Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia [1 ed.] 9781784913809

Citation preview

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia Mirosław Masojć

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia Mirosław Masojć

Archaeopress Archaeology

Archaeopress Publishing Ltd Gordon House 276 Banbury Road Oxford OX2 7ED

www.archaeopress.com

ISBN 978 1 78491 379 3 ISBN 978 1 78491 380 9 (e-Pdf )

© Archaeopress and Mirosław Masojć 2016 English translation - Bartłomiej Madejski English revision - Rebecca E. Hill Front cover illustration: Gramstrup I, Thy. Backed knife from the cult object (Photograph by M. Jórdeczka) Back cover illustration: Barrows, vicinity of Sundstrup, Hjarbæk Fjord, central Jutland (Photograph by the author) The monograph is the result of research which was made possible thanks to financing from the Polish National Science Centre (project number 2012/05/B/HS3/03829), a grant from the Danish Government and two grants from the Kazimierz Salewicz and Marit Jensen Foundation.

All rights reserved. No part of this book may be reproduced, or transmitted, in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior written permission of the copyright owners.

This book is available direct from Archaeopress or from our website www.archaeopress.com

Contents List of Figures������������������������������������������������������������������������������������������������������������iv List of Plates��������������������������������������������������������������������������������������������������������������vi List of Tables�������������������������������������������������������������������������������������������������������������vii Introduction��������������������������������������������������������������������������������������������������������������ix 1. On the phenomenon of flintworking at the end of the Bronze Age and Beginning of the Iron Age �������������������������������������������������������������������������������������������������������������� 1 2. Temporal and spatial framework, concept of the work, and methods�������������������� 3 3. Cult houses: definition, idea, chronology������������������������������������������������������������� 11 3.1. Cult houses from the stone age������������������������������������������������������������������������������������������������11 3.2. Cult features of the Early Bronze Age�������������������������������������������������������������������������������������15 3.3. Definition�������������������������������������������������������������������������������������������������������������������������������������18 3.3.1. Location������������������������������������������������������������������������������������������������������������������������������18 3.3.2. Construction����������������������������������������������������������������������������������������������������������������������19 3.3.3. Character of the artefacts������������������������������������������������������������������������������������������������21 3.4. Idea�������������������������������������������������������������������������������������������������������������������������������������������������22 3.5. Chronology����������������������������������������������������������������������������������������������������������������������������������24 4. Specification of source information��������������������������������������������������������������������� 27 4.1. Current state of the research�����������������������������������������������������������������������������������������������������27 4.2. Flint assemblages of the Late Bronze Age – contexts of occurrence��������������������������������31 4.2.1. Settlement���������������������������������������������������������������������������������������������������������������������������33 4.2.1.1. General picture, on the basis of Thy�������������������������������������������������������������������33 4.2.1.2. Bjerre, Thy����������������������������������������������������������������������������������������������������������������36 4.2.1.3. Bulbjerg Troldsting, Thy���������������������������������������������������������������������������������������43 4.2.1.4. Fragtrup, Vesthimmerland������������������������������������������������������������������������������������45 4.2.1.5. Højby, Funen�����������������������������������������������������������������������������������������������������������48 4.2.1.6. Skamlebæk, Zealand����������������������������������������������������������������������������������������������49 4.2.1.7. Vinde Helsinge, Zealand���������������������������������������������������������������������������������������51 4.2.1.8. Voldtofte, Funen�����������������������������������������������������������������������������������������������������53 4.2.2. Places of acquisition of the raw material and flintworking workshops�����������������57 4.2.2.1. Knudshoved, Funen�����������������������������������������������������������������������������������������������60 4.2.2.2. Stagstrup, Thy���������������������������������������������������������������������������������������������������������62 4.2.3. Flint deposits����������������������������������������������������������������������������������������������������������������������64 4.2.4. Flints as funerary finds�����������������������������������������������������������������������������������������������������67 The Sandagergåd cult house�������������������������������������������������������������������������������������������������������������71 Sickle from the Stenild peat bog������������������������������������������������������������������������������������������������������73 5. Diversity of cult features from the Late Bronze Age in Jutland������������������������������ 74 5.1. Cult houses�����������������������������������������������������������������������������������������������������������������������������������79 5.2. Semicircular cult features�����������������������������������������������������������������������������������������������������������82 5.3. Torup Høje and successive types of cult features������������������������������������������������������������������83 5.4. Tetrahedral stone installations from southern Jutland��������������������������������������������������������84 i

5.5. Remaining features����������������������������������������������������������������������������������������������������������������������85 6. Cult features with flint assemblages from northern Jutland��������������������������������� 87 6.1. Grydehøj, Thy������������������������������������������������������������������������������������������������������������������������������88 6.1.1. Description of the assemblage����������������������������������������������������������������������������������������94 6.1.2. Raw material����������������������������������������������������������������������������������������������������������������������94 6.1.3. Technology�������������������������������������������������������������������������������������������������������������������������96 6.1.4. Cores������������������������������������������������������������������������������������������������������������������������������������98 6.1.5. Blanks������������������������������������������������������������������������������������������������������������������������������� 100 6.1.6. Tools��������������������������������������������������������������������������������������������������������������������������������� 102 6.1.7. Ad hoc tools�������������������������������������������������������������������������������������������������������������������� 106 6.1.8. Spatial arrangement������������������������������������������������������������������������������������������������������� 107 6.2. Høghs Høj, Thy������������������������������������������������������������������������������������������������������������������������ 114 6.2.1. Description of the assemblage������������������������������������������������������������������������������������� 117 6.2.2. Raw material������������������������������������������������������������������������������������������������������������������� 119 6.2.3. Technology���������������������������������������������������������������������������������������������������������������������� 119 6.2.4. Cores and blanks������������������������������������������������������������������������������������������������������������ 120 6.2.5. Tools��������������������������������������������������������������������������������������������������������������������������������� 123 6.3. Gramstrup I, Thy��������������������������������������������������������������������������������������������������������������������� 128 6.4. Ginnerup, Thy�������������������������������������������������������������������������������������������������������������������������� 131 6.5. Torup Høje, Fjelsø�������������������������������������������������������������������������������������������������������������������� 134 6.5.1. Arrangements of vertically positioned boulders������������������������������������������������������ 136 6.5.2. Ditch features������������������������������������������������������������������������������������������������������������������ 137 6.5.3. Pairs of posthole features���������������������������������������������������������������������������������������������� 138 6.5.4. Mounds of cobbles��������������������������������������������������������������������������������������������������������� 138 6.5.5. Exploration sectors�������������������������������������������������������������������������������������������������������� 141 6.5.5.1. Sector A����������������������������������������������������������������������������������������������������������������� 142 6.5.5.2. Sector B������������������������������������������������������������������������������������������������������������������ 143 6.5.5.3. Sector C����������������������������������������������������������������������������������������������������������������� 143 6.5.5.4. Sector D����������������������������������������������������������������������������������������������������������������� 143 6.5.6. Description of the assemblage������������������������������������������������������������������������������������� 143 6.5.7. Spatial arrangement������������������������������������������������������������������������������������������������������� 144 6.5.8. Raw material������������������������������������������������������������������������������������������������������������������� 148 6.5.9. Technology���������������������������������������������������������������������������������������������������������������������� 149 6.5.10. Cores������������������������������������������������������������������������������������������������������������������������������ 149 6.5.11. Blanks����������������������������������������������������������������������������������������������������������������������������� 151 6.5.12. Tools������������������������������������������������������������������������������������������������������������������������������� 152 7. Issues in flintworking technology and typology���������������������������������������������������159 7.1. Technological aspects�������������������������������������������������������������������������������������������������������������� 159 7.2. Dynamic view of assemblages from ritual zones���������������������������������������������������������������� 160 7.2.1. Raw material (I)������������������������������������������������������������������������������������������������������������� 160 7.2.2. Early phase of reduction (II)��������������������������������������������������������������������������������������� 161 7.2.3. Advanced reduction (III)��������������������������������������������������������������������������������������������� 162 7.2.4. Final exploitation (IV)�������������������������������������������������������������������������������������������������� 164 7.2.5. Other flakes and unidentified specimens (V)���������������������������������������������������������� 164 7.2.6. Tools (VI)������������������������������������������������������������������������������������������������������������������������ 164 ii

7.3. Tool Classification������������������������������������������������������������������������������������������������������������������� 166 7.3.1. Backed knives������������������������������������������������������������������������������������������������������������������ 168 7.3.2. Burins�������������������������������������������������������������������������������������������������������������������������������� 174 8. Examination of wear traces on flint artefacts from ritual zones���������������������������178 8.1. Objects of study and method of examination�������������������������������������������������������������������� 180 8.2. Preservation status of flint artefacts������������������������������������������������������������������������������������� 181 8.3. Characteristics of cultural damage on artefacts����������������������������������������������������������������� 182 8.3.1. Technological traces������������������������������������������������������������������������������������������������������ 182 8.3.2. Diversity of traces of use����������������������������������������������������������������������������������������������� 182 8.3.3. Backed knife from the Gramstrup cult house���������������������������������������������������������� 191 8.4. Interpretation of microtraces of wear – functional specificity of tools from cult houses�� 191 9. Grydehøj cult house. Results of biological analyses���������������������������������������������194 9.1. Anthropological analysis�������������������������������������������������������������������������������������������������������� 195 9.2. Geochemical analyses�������������������������������������������������������������������������������������������������������������� 195 9.2.1. Phosphorus���������������������������������������������������������������������������������������������������������������������� 196 9.2.2. Copper����������������������������������������������������������������������������������������������������������������������������� 198 9.3. Macrobotanical remains��������������������������������������������������������������������������������������������������������� 199 9.4. Palynological analysis�������������������������������������������������������������������������������������������������������������� 200 9.5. Radiocarbon chronometry����������������������������������������������������������������������������������������������������� 202 9.6. Conclusions������������������������������������������������������������������������������������������������������������������������������� 204 10. Ritual flintworking��������������������������������������������������������������������������������������������206 10.1. Remarks on the significance of the construction of cult houses���������������������������������� 206 10.2. Ritual flintworking or flints in cult houses? Sacrum versus profanum������������������������ 211 10.3. Ritual���������������������������������������������������������������������������������������������������������������������������������������� 216 11. Concluding remarks������������������������������������������������������������������������������������������219 Bibliography�����������������������������������������������������������������������������������������������������������222 Plates���������������������������������������������������������������������������������������������������������������������235

iii

List of Figures Figure 2.1. Basic spatial framework of the work ������������������������������������������������������������������������������������������������ 3 Figure 2.2. Southern Scandinavia. ���������������������������������������������������������������������������������������������������������������������� 4 Figure 2.3. North-western Jutland – the Thy region, the island of Mors ������������������������������������������������������������ 5 Figure 2.4. Basic chronological divisions of the late Bronze Age for the area comprised by the study �������������� 7 Figure 3.1. Skateholm, Sweden. Plan and profiles of a supposed cult house. �������������������������������������������������� 12 Figure 3.2. Tustrup, Djursland, eastern Jutland. Cult house from the Funnel Beaker culture.�������������������������� 13 Figure 3.3. Outlines of cult houses from Funnel Beaker culture in northern Jutland. �������������������������������������� 14 Figure 3.4. Reconstruction of a cult house Tustrup, Djursland ������������������������������������������������������������������������� 15 Figure 3.5. Sennels, Thy. Cult structure from the early Bronze Age ����������������������������������������������������������������� 17 Figure 3.6. Orientation of selected cult objects from the late Bronze Age ������������������������������������������������������� 18 Figure 3.7. Hågahagen, Uppland, eastern Sweden. Cult house from the early Bronze Age (II-III period EB).��� 20 Figure 3.8. Hågahagen, Uppland, eastern Sweden. ������������������������������������������������������������������������������������������ 21 Figure 3.9. Solar motifs from bronze razors from the vicinity of Viborg in Jutland.������������������������������������������ 23 Figure 4.1. Tossene Raä, site 446:2-3, Sweden. ������������������������������������������������������������������������������������������������ 28 Figure 4.2. Tossene Raä, site 63:1-2, Sweden. �������������������������������������������������������������������������������������������������� 29 Figure 4.3. An erratic boulder in site Machary, north-western Poland ������������������������������������������������������������ 30 Figure 4.4. Bjerre 7, Thy. Trench 76 during excavations. ���������������������������������������������������������������������������������� 37 Figure 4.5. Bjerre 7, Thy. Eastern part of trench 76. ����������������������������������������������������������������������������������������� 37 Figure 4.6. Bjerre 7, Thy. Core with changing orientation �������������������������������������������������������������������������������� 40 Figure 4.7. Bjerre 7, Thy. Backed knife (type C) ������������������������������������������������������������������������������������������������ 41 Figure 4.8. Bulbjerg Troldsting, Thy. Backed knives in the museum in Copenhagen ���������������������������������������� 44 Figure 4.9. Fragtrup. Plan of the site with residential objects �������������������������������������������������������������������������� 46 Figure 4.10. Fragtrup. House I. Backed knife ���������������������������������������������������������������������������������������������������� 46 Figure 4.11. Fragtrup. House II. Flake with natural backed edge and point. ���������������������������������������������������� 48 Figure 4.12. Højby. Reconstruction of a three-aisled house located in Odense, ���������������������������������������������� 49 Figure 4.13. Skamlebæk, Zealand. Backed knives ��������������������������������������������������������������������������������������������� 50 Figure 4.14. Vinde Helsinge, Zealand. Backed knives ��������������������������������������������������������������������������������������� 52 Figure 4.15. Voldtofte, Funen (Kirkebjerg). ������������������������������������������������������������������������������������������������������ 54 Figure 4.16. Voldtofte, Funen (Kirkebjerg). ������������������������������������������������������������������������������������������������������ 55 Figure 4.17. Knudshoved in Funen. Mass material from a flint workshop. ������������������������������������������������������ 60 Figure 4.18. Wierzbica ‘Zele’ in central Poland. Shaft no. 19. �������������������������������������������������������������������������� 61 Figure 4.19. Knudshoved in Funen. ������������������������������������������������������������������������������������������������������������������ 61 Figure 4.20. Knudshoved in Funen. Flint blanks lying on the beach������������������������������������������������������������������ 62 Figure 4.21. Stagstrup, Thy. Big backed knives from a private collection���������������������������������������������������������� 63 Figure 4.22. Sun chariot from Trundholm in Zealand. Dated to the early Bronze Age (1700-1300 BC). ���������� 65 Figure 4.23. Torup Høje, Jutland. Urn grave in its secondary location at the foot of the barrow. �������������������� 68 Figure 4.24. Torup Høje, Jutland. Urn from grave K49 containing unburnt flint products. ������������������������������ 69 Figure 4.25. Torup Høje, Jutland. End-scraper on blade from the urn from grave K49 ������������������������������������ 70 Figure 5.1. Northern Jutland. The vicinity of Thisted ��������������������������������������������������������������������������������������� 74 Figure 5.2. The vicinity of Skjern in central Jutland.������������������������������������������������������������������������������������������ 75 Figure 5.3. Thy and the island of Mors with barrow mounds marked in the area (points). ����������������������������� 76 Figure 5.4. Barrows constituting a permanent element of Jutland landscape. ������������������������������������������������ 78 Figure 5.5. Cult houses preserved in fragments at the foot of the barrows from Thy ������������������������������������� 80 Figure 5.6. Cult fluted objects from the vicinity of Viborg and Holstebro. ������������������������������������������������������� 82 Figure 5.7. Jernhyt near Haderslev. Southern Jutland. ������������������������������������������������������������������������������������� 84 Figure 5.8. Potential cult houses from the vicinity of Viborg: �������������������������������������������������������������������������� 85 Figure 6.1. Grydehøj, Thy. The barrow and the cult house during exploration. ����������������������������������������������� 88 Figure 6.2. Grydehøj, Thy. ��������������������������������������������������������������������������������������������������������������������������������� 89 Figure 6.3. Grydehøj, Thy. Layers of the cult house’s fill during exploration ���������������������������������������������������� 90 Figure 6.4. Grydehøj, Thy. Cult house during exploration �������������������������������������������������������������������������������� 92 Figure 6.5. Grydehøj, Thy. Cult house during exploration �������������������������������������������������������������������������������� 92 Figure 6.6. Grydehøj, Thy. Cult house during exploration �������������������������������������������������������������������������������� 93 Figure 6.7. Grydehøj, Thy. Structure of the flint assemblage from the cult house ������������������������������������������� 94 Figure 6.8. Grydehøj, Thy. Hammerstones from the interior of the cult house (N18) �������������������������������������� 95 Figure 6.9. Grydehøj, Thy. Cores from the cult house �������������������������������������������������������������������������������������� 97 Figure 6.10. Grydehøj, Thy. Sizes of blanks.����������������������������������������������������������������������������������������������������� 102

iv

Figure 6.11. Grydehøj, Thy. Perforators ���������������������������������������������������������������������������������������������������������� 103 Figure 6.12. Grydehøj, Thy. Perforators ���������������������������������������������������������������������������������������������������������� 104 Figure 6.13. Grydehøj, Thy. Notched tools ����������������������������������������������������������������������������������������������������� 104 Figure 6.14. Grydehøj, Thy. Burins ������������������������������������������������������������������������������������������������������������������ 105 Figure 6.15. Grydehøj, Thy. knife-like flakes ��������������������������������������������������������������������������������������������������� 106 Figure 6.16. Grydehøj, Thy. Distribution of pottery inside the cult house ����������������������������������������������������� 108 Figure 6.17. Grydehøj, Thy. Cultural layer inside the cult object �������������������������������������������������������������������� 109 Figure 6.18. Grydehøj, Thy. Quantity ranges of flint products within the cult house ������������������������������������� 109 Figure 6.19. Grydehøj, Thy. Weight ranges of flint products within the cult house ��������������������������������������� 110 Figure 6.20. Grydehøj, Thy. Quantity ranges in the category of chips from the cult house ��������������������������� 111 Figure 6.21. Grydehøj, Thy. Quantity ranges of blanks from the cult house �������������������������������������������������� 112 Figure 6.22. Grydehøj, Thy. Frequency of cores in individual quarters of square metres ������������������������������ 113 Figure 6.23. Grydehøj, Thy. Occurrence of tools in individual quarters of square metres ����������������������������� 114 Figure 6.24. Høghs Høj, Thy. Cult house at the foot of the barrow ��������������������������������������������������������������� 115 Figure 6.25. Høghs Høj, Thy. Cult house ��������������������������������������������������������������������������������������������������������� 116 Figure 6.26. Høghs Høj, Thy. Cult house. View from the barrow �������������������������������������������������������������������� 116 Figure 6.27. Høghs Høj, Thy. Fragment of flint collection from the cult house ���������������������������������������������� 118 Figure 6.28. Høghs Høj, Thy. Structure of the flint assemblage from the cult house. ������������������������������������ 120 Figure 6.29. Høghs Høj, Thy. Selection of cores ���������������������������������������������������������������������������������������������� 122 Figure 6.30. Høghs Høj, Thy. End-scrapers ����������������������������������������������������������������������������������������������������� 123 Figure 6.31. Høghs Høj, Thy. Perforators �������������������������������������������������������������������������������������������������������� 124 Figure 6.32. Høghs Høj, Thy. Burins ���������������������������������������������������������������������������������������������������������������� 126 Figure 6.33. Høghs Høj, Thy. Ad hoc tools – a backed knife on chunk ������������������������������������������������������������ 127 Figure 6.34. Gramstrup I, Thy. Barrow and cult house. ���������������������������������������������������������������������������������� 129 Figure 6.35. Gramstrup I, Thy. Outlines of the cult house and the barrow during exploration ���������������������� 130 Figure 6.36. Gramstrup I, Thy. Backed knife from the cult object ������������������������������������������������������������������ 130 Figure 6.37. Ginnerup, Thy. Plan of the barrow and the cult object �������������������������������������������������������������� 132 Figure 6.38. Ginnerup, Thy. Cult object on from the barrow’s eastern side ��������������������������������������������������� 132 Figure 6.39. Ginnerup, Thy. Selected artefacts from the cult house. ������������������������������������������������������������� 133 Figure 6.40. Ginnerup, Thy. Flint assemblage within the cult house during exploration ������������������������������� 134 Figure 6.41. Torup Høje, Fjelsø. Barrow during exploration in 2002 �������������������������������������������������������������� 136 Figure 6.42. Torup Høje, Fjelsø. Plan of the barrow with the location of secondary burials��������������������������� 137 Figure 6.43. Torup Høje, Fjelsø. Cross-section of the barrow and accompanying zones. ������������������������������� 138 Figure 6.44. Torup Høje, Fjelsø. One of the types of cult objects at the barrow �������������������������������������������� 139 Figure 6.45. Torup Høje, Fjelsø. Postholes occurring in pairs) ������������������������������������������������������������������������ 140 Figure 6.46. Torup Høje, Fjelsø. Location of sectors A-D within cult objects. ������������������������������������������������� 142 Figure 6.47. Torup Høje, Fjelsø. Structure of the flint assemblage from the barrow’s ritual zone. ���������������� 144 Figure 6.48. Torup Høje, Fjelsø. Stratigraphy within sector ���������������������������������������������������������������������������� 145 Figure 6.49. Torup Høje, Fjelsø. Zones of the barrow arbitrarily determined ������������������������������������������������ 146 Figure 6.50. Torup Høje, Fjelsø. Distribution of classes of artefacts ��������������������������������������������������������������� 147 Figure 6.51. Torup Høje, Fjelsø. Flint mass material from one of the explored sectors ��������������������������������� 149 Figure 6.52. Torup Høje, Fjelsø. Single-platform cores ����������������������������������������������������������������������������������� 150 Figure 6.53. Torup Høje, Fjelsø. knife-like flakes, with one cutting edge �������������������������������������������������������� 152 Figure 6.54. Torup Høje, Fjelsø. Sizes of blanks.���������������������������������������������������������������������������������������������� 153 Figure 6.55. Torup Høje, Fjelsø. 1-2 – burin, 3-4 – end-scrapers �������������������������������������������������������������������� 153 Figure 6.56. Torup Høje, Fjelsø. Perforator ����������������������������������������������������������������������������������������������������� 154 Figure 6.57. Torup Høje, Fjelsø. Flakes with traces of use ������������������������������������������������������������������������������ 155 Figure 6.58. Torup Høje, Fjelsø. Perforators ��������������������������������������������������������������������������������������������������� 156 Figure 7.1. Percent contribution of flint products from ritual zones �������������������������������������������������������������� 165 Figure 7.2. Size of the tool categories in individual assemblages of ritual zones in northern Jutland.����������� 166 Figure 7.3. Two types of big backed knives identified in site Fosie IV. ������������������������������������������������������������ 167 Figure 7.4. Big backed knives: ������������������������������������������������������������������������������������������������������������������������� 168 Figure 7.5. Big backed knives: ������������������������������������������������������������������������������������������������������������������������� 170 Figure 7.6. Wierzbica, province of Mazowsze. Backed knife of the Zele type ������������������������������������������������ 172 Figure 7.7. Zakrzów 41, province of Opole. Flake backed knives �������������������������������������������������������������������� 173 Figure 7.8. Zakrzów 41, province of Opole. Flake backed knives with traces of polishing from harvest use � 173 Figure 7.9. Modlniczka 2, province of Małopolska. Lusatian culture burins ��������������������������������������������������� 175 Figure 7.10. Grydehøj, Thy. Single-blow burin doubled on chunk ������������������������������������������������������������������ 176 Figure 8.1. Technological microtraces: ����������������������������������������������������������������������������������������������������������� 183 Figure 8.2. Microtraces on end-scrapers and side-scrapers ��������������������������������������������������������������������������� 185 Figure 8.3. Microtraces on perforators and borers ����������������������������������������������������������������������������������������� 186

v

Figure 8.4. Microtraces on perforators and burins ����������������������������������������������������������������������������������������� 187 Figure 8.5. Microtraces on perforators and burins: ���������������������������������������������������������������������������������������� 189 Figure 8.6. Microtraces on notched tools and flakes and on the blade knife: ������������������������������������������������ 190 Figure 9.1. Grydehøj, Thy. Distribution of phosphorus within the cult house. ����������������������������������������������� 197 Figure 9.2. Grydehøj, Thy. Content of copper in samples collected in the cult house and its surroundings.�� 199 Figure 9.3. Grydehøj, Thy. Results of palynological analysis. �������������������������������������������������������������������������� 201 Figure 9.4. Examples of herbaceous plants.���������������������������������������������������������������������������������������������������� 201 Figure 9.5. Grydehøj, Thy. Calibration of radiocarbon dating. ������������������������������������������������������������������������ 204 Figure 10.1. Grydehøj, Thy. Reconstruction of the cult house’s outlines ������������������������������������������������������� 207 Figure 10.2. Torup Høje, Fjelsø. Cult objects. ������������������������������������������������������������������������������������������������� 208 Figure 10.3. Grydehøj, Thy. Cult house ����������������������������������������������������������������������������������������������������������� 210

List of Plates Plate 1.������������������������������������������������������������������������������������������������������������������������������������������������������������ 235 Plate 2.������������������������������������������������������������������������������������������������������������������������������������������������������������� 236 Plate 3.������������������������������������������������������������������������������������������������������������������������������������������������������������� 237 Plate 4.������������������������������������������������������������������������������������������������������������������������������������������������������������� 238 Plate 5.������������������������������������������������������������������������������������������������������������������������������������������������������������� 239 Plate 6.������������������������������������������������������������������������������������������������������������������������������������������������������������� 240 Plate 7.������������������������������������������������������������������������������������������������������������������������������������������������������������� 241 Plate 8. ������������������������������������������������������������������������������������������������������������������������������������������������������������ 242 Plate 9. ������������������������������������������������������������������������������������������������������������������������������������������������������������ 243 Plate 10. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 244 Plate 11. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 245 Plate 12. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 246 Plate 13. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 247 Plate 14. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 248 Plate 15. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 249 Plate 16. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 250 Plate 17. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 251 Plate 18. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 252 Plate 19. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 253 Plate 20. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 254 Plate 21. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 255 Plate 22. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 256 Plate 23. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 257 Plate 24.����������������������������������������������������������������������������������������������������������������������������������������������������������� 258 Plate 25. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 259 Plate 26. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 260 Plate 27. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 261 Plate 28. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 262 Plate 29. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 263 Plate 30. ���������������������������������������������������������������������������������������������������������������������������������������������������������� 264

vi

List of Tables Table 2.1. List of main sites discussed in this work. ������������������������������������������������������������������������������������������ 10 Table 4.1. Bjerre 7, Thy. Structure of flint assemblage�������������������������������������������������������������������������������������� 39 Table 4.2. Bjerre 7, Thy. Sizes of backed knives ������������������������������������������������������������������������������������������������ 42 Table 4.3. Bulbjerg Troldting, Thy. Sizes of backed knives according to types A-D. ������������������������������������������ 44 Table 4.4. Fragtrup. Structure of flint assemblage�������������������������������������������������������������������������������������������� 47 Table 4.5. Skamlebæk, Zealand. Sizes of backed knives according to types A-D. ��������������������������������������������� 51 Table 4.6. Vinde Helsinge, Zealand. Sizes of backed knives according to types A-D. ���������������������������������������� 53 Table 6.1. Grydehøj, Thy. Structure of flint assemblage from two stratigraphic units ������������������������������������� 94 Table 6.2. Grydehøj, Thy. Flint raw material within stratigraphic units ������������������������������������������������������������ 96 Table 6.3. Grydehøj, Thy. Cores from two stratigraphic units (N17, N18) within the cult house���������������������� 98 Table 6.4. Grydehøj, Thy. Sizes of individual classes of artefacts within stratigraphic units ����������������������������� 99 Table 6.5. Grydehøj, Thy. Blanks from two stratigraphic units ����������������������������������������������������������������������� 100 Table 6.6. Grydehøj, Thy. Presence and manner of preparing butts of blanks and tools) ������������������������������ 101 Table 6.7. Grydehøj, Thy. Tools from two stratigraphic units (N17, N18) within the cult house��������������������� 103 Table 6.8. Høghs Høj, Thy. Structure of the flint assemblage ������������������������������������������������������������������������� 117 Table 6.9. Høghs Høje, Thy. Cores within the cult house��������������������������������������������������������������������������������� 121 Table 6.10. Høghs Høj, Thy. Blanks within the cult house������������������������������������������������������������������������������� 122 Table 6.11. Torup Høje, Fjelsø. Structure of the flint assemblage������������������������������������������������������������������� 143 Table 6.12. Torup Høje, Fjelsø. Cores�������������������������������������������������������������������������������������������������������������� 150 Table 6.13. Torup Høje, Fjelsø. Blanks������������������������������������������������������������������������������������������������������������� 151 Table 6.14. Torup Høje, Fjelsø. Tools��������������������������������������������������������������������������������������������������������������� 154 Table 6.15. Torup Høje, Fjelsø. Presence and manner of preparing butts of blanks and tools����������������������� 155 Table 6.16. Torup Høje, Fjelsø. Sizes of individual classes of artefacts ����������������������������������������������������������� 157 Table 7.1. Cores and products of preparation from the early phase of coring.����������������������������������������������� 161 Table 7.2. Percent comparison of the most numerous categories of cores and blanks ��������������������������������� 163 Table 7.3. Burins from cult objects in northern Jutland.��������������������������������������������������������������������������������� 176 Table 8.1.Number of analysed artefacts and the maximum number of used products���������������������������������� 184 Table 9.1. Anthropological analysis of the osteological material �������������������������������������������������������������������� 195 Table 9.2. Grydehøj, Thy. Macrobotanical remains from the cultural layer of the cult house������������������������ 199 Table 9.3. Grydehøj, Thy. Radiocarbon chronometry of samples from the cultural layer ������������������������������ 203 Table 10.1.Basic differences perceived between stone assemblages form the cult houses and ritual zones� 211

vii

viii

Introduction In 2009 Swedish archaeologist Anders Högberg published his doctoral thesis in the British Archaeological Reports (BAR) International Series, which was devoted to the examination of flintworking during the Scandinavian Late Bronze Age. The book was already in print a season earlier when, during a meeting at the Malmö Museum where Anders was working, we were discussing the occurrence of flint artefacts in various cultural contexts during the Metal Ages. While sharing my observations concerning the assemblages from northern Jutland, I mentioned the presence of flint artefacts in cult structures of the Thy region. We soon realised that he had overlooked this phenomenon in his outstanding1 thesis (Högberg 2009). Even though the first flint collections from cult houses had been stored by the Malmö Museum (the earliest flint assemblage was discovered in 1978 during the excavation of the Høghs Høj cult house in northern Thy), none had been analysed, and no reports had been published. Knowledge of these flint collections was confined to the relatively small circle of individuals who had acquired them. Publications concerning the structures described as cult houses did not include any information about the density of flint artefacts found in some of them (Nielsen 1999, 2000; Nielsen & Bech 2001; Victor 2002; Kaul 1993, 2006). The presence of flint artefacts typical’ of the Late Bronze Age was only briefly mentioned in Bjarne Henning Nielsen’s and Jens-Henrik Bech’s (2001) publication that focused on the cult houses in the Thy region. From the conversation in Malmö there was only one step to my involvement in the research of flint assemblages occurring within the cult structures of the Late Bronze Age in northern Jutland and resigning from the original purpose of my work in Scandinavia, which was the comparative study of flintworking at the turn of Metal Ages in southern Scandinavia and central Europe emphasizing areas in modern-day Poland. The results of the research are presented in this book, and the interim results were also presented elsewhere (Masojć, Bech 2011; Masojć, Bech, Kufel-Diakowska 2013). My interest in flintworking in southern Scandinavia has its beginnings in rescue excavations conducted by the Wrocław Section of the Institute of Archaeology and Ethnology of the Polish Academy of Sciences during the construction of the A4 motorway in the province of Opole. The excavations were carried out in collaboration with Jarosław Bronowicki (Bronowicki, Gediga, Masojć 2002). At that time my research interests primarily focused on prehistoric hunter-gatherer communities (Masojć 2004). When the excavations revealed disparate and specific flint artefacts occurring at practically every site within the context of settlements from the Lusatian 1

Reviews: McLaren A. P. 2011; Werra D. 2009, 2011.

ix

culture, we studied them with great interest. Our analysis of the Lusatian flints and the resulting interpretation of the cultural attribution of those very ‘late’, as we then thought, flint assemblages was viewed with moderate optimism by professor Bogusław Gediga. It was only after we found two backed knives situated near an overturned vessel from the Lusatian culture in feature 91B from Zakrzów 41 in the district of Krapkowice (Bronowicki, Masojć 2010, Fig. 8) that our principal investigator became convinced as to the accuracy of our observations. Reports on the Lusatian flint material from the province of Opole were favourably received by our colleagues specialising in ‘stone’ matters. While presenting the results of our work at the 15th Congress of the International Union of Prehistoric and Protohistoric Sciences (UISPP) in Lisbon we realised that our finds fit perfectly within the framework of flintworking production from the Late Bronze Age in northern Europe (Bronowicki, Małecka-Kukawka, Masojć 2006). For example, artefacts analogous to ours had been found in the British Isles (Ballin 2010), Scandinavia (Eriksen 2010) and even in the Near East (Rosen 1996). Thus, my studies on flintworking associated with the cult structures in Jutland arose from the desire to examine closely the Scandinavian assemblages and compare them with the material from Zakrzów. All this became possible thanks to the Danish Government Scholarship (2008) and the scholarship from the Kazimierz Salewicz Foundation (Kazimierz Salewicz og hustru Marit Jensens Studiefond) awarded twice – in 2010 and 2012. Palaeoecological analyses of the Grydehøj cult house, radiocarbon dating and the analyses of the function of the flint artefacts were carried out with the support of an annual grant from the Polish National Science Centre in the third edition of the Opus programme (2013). The project would not have been completed without the exceptionally friendly attitude of Danish archaeologists with whom I had the privilege of cooperating. First and foremost I should mention two prehistorians from museums in Jutland – Jens-Henrik Bech from the Thisted Mueum and Martin Mikkelsen from the Viborg Museum, who allowed me access to the flint materials and guided me through the archaeology of the Bronze Age. I am deeply impressed by their academic knowledge combined with many years of field experience. I would also like to thank my colleagues Klavs Randsborg (University in Copenhagen), Poul Otto Nielsen and Flemming Kaul (National Museum in Copenhagen), Jørgen A. Jacobsen and Mogens Bo Hendriksen (Odense City Museums), Mikkel Kieldsen (Viborg Museum) and Anne Loise Haack Olsen (Thisted Museum). They were all very helpful during my Scandinavian studies. I would like to express my special thanks to Lisse and Knud Hove (Thisted) for their hospitality during my stays in Thy and also for their kindness. I am also indebted to Jerzy Piekalski from my university for his continuous encouragement to accomplish this work. Without the professional and administrative assistance of the people and institutions mentioned above I would have been unable to accomplish the task which I set for x

myself in this work. Of course all the potential errors and shortcomings are only mine. I also hope that this monograph, which is the culmination of a few years of study, is not going to be the last stage of my Danish and more generally Scandinavian encounter with prehistory.

xi

xii

1. On the phenomenon of flintworking at the end of the Bronze Age and beginning of the Iron Age For hundreds of thousands of years of human prehistory, until the technology of the use of metals emerged and subsequently became widespread, fissile rock had been the most important raw material in tool production and economy of prehistoric societies. From the earliest records of the use of stone as a tool by early hominids, it had a basic practical function, but it also carried cultural and symbolic meanings. The symbolism and extra-utilitarian character of knapped stone implements are seen in the Stone Age (Fiedorczuk et al. 2007; Płonka 2012). They constitute permanent and essential elements of funerary rites and many other extra-utilitarian aspects of life (Gamble 1999). It is especially apparent in the Neolithic and even more so in the Early Bronze Age, when ‘fanciful’ bifacially retouched flint daggers and sickles were produced and distributed in Europe (Apel 2001; Libera 2001). It is a common conviction that the end of the Bronze Age in Europe brought about a slump and decline in the technology of flint production. The simplest explanation would be the increase in the dynamics of circulation of metal tools, effectively replacing ‘obsolete’ stone tools. Some proponents of this opinion argue that popularity of metal objects hindered production of stone tools, in principle used only for utilitarian purposes, depriving them of aesthetic and symbolic qualities (Humphrey 2004). Because of the use of metal, advanced flintworking techniques were to sink into oblivion. Flint was still used by Late Bronze Age communities, but reduced solely to opportunistic, ad hoc forms. The range of formal tools gradually decreased to a minimum with flint products retaining a strictly functional and utilitarian character, while more technologically advanced metal tools acquired symbolic meaning (Edmonds 1995). The arguments and conclusions presented in this work go against the thesis presented above. I will try to prove that the indisputable technological change in flintworking, which occurred between the older and younger Bronze Age, did not deprive it of its metaphysical qualities. It is quite possible that the opposite happened, as it is during this time that the presence of dense flint assemblages in ritual contexts is observed. I am of the opinion that despite the widespread belief that the significance of flint implements was only of a marginal nature, their presence in and impact on the culture of the societies at the end of Bronze Age and the beginning of the Iron Age is a complex and multifaceted phenomenon – from a purely utilitarian character (it is an element of the economy of that time with its specific system of raw material acquisition, its distribution, flint production and use of flint implements) to the extra-utilitarian aspect (presence of flint in symbolic culture), exceeding the limitations of ordinary 1

2

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia consumption and economy (Lech 1997a). With flintworking during this transitional time during the Metal Ages and the assumption of its multifaceted nature in mind, I will try to familiarise the reader with a so far unknown aspect of flint production from the Late Bronze Age in Scandinavia – the ubiquity of flint artefacts in ritual contexts, including the so-called cult houses, situated at the foot of the barrows of northern Jutland. I will try to illustrate the significant role that flint played in community life, both as raw material and artefact, not only as the material used in everyday activities of daily life, but also as an important component of rites and observances strongly connected with the religion and rituals of the end of the Bronze Age in Scandinavia.

2. Temporal and spatial framework, concept of the work, and methods If treated in very general terms, the scope of this work is concerned with Late Bronze Age flintworking in southern Scandinavia. Yet, such a chronological and geographical framework requires further clarification, as does its theoretical framework, especially considering that I aim to go beyond these broadly defined boundaries. The work by A. Högberg (2009) mentioned in the introduction examined the same time period and geographical area. It will be frequently referred to, as it is relevant to the subject matter of this work also in terms of its findings. It would be a mistake to ignore this invaluable work, not least because, having been written by an author from Scandinavia, it presents an emic perspective. It is also important in that I found it very helpful, thanks to its numerous references to the Scandinavian literature devoted to the subject, rarely published in major languages, with which I had to struggle. To be more precise about the context of my research, the physical setting is northern Jutland in modern-day Denmark (Figure 2.1) – one of the five main administrative units of the country. All of the flint assemblages analysed in great detail come from this area, or more specifically – from its western part. Cult houses can be found in the region of Thy, which is situated between the North Sea and the Limfjord. The

Figure 2.1. Basic spatial framework of the work, in the rectangle, against the background of the extent of cultural traditions of the late Bronze Age in northern and central Europe (after: Kristiansen 1998).

3

4

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 2.2. Southern Scandinavia. The big ellipse in the area of northern Jutland marks the core of geographical extent of the work. The remaining small ellipses in Funen and Zealand show the location of additionally analysed sites. The circle linking Zeeland and Scania marks the geographical extent of A. Högberg’s study (2009).

remaining cult features are found to the east and south of the Limfjord, including the island of Mors, mainly in the vicinity of Viborg (Figure 2.3). Cult features in the form of the so-called cult houses are present in the widely conceived area of southern Scandinavia, in the belt whose northern boundary is formed by a line stretching from northern Jutland in Denmark to Uppland in Sweden, while its southern boundary runs from the southern shores of the Danish islands Funen and Zealand to the southern stretches of Scania in Sweden (Victor 2002) (Figure 2.2). The

2. Temporal and spatial framework, concept of the work, and methods

Figure 2.3. North-western Jutland – the Thy region, the island of Mors and the vicinity of Viborg (the area marked with the ellipse in the previous illustration). Main sites discussed in the work. 1-3 settlements; 4-8 barrows and their ritual zones. 1 - Bjerre;, 2 - Bulbjerg Troldting; 3 - Fragtrup; 4 - Grydehøj; 5 - Høghs Høj; 6 - Ginnerup; 7 Gramstrup; 8 - Torup Høje (drawing by the author).

area thus defined comprises the ritual contexts analysed in this work; however, the cult features featuring homogenous flint assemblages were observed only in northern Jutland. The remaining areas provided flint material, if at all, in the form of isolated artefacts. Therefore, Jutland’s western stretches distinctly differ from the greater area of occurrence of Late Bronze Age cult houses. Archaeological sites of another character, whose artefacts constitute an integral part of this work, have also been observed in this area. I decided that the discussion of flint

5

6

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia assemblages from the cult features should be studied within the larger context of Late Bronze Age flintworking, and I included additional collections in my research. Three of these come from within the geographical limits of the study mentioned above (Bjerre, Fragtrup and Troldsting), three are from Funen (Højby, Voldtofte and Knudshoved) and another two are from Zealand (Skamlebæk and Vinde Helsinge). Assemblages from these sites provided the basis for my consideration of the larger question of flintworking during the Late Bronze Age. They are by no means the main subject of this work – as they only constitute the background for the research as indicated in the title. Yet another site, situated even further away in the continental part of central Europe, will be discussed in even more general terms – site Zakrzów 6 in southwestern Poland. The material from this site will be referenced in the discussion of select aspects of flintworking (Bronowicki, Masojć 2010). I analysed the lithic collections from the abovementioned sites located in modern-day Denmark. The reader will also encounter references to sites and analyses which I know solely from the literature. Sites and collections located within the geographical centre of the work will be treated in greatest detail; the further away from northern Jutland, the more general the remarks and conclusions will be. The geographical centre of the work is indicated by the ellipse in Figure 2.2. Sites located outside northern Jutland, constituting only a portion of my analyses, are marked with smaller red circles. The geographical extent of the work coincides with the extent of A. Högberg’s (2009) publication based on the material from Scania in Sweden and eastern Zealand in Denmark. In terms of its subject matter, Högberg’s work is devoted to other phenomena. It is a multifaceted study principally devoted to one category of artefacts identified as ‘large backed knives’, while this work is concerned with flint assemblages that do not contain such artefacts. On the other hand, both works deal with the well-known issue of Late Bronze Age flintworking in Scandinavia. Even though the two studies do not overlap in terms of their areal extent or subject matter, they complement each other in their treatment of flintworking during the same time period. The phenomena analysed in the work date to the younger part of the Bronze Age (periods IV – VI), which in absolute terms corresponds to 1150/1000 – 500 BC. The work’s time span comprises about 500 to 600 years, when the communities from the Nordic tradition/circle (Figure 2.1) deposited cremated burial graves in the mounds of the Early Bronze Age barrows, simultaneously constructing cult features of diverse forms at their foot. In central Europe this period coincides with the Lusatian culture and corresponds to the Ha A1/2 – Ha D1/2 phases according to Reinecke’s chronological system (Hvass, Storgaard 1993, chronological table after page 312; Kristiansen 1998, Figure 13; Earle, Kristiansen 2010b, 22) (Figure 2.4).

2. Temporal and spatial framework, concept of the work, and methods

In sum, I aim in my work to characterise and interpret a select, so far unknown aspect of flintworking from the Late Bronze Age in northern Europe, i.e. the flint assemblages present within the cult structures. They were observed in northwestern Jutland, which constitutes the geographical centre of this study (Figure 2.3). Whether it is a local phenomenon or present at the supraregional scale will be discussed in more detail later on in this work. The phenomenon in question did not function in isolation; it was an element of everyday life of Late Bronze Age communities and as such it should be presented within the larger cultural context. Therefore it seems sensible to present a general picture and a specific characterization of flint production during that period. Thus, apart from the flint assemblages from ritual contexts, this work features contemporary sites with flint assemblages of a different character, Figure 2.4. Basic chronological e.g. those from secular contexts. They come divisions of the late Bronze Age for both from the areas in the immediate vicinity the area comprised by the study (after: Kristiansen 1998). of the occurrence of the discussed phenomena and from more distant regions, such as the Danish islands. The picture is completed with remote analogies of assemblages from the continent’s central part. The nature and significance of the above-mentioned phenomenon will be presented on the basis of its detailed characteristics. Since we are confronted with widespread stone ‘breaking’ within the so-called cult houses and other features of indisputably cult character, in my opinion we may assume at the very early stages of the work that this instance of flintworking is potentially of a ritual nature. I will attempt to provide answers to a series of questions that emerged during the analysis of flint sources and the context of their occurrence, concerned with the causes of such ubiquity of flint in cult houses. The nature of activities performed within these structures will be analysed, and they will be compared with typical economic activities, similar to those that took place in settlements. This will help establish whether they were houses and structures where economic or alternatively ceremonial activities were executed. Mass reduction of flint nodules and production of artefacts will be considered in relation to its potential contribution to the rites and rituals taking place there. Evidence of use of the flint aretefacts from within the structures will be described to explain their function and

7

8

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia potential use of objects displaying traces of work in ceremonial activities. In this work I will undertake to address these questions, along with many others. It would be worthwhile to explain the phrase ‘ritual zones’ quoted several times already. I use this term to include in one definition both the structures constituted by the so-called cult houses and other cult features present at the foot of the barrows which do not form such dense constructions. They will all be described further on in this work. In order not to mention repeatedly ‘cult houses and other cult features situated directly at the foot of the barrows’ where flint artefacts occur, I refer to all these areas as cult or ritual zones. This work does not address the issue of the occurrence of flint artefacts in Late Bronze Age graves. These are the artefacts delivered to the sacral zone and deposited there. So far no vestiges of their production have been found within them. They also do not differ, contrary to the flintworking from the ritual zones discussed here, from the analogous artefacts produced and used in settlements or workshops. The section devoted to the flint artefacts found in graves aims at presenting a full spectrum of flint presence in the culture of the Late Bronze Age communities; yet I distinctly distinguish between flint production in the ritual zones and isolated flint implements present in graves. The sources analysed in the work exclude any direct relationships between them. The example of flint assemblages from the ritual zone of the Torup Høje near Viborg and the flint artefacts present in the grave urns deposited in the barrow (cf. paragraph 4.2.4) seem to confirm their disparate nature. One question may be dispelled already at this stage of the work. Cult constructions from northern Denmark, where the predominating category of artefacts is constituted by flint products, prove the significant role of this raw material in the economy and awareness of Late Bronze Age societies. This contradicts the thesis that flint products, whose practical and symbolic functions were so significant in the Stone Age and the Early Bronze Age, cease to carry a symbolic meaning in the Late Bronze Age, becoming reduced to performing solely utilitarian functions, which has been proposed as an explanation of the change in flintworking technology of that period (Edmonds 1995, 188; Humphrey 2004: 244-245). Late Bronze Age flintworking is a complex phenomenon. Undoubtedly it is not an example of the highest quality of production, especially in the case of ad hoc products ubiquitously present in settlements. However, their extensive presence in the sites, their technological dualism and, most importantly, the multitude of purposes of their use by no means diminishes in relation to the times when stone was the leading raw material used in the production of tools. The flint assemblages presented in this work have been acquired by Danish museums during the last forty years of excavations. The assemblages, which I comprehensively

2. Temporal and spatial framework, concept of the work, and methods

analysed, constitute a complete source base for study of flintworking in the ritual zones in southern Scandinavia defined above. I also analysed the collections constituting the background for presenting the full picture of flintworking during the Late Bronze Age. The majority of these assemblages have not yet been published. My analysis took into account stylistic and morphological features. For the assemblages displaying a homogenous character I used a dynamic classification. Selected artefacts were examined with a microscope to identify and interpret traces of their use. In the case of the two best preserved sites – Grydehøj and Torup Høje – I also analysed the spatial arrangement of flint artefacts. I took into account both mutual spatial relationships between individual artefacts and the nature of distribution of the assemblage in relation to the cult structures – the interior of the Grydehøj cult house and the Torup Høje cult features, which provided interesting data concerning their homogeneity and stratigraphic position. Diverse research methods were used in the case of the best examined structure presented in this work – the Grydehøj cult house. The aim was to illustrate the context in which the flint assemblage from the interior of the house was situated, as well as arrive at the exact temporal relationship between the graves situated near the structure and the structure itself. Most analyses were carried out for the needs of this work. Before they commenced, only an osteological analysis of the faunal remains discovered in the structure’s cultural layer and of the burials deposited near the house had been carried out by Sign Andersen from the National Museum in Copenhagen. The grant from the Polish National Science Centre provided for additional analyses: palynological (Department of Palaeobotany, Institute of Geological Sciences, University of Wrocław), macrobotanical (Moesgård Museum), geochemical, (Chair of Plant Breeding, Wrocław University of Environmental and Life Sciences) and radiocarbon (AMS14C Dating Centre, Department of Physics and Astronomy, Aarhus University). During my research I enjoyed access to the collections of four Danish museums: the National Museum in Copenhagen (Nationalmuseet, Københaven), the Odense City Museums (Odense Bys Museer), the Viborg Museum (Viborg Stiftsmuseum) and, most importantly, the Thisted Museum (Museet for Thy Og Vester Hanherred, Thisted). So that the reader can access information on each site discussed in the work, Table 2.1 presents the numbering of the sites, which enables one to find site specific information in the database of Danish archaeological sites available at www.kulturarv.dk, where exhaustive information on research and the location of the available material may be found. This manuscript is organized in such a way that with each chapter the reader will be able (I hope) to familiarise himself or herself with selected illustrations in the form of photographs of the flint artefacts and drawings without the need to refer to detailed tables presenting comprehensive descriptions of the discussed flintworking. These are

9

10

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia featured at the end of the work for the reader interested in more detailed information about the flint material discussed here. Additionally, the main text is accompanied by two inserts - illustrations with captions presenting in more detail two objects of great significance for this work: the most interesting cult house in Scandinavia – Sandagegård and a complete Late Bronze Age sickle, the best preserved among the ones known so far, from the Stenild peat bog. Both objects are frequently referred to in the pages of this book; therefore, their characteristics do not entirely seem out of place.

Table 2.1. List of main sites discussed in this work. Numbering of sites enables their identification in the database of Danish archaeological sites after Fund og Fortidsminder available on the webpage www.kulturarv.dk. Codes of museums: THY – Museum in Thisted (Museum for Thy og Vester Hanherred, Thisted); NM – National Museum in Copenhagen (Nationalmuseet, København); OBM – Museum in Odense (Odense Bys Museer); GM – Museum in Gilleleje (Gilleleje Museum); VSM – Museum in Viborg (Viborg Stiftsmuseum) Site name

Commune

Region

Site number

Local museum accession number

Thy

No. in the system 128575

Bjerre

Thisted

Bulbjerg Troldsting

Thisted

110211-32

THY 3718

Thy

22713

110707-15

NM B9853,K/492-K/908

Fragtrup Ginnerup

Farsø Thisted

Vesthimmerland Thy

23731 21254

120204-135 110605-58

NM 766/56

Gramstrup

Thisted

Thy

21665

THY 3874

Grydehøj

Thisted

Thy

21939

110612-58 110612-322

Høghs Høj

Thisted

Thy

19036

110310-6

THY 1417

Højby

Odense

Funen

124012

080804-15

OBM 4640

Knudshoved Sandagergård

Nyborg Ferslev

Funen

146821 92893

090611-101 010202-262

OBM 7446

Zealand

Skamlebæk

Odsherred

Zealand

030403-531

NM 1040/75

Stagstrup Torup Høje

Thisted

Thy

165314 164511

Viborg

N Jutland

38449

Kalundborg

Zealand

Assens

Funen

Vinde Helsinge Voldtofte

THY 5055 THY 3759

GM 3063

110111 - 148

THY 1962 VSM 254G

100797

130901-60 030209-58

NM 685/39

167458

080205-335

OBM 1625

3. Cult houses: definition, idea, chronology Before beginning the discussion of the issue of flintworking in the cult zones in northern Jutland, some attention should be devoted to the cult features themselves, and especially the cult houses, within which intensive flintworking activity was observed. The remains are flint assemblages, which, incidentally, were not initially interpreted as potentially connected with ritual activities taking place within the cult structures. The first flint assemblages from Høgs Høj and Grydehøj in Jutland were so surprising for their discoverers that initially their presence and function were interpreted as identical to that of the rubble and pebbles constituting the houses’ construction elements and the surrounding stone benches. However, closer examination proved the intentional nature of the flint assemblages, easily identifiable in the construction rubble of the stone benches. The features referred to as the so-called cult houses were identified relatively early on, i.e. at the turn of the 20th century. The first building from the Bronze Age discovered in southern Scandinavia at the beginning of the 20th century was the cult house in Bods near Enköping in Sweden, erroneously interpreted as a residential structure (Almgren 1912). 3.1. Cult houses from the stone age In Scandinavia the well documented tradition of erecting cult buildings dates back to the late Neolithic (Becker 1993). It may have been a continuation of an even older Mesolithic tradition. Site Skateholm II in Sweden, one of the best known European cemeteries from the Middle Stone Age, yielded a feature that Lars Larsson (1988) interpreted as the remains of a ceremonial building from the Ertebølle culture. It was a quadrangular post construction situated in the cemetery’s central part, constituting the most exposed place in the area. The construction was surrounded by a narrow strip of ochre and covered by a layer of the same dye. The fill of the postholes contained burnt matter (Figure 3.1). The building’s interior contained flint implements and bone fragments of various animal species, including the rarely encountered remains of an eel. L. Larsson concluded that the structure had been used for meals as an element of ritual ceremonies. The author also refers to another residential structure from the Ertebølle culture in Bredasten (M. Larsson 1986), maintaining that its ceremonial character is substantiated by the presence of a young dog’s burial inside it, which, however, does not seem a very convincing argument for the building’s ritual character (L. Larsson 1988). According to I. J. Thorp (1999), the feature from Skateholm II may evidence the presence of a kind of Mesolithic cult house in Scandinavia, though for him the data 11

12

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 3.1. Skateholm, Sweden. Plan and profiles of a supposed cult house. 1 – sand with abundant admixture of red ochre; 2 – sand with admixture of red ochre; 3 – sand mixed with burnt matter; 4 – sand with abundant burnt matter; 5 - sand; 6 – colour of wood; 7 – post objects; 8 - stones. In the part of the object filled with sand with admixture of ochre concentration of flint material sometimes exceeding 100 artefacts per one metre was recorded (after: L. Larsson 1988, amended by the author).

3. Cult houses: definition, idea, chronology

above are insufficient premises for a clear-cut distinction between ceremonial and secular buildings. The author references the example of residential buildings containing, among other things, skeletal remains and sculptures of fantastic creatures (hybrids of fishes and people) from the Middle Stone Age in the area of the Iron Gates at the boundary between Serbia and Romania (e.g. Lepenski Vir, Vlasac, Padina – cf. Srejović 1972). In his opinion this proves that such an unequivocal division may not be justified in the case of the Mesolithic (Thorpe 1999: 24-26, 78). Its absolute uniqueness and absence of a greater number of features of such character may additionally speak against recognising the building from Skateholm as a Mesolithic ceremonial house. To return to the Scandinavian features of the younger Stone Age, one of the first Neolithic cult houses discovered in the 1950s was a quadrangular structure situated in the group of megaliths from the Funnel Beaker culture in Tustrup in Djursland Peninsula in eastern Jutland (Figure 3.2) (Kjærum 1955; Jensen 2001). It was a horseshoe-shaped stone construction containing about thirty Funnel Beaker culture vessels. It is not clear whether it had a roof. According to the author of the research, apart from the massive stone casing, it had a gable wooden roof and post construction walls (Kjærum 1955, Figure 12); the latter was substantiated by rows of postholes arranged along the interior of the walls (Figure 3.2, grey hatching). The Tustrup cult house is a unique building among other Funnel Beaker culture cult houses in Jutland, which might result from its spatial and functional relationship with megalithic features. Most Neolithic structures of this nature are less monumental

Figure 3.2. Tustrup, Djursland, eastern Jutland. Cult house from the Funnel Beaker culture. Circles denote the vessels discovered inside the construction. A hole without artefacts was located in the central part. Hatching along the walls indicates the location of rows of post objects. The photograph shows the location of the house within the group of megalithic graves. A dolmen is seen in the background. Outside the photograph, there is a passage grave to the right of the object and another dolmen several dozen metres further (drawing after: Kjærum 1955; photograph by the author, Nov. 2012).

13

14

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 3.3. Outlines of cult houses from Funnel Beaker culture in northern Jutland. Hatching marks post objects and destroyed areas. Note differentiated orientation of the objects. Objects from the left top row: Tustrup, Herrup XXVI, Trandum Skovby II, Søndermølle III, Herrup XLIII, Herrup XLIV, Engedal, Foulum (after: Kjærum 1955; Jensen 2001).

(Becker 1993, 111). Eleven such structures are known, and they occur solely in central Jutland, concentrated in the area south of the Limfjord, mainly between Viborg and Holstebro ( Jensen 2001: 397). These have survived mainly in the form of quadrangular foundation trenches with one open side, in some cases as a complete quadrangle with two extended, parallel sides forming a kind of vestibule (Figure 3.3). The presence of isolated postholes and pebbles in the foundation ditches is interpreted as the remains of the constructions supporting the roof (Figure 3.4). The structures’ sizes vary; the length of side walls range between 9 (the longest) and 1.5 metres in the case of the smallest structures. Similarly as in Tustrup, the remaining houses yielded vessels representative of the Funnel Beaker culture accompanied by ceramic cult spoons, not normally encountered in the settlements from this culture (Becker 1993: 110-111).

3. Cult houses: definition, idea, chronology

Figure 3.4. Reconstruction of a cult house Tustrup, Djursland (after: Andersen 2000).

There are no dense flint assemblages to be found in the cult houses from the Funnel Beaker culture, and neither were any burials found within their confines. While they display no tendency to orientate in relation to the cardinal directions, they remain in a close spatial relationship to the graves or cemeteries situated in the vicinity, which seem to constitute their sole spatial determinant. Most cult houses from the Funnel Beaker culture were deliberately deconstructed, possibly by their users when they ceased to perform their ritual function. The buildings were dismantled, and their remains were covered with the layers of pebbles (Becker 1993). The spatial relationship of the cult houses from the Funnel Beaker culture to the graves prompted the researchers to associate these Neolithic cult features with the worship of the dead (Fabricius, Becker 1996; Jensen 2001; Kjærum 1955). The interpretation of the function performed by the above-mentioned features takes into account the function of subsequent cult houses from the Bronze Age. Similarly as with most cultural manifestations of the Late Neolithic in Scandinavia, which lend themselves to archaeological interpretation, also the idea of the Neolithic cult houses persisted in the Bronze Age ( Jensen 1987: 156). 3.2. Cult features of the Early Bronze Age The literature devoted to the issue in question mentions an apparent discontinuation in the occurrence of cult features between the Neolithic and the Late Bronze Age. In principle in Jutland there are no records of the Early Bronze Age cult features analogous to those from the Neolithic. This gap is partly filled by cult features from Scania and more remote regions of Sweden, where stone structures interpreted as cult houses are frequently dated to the Early Bronze Age (Victor 2002). However, as I have already mentioned, this absence of cult features from the Early Bronze Age in Jutland is only apparent, resulting from the stage of the research and the scope of interpretation of the existing publications. When compared with the Neolithic cult houses, ritual features from the Early Bronze Age do not display such consistent standardisation in their construction. Their outlines

15

16

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia are seen from concentric or quadrangular stone arrangements, in which they also differ from the Neolithic features mainly discernible as quadrangular foundation ditches. Their spatial relation to graves is analogous, though the Early Bronze Age features usually directly adjoin the embankments of the contemporary barrows, while Neolithic features may be situated at a certain distance from the burials. In her work Benita Clemmensen (2005) mentions several cult features, e.g. Jernhyt near Haderslevt, whose chronology may be determined more precisely on the basis of stratigraphic premises for the end of the Neolithic and the beginning of the Early Bronze Age. They were four-walled stone constructions to which a few words will be devoted in the chapter on the diversity of cult features. Additional possible examples of such features from the Early Bronze Age may be found in the volumes concerning the areas of Thisted (volume XI) and Viborg (volume XII) from the twenty-volume series devoted to the Early Bronze Age finds from modern-day Denmark, Schleswig and Lower Saxony (Aner et al. 2001, 2008). In their majority they are the so-called stone box graves of symbolic character as they do not contain any human remains or elements of furnishing. They are most frequently found at the foot of the Early Bronze Age barrows. Examples will be presented below of features from the vicinity of Viborg and Thy, which in my opinion are quite probably the remains of cult buildings analogous to the Neolithic cult houses, even though in no publication were they interpreted as such. In the 1970s J. Christoffersen excavated at the Sennels site what he thought were two overlapping barrows (Figure 3.5). In fact the features should be interpreted as an Early Bronze Age barrow with an adjoining, badly preserved stone construction on its eastern side – probably the remains of a cult building. The barrow did not contain the central burial. In its southwestern side four graves were found (Grab A-D), of which grave A provided no finds. It was situated at the foot of the barrow adjacent to grave D of a male, where a bronze sword was found, among other objects. In association with graves D and A situated at the foot of the barrow a stone structure was identified, originally determined as the remains of yet another barrow. It may be a cult building connected with grave D, while grave A, also symbolic, constitutes an element of this structure (Aner et al. 2001: 72-75, Figure 49). Both the graves and the barrows together with the stone construction were dated to the Early Bronze Age. Other objects which may potentially constitute the remains of cult structures in the region of Thy are found at the Egshvile site(Thy 2554), where stone concentrations (Aner et al. 2001: 78-80, Figure 55) were located on the barrow’s eastern side (no. 12) or in Torup (Thy 1697) explored by M. Mikkelsen in 1976, where a construction determined to be a stone circle was also identified on the barrow’s eastern side (12) (Aner et al. 2001: 129-130, Figure 106). Another example may be a box grave situated at the foot of the analyzed Torup Høje barrow in the vicinity of Viborg (Figure 6.42 – Stenkiste). It was a symbolic feature without furnishings adjoining the barrow.

3. Cult houses: definition, idea, chronology

Figure 3.5. Sennels, Thy. Cult structure from the early Bronze Age (bright background) at the foot of the barrow (grey background) (after: Aner at al. 2001, amended by the author).

Volume XII of the work by Aner et al. devoted to the region of Viborg presents our next example from the site of Bjerrevej (VSM 440), where on the barrow’s eastern side a structure described as a concentration of postholes was identified (Aner et al. 2008: 216-219, Figure 130). In fact it was the remains of a stone structure (negatives of the removed stones were interpreted as postholes) – potentially a cult feature. Another example comes from the site of Korsøgard (VSM 712), where a few elements potentially constituting the remains of cult structures were encountered near the barrow (no. 208). A distinctive feature is a fragmentarily preserved furrow (foundation) situated on the barrow’s southeastern part, analogous to the features known from the Late Bronze Age (Aner et al. 2008: 249-252, Figure 143). The above examples of potential cult buildings from the Early Bronze Age fill the gap between the Neolithic features relatively well described in the literature and those dated to the Late Bronze Age. Their chronology is primarily based on stratigraphic premises. None of the published descriptions of cult features mentioned the presence of flint assemblages.

17

18

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia 3.3. Definition The increasing number of structures interpreted as cult houses over the years necessitated attempts at determining differences between the buildings of secular and ritual character. The difference became apparent more or less in the middle of the last century when the number of sources and discoveries of the remains of residential buildings increased. In his work devoted to one of the first examined Neolithic cult structures in Scandinavia – in Tustrup – Poul Kjærum proposed three criteria for defining the building of a ritual character (1955: 7-35). They were recently developed in B. Clemmensen’s publication (2005). According to Poul Kjærum, the ritual character of a building is determined by three criteria: 1. specific location, 2. construction, 3. character of material remains found within the object. The criteria may also be used in identifying cult structures from younger periods of the Bronze Age. Their application in defining cult houses and other structures from the Late Bronze Age will be discussed in more detail below. The structures’ characteristics will be presented in chapter 5. 3.3.1. Location As it was mentioned above, cult houses are usually found in the direct vicinity of graves or cemeteries (Kaul 1987, Victor 2002). This sepulchral context is emphasized by the common location of cult structures at the foot of barrows that predate the later constructions, in the places that were used to deposit cremated burials from the Late Bronze Age (Clemmensen 2005; B. H. Nielsen, Bech 2004). Another important factor is the orientation of the secondary burials and the accompanying cult structures in relation to the cardinal directions. The dominating side of the barrow, where graves and cult houses/features are situated is its eastern

Figure 3.6. Orientation of selected cult objects from the late Bronze Age located at the barrows, including the cult houses from Thy (after: Clemmensen 2005). Figures mark the number of cult objects for individual orientations. Yet, there are examples known form Jutland, e.g. from barrows Høghs Høj and Torup Høje within which the cult zone is located at the southwestern part of the barrow’s base.

3. Cult houses: definition, idea, chronology

side (Figure 3.6). Such an orientation is commonly connected with the phenomenon of sunrise, and thus with the solar cult constituting one of the central premises of the religion from the Bronze Age (Kaul 1998, 1999) as well as with life-giving and resurrecting forces (Turner 2010: 86). I will return to the question of religion later in the chapter. 3.3.2. Construction Despite the abovementioned initial erroneous interpretations of constructions of residential buildings based on the analogy to the cult house from Boda, at present the difference between secular buildings known from the Bronze Age farmsteads (Gröhn 2004: 280-332; M.L.S. Sørensen 2010: 127-135) and cult houses is quite evident (B.H. Nielsen, Bech 2004; Victor 2002). The character of settlement in the Late Bronze Age and the built environment of human settlements are briefly discussed later on in this work. Cult constructions will be described in detail alongside the presentation of the flint material. The difference is basic, to the extent that when looking at a residential building and a cult house from the Bronze Age it is difficult to find many common elements. Residential buildings were elongated three-aisle structures (Figure 4.12), while cult houses are quadrangle or oval forms of unidentified construction (cf. remarks on the construction of the cult house – chapter 10.1). A cult house is a notion which does not necessarily have much in common with the literal understanding of the word house as a construction. It is worth noting, however, that two selected elements clearly distinguish secular buildings from cult houses. In the monograph devoted to Swedish cult houses, Helena Victor points out the significant thickness of the stone walls of the structures and the absence of clearly defined door openings (2002: 65-67, Figure 20, 21)1 (Figure 3.7, 3.8). While so far these features have not been found in the cult objects in Jutland (the analogy may be the considerable thickness of peat walls of the cult houses in Thy), but they constitute a good example of their dissimilarity to secular buildings, whose walls are made from organic materials, and their thickness is not substantial. However, a specific feature of the structures from Jutland is the fact that in several cases semicircular furrow features constituting the remains of the buildings’ separated fragment were recorded inside, usually in their more remote parts (Figure 5.6: B; 6.2: C). They are interpreted as especially important or hallowed places (B.H. Nielsen, Bech 2004, Figure 9, 14; Victor 2002: 183-185). This construction element was observed not only in the cult houses in Thy but also within the so-called cult furrow features (cf. the chapter devoted to cult features) (Clemmensen 2005, Hornstrup 1999).

Victor introduces a classification of cult houses from Sweden into the buildings of stone construction and post structures or smaller cult houses (2002: 66-67), which is not exactly applicable in the case of ritual houses and structures from Jutland.

1

19

20

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 3.7. Hågahagen, Uppland, eastern Sweden. Cult house from the early Bronze Age (II-III period EB). Outlines of a cult house and marked extent of excavations (gray background) and probable locations of the house’s walls (dotted line). Big hearths were located outside both gable ends of the house. Width of stone walls was ca. 2.5 m. The space between the walls was ca. 1.8 m in width. The artefacts were outside the house and in its walls. Outside the object there was a concentration of graves from the early Bronze Age and individual burials from the late Bronze Age (after: Victor 2002). No assemblages of flint artefacts were recorded.

3. Cult houses: definition, idea, chronology

Figure 3.8. Hågahagen, Uppland, eastern Sweden. Reconstruction of a cult house (after: Victor 2002).

The construction of cult houses will be analysed later. It is noteworthy that it differs considerably from the buildings of secular character. Apart from their function, main differences can be seen in their location, outlines and construction, as well as the nature of artefacts found in both types of buildings. 3.3.3. Character of the artefacts According to Poul Kjærum, the artefacts from a cult structure should testify to its specific function, i.e. they should differ form those found in the secular buildings (Kjærum 1955: 20-25). F. Kaul supports this opinion, pointing out that there is no evidence of an everyday type of activity to be found in the structures identified as cult houses (Kaul 1998: 42). Complete absence of artefacts and cultural layers within the confines of cult houses, not so infrequent, is also considered a manifestation of their symbolic character (Clemmensen 2005: 292-294; Victor 2002: 65-67). A perfect example of the original character of the artefacts found in a cult house is constituted by the classic example from Sandagergård in Zealand, where cremated burials and the remains of bronze metallurgy, among others, were found within the confines of the house (cf. the insert Cult house Sandagergård) (Kaul 1987). Most Scandinavian cult houses lack flint assemblages, and structures containing isolated flint artefacts occur only infrequently, e.g. the Ballermosen cult house in Zealand (Lomborg 1956). Presence of dense flint assemblages in the cult structures

21

22

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia from Jutland was recorded relatively late, i.e. in the late 1970s, during the excavations of the Høgs Høj barrow. These assemblages differ from their counterparts from secular structures in terms of their technology, typological composition and the function of the flint artefacts, which will be discussed in more detail in the final pages of this work. The three aspects constituting the characteristics of cult houses and other cult objects referred to above and discussed in more detail later in this study enable verbalising their definition. More on this subject can be found in H. Victor’s monograph of cult houses (2002, 64-66). For the purpose of this work, devoted to one specific aspect of cult houses, i.e. flint assemblages, an adequate definition, similar to that proposed by P. Kjærum, is their co-occurrence with cremated burial graves at the foot of barrows (1), their constructional dissimilarity in relation to the residential buildings of that era (2) and occurrence of sources of other character than domestic within them (3). 3.4. Idea Cult houses and structures of analogous character have from the beginning been perceived as the elements of the cult and ceremonial sphere, remaining in the realm between the activities of everyday life and the spirituality and extramaterial aspects. The concepts, differing only in slight degree, primarily link cult houses with the cult of the dead (Kaul 1987: 43; 1993: 159; 1998: 41-45; 2004: 105-126; 2006), sometimes defined as the ancestral cult (Victor 2002: 41-43) or, less frequently, to sun worship (Clemmensen 2005: 294; Kaul 1999). F. Kaul was the most consistent in developing the concept of the cult house as an element of the religion of the Bronze Age. Drawing on the Sandagergård cult house, which he researched, and the Scandinavian analogies known at the time, he sees an unambiguous manifestation of the cult of the dead in the relation between cult houses and graves (Kaul 1998: 43).2 Considering functions and rites which may have taken place in cult houses, Kaul suggests that the rituals may have adopted the form of initiation and transformation. They may have been connected with the acquisition of knowledge and transformation of their participants from childhood into adulthood. Kaul writes that during the rites taking place in the cult houses, the living, experiencing the ‘presence’ of the dead, acquired knowledge of the past, of the old myths and maintained centuries-old ideologies concerning mythological ancestors. Surrounded by the symbolism of death, young people acquired knowledge of the universe and rites, while at the same time they acquired practical skills. Kaul proposes the above as an explanation for the presence of vestiges of bronze smelting in the cult house – the remains identical to those found in the sites of secular character. In the publication mentioned above F. Kaul lists a number of cult houses, including two from Sweden – Koarum (Arne 1925) and Tofte Högar (Burenhult 1983). The current, complete catalogue of finds is featured in H. Victor’s work (2002).

2

3. Cult houses: definition, idea, chronology

Bronze, of undeniably practical significance during the Bronze Age, also had strong symbolic associations. Initiation of young people and their transformation into adults may have taken place in a cult house and assumed the form of practical training in bronze smelting (Kaul 1998: 44). According to Kaul, bronze razors, frequently decorated with solar representations and representations of ships (Figure 3.9), offered during ceremonies may have performed such initiation functions as the objects of strong symbolic associations. Also flint products, and especially small flakes with sharp edges, might just as well have been used in the transformation rituals.3

Figure 3.9. Solar motifs from bronze razors from the vicinity of Viborg in Jutland. According to Kaul, the position of the sun within the drawing enables locating the motif in the cycle of sun’s passage between the day and night. The motif above refers to an early morning, the motif below – to the noon. The sun is carried by solar horses (Kaul 1998, 262).

Joakhim Goldhahn is even more radical in his interpretations, suggesting that during the Bronze Age smithery and handicraft were exceptionally strongly connected with the material and extramaterial worlds. In his view, the cult house in Sandagergård may have been a place where people met to hold ceremonies and initiation rituals involving all kinds of craft, e.g. smithery. According to Goldhahn, the latter was strongly connected with metaphysical and ritual aspects of life. The gatherings may have accompanied exceptional occasions of death and burial. The role of a blacksmith in the rites approximated that played by a priest. Using fire, he ‘destroyed, constructed and liberated,’ becoming the master of the ceremony of transformation (Goldhahn 2007: 279-314). The rites involving bronze smelting in the case of Sadagergård and other activities in the other cult houses may thus have introduced young people into the world of adults, providing them with awareness of their roots and religion and thus combining the cult of the dead with the initiation of new generations (Kaul 1998: 44-45). In his interpretation of cult houses from Sweden, H. Victor treats the cult of the dead as a way in which social elites manifested their ties with the ancestors (Victor 2002: 181189; 2006), which elevated them in their communities. Cult houses would have been seats of the ancestral spirits and this would account for the absence of artefacts within 3

The opinion of Flemming Kaul from the National Museum in Copenhagen communicated verbally.

23

24

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia the structures – they were not meant for earthly matters. Their specific construction – solid stone foundations, thick walls and absence of the roof – were designed with seats of ancestors in mind, only slightly reminiscent of residential buildings for the living. Absence of clearly visible entrances was to prevent the spirits of the ancestors from entering the world of the living. Significant differences in the construction of houses from Scania and other parts of southern Scandinavia in relation to Jutland are responsible for the fact that H. Victor’s interpretation is hardly justified in the case of the structures from western Denmark. Possibly a chronological factor is at play here – most of the stone constructions thus interpreted are dated to the Early Bronze Age. Cult structures from Jutland constructed in the Late Bronze Age do not display similar features, while their character prompts that they were used solely by the elites. The interpretation of a cult house as the space dedicated to depositing the bodies of the dead before they were buried appears in the publications devoted to the Neolithic examples. According to this concept a cult house constituted a kind of platform, a transitory medium enabling the deceased to transform from death to rebirth. A corpse was left in a cult house and only after some time, e.g. after soft tissues naturally disappeared, was it deposited in a grave (Andersen 2000: 23-24; Jensen 2001: 392-398). Orientation of some cult houses is interpreted as evidence of the solar cult (Clemmensen 2005: 294-295). As it was mentioned above, the Late Bronze Age structures in Jutland tend to be situated on the barrows’ eastern sides (Figure 3.6). It may be assumed that such a location was not a coincidence, taking into account the fact that the sun was one of the central elements of religion during the Bronze Age (Brøndsted 1938; Kaul 2004; Müller 1903). F. Kaul perceives solar motifs from the depictions of ships on bronze razors as manifestations of the concept of the sun’s passage, which is also interpreted as the cycle of life from birth to death (Kaul 1998: 259-263; 1999). The sunrise as the beginning of the cycle was identified with the rebirth of the deceased, where a cult house may have played a significant role (Figure 3.9). Scandinavian cult houses constituted an element of the system of religious beliefs. The rites taking place inside failed to leave archaeological traces that would lend themselves to unambiguous interpretation. The houses display considerable diversity in their construction, which may testify to the diversity of the rites taking place inside. What remains unquestioned is that cult houses and structures were dedicated to ritual practices, distinctly differing from contemporary secular activity. 3.5. Chronology Dating cult structures frequently causes considerable difficulties, resulting from several reasons. As symbolic structures, they hardly ever provide sufficient archaeological contexts lending themselves to precise dating. The Neolithic cult houses mentioned

3. Cult houses: definition, idea, chronology

above, such as the structure in Tustrup, usually contain numerous pottery shards, which enable easy dating of the structures. In the case of the Late Bronze Age, frequently the only basis for dating is the spatial relationship of burials with cult houses. The material from inside cult buildings or features less frequently constitutes the basis for dating, while the instances of determining the absolute age of cult features are absolutely unique (Victor 2006: 114-122). An example of a building of well-established chronology is the Hågahagen cult house from Sweden (Figure 3.7 and Figure 3.8). Excavations revealed its approximate construction and history, together with a wide cultural context of its occurrence supported by numerous radiocarbon dates confirming its functioning between the II and V periods of the Bronze Age (Victor 2001: 153-179). However, this case is unique in the context of the whole geographical range of occurrence of cult houses. The chronology of cult houses from the Late Bronze Age in Jutland has so far been based solely on the graves remaining in spatial relationship with them and on the occasional pottery found in them. In numerous cases the pottery present in the cult houses, as well as the vessels performing the function of grave urns, are so uncharacteristic that they may be very generally related chronologically only to the Late Bronze Age (e.g. grave K49 from Torup Høje, Figure 4.23). Constructions of the graves secondarily deposited in the mounds of barrows are sometimes helpful in determining chronology, as was the case of the Grydehøj cult house accompanying three burials situated in the barrow. While two of them are only roughly dated to the Late Bronze Age, the cremated burial grave in a stone casing displays analogies to the features in Thy, relatively dated to the first half of the Late Bronze Age (B. H. Nielsen, Bech 2004: 134-135). The few pottery shards from the cult house itself were also generally dated back to the IV and V periods of the Bronze Age. Presence of flint assemblages in the structures from Jutland renders flintworking a dating element. If a direct chronological relationship between flint artefacts and a structure is demonstrated, the flint material also testifies to the building’s relative chronology. Technological and, to a lesser extent, typological features of flint implements enable dating them generally back to the Late Bronze Age. Another difficulty preventing precise dating of cult features from the Bronze Age is the flattening of the calibration curve from roughly the second half of the Late Bronze Age, i.e. 700-400 BC (the so-called Hallstatt plateau – cf. the calibration curve graph in the period in question in Figure 9.5 in chapter 9.5). As a result, after calibration a wide range of calendar dates may be assigned to the radiocarbon dates corresponding to this period, which renders them highly imprecise (Harding 2000: 16-17; Kristiansen 1998).

25

26

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia Helena Victor, whose proposed division of cult houses into constructions of stone foundations and smaller post-built features is mainly suitable for Sweden, also perceives their chronological variability. More massive structures were erected in the Early Bronze Age and may have functioned for a long time, until the Late Bronze Age. An example here may again be the Hågahagen cult house.4 Smaller posthole forms were erected later, most probably at the end of the Bronze Age (periods V-VI) and in the early Iron Age (Victor 2002: 146-147). Here, flattening of the calibration curve also prevents more precise dating. A degree of chronological variability is also perceived in Jutland. The cult houses in Thy tend to be associated with the first half of the Bronze Age, while the smaller forms from the vicinity of Viborg and other parts of Jutland are associated with its second half. However, these findings require thorough verification during field studies, and especially an acquisition of new data in the form of absolute chronology of structures, at least those from the older part of the Late Bronze Age. An attempt at chronological correlation – with the use of radiocarbon dating of the material from a cult house – was made in the case of the Grydehøj example. The results are presented in the chapter devoted to the multifaceted analyses carried out for this structure (chapter 9).

A difference in the object’s function during the Bronze Age was recorded. In the early period the house’s interior remained beyond human activity and only in the Late Bronze Age do artefacts appear in its interior, which testifies to the fact that the space around and inside the cult house was treated differently with the passage of time (Victor 2002: 147).

4

4. Specification of source information 4.1. Current state of the research This monograph examines flint assemblages discovered in association with Late Bronze Age cult features in northern Jutland. The state of research for this topic may be summarized, and at the same time, exhausted, with a reference to the title of the author’s lecture delivered in April 2012 at a session titled: On the research of the late flintworking in central Europe1 organised by professor Jacek Lech. The paper devoted to the subject of this monograph was titled: Unknown flintworking… because until detailed studies of the relationship between cult features and flint assemblages were undertaken, only a few people directly involved in the research of cult features containing small collections of flint artefacts were aware of the importance (Masojć, Bech 2011), with the exception of one notable mention in the literature, that in addition to pottery, flint artefacts were also found inside the Grydehøj cult house (B. H. Nielsen, Bech 2004: 139). In principle, this scenario is true for the whole of Scandinavia. Absence of information pertaining to flint assemblages in cult features certainly results from the current state of research rather than from the actual absence of flint artefacts. In the only monograph on cult houses in Scandinavia published so far, H. Victor (2002: 66-67) observes that the cult houses excavated in Sweden are characterised by the absence of a cultural layer with only a small number of finds (isolated pottery shards, bone fragments, and the occasional flint artefacts). The author does not specify whether the finds that she mentions come from the older or younger period of the Bronze Age. The flint artefacts from the cult houses in her sample include Early Bronze Age daggers and other isolated and unspecified artefacts, such as those from the Broby cult house, Börje (Victor 2002: 108-111) or the Småland cult feature (Victor 2001: 142-144). The cult house in Håga in Uppland, systematically examined by Victor and described in detail in her work, failed to yield any flint artefacts (Victor 2002: 153-179). However, recent investigations of rock art in Sweden resulted in interesting discoveries. The places featuring stone carvings appear to be the locations where rituals were practised, rituals in which stone and flint materials played a substantial role (Bengtsson. Ling 2007; Ling, Ragnesten 2009; Bengtsson 2010). On the Sotenäset peninsula, in the vicinity of Torp, sites with numerous examples of rock art have been correlated by radiocarbon dating with the occupation from 1500-300 BC. The presence of a cultural layer around and between panels with rock art was the reason two of the many sites present in the area were selected for excavation: Tossene Raä 446: 2-3 and Tossene Raä 446: 1-2 (Bengtsson, Ling 2007; Ling Ragnesten 2009). On the research of flintworking in central Europe, II Scientific Conference of the Autonomous Section of Prehistoric Flint Mining, Institute of Archaeology and Ethnology of the Polish Academy of Sciences, Warsaw 26 April 2012; the full title of the lecture was: Mirosław Masojć, Unknown flintworking. Scandinavian cult houses from the Bronze Age.

1

27

28

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 4.1. Tossene Raä, site 446:2-3, Sweden. A crevice in a stone panel with rock carvings incorporating vertically pushed flint artefacts, among others (after: Ling, Ragnesten 2009).

At the former, also dated to the Late Bronze Age, two rock panels with representations of humans and what have been identified as cup-marks were partially covered by the cultural layer. In the space between the panels, which measured approximately one metre wide, there were two layers of stone paving together with a concentration of burnt stones. Below the paving a large quantity of flint flakes (the authors do not specify their character) and pottery fragments were found. Around the panels and in the space between them there were also quartzite hammerstones and several kilograms of flint flakes, the majority of which were burnt. The carvings themselves were surrounded by rock fissures measuring up to 0.5 metre in width and depth (Figure 4.1). The fissures also contained fire-cracked stone, quartzite, lumps of construction daub, pottery, and flint artefacts, some of which could be refitted. Some of the flakes were pressed into the narrow bottoms of the fissures. In one of the niches the layer containing artefacts was covered by a layer of compacted clay (Bengtsson, Ling 2007: 41-44). At a nearby site two hearths containing flint artefacts and pottery were discovered in the space between two rock panels (Figure 4.2: A and B). The authors interpret the finds discovered around and within the rock panels with carvings as deposits related to ritual

4. Specification of source information

Figure 4.2. Tossene Raä, site 63:1-2, Sweden. Cultural layer (A) containing numerous flint artefacts, partly covering a rock panel with a carving of a human figure in an acrobatic pose and the hearth between the panels (B) (after: Bengtsson, Ling 2007).

activities. They do not specify whether the ritual activities took place because of the presence of the rock carvings or if it was the other way around (Bengtsson, Ling 2007: 45-47). The deposits in the fissures near the rock panels seem especially interesting, as they played an important role in the rituals, sacrifices or celebrations taking place in the vicinity of the rock carvings. The research in Tossene Raä shows that flint artefacts were a fixed element of the rituals. On the basis of the excavations of a spacious cultural centre in the vicinity of Enköping in Uppland in Sweden, L. Karlenby formulated an interesting concept of stone as the symbolic element of the universe (2011). In addition to the numerous cult features, rock carvings and graves, there are piles of considerably burnt pebbles and flint artefacts (unspecified). The presence of pebbles was also recorded in other cult houses in Sweden, where they are thought to have been used in the ritual pulverising and crushing of cremated osteological remains. However, the work does not provide a more detailed description of the morphological characteristics or function of the flint artefacts (Karlenby 2011).

29

30

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 4.3. An erratic boulder in site Machary, north-western Poland (after: Gralak, Wyszyńska-Gralak 2010). Flint artefacts from the late Bronze Age discovered during the excavations around the boulder (drawing by the author).

The results of excavations carried out around boulder erratics found at two sites – Machary and Żabicko in north-western Poland (Gralak, Wyszyńska-Gralak 2010) may provide an analogy for the examples of occurrence of flint artefacts in the abovementioned symbolic contexts. Flint artefacts characteristic of the Lusatian culture were found around both boulders (Figure 4.3: photograph). A larger assemblage from Żabicko, comprising nearly fifty artefacts, provided amorphous, lump-like rock blocks displaying randomly distributed flake scars. There is no evidence of preparation treatment (Figure 4.3: 1). Stone pebbles were also found, displaying evidence of strong blows, resulting in angular shatter. With the exception of a tool resembling a sidescraper with abrupt, blunted retouch on the longer edge of a chunk (Figure 4.3: 2), the remaining tools are opportunistic forms. The manner of their production, with no attention paid to the form, suggests that they were meant for immediate purposes. Two perforators were made with the use of irregular retouch. One of these has a relatively well-defined point on chunk, which was microretouched (Figure 4.3: 3). Evidence of human activity around boulder erratics is usually associated with the sphere of spiritual culture, i.e. with beliefs and magic (Woźny 2008). Erratics, constituting prominent landmarks on the landscape, may have been convenient places for various kinds of gatherings and ceremonies, during which flint artefacts may have been used for any number of purposes. Interestingly, flint deposits in the form of assemblages of large backed knives were placed in the vicinity of similar erratics in southern Scandinavia in the Late Bronze Age (cf. chapter 4.2.3). To summarise, we must conclude that the relationship between flintworking and Late Bronze Age cult features is a so-far unknown phenomenon, in principle lacking any

4. Specification of source information

history of research. The examples presented above show that it offers a great, multifaceted potential for research, exceeding the issues discussed in this work. The history of the research of flintworking during that period is relatively long (Högberg 2009; Lech, Piotrowska 1997). Cult features, or more precisely, cult houses from Scandinavia, similarly have a long history of research (Kaul 1987; B.H. Nielsen, Bech 2004; Victor 2002). To provide the context for the topic of this monograph, I will try to characterise both phenomena, so far studied completely independent of each other. Because topics treated separately do not constitute the sensu stricte immediate subject of this work and have already been extensively researched and discussed, I will only provide an overview. 4.2. Flint assemblages of the Late Bronze Age – contexts of occurrence At the end of the Bronze Age flint was still in universal use in Scandinavia. The first publications devoted to late flint artefacts found in association with excavated settlements (Müller 1919) or bog deposits (Blinkenberg 1898) are more or less from the last century. The phenomenon has probably been better researched and is understood to a greater extent in Scandinavia than elsewhere (Eriksen 2010; Högberg 2009, 2010). Unlike in other parts of Europe, there have never been any doubts concerning the relationship between flint artefacts and such late settlement from the period between the end of the Bronze Age and the beginning of the Iron Age (Gedl 1975: 63-64), which in Poland was overcome with the spectacular discovery of a Lusatian culture flint pit, workshops and homogenous flint assemblages (Lech 1997b: 11-12). Despite that, Danish and Swedish researchers admit that this aspect of Late Bronze Age culture was neglected, and substantial collections of flint artefacts were largely ignored. Berit Valentin Eriksen (2010: 81) called flintworking from the Bronze Age in Denmark the ‘Cinderella of flintworking research’ referring to a phrase from an article by A. van Gijn and M. Niekus (2001). This attitude is partly explained, after Henrik Thrane (1985: 146), by Anders Högberg who, in reference to large backed knives, writes that flint artefacts from the Late Bronze Age – as lithic artefacts – were traditionally stored by the museums together with the collections from the Stone Age, where in turn they were ignored and forgotten because they had no connection with latter (2009: 49-51). While undergoing the process of technological deterioration, at the same time Late Bronze Age flintworking assumes a surprisingly unified character, observed in many areas in Europe (Ballin 2010; Bronowicki, Masojć 2010; Edmonds 1995; Eriksen 2010, Högberg 2004; Lech, Piotrowska 1997). In brief, it is characterised by the predominant use of a hard hammerstone and the exclusive exploitation of flake cores, mainly amorphic or multidirectional and devoid of preparation. This type of flaking produced large, largely cortical flakes with a significant amount of chunks from broken up cores. Blades, whose manufacture is unintentional, occur occasionally. Production of ad hoc tools increased, and the use of the sharp edges of flakes was widespread as many classes of formal tools were clearly abandoned (Masojć, Bech 2011: 204).

31

32

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Yet, such a picture is incomplete and oversimplified, as Late Bronze Age flintworking assumes another form and is characterised by a distinct technological dualism. Its other aspect is an exceptional invention of that time – large backed knives, which during the Late Bronze Age became widespread in northern Europe. They had many practical applications and were the basic tool used in harvesting. According to A. Högberg they are not an evolutionary stage in the development of harvesting tools, but replace sickles from the Early Bronze Age made with the use of pressure technology (Högberg 2010). In central Europe, analogous, multifunctional backed knives are made on flakes (Bronowicki, Masojć 2010), Figure 6), even though large blade forms similar to large backed knives from Scandinavia also occur (Kruk 2005; Lech, Lech 1984). However, it is the context of the occurrence of Late Bronze Age flintworking that is a decisively more complex issue than its technology. Even though the treatment of the flint raw material had become considerably simplified, its use in various aspects of human activity were not limited. In this respect its diversity equals that of the preceding periods. In southern Scandinavia the use of flint at the end of the Bronze Age occurs in several different cultural contexts. The utilitarian use of flint artefacts in settlement sites is easily recognizable in the archaeological record. Its common occurrence in settlements testifies to the complex system of the acquisition of flint raw material, its adaptation to the users’ needs and its distribution and use. The system’s complexity is intensified by the symbolic aspect of flint’s occurrence as exemplified by flint votive deposits or the presence of flint artefacts in graves (Ebbesen 1981). When considering their presence within barrows and cult structures surrounding them, from the archaeological point of view it does not seem possible to separate their practical and symbolic significance (cf. comments by Warren, Neighbour 2004: 92-93). Despite ad hoc technology and obvious lack of aesthetic, the Late Bronze Age flintworking displays a complex nature. It was still vital for the communities from the Late Bronze Age and the early Iron Age, both in its utilitarian and extra-utilitarian aspects. To summarize, presence of flint artefacts in southern Scandinavia during the Late Bronze Age may be seen in its practical and symbolic aspects in four different cultural contexts (1-4). The practical use of flint is especially well represented in settlement sites (1), as well as in the places where the raw material was acquired and in workshops (2). The symbolic use of flint at the end of the Bronze Age is represented by stone votive deposits and the inclusion of flint materials in graves (3). Cult houses and other cult features from northern Jutland, together with large flint assemblages produced there, constitute another, so far unknown, aspect of the presence of this material in the symbolic sphere (4). Moreover, as will be seen in the following chapters of this work, cult houses and other cult features manifest the ritual activities

4. Specification of source information

involving its use. Before considering the latter, I would like to discuss in greater detail the first three aspects mentioned above by providing a few examples. When possible, the discussion will include the material remains that I have analyzed, most of which has not been previously published. 4.2.1. Settlement A number of settlements dating to the end of the Bronze Age where the largescale use of flint was common have been recorded in southern Scandinavia. So far only a small amount of this material has been published, most has only been briefly mentioned, such as that from Fragtrup (Draiby 1984) or Voldtofte (Berglund 1982). Swedish examples of extensively researched settlement sites include Lockarp 7E, a site excavated as part of the Öresundsförbindelsen project2 (Högberg 2001), and Fosie IV (Björhem, Säfvestad 1993), while Danish examples include the study of flint artefacts from homesteads in Krogstrup in Jutland (Olesen, Eriksen 2007) and the analyses of flint from Jutland sites excavated as part of the research project examining the Bronze Age settlement around Viborg (Christiansen1996). The most comprehensive study of Late Bronze Age flintworking in Scandinavia to date is the previously mentioned work by A. Högberg (2007), even though it is concerned with only one aspect of this phenomenon – large backed knives. This study also includes a detailed history of the research of flintworking during the Late Bronze Age in southern Scandinavia. 4.2.1.1. General picture, on the basis of Thy Settlement in southern Scandinavia during the Late Bronze Age, as well as during its subsequent stages, was characterized byindividual homesteads consisting of one elongated residential building of post construction with three aisles, usually accompanied by accessory structures. The homesteads, scattered across the countryside, were the basic settlement units. The distance between them was at least several hundred metres, although occasionally it was smaller. The settlement from Bjerre Enge in Thy provides an example of a cluster of at least two to three homesteads in an area covering less than one square kilometre. In the case of Thy a pattern of two homesteads per square kilometre on the average was adopted, which would amount to over one thousand inhabitants in an area measuring 50 km2, assuming that each homestead included 12 residents (Artursson 2010, 95-98). The site in Bjerre will be one of the sites highlighted in the discussion of flint assemblages later in this work. The intensive settlement in Thy differs substantially from the strongly hierarchical settlement in Funen, where during the Late Bronze Age there were rich hamlets connected with the emerging strong centre of authority (Kristiansen 1998, 175-170). At that time the spacious, richly furnished barrows of elite appear, which represent a direct manifestation of power (Thrane 1984, The construction of the bridge over the Sound strait connecting Zealand with Scania, i.e. indirectly Copenhagen with Malmö.

2

33

34

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

1993). The flintworking from one of the settlements in Funen will be discussed in the following chapters of this work. To an extent, the dispersion of homesteads across the Thy landscape reflects the distribution of funeral rites. Thousands of barrows appeared on the hills and other exposed places during the Early Bronze Age. They differ in their sizes and grave furnishings, which testifies to the hierarchical organisation of the society. This hierarchy is no longer visible during the Late Bronze Age. While in general terms the settlement pattern does not change - no new barrows appear, and the dead are buried in the barrow mounds in the proximity of their homesteads. Small secondary burials occur in the barrows, which probably reflect a social group or a family (Earle, Kolb 2010: 61-67). Except for the house sizes and manner of burying the dead, the differences in the utilization of the surrounding space remain quite insignificant throughout the Bronze Age. The homesteads were self-sufficient farms. The houses functioned for a certain period of time, estimated to be 30-60 years, i.e. one to two generations. They remained unchanged, even though traces of repair, extensions or other modifications are occasionally observed (M. L. S. Sørensen 2010: 128-129). The houses, usually with two entrances on the longer walls, had their own internal, functional division. In general terms, one part of the house was utilized for domestic purposes, while the remaining ones were a barn and a cowshed. Inside the houses were hearths and cooking pits. Ad hoc flintworking took place in the homesteads. Farming tools made of flint were produced both outside and inside the houses. Early Bronze Age homesteads may have housed flintworking workshops, producing specialised tools – asymmetrical sickles. An example of a workshop from the Early Bronze Age where sickles were produced on a large scale was documented at the site in Vilhøj, in the east of Thy (P. Mikkelsen 1996). No traces of specialised flintworking workshops have been encountered inside the houses dated to the Late Bronze Age. Neither has any evidence of the large scale production of an ‘export’ product of the Late Bronze Age – large backed knives – been found. They were probably produced beyond the settlement zones, in the workshops in the vicinity of the areas where the flint raw material was acquired, which will be discussed in more detail in the next chapter. On the other hand, ad hoc production, sometimes on a relatively large scale, took place within the homesteads, which is exemplified by Bjerre mentioned above. It is quite possible that such opportunistic production, however on a greater scale, also took place at a certain distance from the houses themselves in order not to litter the residential areas with flint debris. The remains of hammerstones, cores and blanks provide undeniable evidence of flintworking satisfying immediate needs of a homestead.

4. Specification of source information

Difficulties in the unequivocal interpretation of the nature of presence of flint aretefacts in the houses result from the specific character of the built environment within homesteads. As it was mentioned above, the use-life of houses is estimated to have been several dozen years. When the period of a house’s use came to an end, a new building was erected at a certain distance from or in the immediate vicinity of the old one. The old house was probably demolished; the posts supporting the roof were taken out of the ground and were probably used to construct the new building (M. L. S. Sørensen 2010: 152). If we assume after Artursson (2010: 97) that during the Bronze Age a single homestead may have had more than thirty consecutive episodes of new building construction, whose locations over time frequently overlapped, then determining whether the flint artefacts from a given homestead were produced within a building or in its surroundings carries a substantial margin of error. A flint assemblage produced several hundred metres from a house so that the debitage would not constitute a problem for the residents may at some stage have been located in the immediate environs of a subsequent house. Other incidences cannot be ruled out, e.g. clearing the debitage by removing it from a house or its vicinity and depositing it at a certain distance from it, or the opposite situation – clearing the area for the construction of a new house and depositing the flint flakes found there at the location of an abandoned house. Such a complicated spatial situation also applies to the sites in Bjerre, where due to exceptionally attractive geomorphological conditions favouring the development of agriculture and animal husbandry in the Bronze Age, an area of approximately one square kilometre accommodated two or three homesteads at the same time (Bech 1991; 1997; Bech et al. to be published). The number of flint artefacts from the site in Bjerre, probably from various stages of the use of individual homesteads, amounts to almost thirty thousand (Eriksen 2010, Table 1). A large number of flakes and the accompanying tools results from ad hoc flint production that was not performed by specialised craftsmen. It is also quite possible that the backed knives from Bjerre, small and to an extent lacking standardisation, were made directly in the settlements, following similar ad hoc flint production processes. (Sofaer 2010: 210-211). Exploration of the remains of individual buildings brings about the discoveries of less numerous flint assemblages; however, the instances of their absence are as frequently recorded. This might support the argument that flint production took place outside the residential buildings. The examples and occurrence of flint artefacts in a sample of Late Bronze Age settlements which I had the opportunity to familiarise myself with while studying the flint assemblages from cult features, are presented later in this work. The characteristics presented below are by no means intended to be exhaustive, but aim at introducing the nature of flintworking in individual contexts where evidence of it is found for the

35

36

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

comparison to the flint production evidenced within cult houses and other features of a similar nature. 4.2.1.2. Bjerre, Thy The agglomeration of sites dated to the Bronze Age in the vicinity of Bjerre Eng in northern Jutland in Thy already mentioned above constitutes an exception on a supra regional scale (Bech et al. to be published). In the 1990s research was carried out in a number of settlements; initially they were rescue excavations – preceding a failed motorway construction project – and later as part of the Thy Archaeological Project – the cooperation of the museum in Thisted with an international research team (Bech 1993; 1997; 2003; Bech, Mikkelsen 1999; Earle et al. 1998; Earle, Kristiansen 2010a). Archaeological deposits from Early and Late Bronze Age contexts in Bjerre yielded well-preserved remains of residential buildings inhabited for over 700 years (1500800 BC). Due to exceptional geomorphological conditions, the cultural layer was preserved – evidence of their intensive and prolonged use (Figure 4.4). The remains of the settlements, houses, farm outbuildings and the cultural layer were covered by Aeolian deposits, which to a great extent helped to preserve the archaeological remains, which included thousands of flint artefacts (Eriksen 2010). Apart from the predominant farms from the II-III periods of the Bronze Age, there is also site Bjerre 7 - a settlement from the V period of the Bronze Age that includes the remains of a residential building and an accompanying zone of economic activity (Olsen to be published). The material from this site became the subject of my analysis. The settlement is situated in the eastern part of the Bjerre region on a slightly elevated landform. The level of the buildings was ca. 5.4 m above sea level, covered by a more than 1-metre thick layer of Aeolian sands. The Aeolian sedimentation is interstratified with biogenic horizons. Out of 7 identified horizons, two deposited directly over the settlement from the Late Bronze Age are dated by thermoluminescence to the end of the Late Bronze Age and the beginning of the Iron Age (horizon 6) and the late phase of the Late Bronze Age (horizon 7 deposited below). According to Anne-Lousie H. Olsen, who excavated site 7 in Bjerre, transitory settlement episodes from the Early Bronze Age were followed by permanent settlement during the Late Bronze Age, which resulted in the formation of the cultural layer. In the final stages of the Late Bronze Age a residential building was erected in the same place, in the area that had earlier been used as an arable field (evidence of intensive ploughing – Figure 4.4). The building did not last until the end of the Bronze Age, and the area was once again transformed into an arable field. Soil cultivation lasted for some time in unfavourable environmental conditions and ended at the end of the Bronze Age and the beginning of the Iron Age with the beginning of intensive dune

4. Specification of source information

Figure 4.4. Bjerre 7, Thy. Trench 76 during excavations. Aeolian sedimentation is seen above the cultural layer from the late Bronze Age (photograph by J.-H. Bech).

Figure 4.5. Bjerre 7, Thy. Eastern part of trench 76. Traces of ploughing in the ceiling part of the cultural layer, under which were the remains of a residential building (photograph by A-L. H. Olsen).

37

38

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

accumulation, interrupted twice by the development of vegetation (at the end of the La Tene period and again during the early Middle Ages). There were no subsequent settlement attempts within the site boundaries. B. V. Eriksen (2010), who is currently analysing the flintworking from Bjerre, is working on a monograph of the settlement there, which will also include the study of all seven flint assemblages (Bjerre 1-7) (Eriksen to be published). The flint assemblages from Bjerre are for the most part dated to the Early Bronze Age; only the Bjerre 7 assemblage is dated to the end of the Bronze Age. Site Bjerre 7, very well preserved with the cultural layer and a complete set of movable sources, testifies to the large scale and common use of flint raw material in the Late Bronze Age (the site provided over 13 000 flint artefacts), thus showing how significant it was for the contemporary economy. The assemblage is homogenous, and it is one of the most complete collections from the Late Bronze Age in Denmark. Because no publications have been dedicated to it (not in 2008 and not until now), the assemblage will be briefly described for the purpose of this study. The flint artefacts preserved from site Bjerre 7 represent two different technological traditions. The predominant tradition is flintworking based on simple means of acquiring blanks. The blanks and retouched flakes are of a distinctly ad hoc nature. The production took place directly at the site and was of a local character. Simple products are accompanied by the technologically advanced production of large backed knives. Despite Joanna Sofaer’s opinion quoted above (2010: 210-211) I did not encounter any evidence of their production in the Bjerre 7 assemblage. Most probably they were made by specialised producers and provided to the farm. This generalised picture of dualism seen in the flintworking is analogous to other Scandinavian sites from the Late Bronze Age (Högberg 2004). The flint raw material used at the site came from erratic surface material commonly available in Thy. It cannot be ruled out that to some extent the remains of the rocks abandoned in the mines from the Neolithic Funnel Beaker culture were also used. It is quite possible that the abandoned mines were to some extent still used in the Early Bronze Age. In the case of Bjerre 7 there is lack of clear evidence supporting this assumption (Earle, Kristiansen 2010c: 219; Eriksen 2010: 87-89). A summary of the assemblage consisting of more than 13,000 artefacts is presented in Table 4.1.3 The flint material from the cultural layer was explored at several strata corresponding to stratigraphic layers. The assemblage consists mainly of debitage and debris (94.66%). Cores represent only 1.5% of the collection, while the remaining 4% are tools. The characteristics of the assemblage results from the specific technological The summary of the assemblage was also presented by B. V. Eriksen (2010, Table 1). There are slight quantitative differences between both presentations.

3

4. Specification of source information Table 4.1. Bjerre 7, Thy. Structure of flint assemblage

Horizon 1 Horizon 2 Horizon 3 I: Precore forms and cores

52

II: Other flakes flakes blades other forms – from cores’ edges flakes from core platform splinter flake chunks chips III: Tools backed knives end-scrapers side-scrapers perforators and borers notched tools ad hoc tools-retouched flakes or blades fragments of unidentified tools strike-a-light-like products hammerstones total

52

45

Total

Horizon 4 36

N

%

185

1,37 94,66 79,84 0,22 0,1 0,02 13,91 0,57 3,97 0,24 0,7 0,37 0,01 1,99 0,08 0,08 0,5 100

3882 8 5 644 7

1804 14 7 1 3 360 37

2805 6 2 476 14

2225 2 386 19

13 14 5

6 21 2

3 11 2

83 3 31

11 49 1 41 2 28 12 6

88 8 16

66 15

12707 10716 30 14 1 3 1866 77 537 33 95 1 50 2 265 12 11 68

4747

2428

3489

2765

13429

character typical of the flintworking from the Late Bronze Age. Exploitation of cores, breaking them up solely with the use of a hard hammerstone (0.5% of the assemblage) resulted in the production of large quantities of flake blanks, usually cortical. Such a treatment of the raw material resulted in breaking the blocks and production of numerous chunks, which made up the second most numerous category of artefacts present in the collection. They were treated equally with flakes, which is manifested by the frequent use of chunks as ad hoc tools. Rare blades were not intentionally produced, but are a result of the specific shapes of reduced flint blocks. Core reduction constituted here the only way of acquiring blanks. Absence of bipolar technique is noticeable. The cores from Bjerre 7, with the exception of prevalent amorphic forms with single flake removal scars, carry traces of multidirectional flake removal (Plate 1: 1-2). Striking platforms were frequently changed; a preceding flaked surface was often selected as a consecutive striking platform, which at the same was accompanied by a

39

40

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 4.6. Bjerre 7, Thy. Core with changing orientation (drawing by the author).

complete absence of preparation and repair (this is also confirmed by a scanty number of other forms). An example of a relatively well organised and advanced reduction is a large core, which after the changing of the orientation provided a number of flakes as a result of a circular reduction of the flaked surface (Figure 4.6). The last of the core’s platforms has a circular character. Some of its part displays deep flake scars indicative of the use of a hard hammerstone. The core carried no traces of preparation. It was abandoned during its reduction, which potentially may have been continued upon utilization of the last flaked surface as a striking platform. Continuing the reduction in this way would have given the core a discoidal shape. Quite frequently the cores in Bjerre 7 were discarded at this stage (Plate 1: 2). Cortex and partially cortical flakes or chunks were used in tool production. The variability of tool categories from Bjerre 7 is very limited, consisting of several classes of formal tools (including knives and end-scrapers) and the predominant group of ad hoc tools – irregularly retouched flakes or flake blanks carrying the traces of use. They

4. Specification of source information

Figure 4.7. Bjerre 7, Thy. Backed knife (type C) (drawing by the author).

include knife-like flakes, which usually have a cortical backed edge and considerably worn working edge (Plate 2: 1, 2). The group of formal tools is dominated by end-scrapers (Plate 3: 1-4) and perforators. Even though they are made on irregular flake blanks described above and on cortical flakes or chunks these groups of tools display a certain degree of standardisation. In the case of end-scrapers it is manifested by the care in making their working edges, which in their majority are rounded. The perforators have distinct and carefully retouched tips. Some of them were made with the use of chipping analogous to that of the Clactonian industry. Backed knives constitute a separate category of tools (Figure 4.7). Made on blades or large flakes whose proportions were modified to those typical of blades, they differ

41

42

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Table 4.2. Bjerre 7, Thy. Sizes of backed knives according to types A-D. L – length; W – width; T – thickness; C – presence of cortex; U – presence of macroscopically visible traces of use Type Number A 15 B 1 C 11 D 6

L max 101 0 93 92

L min 46 58 64

L med. 71,7 75 69,6 77,5

W max 46 35 42

W min 23 24 22

W med. 32,66 33 30,18 31,66

T med. 13,8 19 11 13,8

C% 100 100 36 66,6

U% 93,3 100 100 100

technologically from the rest of the assemblage. It is almost certain that they were not made at the site. The manner of their production differed from reduction practised at the site. The specific character of lithic reduction practised at the site prevented the acquisition of flake blanks appropriate for that purpose. In comparison with the published assemblages from eastern Denmark and Scania, the backed knives from Bjerre 7 differ in their proportions (Table 4.2). They are slightly smaller and do not fit the size brackets of the types proposed by Björhem and Säfvestad (1993). For the purposes of this study, the backed knives can be subdivided into two types, those with a natural backed edge and those with a retouched backed edge (type A, C – cf. chapter 7.2) (Figure 7.4 and 7.5, Plate 2: 3-6). A few of the backed knives, with both natural and modified backed edges, exhibit additional retouch of the working edge’s base (Type B, D) (Figure 7.4 and 7.5). The average size of the knives from Bjerre 7 does not exceed 8 cm in length and 3.5 cm in width. The shape of the knives’ backed edges is often fragmentary. Only in some cases was the modification applied to the entire backed edge. Among the knives of the C type as many as 36% have the natural surface preserved on the backed edge – possibly constituting the evidence for hafting. Nearly all the knives from Bjerre 7 have very distinct macroscopic traces of use on the edges, i.e. various types of chippings, crumbling and – less frequently – wear. Some display wear of the base of the backed edge, which might be the evidence of hafting. Apart from their sizes, the remaining features of the knives from Bjerre 7 do not differ from those of the backed knives from the Late Bronze Age from other regions of southern Scandinavia (Högberg 2009). Poor condition of the artefacts, damaged by the postdepositional processes, prevented a more comprehensive functional study at site no. 7. Helle Juel Jensen (in print) examined in more detail two categories of artefacts – 22 massive backed knives and 4 strike-a-lights. According to the author, the traces of use-wear on the backed knives indicate that they were used to cut stalks of corn (12 specimens) and reed (2 specimens). The pattern of

4. Specification of source information

wear and linear traces indicate a parallel or oblique direction of cutting, while presence of wear at the backed edges did not result from cutting but is possibly the evidence of continuous wiping of the knife’s back against the growing stalks and ears during harvest. Similar traces of use and the pattern of wear was present on the massive backed knives from Spjald (Becker 1990) and on a flint knife from a sickle discovered in the Stenild peat bog (cf. the insert The sickle from the Stenild peat bog). Six knives from Spjald were interpreted as tools used to cut reeds, while the knife in a wooden handle from Stenild displayed the traces indicating that it was used to cut corn ears ( Juel Jensen 1990, 1994). The knives from Bjerre 7, similar to those from the two sites mentioned above, display wear along two edges, of which one was active (cutting), while the other was passive (the tool’s back). Of the seven completely preserved knives from Bjerre, as many as six carried the traces of hafting. Their similarity to the microtraces seen on the knife from Stenild suggests an analogous manner of mounting in a wooden handle ( Juel Jensen 1990). Some fragments of knives display no wear, i.e. no evidence of cutting, which resulted from securing them in their shaft. However, these parts which do not carry the traces of cutting display wear resulting from mounting them in their handle. The traces of cutting run along the working edge and the backed edge towards the top, which suggests that the handle covered a large part of the backed edge, in the same way as it did in the case of the sickle from Stenild. In its upper part the sickle has a characteristic widening which enabled fixing an insert and ensuring a stable position of the blade during work. The similarity of the traces of fastening to the knife from Spjald suggests that the manner of fastening known from the sickle from Stenild (Blinkenberg 1898) was common in the Late Bronze Age, at least in northern Jutland. According to H. Juel Jensen (in print), the multifunctionality of the massive backed knives from Bjerre 7 is substantiated by the presence of macro crumblings – a result of cutting hard material – on those knives which were proved to have been used in harvest. 4.2.1.3. Bulbjerg Troldsting, Thy The site Bulbjerg Troldsting situated in the northern part of Thy is one of the four sites featured in one of the first monographs devoted to Bronze Age settlement in Denmark written by Sophus Müller and published shortly after the end of World War I (1919). The site, situated at the coast of the North Sea on top of Bulbjerg, a calcareous moraine hill, was excavated by Carl Neergaard in 1913. Apart from pottery and numerous flint artefacts discovered in the partly destroyed cultural layer, it also yielded numerous unprocessed lumps of amber, the remains of bronze metallurgy in the form of moulds and lumps of bronze, as well as numerous animal remains. This settlement was dated to the Late Bronze Age and mainly within period V (Jansen 1965: 48). Lack of the documented remains of post-built houses, and presence of numerous features, including post holes, may be explained by the fact that the excavations were carried out long before any remains of residential developments were discovered in Denmark (Bech 1997: 3).

43

44

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 4.8. Bulbjerg Troldsting, Thy. Backed knives in the museum in Copenhagen (photograph by the author).

Sophus Müller’s publication was one of the first to present an illustration of a massive backed knife – a typical flint artefact from the Late Bronze Age – as well as an extensive description of the associated flint assemblage. (Müller 1919, Figure 17). Müller is absolutely positive that the flint assemblage is related to the settlement from the Late Bronze Age. He emphasises that Troldsting constitutes important evidence of the common use of flint in that period, which must have had a great impact on how the phenomenon of flintworking at the end of the Bronze Age and the beginning of the Iron Age in Denmark was perceived, considering that together with Oskar Mentelius, Müller was the central personality of Nordic archaeology (Gräslund 1987). Müller mentions a collection of more than seven thousand flint artefacts recovered from the cultural layer and outside of it. Apart from the debitage resulting from core reduction, Table 4.3. Bulbjerg Troldting, Thy. Sizes of backed knives according to types A-D. L – length; W – width; T – thickness; C – presence of cortex; U – presence of macroscopically visible traces of use Type A B C D

Number 15 3 40 15

L max 120 120 119 112

L min 84 70 89 86

L med. 104 95 93,6 96,9

W max 35 33 40 43

W min 23 30 25 25

W med. 29 31 31,3 32,1

T med. 10,6 13,3 11,3 14

C% 66,6 100 47,5 33,3

U% 80 100 65 60

4. Specification of source information

the collection mainly included numerous backed knives, end-scrapers and perforators (Müller 1919: 66-67). The large assemblage described by Müller has survived in a very incomplete form housed in Ørholm by the National Museum in Copenhagen. Most of the preserved artefacts (stored in only five boxes) are backed knives (Figure 4.8) with a few other tools, mainly end-scrapers and perforators. The knives from Troldsting are made from Scandinavian Senonian flint (Tables 4 and 5). The examples with retouched (or partially retouched) backed edges predominate. Similar to the knives from Bjerre 7, they are smaller than the specimens from Zealand or Scania. Their average length approximates 10 cm, while average width is 3 cm. Many exhibit distinct traces of use along the edges. Their proportions, consistently smaller than those of the knives from the Danish islands, suggest their certain regional, specific character typical of Jutland (Tables II and III). 4.2.1.4. Fragtrup, Vesthimmerland Situated three kilometres to the west of Farsø in Jutland, the site Fragtrup was the first settlement in Scandinavia where the remains of three-aisled long houses were discovered. The results of excavations carried out at the site in the 1950s and the 1960s by the National Museum in Copenhagen (among others) were only recently partly published (Draiby 1987). The remains of the houses from Fragtrup were identified as residential structures from the Late Bronze Age, which in the subsequent decades were commonly discovered in Denmark and the whole of Scandinavia (Müller-Willie 1977). Inside they had two rows of posts supporting the roof and walls of post construction with rounded gables (Figure 4.9). Their exceptional character in the case of Fragtrup was the relationship of the two contemporaneous housessituated next to each other (I and II) as well as the preservation of the entire site; similar to the case of Bjerre, the structures were preserved below the Aeolian deposits. In addition to the well-preserved remains of the two houses (both burnt down), the remains of other homesteads were also excavated (III and IV). The collection of the artefacts from the site is quite impressive, including numerous flint artefacts. On the basis of the pottery the chronology was determined to be the V-VI period of the Bronze Age; however, a great majority of fragments from the documented stratigraphic units was dated to the V period of the Bronze Age. Absolute radiocarbon dating determined the time of the settlement as the period from the mid-9th century BC (650±75 BP, K-3795) till the mid-6th century BC (407±75 BP, K-3802) (Draiby 1987). The artefacts from Fragtrup are kept in the National Museum in Copenhagen. The flint collection includes nearly seven hundred artefacts (Plate 6)4 from individual 4

The size of the collection accessible to the author differs slightly from the one published by B. Draiby (1987: 205).

45

46

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 4.9. Fragtrup. Plan of the site with residential objects (after: Draiby 1984).

Figure 4.10. Fragtrup. House I. Backed knife (drawing by the author).

4. Specification of source information Table 4.4. Fragtrup. Structure of flint assemblage I: Precore forms and cores II: Other flakes flake blade chunk chip III: Tools backed knives end-scrapers perforators and borers notched tools ad hoc tools – flakes or retouched blades fragments of unidentified tools strike-a-light total

House I 7 127 102 1 19 5 16 2 5 1 8 150

House II 14 390 348 4 38 44 5 19 2 16 1 1 448

House IV 3 57 36 3 18 12 6 3 3 72

houses (I – III). The remains of house III were the least discernible at the excavated site. This context also provided the smallest number of flint artefacts. The artefacts were made from the erratic flint available locally. The quantity of the artefacts is much lower than that of the assemblages from Bjerre or Troldsting, which may result from the better condition in which the latter were preserved. The flint assemblages from Fragtrup display typical features of flint production from the Late Bronze Age. The technological profile is based on ad hoc strategies of the production of blanks, with only the use of a hard hammerstone. The cores lacking preparation are usually circularly reduced with frequent changes of the striking platform (Plate 6: 4). The forms produced ad hoc with irregularly retouched or unmodified edges predominate among the tools (Plate 6: 5, 7-9). Formal tools were reduced solely to end-scrapers and backed knives (Table 4.4). Backed knives – individual items – were present in each homestead (Figure 4.10; Plate 6: 10). The knives were accompanied by end-scrapers – also made ad hoc on relatively irregular cortical flakes, while their worked edges are limited to a small segment of the edges of the flakes (Plate 6: 1-3). An interesting element of the assemblages from the houses in Fragtrup is a knife-like tool made on flakes and chunks (Figure 4.11). These are ad hoc tools made on debitage produced with the use of opportunistic methods. They usually have natural/cortical backed edges, while modifications occur less frequently. The opposite edge, performing the function of a cutting edge, carried macroscopic traces of use. Each of the three assemblages included numerous hammerstones.5 5

The hammerstones are not presented in tables.

47

48

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 4.11. Fragtrup. House II. Flake with natural backed edge and point. Considerably worn-out (drawing by the author).

4.2.1.5. Højby, Funen Højby, situated in Funen several kilometres south of Odense, is an example of a settlement from the V period of the Bronze Age; however, it is not preserved as well as Bjerre or Troldsting (Hertz 1987, Figure 565). Rescue excavations carried out in the 1980s ahead of road construction concentrated on the remains of a three-aisled residential building measuring 15 x 12 x 22 m, typical of the Late Bronze Age (Figure 4.12). The excavations yielded 364 flint artefacts from within the confines of the house. Despite the fragmentary character of the collection, resulting from the poor preservation of the feature and the absence of a cultural layer, the assemblage contains types that are both technologically and typologically representative of flintworking from the Late Bronze Age. The collection is dominated by irregular flakes (138 specimens) and chunks (83 specimens) from the frequent breaking of flint blocks with a hard hammerstone. There are only seven formal tools, which include end-scrapers (Plate 7: 1-2) and a single sidescraper (Plate 7: 3). There also are two artefacts resembling strike-a-lights (Plate 7: 7, 8). The remaining tools are irregularly retouched flakes and chunks, as well as artefacts carrying the traces of use without secondary modification (16 specimens – Plate 7: 4-6, 9-10). The collection is complete with a large hammerstone. There are only four cores in the assemblage, two unidirectional flake cores (Plate 7: 11, 13) and two nonprepared disc cores (Plate 7: 12, 14). Over 23% of the assemblage, as it happens only blanks and other flakes, carry the evidence of the use of fire, which might indicate that there was a hearth in the house. The flint material from Højby did not include large backed knives, present in many other settlement sites.

4. Specification of source information

Figure 4.12. Højby. Reconstruction of a three-aisled house located in Odense, close to its original location and the drawing of its original outlines (photograph by the author, drawing after: Odense Bys Museum).

4.2.1.6. Skamlebæk, Zealand The excavations at the site Skamlebæk on the north-western coast of Zealand resulted from a large project constructing the infrastructure of a radio station, covering nearly a square kilometre. The preliminary work was carried out in agreement with the National Museum in Copenhagen by an engineer directly involved with the construction work. The site, excavated by Søren Gregersen, transpired to be a large settlement from the Late Bronze Age, where the remains of residential buildings from the Bronze Age were discovered for the first time in Denmark outside Jutland (Gregersen 1960; 19631964). This led to exceptional interest in the site resulting in the excavations carried out from 1976-1978 by the National Museum in Copenhagen (Lomborg 1977, 1979). In addition to artefacts from the Bronze Age, the site also provided evidence of a Neolithic settlement, but the excavated area (ca. 40 acres) primarily revealed a clearly visible portion of a settlement from the Late Bronze Age. The features, which escaped damage from agricultural activities, were preserved below the cultural layer, whose thickness in places exceeded 30 cm. The layer was filled with pottery, flint products and occasional bronze artefacts. The remains of a residential building, analogous to those already known from Jutland, 16 m long and 6 m wide, displayed clear evidence of the use of fire. The feature included two rows of post holes and a hearth located in the house’s central part. The entrances to the building were probably situated in the house’s longer walls. The artefacts – the pottery and flint materials – also displayed the traces of fire. To the east of the building there were postholes – possibly the remains of yet another building, while to the southwest there was a row of small postholes – possibly the remains of a fence. According to E. Lomborg, at least a part of the building was decorated with paintings, which is substantiated by the remains of clay covering the building’s walls that contained the traces of a white die (Lomborg 1979: 9).

49

50

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 4.13. Skamlebæk, Zealand. Backed knives (photograph by the author).

Two wells and a furnace for firing pottery were situated around the building. Lomborg concluded that their functions differed depending on their location in relation to the building (Lomborg 1977: 127). Small, flat hearths and other cooking pits were situated near the building. Further away from it features of different functions appear, the largest and the deepest of which carry the evidence of the use of fire (burnt layers, burnt pebbles). These pits were situated at the greatest distance from the house – over 40 m. The artefacts, both from the layer and the features, were dominated by pottery dated to the IV-V periods of the Bronze Age. The flintworking from the settlement also corresponded to the end of the Bronze Age (Lomborg 1977, Figure 4), including a small number of Neolithic artefacts and artefacts from the Early Bronze Age (Lomborg 1977, Figure 5). Apart from the assemblages from the Bjerre site, the assemblage from Skamlebæk – counting many thousands of artefacts – constitutes the largest collection of the artefacts from the Late Bronze Age in Denmark. It is now stored in the National Museum in Copenhagen (Ørholm, Lyngby).

4. Specification of source information Table 4.5. Skamlebæk, Zealand. Sizes of backed knives according to types A-D. L – length; W – width; T – thickness; C – presence of cortex; U – presence of macroscopically visible traces of use Type Number A 2 B C 8 D 4

L max 129 132 -

L min 91 103 -

L med. 110 118 -

W max 37 44 39

W min 36 33 32

W med. 36,5 39,1 35,7

T med. 20 18,1 16,2

C% 100 25 25

U% 100 87 100

The assemblage from Skamlebæk exhibits typical features of the flint industry of its time – the end of the Bronze Age and the beginning of the Iron Age. Its characteristic element is the presence of large backed knives (Figure 4.13). These measure between ten and twenty centimetres in length. Their backed edges were made with the use of abrupt blunted retouch (Plate 8: 1). However, their course is either irregular, partly natural or consists of natural backed edge with retouch on the ventral face (Plate 8: 3, 4). It might also exhibit the form of a knife’s base, comprising its ventral face and shaping it in the form best lending itself to being hafted (Plate 8: 2). The standardised backed knives are accompanied by ad hoc tools, including large wide flake-like blades carrying the traces of use (Plate 9: 3, 4). The cores present in the assemblage testify to the manner in which flake blanks were produced, with the use of either discoidal reduction with a prepared platform (Plate 9: 1, 2) or single-platform reduction with a crude or prepared platform to make large flake blanks with the sole use of a hard hammerstone. I did not record in the assemblage the presence of cores used in the production of blanks suitable for making such large backed knives. The latter were probably produced outside the settlement in specialised workshops. Despite clear similarities to contemporary assemblages, the collection from Skamlebæk differs from that from the settlement in Bjerre. The knives from Skamlebæk are decisively larger than the knives from Thy. As may be seen in the table, the average sizes of knives of particular types considerably exceed the respective measurements of the knives from Bjerre (Table 4.5). 4.2.1.7. Vinde Helsinge, Zealand Vinde Helsinge Hals (also known as Helsinge Kirke) is a Late Bronze Age site where characteristic flint artefacts were commonly present. The site also features a thick cultural layer covering pit features. Vinde Helsinge is an archival site. The store rooms of the National Museum in Copenhagen house a small collection accompanied by a handwritten note by A. Søndergaard stating that he had handed over the materials from his 1935 excavations

51

52

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 4.14. Vinde Helsinge, Zealand. Backed knives (photograph by the author).

to the museum in Copenhagen.6 According to Søndergaard’s description, the cultural layer present below the arable layer was from 0.4 to 1 metre thick. The excavations comprised only a few square metres (between 5 and 8). In this area and below the cultural layer two hearths and two storage pits were excavated. Both the cultural layer and the excavated features contained artefacts, including pottery and flint. The museum’s inventory includes a general description of the artefacts that it received,7 mentioning two fragments of a bronze sword, pottery originally coming from at least four vessels, faunal remains and a dozen different flint artefacts – tools, cores, and blanks. In his note Søndergaard wrote that he had only worked for 3-4 days at the site, but that in his opinion it deserves systematic excavations. Despite its archival character, the small collection of artefacts and the obvious imperfection of the amateure nature of the work and the records made, the acquired Payment for the artefacts to their owner or the owner of the field where the excavations took place – K. Rasmussen – 20 Danish crowns (in 2015 this was an equivalent of 3 Euros). 7 Inventory numbers 618/39 and 685/39. 6

4. Specification of source information Table 4.6. Vinde Helsinge, Zealand. Sizes of backed knives according to types A-D. L – length; W – width; T – thickness; C – presence of cortex; U – presence of macroscopically visible traces of use Type Number L max A 5 136 B C 5 159 D 2 107

L min 100 108 102

L med. 119,4 129,6 104

W max W min W med. 39 27 34 39 30 34,4 40 35 37,5

T med. 15 18,1 20

C% 40 40 0

U% 100 100 100

flint artefacts deserve a closer look, because there are very few examples of sites from the Late Bronze Age where the cultural layer protected the artefacts deposited within and beneath it. If the description is to be believed, Vinde Helsinge’s potential of preserving the remains of a settlement might equal that of the settlements in Bjerre or Skamlebæk. The museum still houses a dozen or so of flint artefact from Vinde Helsinge. The collection is incomplete, containing only the tools (Figure 4.14; Plate 10, 11): 12 knives and 4 end-scrapers. The knives from Vinde Helsinge are large forms made solely from single-platform cores. The largest measures 16 cm in length (Plate 10: 1). Their proportions, similar to those of the artefacts from Skamlebæk, clearly exceed the proportions of the backed knives from Bjerre. In the case of Vinde Helsinge the average length of knives ranges from 10.4 to nearly 13 cm depending on the type of knife, while the average length for all the knives would be 11.7 cm (Table 4.6). The knives from Bjerre occasionally exceed 10.0 cm, while their average length does not exceed 8.0 cm. Average width and thickness of the knives from both sites and the radio station in Skamlebæk suggest that large backed knives from Zealand are decisively larger forms than analogous products known from the sites in Jutland. In comparison to the knives from Bjerre they display wear relatively more frequently (both on the edges/cutting edge and on the other sides, which probably evidences the fastening in their mounts) apart from other traces of use always present on the edge. All types of knives occur, including natural flat blades and blades with a natural cortical backed edge (Plate 11:2), forms with continuous abrupt retouch of the backed edge (Plate 11: 1), and forms which apart from the retouch of the backed edge were additionally retouched on the sides (Plate 11: 4). One of the end-scrapers is reminiscent of a large backed knife (Plate 10: 2), while another is an ordinary, large form with a rounded worked edge (Plate 10: 3). 4.2.1.8. Voldtofte, Funen A thick cultural layer covering the remains of outbuildings was also preserved at another site – Voldtofte (Kirkebjerg). The excavations carried out in the 1970s yielded over two thousand flint artefacts displaying features typical of flintworking from the Late Bronze Age (Berglund 1982: 60; Figure 11).

53

54

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 4.15. Voldtofte, Funen (Kirkebjerg). Extent of the cultural layer and location of the trenches (after: Berglund 1982).

4. Specification of source information Voldtofte is primarily known for a find from the beginning of the 20th century, when a feature with five large vessels standing next to each other and containing grains of corn was discovered during the excavations carried out in a small area ( Jensen 1967, Figure 3; Thrane 1980, Figure 2). Barley proved to be the main type of grain found both in the vessels and during the subsequent excavations of the cultural layer (in 1976-1978) (Berglund 1982, Table on page 62).

Figure 4.16. Voldtofte, Funen (Kirkebjerg). Thickness of the cultural layer located above the objects from the late Bronze Age (photograph from the archives of the Odense Bys Museum).

Systematic excavations, although only covering small areas, were carried out in the subsequent years (1909-1911, 1915-1916), and the finds were first made public by S. Müller in the first monograph devoted to the Bronze Age in Denmark mentioned above (Müller 1919). The spatial arrangement of the whole site was clarified by the excavations carried out in the 1970s. Covering more than a hectare, they provided a picture of the extent of the cultural layer in the area still free from the 20th century settlement (Figure 4.15). The thickness of the cultural layer varied between 15 cm to over 1.5 m (Figure 4.16). It contained charred grains of corn, burnt pebbles, pottery, flint artefacts, lumps of clay, pieces of charcoal and bronze artefacts. The fragments of pottery and flint artefacts were mostly concentrated in the upper levels of the cultural layer, decreasing dramatically below a depth of 40 cm and disappearing completely at the layer’s bottom. The layer’s thickness decreased towards the north-east, which in itself shows that the layer itself constitutes the remains of intensive settlement situated beyond the excavated area, further to the west. Nearly 100 pit features were excavated into the natural subsoil, most of them contained traces of burning with numerous pieces of charcoal and burnt rock pebbles. Pottery and flint artefacts were found only in the upper layers of the pit features, which was interpreted as their secondary position following the original deposition on the living surface. These pit features were identified as furnaces possibly used for drying grain.

55

56

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

A group of post-hole features and numerous lumps of clay present in the excavated area below the cultural layer were interpreted as the remains of a residential building with a clay floor and walls; however, the excavations failed to confirm unequivocally that they were the remains of a dwelling. The pottery from the site, represented by a dozen thousand sherds, are mostly from large storage vessels and to a lesser extent from tableware – more finely made and more richly decorated, mainly from the V and VI periods of the Bronze Age. The fragments of clay mentioned above displayed the remnants of decorations in the form of white, brown and black colouring of the flat surfaces of clay/plaster, which might suggest that the walls of the residential buildings in Voldtofte were partly painted (Berglund 1982: 59). An identical way of decorating walls was observed in the remains of buildings in Skamlebæk in Zealand. Moving on to flint artefacts constituting the main object of our interest here, it should be emphasised that their greatest frequency, corresponding to the vertical distribution of the pottery, was recorded in the upper part of the cultural layer; their presence in the lowest part of the layer was only very occasional, similar to the upper part of the pit features excavated into the subsoil. Horizontally, the flints were present across the cultural layer. The assemblage from Voldtofte consists of more than 2300 flint artefacts. The general characteristics do not differ substantially from other known settlement assemblages from the Late Bronze Age. Berglund writes that the flintworking in Kirkebjerg lacked subtlety. It consisted in fragmenting flint blocks by hitting them against other hard rocks. The consequence of such a treatment of the raw material as well as the use of a hard hammerstone to break up concretions is that a considerable part of the assemblage consists of flint chunks which do not display formal traces of intentional reduction (impact point, bulb of percussion etc.). Eighteen artefacts interpreted as blades and 29 pieces carrying the traces of flake removal were identified (Berglund 1982: 60). Because Kirkebjerg is one of the few Late Bronze Age sites where more than one sentence was devoted to flint artefacts, I relied on Berglund’s published description, with only the occasional reference of the flint collection to confirm the description’s accuracy. The collection’s limited number of retouched artefacts is dominated by massive end-scrapers on flakes with rounded worked edges (34 specimens, including two end-scrapers on blades according to Berglund), which occasionally display additional retouch of the ridge (Plate 12: 1-3). Four perforators on flakes have clearly discernible unifacial or bifacial points. I did not manage to find the three large backed knives, mentioned by Berglund. A photograph of one of them included in the publication seems to confirm that it really was a backed knife, yet in terms of its size it is more closely related to the knives from Bjerre than those from Skamlebæk or Vinde Helsinge (Berglund 1982, Figure 11). The remaining tools are dominated by ad hoc forms, including a few artefacts resembling strike-a-lights. Flintworking also utilised quite large quartzite pebbles, clearly chipped on their perimeters. A dozen or so of these are present in the assemblage.

4. Specification of source information

The prevalent chunks and less frequent blanks were accompanied by cores (29 specimens), which included irregular, amorphic cores ‘typical’ of the flakes produced when the object pieces were broken up as well as unidirectional cores with prepared striking platforms (Plate 12: 4-6). The latter prove that the producer occasionally tried to shape the cores to make the product of their reduction more predictable. The settlement of Kirkebjerg from the Younger Bronze Age has a number of specific features indicating that it was a regional centre of authority at that time in Funen in the vicinity of Voldtofte. Its best visible manifestation were 40-metre-wide barrows of the ruling elites situated in the settlement’s vicinity (Thrane 1984). The importance of the site is also substantiated by the thick and rich cultural layer, painted decorations on the walls of the residential buildings, an exceptional grain storage and a surprising composition of the excavated animal specimens. In the collection of over 3300 bone fragments ca. 82% of those identified belonged to cattle and 11% to pigs (Berglund 1982, Table on page 62). In the Late Bronze Age sheep and pigs predominated among the domesticated animals in the settlements from southern Scandinavia, and the undeniably greater contribution of cattle renders Voldtofte an exceptional place. According to K. Kristiansen this fact indicates that it was a centre of authority, where exceptional burials of the elite in the barrows are accompanied by deposits with a predominant contribution of the osteological remains of cattle – prestigious domestic animals at that time. The monopoly of cattle breeding is a manifestation of the dominance of the centre of authority over the local economy (Kristiansen 1998: 179-180). The cultural layer and the features at Kirkebjerg yielded a relatively small number of bronze artefacts (about 30 small fragments). The ubiquitous presence of flintworking in the settlement constituting a centre of civilisation of the Younger Bronze Age shows that flint still played a basic role in the economy, not only for the communities of the Danish islands or northern Jutland, but also for the main and richest cultural centres of the age. This is also indirectly substantiated by the presence of flint artefacts in the Lusehøj barrows of dukes in Voldtofte (Thrane 1984: 37, 108). 4.2.2. Places of acquisition of the raw material and flintworking workshops During the Late Bronze Age in southern Scandinavia flint was acquired locally. Flint assemblages do not show any evidence that raw material was traded over long distances. The last glaciation covered the greater part of the discussed area with moraine deposits containing high quality erratic flint. Blocks of flint found their way to the farms from the moraine deposits situated nearby or even from the fields covered with erratic material. The beaches with their redeposited flint were also a very good source of the raw material (Eriksen 2010; Högberg, Olausson 2007). Ad hoc flintworking took place within the homesteads. Whether it happened at a distance from the residential buildings or in their immediate vicinity still remains an

57

58

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

issue to be resolved, though different individual scenarios are possible (cf. the comments in 4.2.1). The ubiquitous presence of flint artefacts associated with settlement from the Late Bronze Age testifies to the common use of this raw material, still a significant element of the economy of that time, which is especially clear in well preserved remains of the homesteads. The nature of flintworking of that time shows that it was not a domain of specialised craftsmen. An opportunistic technology and lack of concern for forms, aimed solely at preparing implements suitable for short-term purposes with minimum effort shows that the skills of the producers were absolutely basic. To quote B. V. Eriksen, the production of such flake tools could have been accomplished by ‘anyone, anywhere and at any time’ (Eriksen 2010: 89). The only category of artefacts exceeding the limitations of ad hoc flintworking were the large backed knives. These were multi-purpose knives made on large flakes or blades, produced outside the settlements in specialised flintworking workshops, reaching the homesteads as ready-made, finished implements. There they were fastened in their settings in a such a manner as that exemplified by the sickle from Stenild (the insert The Sickle from the Stenild peatbog). Functional analyses showed repetitiveness in fastening the backed knives in their organic settings ( Juel Jensen 2006). It may be assumed that it was a common way of producing sickles during the Late Bronze Age in southern Scandinavia. The producers of the knives must have had better skills than the farmers producing flakes and tools for their own, everyday needs. However, the skills indispensable for becoming successful mass producers of large backed knives were clearly poorer in comparison to the mastery required for the production of bifacial asymmetric sickles and blades from the Early Bronze Age. It is, however, possible that the production of backed knives might have also taken place within the homesteads with the use of ad hoc methods, as was observed in the Bjerre 7 settlement (Soafer 2010: 210-211). This possibility requires additional study. The multifaceted characteristics of large backed knives was presented by A. Högberg (2009). His publication also characterised for the first time the workshops where the knives were produced on a large scale (2009: 196-215). They were situated close to the abundant deposits of flint. There are no mine sites in Scandinavia from the Late Bronze Age. It is possible that to a certain degree the raw material available in the mines from earlier periods was used. This may have been an option in the region of Thy, where Early Bronze Age local communities, including the site of Bjerre, were able to utilise mining spoil deposits (containing refuse in the form of blocks, chunks, cores and massive blanks left by

4. Specification of source information

Neolithic miners and producers). However, the presence of mining has not yet been proved. All the more so, it did not take place in the Late Bronze Age. A similar phenomenon was observed in the large Neolithic mines in Scania in Sweden in the region of Södra Sallerup, where numerous flint artefacts from the Late Bronze Age occur in thick cultural layers despite the lack of any evidence of mining exploitation of rock material. The mines are connected with the mass production of backed knives from Scandinavian Senonian flint and include large flakes used in the production of knives as well as fragments of knives and their semi-finished products (Högberg 2009: 202). In addition to the workshop from Södra Sallerup, A. Högberg mentions the sites Tommerstrup in Zealand, Hanaskog in Sweden and two places in Rügen in northern Germany: Reidervitz and Ruschvitz (2009, 196-215), which make a complete list of places known to date where large backed knives were produced. Considering how common these tools were at sites from the Late Bronze Age and assuming that they were usually produced in specialist workshops, the list should be much longer. Wierzbica ‘Zele’ in central Poland is an exceptional site on the European scale, where mining of flint and production of backed knives on the spot were observed. (Lech, Lech 1984; 1997a; Lech et al. 2011). The deposits of chocolate flint, also known as ‘Zele’, were exploited here. So far 81 mineshafts have been identified. They were usually wide and open, reaching the depth of 7m. Radiocarbon dating showed that the oldest shafts (e.g. shaft no. 17) were exploited during the Early Bronze Age, while the largest of the dated shafts (no. 19 and 28) were exploited during the Late Bronze Age (Lech et al. 2011, Figure 5) (Figure 4.18). The Late Bronze Age miners looked for large concretions of flint lending themselves to the production of blanks and tools of substantial sizes. The most characteristic backed knives of the ‘Zele’ type produced at the mine were accompanied by large cores carrying the traces of the removal of large flakes and blades. To an extent, the knives of the ‘Zele’ type are similar to the southern Scandinavian large backed knives. Mining in the wilderness of ‘Zele’ in Wierzbica ended during the Early Iron Age. Interestingly, according to the authors of the research in Wierzbica, mining and flintworking at the mine did not necessarily have to be motivated solely by economic factors. The authors point out that sinking such deep shafts ‘to reach flint-bearing horizons appropriately sanctified by the custom’, which required considerable organisational effort, may have been motivated by ritual factors. Acquiring blocks of flint from considerable depths, which required great effort, may have been a symbolic activity, while producing from them a very limited number of implements might testify to their ritual purpose (Lech, Lech 1997b: 111-112). The analyses of the function of the flint artefacts from the ‘Zele’ mining fields considering their potential ritual use did not confirm the above hypothesis (Małecka-Kukawka, Werra 2011: 141-143). The authors of the analyses observed that short-lived rituals may have failed to leave any

59

60

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

traces on the flint, which has also been confirmed by the examination of the assemblages constituting the subject of this monograph. Quite importantly, the authors also point out that the context of the discovery of flint artefacts may be crucial in recognising them as ‘the products of special purpose’ (Małecka-Kukawka, Werra 2011: 143). These conclusions are fully confirmed by Scandinavian flint assemblages from ritual zones. The issue of function and potential ritual activities carried out in these zones is discussed in a separate chapter of this work (cf. chapter 8). 4.2.2.1. Knudshoved, Funen Site Knudshoved is an example of a flint workshop, similar to the ones mentioned earlier. The site is situated in the commune of Nyborg, about 30km from Odense, on Funen’s east coast, which is abundant in flint rubble (Masojć, Bech 2011: 208-210). A nearby cliff contains layers of good quality, slate flint. The flint deposited on the beach includes concretions of considerable sizes (Figure 4.19). Artefacts occur at a section measuring 1 km in length, including large cores with clearly visible flake removal scars, large flakes, fragments of semi-finished backed knives and other flakes (Figure 4.20). A considerable collection gathered from the site’s surface is kept in the museum in Odense, including numerous cores and several hundred specimens of blanks: large flakes, blades and semi-finished large backed knives. The material is considerably

Figure 4.17. Knudshoved in Funen. Mass material from a flint workshop. Big flakes and other semi-finished backed knives. The collection of the Odense Bys Museum (photograph by the author).

4. Specification of source information

Figure 4.18. Wierzbica ‘Zele’ in central Poland. Shaft no. 19. Scale in the shaft of 3 metres in length (after: Lech et al. 2011).

Figure 4.19. Knudshoved in Funen. A – View of the site and rock rubble on the seashore at the foot of a cliff; B – original location of flint raw material within the gradually washed out cliff; C – flint blanks in the collection of the Odense Bys Museum (photograph by the author).

patinated (Plate 13; Figure 4.17 and 4.19; C). Site Knudshoved is the only workshop in Funen known so far where backed knives were produced during the Late Bronze Age for the needs of local farms.

61

62

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 4.20. Knudshoved in Funen. Flint blanks lying on the beach. Massive cores and flakes are seen among the rubble (photograph by the author).

4.2.2.2. Stagstrup, Thy So far no flint workshops specialising in the production of backed knives have been found in Jutland. However, one of the sites in Thy, unfortunately only superficially excavated, provided a collection of flint artefacts that are likely related to such a workshop (Masojć, Bech 2011: 208-211) Stagstrup8 is a site known only from the material collected from the surface. The collection consists of over a hundred characteristic backed knives from the Late Bronze Age (Figure 4.21). It does not include any artefacts that might suggest that they were produced at the site, i.e. characteristic blanks and semi-finished products. This may be explained by the fact that the collection is quite random, collected by the owner of the local fields and an enthusiastic amateur interested in the prehistory of the region. Yet, the number of artefacts is puzzling. Despite the fact that the artefacts were collected from the surface, the number is several times greater than the number of the knives discovered elsewhere, e.g. the well preserved site Bjerre 7. 8

It is known locally as Mølle Bakke. The artefacts are in Aksel Kristensen’s private collection.

4. Specification of source information

Figure 4.21. Stagstrup, Thy. Big backed knives from a private collection, possibly the remains of a specialised flint workshop located in the settlement (photograph by the author).

Small probing excavations were carried out at the site, revealing the existence of a cultural layer from the Late Bronze Age.9 Even though the site has been excavated only very cursorily, which prevents any serious conclusions concerning its character, two interpretations seem quite probable. It could be assumed that the site was a place where knives produced in a nearby workshop of an unknown location, were stored and later delivered to the local farms. It would thus have been a homestead specialising in storing and distributing ready-made products in the form of sickle inserts. This would suggest that the system of production of backed knives was not based only on a simple relationship between a flint workshop and settlements, but that the distribution network also included the existence of middlemen. A less far-fetched hypothesis would be the assumption that a workshop specialising in the production of backed knives in Stagstrup functioned within a residential 9

Jens-Henrik Bech’s verbal communication.

63

64

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

homestead. I have already mentioned that during the Early Bronze Age, residential homesteads specialising in the production of asymmetric bifacial sickles have been identified on the cliff ’s edge in Ås/Villhøj situated no more than 10 km away (P. Mikkelsen 1996: 113), where very numerous semi-finished products and fragments of sickles, accompanied by debitage, were found. The location of the site in Stagstrup is similar – on the edge of a cliff over a fjord, close to abundant deposits of good quality flint. This geographical similarity might not be coincidental. There are premises for adopting the assumption that it might be the first workshop of this type within a Late Bronze Age homestead. Summarising the discussion of workshops and places where the raw material was acquired, it should be assumed that there must have been many places where backed knives were produced. There may have been a few of them functioning in a given settlement area, providing a sufficient number of finished products to meet the needs of the local population. Their presence is a proof of the existence of a complex system of production and transportation of the raw material, semi-finished and finished products from workshops to settlements. The manufacture of knives required advanced flintworking skills. Stagstrup may be the first known workshop possibly operating within a residential homestead. Basic flintworking, meeting everyday needs, took place in the settlements. There are no known instances of the mining of flint deposits in southern Scandinavia during the Late Bronze Age. 4.2.3. Flint deposits Votive Bronze Age deposits constitute a very satisfying and attractive theme in the prehistory of Scandinavia. They are often very spectacular, as are the so-called treasures or hoards that usually contain metal implements and prestigious objects. These include intentionally deposited helmets, swords, lurs, daggers or jewellery as a manifestation of religious beliefs and a consequence of ritual activities. However, even though the deposits and treasures are one of the most frequently discussed topics by the researchers of the Bronze Age, according to A. F. Harding its ideological and religious background is still relatively unknown (2000: 354). The most spectacular find of this type is probably the so-called sun chariot from Trundholm, found more than one hundred years ago during agricultural work in a bog in north-western Zealand (Kaul 1998, 3034). The chariot, on display in the National Museum in Copenhagen, is the pride of the exhibition devoted to the prehistory of Denmark (Figure 4.22). One aspect of the issue of deposits, known much less than the metal artefacts, is the similar deposition of flint objects. The works devoted to the religion and deposits from the Bronze Age in principle ignore this category of finds (Levy 1982, M. L. S. Sørensen 1989). While working with Late Bronze Age artefacts in Danish museums, I had no opportunity to study flint deposits as they are not commonly found. Only two have been discovered in Jutland. The best known is the deposit of knives from Spjald

4. Specification of source information

Figure 4.22. Sun chariot from Trundholm in Zealand. Dated to the early Bronze Age (1700-1300 BC). In the collection of the National Museum of Denmark in Copenhagen (photograph – courtesy of the Museum’s archives).

(Becker 1972, 1990; Juel Jensen 1990). Individual deposits have also been found in Poland (Kruk 1994, 2005). The deposit from Spjald in Jutland is one of the best studied finds so far of this type (Becker 1990, Figure 5a, 5b). The context of its discovery differs from the other examples. It was found during the excavations of a post hole of a house from the Late Bronze Age at a site where there were remains of residential buildings from both the Late Bronze Age and the Iron Age. The deposit consisted of six large backed knives accompanied by a stone trapeze and several other artefacts, including fragments of petrified resin, a fragment of a wooden vessel and pottery. According to C. J. Becker the collection did not represent a single act of deposition. During the time of the building’s occupation various objects were placed several times into a hole where there was a post supporting the roof (Becker 1990). An analysis of microtraces present on the knives proved that they were used to cut reed (Juel Jensen 1990). They also displayed the traces of setting, identical to those identified on the knife from Stenild (Juel Jensen in print). In his monograph devoted to Scandinavian backed knives A. Högberg presents a comprehensive analysis of other known flint deposits (2009: 172-196). He discusses more than twenty assemblages similar to the deposit from Spjald from eastern Denmark and southern Sweden as well as one discovery from Norway (2009, Figure 2.104; Table 2.63). It is intriguing that nearly one half of the deposits were discovered prior to the beginning of the 20th century and others before 1950, which adversely affected the documentation. For the same reason the collections are frequently incomplete. No other bog deposits were found mainly because exploitation of peat was abandoned. Only two deposits came from excavations; the rest were found by coincidence during various kinds of earthworks. The data presented by Högberg show that the deposits usually contain several flint artefacts (2 -10 specimens), although some included several dozen of them – the maximum being 60 as in the case of the Gamla Köpstad site in Sweden. The largest ritual deposit found in Denmark from Dalegård in Bornholm

65

66

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

contains 20 flint artefacts (Ebbesen 1981; F. O. Nielsen 2005). It was discovered during the second half of the 19th century as a deposit of knives and semi-finished knives placed next to each other near a huge erratic stone. Comparing the contexts of discoveries of the deposits A. Högberg concludes that contrary to what might be assumed, flint deposits were not usually placed in peat bogs or bogs (2009, Table 2.69). They were most frequently deposited in the mineral context (8 deposits) and near large boulders, e.g. the case of Dalegård. Only two deposits came from the marsh environment, while the deposit in the posthole feature from Spjald in Jutland remains unique. The arrangement of knives in the deposits suggests that they were probably wrapped up or bundled together (e.g. with a piece of leather, sack, or leather strap) and placed more or less vertically; therefore, some researchers concluded they were stuck into the feature. Because they were placed so close to each other, complete sickles, including flint inserts, could not be deposited. Even though many of the knives exhibited traces of use, the object of the deposit was the flint knife itself and not a composite tool, i.e. a sickle (Högberg 2009: 193-194). A recently published deposit of flint artefacts from north-western Poland, from the island of Wolin – site no. 8 ‘Wzgórze Młynówka’ – is closely related geographically with the Scandinavian ones (Kruk 2005). There are premises for interpreting it as resulting from not only utilitarian strategies. Although the find is relatively new – from the 1980s – in its character it is similar to the deposits from the north. It was found by chance in a wall of a gravel pit. There is no information concerning mutual spatial relationships between the individual elements of the deposit. Karol Kruk relates it to the Lusatian culture, as the remains of a settlement from this culture were also found in the gravel pit (2005: 210-213). Thus the deposit comes from a mineral context. It may have been in a feature that was part of the settlement. The deposit includes nine flint artefacts – three backed knives and six blanks of with blade-like proportions. Some of the blades have retouched edges, and according to the author, they also display utilitarian retouch, including traces of wear resulting from harvesting activities. All the specimens were made from erratic Baltic flint. The site is dated from the V period of the Bronze Age to the C-D Hallstatt period. K. Kruk reports another flint deposit found in Poland, i.e. the deposit from site Rudnik 10 near Stalowa Wola (1994, 2005). In my opinion it would be difficult to unambiguously determine its character. The deposit was found in a settlement within a feature and may have just been a collection of flint artefacts used in everyday activities. Another issue concerning the deposits is that of determining their character, i.e. determining unambiguously whether they constitute ritual or non-ritual manifestations of human activity. It would be an oversimplification to assume that in all instances

4. Specification of source information

the discovered deposits, especially those found in bogs, are obvious remains of ritual activities. In her work devoted to the ritual deposits from the Bronze Age, Janet E. Levy proposed criteria for determining whether a deposit has a religious association or not (1982: 17-44). Drawing mainly on ethnographic analogies from the communities of diverse cultural and geographical backgrounds, she observes that the nature of a deposit is determined by such factors as its contents, arrangement and context. The deposits found in the mineral environment, shallowly deposited and without any specific arrangement would be, according to the criteria adopted by the author, the finds of a secular character. As unambiguously ritual she considers such finds that come from the bog environment, include weapons and display a certain spatial arrangement (e.g. were deposited inside a vessel) (Levy 1982: 24). However, J. E. Levy’s work considered metal artefacts, while flint artefacts were not featured in her work. Without attempting to determine the character of individual deposits of flint artefacts from the Late Bronze Age, it seems that their contents: semi-finished and completed backed knives, such as the deposit from Spjald, constitute an element of a wider concept of religious or ceremonial manifestation in the Late Bronze Age. They are not its most spectacular element, but unequivocally prove that flint and flint objects were present not only in the utilitarian sphere of economy but were also an element in a more metaphysical perception of the reality. 4.2.4. Flints as funerary finds The presence of flint objects in Late Bronze Age graves is an issue yet to be studied in great detail. Even though their occurrence in the graves from that time is by no means anything unusual or rare, there are very few publications that do not limit themselves to the mere mentioning of flint while describing other elements of the burial. This tendency is seen both in Scandinavia (Högberg 2009) and in central Europe (Lech, Piotrowska 1997). Studies devoted to flint objects in graves and cemeteries are still very few in Poland (e.g. Mogielnicka-Urban 1997; Piotrowska 1997, 2000) and in Sweden (Högberg 2009; Rudebeck, Ödman 2000), and there are no such studies in Danish literature. The homogeneity of flint finds and their relationship to burials is still debated. The artefacts displaying features typical of flintworking from the Stone Age are interpreted as mechanical admixtures originally unconnected with the graves (Kobusiewicz 1988). Some interpretations speak of the possibility of the intentional deposition of flint artefacts from the preceding ages in the graves by the communities from the end of the Bronze Age (Piotrowska 2000: 308-309). Analysing finds from a few cemeteries in Scania in Sweden, A. Högberg drew interesting conclusions (2009: 256-261). On the basis of the finds from cremated burial graves from four sites dated to the V and VI periods of the Bronze Age he observed that they contained flakes that carried no traces of burning. These flakes were made in the same

67

68

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

manner as the flint artefacts recovered from the residential settlements, possessing at least one naturally sharp edge. According to Högberg, they were selected because of this sharp cutting edge. Absence of evidence of burning indicates that they were added to the features after the cremation and may have performed certain functions during the funerary rituals and ceremonies (2009: 260-261). The graves and cemeteries from the Late Bronze Age feature flint objects corresponding in their character to the flintworking known from settlements and residential homesteads. These are flake blanks produced ad hoc and sometimes tools, such as endscrapers and retouched flakes. Högberg observes that the graves do not feature large backed knives. The only example of a backed knife from a grave is a burnt fragment from the site Lebedow near Greifswald in Mecklenburg-Western Pomerania in Germany. It was found in one of the graves along with fragments of burnt bones and fragments of burnt stones. The knife’s proximal end was preserved, measuring 8 cm in length (Högberg 2009: 260 and Figure 2.195). Although flint assemblages from the cult features occur in the proximity of cult features in northern Jutland and are in direct relationships with them, no common presence of flint artefacts within the graves has been observed so far. The only example of which I am familiar of the presence of flint within cremated burial graves associated with a cult feature comes from the barrow in Torup Høje. It is a cemetery consisting of a

Figure 4.23. Torup Høje, Jutland. Urn grave in its secondary location at the foot of the barrow. The urn contained burnt human remains and two stones which did not display traces of fire. The urn had a double closing in the form of a flat pebble stone and a fragment of the bottom part of another vessel (drawing by M. Mikkelsen).

4. Specification of source information

dozen (originally possibly more) cremated burials surrounding the foot of the barrow, where flintworking activities took place. The cremated burial grave (K49) containing a grave urn with flint objects was found in the southwestern sector of the Torup Høje barrow. It was placed in the barrow’s base. One of the twelve cult features – a small mound of pebbles – was situated about a metre away from the grave. These features, together with the accompanying flint assemblages, will be discussed Figure 4.24. Torup Høje, Jutland. Urn from grave in the following chapters of the K49 containing unburnt flint products. book. This part of the barrow’s Note a pebble pressing on the bottom of the vessel ritual zone was explored without closing the urn and binder, probably resin, sealing the use of a meter grid; therefore, the space between the lid and the urn (photograph there is no information about a – courtesy of the Viborg Museum). flint assemblage accompanying the cult feature in question. The grave contained a completely preserved urn (Figure 4.23 and 4.24). The vessel escapes unambiguous dating; Martin Mikkelsen determined it to be generally connected with the Late Bronze Age. After it was documented at the site, the urn was taken to the museum, where it was investigated in detail. The urn had a double ‘lid’; a fragment of the bottom of a small vessel (height 5cm; diameter 5-6cm) and a flat pebble measuring 6 cm in thickness and 13 cm in diameter. The urn contained burnt human remains weighing about 350 grammes. The urn was accompanied by two flint artefacts exhibiting no traces of burning: a small cortical flake and a complete retouched flint blade (Figure 4.25). While the flake is a common type whose chronology is difficult to determine, the blade is by no means affiliated with Late Bronze Age flintworking. It is a Neolithic form, possibly connected with the Early Bronze Age. It was made from a regular single-platform core with a prepared striking platform. Scars from two previous blade removals can be seen on the dorsal surface. Flaking was not performed with a hard hammerstone, but possibly with either a soft hammerstone or a punch. One of the longer edges was retouched with continuous semi-abrupt retouch. An oblique rounded edge was made on the blade’s distal end. It does not display any wear or crumbling that would testify to its use or attempts

69

70

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

of fastening in a setting. The condition of both flint artefacts suggests that they were not an element of the cremation, but were placed in the urn together with the burnt remains. They were only an element of the grave’s furnishing. It is interesting that an older artefact, possibly hundreds years older, was put into the urn instead of the commonly used contemporary ones. The grave from Tarup Høje confirms Danuta Piotrowska’s thesis that the communities from the end of the Bronze Age intentionally deposited the flint objects from the preceding ages in the graves (Piotrowska 2000: 308-309). Unlike the blade tool, the cortical flake may have been associated with flintworking contemporary to the burial. It is an amorphic cortical form, one of the Figure 4.25. Torup Høje, Jutland. many that were produced in the ritual zone during the End-scraper on blade from the Late Bronze Age. Unfortunately, as it was mentioned urn from grave K49 above, the cult feature associated with the grave was (drawing by the author). not systematically investigated. If this had been the case, theoretically it would have been possible to connect the flake with the flint cores reduced near the feature. Then perhaps it would have been possible to determine a direct chronological and spatial relationship between the cult feature’s flint assemblage and the flint deposit in the grave. This example illustrates the great potential of the analyses of flint artefacts in explaining the phenomena connected with the burial rites in the periods when flint is assumed to have been only a marginal aspect of the culture. The remaining burials from the Torup Høje barrow in association with the cult features and the accompanying flint assemblages did not contain flint objects in the urns or grave pits. The cult features from northern Jutland in principle only feature flint objects related to the flintworking known from the settlement sites, but there is one exception – one large backed knife was found within the confines of the cult house Gramstrup I (Masojć, Bech 2011, Figure 20). It is a well-preserved, complete tool with additional traces of use. It will be discussed further on in this work, together with other artefacts from the cult houses. Inadequacy of interpretations of the occurrence of flint objects in the Late Bronze Age graves is symptomatic. The interpretation proposed by D. Piotrowska (2000: 319) concerning the ritual character of the presence of flint objects in graves and their use during funeral ceremonies in cemeteries seems fully justified, especially if viewed from the perspective of the issues discussed in this book. It sets the direction for further studies of the nature of the presence of flint products in the graves and cemeteries from the end of the Bronze Age and the Early Iron Age.

4. Specification of source information

The Sandagergåd cult house The Sandagergåd cult house in the commune of Ferslev in northern Zealand was discovered in 1985 during the construction of a local gas network. This is an area of the greatest concentration of barrows and megaliths in Denmark. Flemming Kaul interpreted the house as ritual on the basis of the presence of surprising features of an undoubtedly cult character and the absence of the remains typical of settlement sites (Kaul 1987, 1993, 1998, 2004). The outlines of the house’s construction consisted of two rectangular rows of pebble stones up to 50 cm in diameter. They were oriented along the south-north axis. Its length was 18.5 m and width – 7.5 m. The inner row of stones was separated from the outer one by 0.7 m. In the building’s northern part the inner row of stones had slightly rounded corners. The entrance to the building was not identified. Perhaps it was situated in the eastern wall, which did not survive. However, examples of comparable cult houses from Sweden show that they usually had no entrances (Victor 2002). Similarly, there is no evidence of the existence of walls or a roof. Pottery sherds and fragments of casting moulds were found inside. The pottery was dated to the IV period of the Bronze Age. Inside the house there were three urn graves from the Late Bronze Age, which – apart from burnt bone fragments – did not contain any furnishing. Additionally, altogether twelve graves associated with the pre-Roman period of the Iron Age, which were probably part of a larger cemetery, were found. However, the graves had no connection with the cult house, which undoubtedly was associated with the older part of the Late Bronze Age. Approximately 3 m to the south of the exterior wall of the cult house there was a group of four flat stones, whose length did not exceed 50 cm. On the flat surface of each of the stones was a carving of a human arm and an open palm, and an arrangement of four parallel horizontal lines towards which the palms stretched. Three stones lay flat, two of them were placed with the carving up and the palms reaching towards the cult house, while the third was placed with the carving down, but the palm also stretched towards the house. The fourth stone was placed almost vertically with the carving facing the building and the palm stretching downwards. The stones placed horizontally were granite, while the vertical one was a reddish sandstone. The stones with the carvings were accompanied by a fifth stone of the same size. It also probably had a carving of a palm and horizontal lines, but due to considerable weathering it had disappeared. Approximately 2.5 m to the west of the stones with the carvings there was an almost vertically placed boulder (menhir), 170 cm high and 100 cm wide. Fragments of another boulder (menhir) were situated at a similar distance on their other side to the east. Thus originally the two menhirs stood on either side of the stones with carvings. There was also a trace of a third boulder of a similar size closer to the stone outlines of the building. Flemming Kaul interprets the five fingers of each palm and the four lines on the carvings as a symbol of a nine-month cycle of birth. The stones with carvings surrounded by the menhirs marked an inaccessible sphere of sacrum, where there were spirits of the dead buried inside the house. Crossing the boundary marked by the stones, one entered the world of spirits. The arrangement also acted in the other direction, preventing the spirits of the dead from leaving the sacred sphere and re-entering the world of the living (Kaul 1987, 43). The remains of bronze casting within the cult house prompted Kaul to interpret its function as a place of ritual practices connected with the cult of the dead. It was also to be a place of initiation practices, introducing young people into maturity by teaching them the skills of producing objects from bronze while interacting with the spirits of the ancestors. Contrary to the well-preserved cult houses from Thy – Grydehøj and Høghs Høj – no flint artefacts were found in the Sandagergåd cult house.

71

72

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Sandagergård cult house, Zealand. Top: outlines of the cult object (after: Kaul 1987) and photograph during excavations – human figures standing in the south (after: Kaul 1993). Bottom: stones with representations of hands; on the left – in situ (three from four representations of hands, the fourth, westernmost stone was situated with the representation of hand downwards) (after: Kaul 1987); on the right – horizontal lines seen above hands (after: Kaul 2004). Drawing below the photographs: free reconstruction of the cult house as seen from the south (after: Kaul 1986).

4. Specification of source information

Sickle from the Stenild peat bog At the beginning of June 1898 in Stenild near Hobro a completely preserved tool consisting of a flint knife set in a wooden handle was found by an individual digging for peat at a depth of approximately 1.1 m below the surface of the peat bog (refer to the illustration on the right). The discovery was published in the same year by C. Blinkenberg (1898, 125-156). To this day it is the only completely preserved wooden sickle from the Late Bronze Age. The flint knife was fixed in a birch handle with two elongated, triangular wooden wedges. The handle is 36 cm long; the handgrip, 2.5 cm thick, is rounded at the bottom. The top widens to approximately 5.5 cm; in its lower part there is a notch and a hole, where the flint knife was fixed. A flint blade (11.6 x 2.8 x 1.2 cm) – an element of the tool – belongs to the category of large backed knives – tools characteristic of the Late Bronze Age. Its backed edge is unmodified and cortical. Only on the distal end was it truncated. The cutting edge is slightly S-shaped; it is chipped and polished as a result of use. Since the time of the first publication, the sickle from Stenild has been the subject of many studies, of which a substantial number was concerned with its function (Steensberg 1943; Wyszomirski 1979). C. Blinkenberg determined its use as a harvesting tool for cutting crops. Analysing the artefact at a high magnification (HPA), Helle Juel Jensen confirmed Blinkenberg’s conclusions, adding that similar knives from the Late Bronze Age also carry the traces of cutting reeds (1990). It was also observed that the wooden handle resulted in characteristic polishing on the surfaces of the blade’s backed edge. A number of instances of similar polishing seen on other knives, e.g. those from the deposit in Spjald (Becker), prompts the conclusion that the manner of hafting the knife in the handle utilized for the knife from Stenild was common during the Late Bronze Age, at least in northern Jutland ( Juel Jensen in print). Recently experiments have been carried out with replicas of the sickle from Stenild to study the microtraces caused by work and by the handle. The experiments have shown that large backed knives, such as the one from Stenild, were mainly used to cut crops and reeds, but they were also used for other purposes, which renders them multifunctional tools (Högberg 2009, 145-164). On the other hand, the experiments did not confirm the suggestions that they could have been used to process wood or cut branches, trees, leaves and nettle (e.g. Wyszomirski 1979). The artefact is exhibited on permanent display in the National Museum in Copenhagen.

Sickle from Stenild. Photograph and drawing of its flint element (after: Juel Jensen 1990)

73

5. Diversity of cult features from the Late Bronze Age in Jutland A characteristic feature of contemporary landscape of southern Scandinavia is the presences of countless barrows practically everywhere. Northern Jutland is the area where their number is especially great. When travelling from Viborg to Thisted one cannot fail to notice dozens of small mounds – landmarks of the local countryside for more than 3.5 thousand years (Figure 5.1). They were constructed during the Early Bronze Age, although the majority has its origins in the final phases of the Neolithic during the period of the Single Grave Culture. As a rule the barrows constructed by Neolithic communities were enlarged during the Early Bronze Age. Since the beginning of archaeological work in the mid-19th century many have been excavated, providing spectacular finds from the Early and Late Bronze Age. Equally many did not survive, destroyed by farming or plunder. To provide an idea of their actual number, one should consult a recently published portion of the orthophotographic record of the area in the vicinity of Skjern in central Jutland, comprising the valley of Skern Å, one of Denmark’s

Figure 5.1. Northern Jutland. The vicinity of Thisted (Photograph by the author).

74

5. Diversity of cult features from the Late Bronze Age in Jutland

Figure 5.2. The vicinity of Skjern in central Jutland. Barrows are marked with black points. Others: 1-2 – agricultural areas; 3 – meadows, marshes; 4 – moors; 5 – lakes; 6 – sea; 7 – forests (after: Johansen, Laursen 2007).

largest rivers, created as part of the programme Basic Cover1 ( Johansen, Laursen 2007). Aerial photography revealed the presence of greater than 30% more barrows than had been previously recorded (Figure 5.2). In the area of Thy 3741 surviving barrows have been recorded (Figure 5.3). Excavations carried out with systematic, modern methods have so far investigated less than 1% of that number, i.e. about 30 features. In a few cases, excavations were also completed in Recently published Basic Cover was in the 1950s a secret programme of the US Air Force. The goal was photographing over 99% of the country’s area from the altitude of 10 thousand feet.

1

75

76

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 5.3. Thy and the island of Mors with barrow mounds marked in the area (points). Based on laser scanning (courtesy of Danish cadastral office, Milijøministeriet, Kort & Matrikelstyrelsen).

the immediate vicinity of the barrows, which resulted in the discovery of cult features at their bases.2 For a long time archaeologists, amateur archaeologists and plunderers were primarily interested in what was inside the barrows, i.e. what was inside the mounds and was most spectacular – burials, including secondary ones, and their furnishings. This explains a great discrepancy between the number of excavated barrows and the number of recorded associated cult features (Clemmensen 2005: 304-305; B. H. Nielsen, Bech 2004: 150-151). While concentrating on the barrows’ central sectors, the excavations ignored their bases and immediate vicinity, where various cult features associated with 2

Information from Jens-Henrik Bech of the museum in Thisted.

5. Diversity of cult features from the Late Bronze Age in Jutland

them are situated. The surrounding areas, unprotected by law, unlike the barrows themselves, were destroyed by farming. However, the Høghs Høj barrow – the first well preserved cult house from Thy – is an example to the contrary; farming sometimes protected a barrow’s immediate vicinity by ploughing around it and thus covering its base with an excessive amount of soil (B. H. Nielsen, Bech 2004: 129). The discovery from more than thirty years ago of the cult house near the Høghs Høj barrow resulted in an increased interest in the areas surrounding barrows. The cult features situated there include cremated burials, usually urn graves, transferred into the mounds during the Late Bronze Age. The issue of the relationship between the cult features and the burials will be elaborated further on in the work. For now, I would only like to emphasise the direct connection between the location of the features known as cult houses and other features of a ritual character at the bases of the barrows and the burials which were in the Late Bronze Age placed inside the barrows themselves. As I mentioned earlier, for reasons of simplicity these areas are referred to in this work as cult zones. Aside from their ritual purposes, cult features may have played the role of spatial markers enabling easy location of burials. This spatial relationship: a grave – a cult feature/house substantiates the interpretation linking the appearance and presence of cult structures with the cult of the dead or the ancestral cult (Kaul 1998: 43). The connection between cult houses and graves is not restricted solely to Jutland. Most of the 80 houses from modern-day Sweden are characterised by the same relationship (Victor 2002, 86% according to Figure 26). Another criterion cited as the motivation for erecting cult houses is the proximity of water and the areas prominently visible in the landscape. In the region that is the focus of this monograph – northern Jutland, the cult features from the Late Bronze Age accompanying the burials dug into the barrows of Neolithic and Early Bronze Age origins differ in their construction, which – as has been observed – is reflected geographically (Hornstrup 2005: 279). The first were discovered only in the 1980s (Hornstrup 1999: 128). I would like to return to the issue of the number of surviving barrows and the presence of the cult features from the Late Bronze Age situated at their bases. I did not establish the number of recorded and systematically excavated barrow mounds in the vicinity of Viborg. According to M. Mikkelsen, these proportions should approximate those from the vicinity of Thy. Knowing an approximate number of systematically excavated features in relation to the discovered cult structures which accompany them, one might be convinced that the relationship is 1:1, which is not the case. Individual barrows frequently occur in small agglomerations, which to an extent reflect the settlement network as well as social and familial ties from the Early Bronze Age, when the basic unit was a homestead. During the Late Bronze Age their residents continued this relationship, selecting a specific barrow for a burial place. It became a small cemetery,

77

78

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

which reflected certain relationships, possibly of a familial nature. Thus not all the barrows were used for depositing secondary burials. It may be assumed that as a rule only a few out of any given number of barrows were used as burial places during the Late Bronze Age. Because cult features emerged in a direct relationship with specific burials or groups of burials, it becomes obvious that they were not constructed near the barrows which had not been selected for this purpose. An example illustrating this situation may be a group of five barrows at Torup Høje near Viborg, of which one will be the subject of detailed analysis in this work. In this group, only out of three excavated barrows (feature no. 60) had numerous secondary burials from the Late Bronze Age. It is these burials that were accompanied by more than ten cult features. It may be safely assumed that the barrows were used as a family cemetery for the residents of a homestead situated almost 300m to the east (M. Mikkelsen in print, Figure 2). It was also the closest of all the barrows to Torup Høje. This may likely be treated as evidence of persistence of family and social memory and continuity of the tradition of burying the dead in the same place for decades or centuries. To summarize, it cannot be assumed that all the Early Bronze Age barrows were used as burial places during the Late Bronze Age. Cult features, constructed in connection with secondary burials, are located where the dead were buried during the Late Bronze

Figure 5.4. Barrows constituting a permanent element of Jutland landscape. The vicinity of Sundstrup, Hjarbæk Fjord, central Jutland (Photograph by the author).

5. Diversity of cult features from the Late Bronze Age in Jutland

Age. On the basis of the research to date, including the number of excavated barrows and the established frequency of the accompanying cult features, it may be assumed that cult zones in the vicinity of barrows were a common phenomenon. The diversity of cult features within ritual zones in Jutland will be presented below. 5.1. Cult houses The features described in the literature as cult houses appear, as I mentioned above, in the Neolithic, and persist throughout the Bronze Age, while the most recent come from the Early Iron Age (Victor 2002: 67-76; L. Larsson 2007: 11-25). They occur mainly in southern Scandinavia, but some were discovered in Germany (Kersten 1936) and Norway (Myhre 1978). Cult houses are also known from the Mediterranean world, e.g. Greek islands (Randsborg 1995) and the vicinity of Troy, where they are dated to ca. 1300 BC (Besedow 2001). This chronology became the basis for the concept of contacts between Anatolia and Scandinavia and the potential adaptation of the idea of cult houses and ritual vessels from the south (T. B. Larsson 1999). Nevertheless, the cult house appears to be a Scandinavian tradition. Their greatest concentration is in southern Sweden, where 80 such structures have been discovered (Victor 2002, Figure 26). They have been known from the literature for a relatively long time (Almgren 1912). The work by Helena Victor cited earlier is the first and so far the only study devoted to cult houses. It is concerned with southern Scandinavia and focuses on the cult houses from Sweden. The author also mentions the features of this type from Denmark, e.g. from Sandagergård, Grydehøj and Ballermosen (Victor 2002: 69-72) The first documented cult house from the Bronze Age, published by Almgren at the beginning of the 20th century, was at the same time the first feature of this type known in southern Scandinavia. It was a famous house from Boda, situated near Enköping. The discovery is responsible for an erroneous conviction, which persisted until the 1960s, that all the Bronze Age houses were small structures with a stone floor (Almgren 1912, Victor 2002: 54). The cult houses from Thy in Denmark seem to have a specific local character, which has no analogies outside Jutland (B. H. Nielsen, Bech 2001: 2004). They escape the classification of cult houses proposed by H. Victor, who on the basis of the structures from Sweden distinguishes their two categories: large houses with stone walls and small post-construction features (2002: 66-67). The best known Danish cult house from Sandagergård in Zealand (Kaul 1987) clearly shows stronger links with Scania and Uppland in Sweden than with the cult houses in Jutland. The cult features from Danish ritual zones around the barrows are not the only cult houses. The features are variable, from cult houses proper to smaller ditch layouts, to

79

80

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 5.5. Cult houses preserved in fragments at the foot of the barrows from Thy: A – Toftumhøjen, B – Sundbyhøjen (after: Nielsen, Bech 2004, amended by the author). Location of secondary burials from the IV period of the Bronze Age is marked with an ellipse within the barrow.

5. Diversity of cult features from the Late Bronze Age in Jutland

occasional very small cult features, such as the remains of altars or individual spatial landmarks. Their detailed characteristics will be presented further on in the work. In their publication from 2004, Bjarne H. Nielsen and Jens-Henrik Bech list more than twenty barrows from the area of Thy, Mors island and the region of Skive and Viborg south of Limfjord, around which the remains of stone structures possibly connected with cult houses or analogous layouts were discovered (B. H. Nielsen, Bech 2004: 150-151). The illustration demonstrating spatial distribution of the features clearly shows that 13 out of 21 objects are the cult houses situated in the vicinity of Thy (one of them on Mors island), while the remaining ones outside Thy, mainly in the area of Viborg and Skive, are semi-concentric ditch features (B. H. Nielsen, Bech 2004, Figure 1). The best preserved cult house in northern Jutland is the feature from Grydehøj (B. H. Nielsen, Bech 2001, 2004) so far relatively dated to the early period of the Late Bronze Age (ca. 1000 BC) and situated at the foot of the barrow. Three graves were located at the barrow’s base – probably the reason why the cult feature was erected. The outline of the construction was rectangular; there was an entrance 1 metre in length in its southern side. The building’s walls consisted of inner and outer parts supported by post construction with peat filling the space between them. The walls’ thickness reached 1 metre. On the house’s eastern and western side there was a 1 metre wide stone paving, which probably continued to the entrance of the house, but in its southern part was later destroyed by ploughing. It is assumed that the roof was supported by the walls and a post situated in the middle of the house (B. H. Nielsen, Bech 2001, 2004). The Grydehøj cult house, a classic construction among the cult features in Thy, was not the first known structure of this type. An analogous feature was examined in the 1970s during the excavations of the Høghs Høj barrow (Masojć et al. 2013; B. H. Nielsen, Bech 2004). The Grydehøj and Høghs Høj cult houses are the basis for the analysis presented in this study. Both provided dense flint assemblages (Masojć, Bech 2011). A detailed description of their constructions and the characteristics of their flint assemblages will be presented in chapters 6.1 and 6.2. The discovery of the Høghs Høj cult house resulted in the search for similar features in Thy and Mors. However, the features found were considerably damaged by farming around the barrows. The remains of cult structures from the Thy region discovered at the foot of the Nørhå, Toftumhøjen and Sundbyhøjen barrows are similar, although their preservation is decisively worse (Figure 5.5). Situated at the eastern and southeastern side of the barrows, they have survived in the form of barely visible fragments of stone structures, which is the reason why it is difficult to determine their actual nature. In all three cases the cult houses were situated in the places where at the foot of the barrows there were secondary burials mainly from the IV period of the Bronze Age (B. H. Nielsen, Bech 2004).

81

82

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Another cult house from Thy, this time preserved in good condition, is the structure at the Gramstrup barrow. Its construction differs from that of the Grydehøj and Høghs Høj houses mentioned earlier (B. H. Nielsen, Bech 2001, 2004). Only a handful of flint artefacts were recovered from within the feature, a few flint specimens and a large backed knife. The specific nature of its construction and the artefacts found there are presented in chapter 6. In 2008 the excavations of the Ginnerup barrow carried out by the Thisted Museum (Museet for Thy og Vester Hanherred) revealed another typical structure adjacent to the barrow. In terms of its construction it is very similar to the Grydehøj and Høghs Høj structures – quadrangular outlines of a cult house limited on two sides with elongated stone benches. It is yet another example of a cult structure from Thy which provided numerous flint products. 5.2. Semicircular cult features Outside Thy, in the neighbouring areas, the ritual function of the cult houses seems to be transferred to features of a different character. They are seen on the ground as foundation ditches forming a semi-circle or the shape of a horseshoe. Their

Figure 5.6. Cult fluted objects from the vicinity of Viborg and Holstebro. A – Fly; B – Kobberup; C – Nr. Dalgaard Syd. Semicircular shading in marks the hypothetical location of the barrows (after: Hornstrup 1999, amended by the author).

5. Diversity of cult features from the Late Bronze Age in Jutland

location is analogous to that of cult houses – they are situated at the foot of the barrows, but their relationship with the secondary burials is not so obvious. A simplified version of cult houses, they were excavated in the 1990s in a few sites in the vicinity of Viborg and Holstebro: Fly, Kobberup; Nr. Dalgaard Syd (Clemensen 2005; Hornstup 1999, 2005). The barrows form the bases of the semi-circular outlines of the ditch features (Figure 5.6). There is a gap at the place where the ditch bends most sharply, which was probably the entrance to the feature. Inside one such feature, in Kobberup (Figure 5.6: B), there was a smaller semi-circular layout, also with a gap at the entrance. A similar construction element was found in the Grydehøj cult house. It was interpreted as the most sacred element of the cult feature, which will be presented further on in this work. The diameter of the ditch features varies between 4.5 and 9 metres; the width of the ditches measures from 0.2 to 1 m. Inside there were stones, which may have been used to support construction elements, such as wooden perches. Post holes occur along the ditches, but they are not distributed regularly. Both these elements suggest that there may have been a structure of some kind, e.g. a hut (Hornstrup 2005: 283). The function of the ditch features seems the same as that of cult houses from Thy. The article by Karen M. Hornstrup mentions flint artefacts present within the burials from the sites mentioned above, but it is difficult to determine whether they are the materials from the end of the Bronze Age or older, connected with the construction of the barrow itself. 5.3. Torup Høje and successive types of cult features The most important, in light of the topic at hand and at the same time the best excavated, site outside Thy is the Torup Høje barrow. At its base a dozen different cult features and associated flint artefacts were recorded. Despite extensive damage to the barrow and its vicinity at least four different types of cult features were recorded within a distinctly visible ritual zone (M. Mikkelsen in print). Within these constructions there was a cultural layer containing a large number of flint artefacts. Therefore, the context itself and the flint material will be discussed elsewhere. Here only the types of cult features recorded during the excavations of the barrow’s base need to be mentioned. The first are the ditch features forming a semi-circle or the shape of a horseshoe, similar to the above-mentioned features published by K. M. Hornstrup (2005). In addition to the two such ditch features (1), the following were identified in Torup Høje: vertically arranged boulders (2), pairs of post-hole features (3), stone mounds (4). The entire ritual zone of the Torup Høje barrow will be discussed in the section devoted to the cult features where flint collections were found. The concentration of the cult features at the Torup Høje barrow has no precedent on the supreregional scale, with its spatial arrangement, direct relationship with Late Bronze Age burials and the large quantity of flint artefacts.

83

84

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

To verify his observations made during the excavations of the Torup Høje barrow, in 2010 M. Mikkelsen excavated another barrow and its immediate vicinity, which seemed to have survived in good condition – Nøragergård Høje II situated ca. 10km east of Viborg (M. Mikkelsen 2011, 2012). The excavations confirmed the presence of cult features comparable to those from Torup Høje, situated at the foot of the barrow, right next to the burials from the younger Bronze Age deposited inside it. However, the cultural layer within them did not contain flint material (M. Mikkelsen in print). 5.4. Tetrahedral stone installations from southern Jutland Other cult constructions are discussed by B. Clemmensen (2005). They are tetrahedral outlines of stones, possibly constituting the remains of more complex structures. They are situated just like the remaining types of the cult features – at the edges of barrows. This type of cult feature is known almost solely from southern Jutland, reaching no farther north than the vicinity of Skanderborg in its central part. A few of them, e.g. the feature from Warringholz, are situated in Schleswig-Holstein in northern Germany. They are absent in Thy and other parts of northern Jutland. The features may assume the form of multiple (up to five) stone squares, adjacent to each other or situated separately on the circumference of the stone paving surrounding the barrows but linked by rows of stones (Figure 5.7) (Clemmensen 2005, Figure 2, 3). They are usually empty. Their chronology determined on the basis of the stratigraphy, i.e. the relationship to the consecutive construction phases of the barrow, is associated with the Early Bronze Age. Only the feature from Vrejlev Kloster has been verified as originating in the younger Bronze Age. The literature does not mention the presence of

Figure 5.7. Jernhyt near Haderslev. Southern Jutland. An example of two four-wall cult objects linked with a stone strip (after: Clemensen 2005, amended by the author).

5. Diversity of cult features from the Late Bronze Age in Jutland

any flint artefacts in this type of cult feature. According to F. Kaul they may have played the role of spatial markers identifying the burial places (2004: 113). 5.5. Remaining features In 2007 the Museum in Viborg carried out rescue excavations of the Duehøj Syd barrow at the north-eastern outskirts of the town. Even though the barrow was completely ploughed over and had no remaining mound, buried contexts survived, including cult features from the Late Bronze Age associated with urn graves, e.g. a semi-circular ditch features of ‘classic’ construction, size and location in relation to the barrow (M. Mikkelsen in print). Quite specific features, possibly cult ones, from the vicinity of Viborg are reported by Martin Mikkelsen (in print). They come from the old excavations carried out by the Museum in Viborg. The first – Langmosehøj excavated by G. Rozenberg, initially interpreted as a residential building from the Roman period, has the shape of a horseshoe, whose open end touches on the barrow’s base (Figure 5.8: A). It has a double stone construction. The interior was filled with the cultural layer containing pottery dated to the period of Roman influence. According to M. Mikkelsen, there are a number of premises to assume that the feature in Langmosehøj is a cult one. The other feature is at the St. Torup barrows excavated in the 1960s, situated north of Viborg (Figure 5.8: B). It is now interpreted as the remains of two phases (two features?) of a cult structure corresponding to the house from Langmosehøj. In both cases there were cremated burials from the Late Bronze Age in direct association with the cult features at the foot of the barrow.

Figure 5.8. Potential cult houses from the vicinity of Viborg: A - Langmosehøj; B - St. Torup (after: Mikkelsen in print).

85

86

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Both features should be interpreted as cult houses. The construction combines the elements of the cult houses from Thy and the semi-circular ditch features from northern Jutland. No flint artefacts have been found inside these features. In conclusion, the cult houses in Jutland seem to be a specific feature of the Thy region, while the features of more variable form occur in northern Jutland but outside Thy. B. Clemmensen mentions the features similar to the cult houses outside Thy: a cult feature in Kragelund near Silkeborg and a feature from Stendis near Holstebro, but these examples are not well documented and their chronology has not been determined (Clemmensen 2005: 300). Because it is still a new aspect for the archaeologists excavating barrows and adjacent structures, more will be known as a result of additional field investigations. A well-documented and interpreted cult house from northern Zealand is Ballermosen, excavated by E. Lomborg in the 1950s and dated to the II period of the Bronze Age on the basis of its relationship to the barrow’s construction. Even though with its dimensions (3 x 3m) it is reminiscent of the already-mentioned small square structures from Germany and southern Jutland, it is so far one of the oldest developments of this type, which is actually the remains of a building of a ritual character (Lomborg 1956: 150-152, Figure 7, 8). Discussing the cult house (house G, related to barrow II) Lomborg writes that flint flakes and a end-scraper on a flake were found within the cultural layer present inside, similar to that constituting the fill of the third phase of the barrow (1956: 151). It is quite significant that Ballermosen is situated only about 15km from the famous cult house Sandagergård (Kaul 1987). The presence of numerous flint assemblages within cult houses is limited to Thy and the neighbouring areas. In light of the current state of the research, with barely more than ten such structures known and only a few having been excavated with the use of modern methods, this should also change during the course of further field work. An example is the Ginnerup cult house in Thy excavated recently (Masojć, Bech 2011: 221-222). According to M. Mikkelsen,3 a certain chronological sequence may be tentatively adopted for the cult features from northern Jutland. The largest, i.e. the cult houses, could be the oldest. In continuation, the smaller the cult features’ constructions became – from extensive semi-circular ditch layouts to the smallest pairs of post holes and mounds of pebbles – the relatively younger they were. This conclusion, however, needs to be verified, for instance by the absolute chronology of at least some of the features that may be dated.

3

Verbal communication.

6. Cult features with flint assemblages from northern Jutland As was mentioned earlier, the occurrence of large flint assemblages within cult houses and other cult structures has so far not been the subject of research. The phenomenon has been mentioned only very recently (Masojć, Bech 2011). The assemblages that are the subject of this research come solely from northern Jutland, while comparable assemblages have not been observed in the rest of Scandinavia or in the areas situated south of Jutland. It is, however, quite possible that just like the case of the material from Thy, such assemblages do exist but their existence is known only by local museums. At the present stage it is best to assume that just as the very forms of the cult houses from Thy are a regional variant of the idea of the cult house from a much larger geographical area, the flint material found inside them is also a manifestation of a regionally-specific character. It is also quite possible that the reason why there are so few well-preserved remains of cult houses containing flint assemblages is the nature of the excavations carried out earlier, which focused on the very barrows or even their selected, central parts. The modern methodology, enabling excavations of entire features and their vicinities, reveals a wider context of their functioning. The excavations carried out in the future will result in the discoveries of new cult houses and features at the foot of the barrows, which during the Late Bronze Age were the burial places. This is confirmed by the cult structures excavated in 2010 at the Ginnerup barrow in Thy. Only preliminary results of the research will be presented here. The cult houses from the Thy region are a local variant, whose occurrence does not stretch beyond the waters of Limfjord. They have never been encountered outside Thy. In the remaining parts of Jutland cult features at the graves dug into the barrows are less monumental. They are smaller but at the same time there are more of them in each individual barrow. Aside from the very idea and location, their common feature may be the presence of flint assemblages. Except for those in Thy, the only known cult feature with a flint assemblage is located at the Torup Høje barrow. This is not proof that the occurrence of this phenomenon outside Thy is exceptional, but it is the result of the combination of many factors mentioned above: the state of the research, its nature and the methodology of investigations. In the Thy region four features featured flint assemblages inside cult houses, Høghs Høj, Grydehøj, Ginnerup and Gramstrup. The rest of Scandinavia is represented by only the feature from Torup Høje in the vicinity of Viborg. All the features will be presented below, both in terms of the constructions of the cult features and the character 87

88

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

of flint assemblages which they yielded. The analysis shows that despite their general uniformity, they differ in their construction and the artefacts found inside them. One recurring theme is the presence of homogenous flint assemblages within the features. 6.1. Grydehøj, Thy The cult structure adjacent to the Grydehøj barrow in southern Thy is the best preserved example of a cult house constructed during the Late Bronze Age in northern Jutland (Figure 6.1; 6.2). The cult house was discovered and investigated in the late 1990s during the rescue excavations of the barrow threatened with destruction by agricultural activity. The excavations resulted in details concerning the cult house’s spatial relationship with the burials secondarily dug into the barrow’s base (Masojć, Bech 2011; B. H. Nielsen 1999, 2000; B. H. Nielsen, Bech 2001; 2004).

Figure 6.1. Grydehøj, Thy. The barrow and the cult house during exploration. View from the south (photograph by B. H. Nielsen).

6. Cult features with flint assemblages from northern Jutland

Figure 6.2. Grydehøj, Thy. Cult house at the barrow. Stages of exploration. A – plan of the barrow and the cult house: 1 – barrow from the Single Grave culture; 2 – ring of pebbles marking the extent of the barrow from the early Bronze Age; 3-4 graves(?); 5 – box-shaped grave; 6 – cremated burial grave in wooden casing; 7 – urn grave; 8-10 elements of cult constructions; 11 – cooking pit; B – cult house before uncovering its base, numerous pebbles inside the house come from the stone footing running along the walls, 1 –box-shaped grave; 2 – cremated burial grave in wooden casing; 3 – urn grave; C – cult house after uncovering its base, 1-3 as in B; 4 – post object constituting an element of the construction supporting the roof(?); 5 – traces of hearth in the form of the base burnt red; 6 – pit; 7 – cooking pit; (after: Nielsen, Bech 2004, drawing by B. H. Nielsen, amended by the author).

89

90

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.3. Grydehøj, Thy. Layers of the cult house’s fill during exploration carried out within individual quarters of square metres. View from the south (photograph by B. H. Nielsen).

The Grydehøj cult house is also one of the best excavated structures. The whole feature was explored with the use of a grid of square metres (also quarters of square metres), which enabled a spatial analysis of the distribution of artefacts (Figure 6.3). The sediments from individual quarters were sifted, thus providing a considerable amount of small debitage. The Grydehøj barrow and its chronology, as well as the location and construction of the cult house have been previously described (B. H. Nielsen, Bech 2004). The barrow was constructed during the time of the Neolithic culture of single graves, and was later – in the Early Bronze Age – reconstructed several times. A burial in a wooden chamber, situated in the centre of a stone circle whose diameter never exceeds 10 m, is one of the surviving elements from the oldest period (Figure 6.2: A, 1). The outline of the Early Bronze Age barrow was nearly twice as large (Figure 6.2: A, 2). Inside it, in its southern part there were three secondary cremated burials from the Late Bronze Age, situated just on the inner side of the stone circle facing the Early Bronze Age barrow (Figure 6.2: A, 5-7). A box grave of stone construction was situated to the east (Figure 6.2: A, 5). About 2 m towards the west were a cremated burial inside a wooden casing (Figure 6.2: A, 6) and an urn grave (Figure 6.2: A, 7). The burials contained burnt bone fragments. Their analysis is presented in the section devoted to the latest excavations of the barrow (refer to chapter 9.1). According to B. H. Nielsen and J. H. Bech their chronology was not determined due to the absence of contexts that could be absolutely dated, but the similarities to the Thy region suggest the first half of the Late Bronze Age, especially in the case of the stone box grave (2004: 134-135). It was during the first half of the Late Bronze Age that at least two cult structures were erected one after another on the southern side, outside the stone face of the barrow from the Early Bronze Age. The oldest – and the best-preserved – were the remains of the cult house (Figure 6.2: A, 9; B; C) partly covered by a younger structure in the form

6. Cult features with flint assemblages from northern Jutland

of a stone ramp or bench (Figure 6.2: A, 8). On the east of the barrow there was a stone structure, possibly fulfilling a ritual function, but because only portions of it survived, it is impossible to more precisely determine its characteristics (Figure 6.2: A, 10). The central part of the older structure was a cult house that measured as follows: 5.5 x 5.7m (the surface of over 30m2). Before it was constructed the area was levelled to create a flat surface for the construction. The feature’s outlines were marked by the remains of walls in the form of a regular, narrow ditch with more or less regularly situated post holes (Figure 6.2: B, C). From the south there was a tunnel entrance to the building, about 1m wide and 1.5-1.7m long, while at the eastern edge of the entrance, next to the ditch, there was a negative of a post or stone, which may have been a construction element barring the entrance to the building (a door?). On the eastern and western sides of the house there were stone structures in the form of an elongated surface of pebbles making up a kind of ramp. The western part was about 7m long, while the eastern was about 5m long. It is quite possible that such features surrounded the building also in the front, reaching the entrance, but they did not survive due to destructive farming activity around the barrow (B. H. Nielsen, Bech 2004: 137-138). The space between the remains of the wall in the form of the ditch and the stone surface was filled with a 1.5m thick wall made of peat. The wall survived to the height of 3040cm, especially along the side of the barrow, which determined its edge. From the south the wall may have stretched on both sides all the way to the building’s entrance, which would account for its considerably elongated, tunnel-like shape. The surface of pebbles surrounding the house and the peat wall preserved on both sides of the building most probably was not covered and according to the researchers may have been used for movement, e.g. as a procession path (?). The artefacts occurred both within the cult house and the ‘procession path’, though in the latter they were much less numerous. The space occupied by the peat wall had no artefacts at all. According to Nielsen and Bech, many pebbles found inside the house were associated with its interior construction. They were not the remains of the floor, because most of them were deposited above the artefacts discovered inside the feature. Instead, they were concentrated along the walls, especially in the better preserved, rear part of the building, which might suggest that they were the remains of a stone bench situated along some of the building’s walls. There was a single, deep post hole in the middle of the cult house, below the level of the pebbles and the level where the artefacts were deposited (Figure 6.2: C, 4). The excavators interpret it as the remains of a pillar supporting the building’s roof (B. H. Nielsen, Bech 2004: 139). The remains of the hearth were identified towards the

91

92

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.4. Grydehøj, Thy. Cult house during exploration (photograph by B. H. Nielsen).

Figure 6.5. Grydehøj, Thy. Cult house during exploration (photograph by B. H. Nielsen).

building’s back wall at the floor level (Figure 6.2: C, 5). It was situated at the top of one of the two arched furrows forming a semi-circular construction – a partly enclosed space based on the building’s hypothetical back wall. Access to the space was provided exactly opposite the entrance to the cult house. This construction may have played a special role in the functioning of the entire structure. The authors refer to it as a sanctum, i.e. a place of special ritual significance.

6. Cult features with flint assemblages from northern Jutland

Figure 6.6. Grydehøj, Thy. Cult house during exploration (photograph by B. H. Nielsen).

In view of the description above, the Grydehøj cult house would have been a roofed building with walls of uniform, compact construction. However, this interpretation is opposed by other researchers, who claim that the feature had no roof, while the walls were low. Additionally, they argue that it did not have a wall along the barrow’s side. These arguments question the idea of the building as a feature remaining in a spatial relationship with the graves. However, these opinions have not yet been published. More about the debate about the possible construction of cult houses, including the one in Grydehøj, will be presented further on in this work (refer to chapter 10.1). Refraining from settling the issue of whether the building had a roof or not, the hypothesis that the feature is a cult house is still well-grounded. Pottery from the IV and V periods of the Bronze Age was found in the fill of the house, in the cultural layer consisting of two separate stratigraphic units (N17 overlying N18). Altogether more than 5kg were found, coming from small vessels – possibly goblets or beakers. For the most part, the fragments of the vessels lay on both sides of the entrance to the semi-circular feature at the building’s rear and in the vestibule of the entrance in the front (Figure 6.16). The authors interpreted the presence of outbuildings in the house’s vicinity and the fragments of beakers or goblets as the remains of ceremonies taking place within or in the vicinity of the cult structure (B. H. Nielsen, Bech 2004: 141, 156). In addition to the pottery, a small number of burnt bones (it has not been established whether they are animal or human remains) and a very large assemblages of flint artefacts were found. The distribution of the flint artefacts and pottery sherds is presented graphically in the illustrations (pottery – Figure 6.16; flint artefacts – Figure 6.18-6.23).

93

94

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Table 6.1. Grydehøj, Thy. Structure of flint assemblage from two stratigraphic units (N17, N18) within the cult house N17

Category raw material

N18

N17+N18

N

%

N

%

N

%

0

0

8

0,1

8

0,12

cores

22

5,2

487

7,6

509

7,45

tools

11

2,7

163

2,5

174

2,54

flakes

127

30

1793

28

1920

28,11

blades chips (less than 2

5

1,2

52

0,8

57

0,83

168

39,8

1766

27,6

1934

28,32

89

21,1

2138

33,4

2227

32,63

422

100

6407

100

6829

100

industrial chunks total

Figure 6.7. Grydehøj, Thy. Structure of the flint assemblage from the cult house (within stratigraphic units N17and N18). I – raw material; II – early phase of coring; III – advanced exploration; IV – final exploration; V – other flakes and unidentified products; VI – tools.

6.1.1. Description of the assemblage The artefacts, including flint artefacts, occurred within the cult structure in Grydehøj both in the cult house (layers N17 and N18) and, in considerably smaller numbers, in the surrounding stone surface interpreted as a procession path (N21, N22, N23, N29). The latter, however, were not explored as systematically as layers N17 and N18 within the building (within individual quarters of square metres). In this way a small assemblage was acquired, which was not analysed. Flint artefacts occurred within nearly all the areas of the building. The flint assemblage from the Grydehøj cult house weighs more than 103kg and consists of nearly 7 thousand artefacts. Only a few hundred artefacts occurred in the house’s uppermost cultural layer (N17), while the majority occurred in the lower layers of the house’s fill (N18). The general typology of the assemblage is presented in Table 6.1 and Figure 6.7. 6.1.2. Raw material The assemblage from Grydehøj was largely made from the most common flint in southern Scandinavia – Late Cretaceous Scandinavian Senonian flint. It is a good quality

6. Cult features with flint assemblages from northern Jutland

raw material, gray or black in colour with thin Cretaceous cortex with occasional light-gray oval spots (up to 25mm in diameter). There are different varieties of the flint (Becker 1988; Vang, Petersen 1933), which are known by many names, e.g. flint of the Bjerre type (Högberg, Olausson 2007). It is commonly found in the Cretaceous formations from Scania to Zealand, to Jutland and to the secondary deposits in the moraine formations on the beaches and cliffs. In the Grydehøj cult house its contribution reaches 90% in all types of the artefacts (Table 6.2). The remaining raw material in the assemblage is the speckled Senonian flint, whose contribution never exceeds 5% of the artefacts (this raw material is very similar to the flint predominating the assemblage and often confused with it or regarded as identical to it), the gray band Danian flint (up to 2.4%) and the brown bryozoan (up to 2%). These are all easily available local raw materials. Their presence in the assemblage obviously results from its availability, while its quality – although quite good – must have played a secondary role. The flint assemblage has been preserved in a good condition, carries no traces of patination and

Figure 6.8. Grydehøj, Thy. Hammerstones from the interior of the cult house (N18) (photograph by J.-H. Bech).

95

96

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

only 0.57% of the artefacts (52 pieces) exhibit traces of heat alteration. The burnt artefacts come from the area where fire was kindled inside the house (Figure 6.2: C, 5). 6.1.3. Technology With the manner of its production and its morphological characteristics, the assemblage from the Grydehøj cult house corresponds to the flint assemblages from the Late Bronze Age settlements mentioned above. The shared technological characteristic of the flint assemblages from the cult houses and features in Jutland is that they were made solely with the use of a hard hammerstone (Figure 6.8). It was used to break up flint blocks, which carry distinct traces of multiple blows. This resulted in numerous cores – negative forms of amorphic shapes, frequently with single flake removal scars. Strong blows caused the blocks to break into many fragments with sharp edges which do not display the characteristics of flakes, even though they were created by the intentional fragmentation of the raw material. This explains the substantial contribution of chunks in the assemblage as well as the fragments identified as cores (Figure 6.9; Table 6.1). The purpose of this method of reduction was the production of flint fragments with sharp edges, and not necessarily flakes. The presence of flintworking tools - quartzite hard hammerstones carrying the traces of use indicates that the assemblage was made on the spot, within the confines of the house.1 Carrying it into the house would have been difficult, because – as it was mentioned above – it weighed more than 100 kg, with one of its main constituents being the category of chips numbering nearly two thousand artefacts. It is quite possible that a small number of hammerstones in the assemblage (5 pieces) resulted from the presence of a great number of cobbles in the layers, which came from the Table 6.2. Grydehøj, Thy. Flint raw material within stratigraphic units (N17, N18) from the cult house Category of artefacts

 

SSF

Stratigraphic unit

N/100%

N

N17

132

125

Blanks

N18

1845

1648

N17

22

20

Cores

N18

487

448

Tools1

N17

11

11

N18

154

138

SSF – Scandinavian Senonian Flint, SpSF – Speckled Senonian Flint, GDF – Gray Band Danian Flint, BB - Brown Bryozoan, UN/B - unidentified/burned

SpSF %

N

89,8

1 94

91,9

0

90,4

0

7 5

% 4,8 1,4 3,0

GDF N 2 29 0 1 0 4

BB %

1,6 0,2 2,4

N 0 32 1 8 0 2

UN/B %

N 4

%

1,5

42

1,7

23

4,8

1,2

5

3,0

1

2,3

Nine tools (4 hammerstones and 5 perforators) were added to the assemblage after its analysis.

1

The first publication to address flintworking from the Grydehøj cult house reported that no flintworking tools were found inside it (Masojć, Bech 2011: 213). Only after the publication of the article did it become apparent that quartzite hammerstones constituted an element of the collection acquired during the exploration of the cult structure.

1

6. Cult features with flint assemblages from northern Jutland

house’s construction (refer to observations concerning the construction). Some of the hammerstones may have been mistakenly identified as construction material, which is indirectly substantiated by the fact that a considerably greater number of hammerstones carrying traces of use was found at Høghs Høj.

Figure 6.9. Grydehøj, Thy. Cores from the cult house (photograph by M. Jórdeczka).

97

98

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

6.1.4. Cores The simple procedure of breaking and exploitation of flint blocks with the use of a hard hammerstone is reflected in the characteristics of the category of cores (Table 6.3), where nearly 40% of the artefacts are what have been classified as amorphic cores, i.e. blocks of raw material exhibiting traces of the unsystematic separation of blanks from the most convenient surfaces . This strategy frequently resulted in breaking a core into a few fragments – chunks, which were later also used as cores. In addition to this most frequent method of reduction, two other categories of cores are common within the cult house: single-platform cores and non-prepared conical cores (Plate 14: 1, 2). As in the case of the amorphic cores, the manner of reducing the blocks was determined by the shape of a flint objective piece. Block-like nodules with one flat surface suitable for a striking platform serve as single-platform cores. The radial cores are primarily blocks, moderately flat and oval. They also sometimes included the cores which had undergone bifacial flaking, resulting in the creation of bi-pyramidal blocks, while the flakes knapped on the circumference displayed prepared striking platforms (Plate 14:3, 4; 15:1). It is strikingly similar to the technique of Levallois reduction of conical cores from the mid-Palaeolithic (Van Peer 1992). In our case, however, bifacial, radial coring was an end in itself. The flakes acquired in this way became the finished products. Table 6.3. Grydehøj, Thy. Cores from two stratigraphic units (N17, N18) within the cult house Total

Category of cores initial core

%

N17

N18

N

1

39

40

7,85

single-platform flake core

9

107

116

22,78

flake core with equivalent flaked surfaces/striking platforms

1

11

12

2,35

blade core with equivalent flaked surfaces/striking platforms

-

3

3

0,58

non-prepared flake disc core

1

82

83

16,4

prepared flake disc core

-

27

27

5,3

block-like core with changing orientation

-

22

22

4,32

flake amorphous core/fragment exhausted core total

10

180

190

37,28

-

16

16

3,14

22

487

509

100

Occasionally reduction was continued by changing a core’s orientation. In such a case either the existing flaked surface or a new unprepared surface were selected as a striking platform. Such cores usually have a block-like form, while the change of orientation was applied several times (Plate 15: 2). A separate category consists of cores with two equal striking platforms (Plate 15: 3-5). However, the exploitation of cores ended after removing a few flakes, at which time a core was abandoned. Advanced reduction, as represented by exhausted cores, was only marginally present.

34

N17, N18 N18 N17, N18 N18 N18 N17, N18 N17, N18 N17, N18 N18 N18 N17, N18

flake core with equivalent flaked surfaces/striking platforms

blade core with equivalent flaked surfaces/striking platforms

non-prepared flake disc core

prepared flake disc core

block-like core with changing orientation

flake amorphous core

end-scraper

perforator and borer

notched tool

burin

ad hoc tool

102

3

9

15

7

56

21

27

75

1

14

9 101

39

N17

N17, N18

initial core

38

N18

N17, N18

preparation products

single-platform flake core  

N17, N18

two-directional blade

18

29

33

23

20

24

33

35

24

30

21

27

25

21

34

27

21

4

12 31

82

N17, N18

cortex blade >50% N17, N18 blade with removals on dorsal face consistent with the direction of blade’s removal N17, N18

multidirectional flake

16

97

18

21 23

11

656

10

flake with removals on dorsal face oblique to/shifted to the direction N17 of flake’s removal N18

20 12 17

37 734

N17 N18

L min

59

N

stratigraphic unit

flake with removals on dorsal face consistent with the direction of N17 flake’s removal N18 flake with removals on dorsal face opposite to the direction of flake’s removal N17, N18

cortex flake >50%  

Category of artefact

80

82

60

116

59

46

84

70

84

86

84

84

60

95

69

55

70

56

54

70

47

62

70

57

74

60

L max

44,3

48,6

63

41,2

28,8

51,8

49,2

55,1

55,1

-

52,4

45,3

43,8

64,1

38,6

40,5

42,3

40,6

31,3

32,4

32,7

35,1

30,9

30,1

32

34,7

L med.

45,1

52,1

41,1

32,4

L Med.

13

26

28

24

17

23

33

27

25

26

25

35

22

9

8

8

7

14

14

15

16

9

15

14

19

W min

28

86

42

58

52

39

101

90

84

101

60

85

73

101

48

14

22

21

70

62

42

46

84

56

76

60

W max

36,3

33

47,1

38,2

26,2

44,5

45,6

45,7

46,3

41 -

45,2

49,1

48,9

36,1

45,7

13,5

31,2

W Med.

23,4

11,7

15,2

13,6

32,6

30,5

28,2

30,9

32,1

29,8

31,6

34,1

S med.

Table 6.4. Grydehøj, Thy. Sizes of individual classes of artefacts within stratigraphic units (N17, N18) (in millimetres); L – length; W – width; T – thickness

3

10

12

6

5

14

16

12

11

20

14

13

18

4

5

2

3

3

4

5

3

2

4

2

5

T Min

26

11

26

16

12

60

48

37

60

43

61

68

25

85

26

9

30

19

28

23

9

19

40

18

69

18

7,9

8,9

T Med.

10,9

10,6

17,8

10,8

8,4

26

29,6

21,5

24,6

29,5 -

28,1

19,2

33,4

11,7

26,4

13,4

7

8,6

8

9,9

9

7

9,7

8,9

7,6

9,9

9,5

T T Max med.

6. Cult features with flint assemblages from northern Jutland 99

100

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

The sizes of cores vary (Table 6.4). In some cases they measure nearly 10 cm in length and even more in width. On average the cores from Grydehøj measure 5 cm both in width and length; they are also relatively flat – on average they do not exceed 5 cm. The plate-like shape of many cores, with their relatively small thickness, determined the radial nature of reduction of some of them. 6.1.5. Blanks All flintworking within the Grydehøj cult house had a flake character (Table 6.5). Blade blanks are coincidental and have only a metric dimension, similarly as very occasional blade cores. The direction of flake scars on the blanks’ dorsal surfaces was adopted as the basic criterion of their division. The methods of core reduction mentioned above are reflected in the character of the blanks. All the blanks have common features: they are usually short, with wide distal ends, very distinct impact points and bulbs. Cortical blanks predominate, accompanied by flakes with flake scars on dorsal surfaces consistent with the direction of the flake’s removal, which confirms that the cores are predominated by single-platform forms. The analysis of the blank butts reveals that the assemblage is predominated by cortical or natural butts (Table 6.6), which includes the categories of cortical flakes, non-cortical flakes and individual flakes of blade proportions. The relatively smallest contribution of Table 6.5. Grydehøj, Thy. Blanks from two stratigraphic units (N17, N18) within the cult house N17+N18

Categories of blanks

N17

N18

N

%

cortex flake >50%

37

814

851

43,06

cortex blade >50%

1

20

21

1,06

preparation products flake with removals on dorsal face consistent with the direction of flake’s removal flake with removals on dorsal face opposite to the direction of flake’s removal flake with removals on dorsal face oblique to/shifted to the direction of flake’s removal

5

33

38

1,92

62

709

771

38,99

1

19

20

1,01

10

118

128

6,5

multidirectional flake blade with removals on dorsal face consistent with the direction of blade’s removal

5

80

85

4,29

2

29

31

1,56

two-directional blade

2

3

5

0,25

unidentified blanks total

7

20

27

1,36

132

1845

1977

100

2

multidirectional blade 121

4

two-directional blade 1782

42

  blade with removals on dorsal face consistent with the direction of flake’s removal

tools

82

blanks in total

8

115

multidirectional flake

5

1

2

2

-

1

23

4

47

38

preparation products

1

714

21

cortex blade >50%

39

N17

  flake with removals on dorsal face consistent with the direction of flake’s removal flake with removals on dorsal face opposite to the direction of flake’s removal flake with removals on dorsal face oblique to/shifted to the direction of flake’s removal

741

N 100%

cortex flake >50%

List of features  

87

1

1

20

47

80

14

536

16

16

523

N18

92

1359

2

3

22

47

88

15

583

20

17

562

N

Cortical/natural

76,1

100,0

75,0

52,4

57,4

76,6

65,4

81,7

52,7

80,9

75,8

%

71,9

%

-

-

-

2

4

2

-

11

1

-

6

N17

25

1

20

32

25

6

121

17

4

172

N

Shaped

398 25

-

1

18

28

23

6

110

16

4

166

N18

-

%

26,3

20,6

25,0

47,6

39,0

21,7

26,0

16,9

44,7

19,1

23,3

%

-

-

-

-

-

-

-

3

-

-

1

N17

4

-

-

-

3

2

2

7

1

-

6

N18

4

25

-

-

-

3

2

2

10

1

-

7

N

Prepared

-

-

-

3,6

1,7

8,6

1,4

2,6

-

0,9

%

Table 6.6. Grydehøj, Thy. Presence and manner of preparing butts of blanks and tools within stratigraphic units (N17, N18)

3,3

1,8

%

6. Cult features with flint assemblages from northern Jutland 101

102

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.10. Grydehøj, Thy. Sizes of blanks.

unprepared butts is seen among the products of core preparation. Shaping is observed in one-quarter of the assemblage, while preparation occurred extremely rarely. The tendencies seen in the blanks and the products of preparation are also observed among the tools (Table 6.6). Here, shaping and preparation were very occasional, which confirms that the ad hoc tools were dominant. Individual formal tools present in the assemblage are usually unifacial crested blades and the forms approximating plunging blades, but they occur very occasionally (Table 6.6). Flakes, and especially cortical ones, are quite large (6-7cm in length; 7-8cm in width). On average the size of flakes slightly exceeded 3cm both in length and width (Table 6.5, Figure 6.10). Thus, in general terms the blanks in the assemblage from Grydehøj, including blades, are not especially large. A hard hammerstone and absence of preparation are responsible for the fact that the blanks are usually amorphic, short and relatively thick (an average flake was nearly 9mm thick). 6.1.6. Tools The category of tools from Grydehøj consists of the products which are to an extent standardised – formal tools and ad hoc forms (Table 6.7). However, even in the case tool categories such as end-scrapers and perforators it is obvious that the selection of blanks was coincidental and that they were produced ad hoc, even though they retained classic features of this type of tool. Most were made on irregular cortical flakes or even on chunks, usually utilising flint fragments whose shape rendered them the easiest to produce a tool. The formal tools include endscrapers (Plate 16: 1-8), perforators (Plate 17: 1-12, Figure 6.11; 6.12), notched tools and denticulated pieces (Plate 18: 1-10, Figure 6.13), as well as burins (Figure 6.14).

6. Cult features with flint assemblages from northern Jutland Table 6.7. Grydehøj, Thy. Tools from two stratigraphic units (N17, N18) within the cult house Categories of tools

N17 + N18

Formal tools

N17

N18

N

%

end-scrapers

2

7

9

5,2

perforators and borers

2

18

20

11,5

notched tools

-

9

9

5,2

-

3

3

1,7

burins Ad hoc tools knives

-

5

5

2,8

flakes/blades/retouched chunks

3

77

80

45,9

functional implements

3

33

36

20,8

unidentified tools/fragments

1

7

8

4,6

hammerstones

-

4

4

2,3

11

163

174

100

Other

total

Figure 6.11. Grydehøj, Thy. Perforators (drawing by the author).

In their majority the end-scrapers were made on sizeable cortical flakes (Plate 16: 1, 3-5), while some were made on chunks (Plate 16: 6-8). Except for a flake with additional marginal retouch (Plate 16: 2), the end-scrapers are short and thick, and a few are circular and semi-circular forms (Plate 16: 6, 8). Rounded fronts predominate. The perforators represent a more complex category (Figure 6.11; 6.12; Plate 17). Selection of good quality raw material was of secondary importance here, which is substantiated by the presence of numerous perforators on chunks. Usually they were produced from the blanks where manufacturing the tip did not require much effort because of the direction of the edge.

103

104

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.12. Grydehøj, Thy. Perforators (photograph by M. Jórdeczka).

Figure 6.13. Grydehøj, Thy. Notched tools (photograph by M. Jórdeczka).

6. Cult features with flint assemblages from northern Jutland

Figure 6.14. Grydehøj, Thy. Burins (photograph by M. Jórdeczka; drawing by the author).

Two methods for producing perforators’ tips are evident. The first is retouch, sometimes alternate, of the edge of a flake or a chunk. Such perforators have more or less penetrating tips (Plate 17: 2, 5, 9, 10). The other method involved shaping the edge by retouching the flakes to form notches. This shows a certain similarity to the Clactonian retouch known from the Palaeolithic assemblages. In this way the tip appeared where the edges met, at the same time constituting an edge between two notches (Plate 17: 1, 6, 7, 10). The notched tools and denticulated pieces were produced in a similar way, and as a result the distinction between these categories of tools is relatively fluid. When the notches formed a sharp tip, the tools were identified as perforators (Plate 17: 6, 7, 10). The notched tools and denticulated pieces were made in the way described above. The edges formed by retouching flakes were concave, less frequently convex, and sometimes there were a few notches on such an edge, which rendered the tool a denticulated piece (Plate: 1, 3, 4). The tools made on chunks predominate among the denticulated pieces. Plate-like chunks with sharp edges produced by breaking up flint blocks were the best suited to create such extensive notched forms (Plate 18). The assemblage also includes tools identified as burins (Figure 6.14). The authors of the interim presentation of sources from the Grydehøj cult house treated this category as problematic (Masojć, Bech 2011: 215), especially in that this type of tool had so far never been observed in the Late Bronze Age assemblages of southern Scandinavia (Högberg 2009: 133). The issue of the presence and nature of burins in the Late Bronze Age assemblages certainly requires further study, but due to the fact that burins are present in other cult houses in Jutland (Høghs Høj, Torup Høje) and in central-European sites of similar chronology (cf. Zakrzów, Modlniczka), there is no

105

106

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

reason to negate the presence of this type of tool in the Scandinavian, Late Bronze Age assemblages. The burins from Grydehøj are solely single-blow forms. Two were made on chunks (Figure 6.14: 1, 3, photo). One of these (Figure 6.14: 3) and the specimen made on a flake (Figure 6.14: 2) are multiple forms with one of the tips made by several burin spalls. 6.1.7. Ad hoc tools The ad hoc tools are the most common type of tool from Grydehøj (Table 6.7). These include flakes and chunks, often large and amorphic, with irregularly retouched edges. These exhibit the traces of use in the form of chippings, frequently on the edges devoid of retouch. In addition to the retouched tools, a group of artefacts devoid of secondary modification by chipping was identified, which displays distinct traces of use – various types of chippings of the edge and gloss on the surfaces. These are flakes and chunks used ad hoc as tools. The tools where the traces of use were macroscopically visible are presented in the tables (Plate 19: 1, 4). The category of ad hoc tools also includes flakes with a retouched or natural backed edge – identified as ‘knives’ or knife-like flakes (Figure 6.15). To an extent, these products correspond to large backed knives produced in a standardised way and present in the assemblages from the Late Bronze Age. They may have emulated such knives or performed their functions; however, their production and technology does not differ

Figure 6.15. Grydehøj, Thy. knife-like flakes (photograph by M. Jórdeczka).

6. Cult features with flint assemblages from northern Jutland

from the remaining ad hoc tools. These are flakes with one edge formed as a backed one (or with a natural but distinctly visible backed edge), while the opposite edge was sharp, often chipped, displaying traces of use (Plate 19: 1, 2, 4). Similar artifacts were discovered at sites in Poland, where the production of flake backed knives was standardised (Bronowicki, Masojć 2008, 2010). Despite their ad hoc character, the ‘knives’ display a certain degree of consistency. No regular, blade backed knives occurring in settlement sites were found in the assemblage. With its character, typological and technological structure the flint assemblage from Grydehøj is to an extent similar to the settlement assemblages from the Late Bronze Age. The differences will be presented below. Some of the artefacts carry the traces of use. Their analysis will provide more insight into what exactly occurred within cult houses. 6.1.8. Spatial arrangement The interior of the Grydehøj cult house was excavated with the use of a metre grid. Each metre was subdivided into quarters, which became the basic spatial units within which the artefacts were stratigraphically documented. The sediments removed from the quarters of metres underwent flotation, which provided a detailed picture of the distribution of archaeological materials in the excavated area, including pottery and the predominating flint artefacts. As it was already mentioned, more than 5kg of pottery was found in the confines of the house. Its distribution examined within the quarters of square metres shows that it was concentrated in the entrance’s vestibule and inside the building around the semi-circular area forming a separate zone in the house. There was no pottery in the building’s central part (Figure 6.16). The spatial distribution of the flint material in the feature does not exactly overlap with the occurrence of the pottery (Figure 6.18-6.23). The flint artefacts, contrary to the pottery, were recovered from all areas of the house, in all the excavated quarters of metres. Their quantity varies between a dozen and more than three hundred in one quarter. The greatest number of flint artefacts recorded in a quarter approximated one thousand specimens. A large number of artefacts were located along the cult house’s whole northern wall, while their greatest concentration was recorded in the north-western corner. Towards the entrance to the building the frequency of flint material decreased, and at a distance of about a metre from it the frequency increased slightly, which overlapped with a greater amount of pottery found there. Flint artefacts and pottery did not occur at the cult house’s vestibule or in the tunnel of the entrance.

107

108

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.16. Grydehøj, Thy. Distribution of pottery inside the cult house; weight in grams (after: Nielsen, Bech 2004; amended by the author).

The frequency of the artefacts in individual quarters corresponds to the weight of the flint material, i.e. the places where the number of the artefacts is the greatest the raw material present is the heaviest. This suggests that the artefacts were not selected and deposited in separate zones. The analysis of the distribution of stones shows that it overlaps with the occurrence of the flint products. There is an obvious spatial relationship between the stones and flint artefacts. Is it a result of what took place inside the house? Maybe the stones were brought and deposited together with the flint blocks in one corner of the house and the pebbles were also used for some purposes in the place where the frequency of flint artefacts is the greatest.

6. Cult features with flint assemblages from northern Jutland Figure 6.17. Grydehøj, Thy. Cultural layer inside the cult object with visible numerous flint artefacts and pebbles (photograph by. J.-H. Bech).

Figure 6.18. Grydehøj, Thy. Quantity ranges of flint products within the cult house in individual quarters of square metres (from both horizons N17 and N18).

109

110

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.19. Grydehøj, Thy. Weight ranges of flint products within the cult house in individual quarters of square metres (from both horizons N17 and N18).

B. H. Nielsen’s and J.-H. Bech’s put forward a hypothesis that the numerous pebbles are the remains of the house’s inner construction – a stone bench running along the walls (or at least part of them). If their interpretation is accepted, the mutual relationship between the pebbles and the artefacts seems clear. The stone bench may have been a workplace, a seat for the people occupied with reduction of flint blocks. Thus flint

6. Cult features with flint assemblages from northern Jutland

Figure 6.20. Grydehøj, Thy. Quantity ranges in the category of chips from the cult house in individual quarters of square metres (from both horizons N17 and N18).

products were naturally deposited in the immediate vicinity of the bench. When the cult house was no longer used, its walls and the whole structure began to undergo the gradual deconstruction process. The stone pebbles probably ‘scattered’ and were deposited among the flint artefacts and above the level of their occurrence and were thus recorded in such a stratigraphic relationship during the excavations.

111

112

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.21. Grydehøj, Thy. Quantity ranges of blanks from the cult house in individual quarters of square metres (from both horizons N17 and N18).

6. Cult features with flint assemblages from northern Jutland

Figure 6.22. Grydehøj, Thy. Frequency of cores in individual quarters of square metres (from both horizons N17 and N18).

113

114

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.23. Grydehøj, Thy. Occurrence of tools in individual quarters of square metres (from both horizons N17 and N18).

6.2. Høghs Høj, Thy Situated near Thisted, the cult house at the Høghs Høj barrow, excavated in 1978 by J.-H. Bech, was the first fully documented cult structure in Thy and in Jutland (B. H. Nielsen, Bech 2004; Masojć, Bech 2011; Masojć et al 2013). The Høghs Høj cult house is a classic example from the area of Thy, even though it is not as well-preserved as the Grydehøj cult house.

6. Cult features with flint assemblages from northern Jutland

Figure 6.24. Høghs Høj, Thy. Cult house (bright background) at the foot of the barrow (darker background) (after: Nielsen, Bech 2004, 131, amended by the author).

The barrow’s diameter is ca. 25m. It was erected over a burial from the III period of the Bronze Age. In its mound there were secondary burials, including an urn grave from the IV period of the Bronze Age, excavated at the foot of the barrow on its south-western side. Linear stone features built directly at the stone ring marking the barrow’s base were found in the immediate vicinity of the cult structure (Figure 6.24). The features included one or two benches built from cobbles, surrounding a square area measuring approximately 6m in length and width. To the north of the bench there was another well preserved, linear stone pavement. The area between the bench and the pavement, measuring approximately 1.5m in width, had no stones but displayed darker oval areas, which may indicate the locations of the

115

116

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.25. Høghs Høj, Thy. Cult house (photograph by J.-H. Bech).

Figure 6.26. Høghs Høj, Thy. Cult house. View from the barrow (photograph by J.-H. Bech).

6. Cult features with flint assemblages from northern Jutland

building’s construction posts (Figure 6.25, 6.26). The southern portion of the feature had been destroyed by farming. It is quite probable that the outer stone pavement continued, surrounding the square inner area (B. H. Nielsen, Bech 2004: 129-131). A large quantity of flint artefacts and pottery dated to the IV and V periods of the Bronze Age were found in association with the stone pavement and the interior of square structure. The construction of the Høghs Høj cult house and its location in relation to the barrow display considerable similarities to the Grydehøj cult house. Both were constructed in the vicinity of a burial, which in the case of Grydehøj was a secondary burial situated at the barrow’s base. Positioning the features of a cult character next to the burials located in the barrows is a rule during the Late Bronze Age, not only in Thy but also in other regions – refer to the remarks concerning Torup Høje in the vicinity of Viborg. 6.2.1. Description of the assemblage The flint assemblage from the Høghs Høj cult house includes nearly 1600 artefacts (Table 6.8; Figure 6.27). It is four times smaller than the assemblage from Grydehøj. Table 6.8. Høghs Høj, Thy. Structure of the flint assemblage within the cult house N

%

cores/nodules of raw material

General structure of assemblage

503

31,5

tools

111

7,0

debitage

944

59,1

other forms

12

0,8

chunks

26

1,6

chips

0

0

1596

100

total Categories of tools Formal tools end-scrapers

18

16,23

perforators and borers

17

15,33

notched tools

1

0,9

side-scrapers

2

1,8

16

14,41

flakes/blades/retouched chunks

24

21,6

knives

4

3,6

functional implements

6

5,4

burins Ad hoc tools

Other unidentified tools/fragments

10

9

hammerstones

13

11,73

111

100

total

117

118

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.27. Høghs Høj, Thy. Fragment of flint collection from the cult house displayed in the museum in Thisted (photograph by author).

It was not as methodically excavated as the latter. The cultural layer containing the artefacts was removed without the use of the metre grid. The artefacts were recovered from the interior of the cult house and the area of the linear cobble benches, which results from the preservation of the house – its collapse and the secondary deposition of the artefacts outside its interior. The excavation methodology directly influenced the composition of the assemblage, which does not include artefacts measuring less than 2cm (chips). This is understandable, because during the excavation of the cultural layer neither sifting nor flotating was used; the artefacts were removed from the matrix by hand. The other, more intriguing aspect is the almost total absence of chunks (1.6% of the assemblage). Either the assemblage from Høghs Høj differs considerably from

6. Cult features with flint assemblages from northern Jutland

other flint assemblages associated with cult structures (32% of chunks in Grydehøj and 21% in Torup Høje), or the artefacts were selected, eliminating those which might have seemed unintentional.2 A preliminary examination suggests that the flint assemblage from Høghs Høj is identical to the assemblage from the Grydehøj cult house. The conclusions regarding the technology and morphological characteristics of the assemblage from Grydehøj have their close analogies here. The treatment of flint blocks and the method of producing blanks are comparable, as is the typological organisation of the assemblage. Breaking up the blocks, common in both assemblages, resulted in the production of a large quantity of chunks. Their almost complete absence in Høghs Høj may be explained by the selection of the assemblage during the excavation or following it. Adopting the same percentage contributions of chips (28%) and chunks (32%) as in Grydehøj, it may be assumed that originally the Høghs Høj assemblage included approximatly 2000 chips and the same number of chunks, and thus it would have numbered approximately 6000 artefacts. Considering that more than five hundred cores were reduced at Høghs Høj (503 specimens) and nearly the same number at Grydehøj (509 specimens) and bearing in mind that they were technologically identical, it may be assumed that both assemblages originally were the same size (i.e., approximately 6000 artefacts). This suggests that only a portion of the lithic assemblage from Høghs Høj has survived. 6.2.2. Raw material Similar to Grydehøj, the vast majority of the studied assemblage was made from Scandinavian Senonian flint, which may have been acquired opportunistically from the nearby places where it occurred, that is picked from the gravel of Scandinavian origin. Due to the fact that it has only partially survived, the assemblage was not weighed. 6.2.3. Technology Flint blocks at Høghs Høj were reduced with the use of a hard hammerstone. Quartzite pebbles carrying traces of chipping on the circumference, some of quite an impressive size, are at Høghs Høj the most numerous off all the assemblages from the Late Bronze Age cult features analysed in this work (13 specimens). Flint concretions without preliminary preparation (only 12 artefacts exhibit variable preparation) were broken into smaller fragments, producing a small number of flakes sensu stricte from a block. Fragments of the concretion were subsequently used to produce chunks and flakes analogous to the subsequent stages. Jens-Henrik Bech confirmed recently (verbal communication) that around 1980-1981, when the material from Høghs Høj was washed, sorted and catalogued, all the fragments which did not exhibit unambiguous characteristics of intentional production, including most chunks, were removed from the assemblage.

2

119

120

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.28. Høghs Høj, Thy. Structure of the flint assemblage from the cult house. I – raw material; II – early phase of coring; III – advanced exploitation; IV – final exploitation; V – other flakes and unidentified products; VI - tools. Note scanty frequency in the category of other flakes (non-cortical chips, chunks), which results from fragmentary state of the assemblage’s preservation

Graphical representation of the sequence of reduction evidenced by the assemblage (Figure 6.28) differs from the similar representation for the Grydehøj cult house (Figure 6.7) and for the ritual zone in Torup Høje (Figure 6.47). In the latter two, with the exception of the category of the debitage, the products of the early stage of core reduction, i.e. flakes and cortical chips (more than 50% of cortex) as well as initial cores, predominate, while the group of advanced core reduction predominates at Høghs Høj where debitage is present in very small amounts (which was explained above). This situation indicates that at Høghs Høj cores were subjected to later stages of reduction resulting in the production of a larger number of non-cortical flakes and blades, which would exhibit dorsal flake scars from later reduction stages, traces of preparation, repair, change of orientation, etc. However, such reduction did not take place at Høghs Høj, and the described situation is a result of absence of chips (cortical ones in this case), which in both assemblages from cult zones constitute one of the main components of group II of the graphical representation (Figure 6.28). As at the other sites, the preliminary reduction of cores was most common at Høghs Høj, leading to the production cortical blanks, after which only very few were not discarded. Advanced reduction of cores was limited to a much smaller group. What attracts attention is the small number of artefacts in the debitage category (noncortical chips, chunks), which results from the fact that the assemblage has been only partially preserved. 6.2.4. Cores and blanks At Høghs Høj, similarly as at Grydehøj, approximately40% of cores are amorphic in shape –evidence that fracturing flint blocks was a coincidental process (Table 6.9). The remaining main categories of cores also correspond to those most commonly

6. Cult features with flint assemblages from northern Jutland Table 6.9. Høghs Høje, Thy. Cores within the cult house Categories of cores

Total N

%

initial core

25

4,9

single-platform flake core

108

21,5

flake core with equivalent flaked surfaces/striking platforms

7

1,4

blade core with equivalent flaked surfaces/striking platforms

-

-

non-prepared flake disc core

95

18,9

prepared flake disc core

17

3,4

block-like core with changing orientation

22

4,4

flake amorphous core/fragment

218

43,3

exhausted core

11

2,2

total

503

100

represented at Grydehøj: discoidal, unifacial unprepared cores (Plate 20: 1) and singleplatform flake cores (Plate 21: 4). These three predominant categories of cores are evidence of the producers’ minimal efforts to produce blanks through the reduction of cores (Figure 6.29). They did not require any preparation treatment. A block’s natural shape determined the manner and form of reduction. Tabular cores (Plate 21) with flat surfaces (or only one surface) were used for single-platform reduction. Discoidal cores, frequently relatively flat, were chipped along their circumference to produce a few cortical flakes (Table 6.9). If they were discoidal and relatively thick, bifacial radial reduction was utilised, which is treated as a core with preparation, but in fact it is an opportunistic use of a flake scar as a striking platform. If a block was neither discoidal nor tabular but amorphic, one or a few flakes were detached in a completely opportunistic manner before it was discarded. The remaining categories of cores account for less than 5% of the entire category of cores. No flake cores were observed. The presence of a several dozen flakes in the assemblage should be not interpreted as attempts of acquiring blanks of appropriate proportions – their blade-like proportions are a coincidence of the shape of a flint block. This is substantiated by cortical blades (greater than 50% cortex on the dorsal side) blade-like flakes contributing as much as 50% to the blade category. The method of reduction is reflected by the category of blanks from Høghs Høj (Table 6.10). It is dominated by flake blanks (90%). Cortical flakes make up half of the debitage (43% in Grydehøj, 45% in Torup Høje). The second largest category includes flakes with flake removal scars on the dorsal surface consistent with the direction of the flake’s removal from the reduction of single-platform cores (27%, Grydehøj – 38%, Torup Høje – 44%).

121

122

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Table 6.10. Høghs Høj, Thy. Blanks within the cult house N17+N18 N %

Categories of blanks cortex flake >50%

473

49,5

cortex blade >50%

41

4,3

preparation products

12

1,3

flake with removals on dorsal face consistent with the direction of flake’s removal

265

27,7

flake with removals on dorsal face opposite to the direction of flake’s removal

14

1,5

flake with removals on dorsal face oblique to/shifted to the direction of flake’s removal

31

3,2

multidirectional flake

71

7,4

blade with removals on dorsal face consistent with the direction of flake’s removal

49

5,1

-

-

956

100

two-directional blade total

Figure 6.29. Høghs Høj, Thy. Selection of cores (photograph by M. Jórdeczka).

6. Cult features with flint assemblages from northern Jutland 6.2.5. Tools The category of tools in the assemblages from the Late Bronze Age is limited to several types of formal tools and a wide range of ad hoc tools. The types of tools persist despite the decrease in the number of forms. They are usually end-scrapers, perforators and borers, notched tools, burins (!) and – typical of the settlement in this region – large backed knives. Differences between individual assemblages result from the absence of certain

Figure 6.30. Høghs Høj, Thy. End-scrapers (photograph by M. Jórdeczka).

123

124

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.31. Høghs Høj, Thy. Perforators (photograph by M. Jórdeczka).

elements or varying proportions among individual categories of tools. The Høghs Høj cult house is a distinct example of differences from the perspective of the settlement type. Ad hoc tools constitute almost 1/3 of all the tools from Høghs Høj (Table 6.8). Formal tools constitute approximately 45%. In an assemblage consisting of over a hundred artefacts, no

6. Cult features with flint assemblages from northern Jutland

more than 1/10 are hammerstones used in flintworking. Three types of tools predominate the category of formal tools: perforators (Figure 6.31; Plate 24), burins (Figure 6.32; Plate 25) and end-scrapers (Figure 6.30; Plate 26). The end-scrapers occur in two forms: short and wide forms, often semi discoidal (Plate 26: 2, 3, 5-9) and large flakes with convex ends, usually made on cortical flakes (Plate 26: 1, 4, 10-12). The quantity of end-scrapers is interesting – their 16% contribution is the greatest in comparison with the remaining assemblages from cult structures (5% in Grydehøj and ca. 2% in Torup Høje). The perforators, consisting of 17 artefacts, displays similarities to those present at Grydehøj and Torup Høje. They usually have prominent tips (Figure 6.31; Plate 24). The tendency to utilise the natural shapes of the blanks (flakes or blades) on which the tip was formed is clearly visible. They are carefully made. The assemblage does not include the perforators made with the use of notched retouch. Retouch along the edges was either unifacial or alternate. The third largest category of formal tools includes the burins (Figure 6.32; Plate 25). So far they are the most numerous collection of this tool category, among both the assemblages from the cult structures discussed here and all sites from the Late Bronze Age in southern Scandinavia. As mentioned above, burins have not been recovered from the settlement sites in the region. Individual burins were found at Grydehøj (3 specimens) and at Torup Høje (1 specimen). At Høghs Høj they constitute a full set, being one of the most numerous categories of tools. The category of burins includes single-blow burins (Plate 25: 3-6, 8-11) and dihedral burins (Plate 25: 1, 2, 7). No burins on retouched ends or burins on broken ends have been recorded. Similar to flintworking as a whole discussed here, including the formal tools (e.g. perforators), the selection of blanks of appropriate shapes and proportions was important. Short, wide flakes or blades with relatively wide sides were used for their production, which also facilitated the detachment of burin spalls. In the assemblage I did not record the flakes that could be unambiguously identified as debitage from the production of burins. Assuming that we are dealing with several dozen burinated tools, it seems that burin spalls should be easily identified in the assemblage, especially given that their disparate proportions would have distinguished them in the collection. This leads one to believe that the burins were not manufactured at, but rather brought to the cult house. No burin spalls were identified in any of the similar assemblages. Notched and denticulated tools, so common in other assemblages, are only occasionally identified at Høghs Høj. This, together with the large number of burins, so far never encountered in such numbers, and an exceptionally represented category of endscrapers indicates that it is a considerably different set of tools. Perforators are the only common element. It is difficult to determine whether or not this is evidence of functional diversity of the assemblages from the cult structures.

125

126

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.32. Høghs Høj, Thy. Burins (drawing by the author, photograph by M. Jórdeczka).

Two artefacts were identified as side-scrapers (Plate 23: 2, 3). They are on similar flakes with one retouched lateral edge. Even though they fall into the category of retouched flakes, their regularity and the nature of their retouch makes them formal tools, as does a single notched tool with four separate notches made on a chunk (Plate 23: 1).

6. Cult features with flint assemblages from northern Jutland

Figure 6.33. Høghs Høj, Thy. Ad hoc tools – a backed knife on chunk (photograph by M. Jórdeczka).

Several artefacts identified as knives belong to the large category of ad hoc tools with flakes displaying traces of use in the form of various types of chipping, which is typical of the Late Bronze Age (Plate 27, Fig, 6.33). These should not be confused with standardised larged backed knives found at other settlements. Here, the artefacts described are not blades, they often are chunks. They have backed edges (usually natural, but there are specimens with backed edges modified by retouch) and a parallel sharp edge, always bearing traces of chipping or polishing. These artefacts are included within the category of ad hoc tools but because of a certain degree of similarity to large backed knives they were treated separately, yet still remaining within the group. Standardised short flake backed knives, which the author observed during excavations in southwestern Poland (site Zakrzów 41 – cf. Bronowicki, Masojć 2010), are unknown in southern Scandinavia. The above-mentioned examples from Høghs Høj are similar to the knives from Zakrzów, despite the ad hoc nature of their production. As was the case for Grydehøj, the Høghs Høj assemblage does not include what have been identified as strike-a-lights and their semi-finished forms, but it does include numerous hammerstones, whose presence indicates that reduction of stone blocks took place within the cult structure, despite the absence of chips in the assemblage. Selected artefacts from Høghs Høj were subjected to optical microscopy analysis. The results of the microscopic analysis examining use wear of a selection of artefacts from

127

128

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

the cult house will be presented in another part of this work together with the analysis of artefacts from other cult features. 6.3. Gramstrup I, Thy The Gramstrup barrow is situated in a slightly hilly area, with an extensive view of a small bay of western Limfjord, Krik Vig, and the North Sea, barely a few hundred metres to the north-west of the Grydehøj barrow. Because of constant threats of ploughing the barrow, rescue excavations were carried out, which revealed a cult house situated at the barrow’s base (Figure 6.34; 6.35). The barrow, constructed during the Neolithic (Single Grave Culture) and reaching a diameter of approximately 25m during the Early Bronze Age (probably in the II or III period of the Bronze Age), contained six to seven phases of use (B. H. Nielsen, Bech 2004: 141-145). It has many similarities to the neighbouring Grydehøj barrow, while at the same time displays certain differences, such as in the construction of the cult house. The cult house, situated on the barrow’s eastern side, was approximately 8 x 13m in size (Figure 6.34: A, B). On the southern and northern sides it was flanked by stone benches. The northern portion was more complete, with clearly visible posthole features on the inside. The structure was built from cobbles. On the eastern side its foundation turns sharply at a right angle, originally forming a stone enclosures of an interior construction. This type of flanking is analogous to that of other cult houses from Thy, including the closest houses at Grydehøj and Høghs Høj. However, contrary to the structures mentioned above, the construction situated inside the stone enclosure at Gramstrup was apsidal in its outlines, not rectangular. The perimeter, whose diameter measured approximately 7m, was marked by the arrangement of cobbles and their negatives. B. H. Nielsen and J.-H. Bech observed that in principle the construction could have been the remains of another, unpreserved barrow if it had not been for its different internal structure and the flanking stone benches (2004: 142). Another question is whether the structure in Gramstrup was actually a building as was probably the case in Grydehøj. Similar to the latter, the remains of an approximately 1.5-metre-wide peat and earthen wall were situated inside the stone enclosure. They were recorded along the inner stone enclosure and from the side of the barrow, overlying its older layers. The stone enclosure along the peat wall could have been a kind of bench, a platform, a place of rest or procession. Unfortunately, it is difficult to reconstruct it on the features’s eastern side, where the entrance to the building was likely located. One concentration of stones (Figure 6.34: B, 6) and a posthole feature with a stone placed inside (B, 4) were recorded inside the cult house’s apsidal structure. The remains of post features, especially those next to the northern stone enclosure (B, 3) and in the house’s central part (B, 4) may indicate that the structure had a roof, as was similarly assumed in the case of Grydehøj. It is also quite possible that the house was an open

6. Cult features with flint assemblages from northern Jutland Figure 6.34. Gramstrup I, Thy. Barrow and cult house. A – projection of the barrow with the cult object on the east; B – projection of the cult object: 1 – cremated burial grave; 2 – negatives of pebbles; 3 – a row of postholes; 4 – post object; 5 – remains of the walls built from peat; 6 – stone mound; 7 – approximate location of the big backed knife (after: Nielsen, Bech 2004, 131, amended by the author).

129

130

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.35. Gramstrup I, Thy. Outlines of the cult house and the barrow during exploration (photograph by B. H. Nielsen).

Figure 6.36. Gramstrup I, Thy. Backed knife from the cult object (drawing by the author; photograph by M. Jórdeczka).

structure without a roof, with low walls and direct access to the grave at the barrow’s base. Unlike the Grydehøj cult house, very few artefacts were found inside the cult house. There were only a few pottery fragments in the vicinity of the northern stone bench.

6. Cult features with flint assemblages from northern Jutland

A large backed knife, a tool characteristic of the Late Bronze Age, and several flint artefacts without any characteristic features were also found there (Figure 6.34: B, 7; 6.36). There was one grave from the Late Bronze Age at the barrow’s base, near the cult feature. It was excavated into the barrow’s mound about a metre to the west of the cult house (Figure 6.34: B, 1). It was a cremated burial in a shallow pit that measured 1.2 x 0.75m. At either end large cobbles were situated at the bottom of the pit. The feature’s fill included a thick layer containing burnt bones. Neither osteological analysis of the faunal remains from the grave nor radiocarbon dating were carried out. The large backed knife from the cult feature was preserved in its entirety. It was made from Scandinavian Senonian flint. Its measurements are 107 x 34 x 18mm (Figure 6.36). It belongs to the category of knives of the C type according to the classification proposed in this work (cf. chapter 7.2). The knife’s backed edge, slightly rounded on the distal end, is retouched along the entire length of its lateral edge with a blunted, abrupt retouch. The remains of the preparation of the striking platform are visible. It also exhibited distinct polishing resulting from its use. Microscopic observations confirmed the traces of its intensive use. It displays traces of wear typical of this type of artefacts known from settlement sites from the Late Bronze Age. This aspect will be discussed in more detail in the chapter devoted to the functions of flint artefacts from ritual zones (chapter 8). 6.4. Ginnerup, Thy In 2008 the Museum for Thy og Vester Hanherred in Thisted carried out the excavations of the Ginnerup barrow in south-eastern Thy, east of the locality of Hurup. The results of the recently completed work are still being prepared, as is the processing of the artefacts, and therefore their full potential cannot be known. However, preliminary observations seem to confirm that the Ginnerup barrow is another example of a cult feature from the Late Bronze Age which had a large flint assemblage. In view of the remarks above, only very general characteristics will be presented here. The Ginnerup barrow was constructed during the Early Bronze Age; no traces from earlier periods were recorded. The central burial, placed in a stone enclosure in the place of cremation was a typical grave from the Early Bronze Age, even though it lacked elements allowing for unequivocal dating. The fieldwork at the foot of the barrow, on its eastern side, revealed arrangements of stones, i.e. the remains of a cult house (Figure 6.37). Two approximately 1-metre-wide benches are clearly visible among the remains(Figure 6.38). The better-preserved bench, on the northern side, is approximately 5 metres long, while the southern bench has survived only at the length of approximately 2 metres. Between them there was an area measuring 5-6 metres, within which there was an arrangement of stones approximating a quadrangle. This structure

131

132

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.37. Ginnerup, Thy. Plan of the barrow and the cult object located on the barrow’s eastern side as well as another cult object on the barrow’s southern side (drawing by Mette Roesgaard Hansen, amended by the author).

Figure 6.38. Ginnerup, Thy. Cult object on from the barrow’s eastern side (photograph by M. R. Hansen).

6. Cult features with flint assemblages from northern Jutland

was separated from the stone benches by a strip without stones whose width was approximately 1 metre. Both the benches and the structure between them came into contact with the barrow, and they were best preserved there. The opposite side, the structures eastern portion, was considerably damaged by farming activities. Loose cobbles were recorded there, whose arrangement could not be unambiguously interpreted. A detailed description of the stone arrangements within individual stratigraphic units has not yet been completed. Based on its similarity to Grydehøj, the structure – a quadrangle arrangement of pebbles between the benches – is most likely the remains of another cult house dating to the Late Bronze Age. It is interesting that no secondary burials were recorded in the barrow that could be directly associated with the cult feature. The described cult structure provided numerous flint artefacts (Figure 6.40). A more detailed analysis of their distribution has not yet been completed. There were Figure 6.39. Ginnerup, Thy. Selected also pottery fragments generally dated to artefacts from the cult house. Perforators and denticulated tools the Late Bronze Age. The experience gained (drawing by the author). during previous investigations resulted in the excavations being carried out in such a way that the entire flint assemblage was collected; however, it has not yet been catalogued. In December 2012 I made a preliminary survey of the flint assemblage. It is curated in 24 cardboard boxes, each containing several hundred artefacts; the entire assemblage may be estimated to include more than ten thousand artefacts. It includes chips and flintworking tools – quartzite hammerstones bearing the traces of use. I have not found many formal tools and I have not seen any large backed knives. In terms of technology the assemblage is identical to the other assemblages from cult houses in Thy. It was made from cobbles and chunks from Scandinavian Senonian flint. This superficial examination revealed the presence of a dozen or so tools, the most interesting of which included perforators, notched and denticulated tools made mainly on chunks (Figure 6.39). I did not find any burins,

133

134

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.40. Ginnerup, Thy. Flint assemblage within the cult house during exploration (on the left) and a small fragment of the assemblage during museum inventory taking (on the right) (photograph by M. R. Hansen).

but this does not rule out their presence. Blanks and chunks exhibiting macroscopically distinct traces of use were identified. The number of cores and large quantity of chunks are similar to assemblages from other cult houses. The preliminary analysis of the flint assemblage and the context of its discovery indicate that the remains of another cult house within which intensive flintworking took place during the Late Bronze Age was located at Ginnerup. When looking at the plan of the barrow and its vicinity, one may see that poorly preserved remains of another stone structure were recorded in the base of the barrow’s southern portion, approximately 10 metres to the south-west of the cult house (Figure 6.37, blue background). The preserved remains include a concentration of cobbles and a fragment of a stone bench measuring approximately 2-3 metres in length, similar to that from the cult house. Both were difficult to interpret. Given their sizes and fragmentary preservation, they are reminiscent of the features discussed above. Its condition prevented a detailed description, but it seems sensible to assume that another cult object was situated there, whose construction has already been recorded several times in Thy. Next to the feature there was a cremated burial grave inside wooden enclosure placed in the barrow’s mound. There is no information about presence of flint artefacts in the area of the ritual feature. 6.5. Torup Høje, Fjelsø The group of Torup Høje barrows is located in central Jutland, approximately 25 km to the north of Viborg, just to the south of Fjelsø. Unlike to the cult features from Thy – cult houses usually occurring individually – in the vicinity of Viborg the location

6. Cult features with flint assemblages from northern Jutland

of many smaller cult features of variable arrangement seem typical (M. Mikkelsen in print). The best excavated example of such a cult feature is the Torup Høje barrow. The ritual zone identified there provided a rich assemblage of flint artefacts (Masojć et al. in preparation). Torup Høje is group of five barrows situated on the slope of a small hill. The discussed barrow (no. 60) is the easternmost, which makes it clearly visible on the landscape from the north, east and south. Because of this it was the most exposed to plundering. At the request of the owner of the land, who wanted the barrow to be removed to use the land for farming, excavations were carried out in 2002. The excavations were carried out by M. Mikkelsen from the museum in Viborg.3 Rich in experience of the excavation of cult structures at the barrows in Thy, he carried out most of the excavations by hand with little use of mechanical equipment. The method resulted in the discovery of many stone structures and secondary burials within the barrow and completely new, until then unknown, structures at the base of the barrow. The discussion of the results of the excavations below is based on the future publication of the research by the author. Before the barrow’s ritual system is described, it is noteworthy to mention that prior to its utilization as a burial place during the Late Bronze Age, it had had a long history of use. A relatively low barrow with a diameter of approximately 15 metres was constructed during the Neolithic Single Grave Culture around a grave in a stone enclosure. There are four additional graves there, belonging to earlier communities from the Funnel Beaker Culture. On top of the barrow, another, steeper barrow with a diameter of approximately 11 metres was constructed around a pit grave, probably associated with the Late Neolithic or the early period of the Bronze Age (period I-II). During the Early Bronze Age it was amplified, measuring 17 metres in diameter. To summarize, prior to the appearance of the structures and graves of the Late Bronze Age, the barrow had undergone at least four phases of use and extension. The excavations showed that the place ceased to perform funerary functions during the Early Iron Age. Numerous urn graves were deposited within the barrow’s mound and in its vicinity during the Late Bronze Age and probably also in the pre-Roman period. Ten grave urns were discovered in the south-western zone of the barrow’s base. There are reports from the 1950s of a large number of urns plundered there. M. Mikkelsen estimates that their number could have exceeded 20. Beyond the barrow, in its vicinity another 6 grave urns of pit cremated burial graves were discovered, but their definitive number cannot be determined. In addition to the burnt human remains, the grave urns contained Complete results of the excavations were presented by Martin Mikkelsen in 2012 during a session in the museum in Viborg in the address ‘Dødekulten i yngre bronzealders lokale kulturlandskab’ (The cult of the dead in the cultural landscape of the Late Bronze Age).

3

135

136

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.41. Torup Høje, Fjelsø. Barrow during exploration in 2002 (photograph by M. Mikkelsen).

only a few other artefacts: five of them contained small fragments of amber and a few contained isolated bronze objects. In grave no. 49, situated in the barrow’s mounds, a grave urn contained two flint artefacts. They were discussed in one of the previous chapters devoted to the flint artefacts from the Late Bronze Age graves (chapter 4.2.4). In the zone of the width of ca. 5 m from the base of the barrow’s mound there were remains of a number of structures of various construction styles, which were interpreted as cult features accompanying urn graves. They can be subdivided into four different types: a b c d

vertically arranged boulders, ditch features (foundation ditches), pairs of posthole features, mounds of cobbles.

A rich flint assemblage was also found in this zone. The mentioned artefacts were in a direct relationship with the urn and cremated graves secondarily placed in the barrow’s base. Their characteristics will be presented below. 6.5.1. Arrangements of vertically positioned boulders In the southern zone of the barrow’s base there was a group of 4-5 vertically arranged boulders (Figure 6.42: 3; 6.44: A), which were excavated into the barrow’s base. Their arrangement displays a slightly arching direction, leaning towards the barrow’s centre. A pile of smaller stones was situated in front of the boulders, while between the two there were isolated pottery sherds from the Late Bronze Age. Behind the boulders two urn graves, contemporary with the pottery, were found. A similar arrangement of vertically positioned boulders occurred in the eastern part of the barrow’s base. (Figure 6.42: 2; 6.44: B). It was considerably damaged by plundering, and as a result the features context was destroyed. Similar to the previous example of the

6. Cult features with flint assemblages from northern Jutland .

Figure 6.42. Torup Høje, Fjelsø. Plan of the barrow with the location of secondary burials and ritual constructions around the barrow. 1-12 – cult objects of the ritual zone from the late Bronze Age (1-3 – vertical arrangements of boulders, 4-6 – piles of pebbles, 7-10 – pairs of post objects; 11-12 – fluted objects);I-III –3 phases of extension of the barrow in the Neolithic and early Bronze Age (højfase); A-D – the oldest graves in the layout from the Funnel Beaker culture; Stenkiste – box-shaped grave from the early Bronze Age; LBA – urn graves and pit cremated burial graves from the late Bronze Age; grey hatching – area destroyed mainly by plundering in the modern times; prima grav – Neolithic grave (Single Grave culture); secundær grav – graves from the early Bronze Age; lerkargrav – grave from the early Iron Age (after: Mikkelsen in print).

arrangement of stones, this one is also slightly curved towards the barrow’s centre. The third such feature occurred in the western zone of the barrow’s base, where a line of boulders was recorded (Figure 6.42: 1). There the context had also been destroyed by plunderers. 6.5.2. Ditch features Two ditch features, intersecting one another, were discovered in the south-western zone, at the foot of the barrow (Figure 6.42: 11, 12). Their mutual stratigraphic relationship

137

138

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.43. Torup Høje, Fjelsø. Cross-section of the barrow and accompanying zones. R Z – ritual zone where cult objects were located (from SW it reached up to 9 m from the barrow’s base); UG Z – zone of ca. 3 m of locating urn burials in the late Bronze Age in the barrow’s base; GZ – zone where other cremated urn (pit) graves occur), considerably destroyed and excavated to a small degree (after: Mikkelsen in print).

was impossible to determine, except for the fact that they were not contemporary. The smaller feature, best preserved on the southern side, was semi-circular. Its width at the barrow was ca. 4 m. It is difficult to say whether a gap in the course of the ditch, from the south, was the remains of the entrance. In the eastern part, the fill of the ditch contained stones. The smaller semi-circular feature (Figure 6.42: 11) cut across, or was cut across, by a similar ditch feature (Figure 6.42: 12). It was well preserved in the southern part, but the rest was destroyed by excavation in modern times. If the interpretation of the surviving portion of the feature on the northern side is correct (adjacent to the cult feature – no. 10 in Figure 6.42), the structure would originally have had a more or less rectangular shape with a width of approximately 9m. The quadrangular arrangement of stones situated inside the feature – a symbolic box grave – was identified as a cult feature from the Early Bronze Age unassociated with the later feature. 6.5.3. Pairs of posthole features In the barrow’s ritual zone there were four features consisting of pairs of posthole features situated next to each other. The author of the research identified their function as the remains of a construction, possibly an altar. They are supposed to be the evidence of the existence of a feature: a wooden board supported by two wooden posts. They were all situated at the barrow’s base; they had similar sizes, depths and fills. The distance between the features in a pair was approximately 0.4 metres. (Figure 6.45). 6.5.4. Mounds of cobbles Accumulations of small stones were found at the barrow’s base to the south and south-west, where no substantial damage was recorded. Their arrangement was often coincidental and unintentional. However, in a few places they formed distinct

6. Cult features with flint assemblages from northern Jutland

Figure 6.44. Torup Høje, Fjelsø. One of the types of cult objects at the barrow – arrangements of vertically placed boulders (photograph by M. Mikkelsen).

concentrations, e.g. a group of stones in the western zone (Figure 6.42: 4). Three places near the barrow were identified where the concentrations of cobbles were interpreted as intentional features (Figure 6.42: 4-6). ***

139

140

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.45. Torup Høje, Fjelsø. Postholes occurring in pairs interpreted as the remains of the cult object (altar?) (photograph by M. Mikkelsen).

The four types of cult features were situated at the barrow’s base. They were interpreted to be associated with the Late Bronze Age. They were accompanied by pottery. A large collection of flint artefacts was found around them, including cores, blanks and tools undoubtedly affiliated with the late period of the Bronze Age. Flintworking took place at the location, which is substantiated by the presence of hammerstones displaying traces of use and a large number of chips that would have been produced during the reduction of cores. When looking at the location of the described cult installations, one can see clearly that they are grouped in the southern zone of the barrow’s base (Figure 6.42). Altogether there are twelve features: three structures of vertically arranged boulders, two ditch features, four pairs of postholes and three concentrations of cobbles. The last three occurred only in the south-western zone, where there was the least post-depositional disturbance. It may thus be concluded that originally there may have been more features in the southern part. This is substantiated by the numerous cobbles found there, which (following deposition?) lost their clear arrangement. If we assume that the pair of postholes and group of stones from the interior of the smaller ditch feature had formed one complex cult structure, there would be only 10 cult features at the Torup Høje barrow. According to N. Mikkelsen, all the described features are ritual installations associated with individual urn graves excavated into the base of the barrow’s mound. This is especially clear in the well-preserved area of the south-western zone, where a few metres behind each cult structure there is at least one urn grave situated on the axis between the centre of the cult features and the centre of the barrow (Figure 10.2). The accuracy of this observation concerning the spatial relationship between a cult feature and an urn grave was later confirmed during the excavations of the Nøragergård

6. Cult features with flint assemblages from northern Jutland

Høje II (M. Mikkelsen in print). Another confirmation comes from thecult houses in north-western Jutland, for example Grydehøj, where the relationship between the cult structure and sepulchral objects excavated into the barrow’s mound is very clear (B. H. Nielsen, Bech 2004). Analysing the spatial arrangement of graves within the barrow’s base and outside it as well as the arrangement of the cult structures at the barrow’s base, Mikkelsen proposes distinguishing three to four main zones in the entire Torup Høje barrow (Figure 6.49): • the barrow’s central part as the zone without secondary graves from the Late Bronze Age, which may result from the special treatment of the barrow’s core; a considerable part of this zone was destroyed by looters’ trenches, but in more than 20% of the well-preserved area no urn graves were found; • the funerary zone, inside the barrow’s southern base – the area of approximately 3 metres in width – within which urn graves were deposited during the Late Bronze Age; • the ritual zone – a strip measuring 5 metres in width at the barrow’s base, where all the described ritual installations were situated; and • the funerary zone: outside the barrow, to the south and south-east from the ritual objects, where urn graves were also found; more precise determination of its range is difficult due to extensive damage of the area (exploitation of sand). Torup Høje is situated approximately 200m to the east of the nearest settlement, where the remains of a three-aisle house from the Late Bronze Age have been identified. The house stood on the same hill on which the group of barrows is situated. The settlement’s residents may have used the Torup Høje barrow as a place for burying the dead. 6.5.5. Exploration sectors During the investigation of the barrow, and especially its ritual zone associated with the Late Bronze Age, a large deposit of flint artefacts around the cult features was observed. The cultural layer within the metre grid was excavated in the places of their occurrence. The layer’s thickness varied between 2-3 and 10cm, relative to the degree of damage by modern activities. The rescue nature of the excavations prevented the examination with this method of the entire area determined to be ritual. The sectors were placed in such a way that they at least partly covered each of the four identified types of cult features (Figure 6.46). Investigation of the cultural layer of the individual square metres within the sectors was carried out with small excavation equipment. The location of the artefacts identified as tools or cores was documented as they were discovered, providing a general plan of artefact distribution. All sediment from the metres was sifted mechanically at the site; some of it was also floated. This contributed to the completeness of the recovered assemblage, which included a large number of the smallest artefacts.

141

142

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

The systematic exploration within the metric grid was carried out in four sectors. In sector A the exploration reached the base of the barrow. The area of the identified ritual zone at the foot of the barrow comprised the remaining sectors. 6.5.5.1. Sector A This sector was situated in the south-western zone, in the vicinity of the cult feature in the form of a concentration of cobbles – one of the three identified features of this type.

Figure 6.46. Torup Høje, Fjelsø. Location of sectors A-D within cult objects. The table presents surface areas of sectors, weight and quantity of flint assemblages in individual sectors including the contribution of burnt artefacts (after: Mikkelsen in print, amended by the author).

6. Cult features with flint assemblages from northern Jutland

It partly comprised the feature, extending into the base of the barrow and reaching the urn grave deposited in it. The excavation covered 4m2. 6.5.5.2. Sector B This sector was situated in the barrow’s southern zone, approximately 5 metres to the south-west of sector A. The sector comprised two cult features: a pair of postholes and another concentration of pebbles. The excavation covered 2.5m2. 6.5.5.3. Sector C This sector was situated approximately 1 metre to the east of Sector B and included the interior of the smaller ditch feature, two pairs of postholes and the area adjacent to the group of vertically arranged boulders. The excavation covered 2m2. 6.5.5.4. Sector D This sector was situated the further east, approximately 1m from Sector C in the place where the two ditch objects intersect, comprising the eastern edge of the smaller one and the western zone of the interior of the larger one and extending to the stone structure of the Early Bronze Age grave. The excavation covered approximately 9m2. 6.5.6. Description of the assemblage The four investigated sectors provided more than 3500 flint artefacts (Figure 6.46; Table 6.11). The flint assemblage weighs 15.03kg. It was acquired from an area measuring approximately 18m2, with an average of 200 artefacts per each explored metre. In Sector A the number of artefacts in two square metres approximated 400, while it exceeded 700 in the richest sector.

Table 6.11. Torup Høje, Fjelsø. Structure of the flint assemblage Zone A

Zone C

Zone B

Zone D

Total

Category

N

%

N

%

NN

%

N

%

N

%

raw material

5

0,3

-

 

3

1

6

0,5

14

0,4

cores

35

2,1

7

2,4

10

3,4

49

3,8

101

2,9

tools

43

2,6

6

2,1

10

3,4

32

2,6

91

2,6

flakes

391

23,5

36

12,6

57

19,6

250

19,6

734

20,9

blades

13

0,8

-

 

3

1

13

1

29

0,8

chips (less than 2 cm)

885

53,1

212

74,5

137

47,2

550

43,1

1784

50,7

chunks

293

17,6

24

8,4

71

24,4

375

29,4

763

21,7

Total 1665

100

285

100

291

100

1275

100

3516

100

143

144

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.47. Torup Høje, Fjelsø. Structure of the flint assemblage from the barrow’s ritual zone. I – raw material; II – early phase of coring; III – advanced exploitation; IV – final exploitation; V – other flakes and unidentified products; VI - tools.

Chips constitute more than half of the collection. Chunks make up 1/5 of the assemblage, which is the result of fracturing the blocks of raw material with a hard hammerstone – an activity typical of the Late Bronze Age, which detached blanks but also produced a large quantity of fragments with sharp edges. As such, they were attractive for the producers. The collection included approximately 100 cores (almost 3% of the collection). The values for the entire collection from Torup Høje quoted above were more or less the same in all 4 explored sectors. The assemblage’s general typology is presented in Table 6.11 and in Figure 6.47. The typology approximates that of the assemblages from the cult houses in Thy, where the groups of preliminarily reduced cores (II) predominates, while the group of debitage (V), consisting mainly of non-cortical chips and chunks, contribute significantly. 6.5.7. Spatial arrangement Because the excavated 18m2 constitutes merely 10-15% of the site’s area, estimating the area where flintworking activities took place is difficult. If it is assumed that they were as intensive within the whole ritual zone as they were in the excavated portion, it is possible that a large flintworking workshop was located in the ritual zone for either a long time or multiple limited periods of time. The data acquired from the excavated metres seems to confirm that. I will return to this issue in the next chapter, remarking now that the produced flint artefacts are a small part of the assemblage from the Torup Høje ritual zone. This is different from the cult structures in Thy, where the investigated zones had a strictly limited range, while flint artefacts occurred outside them only occasionally. Sector A seems the most interesting in terms of the distribution of artefacts (Figure 6.46; 6.49; 6.50). The number of artefacts in the richest metre number more than 700 specimens, including 15 cores, over 20 ad hoc tools, perforators and hammerstones. The next metre towards the urn grave (and the barrow’s centre) is clearly poorer – 362 artefacts.

Figure 6.48. Torup Høje, Fjelsø. Stratigraphy within sector A. Grey represents the thickness of cultural layer from the late Bronze Age. It considerably increases towards the centre of the barrow. Frequency of flint products increases with the increase of the layer’s thickness (drawing by M. Mikkelsen).

6. Cult features with flint assemblages from northern Jutland 145

146

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.49. Torup Høje, Fjelsø. Zones of the barrow arbitrarily determined Fredet zone – central zone, devoid of secondary burials from the late Bronze Age (LBA) and possibly protected before the burials were placed; Begravelsezone – funeral zone with numerous LBA secondary burials; Rituel zone – ritual zone with cult objects and LBA flint assemblages. Blue circles represent frequency of flint artefacts within explored sectors.

The distribution of artefacts in this sector may indicate that the metre grid utilized for excavation covered the centre of a flintworking workshop situated between the cult feature and the urn, adjacent to the barrow’s base. It could have been a concentric structure, with the greatest density of artefacts in its centre – the central place of flintworking and possible use of its products. Two richest explored metres adjoin the cult features. With a certain degree of certainty it may be assumed that it could be the workshop’s central part. The stratigraphy observed in Sector A, with cultural layer’s thickness increasing towards the barrow’s centre, partly accounts for the increase in the number of artefacts in the same direction (Figure 6.48). However, it appears only to have been a manifestation of the thickness of the cultural layer; the metre situated the closest to the barrow’s centre

6. Cult features with flint assemblages from northern Jutland

Figure 6.50. Torup Høje, Fjelsø. Distribution of classes of artefacts confronted with their frequency in individual explored sectors of the ritual zone from the late Bronze Age.

should display the greatest frequency. However, the frequency of flint artefacts in this metre is significantly smaller than in the preceding one. Sector B also displays an increase of the number of the artefacts in the area situated closest to the cult features – in this case two pairs of posthole objects (altars?) and the edge of the smaller of the two ditch features (Figure 6.50). The excavated area is even smaller than in Sector A, but between the two pairs of postholes in the metre’s central part there were 130 artefacts, such that during the complete excavation of the metres approximately 300 artefacts could be expected. The artefacts included cores (4 specimens), perforators (2 specimens), a notched piece and a retouched flake. This would suggest the greatest activity associated with production and use of flint occurred in the vicinity of the cult features interpreted as altars. Sector C includes only 2 m2 of the investigated cultural layer, but it is situated right in the centre of the complex of cult features (or one complex structure). The square metre situated closest to the barrow’s base, comprising the remains of a pair of posthole

147

148

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

features, provided more flint artefacts (Figure 6.50). It is also situated directly in front of the complex of vertically arranged boulders. The artefacts included a hammerstone, a perforator, 5 ad hoc retouched tools and 7 cores. Sector D is difficult to interpret. It is situated at the place where two foundation ditches intersect, including the western portion of the larger of the two(Figure 6.50). The stratigraphic relationship between them is not clear. Additionally, considerable damage to assemblage in its eastern part resulting from looters’ trenches affected its size and condition. The frequency of the artefacts in this area is high, relatively regular and varies between 100 and 200. They include both cores and tools. If the larger ditch feature was an enclosed space, it could be expected that flint material lay on its entire surface – like in the cult houses in Thy. As was mentioned earlier, the excavations covered a small area. Did the entire area of the ritual zone contain a comparable number of flint products? We cannot confirm this with the available data, but the distribution of known artefacts shows that these concentrations are spatially related to the cult features – the square metres closest to the features are the richest in artefacts, while the areas situated the furthest away are the poorest, which is well exemplified by sectors A, B and partly C. 6.5.8. Raw material The collection from Torup Høje weighs six times less than the twice as numerous flint assemblage from Grydehøj. This ratio reflects the difference in the size of the raw material used at both sites. The raw material from Torup Høje – locally common Scandinavian Senonian flint, probably collected at the closest moraine denudations, was present at the site in the form of small concretions, which is substantiated by the small sizes of blanks and cores (Figure 6.51). Perhaps the raw material was acquired from the hill of Torup Høje, where the group of five barrows is situated. Modern gravel pits, commonly present around the barrow, partly destroyed its structure and exposed numerous natural concretions similar to those used at the site by the Late Bronze Age communities. The differences in the proportions between both assemblages (Grydehøj and Torup Høje) are not great, but clear in comparisons. The cores from the assemblages in Thy areas on average are approximately . 1.2cm longer. The largest cores from Grydehøj exceeded 10cm, while the cores from Torup Høje rarely reached 6-7cm in length. Comparison of both assemblages clearly shows the differences in their proportions (Grydehøj – Table 6.4; Torup Høje – Table 6.16). The flint assemblage from Torup Høje is preserved in good condition; it carries no traces of patina, and only 1.27% (45 specimens) bear traces of heat alteration. These values prove that the contact of the artefacts with fire was coincidental. In Grydehøj the burnt artefacts were found in the area where kindling fire in the cult house was documented, but in Torup Høje their distribution did not display such a spatial relationship.

6. Cult features with flint assemblages from northern Jutland

Figure 6.51. Torup Høje, Fjelsø. Flint mass material from one of the explored sectors and a hammerstone with visible traces of use (photograph by author).

6.5.9. Technology In terms of the means of production and morphological characteristics, the flint assemblage from Torup Høje is similar to the assemblages from the Late Bronze Age settlements. It is also analogous to the known flint assemblages from the cult houses in Thy (refer to the remarks concerning Grydehøj – chapter 6.1 and Høghs Høj – chapter 6.2). Here reduction also took place within the cult houses. Flint concretions were broken, reduced and abandoned in individual cult features. The assemblage contains hammerstones displaying traces of use. 6.5.10. Cores Similar to the rest of the assemblage from Torup Høje, the cores also do not differ substantially from other collections from the Late Bronze Age, including the assemblages accompanying the cult structures (Figure 6.52; Plate 28; Table 6.12). However, there is a significant difference in comparison with the assemblages from Grydehøj and Høghs Høj. In my opinion it is a consequence of the differences in the proportions of the artefacts mentioned above, which in turn results from the differences in the sizes of the flint concretions available. Here the reduction of single-platform cores predominates, represented by more than 43% of all the cores (22% in Grydehøj and 21% in Høghs Høj), while blade cores are absent and disc cores (both with and without preparation)

149

150

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.52. Torup Høje, Fjelsø. Single-platform cores (photograph by M. Jórdeczka). Table 6.12. Torup Høje, Fjelsø. Cores Zone A Zones B Zone C

Categories of cores

Zone D

Total % N

-

-

3

2

5

5

13

4

3

23

43

43

flake core with equivalent flaked surfaces/striking platforms

- 

-

1

 -

1

1

blade core with equivalent flaked surfaces/striking platforms

-

-

-

-

-

0

non-prepared flake disc core

3

 -

1

2

6

6

prepared flake disc core

 -

- 

- 

1

1

1

block-like core with changing orientation

1

1

- 

1

3

3

initial core single-platform flake core

flake amorphous core/fragment

17

2

2

19

40

39

exhausted core

1

 -

- 

1

2

2

35

7

10

49

101

100

Total

are only occasionally present. Significantly larger blocks from Thy were much easier to reduce radially, both unifacially and bifacially. Small concretions from Torup Høje probably broke when radial reduction was attempted; therefore, this technique was rarely used here (altogether 7 discoidal cores). The contribution of the amorphic cores is similar, accounting for approximately 40% of the assemblages mentioned here. They carry the traces of informal reduction of a block, without preliminary preparation, aimed at producing one or a few flakes or chunks with sharp edges. The absolute absence of blade cores is a consequence of the size of the raw material and the technology used. Preparation and repair were occasionally carried out, which is substantiated by the contribution of the products of preparation, less than 2% of the raw material, both in Torup Høje and the assemblages from the cult houses in Thy (Table 6.15).

6. Cult features with flint assemblages from northern Jutland 6.5.11. Blanks The entire production in the investigated area of Torup Høje had a solely flake character (Table 6.13), to even a greater extent than the features from Thy. Blade blanks are coincidental and have only a metric dimension; they were produced only in the cortical form (more than 50% of cortex). All the blanks display common features: they are short and wide, with very distinct impact points and bulbs. Cortical blanks predominate, followed by secondary flakes with flake scars on the dorsal surface consistent with the direction of flake’s removal, which confirms the predominance of single-platform core reduction. Flakes with removal scars on the dorsal surface other than consistent with the direction of flake’s removal do not exceed 4% of the debitage, while the contribution of flakes with unidirectional removal of flakes is 45%. The analysis of the blanks’ butts and tools shows that cortical or natural butts predominate the entire assemblage (Table 6.15), which confirms the observations from the features in Thy. What differs is the absence of the preparation of blanks and tools from Torup Høje. Preparation is seen on approximately 30% of the debitage. This tendency is also seen on tools, where preparation was visible on only 14% of tools; more than 85% of tools do not display any preparation. In this view the assemblage from Torup Høje is only slightly more informal than the assemblages from Thy, which – considering the analogous technology of reduction – Table 6.13. Torup Høje, Fjelsø. Blanks Categories of blanks

Zone A

Zone B

Zone C

Zone D

Total N

%

cortex flake >50%

207

19

40

83

349

45,8

cortex blade >50%

3

 -

2

5

10

1,3

preparation products

1

2

3

9

15

1,9

flake with removals on dorsal face consistent with the direction of flake’s removal

170

14

11

142

337

44,2

flake with removals on dorsal face opposite to the direction of flake’s removal

 -

1

1

3

5

0,7

flake with removals on dorsal face oblique to/ shifted to the direction of flake’s removal

10

- 

1

9

20

2,6

multidirectional flake

3

-

1

4

8

1

blade with removals on dorsal face consistent with the direction of flake’s removal

10

 -

1

8

19

2,5

-

-

-

-

-

0

404

36

60

263

763

100

two-directional blade Total

151

152

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Figure 6.53. Torup Høje, Fjelsø. knife-like flakes, with one cutting edge and the other side in the form of natural backed edge (photograph by M. Jórdeczka).

results from the difference in the assemblage’s proportions. The producers with similar skills had more problems handling smaller flint material from Torup Høje. As previously mentioned, the proportions of the assemblage from Torup Høje are smaller than those from Thy. The average length of flakes varies between 20 and 30mm; the width is approximately 25mm (Table 6.16; Figure 6.54). A comparison of the sizes of blanks with those from Grydehøj shows that average values are approximately 1 cm larger in the assemblages from Thy. In general terms the collection from Torup Høje is an assemblage of considerably smaller blanks, including blade blanks. The use of a hard hammerstone and the absence of core preparation resulted in the fact that blanks were amorphic, short and relatively thick. Despite the differences in the proportions of blanks, the thickness of artefacts in both assemblages is similar. 6.5.12. Tools The concentration of tools from the ritual zone in Torup Høje displays representative characteristics of tools from the Late Bronze Age, such as the lack of consideration for the form of the tools, informal reduction, frequent use of chunks or cortical flakes and a limited number of tool types. The category of formal tools from Torup Høje, consisting of 16 specimens, includes a burin (Figure 6.55: 1, 2), end-scrapers (Figure 6.55: 3, 4), perforators (Plate 29; Figure 6.56) and notched tools (Plate 30: 1, 2, 5, 10, 11). The category is even more

6. Cult features with flint assemblages from northern Jutland

Figure 6.54. Torup Høje, Fjelsø. Sizes of blanks.

Figure 6.55. Torup Høje, Fjelsø. 1-2 – burin, 3-4 – end-scrapers (photograph by M. Jórdeczka, drawing by the author).

simplified than the tools from Grydehøj and Høghs Høj. Ad hoc tools, represented by flakes and partially retouched chunks, as well as blanks exhibiting traces of use, were more numerous in the assemblage than formal tools (Figure 6.57). Altogether the category of tools, complete with 3 hammerstones and tool fragments, did not exceed 100 specimens (Table 6.14). A burin appears once again in the assemblage (Figure 6.55: 1, 2) – a form so far unknown in the flintworking of that period in Scandinavia. The facts that burins appear in the cult houses in Thy and that in Høghs Høj they are a quantitatively significant component prompt a conclusion that it is a common element of flintworking from the Late Bronze Age cult features. The artefact from Torup Høje was made on a chunk, which may be identified as a rule. Plate-like chunks were used to produce burins, making use of their

153

154

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Table 6.14. Torup Høje, Fjelsø. Tools Categories of tools

Total

Zone A

Zone B

Zone C

Zone D N

%

end-scrapers

-

-

-

2

2

2,2

perforators and borers

5

3

1

1

10

11

notched tools

-

1

- 

2

3

3,3

burins

-

-

-

1

1

1

flakes/blades/retouched chunks

14

2

5

17

38

41,9

functional implements

21

 -

4

6

31

34

unidentified tools/fragments

1

 -

- 

2

3

3,3

hammerstones

2

 -

- 

1

3

3,3

43

6

10

32

91

100

Formal tools

Ad hoc tools

Other

Total

Figure 6.56. Torup Høje, Fjelsø. Perforator (photograph by M. Jórdeczka).

natural thickness to produce burin ends. The referenced specimen is a double singleblow burin. One of the ends was formed with two burin spalls. The burin was analysed in terms of the preserved traces of use (refer to chapter 8 devoted to the functions of the artefacts).

6. Cult features with flint assemblages from northern Jutland Table 6.15. Torup Høje, Fjelsø. Presence and manner of preparing butts of blanks and tools List of features  

Cortical/natural

N 100%

N

%

cortex flake >50%

230

192

cortex blade >50%

10

preparation products

Shaped N

%

83,5

38

16,5

10

100

 -

-

14

8

57

6

43

248

166

67

82

33

4

1

25

3

75

18

15

83,3

3

16,7

multidirectional flake

7

6

85,7

1

14,3

blade with removals on dorsal face consistent with the direction of flake’s removal

19

12

63,1

7

36,9

 blanks in total

550

410

tools

64

55

flake with removals on dorsal face consistent with the direction of flake’s removal flake with removals on dorsal face opposite to the direction of flake’s removal flake with removals on dorsal face oblique to/shifted to the direction of flake’s removal

%

70,5

Prepared %

9

%

%

-

-

-

-

-

-

29,5

140 86

N

14

End-scrapers are represented by two specimens. The examples are short and wide, and they were made on flakes; one of them displays additional retouch of the edge (Figure 6.55). The perforators are the most complex group (Figure 6.56; 6.58; Plate 29). Similar to Grydehøj, they include the tools whose tips were made with retouch blunted to various degrees (Plate 29: 3, 6, 7) and the forms shaped with a series of notches (Plate 29: 1, 2, 4, 5, 8, 9). Disregard for the tool blanks is also common in this category. Chunks, amorphic flakes or cores were selected, and in one case a perforator was made on a considerably aeolised fragment of flint gravel (Figure 6.56: right bottom). However, a

Figure 6.57. Torup Høje, Fjelsø. Flakes with traces of use (photograph by M. Jórdeczka).

155

156

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

few display quite careful preparation of tips; these include two perforators on chunks, with quite penetrating tips and a large area of partial bifacial flaking along the edges so that they converge into a wide tip (Figure 6.58).

Figure 6.58. Torup Høje, Fjelsø. Perforators (drawing by the author; photograph by M. Jórdeczka).

6. Cult features with flint assemblages from northern Jutland

The remaining tools from Torup Høje are notched pieces (Plate 30: 1, 2, 5, 9-11) and partially retouched tools or tools displaying macroscopic traces of use. Notched tools usually have single notches; if their number is greater, edges assume notcheddenticulated form (Plate 30: 1, 5). As mentioned above, the barrow under consideration had a long history, preceding the episodes from the Late Bronze Age. The evidence of this history were seen not only in consecutive reconstructions of the mound and successive interments, but also in the produced flint material. The analysed assemblage from the ritual zone included at least two artefacts of older, Neolithic or possibly Early Bronze Age chronology (Plate 30: 3, Table 6.16. Torup Høje, Fjelsø. Sizes of individual classes of artefacts (in millimetres); L – length; W – width; T – thickness Category of artefact cortex flake >50%

L L min. max.

∑

L med.

L W med. min.

W max.

W med.

W T med. min.

T max.

T med.

202

15

74

27,2

15

48

26,7

3

33

8,3

flake with removals on dorsal face consistent with 212 the direction of flake’s removal 

16

47

26,6

11

58

25,8

2

27

7,4

flake with removals on dorsal face opposite to the direction of flake’s removal

5

20

32

25,2

14

32

22,4

6

9

7,8

flake with removals on dorsal face oblique to/ shifted to the direction of flake’s removal 

19

16

38

24,1

21

35

27,3

3

12

6,9

multidirectional flake

26

25,5

8

21

33

27

18

33

25,7

4

12

7,3

cortex blade >50%

10

27

44

32,8

5

12

9

3

12

5,6

blade with removals on dorsal face consistent with the direction of blade’s removal

18

23

38

30,1

5

15

10,1

9,5

2

9

4,6

-

-

-

25,7

5

21

two-directional blade

-

-

-

preparation products

14

18

52

4

32

55

41

20

60

initial core single-platform flake core  flake core with equivalent flaked surfaces/striking platforms blade core with equivalent flaked surfaces/striking platforms

1 -

27

-

-

-

14

45

45,2

36

56

46

12

32

23,7

34,9

20

70

42,4

8

45

20,5

35,1

-

35

-

-

-

-

34

72

51,6

56

-

non-prepared flake disc core

6

prepared flake disc core

1

block-like core changing orientation

2

33

38

35,5

flake amorphous core

with

31,4

40,6

-

30 28

40,2

-

-

-

-

12

54

28

15

-

57

41,8

33

-

31

29,5

11

13

12 22,7

21

20

60

35,9

26

70

41,5

8

46

2

29

31

30

26

21

23,5

7

11

9

perforator and borer

6

26

52

42

17

58

34,6

5

22

14

2

43

47

burin

1

ad hoc tool

55

45 18

45

38,9

38

71

38,7

40 37

11

39

31

8

48

26,9

14 17

4

5,1

-

end-scraper notched tool

7,5

11,3

8

-

T med.

11 -

33

9,2

21,4

10,8

157

158

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

4). In addition to these, a cordiform arrowhead and a massive, very regular 15-cm-long fragment of a blade also connected with the oldest, Neolithic period of the barrow’s use were found in the area outside the subject sectors. A strike-a-light, possibly from the Late Bronze Age, was also found outside the investigated zones.

7. Issues in flintworking technology and typology 7.1. Technological aspects Across much of Europe, Late Bronze Age flintworking assumed a surprisingly uniform character (Ballin 2010; Bronowicki, Masojć 2010; Edmonds 1995; Eriksen 2010; Högberg 2004, 2009; Lech, Piotrowska 1997; Libera 2006). Comparable uniformity is also seen further away from central Europe, e.g. in the Middle East (Rosen 1996, 1997). In short, the uniformity as seen in Scandinavia is characterised by the following: • exclusive use of a hard hammerstone in flintworking, • simplified reduction of only flake cores, mainly amorphic or multidirectional, devoid of preparation, • production of flakes accompanied by a considerable number of chunks, • abandonment of many classes of formal tools, • production of ad hoc tools, and • use of sharp edges of flakes without further modification. This type of flintworking, frequently referred to as ad hoc, appears with a variant, characterised by considerable technological advances and formal care. This dualism is seen at sites in Poland, northern Germany, and especially at sites in southern Scandinavia (Högberg 2010). This variant is characterized by the production and use of the backed knife (it is referred to as a large backed knife to distinguish it from shorter forms) – a flint tool specific to the Late Bronze Age. With the development of the research of the European Late Bronze Age flintworking, this relatively simple technological dualism will probably assume a more complex form, complemented by a number of local technological variations. Examples of such local variations include two sites from southern Sweden – Apalle and Hallunda, where a technology different from the ones described above has been observed. At both sites, small flakes were removed with the use of a hard hammerstone from small cores with prepared flake surfaces characterised by a prepared rhomboidal butt (Högberg 2009: 253-256). As seen in the presentation of the settlement assemblages, large backed knives are common in residential homesteads from the Late Bronze Age in southern Scandinavia. Further south they are not so common. In central Poland large backed knives are exemplified by the Zele type knives. This issue will be discussed in greater detail in the section of the chapter treating the typology of the flint artefacts. It is noteworthy that a large backed knife is not a stage in the evolutionary development of harvesting tools replacing the Early Bronze Age sickle, but that it constitutes a new quality, being an independent ‘invention’ of the Late Bronze Age (Högberg 2010, Figure 3). 159

160

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Large backed knives are absent in the assemblages from the ritual zones in southern Scandinavia, where flintworking assumes an exclusively ad hoc, opportunistic form. The flint assemblages analysed in this work are homogenous. They occur in spatially restricted areas and include all phases of stone tool production, from natural flint concretions to the tools displaying traces of use. Flintworking took place directly within the ritual zones, where the final products (tools) were used and finally deposited. The flint assemblages were subjected to the dynamic technological analysis. Classification of flint artefacts according to the stages of their production, within consecutive sequences, on the one hand showed that the analysed assemblages are complete, and on the other, it enabled verification of their homogeneity. It is a method commonly used in the analysis of flint assemblages from various time periods (Schild 1980). A hierarchical classification of artefacts, based on their technological features and according to the sequences of their production, was devised for the needs of the analysis of ritual zone assemblages. The classification takes into account the places of flint products in the production cycle and includes the categories of individual artefacts present in the assemblages, from cores to tools, together with their types. The classification of individual assemblages as seen from the dynamic perspective are presented in the tables at the end of this work. The tables also include the data concerning the sizes of individual categories of tools, nature of butts, etc. This information was also included in the text in the presentation of individual assemblages from ritual zones. 7.2. Dynamic view of assemblages from ritual zones 7.2.1. Raw material (I) I.1. Nodules, concretions Flint material in the form of unmodified concretions was most likely transported to ritual zones in considerable amounts. Preliminary selection at the place of acquisition of the raw material probably did not take place. To produce flint pieces with sharp edges there was no point in eliminating cracked flint nodules, because they performed the desired function perfectly well. The great majority of the flint nodules transported to the ritual zones were used, which is substantiated by the small number of unworked nodules (8) in relation to cores (509) recovered from the Grydehøj cult house. Of the examined cultural zones, flint is the only raw material that was reduced in the production of stone tools. Quartzite hammerstones were utilized , the majority of which were quite large and heavy. The acquisition of the raw material approximated the manner of its exploitation – an ad hoc, opportunistic attitude predominated, resulting in the collection of the raw material from the places situated closest to the location of planned reduction activities. In northern Jutland flint is common and easily accessible,

7. Issues in flintworking technology and typology

occurring both in primary and secondary deposits. Flint nodules and cobbles lying in their secondary positions in the vicinity of cult houses were thus probably collected. There was no mining. The most frequently used was Scandinavian Senonian flint. Differences in the sizes of nodules from individual assemblages are evident, as reflected in the size distribution of flint artefacts. For example, both the raw material and chipped stone artefacts from Grydehøj display much larger dimensions than the assemblage from Torup Høje, which results from regional diversity of secondary flint deposits. 7.2.2. Early phase of reduction (II) II.1. Initial cores and pre-cores II.2. Cortical blanks II.2.1. Cortical flakes>50% of cortex II.2.2. Cortical blades>50% of cortex II.3. Products of preparation II.3.1. Crested blades II.3.1.1. Unifacial crested blades II.3.1.2. Bifacial crested blades II.3.2. Other II.4. Cortical chips The acquired flint nodules were reduced, which involved fracturing them into pieces using a hard hammerstone, usually resulting in several amorphic fragments of considerable sizes. The evidence of fracturing includes the scars from multiple blows found on cores and the resulting large chunks, which were later used as cores of smaller sizes. As a rule, preliminary preparation was not carried out. The nodule’s most suitable, natural surface or a flat platform acquired after fracturing the concretion was selected as a striking platform. Frequently it was not an intentionally produced surface sensu stricte, i.e. a flake scar , but a surface resulting from fracturing a nodule with a strong blow. Unprepared concretions or chunks obtained by fracturing the nodules were knapped to produce one or several cortical flakes, after which the core was usually abandoned. On average these constitute 5-7% of all the cores (Table 7.1).

Table 7.1. Cores and products of preparation from the early phase of coring. Site

Products of preparation

Initial cores

Unifacial Bifacial crested blades crested blades sum sum %1 7,85 22 8

Grydehøj

sum 40

Høghs Høj

25

4,9

7

Torup Høje

5

5

9

1 2

Proportion in the category of cores. Proportion in the category of blanks.

1,92 %2 1 0,8%2 1,9%2

Other/ Unidentified sum 3

6

161

162

Late Bronze Age Flintworking from Ritual Zones in Southern Scandinavia

Less common types include crested blades, usually unifacial, often exhibiting cortical surfaces. These are simplified forms in comparison to the crested blades known from the industries with advanced preliminary preparation. Their crests consist of scar from a single flake removal. There is also a possibility that the edge of a bifacial discoidal core might be interpreted as a crest if the former were removed. The early reduction phase produced flakes, and less frequently, completely cortical blades and blades with more than 50% cortex on the dorsal face. It is one of the most numerous categories in the typology of the assemblages. In the case of the cult houses it represents 44% of all the blanks from Grydehøj, 53% from Høghs Høj and 47% from the ritual zone of the Torup Høje barrow. These blanks were often used in tool production, e.g. end-scrapers were commonly made on completely cortical flakes. Therefore, the production of large cortical flakes constituted the final stage of core exploitation. The described treatment of flint raw material, i.e. fracturing it with numerous strong blows, especially given the specific spatial context where it took place, may indicate that flint reduction constituted an element of a ritual. Spectacular fracturing of large nodules into a number of smaller chunks might have been one of the ritual activities carried out in the cult houses. 7.2.3. Advanced reduction (III) III.1.Cores III.1.1. Single platform cores III.1.1.1. flake cores III.1.1.2. blade cores III.1.2. Bidirectional cores1 III.1.2.1. flake cores III.1.2.2. blades cores III.1.3. Discoidal flake cores III.1.3.1. unifacial (without the preparation of the striking platform) III.1.3.2. bifacial (with the preparation of the striking platform) III.1.4. Tabular flake cores with changed orientation III.1.5. Amorphic flake cores, multidirectional III.2. Blanks III.2.1. Flakes (cortical