Canyon Petroglyphs.” Salt Lake Tribune, 12 September 2017. Gillin, J. 1938. “Archaeological Investigations in Nine Mile Canyon.” Bulletin of the University of Utah 28(11): 1–48. Macintosh, N. W. G. 1977. “Beswick Creek Cave Two Decades Later: A Reappraisal.” In Form in Indigenous Art, ed. P. J. Ucko, 191–7. Canberra: Australian Institute of Aboriginal Studies. Matheny, R. T., D. G. Matheny, P. Miller, and B. Miller. 2004. “Hunting Strategies and Winter Economy of the Fremont as Revealed in the Rock Art of Nine Mile Canyon.” In New Dimensions in Rock Art, ed. Ray T. Matheny, 145–93. Provo, Utah: Museum of Peoples and Cultures Occasional Papers Series No. 9. Matheny, R. T., T. S. Smith, and D. G. Matheny. 1997. “Animal Ethology Reflected in the Rock Art of Nine Mile Canyon, Utah.” Journal of California and Great Basin Anthropology 19(1): 7–103. Montgomery, H. 1894. “Prehistoric Man in Utah.” The Archaeologist 2(8): 227–342. Munn, N. 1973. Walbiri Iconography. Graphic Representation and Cultural Symbolism in a Central Australian Society. Ithaca: Cornell University Press Rappaport, R. A. 1999. Ritual and Religion in the Making of Humanity. Cambridge, UK: Cambridge University Press. Reagan, A. B. 1933. “Anciently Inhabited Caves of the Vernal (Utah) District, with Some Additional Notes on Nine Mile Canyon, Northeast Utah.” Transactions of the Kansas Academy of Sciences 36: 41–70. ———. 1934. “Evidence of a Possible Migration in the Very Dawning Period of PUEBLO culture.” Primitive Man 7(1): 12–14. Ross, J., and I. Davidson. 2006. “Rock Art and Ritual: An Archaeological Analysis of Rock Art in Arid Central Australia.” Journal of Archaeological Method and Theory 13(4): 305–41. Schaafsma, P. (1971) 2002. “The Rock Art of Utah: From the Donald Scott Collection.” In Papers of the Peabody Museum No. 65. Cambridge, MA: Harvard University. Reprint 2002, Salt Lake City: University of Utah Press. Simms, S. R. 2010. Traces of Fremont. Society and Rock Art in Ancient Utah. Photographs by F. Gohier. Salt Lake City: University of Utah Press. Spangler, J. D. 2011a. Formal Site Documentation and Analysis of Visitor Impacts at Warrior Ridge (42Dc1), Duchesne County, Utah. Ogden, Utah: Colorado Plateau Archaeological Alliance. ———. 2011b. Of Owls and Cranes: A Cultural Resource Inventory of Section 35, Township 11 South, Range 14 East, Middle Nine Mile Canyon, Duchesne County, Utah. Ogden, Utah: Colorado Plateau Archaeological Alliance. ———. 2011c. The Pete’s Canyon Complex: Formal Site Documentation and Analysis of Visitor Impacts in Nine Mile Canyon, Duchesne County, Utah. Ogden, Utah: Colorado Plateau Archaeological Alliance. ———. 2012. The Upper Fringe: Archaeological Inventory in Upper Nine Mile Canyon, Carbon County, Utah. Ogden, Utah: Colorado Plateau Archaeological Alliance. ———. 2013. Nine Mile Canyon: The Archaeological History of an American Treasure. Salt Lake City: University of Utah Press. ———. 2014. The Devil’s Due: Class III Cultural Resources Inventory in the Devils Canyon Area of Nine Mile Canyon, Carbon County, Utah. Ogden, Utah: Colorado Plateau Archaeological Alliance. Strevell, C. N., and C. S. Pulver. 1935. Report of the Utah State Museum Association Archaeological Expedition. Salt Lake City: Marriott Library Special Collections, University of Utah.

==== •

•

•

19. INTERPRETING SCENES IN THE ROCK ART OF THE CANADIAN MARITIMES Bryn Tapper and Oscar Moro Abadía

Introduction Since the beginnings of rock art research, archaeologists and art historians have been particularly interested in the interpretation of scenes, that is, those representations that seem to depict actions, events, or activities. Because of the narrative nature of these images, Western interpretations of rock art scenes have often reflected the biases and prejudices that have oriented rock art research during the last hundred and fifty years. It is well known, for instance, that Henri Breuil described the “White Lady of Brandberg,” a Namibian pictograph representing a shaman performing a ritual dance, as “a young woman with a typically Mediterranean, perhaps Cretan profile” (Breuil 1952: 236). This kind of ethnocentrism remained largely unchallenged until the last decades of the twentieth century. However, during the past twenty years a number of theoretical developments have promoted a less biased understanding of rock art scenes. In the first place, given the postmodernist focus on subjectivity and relativism, Western scholars have become increasingly conscious of the many prejudices that orient their interpretations. In the second place, new theoretical frameworks such as “new” animism1 and the “ontological turn”2 have provided space for alternative readings of rock art images that engage Indigenous theories (Alberti 2016; Ingold 2000). With these developments in mind, we examine several scenes from the rock art of the Canadian Maritimes. First, we briefly present the context and the chronology of the petroglyph traditions of the Eastern Algonquian peoples in this region. We also examine the history of research in the region in order to show how these images have traditionally been interpreted. Second, we suggest an alternative interpretation of these images influenced by new interpretive frameworks such as ontology and animism (Jones 2017). To begin, we describe the representations of French Lake (New Brunswick), a pre-contact rock art site that is stylistically related to the Algonquian Woodland period rock art known from many places of Central and Eastern Canada and the northeastern United States. We then examine a number of scenes from the post-contact sites of Kejimkujik and McGowan Lakes (Nova Scotia). We argue that the content, style, and purpose of these pre- and post-contact petroglyph traditions changed as a result of the social, political, religious, and economic upheaval that accompanied European colonization. While both traditions were grounded in relational ontologies, the nature of the relationships reflected in the scenes appears to have been reconfigured as a result of colonization. In the pre-contact rock art, the artist, the imagery, the site, and the landscape are implicated in an esoteric performance negotiated by ritual specialists for social benefit. The post-contact rock art emphasizes individual and communal biographies, as well as narratives concerned with recording a rapidly changing Indigenous way of life; its content was a blend of contemporary concerns interspersed with fragments of cultural memory and traditional knowledge.

The Algonquian Rock Art of the Canadian Maritimes: Overview and History of Research. The rock art of the Algonquian speaking peoples of Canada has been documented since the nineteenth century (e.g., Mallery 1894; Schoolcraft 1851–57). In particular, the pictograph traditions of the Canadian

296

BRYN TAPPER AND OSCAR MORO ABADÍA

Shield and Great Lakes have been the object of numerous studies during the last six decades (e.g., Arsenault 2004; Conway 2010; Dewdney and Kidd 1962; Rajnovich 1994; Vastokas and Vastokas 1973; Zawadzka 2016). Less well known are the petroglyphs of the Eastern Algonquian peoples of the Canadian Maritimes, a region comprising the provinces of Nova Scotia, New Brunswick, Prince Edward Island and those parts of Québec on the Gulf of St. Lawrence. In this region, fifteen sites have been recorded across eight locations. The majority of the sites are in southwestern Nova Scotia, where some of the largest known Eastern Canadian concentrations of historic period rock art are found on the shores of Kejimkujik and McGowan Lakes. A number of smaller sites, including those around Bedford Basin on the Atlantic coast and in southern New Brunswick on the Saint John River watershed, outlie these larger concentrations. This region constitutes a privileged space for examining important questions concerning the Algonquian-speaking peoples of the Canadian Maritimes, including the impact of European colonization from the early seventeenth century onward. This is related to the fact that the rock art in the region falls into two chronological periods. First, archaeologists have identified a pre-contact rock art tradition that is stylistically and technically related to Algonquian Woodland period (2800–500 years BP) rock art found elsewhere in the northeast, in Maine, Québec and the Canadian Shield. The available dates for this pre-contact tradition seem to indicate that many pictographs and petroglyphs were produced within the past three millennia (Arsenault et al. 1995: 27; Hedden 2004; Lenik 2002: 66; Rajnovich 1994: 41; Vastokas and Vastokas 1973: 24–25). In the case of the Canadian Maritimes, evidence for pre-contact petroglyphs is currently limited to the site of French Lake, in southern New Brunswick. Second, archaeologists have also identified a postcontact tradition in the Atlantic Northeast which belongs to the period following European contact, and which differs in manufacture technique, style and subject matter from the earlier tradition (Hedden 2002; Lenik 2002; Molyneaux 1984a, 1988, 1989). The history of archaeological research into this area follows a trajectory similar to that found in other rock art regions of North America. The formalist analyses prevalent during most of the twentieth century were largely based in cultural history and typically overlooked or ignored the ideological foundations of this art (Schaafsma 1985; Vastokas 1990b: 4–6). Furthermore, the motives underpinning the interpretation of rock art creation tended to be framed within Western notions of artistic expression. A number of examples illustrate this point. For instance, J. Walter Fewkes (Fewkes 1890: 996) suggested that the reason for the production of the petroglyphs at Fairy Bay on Kejimkujik Lake was “an almost involuntary tendency to add one’s name to a wall where others have placed theirs.” Similarly, the naturalist W. F. Ganong (1904: 177) asserted that if the petroglyphs at French Lake were Indigenous, they were “carved by Indian youths in moments of leisure, just as our people carve their names upon prominent places.” Even Garrick Mallery perceived the petroglyphs as art for art’s sake when he commented on the Kejimkujik sites: “. . . the Micmacs3 . . . had gained the idea of practicing art for itself, not merely using the devices of pictography for practical purposes, such as to record the past or to convey information” (Mallery 1894: 469). Such ethnocentric statements positioned the petroglyphs within the evolutionary hierarchy of Western art history aesthetics, meaning that the cultural differences that exist in the motivations for creating rock art were often ignored or distorted. This situation began to change from the 1970s onward when some authors suggested that the rock art of the Algonquian-speaking peoples had links to linguistic traditions and served to articulate, synthesize, and impose order on the practices that reflected Algonquian worldviews (Molyneaux 1980; Rajnovich 1989: 180; Vastokas and Vastokas 1973: 33). This reflected the recognition among researchers that Algonquian peoples enact beliefs that socialize the landscape with those beings—humans and other-than-humans— that inhabit it (Arsenault 2004; Buggey 1999; Hallowell 2010 (1960); Hornborg 2008; Paper 1990; Taçon 1990; Vastokas 1990a). Within this animist framework, the physical environment is charged with spiritual energy. This power is concentrated at specific locations, often remarkable or dramatic parts of the landscape such as mountains, cliffs, caves, boulders, and evocative rock formations as well as lakes, waterfalls, and rapids (Buggey 1999: 6; Paper 1990: 45; Taçon 1990: 13; Vastokas 1990a: 59; Vastokas and Vastokas 1973: 48). Ethnographic evidence indicates that animism, including the recognition that spiritual power resides in unusual and distinctive rocky places or with the beings inhabiting such places, was shared by Eastern Algonquian peoples, including the ancestral Mi’kmaq, Wolastoqiyik, and Peskotomuhkati of the Canadian Maritimes (Hagar 1896: 173; Hoffman 1955: 426; Hornborg 2008: 37–38, 92; Johnson 1943: 67; Leland

INTERPRETING SCENES IN THE ROCK ART OF THE CANADIAN MARITIMES

297

1884: 344–45; Mechling 1959; Parsons 1925: 86-87; Sable 2011; Sable and Francis 2012: 43; Wallis and Wallis 1955: 154–5; Whitehead 2006: 4). The renewed focus on animism has converged with a pressing need to recognize Indigenous realities when interpreting the archaeological record (Alberti and Marshall 2009). As an example of this approach, some authors have suggested that in the context of the ontology of Algonquian rock art, liminal places in the landscape are those that mark the interface of water, rock/earth, and sky, which equate to the multiple and tiered realms of the ancestral Algonquian shamanistic cosmos (Creese 2011; Hornborg 2006; Rajnovich 1994: 35; Taçon 1990: 13; Zawadzka 2016: 286). These junctions are considered places that link the physical and spiritual worlds, locales where communication with spiritual beings and powers was possible and most easily achieved. Deemed to be the most appropriate places to undertake shamanic vision quests and other rituals, at some sites they were manifested as rock art. The rock art enabled ritual specialists (and others) to access and activate the inherent spiritual energy of these places and to solicit the help of local spirit beings and guardians, thus reinforcing the underlying ontological principles and ethical codes governing concepts of belonging with the land and with others within the structure of the cosmos (Arsenault 2004: 301; Arsenault and Zawadzka 2014; Buggey 1999: 8; Creese 2011: 16; Taçon 1990; Vastokas and Vastokas 1973: 48). Recognition of this ontological position offers the means to demonstrate how the region’s petroglyph sites and their content may be framed in a way that retains the integrity of the original Indigenous valorization of the artwork embedded in its landscape and cosmological context (Zawadzka 2016: 305). In this sense, scenes need not be limited to the imagery but may extend to the relations inherent in the artist, the landscape and the beings (human and other-than-human) therein. Scenes are unlikely to have been understood in literal terms; rather, they would have been comprehended as nodes in Indigenous places and cosmographic networks extending through society and across landscape.

The site as the scene: French Lake in New Brunswick The only confirmed pre-contact rock art site in New Brunswick is located on the southwestern shore of French Lake (Figure 19.1), a small lake on the lower Oromocto River and on the traditional boundary of Wolastoqiyik and Peskotomuhkati territories. The petroglyphs, which were first documented and photographed by W. F. Ganong in 1903, comprise an anthropomorph, a beaver, two white-tailed deer and two amorphous shapes pecked in an iconic and abstract style on a small in situ quartzite boulder located in the littoral zone of the lakeshore (Ganong 1904) (Figure 19.2). The date of the petroglyphs remains imprecise; artifacts recovered from a number of places around the lake suggest a Middle-Late Archaic and Woodland chronology for the area (Allen and Forbes 1974). The anthropomorph is shown front-on and is located on the highest part of the water-smoothed boulder. The figure is stylistically similar to anthropomorph motifs found in the Algonquian rock art of the Canadian Shield, Québec, and Maine. In particular, it exhibits motif attributes similar to those found in the sequence of anthropomorph motifs recorded in the extensive tradition of rock art at Machias Bay in eastern Maine, motifs that exhibit postures and gestures thought to characterize shamanic practice during the Late Woodland (1000–500 years BP) and Protohistoric periods (Hedden 2004). The hourglass body shape and unidirectional feet denote an “active” posture, and raised arms bent at the elbows with extended splayed fingers are commonly interpreted as the Algonquian sign for giving or receiving “gifts” (i.e., “medicine”) (Arsenault et al. 1995; Hedden 2004; Rajnovich 1994; Vastokas and Vastokas 1973). The elongated nature of the figure’s left hand may suggest an object is held, perhaps a rattle or some form of hunting weapon. Below the splayed fingers of the figure’s right arm a cluster of peck marks form an amorphous globular shape; the iconicity of this motif is unclear although it may represent a medicine bag. To the left of the anthropomorph, a beaver motif in plan view with a head, four limbs, and a broad tail is clearly discernible. The motif is aligned between natural clefts on a smooth part of the boulder’s surface that slopes gently downward to the north. To the right of the anthropomorph, two white-tailed deer are arranged side by side in profile. They appear to overlap slightly, and both face northward toward the lake. Both motifs exhibit the characteristic raised curved tail indicative of the deer’s alarm posture, and the crooked limbs of the rear motif appear particularly

298

BRYN TAPPER AND OSCAR MORO ABADÍA

indicative of the deer’s distinctive leaping action. Stylistically, the French Lake site could be considered an outlier or satellite to the tradition found at Machias Bay. The site’s location, adjacent to a major traditional travel route along the Oromocto River, fits a wider Algonquian pattern of rock art site distribution. Most are found on the shores of lakes and rivers that served as the principal communication networks through the landscape (e.g., Zawadzka 2013). Within this general pattern, certain places may have been sought out for their phenomenal qualities; places considered by the artist to be suffused with significant spiritual charge. Although an accessible if not an especially prominent or otherwise remarkable place in the physical landscape, the shoreline location of the petroglyph boulder would appear to echo this Algonquian recognition of the liminality of such places. The petroglyphs also appear to reference landscapes at a different scale, where the micro-landscape of the petroglyph boulder relates to the macro-landscape of the lake and its environs (eg., Gjerde 2013). The motifs articulate together in a scene that can most easily be “read” from the south, from the shoreline facing northeast onto the lake. The juxtaposition of the motifs suggests that they are arranged on the surface of the boulder in such a way so as to reflect the topography of the immediate landscape. The surface of the boulder dips away to the north (toward the water) and the beaver is positioned on a slight slope in the boulder as if accessing the lower part of the boulder. The precise placement of the motif may signify the metaphorical entry of the beaver into the water of the lake. The anthropomorph and the deer motifs are arranged at the top of the slight slope on the higher southern part of the boulder (the break in slope is marked by a series of deep natural clefts), and which may conceivably represent elevated terrain. In this scenario, the “active” posture of the anthropomorph might be interpreted as a depiction of the honoring and petitioning of the spirits of game animals of water (beaver) and of land (deer) from the liminal shore. The zoomorphs appear to be orientated (cardinally) toward the habitats the animals occupy in the physical landscape (beaver to the north and the lake, deer to the east and the woods).

FIGURE 19.1. View from the position of the petroglyph boulder to the southeast, along the southern shore of French Lake, Oromocto River, New Brunswick. Copyright Bryn Tapper and Oscar Moro Abadía.

INTERPRETING SCENES IN THE ROCK ART OF THE CANADIAN MARITIMES

299

FIGURE 19.2. The French Lake petroglyph boulder artificially lit using the Highlight-Reflectance Transformation Imaging technique. Copyright Bryn Tapper and Oscar Moro Abadía.

When approached in relational terms, the petroglyph boulder cannot be understood simply as a static and isolated physical object on the lakeshore, divorced from its social and environmental contexts. Rather, the boulder and the petroglyphs are situated in a practice, in which the process of making the motifs is embedded in a ritual performance undertaken by the artist (see Vastokas 1992). In this sense, the images formed an integral component of a practice rather than simply being a later representation of the practice undertaken; the process of making rock art, pecking at the surface of the rock, was just as important for invoking communication with spiritual beings as the imagery itself (Lewis-Williams and Dowson 1990). In this way, the petroglyphs resist categorization as a representational scene in the Western artistic sense. Rather, the performer, the petroglyphs, the boulder, the lake, and the wider topographies (both physical and intangible) coalesce to form “the total interactive field . . . the cosmos as a whole”; effectively a scene that is the node of various related threads of a network that come together at the moment of the performance (Vastokas 1992: 27). By extension, the motifs comprising the scene are not merely representations of beings, although their iconography facilitates their identification, but agents participating in the performance and entering into social relationships with each other (Vastokas 1992: 19; Zawadzka 2016: 202). In this sense, the petroglyph boulder may be understood as a microcosmic metaphor of the Algonquian macrocosm “performed” at the liminal juncture (axis) where the tiers of the cosmos meet and where communication between realms, and the spiritual beings that inhabit them, was most easily achieved and “medicine” obtained. In these terms, it is conceivable that the petroglyphs constituted part of a ritual concerned with maintaining reciprocal relations with animals for the purposes of sustaining hunting success.

300

BRYN TAPPER AND OSCAR MORO ABADÍA

Changing scenes: Mi’kmaq and colonialism at Kejimkujik and McGowan Lakes, Nova Scotia In the Mi’kmaw district of Kespukwitk, the petroglyphs at Kejimkujik and McGowan Lakes depict various aspects of the lives of Kiskukewe’k L’nuk (historic Mi’kmaq). The rock art is located on glacially scoured and polished bedrock of metamorphosed slate that outcrops along the eastern shorelines of the lakes, although those at McGowan Lake are now submerged following dam construction in the 1940s. The petroglyphs are incised using sharp lithics or metal tools, producing detailed but faint petroglyphs easily eroded by ongoing wave and ice action. The degree and rate of erosion makes it difficult to identify the date and longevity of rock art production, and untangling the superimposition of many of the visible petroglyphs is challenging (Molyneaux 1988: 52). It is possible that petroglyphs dating to earlier periods existed but have subsequently disappeared (Mallery 1894: 39). Nevertheless, some superimposition and juxtaposition, in combination with subject matter, allows archaeologists to establish a tentative relative chronology for many of the visible petroglyphs (Molyneaux 1989). The Indigenous presence in the Kejimkujik area spans the past four millennia, and the lake lies on the Mersey River, the traditional travel route linking the Bay of Fundy to the Atlantic coast (Ferguson 1993). However, the surviving petroglyphs at Kejimkujik and McGowan Lakes, similar in cultural content, style and manufacture technique, date mainly from the seventeenth to nineteenth centuries. From the late eighteenth century onward, settlement of the interior increased rapidly and land use intensified as immigrant populations arriving with British colonial rule expanded. By the mid-nineteenth century, farms and small towns were established around the lakes, logging was a major industry, and game animals and fishing drew large numbers of American and European sports hunters (Morrison 1977; Parker 2004). A Mi’kmaw presence in the area continued throughout this period, and families and individuals were drawn into farming, heavy industries, and commercial opportunities such as guiding and commodity-artwork that came with the opening up of the interior. Many of the petroglyphs depict the Mi’kmaw experience of these changing realities. The petroglyph sites have been the object of survey and conservation programs since the late nineteenth century, and analyses of almost seven hundred petroglyphs identified as Mi’kmaq in origin have established a broad suite of socioeconomic, political, and religious themes, including numerous scenes showing hunting and fishing activities, maritime-related subsistence and travel, and religious activities incorporating iconography relating to Christianity and Indigenous spiritualism (e.g., Fewkes 1890; Mallery 1894; Molyneaux 1984a, 1984b; Myers 1972, 1976). These petroglyphs differ markedly in style and content from the pre-contact art traditions we have discussed. For instance, many of the animal motifs depicted by the carvers have moved away from the abstract forms characteristic of zoomorphic motifs found at pre-contact sites such as French Lake, and Embden and Machias Bay in eastern Maine (Hedden 1989: 25). Rather, they are characterized by the emergence of distinct new configurations of motifs that are often naturalistic, animated and biographical in nature. This change in style, with an emphasis on naturalism, is likely to reflect the Indigenous artists’ increasing familiarity with nineteenth-century Western concepts and conventions of artistic expression (Hedden 1989: 27). The imagery created by Mi’kmaw people at Kejimkujik and McGowan Lakes is likely to be associated with their communal oral histories rather than the conventional inventory of motifs associated with the esoteric and private knowledge characteristic of Algonquian shamanism. Generally, the imagery better fits the category of public or “secular” rock art by which Mi’kmaw people recorded their daily practices, and through which they communicated and commemorated traditional knowledge, identity, and values (Hedden 1989: 27; Mallery 1894: 42). Essentially, the petroglyphs chronicle how some Mi’kmaw people adapted to challenges presented and opportunities afforded by a rapidly changing way of life. Yet, within the petroglyph corpus there is evidence of an enduring Mi’kmaw cultural memory.

Scenes of Mi’kmaw cultural change and adaptation Scenes are common in the petroglyphs at Kejimkujik and McGowan Lakes, and certain panels exemplify the increasingly biographical trend in the rock art of the historic period. Compared to the pre-contact arrangements, these historic scenes are mainly representational and narrative, acting as a record of action rather

INTERPRETING SCENES IN THE ROCK ART OF THE CANADIAN MARITIMES

301

than the performance of action. At McGowan Lake, Brian Molyneaux has used a complex palimpsest of Mi’kmaw petroglyphs (Figure 19.3) to examine the changing conceptual relationship that occurred between Mi’kmaq and animals during the historic period (Molyneaux 1989). Molyneaux critically assessed Cheska’s interpretations of the petroglyphs, in which she argued that long sustained relationships between Indigenous peoples and animals progressively deteriorated and degenerated during the historic period, and that this was indicative of the breakdown and fragmentation of the pre-contact Mi’kmaw worldviews (Cheska 1981). The palimpsest consists of three strata, each comprising a scene of related images. The earliest stratum contains an abstracted human figure next to a European-style sailing vessel whose hull shape and rigging suggest it came from the late seventeenth century, and the French colonial period; this scene may record the experience of early Mi’kmaq-European contact (Molyneaux 1989: 201, Figure 8.2). The middle stratum comprises three ranks of at least six moose motifs, both cow and bull, etched to portray behaviors associated with the autumnal rut. The motifs are incised in a naturalistic and sketchy style similar to the nineteenth-century quadruped petroglyphs at Grand Lake Stream in eastern Maine (Hedden 1989). In Molyneaux’s relative chronology, the depiction of the moose reflects traditional Mi’kmaw concerns about hunting, and he places the creation of the glyphs sometime during the period of British colonization (i.e., the early eighteenth to early nineteenth century). At this time, the Mi’kmaq still maintained a semblance of their seasonal round, and despite encroachment, the interior of southwestern Nova Scotia served as a winter hunting territory for groups on the coast (Hoffman 1955: 172; Molyneaux 1989: 203). The appearance of these moose motifs may be related to game depletion (especially of moose), first reported in the seventeenth century and particularly pronounced by the mid-late eighteenth century, when major waves of European and other immigrant settlers arrived (Reid 1995: 27–28; Prins 1996: 106). The upper stratum comprises a number of human figures, perhaps a small family group, clothed in typical nineteenth-century Mi’kmaw attire, and also includes an earlier cow moose rendered into a woman on horseback. This scene is thought to commemorate an important Roman Catholic occasion observed by the Mi’kmaq (Molyneaux 1984b: 251; 1989: 204–5, Figs. 8.4 and 8.5).

FIGURE 19.3. Palimpsest of petroglyphs from McGowan Lake, Nova Scotia (see Molyneaux 1989: 20, Fig. 8.1) Copyright Ethnology Collection, Nova Scotia Museum (Ref. BcDg-4, 1,177).

302

BRYN TAPPER AND OSCAR MORO ABADÍA

The changes in content and style across these three layers, transitioning from static, generalized, and passive representations of humans and animals in the first two strata to highly ornamented, animated and individualized human figures in the upper stratum seems to reflect conceptual changes in Mi’kmaw society, in which the balance between social and individual worldviews was adjusted. Generally, the artistic expression becomes increasingly representational, active, and personalized. The appropriation of the moose motif is seen as particularly indicative of the changing attitude toward animals; the traditional community of humans and animals had broken down to be replaced by specifically human and individualized concerns (Molyneaux 1989: 211). This form of alteration and incorporation of static zoomorph motifs into animated scenes is evident elsewhere in the rock art of Kejimkujik Lake, as a number of hunting scenes attest (e.g., Molyneaux 1984b: 87; 1989: 207, Figure 8.7; Myers 1976: 123–25). At Fairy Bay, a zoomorph appears to have been variously rendered as a cow moose, a caribou with antlers, and even a horse with a long tail. At some point the quadruped was incorporated into a hunting scene in which a hunter (with a Mi’kmaw double-curve motif for its head) armed with a rifle is accompanied by two dogs (Figure 19.4). For the Mi’kmaq, hunting was a socially prescribed behavior with both spiritual and physical parameters; it was a pact, articulated in the Mi’kmaw concept netukulimk,4 between inter-dependent persons (humans and animals), intended to guide and foster respect and sustainable survival for all (Hoffman 1955: 503; Prosper et al. 2011: 5–6). During the post-contact period, however, the Mi’kmaw relationship with animals was jeopardized and became increasingly ambivalent under the stress of colonialism’s new economic models. The intensification of land clearance and forest harvesting led to widespread habitat destruction, while the commercial exploitation of wildlife and fisheries by settlers and game hunters (Gwyn 2003), abetted by colonial regulations restricting Mi’kmaw rights to hunt, meant that Mi’kmaw capacity to sustain netukulimk was “assailed” and long-held practices sanctioning the acquisition of resources eroded (Prosper et al. 2011: 7, 9–10). In this context, it is of interest to note that traditional Mi’kmaw practices regulating the disposal of animal remains following the hunt in the early twentieth century were seen to be “a weak survival of early Micmac

FIGURE 19.4. Hunting scene from Fairy Bay, Kejimkujik Lake. The quadruped appears to have been rendered variously as a moose, a caribou, or even a horse in the past. Copyright Bryn Tapper and Oscar Moro Abadía.

INTERPRETING SCENES IN THE ROCK ART OF THE CANADIAN MARITIMES

303

practice” recorded for the late seventeenth century (Wallis and Wallis 1955: 107). Nevertheless, Molyneaux contests Cheska’s assertion that by the nineteenth century the Mi’kmaq-animal relationship was one of abuse or even dislocation from the Indigenous past, preferring instead to regard the changing nature of the petroglyph scenes comprising the palimpsest as an expression of Mi’kmaw pragmatism in the face of a rapidly changing world. The colonial expansion along the coasts and rivers of southwestern Nova Scotia deprived Mi’kmaq of their traditional access and rights to their land, while the establishment of reserves from 1821 onward forcibly settled many Mi’kmaq inland, in places such as Kejimkujik Lake (Wicken 2012). Impoverished and marginalized, Mi’kmaq entered individualized economic relationships with animals in which the ancestral spiritual and metaphorical social relationships were reconfigured (Molyneaux 1989: 211–12).

Maritime scenes and the persistence of Mi’kmaw worldviews Despite their geographical marginalization, the Mi’kmaq of the eighteenth and nineteenth centuries at Kejimkujik Lake appear to have retained their strong cultural relationship with the ocean. Many of the petroglyphs at Kejimkujik and McGowan Lakes include motifs and scenes that relate to maritime economic activities. Canoes, sailing vessels, and the hunting of marine mammals and fish contribute substantially to the corpus of petroglyphs at these sites. Several scenes of cetacean hunting and fishing from distinctive ocean-going canoes with the high arched gunwales suggest that many Mi’kmaq continued to follow seasonal patterns of movement, even though the petroglyphs themselves are located on interior lakes. Meanwhile, a large number of sailing ship glyphs indicate an active Mi’kmaq participation in marine mercantile industry and travel (Molyneaux 1985, 1988). The sequence of European sailing vessel designs in the petroglyph corpus span the seventeenth to nineteenth centuries and help to establish a relative chronology for associated petroglyphs. The inscription of Mi’kmaw names on the hulls of ten of the sailing ship glyphs at Kejimkujik Lake suggests that many other ship glyphs can be reasonably attributed to Mi’kmaw artists (Committee for the Kejimkujik Petroglyphs 1994: 26). The precise form and intricate detail of many of the sailing vessels, demonstrated in the attention paid to hull and rigging designs, indicate familiarity with their technology and navigation; indeed, ever since the early historic period the Mi’kmaq had been widely recognized for their skillful seamanship (Molyneaux 1985: 6). It is also evident in some of the iconography at Kejimkujik Lake that traditional spiritual beliefs associated with the sea and/or water persisted in the cultural memory of the artists who incised canoes and sailing vessels on the lakeshore rocks (Molyneaux 1988: 63). This persistence is perhaps reflected in the toponymy of one of the small coves, Fairy Bay, at which concentrations of petroglyphs are found on the lake. A widespread belief recorded ethnographically among Eastern Algonquian groups is that other-than-human spiritual beings—“little people” or “fairies”5—live along the shorelines of lakes and rivers and are responsible for making petroglyphs, some of which serve as warnings about dangers that might threaten travelers (Parsons 1925: 96; Stamp 1915; Walker 1996). Despite early conversion to Roman Catholicism (indeed, much of the religious iconography found in the petroglyphs at Kejimkujik and McGowan Lakes, including churches, crosses, altarpieces and stars, indicates overt and committed Catholic observance), some images, such as serpent-like forms, may suggest that a degree of religious syncretism also existed in post-contact Mi’kmaw society (Bilodeau 2013; Chute 1992; Hornborg 2006: 314; Krieger 2002). The association of serpents and medicine is widely documented in Algonquian cultures, and references to the horned serpent, Jipijka’m, as the keeper or guardian of powerful medicine are common in Mi’kmaw legends (Hagar 1896; Parsons 1925; Sable 1997). Usually, the horned serpent other-than-human being is associated with water bodies and the Underwater world, and spiritual power is concentrated in its horns. This being could be propitiated with gifts or ritual to ensure benevolent actions or fortune, especially before journeys across potentially hazardous bodies of water. There are a number of reasons to interpret the sinuous motifs as Mi’kmaw representations of serpents, and even as other-than-human beings, rather than the eels that were historically fished on the lake and its rivers. Most images are depicted with the distinctive hatching or cross-hatching style indicative of Mi’kmaw artistic technique, and some of these display exaggerated morphological traits including horns, limbs, bifurcated tongues, and tooth-filled jaws. Many are also large, resisting any realistic sense of scale in relation to the crafts and vessels with which they are associated. Furthermore, in landscape terms, the shoreline loca-

304

BRYN TAPPER AND OSCAR MORO ABADÍA

tion of the Kejimkujik Lake rock art echoes the Algonquian recognition of liminal junctions: creating images of the other-than-human on the dipping rocks at the lake’s edge, where they might be lapped or submerged by changing water levels, might reflect the capacity of the images, indeed of the other-than-humans, to operate between different worlds. Of further interest is the way in which some of these petroglyphs also appear to articulate with natural features and other anomalies inherent in the rock surface. At Fairy Bay, at least two serpent motifs appear to emerge (or disappear) from natural clefts in the bedrock, again alluding to the these beings’ ability to move between the Underwater and Underground worlds and into the world of people (Whitehead 2006: 4). A precedent is found in Mi’kmaw ethnography, which conceives of cracks and holes in some large rocks as portals through which “little people,” the resident spirit beings, move between cosmic realms (Hagar 1896: 171–2). Of particular interest here is the juxtaposition and superimposition of sailing vessel motifs with serpent-like forms. The association of canoe and serpent motifs has been observed at many Algonquian rock art sites across the Canadian Shield and throughout the Northeast (Dewdney and Kidd 1962; Hedden 2002: 13; Lenik 2010). Molyneaux has noted similar arrangements of serpents with canoes and sailing ships at Kejimkujik Lake (Molyneaux 1980; 1985: 9–10; 1988: 60, Figure 8). Such combinations likely reflect acknowledgment of the other-than-human serpents inhabiting the lake (and other water bodies), and the appeasement necessary, through gifts or respect, to ensure safe travel (Molyneaux 1988: 62; Rajnovich 1994: 110). There are at least eight instances at different locales on Kejimkujik Lake where petroglyphs of canoes or sailing vessels articulate, either in superimposition or juxtaposition, with serpent motifs. At Fairy Bay, a petroglyph of two figures paddling a canoe is juxtaposed with a petroglyph of a large serpent with a bifurcated tongue (Figure 19.5) (Molyneaux 1985: 10; 1988: 52, Figure 1). Both petroglyphs are hatched and cross-hatched in the Mi’kmaw style, and despite the damage done by modern graffiti, the figures in the

FIGURE 19.5. Canoe motif with two occupants, juxtaposed with a serpent motif at Fairy Bay, Kejimkujik Lake. The canoe is partially obscured by modern graffiti. The inset shows the line drawing interpretation of the two motifs discussed. Copyright Bryn Tapper and Oscar Moro Abadía.

INTERPRETING SCENES IN THE ROCK ART OF THE CANADIAN MARITIMES

305

canoe still represent a Mi’kmaw man and woman, the man wearing a top hat in the stern, paddling, and the woman in the bow, wearing a distinctive peaked cap; the serpent appears to precede the canoe and its occupants on their journey. The arrangements at Kejimkujik Lake are still further of interest because of the way in which the majority of serpent motifs appear to be associated with sailing ships rather than canoes, suggesting that traditional spiritual concepts accompanied Mi’kmaw people in the emerging maritime economies in which they participated. Also at Fairy Bay, an unfinished petroglyph of a small sailing vessel is juxtaposed with a serpent petroglyph with a bifurcated tongue and tooth-filled jaws (Figure 19.6). Again, the distinctive cross-hatching on the serpent and along the bulwark of the vessel’s deck strongly suggests Mi’kmaw authorship, and the superimposed cross-hatched handprint reinforces this interpretation. Part of the rigging of the sailing vessel is superimposed by an inscription that includes the date 1886, indicating that both the vessel and the serpent petroglyphs were created before this time. As seen elsewhere, the way in which the serpent motif precedes the vessel, in this instance respecting the bowline and bowsprit, alludes to its guardianship of the voyage. Many of the maritime scenes depicted in the petroglyphs of Kejimkujik Lake are literal in nature and document historical practices. For example, Mi’kmaq engaged in the porpoise hunt, a commercial activity undertaken by canoe in the Bay of Fundy from the 1830s onward, and probably earlier (Wallis and Wallis 1955: 30–31). However, other scenes resist literal interpretation: the association of serpents and vessels reminds viewers of the persisting cultural memory of the Mi’kmaw artists, suggesting that even some historic period petroglyphs echo a sense of the performance found in the pre-contact petroglyphs of the region.

FIGURE 19.6. A reproduction of the petroglyph tracing of a small sailing vessel depicted in association with a serpent motif with bifurcated tongue and teeth-filled jaws from Fairy Bay, Kejimkujik Lake. Copyright Bryn Tapper and Oscar Moro Abadía.

306

BRYN TAPPER AND OSCAR MORO ABADÍA

Some Concluding Thoughts The analysis of the Maritime Algonquian petroglyphs illustrates some general problems related to the interpretation of rock art scenes. The case studies presented here illustrate the need to overcome the “contemplation model” when examining non-Western images. This model, which has defined the Western artistic experience since the eighteenth century, states that “what defines a work of art is its status as an object to be ‘contemplated’ and contemplated ‘disinterestedly,’ that is, attended to ‘as such,’ for its own sake” (Abrams 1985: 8). However, as the example of the petroglyphs at French Lake illustrates, these images may be better understood as agentive participants in performances designed to establish and sustain meaningful relationships between people and the other-than-human actors with whom they shared existence. The scenes in this rock art extend along the social networks implicated in the performance. Whereas the pre-contact sites typically display an abstracted iconic imagery that has its origins in the esoteric and private knowledge practices associated with traditional shamanic rituals (albeit for wider social benefit), the petroglyphs of the postcontact sites appear to have had a far more public orientation, in that the increasingly biographical motifs and scenes of Mi’kmaw life were designed to evoke remembrance in the mind of the viewer. The change in the content and style of the petroglyphs is related to the impact of European contact, which shattered many aspects of Indigenous life. By the eighteenth and nineteenth centuries, the Mi’kmaq were facing stark socioeconomic and political conditions, coupled with plummeting population and geographic marginalization. This situation is likely to have given impetus to the creation of petroglyphs as a means of retaining cultural identity. Certainly, some petroglyphs from the historic period speak to the strength and persistence of Mi’kmaw cultural memory, and in a number of the panels at Kejimkujik and McGowan Lakes we see tangible evidence of the syncretism observed in Mi’kmaw religion, at least as expressed by some people. But this Indigenous syncretism is feasible in part because of Indigenous tolerance of ontological alterity and ambiguity, rather than any particular permissiveness on the part of the Roman Catholic Church or colonial administrators (Bilodeau 2013: 4; Reid 1995: 87–88). In our assessment, the multifaceted rock art tradition thereby produced an assemblage that documents the Mi’kmaw experience of a changing world and drew upon elements of Western artistic techniques and conventions, yet retained an essentially Mi’kmaq attitude in terms of the identity and cultural memory expressed. It was not organized according to a prescribed template such as might be observed in the pre-contact shamanic rock art of Machias Bay or French Lake, but instead emerged and accrued as a result of ongoing Mi’kmaw negotiation, collectively and individually, of the impacts and opportunities that arose with the intensification of Euro-Canadian land use in interior southwest Nova Scotia. The palimpsests in the historic rock art illustrate the creators’ preparedness to “build” upon earlier images, and the accumulation of imagery and meaning over time reflects the evolving and changing relationship between Mi’kmaw people and their land. As the petroglyph sites became historicized as places in their own right, they moved beyond the intentions of their original creators to be incorporated, reconfigured, and reinvented by subsequent populations who, as users and caretakers, applied further levels of social meaning (Norder 2012: 386). Indigenous perceptions of the art’s dynamic role persisted: it was not considered a static, rigid, or inviolate medium, but one that reflected the Mi’kmaw peoples’ own ongoing transition and adaptation. As these examples illustrate, we cannot develop a better understanding of the Indigenous realities unless we are able to contribute, in the long run, to improving the understanding of rock art scenes. In an Algonquian context, this requires a commitment on the part of archaeologists to continue to undertake “relational rock art studies” that not only investigate the ontologies of rock art producers in the past but also critically examine the conceptual frameworks within which the research is itself pursued (Alberti 2016; Jones 2017; Zawadzka 2016). Bryn Tapper is a doctoral candidate in the Department of Archaeology, Memorial University of Newfoundland. His current research focuses on the rock art of the Mi’kmaq, Wolastoqiyik and Peskotomuhkati First Nations in Atlantic Canada. Oscar Moro Abadía is Associate Professor in the Department of Archaeology, Memorial University of Newfoundland. His primary research interests lie in the study of prehistoric and Indigenous arts, with a particular focus on the history and epistemology of prehistoric images.

INTERPRETING SCENES IN THE ROCK ART OF THE CANADIAN MARITIMES

307

Notes 1. “New” animism refers to a reconfiguration of the traditional concept of “animism” as a sort of primitive religion. In this new sense, and in an Algonquian context, the term refers to a type of relational ontology in which humans and other-than-humans ( including spirits, plants, animals, natural phenomena, and elements of the physical environment ) have personhood (Ingold 2000: 89-110). 2. In disciplines such as anthropology and archaeology, the “ontological turn” refers to the impact of a number of theoretical approaches that seek to reconfigure theory on the basis of Indigenous viewpoints. Central to these approaches is the idea that different groups of people not only perceive reality differently, but in fact live in different realities. 3. The Mi’kmaq call themselves L’nu’k meaning “the people”. The historic (British) ethnonym Micmac is, in both spelling and pronunciation, a corruption of the Mi’kmaw word nikmak, meaning “my kin-friends.” Today, the term Mi’kmaq is the preferred spelling of the ethnonym. 4. Netukulimk is a Mi’kmaw concept that operates as a “guide to responsible co-existence and interdependence with natural resources, each other and other than human entities. It is a . . . body of living knowledge, which underpins the moral and ethical relationships that explain [the Mi’kmaw] world in the past and provides for the present by sustaining the future” (Prosper et al. 2011: 3). Hunting, within this ontological framework, was “guided by actions and values intended to keep a respectful co-existence and co-dependence between human and animal.” (Prosper et al. 2011: 5). 5. Known as Wiklatmu’j to the Mi’kmaq (Wallis and Wallis 1955: 155).

References Abrams, M. H. 1985. “Art-as-Such: The Sociology of Modern Aesthetics.” Bulletin of the American Academy of Arts and Sciences 38(6): 8–33. Alberti, B. 2016. “Archaeologies of Ontology.” Annual Review of Anthropology 45(1): 163–79. Alberti, B., and Y. Marshall. 2009. “Animating Archaeology: Local Theories and Conceptually Open-Ended Methodologies.” Cambridge Archaeological Journal 19(3): 344–56. Allen, P., and G. Forbes. 1974. “Oromocto River Survey.” Unpublished manuscript. Fredericton: Archaeological Services, Government of New Brunswick. Arsenault, D. 2004. “From Natural Settings to Spiritual Places in the Algonkian Sacred Landscape: An Archaeological, Ethnohistorical and Ethnographic Analysis of Canadian Shield Rock-Art Sites.” In The Figured Landscapes of Rock-Art: Looking at Pictures in Place, ed. C. Chippindale and G. Nash, 289–317. Cambridge: Cambridge University Press. Arsenault, D., L. Gagnon, C. Martijn, and A. Watchman. 1995. “Le projet Nisula : recherche pluridisciplinaire autour d’un site à pictogrammes (DeEh-1) en Haute-Côte-Nord.” In Archéologies québécoises, ed. A.-M. Balac, C. Chapdelaine, N. Clermont and P. Desjardins, 17–57. Montreal: Recherches Amérindiennes au Québec. Arsenault, D., and D. Zawadzka. 2014. “Spiritual Places: Canadian Shield Rock Art within Its Sacred Landscape.” In Rock Art and Sacred Landscapes, ed. D. Gillete, M. Greer, M. H. Hayward, and W. Breen Murray, 117–37. New York: Springer. Bilodeau, C. 2013. “Understanding Ritual in Colonial Wabanakia.” French Colonial History 14: 1–31. Breuil, H. 1952. “The Influence of Classical Civilizations on the Cave Paintings of South Africa.” In Proceedings of the First Pan-African Congress on Prehistory, ed. L. S. B. Leakey, 234–37. Oxford: Blackwell. Buggey, S. 1999. An Approach to Aboriginal Cultural Landscapes. Ottawa: Parks Canada, Historic Sites and Monuments Board of Canada. Cheska, A. T. 1981. “The Micmac Petroglyphs: Evidence of Self/Nature Concept Changes.” Nova Scotia Historical Review 1(1): 54–66. Chute, J. E. 1992. “Ceremony, Social Revitalization and Change: Micmac Leadership and the Annual Festival of St Anne.” Papers of the 23rd Algonquian Conference, 1991, ed. W. Cowan, 45–61. Ottawa: Carleton University. Committee for the Kejimkujik Petroglyphs. 1994. “Mi’kmaq Culture History, Kejimkujik National Park, Nova Scotia.” Unpublished manuscript. Ottawa: Historic Sites and Monuments Board of Canada. Conway, T. 2010. Spirits on Stone: Lake Superior Ojibwa History, Legends and the Agawa Pictographs. Sault Ste. Marie, Ontario: Heritage Discoveries. Creese, J. L. 2011. “Algonquian Rock Art and the Landscape of Power.” Journal of Social Archaeology 11(1): 3–20. Dewdney, S., and K. Kidd. 1962. Indian Rock Paintings of the Great Lakes. Toronto: University of Toronto Press. Ferguson, R. 1993. “Kejimkujik Petroglyphs, Kejimkujik National Park, Nova Scotia: A Preliminary Report.” Unpublished manuscript. Halifax, N.S.: Parks Canada. Fewkes, J. W. 1890. “A Pictograph from Nova Scotia.” The American Naturalist 24(287): 995–99. Ganong, W. F. 1904. “Upon Aboriginal Pictographs Reported from New Brunswick.” Bulletin of the Natural History Society of New Brunswick 5: 175–8. Gjerde, J. M. 2013. “Stone Age Rock Art and Beluga Landscapes at River Vyg, North-Western Russia.” Fennoscandia archaeologica 15: 37–54.

308

BRYN TAPPER AND OSCAR MORO ABADÍA

Gwyn, J. 2003. “The Mi’kmaq, Poor Settlers, and the Nova Scotia Fur Trade, 1783–1853.” Journal of the Canadian Historical Association / Revue de la Société historique du Canada 14(1): 65–91. Hagar, S. 1896. “Micmac Magic and Medicine.” The Journal of American Folklore 9(34): 170–77. Hallowell, A. I. 2010 (1960). “Ojibwa Ontology: Behaviour and World View.” In A. Irving Hallowell, Contributions to Ojibwe studies: essays, 1934-1972, ed. J. S. H. Brown and S. E. Gray, 535–568. Lincoln: University of Nebraska Press. Hedden, M. 1989. “Petroglyph Evidence for a Possible 19th Century Survival of Algonkian (Passamaquoddy) Shamanism in Eastern Maine.” Maine Archaeological Society Inc. Bulletin 29(1): 21–28. ———. 2002. “Contact Period Petroglyphs in Machias Bay, Maine.” Archaeology of Eastern North America 30: 1–20. ———. 2004. “Passamaquoddy Shamanism and Rock-Art in Machias Bay, Maine.” In The Rock-Art of Eastern North America: Capturing Images and Insight, ed. C. Diaz-Granados and J. R. Duncan, 319–43. Tuscaloosa: University of Alabama Press. Hoffman, B. G. 1955. “The Historical Ethnography of the Micmacs of the Sixteenth and Seventeenth Centuries.” PhD dissertation. Berkley: University of California. Hornborg, A.-C. 2006. “Visiting the Six Worlds: Shamanistic Journeys in Canadian Mi’kmaq Cosmology.” The Journal of American Folklore 119(473): 312–36. ———. 2008. Mi’kmaq Landscapes: From Animism to Sacred Ecology. Aldershot, UK: Ashgate Publishing Limited. Ingold, T. 2000. The Perception of the Environment: Essays on the Livelihood, Dwelling and Skill. London and New York: Routledge. Johnson, F. 1943. “Notes on Micmac Shamanism.” Primitive Man 16(3/4): 53–80. Jones, A. M. 2017. “Rock Art and Ontology.” Annual Review of Anthropology 46(1):167–81. Krieger, C. J. 2002. “Culture Change in the Making: Some Examples of How a Catholic Missionary Changed Micmac Religion.” American Studies International 40(2): 37–56. Leland, C. G. 1884. The Algonquin Legends of New England, or, Myths and Folk Lore of the Micmac, Passamaquoddy, and Penobscot Tribes. Boston: Houghton, Mifflin and Company. Lenik, E. J. 2002. Picture Rocks: American Indian Rock Art in the Northeast Woodlands. Hanover and London: University Press of New England. ———. 2010. “Mythic Creatures: Serpents, Dragons, and Sea Monsters in Northeastern Rock Art.” Archaeology of Eastern North America 38:17–37. Lewis-Williams, D., and T. A. Dowson. 1990. “Through the Veil: San Rock Paintings and the Rock Face.” South African Archaeological Bulletin 45(151): 5–16. Mallery, G. 1894. Picture-Writing of the American Indians, Extract from the Tenth Annual Report of the Bureau of Ethnology (1894). Washington, DC: Government Printing Office. Mechling, W. H. 1959. “The Malecite Indians, with Notes on the Micmacs.” Anthropologica 8: 161–274. Molyneaux, Brian Leigh. 1980. “Landscape Images: Rock Paintings in the Canadian Shield.” Rotunda 13(3): 6–11. ———. 1984a. “An Analysis and Interpretation of the Micmac Petroglyphs of Kejimkujik National Park (3 Vols.).” Unpublished manuscript. Halifax, N. S.: Parks Canada. ———. 1984b. “The McGowan Lake Petroglyph Recording Project, Queens County, Nova Scotia, 1983.” Unpublished manuscript. Halifax: Nova Scotia Museum. ———. 1985. “Floating Islands: The Micmac Vision of Sailing Ships.” Rotunda 18(1): 6–13. ———. 1988. “Images de la mer dans l’art rupestre des Micmacs.” In Les Micmacs et la mer, ed. C. Martijn, 49–64. Montreal: Recherches Amérindiennes au Québec. ———. 1989. “Concepts of Humans and Animals in Post-contact Micmac Rock Art.” In Animals into Art, ed. H. Morphy, 193–214. Boston: Unwin Hyman. Morrison, J. 1977. Kejimkujik National Park: “We’ve Held Our Own”; An Oral and Documentary History of Northern Queens & Southern Annapolis Counties, Nova Scotia, Manuscript Report Number 238. Ottawa: National Historic Sites and Parks Branch, Parks Canada. Myers, H. B. 1972. “Report on the Recording and Mapping of Petroglyphs Located at Fairy Bay, Kejimkujik National Park (2 Vols.).” MS on file. Ottawa: National and Historic Parks Branch, Parks Canada. ———. 1976. “The Petroglyphs of Kejimkujik National Park, Nova Scotia: A Comprehensive Report.” Unpublished manuscript. Halifax: Parks Canada. Norder, J. 2012. “The Creation and Endurance of Memory and Place Among First Nations of Northwestern Ontario, Canada.” International Journal of Historical Archaeology 16(2): 385–400. Paper, J. 1990. “Landscape and Sacred Space in Native North American Religion.” In Perspectives of Canadian Landscape: Native Traditions, ed. J. M. Vastokas, 44–54. North York, Ontario: York University, Robarts Centre for Canadian Studies. Parker, M. 2004. Guides of the North Woods: Hunting and Fishing Tales from Nova Scotia 1860-1960. Halifax, N.S.: Nimbus Publishing. Parsons, E. C. 1925. “Micmac Folklore.” The Journal of American Folklore 38(147): 55–133. Prins, H. E. L. 1996. The Mi’kmaq: Resistance, Accommodation, and Cultural Survival. Fort Worth, TX: Harcourt Brace. Prosper, K., L. J. McMillan, A. A. Davis, and M. Moffitt. 2011. “Returning to Netukulimk: Mi’kmaq Cultural and Spiritual Connections with Resource Stewardship and Self-Governance.” The International Indigenous Policy Journal 2(4): 1–17.

INTERPRETING SCENES IN THE ROCK ART OF THE CANADIAN MARITIMES

309

Rajnovich, G. 1989. “Visions in the Quest for Medicine: An Interpretation of the Indian Pictographs of the Canadian Shield.” Midcontinental Journal of Archaeology 14(2): 179–225. ———. 1994. Reading Rock Art: Interpreting the Indian Rock Paintings of the Canadian Shield. Toronto: Natural Heritage/Natural History Inc. Reid, J. 1995. Myth, Symbol, and Colonial Encounter: British and Mi’kmaq in Acadia, 1700–1867, Religions and Beliefs Series, No. 4. Ottawa: University of Ottawa Press. Sable, T. 1997. “Multiple Layers of Meaning in the Mi’kmaw Serpent Dance.” Papers of the 28th Algonquian Conference, 1996. Winnipeg: University of Manitoba. ———. 2011. “Legends as Maps.” In Ta’n Wetapeksi’k: Understanding from Where We Come. Proceedings of the 2005 Debert Research Workshop, Debert, Nova Scotia, Canada, ed. T. Bernard, L. M. Rosenmeier, and S. L. Farrell, 157–71. Truro, Nova Scotia: Eastern Woodland Print Communications. Sable, T., and B. Francis. 2012. The Language of This Land, Mi’kma’ki. Sydney, Nova Scotia: Cape Breton University Press. Schaafsma, P. 1985. “Form, Content, and Function: Theory and Method in North American Rock Art Studies.” Advances in Archaeological Method and Theory 8: 237–277. Schoolcraft, H. R. 1851–1857. Historical and Statistical Information respecting the History, Condition and Prospects of the Indian Tribes of the United States. Philadelphia: Lippincott, Gambo & Co. Stamp, H. 1915. “The Water-Fairies.” The Journal of American Folklore 28(109): 310–16. Taçon, P. 1990. “The Power of Place: Cross-Cultural Responses to Natural and Cultural Landscapes of Stone and Earth.” In Perspectives of Canadian Landscape: Native Traditions, ed. Joan M. Vastokas, 11–43. North York, Ontario: York University, Robarts Centre for Canadian Studies. Vastokas, J. M. 1990a. “Landscape as Experience and Symbol in Native Canadian Culture.” In Perspectives of Canadian Landscape: Native Traditions, ed. J. M. Vastokas, 55–82. North York, Ontario: York University, Robarts Centre for Canadian Studies. ———. 1990b. Perspectives of Canadian Landscapes: Native Traditions. York University: Robarts Centre for Canadian Studies. ———. 1992. Beyond the Artifact: Native Art as Performance. York University, North York, Ontario: Robarts Centre for Canadian Studies. Vastokas, J. M., and R. K. Vastokas. 1973. Sacred Art of the Algonkians: A Study of the Peterborough Petroglyphs. Peterborough, Ontario: Mansard Press. Walker, W. 1996. “Wabanaki ‘Little People’ and Passamaquoddy Social Control.” Papers of the 27th Algonquian Conference, 1995, ed. D. H. Pentland, 353–61. Winnipeg: University of Manitoba. Wallis, W. D., and R. S. Wallis. 1955. The Micmac Indians of Eastern Canada. Minneapolis: University of Minnesota Press. Whitehead, R. H. 2006. Stories from the Six Worlds: Micmac Legends. Halifax, N. S.: Nimbus Publishing Limited. Wicken, W. C. 2012. The Colonization of Mi’kmaw Memory and History 1794–1928: The King vs Gabriel Sylliboy. Toronto: University of Toronto Press. Zawadzka, D. 2013. “Beyond the Sacred: Temagami Area Rock Art and Indigenous Routes.” Ontario Archaeology 93: 159–99. Zawadzka, D. 2016. “Cultivating relations in the Landscape: Animism and Agency in the Rock Art of Temagami Region, Northeastern Ontario.” PhD dissertation. Montreal: Université du Québec à Montréal.

==== •

•

•

20. THE BLACK SERIES IN THE HUNTING SCENES OF CUEVA DE LAS MANOS, RÍO PINTURAS, PATAGONIA, ARGENTINA Carlos A. Aschero and Patricia Schneier

Introduction This chapter analyzes the Black series, which is the total scene in black color from the Cueva de las Manos site, inscribed in the UNESCO World Heritage List. One of the earliest collective hunting scenes at Cueva de las Manos, it dates to before 8800 BP and has a set of specific characteristics, many of which influenced later scenes up to 7400 BP. This total scene, together with the Ochre series, which precedes it in dating and is painted beneath it, features an extensive spatial development on an ignimbrite wall 340 meters long. We define a “scene” as a group of representations that are spatially close to each other and among them have a clear anecdotal nexus or action that can be inferred by the observer. Each scene has a dominant monochrome unity (in this case black) that allows the observer to follow the scene even in cases of superposition by other scenes of different color. The concept of “total scene” applies when we observe an articulation between different compositions or partial scenes that extend over two or more rock panels along the 340-meter wall. These partial scenes represent the different actions of a collective hunt, the complete development of which is observed in the total scene. Therefore, we refer to “partial scenes” contained in a total scene that was conceived as a “unified visual image.” The interconnected scenes constitute a narrative, and their spatial display requires the observer’s movement through the set of actions portrayed in order to decode the complete visual narrative. The underlying theme in the Black series appears to be that of a collective hunt and the various actions required for its completion—that is, it does not commemorate a specific hunt or event. This may be inferred by the level of detail in the actions and weapons utilized, as by the usage of the micro-topography of the rock medium as a virtual scene in which the real topography of the hunting ground is portrayed, and by the inclusion of deer in the representation of a hunting scene otherwise completely dominated by the presence of guanacos—a combination that does not occur in the normal behavior of these species. This chapter is therefore an analysis of the total scene of a collective hunt that corresponds to the A2 Style, one of five styles of hunting scene paintings superimposed on the Cueva de las Manos wall.

Eco-Landscapes and Archeological Landscapes in the Río Pinturas Area Eco-landscapes The Rio Pinturas Canyon marks the Western limit of the Central High Plateau of Santa Cruz Province (Argentina). The river’s permanent flow is fed in part by its main tributary, the Ecker. Initially the Pinturas runs in a north-south direction; then it bends to take a south-north direction (middle and low waters) before finally flowing into the Deseado River (Figure 20.1). Cueva de las Manos, located in the upper banks of the Río Pinturas, is one of the archaeological sites containing rock art that is inscribed in the 1999 UNESCO World Heritage List.

THE BLACK SERIES IN THE HUNTING SCENES OF CUEVA DE LAS MANOS, RÍO PINTURAS, PATAGONIA, ARGENTINA

311

FIGURE 20.1. Northwestern Santa Cruz archaeological sites, Patagonia, Argentina. Copyright Carlos Aschero.

The Pinturas riverbed is bordered by ignimbrite cliffs over two hundred meters high, with slopes made of sand and debris from landfalls, and also some low terrace-like reliefs (Figure 20.2). The Cueva de las Manos and other rock shelters with rock art manifestations used by ancient communities of hunter-gatherers are found at the foot of these cliffs, where the slopes begin. Protected by the cliffs, a microclimate was formed in the level between 400 and 500 MASL, which during the Holocene offered abundant grasslands for local fauna to graze on, as well as various species of useful bushes like the “molle” (Schinus molle), which was not just firewood but also a source of resins for making adhesives. Inhabitants also found long useful molle branches from which to make wood artifacts and tall grasses for making into long sheaves to line the sandy floors of their shelters. There were abundant fruits like the “calafate” (Berberis sp.), tubers like the rush root, and other vegetables gathered for food and medicinal uses.

312

CARLOS A. ASCHERO AND PATRICIA SCHNEIER

FIGURE 20.2. View of Río Pinturas Canyon, Santa Cruz. Copyright Carlos Aschero.

In the fall and winter months, the Rio Pinturas Canyon and its immediate periphery (pampas and gorges) were host to a large concentration of guanaco herds seeking shelter. This hunting season was crucial to the survival of the hunter-gatherers. The guanaco (Lama glama guanicoe), a species of large animals with gregarious habits, provided native peoples with abundant meat as well as furs, bones, and wool. Hunter-gatherers were able to access the canyon by following certain gorges and ravines from the higher relief (600–700 MASL) in the east and west. These points of entry also served as effective “natural traps” for capturing guanaco by hunting in large groups or families. In these high reliefs of the pampas and gorges, guanacos were also hunted in spring, summer, and early fall. To the west-southwest of the Río Pinturas, the mountainous-lacustrine environment on the Andes’ eastern mountainsides offers a variety of altitudinal floors with good vegetable resources in the steppes and southern Nothofagus sp. forests, with guanaco and South Andean deer (Hipocamelus bisulcus) present as large game for hunting. Particularly at the higher level of this environment (900 MASL), the area containing the Belgrano and Burmeister lakes (currently Perito Moreno National Park) was occupied in the early Holocene by hunter-gatherers who took advantage of the numerous rock shelters and caves in the rhyolite formations of Cerro Casa de Piedra (Aschero et al. 2005).The so-called “Pampa del Asador” quarries, a vast area providing obsidian stone with a wide range of artifactual distribution in archeological sites of southern Patagonia over a range of 500 kilometers, is located on one of the old paths from Cueva de las Manos to Cerro Casa de Piedra. This black obsidian was used in production of artifacts throughout the entire Holocene (Espinosa and Goñi 1998). Along the same path, flint nodules of excellent quality could also be found in Laguna Cisnes, in the direction of Cueva de las Manos-Pampa del Asador (Figure 20.1). We therefore count three eco-landscapes exploited by these “people of the Pinturas” since the early Holocene. We understand an “eco-landscape” as the total set of physiographic components of each of these different altitudinal areas: (1) The Rio Pinturas Canyon and the Pampas, hills, and gorges in its immediate periphery, between 700 and 500 MASL; (2) The Andes mountainous-lacustrine environment, with the eco-

THE BLACK SERIES IN THE HUNTING SCENES OF CUEVA DE LAS MANOS, RÍO PINTURAS, PATAGONIA, ARGENTINA

313

tone between forests and steppes, at 800–900 MASL; and (3) The western area of the Central High Plateau, at 800–900 MASL.

Archaeological Landscapes Around 9500 BP (early Holocene), we find the beginnings of a seasonal nomadic circuit through these three eco-landscapes. Cueva de las Manos and the Cerro Casa de Piedra-7 site near Burmeister Lake were important landmarks on this circuit. The highlands to the east of Río Pinturas in the Central High Plateau at 900 MASL also formed part of the circuit, with summertime camps for hunting guanaco near temporary lakes that attracted the animals. This was so in the case of Cerro Bayo sites 1–3 (Figure 20.1), where we recently found lithic artifacts closely resembling those in Cueva de las Manos, layer 6, and the Cerro Casa de Piedra-7 lower occupational layers (Aschero and Civalero 2016). This seasonal nomadism made use both of the vegetable resources in each of these altitudinal zones and the different birth seasons of female guanacos in relation to these altitudinal differences. Hence, hunters had access to “chulengos” (newborn guanacos) in November around Cueva de las Manos, and in December and January in Cerro Casa de Piedra and the nearest highlands. Newborn guanaco furs were highly valued for the warm clothing and capes produced by the southern Tehuelche people (Aonikenk) in the historic period. In the period between 9500 and 6400 BP, human exploitation of these ecological landscapes established places and paths that constituted an initial “archeological landscape” (Ingold 2000). From 9500 to ca. 8000 BP, only Cueva de las Manos had rock art in this landscape, in the form of two styles of hunting scenes that we call A1 scenes (Ochre series) and A2 scenes (Black series). After 8800 BP, these early hunting scenes of Cueva de las Manos extended to other sites. In this period, this group of sites continued to form part of the seasonal nomadism circuit of the early Holocene. They are marked by rock art representations corresponding to styles A3, A4, and A5, together with the Cueva de las Manos collective hunting scenes (Aschero 2010, 2018).

Chronology The chronology of styles in the rock art scenes of Cueva de las Manos (a subject we will return to at the end of this chapter) is the following: • Styles A1 and A2 (ochre chromatic series and black chromatic series, respectively): ca. 9400–8800 uncalibrated years BP. • Styles A3 and A4 (red chromatic series and purple-red chromatic series, respectively): ca. 8800–7700 uncalibrated years BP. • Style A5 (white chromatic series): after ca. 7700 uncalibrated years BP, until the H1 eruption of the Hudson volcano, dated 6890+100 uncalibrated years BP (Prieto, Stern and Estevez 2013). Together, the Cerro Casa de Piedra sites 5–7, Cerro Bayo sites 1–3, and Cueva de las Manos constitute compound rock art site(s) (CRAS) where caves and rock shelters are complemented by temporary camps for the aggregation of large bands or extended family groups (Aschero 1996; Aschero et al. 2005). These aggregation camps were places (sensu Ingold 2000) where the best resources for hunting and gathering were concentrated. The topographical resources in the vicinity of the CRAS were another factor favoring such aggregation camps. They facilitated the collective hunts of guanaco, in which animals were intercepted and driven into gorges and lake and lagoon coasts limited by abrupt slopes, considered “natural traps.” This is very accurately expressed in the Black series of the hunting scenes treated below, where the micro-topography of the rock support works as a virtual setting portraying the use of real topography. In these CRAS we find representations of guanacos, human figures, hand stencils, and non-figurative signs constituting hunting scenes. Hand stencils also appear independently in a number of rock shelters and small caves in the Río Pinturas area and its environs. These hand stencils persist in the area until 2000 BP, that is, the period subsequent to the disappearance of the hunting scenes. Those seem to be linked to residential camps of reduced family groups or camps for limited activities.

314

CARLOS A. ASCHERO AND PATRICIA SCHNEIER

The Black Series, amongst Other Hunting Scenes As described earlier, the two oldest hunting scene styles A1 (Ochre series) and A2 (Black series) are only found in Cueva de las Manos, while the later styles A3, A4, and A5 are more extensive and thus are found in different sites in the vicinity of Río Pinturas, the Central High Plateau, and the Andes mountainouslacustrine area. That is, the same rock art styles appear at sites of different ecological and altitudinal areas within a maximum radius of 140 linear kilometers toward the southwest and 60 kilometers toward the southeast of Cueva de las Manos. In what we call the “Black series,” we speak of a chromatic series that uses black color for its representations in 153 cases (98.4 percent) and dark purple color in 4 cases (1.6 percent), considering the complete set of representations included in the entire scene. The producers implemented this visual strategy in order to distinguish this scene from the earlier ochre scene, on which the Black series is superimposed. The use of this black color (which is not repeated in the styles of later scenes) allows the observer to associate the different biomorphic or simple geometric representations that compose the scene, and to re-create the “thematic or anecdotal link” that exists between these components (sensu Gradin, Aschero, and Aguerre 1976). This thematic link between two or more representations, resolved with a unifying color monochromy, is combined with the repetition of guanaco figures (the most painted hunted animal) and human figures in a visual presentation that allows the observer to follow the scene through its full development. This unity of meaning—this visual association through monochromy—is what we originally called a “tonal group” (Gradin et al. 1976; Gradin 1983), but here we call it a “chromatic group.” In Cueva de las Manos, we find different chromatic groups that allow us to understand each of the actions contained in the scenes of collective hunting, including the Black series. It is worth pointing out that the two concepts—chromatic group and chromatic series—define visual resources that the producers originally used to distinguish their work from previous ones. In Río Pinturas and neighboring areas of research, these visual resources continued to be used in later styles. From what we know about rock art west of meridian 70°, distinguishing the superimposed scenes through different colors is what led to the creation of a particular mode of execution used on a regional scale. It is precisely the Black series that introduces this mode of execution at the expense of the previous Ochre series. In the analysis of rock art and portable art, these representations are regarded as discrete units, in the sense of an execution independent of the others found on the same space and rock-support, with no ties or articulation between them. A circle or a vertical line traced digitally with paint, an open or a closed contour that may be painted or carved, the negative stencil of a hand, and the complete depiction of a geometric figure as a schematic labyrinth (Aschero 1973) are examples of representations on surfaces of rock-supports that may be non-transportable (rock art) or transportable (portable art). The A1 through A5 classification of styles paid particular attention to monochrome palettes used in the execution of each scene, as well as in the superimpositions between the series and the differences observed between the representation of guanaco types. Initially, in 1976, when the relation between the hunting scenes of Cueva de las Manos and other sites of the region was unknown, they were all grouped under the name “stylistic group A” (Gradin et al. 1976). This implied that there existed between them a common theme (hunting scenes) and a series of variations in the ways they were represented. In 2010, when stylistic group A was classified into five different styles (Aschero 2010), we noted the following sets of characteristics that distinguish each style: • Different chromatic series: use of ochre (predominant) and dark red for style A1 (Ochre series); use of black (predominant) and dark purple for style A2 (Black series); use of red for style A3 (Red series); use of purplish red and dark red for style A4 (Purplish/Dark Red series); and white (predominant) and yellow-ochre for style A5 (White/Yellow series). • Superimpositions amongst the different chromatic series: A2 over A1; A3 over A2 and A1; A4 over A3 and A2; A5 over all the others. • Different canons and patterns in the representations of guanaco and, in a few cases, in the representations of the huemul (South Andean deer). • Different lengths of the total scenes, with the A1 and A2 styles being the largest (340 m long), and the A4 style being the smallest (3–1.5 m long). • Variation in the system of hunting weapons included in each scene.

THE BLACK SERIES IN THE HUNTING SCENES OF CUEVA DE LAS MANOS, RÍO PINTURAS, PATAGONIA, ARGENTINA

315

In 2015, excavations carried out in collaboration with Aguerre (Aguerre et al. 2016) allowed us to broaden the chronological base for both the Black series and the later Red series. We will comment on these datings when dealing specifically with the chronology of the Black series (Aschero et al. 2019b).

Principles guiding the analysis: The total scene The total scene is divided into different compositions connected by the anecdotal links that characterize the scene or part of it. Distributed throughout the four sites of the Cueva de las Manos CRAS (see Table 20.1), these compositions take into account the visual field of a standing observer and make use of the topographic irregularities of the rock-support—crevices, fissures, protrusions—to establish its limits. Each composition has, in the rock-support, a certain virtual framing of its representations: a certain length and width inside which these representations are distributed, and its height from the floor where the composition begins. Each composition can include one or more acts of collective hunting, like driving the animals, intercepting them, and so on. Between the different compositions are figures of guanacos that function as visual links designed to attract the observer’s attention. Due to their morphological characteristics (the constructive canon and pattern of each figure), these guanacos are clearly distinguishable from the guanaco types in the scenes of other chromatic series. By “canon” we mean the distinct way in which the body of the guanaco is portrayed, and its constructive order, which was inferred from some unfinished representations that show the preforms of the guanaco bodies. In the case of the Black series guanacos, the body is an oval or elliptical shape to which the remaining features are added in a certain order: first the neck and head, then the legs and tail. We call “morphological traits” the morphological characteristics of these features (the proportions of the body compared to the real model, the thin legs on which hoofs are depicted, the dynamic attitude of the figure, the definition of the head, etc.); we call “patterns” the specific combination of these traits. These different patterns are what allow us to distinguish the guanacos of the Black series as a different type compared to the guanacos of other series (Aschero 2010, 2018). By “total scene” we refer to the maximum display reached by the Black series in the different sites of the ignimbrite wall on which Cueva de las Manos is painted. It consists of an area 340 meters long with two painted areas at the beginning and the end of the expanse (sites II and IV), with empty spaces and intermediate visual links (site III). Three of the five rock art sites are distributed in this area, with the different styles recognized in Cueva de las Manos (Gradin et al. 1976, 1979). But we could divide this total scene into two main scenes with different production phases: one in Paredón de las Escenas and the Cave, the other in the Alero de las Escenas. Taking these two particular scenes into account, we can observe certain characteristics that lead us to believe that the scene of the Paredón de las Escenas is older than the one developed in the Alero de las Escenas, namely: (1) There are numerous representations of faded black guanacos. (2) Several of these guanacos were restored and repainted. (3) “Impact marks” are present at different heights, which is very common in the Paredón de las Escenas but very rare in the Alero de las Escenas, where it appears only in the bottom section (Table 20.3). These “impact marks,” which have a sub-circular shape and are 3 to 5 centimeters in diameter, are produced by throwing a rock or stone ball wrapped in fur or leather soaked in black paint. The particularity of

TABLE 20.1. Distribution of Black series in Cueva de las Manos sites. Table created by Carlos Aschero. Site IV Alero de las Escenas

Site III Alero del Derrumbe

Site II b-c Cave vestibule and interior

Site II a Paredón de las Escenas

(exterior rock-shelter) 40m

40m

15m

30m

Black series 2nd phase

Black series Visual Links

Black series 1st phase

Black series 1st phase

316

CARLOS A. ASCHERO AND PATRICIA SCHNEIER

these “impact marks” is that they never occur over other representations of the scene—an indication that they were produced before the other representations. They may be interpreted as an operation intended to “kill” or “nullify” the power of the rock-support and its negative effects on the scene taking place. Such impact marks are common in the first Ochre series (A1 style) and also later, in the red chromatic series (A3 style). Thus we find a temporal difference between the Cave Paredón de las Escenas and the Alero de las Escenas. The first could be earlier than the second, and the two can be distinguished as two different phases of rock art production. This does not contradict the concept of total scene, since the second scene adds a final action to the previous one, which is a guanaco slaughter or a human fence surrounding the prey. Each scene is connected to the others by certain visual links that replicate the representations contained in the scenes and allow the observer to follow the development of the total scene throughout the promontory. The visual links of site III (Alero del Derrumbe) are outside the collapsed section of the rock shelter. The collapse took place after 3400 BP, according to the 2015 excavation of the site. There are few depictions of guanacos of the same canon and pattern as those of the Paredón de las Escenas and Alero de las Escenas, but three purple guanacos are added in a horizontal position, one over the next in a vertical line, and a faded and unfinished black guanaco. The names, extensions, chromatic series, and phases of each site are shown in Table 20.1. The order of succession of the sites runs from the right to the left of the observer, and the total scene seems to follow this same order, a fact that we can deduce from the direction imposed on most of the guanaco representations, and also from the thematic and temporal link between the actions depicted in each series. This “reading order” of the scene requires the observer to move along the walls of the sites in order to understand the different thematic links. The first chromatic group, executed in black, is found in the Paredón de las Escenas. Thereafter, in the vestibule and interior of the cave itself, the figure of the pregnant female is replicated in a larger size than the rest of the representations of the Paredón de las Escenas. The only pregnant female inside the cave has a rope or leash tied to her neck (Table 20.2: fig. 5). Nine pregnant guanacos are represented in the vestibule: in this way, the representation of pregnant females operates as a visual link because they are depicted in the interior of the cave and its vestibule, as well as at the beginning of the scene in the far right of the Paredón de las Escenas. They open and end this part of the total scene.

Weapons and actions depicted The three weapons represented in the total scene are (1) the dart/spear-thrower, (2) the “bola perdida” (lost ball), and (3) the “lazo-bola” (lasso-ball). The dart/spear thrower is widely known, but there is little archeological evidence of it in Patagonia. It allows the hunter to strike from a distance. The usage of this weapon is deduced by tracing the possible trajectory of the darts, as well as the presence of some “cephalic garments,” which coincides with data recorded in the chronicles of seventeenth-century Patagonian explorers who said hunters carried the distal end of shafts with projectile points, on leather cords attached to their heads. Asymmetry of the arms or elongation of one of them (as if using a spear-thrower) in depictions of hunters driving or intercepting animals (Aschero 2018) also represents this weapon. The “bola perdida” (“lost ball”) consists of a round, heavy stone tied to a leather rope about 1 to 1.5 meters long with a transverse handle on the free end of the rope (Figure 20.3). It was flung to strike and take down the animal, which would later be killed by the same hunter or a different one. Use of this weapon requires a hiding place that allows the hunter to be as close as possible to the prey. The “lazo-bola” {“lasso-ball”) is a lighter round stone tied to a leather rope 5–6.5 meters long. It was flung, without releasing the rope, at the legs of an animal in order to entangle it and bring it down to be killed later by another hunter. This weapon is more common during the latest A5 style. As a hand-held weapon, the lasso-ball is carried in a roll on one shoulder, according to different representations in later styles (Mena 1991). Experiments have proven that when flung to the ground next to a running pack of guanacos, the

THE BLACK SERIES IN THE HUNTING SCENES OF CUEVA DE LAS MANOS, RÍO PINTURAS, PATAGONIA, ARGENTINA

317

TABLE 20.2. Representations of Biomorphology of Animals and Humans. Table created by Patricia Schneier.

318

CARLOS A. ASCHERO AND PATRICIA SCHNEIER

FIGURE 20.3. Interception of prey with lasso-balls with handle (bottom left). Paredón de las Escenas, Cueva de las Manos. Copyright Carlos Aschero.

THE BLACK SERIES IN THE HUNTING SCENES OF CUEVA DE LAS MANOS, RÍO PINTURAS, PATAGONIA, ARGENTINA

319

bola bounces and is easily tangled around the legs of the animal.1 A second hunter is required to approach and kill the animal, so again, the hunter needs a hiding place as close as possible to the running pack of guanacos being driven by the hunters (Figure 20.3 and Table 20.2: fig. 13). The actions portrayed in the total scene are as follows: from the standpoint of the observer, the animals are driven right-to-left. At the Paredón de las Escenas there are three actions of human figures driving animals— there is one human per drive—and a total of 12 adult guanacos and one newborn. In the Alero de las Escenas there is only one drive with 1 human, 16 adult guanacos, and 3 newborns. A total of 4 humans, out of a total of 21 humans in the series, are represented in these droves. There are 28 adult guanacos in total and 4 newborns (Table 20.3; Table 20.2: figs. 14 and 12). A representation of the interception of prey is on a flat rock-support without the use of micro-topographical features. It might be called “interception in the open field” and involve the use of a dart-thrower and darts. Seen only in the Paredón de las Escenas, the scene includes 1 human, 5 guanacos, and 7 darts (Table 20.3). TABLE 20.3. Frequency of motif types within the site. Table created by Patricia Schneier.

320

CARLOS A. ASCHERO AND PATRICIA SCHNEIER

The darts are shown as curved parallel lines, and it is through that curvature that the parabolic movement of the thrown darts is captured. In this series (as in other, later ones) the carrying of a dart-thrower is represented by an elongation of one of the hunter’s arms. This elongation is always applied to the right arm. Since no micro-topographical features are used (though they will be in the next action), we are led to believe that this interception occurs on flat topographical terrain (pampas). Interception achieved by using micro-topographical resources may involve use of natural traps and “lasso-balls with handles” as hunting weapons. Unlike the previous action described, this strategy uses cracks or crevices in the rock as a sort of virtual eco-landscape of ravines and gorges, which were used as traps to drive the animals and ambush them using “bolas perdidas” with transverse handles. There are three representations of these weapons. The top end with the “bola” is depicted, joined by a vertical line (the cord) and a perpendicular line (the “handle”) at the bottom end.2 They are displayed as vertical parallels and are not associated with any human figure. Further on we will discuss the characteristics of the weapons mentioned (Figure 20.3). A guanaco slaughter, or “human fence” surrounding caught prey, occurs in the Alero de las Escenas, at the left end of the development of the total scene, which also features a guanaco inside a fence of human figures that is the last of the Black series, and the largest in size. Nine human figures are surrounding the guanaco. There is a use of hierarchic perspective, with the lower central figure being larger than other figures and superimposed on the animal’s body. Also, there is an aerial perspective in the distribution of the human figures up and under the central guanaco. The human figures with “cephalic garments” look markedly different from those without them (center top, above the animal’s back). Beneath the slaughter to the right are two simplified human figures followed by dotted lines that continue the direction of the figures’ movement, resembling their footprints, as occurred with the humans featured in the previous action (Table 20.2: fig. 11 and Figure 20.4). Regarding the composition of the Paredón de las Escenas scene, let us point out that it starts with an animal drive from right to left at the top of the panel, surrounded by impact marks (Figure 20.4). It is composed of a single human figure with an elongated right arm, indicating the carrying of a dart-thrower. At the bottom right of the panel is a large, pregnant guanaco three or four times larger than the guanacos depicted

FIGURE 20.4. Animal drive during a hunt, with the hunter with a long single arm. Paredón de las Escenas, Cueva de las Manos, trat. Enhanced with DStretch. Copyright Carlos Aschero.

THE BLACK SERIES IN THE HUNTING SCENES OF CUEVA DE LAS MANOS, RÍO PINTURAS, PATAGONIA, ARGENTINA

321

around her, and two groups of three hand stencils painted in black (Table 20.2: fig. 16). Toward the center of the panel, the composition shows an action of interception, with 7 thrown darts, 1 hunter, and 5 guanacos. Toward the top left of the panel, a crack in the wall has been used to represent a gorge through which 4 guanacos and 4 hunters move up and down, with their footprints marked by dots (Figure 20.3). To the left of the crack, beneath a ledge in the rock, is a moving guanaco that is being struck by three “bolas perdidas” with handles, coming from the bottom of the panel. The hunters are not represented; it is as if they were hidden behind a second ledge on the wall. At the lower left of the panel are two lines of guanacos to the right and more hand stencils, including an inverted one. The lower part of this scene begins 1.20 meters from the current ground and 1.50 meters from the bedrock where black paint stains appeared in the excavation done at the bottom of the Paredón de las Escenas in 2015, right at the base of the black hand stencils of the scene. The visual framing of this scene is 9 meters long and 2.90 meters high, and its middle and top sections are out of a standing person’s reach. Considering its height, it is unclear how this artwork could have been made (Table 20.2: fig.17). A possibility is the use of hand scales and more than one person helping the painter. If this was the case, it means that these rock art scenes were a collective production. The Alero de las Escenas scene begins with an animal drive composed of 1 hunter, 16 guanacos predominantly heading toward the left, and 1 newborn guanaco (Figure 20.5). The guanacos exhibit dynamic attitudes that are very well depicted and differ from those in the Paredón de las Escenas in the slenderness of their extremities. The framing of the animal drive is 1.35 meters long and 0.70 meters wide. The base of the framing is 2.10 meters from the rocky ground, leaving the scenes out of a standing person’s reach.

FIGURE 20.5. One hunter and sixteen guanacos running toward the left. Alero de las Escenas, Cueva de las Manos. Copyright Carlos Aschero.

322

CARLOS A. ASCHERO AND PATRICIA SCHNEIER

Above this action are three unfinished guanacos that indicate the order in which the representations were constructed, starting with the oval shaped body, to which the neck and legs were added, in that order (Table 20.2: fig. 7).This action and that of the slaughtering or human fence surrounding the prey are separated by a distance of 10 meters, a section in which we observe 45 guanacos and 2 simplified human figures, one of which is smaller than 3 centimeters (Table 20.2: fig. 11); the other has a trail of dots. The slaughtering or human fence occurs at the far left, after the distribution of the Black series. It depicts a guanaco surrounded by 9 human figures whose features will be detailed below. The frame is 1.25 meters long and 0.60 meters high, and as is also the case at the Paredón de las Escenas, it includes several superimpositions of different styles of scenes. At the side and above the human fence are hand stencils and three animal figures not found in the Paredón de las Escenas: 1 huemul with its offspring (Table 20.2: fig. 8), 1 canine (fox or dog; Table 20.2: fig. 9), and an undifferentiated animal (Table 20.2: fig. 10). The total frame of the Alero de las Escenas would be 12 meters long by 2.20 meters high, but the base of the frame varies in height from the floor, from 2.10 meters to 1.50 m at the left end.

The representations of guanacos and humans As we have seen, the preforms of guanacos found in the top section of the Alero de las Escenas show that their construction started with an oval shape, to which the neck and extremities were added. This is what one of us has called “canon A,” which characterizes most of the guanaco representations in styles A1, A2, A3, and A4, with room for variations in the original shape of the body and the thickness of the legs. These variations are the different patterns that characterized the different types of guanacos, as we have said (Aschero 2010, 2018). But a different method was used to build the shape of pregnant guanacos, starting with an “almond” shape to which are added the neck and then the front and hind legs as a “block” or joined legs. This preform is observed in the vestibule of the cave, above the group of pregnant females. It allowed us to identify another constructive canon, called “canon B,” which originated in the Black series, continued in the Red series (A3 style), was afterward reactivated at the time of the A5 style or White series, and remained in use at Río Pinturas after the disappearance of the scene styles before and after 6890 BP and the H1 Hudson eruption (Aschero et al. 2019b). Regarding the human figures, there are actually three patterns of representation related to a canon in which the height of the body is 4 or 5 times the size of the head, with one short arm and another elongated one likely to represent the carrying of a dart-thrower, as the same characteristic is replicated in styles A3 and A4 in what are clearly hurling postures. The legs are short in proportion to the body. The figures are depicted facing front or three-quarters sideways, with the legs facing front and the body facing sideways (Table 20.2: fig. 11 and Figure 20.4). Among the 9 human figures surrounding the prey there were cephalic garments, which should be interpreted as a way to carry the distal end of shafts, on leather cords attached to the heads. Also present are small figures without cephalic garments and with only a circle for the head (which could indicate a different sex), and another figure, still smaller than the other, that might represent a child. The central figure, superimposed on the guanaco, deserves particular attention. The body presents a possible dress or cape (like those used by the Aonikenk and also the Selknam of Tierra del Fuego). It also shows a lasso-ball on the right side of the shoulder.

On the chronology of the Black series The results obtained in the excavations undertaken in 2015 at the foot of the Paredón de las Escenas in Cueva de las Manos allow us to present stratigraphic evidence that could be used to date rock art. The new radiocarbon dates must be added to the ones previously published, that is, 9300+90 (CSIC-385) and 9320+90 BP (CSIC-1238) for the lower and medium layer 6, in which we found a lithic artifact impregnated with goethite ochre mixed with gypsum hemihydrate. The same mix is used to paint the ochre chromatic series (style A1), the earliest of the hunting scenes (Gradin et al. 1976; Iñiguez and Gradin 1977; Wainright et al. 2002; Aguerre 1977).

THE BLACK SERIES IN THE HUNTING SCENES OF CUEVA DE LAS MANOS, RÍO PINTURAS, PATAGONIA, ARGENTINA

323

The evidence from the 2015 excavation included lithic and bone artifacts similar to those registered in the lower layer 6 from the excavations undertaken in the 1970s. This is added to a series of three fire pits, around which large amounts of mineral pigments of ochre and red colors (exclusively) were recovered, with a low presence of bone fragment specimens and stone debris. The repetition of these findings in all three fire pits (layer 3(1), layer 3(2), and layer 4) suggests that their function was to thermally convert the ochre mineral pigments into red ones that could be used in the production of the red chromatic series (styles A3 and A4), in the same Paredón (Aschero et al. 2019a). This situation turns out to be similar to the one documented at the site Cerro Casa de Piedra 5 (Perito Moreno National Park), some 130 geodesic kilometers away from Cueva de las Manos. At that site, X-ray diffraction tests allowed us to establish that this thermal alteration of the ochre goethite (a hydrated iron oxide) to obtain intense red tones (hematite) or violet tones (maghemite) from the yellow-brownish goethite used in the rock paintings at Cueva de las Manos (Wainright et al. 2002) was intentionally produced in the fire pits of the site (Aschero 1983–1985; Rial and Barbosa 1983–1985). Various experiments that we later had opportunities to undertake in our campfires confirmed this sequence for Cerro Casa de Piedra 5 and Cueva de las Manos. Beneath the layer 4 fire pit and above the basal rock at the foot of the Paredón, three sets of black paint stains were documented in the bedrock, two of which coincided with two of the hand stencils that accompanied the black scenes of the A2 style superimposed on the ochre scenes of the first style. This layer 4 fire pit produced a date of 8890+/−120 BP (LP-3388), and the immediately superior level 3(2) produced a date 9030+/−100 BP (LP-3394). Both 14C dates are equivalent when the corresponding standard deviations are considered. This implies that the black chromatic series of the A2 style scenes was executed before these chronologies. Therefore, the two earliest styles of the hunting scenes (A1 and A2) might have developed in the period between ca. 9400 and 8800 BP, and the Black series specifically between ca. 9000 and 8800 BP. A third firepit (firepit No. 1), at the foot of the Paredón de las Escenas and also linked to the thermal alteration of ochre pigments into red pigments, was dated at 7710+80 BP (LP-3378). This result could be used to date the beginning of the white chromatic A5 style (after ca. 7700 BP), which ended with the abandonment of the Río Pinturas area following the H1 eruption of the Hudson volcano around 6890 uncalibrated years BP (Prieto et al. 2013).

Contribution of the Black series to later styles Taking into account the temporal depth displayed by the rock art in Río Pinturas, the Black series has the following characteristics, many of which contributed to the subsequent styles: 1. Like its prior A1 style (ochre), the A2 style of the Black series is only found in Cueva de las Manos and has not been detected in any other site of the area. This implies that, in the time period previously noted, this site operated as a node of attraction for production of rock art. 2. The Black series can be visualized as a total scene that associates two individual scenes through guanaco types as visual links. 3. The Black series differs from the previous Ochre series, as it introduces contrasting colors (black and purple) in order to distinguish the different representations. This method was adopted by later styles. 4. It introduces the theme of pregnant females and the “B” canon in the construction of these images. 5. It shows three hunting weapons, of which the lasso-ball and the “bolas perdidas” (with handle) were not present in the previous Ochre series. 6. It shows a more objective rendering of the guanaco’s body and movements. It also maintains, in relation to the Ochre series, the use of flat, monochrome, strictly two-dimensional silhouettes. 7. It introduces the hierarchic perspective and the aerial perspective through replication of figures (in the representation of the human fence). 8. It introduces the use of the micro-topography of the rock-support, taken as “virtual topographic features” that reproduce the real topographies used for the hunt. 9. It introduces the concept of a “trail” marked by the direction followed by the human figures.

324

CARLOS A. ASCHERO AND PATRICIA SCHNEIER

This is reprised by later styles and particularly by the carved figures of the “stylistic group D,” also called the “footprint style” (Gradin et al. 1979).

Towards an interpretation of the Black series More generally, we want to present some remarks within the interpretative field of the Black series. From what is currently known, we can see that the emphasis of the Black series in the pregnant guanaco female figures (together with Cerro Casa de Piedra sites 5 and 7) is unique in northwestern Santa Cruz rock art. These representations, and the presence of the huemul figures in the scenes of collective hunting of guanacos (which did not occur in real time), suggest the possibility that the paintings had mythographic and didactic meanings explaining the origin of the guanaco as well as the different actions, numbers of hunters, weapons, and topographic situations used in these collective hunting activities. They also serve as reminders that everything depends on the reproductive role of female guanacos, tied to a primeval mythographic narrative that linked the female guanacos to the cave. This is why the pregnant female guanacos of the vestibule and the cave interior deserve special attention. They are situated in the most protected area of the cave, and the one found in the interior is painted with a rope around her neck. They are “unique” representations, in which the relationship between their pregnancy, the protection of the cave, and their movement toward the exterior reminds us of Escalada’s 1949 account of a myth told by the Aonikenk people (otherwise called Metcharnuwe people, a subgroup of these Southern Tehuelches) that the native peoples found in this area of Patagonia during the first contact with Europeans (Escalada 1949: 127 and 129). According to the myth, the benevolent spirit Seecho created guanacos inside a cave. In the words of Escalada: “Aóni-kenk sources conserve the ancient belief of the cave . . . from which all people and animals originated.” The first being to come out of the cave was a woman “endowed with supernatural powers. She was the owner and master of the Earth and all the animals and plants that inhabited it.” (Escalada 1949: 127 and 129). Another aspect to highlight is that the superimposition of many representations of the Black series, without the obliteration of the previous Ochre series, implies a “memory of space” that, in the collective memory of these hunter-gatherers, encourages the reactivation of certain visual themes from previous generations, particularly hunting as a social practice. This function of reproducing social memory was not fulfilled by the representation of individual animals or small scenes, but by a series of interconnected scenes making up a narrative. The spatial display of this series, requiring the observer’s movement through the set of actions portrayed, reinforced this visual narrative. In this sense, sites like these should be considered true “places of memory” where this action of “returning” operates positively in the reproduction of social memory. To sum up, the practices of collective hunting were important events in the social life of these ancient hunter-gatherers. This suggests seasonal congregation of families as well as bands, which are inferred by the number of hunters included in the rock art scenes, and also indicates that for those who produced the Black series, the visual testimony of these practices operated as a kind of ritual originated by their ancestors (the authors of the Ochre series). As such, this ritual gave meaning to the replication of the images, while also retaining a temporal identity through the color of each superimposed image. If by ritual we understand different types of action taken from ordinary life and used with a communicative function and a specific meaning by way of a certain symbolic framework, while also marking the position of each individual in their society (J. P. Mitchell in Barnard 2002), then these rituals joined to the painting of hunting scenes in rock art denoted not only individuals but the social groups that returned to this place and produced them. This production and associated rituals positioned them as a social collective distinct from other groups, while also helping to achieve social integration of its members. Cueva de las Manos and Cerro Casa de Piedra were places that, as Ingold (2000: 192) put it, were imbued with the personal experiences of the individuals that occupied that space. A place owes its character to the experiences it affords to those who spend time there—to the sights, sounds, and indeed smells that constitute its specific ambience. And these, in turn, depend on the kind of activities in which inhabitants engage. It is in the relational context of people’s engagement with the world, in the business of dwelling, that each place

THE BLACK SERIES IN THE HUNTING SCENES OF CUEVA DE LAS MANOS, RÍO PINTURAS, PATAGONIA, ARGENTINA

325

draws its unique significance. With the addition of these visual images, inaugurated by the ancestors, these also could become places of memory, places that are “. . . units of meaning, of material or ideal order, transformed by human will or the work of time into a symbolic element of a certain community” (Candeau 2002: 112). If we accept that these collective hunting scenes of styles A1–A5 were painted over each other in Cueva de las Manos between ca. 9400 and 6900 uncalibrated years BP, that is, throughout a period of 2500 years, then it is highly probable that these visual images helped sustain the social memory of these huntergatherers. In Candeau’s (2002: 104) words: “. . . all human societies may disappear, but the representations we make of our tradition, our customs and our rituals, seem to indicate that we believe the opposite is true.” Carlos Aschero obtained his degree in Anthropology from the Universidad de Buenos Aires (UBA), Argentina. He is currently a Principal Researcher in the National Council of Scientific and Technical Research (CONICET-ISES, CCT-Tucumán, Argentina). As a professor at the Universidad de Buenos Aires and the Universidad Nacional de Tucumán from 1983 to 2016, he directed eleven doctoral theses at both universities, as well as different research projects (PIP-CONICET and PICT-FONCyT) in the Argentinian Puna and Patagonia. He has published more than seventy papers and articles in books on hunter-gatherers’ rock art, lithic technology and subsistence, and mobility patterns in the mentioned areas. Since 2016, he has directed two research programs in Antofagasta de la Sierra and Coranzuli (Puna) and northwestern Santa Cruz (Southern Patagonia). Patricia Schneier gained her Anthropology degree at the Universidad de Buenos Aires (UBA), Argentina. She has done research with the Toba communities in Chaco (Argentina), and since 2015 she has collaborated in research on rock art in Río Pinturas, Santa Cruz, Patagonia. She is currently the coordinator of the ethnographic exhibit at the Ethnographic and Historic Museum at Perito Moreno, Santa Cruz, Argentina, which includes the hunter-gatherer societies of Patagonia.

Notes 1. We owe this information to the expertise of Ricardo Vázquez of the Museo C. J. Gradin in Perito Moreno, Santa Cruz. 2. We use the Spanish term “bola” to refer the round, abraded and polished lithic artifact that was used alone (“bola perdida” or “lost ball”) or in a compound of three bolas forming the “boleadora pampa.” Both were used by historic Tehuelches in north and south Patagonia.

References Aguerre, A. M. 1977. “A propósito de un nuevo fechado radiocarbónico para Cueva de las Manos, Alto Río Pinturas, Provincia de Santa Cruz.” Relaciones de la Sociedad Argentina de Antropología 11: 129–142. Aguerre, A. M., C. A. Aschero, M. V. Isasmendi, and V. Ucedo. 2016. Nuevas excavaciones en Cueva de las Manos: el Alero del Derrumbe y el Paredón de las Escenas. Informe a la Dirección de Patrimonio de la Pcia. de Santa Cruz. Aschero, C. A. 1973. “Los motivos laberínticos en América.”Relaciones de la Sociedad Argentina de Antropología, NS.7, 259–75. Aschero, C. A. 1983–1985.“Pinturas rupestres en asentamientos cazadores-recolectores. Dos casos de análisis aplicando difracción de rayos-X.” Cuadernos del Instituto Nacional de Antropología 10: 291–306. Aschero, C. A. 1996. “¿A dónde van esos Guanacos?” In Arqueología. Solo Patagonia. Ponencias de las II Jornadas de Arqueología de la Patagonia, ed. J. Gómez Otero, 152–53. Puerto Madryn, Argentina: CENPAT-CONICET. Aschero, C.A. 2010. Las escenas de Caza en Cueva de las Manos. Una Perspectiva Regional (Santa Cruz, Argentina). Simposio Internacional de Arte Rupestre IFRAO 2010. Tarascon-sur Ariegè, Francia. CD Pre-Actes. Aschero, C. A. 2018. “The Hunting Scenes in Cueva de las Manos: Styles, Content and Chronology (Río Pinturas, Santa Cruz-Argentinian Patagonia).” In Interpreting Rock Art, ed. A. Troncoso and E. Armstrong. Cap. 9. London: Routledge. Aschero, C. A., and M. T. Civalero. (2016). Informe final Proyecto PICT—FONCyT 1927. Agencia Nacional de Promoción Científica y Tecnológica. Ministerio de Ciencia y Técnica, Buenos Aires. Aschero, C. A., R. A. Goñi, M. T. Civalero, R. L. Molinari, S. Espinosa, A. G. Guráieb, and C. T. Bellelli. 2005. “Holocenic Park: arqueología del Parque Nacional Perito Moreno (PNPM).” Anales de la Administración de Parques Nacionales 17: 71–119.

326

CARLOS A. ASCHERO AND PATRICIA SCHNEIER

Aschero, C. A., M. V. Isasmendi, R. V. Ucedo, and A. M. Aguerre. 2019a. “Aportes a la cronología y contexto de las escenas de caza tempranas en Cueva de las Manos (9400–7700 años AP): Alto Río Pinturas, Santa Cruz.” In Arqueología de la Patagonia: El pasado en las arenas, ed. J. Gómez Otero, A. Svoboda, and A. Banegas, 41–51. Puerto Madryn, Argentina: CONICET-IDEAUS. Aschero, C. A., Bozzuto, D., Civalero, T., Nigris, M. D., L. Fernández, N., N. Maveroff, & M. Sacchi. (2019b). “Se nos viene la noche. El volcán Hudson y su influencia en el noroeste de Santa Cruz: integrando perspectivas arqueológicas.” In Arqueología de la Patagonia: El pasado en las arenas, ed. J. Gómez Otero, A. Svoboda, and A. Banegas, 239–49. Puerto Madryn, Argentina: CONICET-IDEAUS. Candeau, J. 2002. Antropología de la Memoria. Buenos Aires: Ediciones Nueva Visión. Escalada, F. A. 1949. El complejo Tehuelche. Estudios de Etnografía Patagónica. Buenos Aires: Imprenta y casa editora CONI. Espinosa, S., and R.A. Goñi. 1998. “¡Viven! Una fuente de obsidiana en la provincia de Santa Cruz.” In Soplando en el viento, ed. INAPL Univ Comahue, 177–88. Actas de las III Jornadas de Arqueología Patagónica. Buenos Aires: Neuquén. Gradin, C. J. 1983. “Algunos aspectos del análisis de las manifestaciones rupestres.” Revista del Museo Provincial Neuquén 1: 120–37. Gradin, C. J., C. A. Aschero, and A. M. Aguerre. 1976. “Investigaciones arqueológicas en la Cueva de las Manos, Alto Río Pinturas, Santa Cruz.” Relaciones de la Sociedad Argentina de Antropología 10: 201–51. ———, C. J., C. A., Aschero, and A. M. Aguerre. 1979. “Arqueología del Área del Río Pinturas.” Relaciones de la Sociedad Argentina de Antropología 13: 183–227. Ingold, T. 2000. “The Temporality of Landscape.” In The Perception of Environment: Essays in Livelihood, Dwelling and Skill, ed. T. Ingold, pp. 189–208. London: Routledge. Iñiguez, A. M., and C. J. Gradin. 1977. “Análisis por difracción de rayos X de muestras de pinturas de Cueva de las Manos, Alto Río Pinturas.” Relaciones de la Sociedad Argentina de Antropología 11: 121–28. Mena, L. F. 1991. “Cazadores recolectores en el Área Patagónica y Tierras Bajas Aledañas (Holoceno Medio y Tardío).” Revista de Arqueología Americana 4: 131–63. Mitchell, J. P. 2002. “Ritual.” In Encyclopedia of Social and Cultural Anthropology, ed. A. Barnard and J. Spencer, pp. 490–93. London: Routledge. Prieto, A., C. Stern, and J. E. Estevez. 2013. “The Peopling of the Fuego-Patagonia.” Quaternary International 317: 3–13. Rial, G., and C. E. Barbosa. 1983–1985. “Análisis mineralógico por difracción de rayos-X de muestras de pinturas del Cerro Casa de Piedra, sitio CCP5 (Prov. de Santa Cruz, Argentina).” Cuadernos del Instituto Nacional de Antropología 10: 307–12. Wainwright, I. N. M., K. L. Helwig, D. Rolandi, C. A. Aschero, C. J. Gradin, M. M. Podestá, and C. Bellelli. 2002. “Identification of Pigments from Rock Paintings Sites in Argentina.” In L’art avant l’histoire. La conservation de l’art préhistorique, ed. 15–24. Paris: 10es Journées d’études de la Section française de l’Institut International de conservation. SFIIC-F77420-Champs-sur-Marne.

==== •

•

•

EPILOGUE Is There More to Scenes Than Meets the Eye? Iain Davidson and April Nowell

Introduction Readers of a book with a long introduction and twenty chapters may think it indulgent for its editors to write one more piece, but we feel that there is room for a few more questions, so please bear with us.

What Have We Just Seen? The chapters in this book are a series of case studies that could be primarily aimed at those interested in a particular time in the past of regions where that rock art exists. We hope that readers welcome that function, and we are happy that all the chapters can be seen as making a contribution to the archaeology of their regions. Yet by putting the studies together in a book with a theme, in this case scenes in rock and cave art, there is a presumption that the chapters together might contribute to an understanding that goes beyond the particular time periods and regions to generate new thought about the theme more generally. What are scenes? And why are they important? The answers to the first question are many and various, and this book holds many examples throughout. There are many definitions of “scene” and, by extension, of what does not quite measure up as a scene. Can an individual figure be a scene? Do images of tracks constitute scenes by any definition, and is there a difference between a group of footprints and a group of handprints? Is it legitimate to identify a scene when it can only be identified after much analysis with all of the modern analytical tools? Is it a scene if individual images are arranged in a composition without any other apparent interaction among them? What about the contexts in which scenes were enacted—why are there so few examples of ground or landscape in any example of rock or cave art? How do we approach the knowledge that in some contexts it is possible to analyze the topography of the rock on which the images were produced as representing the environment of action of the images represented? How can we take account of the knowledge that some recorded societies include art in their stories, which incorporate individual sites and their contents into narratives about interconnectedness across the landscape—that is, the whole landscape is the scene? Should we routinely distinguish between different types: scenes that are snapshots of the moments in the life of the society of the artists; scenes that tell a visual story that moves through (brief) time; scenes that capture a process within the lives of the society? How were choices made about which aspects of life to represent in scenes? Why do some rock art scenes represent the daily activities of the society and others represent activities that are less central, such as hunting scenes produced by agricultural societies? Why are there so few scenes of hunting in hunting societies? How do we deal with the fact that while some sites contain panels that are interpretable as scenes, others, probably made within the same society, do not represent scenes? How do we interpret the fact that scenes were represented in some contexts of image production but not in others? To what extent are these questions, posed about art made in societies other than our own, constrained by our own cultural baggage, whatever that is? Are there innate cognitive propensities to make and see representations of scenes, or are all such perceptions culturally embodied? All of these questions are addressed in one way or another within the chapters of this book.

328

IAIN DAVIDSON AND APRIL NOWELL

As a consequence of the thematic nature of the book, the chapters devote little attention to rock art that does not appear to be a representation of a scene—and the earliest claims for art, whether in Asia, Africa, or Europe, fall into that category. It emerges that, even at the time of their production, many of the panels of rock art required the producer to tell stories about them. That would especially have been the case with any non-representational marks. All of this emphasizes that the rock or cave art cannot be understood without moving beyond the image(s) to consider who made them, who first saw them in the context defined by the producer(s) and heard the stories, and of course what was represented, that is, the things through which stories were told. Story and interpretation were intrinsic to the earliest production of meaningful marks. But images made on rock surfaces are often visible long after they were produced. Eventually the connection to the original producers, and then to the audience informed by the producers, was broken, and the stories were lost. Today, the persistence of the images allows untrained or uninformed people to observe them within their own, probably different, cultural context. Take for example the exceptional circumstance (within which this book is by and large situated) of the context of “Western-trained archaeologists,” all untrained or uninformed about the culture of the producers. If the marks were zigzags or cross-hatching or nonrepresentational paint marks on a cave wall, new stories constructed by untrained eyes would likely be dissimilar to the originals. Scenes, however defined, offered an opportunity to make some sense of those images, because even the untrained could witness the agency of the creatures represented in a panel—not the images, but the animate objects they represent. Meanwhile, some rock or cave art consists only of apparently unrepresentational marks. Given the data presented in some of the chapters, it might be argued that even this art could have contributed to the work that is done by a scene. What do we mean by that? It is possible, through detailed and painstaking analysis of such marks, to work back to the relationship between the producers and the rock on which they produced the marks, and through that to the relationships among different producers. Such analysis therefore demands that we ask whether it is helpful to divide the world of images into those that look like scenes in our set of cultural dispositions, and those that do not. One answer to that question would be that most people, whether close to the time of production or to the present, do not undertake the sort of painstaking analysis that reveals such cryptic relationships. Without the analysis, the relationships would have remained as opaque to the uninformed then as they would to modern observers now. The exception might be the reading of representations of tracks, which appears to be possible when the images are interpreted by people who live by hunting animals similar to those known by the artist (Pastoors et al. 2015). The “scene-ness” of images, therefore, is informed by important qualities that allow a narrative to exist after, and without, the presence of a narrator. We would argue that it created a sense of historical narratives about lost times that could not have been achieved in any other way. At the same time, there obviously is uncertainty about the timing of the production of separate images that now constitute a scene. This time transgressive production can lead to the possibility of false “scene-ness.” In some examples, the argument that related images were produced as a single composition is strong, despite certainty that they could not have been produced in an instant. In other examples, it may be that what we now see as a composition or a deliberate scene was the outcome of production of images that accumulated over time. Such an observation is complicated by the fact that the person who produced the last image on such a panel could see all of the earlier ones and thus might have been able to see a scene that was not visible to their predecessors. Finally, since we embarked on this project to look at the idea of scenes in rock and cave art, the new discovery of the earliest representational art in the world, at Leang Bulu’ Sipong 4 in Sulawesi, shows scenes of human/animal composite figures (therioanthropes) interacting with identifiable animals older than 43.9 thousand years old (Figure 21.1) (Aubert et al. 2019). Such scenes were not present in the almost contemporary representational art at Lubang Jeriji Saléh in Borneo (Aubert et al. 2018) or in the slightly later art at Leang Bulu Bettue in Sulawesi (Brumm et al. 2017), but a line below a babirusa image at Leang Timpuseng older than 35 thousand years old may be the makings of a scene (Aubert et al. 2014). On present evidence, it appears that scenes existed very early in the production of representational art. This suggests that the apparent rarity of scenes from Paleolithic art on cave walls in Europe is a further indication that that art was about something else—as follows from the presence of scenes on moveable objects in Paleolithic sites. Moreover, the apparent representation of therianthropes in the Leang Bulu’ Sipong 4 scene suggests that, as things stand at present, the imaginary life of the painters was rich and complex prior to the production of the imagery

EPILOGUE

329

to go with it—as we can also infer from the early production of therianthropic figurines in Europe. Between Blombos and Leang Bulu’ Sipong 4, there was a great distance of travel and an equally great symbolic journey.

What Does It Mean? The answer to the second question “Why are scenes important?” depends on the scale of the inquiry. If it is only a question about why scenes are important to the particular culture history of the time and region, then we will fall back on the usual tropes about art reflecting the lives of the societies of the artists (with the anomalies we highlight here) and the ability of observers from our cultural context to understand what was being represented. That is important enough. But as we hinted in the previous section, issues arising from the persistence of images on rock surfaces signal that the importance penetrates beyond that limited time and cultural frame. All rock art shares certain key characteristics. It was created on a rock surface. The rock surface is fixed to a location in space and in a particular geology and ecology. It has persisted since the time when it was created. It has a particular subject matter that may or may not be related to the ecology of the time it was produced. It was made by people who existed in social relations with other people at the time. It has been and can be seen by people unrelated to the producers culturally or socially. The images themselves can be seen and appreciated as objects—and thus can be confused with art—but given their persistence and location, they also invite analysis of relationships. Those relationships might be between the variation in images and the location and ecology, between images and other similar or different images in space, and/or between images and the society of the people who made them. The functions of things that can be called art vary from place to place, society to society, and time to time. They have ranged from ritual and other symbolic functions—including initiation of children—to ephemeral graffiti produced while enhancing narratives in small groups, or indeed, produced by a bored and lonely shepherd watching for an imaginary wolf. It is possible to name a wide range of visual materials from our own societies (wherever they may be) that might be called art, from Brassempouy to Brancusi, doodles to Dalí, fabric styles to photography, graphic designs to graffiti, kitchen wares to kitsch, Caravaggio to spray can murals, Velazquez to Emily Kame Kngwarreye, wrapping paper to wallpaper. Asking why the visual design on them is important or what it means might lead to madness—but it might not, when it comes to scenes in rock and cave art. A case can be made that the production and elaboration of scenes in rock art, devoid of ground or landscape, made possible the awareness that visual materials can tell a story. Paradoxically, awareness that the creation of visual materials could tell a story may have made possible the variety of visual enhancement that we celebrate in all those examples from Brassempouy to wallpaper. Marks and sculpture were material parts of the narrative but could not narrate anything themselves. People could use them as mnemonics, and in societies with oral traditions, such material objects were undoubtedly contributors to the relatively faithful repetition of the narrative. But such narratives inevitably degrade over time. Scenes reduced the extent of degradation and made it possible for narratives to endure and be reinforced beyond the short term. They make it possible to answer the question, what does it mean? at one level if not any other. We think the chapters in this volume force us to think of the variety of ways in which that ubiquitous question can be answered.

Where to Next? Scenes emerged early in the history of rock art—later in cave art. They were part of the variation in ways in which people communicated with each other in ritual, symbolic, and banal contexts. Like all rock and cave art, the imagery persisted to be seen beyond the context of the original communication. Some, like the Paleolithic cave art of Western Europe—still exceptional in the cryptic nature of its representation of scenes—disappeared and made no contribution to the history of art. Other examples of images made on the permanent and public medium of rocks in the landscape were produced all over the world at many times up to and including the twentieth century. Rock art persisted in particular locations and inevitably influenced how people interacted with the ecology and with each other. All of these generalizations are important aspects of the behavior associated with rock and cave art. By isolating scenes as a theme to consider at different times and different places, we teased out more aspects of

330

IAIN DAVIDSON AND APRIL NOWELL

FIGURE 21.1. Hunting scene (detail) from Leang Bulu’ Sipong 4 (Photo credit Ratno Sardi) with permission from Maxime Aubert and Adam Brumm.

EPILOGUE

331

the variation in this type of art. By looking at the different characteristics that follow from making images on rocks, we are able to investigate the ecological and social context of image-making through ritual and symbolic aspects of the art, and particularly through the ideas interpretable from scenes. The next phase of analysis needs to separate out these themes and contexts and consider them together.

Acknowledgments We thank all our authors for their forbearance during the long process of getting these chapters to press. We hope that they think the wait has been worth it, as we do. We thank John Robb for his insightful comments during our initial discussions in Vancouver and further interactions in Paris in 2018. We both thank our spouses, Helen Arthurson and Jon Miller, who put up with our prolonged preoccupations and frustrations. Iain Davidson has worked on Spanish Upper Palaeolithic Art, archaeology and ethnography of Northwest Queensland, Australian rock art, colonization of Sahul, language origins, and cognitive evolution. He has taught in England, Australia, the United States, and Chile, and has worked with ten different Aboriginal groups and undertaken major archaeological consultancy for many of Australia’s leading industries. His publications include 4 books, 7 edited volumes, more than 60 chapters and 80 articles in journals. He is a Fellow of the Australian Academy of Humanities and was awarded the Rhys Jones Medal of the Australian Archaeological Association 2010. He held the Chair of Australian Studies at Harvard University from 2008–09. April Nowell is a Paleolithic archaeologist and Professor and Chair of Anthropology at the University of Victoria. She directs an international team of researchers in the study of Lower and Middle Paleolithic sites in Jordan and collaborates with colleagues on the study of Pleistocene rock art in Australia and France. She is known for her publications on cognitive archaeology, Paleolithic art, the archaeology of children, and the relationship between science, pop culture, and the media. She is the author of Growing Up in the Ice Age: Fossil and Archaeological Evidence of the Lived Lives of Plio-Pleistocene Children (2021). Her co-edited volumes include Stone Tools and the Evolution of Human Cognition (2010) with Prof. Iain Davidson, Archaeology of Night: Life After Dark in the Ancient World (2018) with Dr. Nancy Gonlin, and the forthcoming Culturing the Body: Prehistoric Perspectives on Identity and Sociality with Dr. Benjamin Collins.

References Aubert, M., A. Brumm, M. Ramli, T. Sutikna, E. W. Saptomo, B. Hakim, M. J. Morwood, G. D. van den Bergh, L. Kinsley, and A. Dosseto. 2014. “Pleistocene Cave Art from Sulawesi, Indonesia.” Nature 514(7521): 223–27. Aubert, M., R. Lebe, A. A. Oktaviana, M. Tang, B. Burhan, A. Jusdi, B. Hakim, Zhao, Jian-xin, I M. Geria, and P. H. Sulistyarto. 2019. “Earliest Hunting Scene in Prehistoric Art.” Nature 576(7787): 442–45. Aubert, M., P. Setiawan, A. A. Oktaviana, A. Brumm, P. H. Sulistyarto, E. W. Saptomo, B. Istiawan, T. A. Ma’rifat, V. N. Wahyuono, F. T. Atmoko, Zhao, Jian-xin., J. Huntley, P. S. C. Taçon, D. L. Howard, and H. E. A. Brand. 2018. Palaeolithic Cave Art in Borneo. Nature 564(7735): 254–57. Brumm, A., M. C. Langley, M. W. Moore, B. Hakim, M. Ramli, I. Sumantri, B. Burhan, A. M. Saiful, L. Siagian, L. Suryatman, R. Sardi, R. Jusdi, A. Abdullah, A. P. Mubarak, H. Hasrianti, A. A. Oktaviana, S. Adhityatama, G. D. van den Bergh, M. Aubert, Zhao, Jian-xin, J. Huntley, Li Bo, R. G. Roberts, E. W. Saptomo, Y. Perston, and R. Grün. 2017. “Early Human Symbolic Behavior in the Late Pleistocene of Wallacea.” Proceedings of the National Academy of Sciences 114(16): 4105–10. Pastoors, A., T. Lenssen-Erz, T. Ciqae, U. Kxunta, T. Thao, R. Bégouën, M. Biesele, and J. Clottes. (2015). “Tracking in Caves: Experience Based Reading of Pleistocene Human Footprints in French Caves. Cambridge Archaeological Journal 25(3): 551–64.

==== •

•

•

INDEX

Page numbers in italics indicate figures or tables Abauntz (Spain), 17, 18, 208, 209, 210, 214 Aboriginal peoples, 7–8: Ancestral Beings, 152, 155–58, 163–64, 167–68, 171; Dreaming/Alcheringa/Songlines/ buwarraja, 10, 12, 22, 152, 155, 156, 163–64, 167–68, 171, 172, 174; as hunter-gatherers, 83, 152; languages/communication practices of, 11, 58, 61; rock art of, 53, 69–70, 240; sand drawings, 22, 155–56, 157; Traditional Owners, 70, 71, 156. See also Arrernte peoples; Kimberley; Walbiri people; Wardaman people; Yanyuwa people Abric del Barranc de la Palla. See de la Palla Abrigo de la Sarga (Spain), 242: II, 236 action, as defining a scene, 3, 16, 32, 55–56, 62, 69–70, 78, 90–92, 109–10, 113, 117, 167, 224, 242; of animals, 224; characteristics, 181; climbing scenes, 236; diagonal vectors/planes, 185, 243; engagement, 181, 186; in European/Eurasian Upper Paleolithic art, 184–85; motion markers, 115, 117–18; non-narrative “snapshots,” 5, 12; “peak action,” 35, 36; perception of, 35, 58; visual markers for a scene, 34–37, 40, 70, 110, 111, 118, 181–83, 184, 185, 186, 187, 242; “specialized action,” 96, 97, 100, 104. See also combat; dancing; death; event; hunting; marching; narrative; pedal narratives; ploughs/ploughing Afghanistan, rock art of, 107 agency, in rock/cave art, 12, 78: actor, 111–12; agent/ agens, 21, 22–23, 96, 97, 98, 100, 110, 185; of animals, 25–26, 188–89; of art itself, 22–23; characteristics, 181; of humans, 26, 28, 188–89; nonhuman, 168, 173; observer, 22, 23, 26, 28, 180, 181; participant, 111–12; patient/patiens, 22, 23, 26, 96, 98, 100; producer, 22; of ritualized humans, 188–89; scenes and, 21–28, 183–84 agriculture, and rock art, 8, 223, 245, 246, 277, 278, 280, 283–86: scenes of, 125, 289. See also farmers Algonquin people (Canada), rock art of, 295, 297, 304: Algonquin animism and, 296–97; Algonquin Woodland period, 296; Canadian Shield and Great Lakes pictographs, 295–96; landscape and, 296–97, 298; shamanism and, 297–99. See also Eastern Algonquin people; Mi’kmaw people/Mi’kmaq; Peskotomuhkati people; Wolastoqiyik people Alpine rock art, 260, 270 Altamira (Spain), 45, 56: finger flutings, 194; publication of images from, 52–53, 60 Alta (Norway), rock art of, 62, 71

Ancient Schematic Art, 245–46. See also Macroschematic Art; Schematic Art Andes (Chile), rock art of, 68 animals, images of in rock/cave art, 4, 9, 79, 81, 83, 84, 85, 91, 96, 104, 122, 188–89, 207, 210, 270, 279, 328: animation of, 225; bats, 166; bat-eared foxes, 81, 82; beavers, 297; bison, 17, 21, 38, 39, 40, 56, 131, 194, 203, 210, 279, 289; bovids, 138, 139, 210; camels/dromedaries, 134, 136, 138, 140, 141, 142, 145, 146; “canonical forms,” 184; caribou, 302; caterpillars, 155; cats/felines, 156, 215; cetaceans, 303; cheetahs, 113, 114, 115, 116, 118; compositional categories of, 19, 20; dimensionality, in cave art, 190; echidnas, 166; elephants, 122, 125, 127; elk, 18, 19, 20, 279, 286, 287, 289; emus, 51, 61, 166; fish, 129, 166; foreign, 10; gazelles, 138; goats, 223, 228; groups/ herds, 17, 26, 56, 139, 184, 224, 232; kangaroos, 23, 62, 154, 157, 166; insects, 236, 242; leopards, 81, 107, 115, 214; in Levantine rock art, 240, 241, 243; lions, 8, 25, 28, 56, 107, 115, 116, 124, 136, 138, 139, 184, 227; lizards, 156; lynx, 227, 228; macropods, 157, 165, 166, 250; mammoths, 194, 201, 203, 208, 210; mating/pre-coupling scenes, 25, 137–38, 227, 229, 230, 301; moose, 301–2; nursing/suckling scenes, 26, 137–38, 145, 226, 227; oryx, 136, 138, 139, 140, 141, 143, 145; possums, 156; predators, 107, 116, 138, 145, 227, 242; quadrupeds, 77, 79, 81, 83, 84, 261, 285, 286, 288, 289, 301; rarity of individual, 19; reindeer, 5, 210, 227; representations vs. contemporaneous, 11–12; rhinoceroses, 55, 69, 184, 185, 194, 197, 199, 203; saiga antelope, 197, 200, 201, 203; snakes/serpents, 156, 166, 167, 170, 171, 173, 281, 291, 292, 293, 303–5; springbok, 81; static, 23; stingrays, 166; thylacines, 154; tracks of, 51, 156– 57; turtles, 166; ungulates, 79, 81, 208; in Valcamonica rock art, 263; wallabies, 23; wild boar, 129, 130, 223, 231, 232, 242, 244, 245, 248; wildebeest, 81. See also bears; bighorn sheep; birds; cows; deer; dogs/canids; guanacos; horses; ibex; imaginary beings; ostriches; tracks; zoomorphic beings Antelas (Portugal), 208 anthropomorphic figures, images of in rock art, 285, 286, 288: colors of, 168, 314; Eastern Algonquin, 297; European Paleolithic, 182, 223, 242; in Iberia, generally, 245; in Iran, 108, 109, 112, 113, 114, 118; Jebel Qurma, 134, 136–38, 140, 142, 143, 144–45; Levantine, 242; Mirabib, 77, 80, 81, 82; San Rafael Fremont: Northern Zone trapezoidal, 279–80, 281,

INDEX

289–90, 291; in Valcamonica, 266; in Wardaman Country, 165, 166, 167, 169. See also humans Apollo 11 (Namibia), rock art of, 11, 45 Aranda peoples. See Arrernte/Aranda peoples archaeology, and rock/cave art, 8–9, 75, 82, 108, 141, 150, 162–63, 174, 186–88, 214, 229, 249, 252, 293, 313, 322–23: “archaeology of intimacy”/“archaeology of the small,” 195 archers, images of in rock art, 136, 143, 235, 243: in combat, 140, 141, 144, 235, 249; as hunters, 113–14, 115, 118, 125, 138–40, 144, 146, 231, 232, 234, 293 Armenia, rock art of, 9, 108: Geghama Mountains, 107 Arnhem Land (Australia), rock art of, 7, 9, 70, 71, 240: combat depictions in, 249; Lightning Man, 54; Nawarla Gabarnmang, 151; scenes in, 45, 51. See also Western Arnhem Land Arrernte/Aranda peoples, art of, 3, 51; tracks/mbatja, 155–56. See also Aboriginal peoples art: anthropological theories of, 21–23; archaeology of, xvii; cognitive evolution and, 41–46; as communication, 100, 150, 196, 240, 254, 329; definition of, 32; eidetic image theory of art’s origin, 42; ethnography and, 6; function of, 329; real life, relationship with, 34; as having agency, 28; traces as, 46; trance origins of, 42–43; Western tradition of, 52–53, 151, 306. See also cave art; figures; images; portable objects; prehistoric art; rock art art historical studies, of rock/cave art, 7, 162, 296 artifacts, and rock/cave art, 8, 82. See also plaquettes; portable objects association, of images, 16, 17, 111, 112–13, 151, 158, 229; cognitive evolution and, 43–44; as compositional rule, 38, 40, 45; as intentional, 42; as “tight” organization, 43, 44; types of, 69, 224. See also iconicity; juxtaposition Australia, rock art of, 4, 91: dating, 150–51; finger flutings, 194; Koonalda Cave, 196; Northern, 45, 70; scenes in, 150–51; South, 60; study of, 150–51; superimposition of motifs, 150–51; traces, 57. See also Arnhem Land; Arrernte/Aranda peoples; Central Australia; Kimberley; Walbiri people; Wardaman Country; Wardaman people axes, 122, 124, 130, 169, 244, 246, 260 Azéma, M., 56 Azerbaijan, rock art of, 108: Gobustan, 107 Azraq (Jordan), 7, 134 Basedow, H., 51 bears, rock art images of, 62: den, 71; tracks, 71 Bhimbetka (India), 45, 122, 124, 125, 126, 129, 130, 131 Bicknell, C., 270–71 bighorn sheep, rock art images of, 20, 286, 289, 292: ewes, 26, 279; lambs, 26, 279; rams, 26, 279 “Biographic” pictorial tradition (United States), 45, 62–63. See also “Late Archaic” pictorial tradition birds, images in rock art, 61, 116, 155, 156, 166, 167: gulirrida/peewees, 163, 164; owls, 292; water birds, 266, 292

333

Black Desert (Jordan): hamad, 135; harra, 134, 135, 136; rock art of, 7, 134, 147. See also Jebel Qurma boats, images of in rock art, 216, 250, 303–5 boomerangs, images of in rock art, 152, 153, 154, 166, 170, 171, 173. See also portable objects; weapons Brandberg-Dâureb Mountain/massif (Namibia), rock art of, 68, 82, 90–91: Brandberg-Dâureb Database, 90, 95, 96, 103, 104n1; data mining analysis of, 95, 96–104; interactions in, 96–97; “White Lady of Brandberg,” 295 Breuil, Abbé H., 7, 194, 295 Bronze Age, 8, 245 Bruniquel (France), 242 camp, image of, 17–18 Canadian Maritime provinces, rock art of, 6, 10 Cantos de la Visera (Spain), 245 Cartailhac, É., 53 Çatalhöyük (Turkey), 208, 214 causality, scene as depiction of, 32, 34, 36, 40, 45 cave art: changing meaning of, over time, 6; colors used, 202, 212; content of, 4–5; creators of, 195; development of, 10, 44–46; diachronic creation of, 19; dimensionality in, 190; European/Eurasian Upper Paleolithic, 179, 186–90; interpretation of, xvii, 18, 22, 194, 203–4, 327; navigational lines, 201, 202, 204; non-scenic to scenic transition, 329; themes, in general, 7; transhistorical significance of, 329. See also cave paintings; parietal art; scenes; and specific sites cave paintings, 20: Altamira Cave, 52–53; Chauvet Cave, 8, 11, 69, 184, 185; Cueva de las Manos, 4, 19–21; in Europe, 108; Lascaux Cave, 70 Central Asia, rock art of, 107–8 Central Australia, rock art of, 6, 11, 51, 154, 235: Ancestral Beings, 155–57; Atalpi, 156, 157; background, 152; cupules, 156; Dreaming/Alcheringa and, 155; Emily Gap/Nthwerrke, 155; engravings, 152, 154, 156; ethnography and, 156, 157, 158; interpretation of, 155, 156, 157–58; Kathleen Domes, 157–58; landscape, interaction with, 157–58; motifs, 154–55; paintings, 153–54, 156–57; scenes in, 155–56, 158; tracks, 155–58; Traditional Owners, 156. See also Arrernte/Aranda peoples Cerro Bayo (Argentina), 313–14 Cerro Casa de Piedra (Argentina), 312–14, 323, 324 Chalcolithic period: in Europe, 260, 261; in India, 125, 126, 128 Chaturbujnath Nala (India), 125, 131; B3, 130; B4, 130; B5, 130; B6, 129, 131; B7, 125, 130; B10, 129; B11, 130, 131; B17, 124, 126; B19, 130 Chauvet (France), 8, 11, 17, 19, 20, 21, 25, 28, 45, 56, 69, 184–85, 187, 194, 203, 204, 230, 185 Chhattisgarh (India), rock art of, 122: ethnography and, 131–32; Kohbar, 130; Lekhamada, 126; Likha Pakhna 2, 129, 130; Singar Pathar, 128; Sita Lekhni 4, 128 children, as creators of cave art, 196–201 Cingle de la Mola Remigia (Spain), 231, 232, 235 circles, in rock art, 292: in Alta art, 71; in Central Australian art, 154–55, 156; in Jebel Qurma art, 144; in Levantine rock art, 18; in Valcamonica rock art, 269

334

INDEX

Circus Gorge (Namibia), 91 Civil (Spain), 235, 242, 245 climbing, scenes of, 236, 248 clothing/headgear, images of in rock art, 124, 129, 130, 131, 152, 153, 154, 166, 167, 242, 244, 322; sandal, 266 Cõa Valley (Portugal), 230 cognitive evolution: art’s role in, 41–46; behavioral shift as evidence of, 43–44; development of communication and, 61; event/scene depiction and, 41–42, 46–47; genetic vs. epigenetic, 42, 44, 47; neuroplasticity, 44 colonialism, 10, 162, 300, 302–3 combat, scenes of, 5, 107, 114, 122, 250, 271, 273, 286, 290: animal vs. animal, 203; human vs. animal, 116, 117, 124; in Jebel Qurma art, 134, 137, 140, 141, 144, 145; in Levantine art, 230, 235–36, 240, 244, 251; raiding, 140, 141, 145, 146. See also archers; warriors communication, 51, 95, 97, 99, 147, 298: art as, 100, 150, 196, 240, 254, 329; evolutionary communications theory, 64; narrativity and, 92; with spiritual beings, 297, 299. See also multimodal systems of communication compositions, as types of rock/cave art, 67–68: as intentional/not intentional, 68, 110–11; perception of, 37–38. See also galleries; panels; scenes contact art, 250–1, 253, 295, 296, 300, 306 contemporary/descendant communities, rock art and, 3, 6, 7–8, 16, 17, 21, 70. See also ethnography cosmologies, 293, 297: of Aboriginal peoples, 10, 163, 164, 168, 173, 177n1; of shamanic societies, 76; of southern San people, 77–78 Cottonwood Canyon (United States): Great Hunt Panel, 26–27, 28 Cougnac (France), 23, 188 Cova dels Cavalls (Spain), 26, 230, 232, 234, 243, 245 Covalanas (Spain), 23 cows/aurochs, 70, 71, 125, 184, 210, 229, 244: bulls, 10, 77–78, 81, 85, 125, 130, 223, 242; oxen, 122, 269 Cueva de las Manos (Argentina), Black Series rock art of, 5, 45, 310, 314–15, 320, 322–24: actions, 319–22; archaeology and, 313, 322–23; chromatic groups/ series, 314, 316, 323; chronology, 313, 315, 322–23; as creating social memory, 324–25; eco-landscapes and, 311–13; ethnography and, 324; figures, rendering of, 322; interpretation of, 324–25; scenes in, 314; styles, 313; themes, 310, 314, 323, 324; total scene, 315–16, 319, 323; weapons motifs, 318, 319–20, 321 Cueva de la Araña (Spain), 232, 233, 236, 245 Cueva del Niño (Spain), 19, 20–21 culture: cultural change and rock/cave art, 249–50, 251–52, 254; cultural contact and rock/cave art, 246, 250, 251, 252–53; cultural knowledge/memory and rock/cave art, 3, 7–8, 10, 72, 324–25; informants and, 4; visual perception and, 33–34, 69, 174–75 cupules, 39, 45, 156 dancers, images of in rock art, 124, 125, 126, 128–30

dancing, scenes of, 122, 124, 125, 126, 127–31, 137, 153, 236, 244, 248, 293, 295: ethnography and, 131–32, 153–54 Daraki-Chattan (India), 39, 45 Darwin, C., 53 Davidson, I., 61 death, scenes of, 244, 248 deer, rock/cave art images of, 26, 107, 115, 116, 210, 223, 229, 242, 244, 245, 248, 279, 289: does, 226–27, 228, 229, 234, 279; fawns, 26, 28, 226, 227, 228, 234, 242; in pitoti, 260, 261; finger flutings of, 194; huemul, 314, 317, 324; herds of, 230, 232; hinds, 20, 26, 28; mating, 226–27; nursing/suckling, 26, 226, 227; stags, 19, 20, 107–8, 226–27, 271, 272; whitetailed, 297 de la Palla (Spain), 231, 243 Desolation Canyon (United States), 292 Dobrez, L., 55 Dobrez, P., 46, 69, 70–71, 92, 109–11, 167, 180–81, 182, 184–185, 190, 203 dogs/canids, 279: in Aranda rock art, 51; in Cueva de las Manos, 317, 322; in Iranian rock art, 8, 113, 114, 116, 118: in Jebel Qurma rock art, 136, 138; in Valcamonica rock art, 260–61, 267, 270, 271, 272–73; in Wardaman Country rock art, 166, 169 Dolni Vestonice (Moravia), 207 dots, in rock/cave art, 19, 61, 126, 130, 136, 139, 155, 203, 204, 290, 292, 317, 320, 321, 322, 323–24: in maps, 212, 213, 215; in pitoti, 259, 261 Eastern Algonquin people: petroglyphs of, 296; pictographs of, 296; post-contact art tradition, 296, 300, 306; precontact art tradition, 296, 300, 306; spiritual beliefs, 296–97, 299, 303. See also Algonquin people; Mi’kmaw people/Mi’kmaq; Peskotomuhkati people; Wolastoqiyik people Ekain (Spain), 19, 20, 23, 24 El Castillo (Spain), 194, 196 Eliot, G., 52, 56 epistemologies: and art, xvii; Indigenous vs. Western, 72; and sight, 175 erotic/sexual scenes, in rock art, 137, 167, 293 ethnography, and study of rock art, 6–8, 70, 71, 141, 150, 218, 224, 324: Australia, 153–54, 155–56, 157, 158, 162–63, 167–71; Canada, 296; India, 131–32; Iran, 108–9, 118; of southern San peoples, 76–77, 85–86. See also contemporary/descendant communities ethology, and rock/cave art, 56, 224, 225, 229 European/Eurasian Paleolithic, art of, 5, 52, 179, 223: animal images in, 240; animation in, 225; archaeology and, 229; associations in, 224, 229; cave paintings, 108, 122, 124; Epipaleolithic, 246; “eternal present,” 230; Final Paleolithic, 251; Franco-Cantabrian rock art, 246; images in, xvii–xviii; parietal art, 203, 229; pictography in, 225; repetition in, 229, 235; scenes in, 17, 45, 227, 229–30, 328, 329; vs. Levantine rock art, 242–43, 246, 253. See also European/Eurasian Upper Paleolithic cave/rock art; Paleolithic

INDEX

European/Eurasian Upper Paleolithic, 215: Aurignacian, 122, 186, 187, 188, 215; Gravettian, 186, 187, 188, 210, 212, 223; Magdalenian, 186, 187, 188, 189, 194, 207, 208, 215, 223, 225, 227, 229, 242, 253; preMagdalenian, 188, 189, 229; Perigordian, 215; scenes in art of, 242; Solutrean, 186, 187, 188, 215, 225, 226, 227, 228 European/Eurasian Upper Paleolithic cave/rock art, xvii, 78, 181, 245: action in, 184–85; agency in, 183–84; animal images, 184, 188–89; archaeology and, 186–88; dating, 186–87; engagement, 186; event participants, 180, 186, 188, 189, 190; ground line in, 10; multiplied features/“visual echoes,” 181, 185; parietal art, 186–89; portable objects, 179, 184, 187–89; scenes in, 179, 181, 186–90 events, 12: depiction of, 60, 62, 63; depiction of and cognitive evolution, 41–42, 44, 46–47; event participants, 180, 186, 188, 189, 190; Narrative Events, 34, 180–83, 189, 191n1; Performative Events, 34, 35, 36, 180, 181; real vs. depicted, 34; scene as, 3, 32, 54. See also action; motion/movement; scenes falconry, scenes of, 107, 116, 118 farmers, 7, 240, 251, 252, 280. See also agriculture figures, in scenes, 4: in pitoti, 260; single figure as scene/ non-scene, 36, 45, 55, 70–71, 92, 109, 183–184, 186, 203, 236; spatial ordering of, 5; “weak avoidance,” 260, 282. See also images finger fluters, in Rouffignac: identification of individuals, 196; adult/child and gender determination of, 197 finger flutings, 6, 12, 212: “archaeology of intimacy”/ “archaeology of the small,” 195; communities of practice, 195; dating, 194; definition, 194; as scenes, 194–95, 198, 199, 200–201, 203–4; study/ interpretation of, 194. See also under Rouffignac France, rock/cave art of, 10, 187: finger flutings, 194; of Upper Paleolithic period, 189. See also specific sites friezes, 3, 67, 152, 184, 187, 201, 224 Fremont Complex rock art, 277, 278, 290: San Rafael Fremont: Northern Zone, 279–80; trapezoidal anthropomorphs in, 279–80, 281, 289–90 French Lake (Canada), rock art of, 300: dating, 297; interaction with landscape, 297–99; as microcosm of Algonquin macrocosm, 299; motifs, 297–98; shamanism and, 297–99, 306 Fuente del Trucho (Spain), 208, 212, 213 Fuente Flores (Spain), 249 galleries, 67 Gallese, V., 51, 58 Gargas (France), 210, 213: finger flutings, 194, 196; maps and, 208, 210–212; as palimpsest, 211 Garonne River, 208, 210, 211 Gell, A. E., 22: Art and Agency, 21 gender roles, rock art images and, 5. See also women gesture/gestural language. See multimodal system of communication

335

Gillen, F. J., 3, 51, 151, 155 Gonds people, 129, 131–32 Grand Lake Stream (United States), 301 Green, J., 59 guanaco, 45, 313, 314–16, 317, 319–23: pregnant females, 316, 320, 322, 323, 324 Halo (United States), 19, 20 Hayonim (Israel), 208, 212 hermeneutics, 33: of intention/reception, 38 Holocene, 240, 250, 251, 252: hunter-gatherers, 79, 82, 83, 86, 311–13 Homo habilis, 42, 43 Homo neanderthalensis, 42, 43 honey collecting, images of, 125, 236, 244, 248 Hornos de la Peña (Spain), 196 horses/equids/wild asses, 8, 20, 23, 24, 112, 113, 115, 122, 124, 125, 136, 138, 139, 142, 184, 208, 210, 212, 227, 229, 230, 242, 302; heads of, 230; mares, 230; mating, 230; in pitoti, 260 human anatomy, images of in rock art, 242, 266: arms, 267, 320, 322; breasts, 125, 129, 130, 154, 166, 235, 236; eyes, 129, 130, 155, 166; feet, 58, 170, 267; fingers, 266; hair/hairstyles, 153; hands, 46, 51, 58, 61, 62, 64, 112, 124, 131, 132, 293, 313, 317, 319, 321, 322; legs, 267, 271, 322; mouths, 124, 128, 130; penises/phalluses, 127, 128, 166, 188, 266, 271; toes, 266; vulvas, 166. See also stencils; traces humans, images of in rock art, 4, 5, 91, 97, 98–101, 122, 188–89, 250, 260, 263, 265, 266, 313, 317, 322: animal-headed figures, 242, 248; animation of, 225; children, 97, 98–101, 128, 235, 242, 322; colors of, 128, 130, 131, 153; Dynamic Figures, 154, 251–52; groups of, 235–36; Gwion figures, 152–54; in Levantine rock art, 240, 243–44, 253–54; men, 97, 98–101, 104, 129, 166, 167, 235, 304; ritualized humans, 188–89; women, 97, 98–101, 104, 128, 129, 155, 166, 235–36, 304–5; zero-marked, 97, 99, 103, 166. See also anthropomorphic figures; archers; clothing/headgear; dancers; human anatomy; hunters; imaginary beings; musicians hunter-gatherers, 76, 83, 85, 86, 208, 218, 223, 245, 246, 248–49, 250, 252, 253, 311–13 hunters, images of in rock art, 26–27, 243, 302, 320. See also archers hunting, scenes of, 8, 16, 125, 147, 154, 169, 250, 286, 289, 302, 313–14, 318, 330: animals hunting animals, 138, 139, 145, 227, 230–31; of aurochs, 244; of bears, 62, 71; of bighorn sheep, 26; of cetaceans, 303; of deer, 230, 231, 232, 234, 244, 245, 271, 272–73, 310; of emu, 51; of gazelle, 138; of guanaco, 5, 45, 310, 319–22, 324; of ibex (all species), 112–14, 138, 232, 233, 244, 245, 248; in Iranian rock art, 107, 108, 109, 116, 117, 118; in Jebel Qurma rock art, 137, 138–40, 144–46; of kangaroos, 62, 154; in Levantine rock art, 230–34, 237, 240, 243–44, 247, 248; of mammoth, 208; in Namibian rock art, 55, 104; of oryx, 138, 139, 140, 141, 143, 145; of ostrich, 138, 139–40, 145; of

336

INDEX

wild ass, 138, 139, 145; of wild boar, 231, 232, 244, 245, 248 huts/tents, images of, 17–18, 125, 207, 208 Iberian Paleolithic rock shelters, art of, 223, 225–30, 236– 37. See also Parpalló; Levantine rock shelters ibex, images of in rock art: Iberian, 19–20, 210, 212, 223, 227, 229, 242, 244, 245, 248; Iranian, 107, 112–15, 117, 118; Nubian, 136, 138 Ice Age. See Pleistocene icon, 26, 51, 57, 58, 62, 64 iconicity, 43, 45–46, 150, 154, 297; conventional iconicity, 261; detailing, 181–83; scenes and, 182; symbolic reference, 182–83; of traces, 51, 57. See also images iconic universals, 51: hand icons, 58; foot icons, 58 iconography, 6, 57, 78, 93, 107–8, 253, 286, 292, 293, 299, 300, 303: iconic images, 11, 45, 46, 56, 61, 70, 181–82, 183, 185, 188, 189 ideograms, in rock art, 90 images: Canonicals, 34, 45, 56, 61; exograms, 37, 42, 43, 44, 46; Narrative Events, 34, 180–83, 189, 191n1; Performative Events, 34, 35, 36, 180, 181 images, in rock/cave art, 67, 175: aspective, 260; categories of, 19, 20; composite view, 63, 64; framing of, 37–38, 59, 61, 315; geometric/non-representational, 6, 11, 22, 212, 245, 269, 281, 285, 286, 292; group of, as not scene, 5; groups/clusters, 264, 266, 278, 281–86, 288; horizonal view, 62; as indexes, 21, 26, 181; mimicking vs. outlining, 61; multiplied features/“visual echoes,” 185; patterning, types of, 38; phosphenes, 42, 46; pictography, 225, 227, 237; plan/ map view, 51; profile/side view vs. frontal view, 35–36; profile/side view vs. plan/map view, 59–63; simple vs. complex, 40; single image as scene/not scene, 3, 19, 45, 70–71; stereotypes/stylized images, 23, 25, 26, 27, 28. See also circles; dots; figure; lines; representation; visual perception; zigzags imaginary beings, 97, 98–99, 100, 101: therianthropes, 81, 82, 328, 329 India, rock art of, 7: Bhonrawali Hill, 208; introduction, 122–24; chronology, 124–25; eastern, 107; interpretation of, 133; Kaimur Range, 122; regions, 123; rock shelters, 122, 125, 127–30, 133; traditional dancing and, 131–32. See also Bhimbetka; Chaturbujnath Nala; Chhattisgarh; Daraki-Chattan; Madhya Pradesh; Penganwa; Urden Injasuthi figures, 55 inscription. See multimodal system of communication intentionality: and art, 38, 40, 42–43; and cave/rock art, 110–11, 203. See also association; juxtaposition interaction, in scenes, 4, 12, 40, 69–70, 91, 101, 109, 115, 167, 203, 224, 286: and agency, 28; of animals and animals, 25, 97, 107, 115, 138; of animals and humans, 5, 55, 104, 116, 138, 140, 293; categories of, 96–97, 98–99; cooperation, 104, 138–39, 246, 249; finger flutings and, 6; of humans and humans, 97, 138, 292–93; of humans and plants, 236; network

representation of, 100, 103. See also combat; hunting; pastoralism intercorporeality, 58 Iran, rock art of, 8, 9: Abdozu, 108; Aps-e Golam, 108; Arasbaran, 108–9; Baluchistan, 107, 108; of central Iran, 107, 108, 112, 113, 114, 115, 116; Dasht-eTous, 108; dating of, 107–8; Domab, 112, 113, 114; ethnography and, 108–9, 118; fertility concerns and, 108; Gharghab, 115; hunting, scenes/non-scenes of, 107, 109, 112–14, 116, 117, 118; interpretation of, 108–9, 114, 118; Islamic art and, 116, 118; of Islamic period, 108, 114; Jargih, 114; Jorbat, 108, 115, 116, 117, 118; Kal-Jangal, 116, 117, 118; Koochari, 116; Kouhbodan-e Jor, 108; Kuhdasht, 107; Lakh-Mazar, 116, 118; Marzbanik Valley, 107; Meymeh, 108; Mount Ernan, 113, 114; of Parthian dynasty, 108, 116; petroglyphs, 9, 107–9, 113, 114; pictographs, 107; Pir-Bareh, 108; Qameshlu, 108, 114; Qarqab, 107–8; Saravan, 115; of Sassanid dynasty, 114; scenes in, 112–17; shamanism and, 108–9; Shmsali rock shelter, 112–13; Sorkheh Lizeh, 108; southeast Iran, 107, 114, 115; Teymare, 107, 108, 115–16; themes/motifs, 107; Zagros Mountains, 107 Iron Age, 8, 260, 261 Jebel Qurma (Jordan), rock art of, 7, 135: anthropomorphic figures in, 134, 136–38, 140, 142, 143, 144–45; combat scenes, 140, 144; description of, 134; epigraphic evidence and, 140, 146; historical context of, 145–46; hunting scenes, 137, 138–40, 144, 145; interpretation of, 146–47; nomadism and, 134, 140, 146; pastoral scenes, 134, 138, 145; raiding and, 140, 141, 145, 146; Safaitic inscriptions, 134–37; scenes, in general, 137; themes in, 137, 138, 145, 146– 47; weaponry in, 140–46; weaponry in, distinctions in use, 145; zoomorphic figures in, 134, 136. See also Black Desert Jordan, rock art of, 11. See also Azraq; Black Desert; Jebel Qurma; Rujum Hani; Wadi Sama Ju/’hoansi (!Kung) people, 76, 77, 85 juxtaposition, of images, 4, 16, 17, 20, 25, 47, 79, 289, 298, 305; as accidental/random, 42, 45; agency and, 26; compositions and, 38, 40, 44, 68, 81, 86; intention and, 19; as “loose” organization, 43; scenes and, 67, 111; superimposition and, 167, 300, 304. See also association Kalahari Desert, 76, 85 Kejimkujik Lake (Canada), rock art of, 296: background, 300; cultural change and scenes, 302–3; Fairy Bay, 296, 302, 303, 304, 305; maritime scenes, 303–6; themes, 300 Kesslerloch (Switzerland), 208, 212 Keyser, J., 62 Kiev-Kirillovskaya (Ukraine), 207, 214 Kimberley, the (Australia), rock art of, 5–6, 11, 158; background, 152; dating, 153; ethnography and 150, 153–54; Dynamic Figures, 154, 158; Gwion (Bradshaw)

INDEX

figures, 152–54, 158; Malauwarra, 152–53; motif relationships, 150, 152, 158; motif superimposition, 150–51; pigments, 153; rock shelters, 152; scenes in, 45, 62, 70, 151–54; stylistic sequence, 152; Wandjina figures, 54 Klassen, M., 70 Kuiseb River (Namibia), 75, 82, 83 !Kung people. See Ju/’hoansi people Land of the Lightning Brothers, 163 landscape: agency of, 12; in later Western art, 53, 60 landscape, in rock/cave art: figures in, 54, 63; ground line/ floor line, 10, 53, 243, 273; human relationship to, 52, 157–58, 218, 296–97; interaction with rock/cave art images, 6, 10, 58, 59, 70, 71, 75, 81, 83, 84, 85, 157, 171–73, 175, 190, 304, 321, 323; land surface, 17; limited number of images of, 2, 4–5, 18, 19, 90; as realworld contextualization, 60; representations, 54–55, 219. See also maps; pedal narratives La Rochette (France), 208, 215, 216 Lascaux (France), 5, 16, 17, 45, 56, 70, 71, 184, 209, 215, 216, 230: “bird man,” 188; Lascaux 4 Well (Shaft) scenes, 28, 55–56, 182, 183, 188, 224 Las Chimeneas (Spain), 194, 196 Last Maximum Low Sea Level, 9 “Late Archaic” pictorial tradition (United States), 45. See also “Biographic” pictorial tradition Laugerie-Basse (France), 242: “Woman under the reindeer scene,” 224 leading, scenes of, 137–38, 141, 143, 145: lead ropes, 142 Leang Bulu’ Sipong 4 (Sulawesi), 328–29, 330 Lenssen-Erz, T., 68–69, 70, 109, 167 Leroi-Gourhan, A., 55 Les Trois-Frères (France): Le Tuc d’Audoubert, 227; “Sorcerers,” 124, 188, 189 Levantine rock shelters, art of, 5, 7, 45, 240: action depictions in, 242–43; age/gender differentiation in, of humans/animals, 242; animation in, 225; archaeology and, 249, 252; Centelles/Valltorta-Gasulla phase, 244–45, 248, 249, 251; Cingle phase, 245, 248, 249; Civil phase, 244–45, 248; combat scenes in, 230, 235–36, 237, 244, 245, 249, 251; cultural change and, 245, 249–50, 251, 252; dancing scenes, 236; dating/chronology, 230, 246, 251, 252; diachronic creation of, 68, 230, 237; engravings associated with, 251; vs. European Paleolithic art, 223, 242, 246, 253; farmers and 7, 240, 251, 252; honey-collecting scenes, 236; humans, images of, 243–44; hunter-gatherers and, 240, 246, 248–49, 252; hunting scenes in, 230–34, 237, 240, 243–44, 247, 248–49; vs. Iberian rock shelter art, 223; innovations of, 240, 242–43, 253; interpretation of, 237, 245, 249; Linear phase, 245, 248, 249; vs. Macroschematic Art, 246; Mas d’en Josep phase, 245, 248; Mesolithic peoples and, 249, 251, 253; motion/movement in, 242, 243, 251; narratives in, 12; Neolithic peoples and, 6, 8, 246, 253; overview, 241– 42; pictography in, 225; plant-collecting scenes, 236; post-Paleolithic, 16, 53; post-Pleistocene, 7; repetitive

337

association, 230, 235; scenes in, in general, 236–37; social interaction scenes, 235–36, 244; subjects, 246, 248; themes, 249, 251; time, representation of, 9, 231–32; weaponry, images of, 246; vs. Western Arnhem Land rock art, 250, 251–53. See also Iberian rock shelter art and specific sites lines, in rock/cave art, 46, 69, 286, 292, 320; Australia, 11, 51, 57, 59, 61, 154–57; European/Eurasian Upper Paleolithic, 61, 188; India, 122, 129, 130; Iran, 114– 15, 117; Jebel Qurma, 136, 144; maps and, 207, 210, 211, 212, 214, 215; in pitoti, 259; Rouffignac, 194, 195, 197, 198–201, 202, 203–4. See also finger flutings Lubang Jeriji Saléh (Borneo), 328 Machias Bay (United States), 297, 298, 300, 306 Macroschematic Art, 245–46, 251, 252. See also Ancient Schematic Art; Schematic Art Madhya Pradesh (India), rock art of, 39, 45, 122: activities represented, 122; Belkandhar 1, 129, 130; Belkandhar 2, 130; Bori, 125, 129, 130; Budni, 125, 126, 129, 130; Chhaterpur, 125, 126, 129; Chiklod, 125, 126, 127, 129; Churna, 124, 126, 127, 130; dancing scenes, 127–31; Deur Kothar, 125, 130; Firangi, 125, 130; Hathitol, 125, 130, 131; Jawara, 125, 126, 129, 130; Jhiri, 125, 126, 129; Kathotiya, 125, 126, 128, 129, 130, 131; Kharwai, 125, 126, 130; musical instruments, images of, 125–27; music scenes, 122, 124, 126–28, 131; Pachmarhi, 124, 125, 126, 127, 129, 130; Panna, 125, 126, 129, 130; Pareba Pahari, 125, 130; Putli Karar, 129, 130; Rewa, 125, 126, 129; Sagar, 125, 126, 129, 130; Samardha, 125, 126, 130; subjects, 122; Tapka Pani, 125, 126, 127, 130. See also Bhimbetka; Chaturbujnath Nala; Daraki-Chattan; Penganwa; Urden maps, in prehistoric art, 11, 17, 217, 261; Abauntz, 210; conceptions of, 207; examples of, 207–8; identification/ classifications of, 208–9; interpretation of, 218–19; mobility routes, 212–13; possession, 210–12; special events, 214–16; symbols, use of in, 218 marching, scenes of, 248 Mas d’Azil (France), 230, 242 maternity, scenes of, 244, 248, 249 Mezhirich (Ukraine), 207, 208, 214–15 McGowan Lake (Canada), rock art of, 296: background, 300; cultural change and scenes, 300–302, 303, 306; themes, 300 Mesolithic period: in Europe, 246, 249, 251; in India: 122, 124, 125, 128, 129, 130, 131 Mi’kmaw people/Mi’kmaq, 296, 300; cultural change and, 301, 302–3, 306; hunting practices, 302, 307n4; rock art of, 300–306; spiritual beliefs, 303, 306. See also Algonquin people; Eastern Algonquin people Mirabib (Namibia), rock art of: archaeology and, 75, 82; dating of, 82; description of, 79, 81; diachronic creation of as ritual, 86; driplines and, 75, 81, 83, 85; images depicted, 79, 80, 81; map, 80; “Rain Bull” imagery, 83, 85; rainmaking/water supply and, 83, 84–85; scenes/ non-scenes in, 79, 81, 86; shamanism and, 85, 86

338

INDEX

mobile art. See portable objects Mont Bego (France), 10, 208, 212, 260, 263 motifs, in rock/cave art, 101: defined, 92–93, 100, 163, 164, 165–67; fluted, 6, 38; geometric, 11, 136; interaction of/relationships between, 69–70, 97, 98–100, 103, 150, 151, 158; lineups, 45; number and arrangement of, as composing scene, 54, 67, 70–71, 91, 94; superimposition of, 150–51. See also association; juxtaposition motion/movement, in rock/cave art, 250, 251: animation, types of, 225; depicted motion, 32; direction of, 242, 243; illusory motion, 34–35; images of, 235; implied motion/representational momentum, 58; as not integral to scenes, 71–72; perception of, as integral to scenes, 32, 40, 55, 58, 69–70. See also action multimodal system of communication, 57–58, 61–62: gestural fixing, 61; gestural mimicry, 61; gesture first hypothesis, 58 musical instruments, images of in rock art: cymbals, 127; drums, 124, 125, 127; flutes, 122, 124; harps, 124, 125–26, 128; pipes, 126–27 musical performance, images of in rock art, 7, 123–26, 127, 128, 130 musicians, images of in rock art, 7, 123–26, 127, 128, 130, 281 mythology, and rock art, 6, 10, 12 Namib Desert (Namibia), 75, 82, 83, 85, 86 Namibia, rock art of: Kuiseb River valley, 75; Noukloof Mountains, 55. See also Apollo 11; Brandberg-Dâureb Mountain/massif; Mirabib; Namib Desert narratives, depictions of, 3–4, 12, 28, 64, 272–74: in Black Desert rock art, 134; categorical attributes of, 181–86; composite/bi-angular, 63; decontextualized, 61; embedded, 61, 64; European/Eurasian Upper Paleolithic cave art and, 189–90; explicit, 56; iconicity and, 182; in Levantine rock art, 243; narrativity and scenes, 55, 56, 70–71, 91–92, 109–10, 203; plan/ map-view, 61; side/profile and plan/map view hybrids, 62–63, 64. See also pedal narratives; profile narratives; scenes Natufian culture, 212 naturalism. See realism Neolithic period, 6, 8: in Europe, 230, 237, 240, 245, 246, 249, 252, 260; in India, 124, 125; Later Stone Age in southern Africa, 75, 82, 86; in Turkey, 208, 214 Niaux (France), 21, 23, 38, 39, 40, 45 Nine Mile Canyon (United States), rock art of, 9, 11: abstract images, 286, 288, 290; agriculture and, 278, 280, 283–86, 293; animal image identification, 279; anthropomorphs, 281, 285, 286, 288–90, 291; backpacker images, 290, 291, 292; clusters of images, 278, 279, 281–86, 288; clusters vs. panels, 279; Cottonwood Canyon, 286; CPAA survey methods, 278–79; Currant Canyon, 282; Desbrough Canyon, 291; Devils Canyon, 281, 283, 284, 286, 288, 289, 290, 292; Family Panel, 283, 292–93; First Canyon, 281, 282, 284, 286, 289, 290, 292; First Site, 285,

288; geometric images, 281, 285, 288–89, 290, 292; Great Hunt site, 286; overview, 277; Owl Canyon, 279, 281, 284, 285, 286, 288, 289, 290, 292; Petes Canyon, 281, 284, 285, 286, 288, 289, 290, 292; quadruped images, 285, 286, 288; Rasmussen Cave, 293; Sandhill Crane site, 282; San Rafael Fremont: Northern Zone rock art, 280; scenes in, 278, 281, 286, 288, 290, 293; study of, 293; superimposition in, 282, 282–84, 286; Warrior Ridge, 18, 19, 20, 26, 278–79, 281, 284–86, 287, 288, 289, 290, 292; zoomorphs, 281 Noble, W., 61 ontologies, 224: of Algonquin rock art, 297, 306, 307n1; art as evidencing differences in, 11; cave ontologies, 190; Indigenous vs. Western, 72; as interpretive framework, 295, 307n2; Mi’kmaw, 307n4; Wardaman, 162, 173, 174, 175 ostriches, 82, 136, 138, 139–40, 145 Pakistan, rock art of, 107 Paleolithic, rock/cave art of, 7: communities of practice and, 195; in Eastern Europe, 214; Lower Paleolithic, in India, 122, 124, 125; scenes in, 1, 16–18, 45; study of, 194. See also European/Eurasian Paleolithic; European/ Eurasian Upper Paleolithic Panaramitee Hills (Australia), 60 panels, in rock/cave art, 92, 102: composition of, 18–21; definition, 67 parietal art, 186–89, 203, 229. See also cave art; portable objects Parpalló (Spain), portable art of, 5, 223, 236–37: anthropomorphic representations, 223, 229; associations, 229; Galerías, 229; maps, 207; narrative in, 229; pictography, 227, 237; plaquettes, 11, 20, 23, 24, 26, 28, 29n5, 228; scenes in plaquettes, 16, 17, 225–30; themes, 223, 227, 229, 230, 232, 236–37; zoomorphic representations, 223, 229 pastoralism, 6, 8, 82, 147, 223: scenes of in rock art, 138, 145; sites, in Iberia, 249 patterning, in rock/cave art, 212: compositional, 242–43, 268, 271; data mining and analysis of, 90, 95, 103; duplicates and, 102; geometric, 290, 292; of image features, 194, 315, 322; interior to images/motifs, 139, 144, 145, 215, 217; pattern classes in Valcamonica rock art, 262; scenes and, 68, 134, 147; spatial, 278, 282, 288; visual perception and, 38, 42, 45, 185 Pavlov (Moravia), 207, 208, 212 Pech Merle (France), 23, 194 pedal narratives, 51–52, 61, 64: hybrids with side/profile narratives, 62–63; as landscape embedded, 56–60; pedal sequencing, 58; plan/map view of, 60; vs. scenes, 60; as situated, 59. See also scenes; traces Penganwa (India), 4, 12, 126, 127, 130 Peskotomuhkati people, 296, 297. See also Algonquin people; Eastern Algonquin people petroglyphs/rock engravings, 5, 11, 17, 78, 85, 86, 223, 274: in Australia, 57, 60, 61, 64, 152, 154, 156, 162, 164, 165–66; in the Black Desert, 134, 147; in Canada, 6, 10, 295–306; in India, 122; in Iran, 9, 107–9, 113,

INDEX

114; of Jebel Qurma, 135–47; Native American, 70; in Nine Mile Canyon, 277, 282, 285; in Spain, 210, 251; weathering and, 19. See also pedal narratives; pitoti; traces phenomenology, 33–34, 47, 69, 94, 109, 110 pitoti: adjacent-figure scenes, 271; as aspective, 260; composite figures, 261, 271; contiguous figure scenes, 270–71; conventional iconicity in, 261; dating, 260, 261; definition, 260; figure clusters, 262, 264, 267, 266; figures in, 260; figures in arrays, 262, 266, 267; figures in graphic designs, 262, 268–69; group, defined, 262; human figures, 265, 266, 267; image types, 260; as maps, 261; measures for scenes, 271–72; narrative scenes, 262, 272–74; overview of, 259–60; placement of figures, 261; scenes, in general, 262; scholarship on, 259; “simple figures,” 262; solitary figures, 262–63; spatial order, 262. See also petroglyphs/rock engravings; Valcamonica Plains Indians, 62 plant collecting, scenes of, 236 plants/vegetation, images of in rock art, 4–5, 242: trefoil, 208; walanu/bush potato, 170, 171 plaquettes, 1, 16, 17, 19: at Parpalló, 20, 23, 24, 26, 28, 29n5. See also portable objects Pleistocene period, 185, 189, 190, 194, 240, 250, 251 ploughs/ploughing, images of in rock art, 122, 137, 147, 269, 270–71 portable objects, 17, 179, 242, 328: animal images on, 207; bone engravings, 11, 16, 19, 212, 214, 215; dating of European rock art and 186, 187, 251; European/Eurasian Paleolithic, 229; lissoirs, 184; maps on, 207, 212, 214–15, 216; scenes on, 188, 189, 224, 229; sculptures, 214; stone tools, 186; weapons, 189. See also Parpalló; plaquettes portable objects, images of in rock art, 224–25, 242, 244. See also boomerangs; weapons Predmost (Moravia), 208 prehistoric art, 52, 53, 179, 219, 240, 241, 242, 244, 250, 254: Megalithic, 207; “Occidental,” 207; rock art, 78, 107, 136, 218, 245, 253, 259, 273, 277; scenes in, 224 Pressa Canyon (United States), 40, 41 profile narratives, 59, 60–61, 63, 64: decontextualized, 60–61; as framed, 60, 61, 64. See also narratives proprioception: body awareness, 58; body schema, 58, 64; motor resonance, 58; proprioceptive awareness, 58, 58, 64 Puebloan peoples, 291 realism/naturalism, 10, 52, 56, 300 riding, scenes of: of camels, 140, 142; of elephants, 122, 125, 127; of horses/equids, 8, 112, 113, 115, 122, 124, 125, 138, 139, 140, 141, 142, 144, 145, 267, 271, 272–73, 301; of oxen, 122 rituals, rock/cave art and, 5, 10, 13, 22, 190, 235, 292, 299, 324, 331: scenes of, 27–28, 154, 188. See also under San people rivers, in prehistoric maps, 210–11 Robe and Ledger art, 62

339

rock art: as accidental, 42; changing meaning of, over time, 6; colors used, 91, 124, 128, 130, 131, 153, 156, 314–16; as communication, 240; composition of, 37–38; content of, 4–5; co-speech hypothesis and, 58, 157; development of, 44–46; diachronic creation of, 4, 9–10, 19, 68–69, 79, 81, 86, 177n2, 230, 237, 282–84, 286, 301, 306; duplicates, 102; gesture and creation of, 57, 61; “iconic” vs. “narrative” traditions, 70; Indigenous vs. Western understanding of, 72, 158, 162, 174–76, 295, 296, 306; information transcription and, 95; interpretation of, xvii, 16, 18, 22, 71, 76–78, 86, 87n1, 92, 108, 110, 113, 150, 154–55, 218–19, 224–25, 251, 327; linguistics and analysis of, 90, 92, 93, 95, 103; as mimetic/non-mimetic, 56, 96, 97, 261; multiple temporalities in, 175; non-scenic to scenic transition, 240, 329; origins of, 61; as palimpsests, 79, 81, 82, 92, 211, 218, 301; perception of, 33; pictographs, 57, 107, 277, 295–96; “picture problem,” 259, 273, 274; site, defined, 67; symbols/ideomorphs in, 218, 223; tectiforms, 197, 198, 203, 204; themes, in general, 7; time, representation of, 9, 231–32, 273–74; transhistorical significance of, 329; universalist approach to, 1, 72, 110. See also Ancient Schematic Art; dots; galleries; landscape; lines; Macroschematic Art; panels; petroglyphs; rock paintings; scenes; Schematic Art; shamanism; stencils; themes; traces; and specific sites and examples rock/cave art studies: art making, theories of, 43; in Australia, 150–51, 162–63; of Black Desert rock art, 134; Brandberg-Dâureb Database, 90, 95, 96, 103, 104n1; in Canada, 295–96; cross-regional comparison, 11–12; data mining and, 90, 95–96, 100–104; diachronic comparison, 9–10, 244; empirical approach, 4; European/Eurasian Upper Paleolithic, 187; geography and, 262; history of, 52–54; in India, 122; in Iran, 107– 9; “new” animism, 295, 296–97, 307n1; ontological turn, 295, 297, 307n2; Paleolithic, 194, 195. See also archaeology; ethology; ethnography rock painting, 20: in Australia, 153–54, 156–57, 162, 163, 164–66, 168–69, 170, 171, 173–74; European/ Eurasian Paleolithic, 184, 187, 223; in India, 122, 124–25, 126–27, 128–30, 133; in Iran, 108, 112; in southern Africa, 4, 45, 75–79, 80, 81–86, 90–91; in Turkey, 214; in United States, 19 rock shelters, 5: Apollo 11, 45; in Central Australia, 156– 58; European/Eurasian Upper Paleolithic, 179, 186–89, 208, 215, 216, 245; in India, 122, 125, 127–30, 133; in Iran, 112–13; in the Kimberley, 152; in Wardaman Country, 162. See also Iberian rock shelters; Levantine rock shelters; Mirabib Romanelli (Italy), 207, 208 Rouffignac (France), art of, 6, 12, 23: dating, 196, 205n1; finger flutings in, 194, 195, 196–204; navigational lines, 201, 202, 204; scenes in, 194–95, 198, 199, 200–201, 203–4; tectiforms, 197, 198, 203, 204 Ruskin, J.: Modern Painters, 52–53 Saami people, 71 Sahara Desert, rock art of, 91

340

INDEX

San Antonio (Spain), 215 sand/ground drawings, 22: trace rock art and, 46, 57, 59–60, 61, 63 San people, rock art of, 53, 68: blood used as pigment, 85; rainmaking/“Rain Bull” and, 75, 77, 85; rituals and, 76, 78, 79, 82, 85–86; shamanism and, 76–78, 79. See also Ju/’hoansi (!Kung) people; Mirabib; /Xam people Sanz de Sautuola, M., 52 scenes, in rock/cave art: xvii–xviii, 112–17, 162, 167, 169–70, 259; categorical attributes of, 181–86; categorical markers of, 179, 180; ceremonial, 250; cognitive psychology and, 54, 109; coherence of, 68, 91, 92, 93–94, 96, 97, 102, 181, 186; composition as synonym for, 68; as cosmographic networks, 297, 299; as culturally determined, 7, 18, 328; data mining and, 90; definitions of, 2–4, 16, 32, 34–37, 45, 47, 55–56, 64, 67–71, 78, 90–92, 109–11, 137, 151, 166, 174– 75, 181–86, 194, 203, 242, 281, 310; development of depictions of, 44–47; diachronic creation of, 68; diachronic vs. synchronic creation of, 69, 328; domestic, 250; emergence of, 45, 329, 331; engagement, 181, 186; as evidence of cultural/cognitive shift, 254; finger flutings as, 194–95, 198, 199, 200–201, 203–4; “focus” of, 68–69, 94–96, 100, 101; gestalt and, 36, 91, 93, 94–95, 102; homogeneity in, 92, 100, 102–3; identification of, 90, 111–12, 224–25; interpretations of, 12–13, 174–75. 295, 306; landscape and, 171–73, 175, 331; narrativity and, 93, 110, 203, 240; as “multi-temporal,” 175; in Paleolithic art, 1, 16–18; patterning and, 68; perception of, 35; as perceptual categories, 179–181, 191n1; pictography, 225, 237; in portable objects, 225–30; real vs. depicted, 32–33; realist understanding of, 52, 56; recent discoveries, 17–18; repetition and, 6, 26, 27, 100, 103, 152, 155, 203, 218, 224, 235, 268, 314; scene gist, 54; “setting” of, 68–69; spatial proximity in, 2, 16, 18, 35–36, 181, 186; “subject/theme” of, 68–69; vs. symbolic signs, 35; “theme” of, 94–95; tracings, 212, 227, 228; “vector” theory, 111–12. See also action; agency; events; figures; iconicity; images; motifs; narratives; pedal narratives scenic markers, 34–37: integration of markers, 36; motion 32, 34; optimal viewpoint, 36; vs. performative, 36; profile, 35; size, 36; space, 35; window principle, 36 Serra da Capivara (Brazil), 19, 20, 45 setting, 95: definitions, 68, 94–95, 104n2; as integral to painting, 52–54; interaction of figure with, 70; lack of in scenes, 55, 64; real world as providing, 60; scenes in rock art, 54, 63 Schaafsma, P., 277, 279–80, 281, 289–90 Schematic Art, 241, 245, 246. See also Macroschematic Art; Ancient Schematic Art shamanism, 5, 6: Algonquin rock art and, 297–99; European/Eurasian Upper Paleolithic cave art and, 188; Indian rock art, and 131; Iranian rock art and, 108–9; southern African rock art and, 76–78, 79, 85, 87n1, 295 Sharpe, K., 196 Shepard, E. H., 23, 24, 29n5 speech. See multimodal system of communication

South Africa, rock/cave art, of: Blombos, 11, 45, 329; Cedarberg, 208; Drakensberg/Maluti highlands, 85; Tandjiesberg, 40 southern Africa, 79, 82, 83, 86 southern Africa, rock art of: 5, 45, 75, 82: interpretations of, 76–77, 86; “Rain Bull” imagery, 53, 77–78, 81, 83; scenes/non-scenes in, 78–79; shamanism and, 76–78, 79, 85, 87n1. See also Apollo 11; Mirabib rock shelter; San people Spain, rock/cave art, 45–46: finger flutings, 194. See also specific sites Spencer, W. B., 3, 51, 151, 155 stencils, 46, 51, 64, 165, 166, 250: of emu, 61; human foot, 58; human hand, 5, 45, 46, 156, 157, 208, 212, 213, 313, 321, 322; whole human body, 58. See also traces subjective/objective binary, 33 Sulawesi, cave art of, 7, 46, 328–29 superposition/superimposition, of rock/cave art images, 68, 79, 81, 86, 150–51, 232, 244, 282–84, 286, 300, 314; as memory of space, 324 Swabian Jura sites, 11 symbolic readings, rock art as, 112–15, 117–18, 183 symbols, 6, 26, 110, 113, 125, 181–83, 195, 214, 218, 223, 229 Texas, rock art of, 45 Tito Bustillo (Spain), 23, 208, 215, 217 themes, as scene component, 68–69, 71, 91, 94–95, 103, 137: defined, 104n2 therianthropes. See under imaginary beings Tjungurrayi, Charlie Marshall: Witchetty Grubs (maku) at Kunatjarri, 57 traces, 46, 57, 129, 155–58, 165, 173, 174, 242, 266, 327: hand and foot traces, 62; as plan/map view, 60; representation of, as scenes, 52, 156, 158; trace-derived petroglyphs, 64; trace-form narratives, 60; trace images, 51, 61, 63, 64; trace-view/map-view, 56, 59; trails, 320, 321, 322, 323–24. See also pedal narratives; scenes; stencils tracks, 52, 62: as iconics, 57; tracks and lines, 57; tracks narratives, 60; tracks sequencing, 56, 58, 64 tracking, 56, 58, 64 trance: and art making, 42–43, 76–78, 79, 85; shamanism and, 131 twisted perspective. See composite view Turner, J. M. W., 53 Urden (India), 125, 129: Urden 9, 130; Urden 19, 128, 130; Urden 28, 126, 130; Urden 29, 124; Urden 40, 130 Valcamonica (Italy), rock art of, 5, 8, 208, 212, 259–74: Bagnolo 2, 268; Bedolina, 261; Foppe di Nadro, 266; grandi mani, 266; Naquane rock 1, 266; Naquane rock 50, 266; oranti, 266; Plas Roccia 1: Capitello dei Du Pini, 268; Seradina 12, 262, 263, 265, 267, 270, 272; statue-stele, 268–69. See also pitoti Valltorta-Gassulla (Spain) rock art, 230–31, 244

INDEX

Van Gelder, L., 196 visual perception: of action/motion, 40, 57, 110, 184; attention and, 37; as biological, 33–34; categorical attributes, 180–83; categorical markers, 179–80; cognitive evolution and, 43–44; cultural conventions of, 69; framing, 37–38, 59; informed narrative vs., 174– 75; mirror neurons, 46; Narrative Events, 34, 180–83, 189, 191n1; neurophysiology of, 32, 35, 42, 69, 109, 179–80; patterning, 38, 42, 45; perceptual categories, 179–81, 191n1; Performative Events, 34, 35, 36, 180, 181; rock/cave art, 33; of scenes, 36–37; as universal (not cultural), 33–34, 47, 69 volcanoes, images of in rock art, 17 Walbiri people, 154, 157 Wardaman Country, rock art of, 6, 9–10, 162–76; archaeological research and, 162–63; buwarraja/ Dreaming and, 163–64, 168–71, 172, 174–75; diachronic creation of, 162, 164, 177n2; ethnography, 162–64, 167–71; Lightning Brothers Earthwatch Project, 163; motifs, 163, 164; multiple temporalities of, 175–76; rock shelters and, 162; Yingalarri and, 164, 165, 167. See also Yiwarlarlay Wardaman people, 162, 164, 165: Ancestral Beings, 163– 64, 167, 168, 171, 175, 177n1; buwarraja/Dreaming, 163–64, 168–71, 172, 174–75; landscape and, 163–64, 171–73, 175; Lightning Brothers mythology, 10, 54, 164, 165, 167–73. See also Yiwarlarlay warfare, scenes of in rock art, 235, 237, 244, 248. See also combat warriors, images of in rock art, 124, 125, 127, 244, 248, 266, 267, 270 water supply: in Central Desert of Australia, 152, 156; in Namibia, 75, 82, 83, 85, 86; in Nine Mile Canyon, 285 weapons, images of, 5, 8, 134, 244: arrows, 26, 112, 113, 114, 115, 117, 122, 124, 125, 129, 130, 136, 141, 142, 143, 144, 145, 242, 246; bola perdida, 316, 321; bows, 26, 113–14, 115, 118, 122, 124, 125, 129, 130, 141, 142, 143, 144, 145, 242, 246; daggers, 260, 263, 266, 268, 269; darts, 319–20; guns, 250, 253, 302; halberds, 263, 268, 269; lances, 140, 141, 142, 143, 144, 145–46, 271; lasso-bola, 316, 318, 319–20, 321; quivers, 129, 130, 141, 142, 143, 242; shields, 122, 124, 125, 129, 130, 131, 141, 142, 143,

341

144, 145, 265, 271, 274; spears, 122, 140, 141, 142, 143, 144, 145, 154, 265, 267, 271, 272–74; spear- or dart-throwers, 152, 230, 316, 319–20; swords, 125, 129, 130, 131, 142, 143, 144, 145, 271, 274. See also boomerangs; portable objects Western Arnhem Land, rock art of: contact art, 250–51, 253; Dynamic figures, 251–52; vs. Levantine rock art, 250, 251–53; motion representations in, 250–51; narrative in, 250, 251; overview, 250; style evolution, 251, 252; styles of, 250. See also Arnhem Land White Shaman Mural (United States), 19 Wilton Industry, 82 Wobst, H. M., 6 Wolastoqiyik people, 296, 297. See also Algonquin people; Eastern Algonquin people women: as authors of Safaitic texts, 148n1; as creators of rock/cave art, 148n1; hunting activities and, 154 worldviews, scenes and, 12–13, 296, 302 Writing-on-Stone (Canada), 62, 70 /Xam people, 76, 77, 85 Yanyuwa people, 72, 168. See also Aboriginal peoples Yiwarlarlay (Australia), rock art of, 164: anthropomorphs, 166, 167, 169; archaeology and, 163, 174; background, 165; buwarraja/Dreaming, 168–71, 172, 174; colors, 165, 168; dating, 174; Delamere 3, 165, 166, 167, 170, 171, 173, 174; diachronic creation of, 174; engravings, 165–66, 171, 173, 174; interaction with landscape, 171–73, 175; Lightning Brothers and, 165, 167–75; motifs, 165–67, 169–71, 173–74; paintings, 165–66, 168–69, 171, 173–74; scenes in, 166–67, 169, 171–74; temporality of, 174; tracks, 165, 173; Wardaman vs. Western understanding of, 174–76; Yirwarlarlay 1, 165–67, 169–71, 173, 174; Yirwarlarlay 2, 165, 166; Yirwarlarlay 4, 166; Yirwarlarlay 5, 166; Yirwarlarlay 6, 166; Yirwarlarlay 7, 166; zoomorphs, 166, 169. See also Wardaman Country; Wardaman people zigzags, in rock art, 51, 245, 246, 292 zoomorphic/biomorphic figures, 81, 82, 107, 112, 114, 115, 116, 118, 134, 136, 166, 169, 184, 223, 302, 317