The Iconography of Magic: Images of Power and the Power of Images in Ancient and Late Antique Magic [1 ed.] 9042948825, 9789042948822, 9789042948839

Table of contents :
CONTENTS
PREFACE
NOTES ON THE CONTRIBUTORS
I. THE “VIGNETTES”OF THE GREEK MAGICAL PAPYRI.VISUAL ELEMENTS OF THE PHARAONICMAGICAL TRADITION AND THE USEOF BILDZAUBER IN THE PGM
II. IMAGES OF TIED VICTIMS IN MAGICAL TEXTS
III. EULAMO VS. SETH.ON THE EQUINEHEADED DEMONREPRESENTED IN THE DEFIXIONESFROM PORTA S. SEBASTIANO (ROME)
IV. BAITMO AND LAMASHTU
V. THE HUNGRY WOLF
VI. THE LATEANTIQUE TRANSFEROF CIRCULAR GEMDESIGNS TO PAPYRI AND FOIL:THE OUROBORUS AND SOLOMON’S SEAL
VII. THE ANGUIPEDE, ITS ORIGINS AND MARKET DIFFUSION
VIII. HEATHEN SERPENTS AND WINGLESS ANGELS? SOME NOTES ON IMAGES IN COPTIC MAGICAL TEXTS
IX. DRAWING AND WRITING. REFLECTIONS ON THE TRANSMISSION OF RITUAL KNOWLEDGE
LIST OF ABBREVIATIONS
INDEXES

Citation preview

Studies in the History and Anthropology of Religion 7

The Iconography of Magic Images of Power and the Power of Images in Ancient and Late Antique Magic

Edited by Raquel Martín Hernández PE E T E R S

THE ICONOGRAPHY OF MAGIC IMAGES OF POWER AND THE POWER OF IMAGES IN ANCIENT AND LATE ANTIQUE MAGIC

ST U D I E S I N T H E H I ST O RY A N D A N T H R O P O L O G Y O F R E L IG IO N

Editors: Jan N. Bremmer and Laura Feldt In recent years, especially after the tragic events of 9/11, religion has increasingly drawn the attention of scholars. Whereas, traditionally, religion was studied by historians, anthropologists and students of the main religious traditions, today religion can be seen as a major factor on the contemporary political stage and is omnipresent in the media. Yet modern developments can rarely be well understood without proper anthropological and historical analyses. That is why we are pleased to announce a new series, Studies in the History and Anthropology of Religion. The editors welcome contributions on specific aspects of religion from a historical and/or anthropological perspective, be it proceedings of conferences or monographs.

1. The Strange World of Human Sacrifice, J.N. Bremmer (ed.), Leuven, 2007 2. The Celtic Evil Eye and Related Mythological Motifs in Medieval Ireland, J. Borsje, Leuven, 2012 3. Dreams as Divine Communication in Christianity: From Hermas to Aquinas, B.J. Koet (ed.), Leuven, 2012 4. Writing Myth: Mythography in the Ancient World, S.M. Trzaskoma and R.S. Smith (eds), Leuven, 2013 5. Priests and Prophets among Pagans, Jews and Christians, B. Dignas, R. Parker and G.G. Stroumsa (eds), Leuven, 2013 6 Marginality, Media and Mutations of Religious Authority in the History of Christianity, L. Feldt and J.N. Bremmer (eds.), Leuven, 2019

The Iconography of Magic Images of Power and the Power of Images in Ancient and Late Antique Magic edited by

Raquel Martín Hernández

PEETERS LEUVEN  PARIS  BRISTOL, CT 2022

Cover illustration: “The Magic Circle” (1886) by John William Waterhouse (1849-1917); Tate, Presented by the Trustees of the Chantrey Bequest 1886 Oil on Canvas, 1829 × 1270mm Photo: © Tate, London. A catalogue record for this book is available from the Library of Congress.

© 2022, Peeters – Bondgenotenlaan 153 – B-3000 Leuven – Belgium ISBN  978-90-429-4882-2 eISBN  978-90-429-4883-9 D/2022/0602/85 All rights reserved. No part of this book may be reproduced or transmitted, in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the publisher.

CONTENTS

Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

VII

Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XVII I

R. Lucarelli, The “Vignettes” of the Greek Magical Papyri. Visual Elements of the Pharaonic Magical Tradition and the Use of Bildzauber in the PGM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

F. Marco Simón and C. Sánchez NatalÍas, Images of Tied Victims in Magical Texts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

15

R. Martín Hernández, Eulamo vs. Seth. On the Equineheaded Demon Represented in the defixiones from Porta S. Sebastiano (Rome) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

35

IV

G. Németh, Baitmo and Lamashtu. . . . . . . . . . . . . . . . . . . . . . . . .

51

V

A. Mastrocinque, The Hungry Wolf . . . . . . . . . . . . . . . . . . . . . .

63

VI

Ch.A. Faraone, The Late-Antique Transfer of Circular GemDesigns to Papyri and Foil: The Ouroborus and Solomon’s Seal

81

II III

VII

W. Shandruk, The Anguipede, its Origins and Market Diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

VIII K. Dosoo, Heathen Serpents and Wingless Angels? Some Notes on Images in Coptic Magical Text. . . . . . . . . . . . . . . . . . . . . . . . . . 117 IX

S. Crippa, Drawing and Writing. Reflections on the Transmission of Ritual Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

List of Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

185

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

187

PREFACE Raquel Martín Hernández

“A picture is worth a thousand words.” This saying, which comes from an ancient Chinese proverb that can be more literally translated as “the meaning of an image can express ten thousand words,” quickly comes to mind when discussing the power of images. It stresses the importance of imagery in the communication process and how these images constitute, in and of themselves, a communicative act. Images express concepts, illustrate thoughts, give form to abstract ideas and have a high power of expression and suggestion. Understood as a language that has its own codes, images possess a meaning that can be understood through translation, interpretation, and their integration in a given context. At times images can indeed have a greater communicative potential than written texts, since sometimes there is no need for the person observing an image to fully unravel its meaning in order to grasp what it refers to; this is especially the case with the use of standardized images such as icons. We can even think of instances in which the image effectively replaces the word. Just think of the incredible popularity of emojis in social networks and instant messaging applications nowadays. But this is neither the moment to expound on the communicative capacity of images, nor is it the proper place to debate their use or how they are perceived by observers. Instead, I will dedicate this brief introduction to outlining the process that led to the production of this volume, the field of research to which it belongs, and some of the findings contained in its articles. The present volume has its origin in the organization of a conference entitled “Drawings, Charaktēres, and Figural Representations in Ancient Magic. New Perspectives,” held on September 15th and 16th of 2016 at the Universidad Complutense of Madrid thanks to funding from the BBVA Foundation and the research projects awarded by the editor of the book.1 The idea 1

BBVA Foundation Leonardo Grant for Cultural Researchers and Creators (2015) and the project “Leyendo vidas. Religión, derecho y Sociedad en los papiros de las colecciones españo-

VIII

preface

for the conference originally arose from a personal interest in exploring the possibilities that the study of the non-textual elements of magical objects and documents could offer to our understanding of the phenomenon of magic in Antiquity. Within the broader framework of new, burgeoning trends in the study of magic, I decided to bring together a group of renowned specialists to discuss the iconography of magical images as well as to explore the meaning of different figurae magicae and their use in acts of ritual power. I aimed to turn the scholarly gaze to the “other texts” on magical items, with those “texts,” of course, being understood figuratively.2 Images found on magical artifacts enhance their pragmatic efficacy, imbue the objects with part of their power, and in so doing, they play a fundamental role in transforming an object into a magical artifact.3 Even images have agency of their own.4 Therefore, it is vital to “translate” and “read” these images, in order to properly understand magical items. Since the appearance of the first scholarly editions of the magical documents (mainly curse tablets and the texts of the Graeco-Egyptian papyri), scholars have amassed a venerable tradition of studies of magical texts in Mediterranean Antiquity.5 The las” (PGC2018-096572-B-C22) [MCIU/AEI/FEDER, UE]. I would like to express my sincere thanks to the editors of the series, Jan N. Bremmer and Laura Feldt, who reviewed the first version of the manuscript. Their corrections and suggestions have greatly contributed to the improvement of the present book. 2 C. Geertz, Negara: The Theatre-State in Nineteenth-Century Bali (Princeton, 1980) 13, led me to reflect on how to understand figurae magicae in an improved way: “arguments, melodies, formulas, maps, and pictures are not so much idealities to be stared at but texts to be read.” 3 See R. Gordon, ‘From Substances to Texts: Three Materialities of Magic in the Roman Period’, in D. Boschung and J.N. Bremmer (eds.), The Materiality of Magic (Paderborn, 2015) 133-76; R. Gordon, ‘Getting it right. Performative Images in Greco-Egyptian Magical Practice’, in M. Arnhold et al. (eds.), Seeing the God. Image, Space, Performance and Vision in the Religion of the Roman Empire (Tübingen, 2018) 101-23; A. Wilburn, ‘Figurines, Images, and Representations Used in Ritual Practices’, in D. Frankfurter (ed.), Guide to the Study of Ancient Magic (Leiden, 2019) 456-506; R. Martín Hernández, ‘The Figural Representations of Victims on Agosnistic Late-Antique Curse Tablets’, Religion in the Roman Empire 7.1 (2021) 96-124 and C. Sánchez Natalías, ‘Paragraphics and Iconography’, in R. Gordon et al. (eds.), Choosing Magic. Contexts, Objects, Meanings: The Archaeology of Instrumental Religion in the Latin West (Rome, 2021) 103-23 at 108-09. On agency and magical figurines see D. Collins, ‘Nature, Cause, and Agency in Greek Magic’, Transactions of the American Philological Association 133.1 (2003) 17-49 at 37-44. 4 J. Johnston, ‘Prolegomena to Considering Drawings of Spirit-Beings in Mandaean, Gnostic and Ancient Magical Texts’, ARAM 22 (2010) 573-82 at 575-79. 5 R. Wünsch, Defixionum Tabellae. Inscriptiones Graecae III.3 Appendix (Berlin, 1897); R.  Wünsch, Sethianische Verfluchungstafeln aus Rom (Leipzig, 1898); A. Audollent, Defixionum tabellae quotquot innotuerunt: tam in Graecis Orientis quam in totius Occidentis partibus praeter Atticas in corpore inscriptionum atticarum editas (Paris, 1904); K. Preisendanz et al. (eds.), Papyri graecae magicae, 2 vols. (Leipzig, 1928-1931, re-edited by A. Henrichs in 2 vols. in 1973-1974).

preface

IX

discipline has experienced a great boom since the mid-80s, when research was still largely focused on texts, though new issues began to come into focus, such as the figure of the magician in Antiquity, the categorization of ritual acts, the debate between magic and religion, and the social conception of the “arcane discipline.”6 Currently, studies on magic in Antiquity are experiencing yet another boom in which research on materiality has been increasingly coming to the fore. The so-called “material turn”7 has also taken root in the study of religions and, especially, magic in Antiquity in recent years. Studies that are currently being published on the subject are a good testament to this shift in focus and are continually proposing new perspec-

6 It would be impossible to list all existing works. I will limit myself to providing information on some of the books and articles that from my perspective have been especially important for the development of the discipline, especially Greek and Roman magic. The list, of course, is far from exhaustive: G. Luck, Arcana Mundi. Magic and the Occult in the Greek and Roman Worlds (Baltimore, 1985); C.A. Faraone and D. Obbink (eds.), Magika Hiera. Ancient Greek Magic and Religion (Oxford, 1991); H.S. Versnel, ‘Some Reflections on the Relationship Magic-Religion’, Numen 38.2 (1991) 177-97; A. Bernand, Sorciers grecs (Paris, 1991); J.G. Gager, Curse Tablets and Biding Spells from the Ancient World (Oxford, 1992); C.A. Faraone, Talismans and Trojan Horses: Guardian Statues in Ancient Greek Myth and Ritual (New York, 1992); M. Meyer and R. Smith, Ancient Christian Magic: Coptic Texts of Ritual Power (Princeton NJ, 1994); W. Brashear, ‘The Greek magical papyri: An introduction and survey; annotated bibliography (1928-1994)’, ANRW II 18.5 (1995) 3380-684; M. Meyer and P. Mirecki (eds.), Ancient Magic and Ritual Power (Leiden and New York, 1995); F. Graf, Magic in the Ancient World (Cambridge, MA and London, 1997, English Translation of the German original); P. Schäfer and H.G. Kippenberg (eds.), Envisioning Magic: A Princeton Seminar and Symposium, Studies in the History of Religions 75 (Leiden, 1997); V. Flint, R. Gordon and G. Luck (eds.) Witchcraft and Magic in Europe, Volume 2 Ancient Greece and Rome (London, 1998); C.A. Faraone, Ancient Greek Love Magic (Cambridge MA, 1999); J.N. Bremmer, ‘The birth of the term magic’, ZPE 126 (1999) 1-12, updated in id., Greek Religion and Culture, the Bible and the Ancient Near East (Leiden, 2008) 235-47, 347-52; D.R. Jordan, H. Montgomery and E.  Thomassen (eds.), The World of Ancient Magic (Bergen, 1999); M.W. Dickie, Magic and Magicians in the Graeco-Roman World (London and New York, 2001); N. Janowitz, Magic in the Roman World. Pagans, Jews and Christians (London and New York, 2001); J.N. Bremmer and J.R. Veenstra (eds.), The Metamorphosis of Magic from Late Antiquity to the Early Modern Period (Leuven, 2002); D. Ogden, Magic, Witchcraft, and Ghosts in the Greek and Roman Worlds. A Sourcebook (Oxford, 2002); P. Mirecki and M. Meyer (eds.), Magic and Ritual in the Ancient World (Leiden, 2002); S. Shaked (ed.), Officina Magica. Essays on the Practice of Magic in Antiquity (Leiden, 2005); E. Eidinow, Oracles, Curses, and Risk among the Ancient Greeks (Oxford, 2007); R. Gordon and F. Marco Simón (eds.), Magical Practice in the Latin West (Leiden and New York, 2010), D. Frankfurter (ed.), Guide to the Study of Ancient Magic (Leiden, 2019); R.G. Edmonds III, Drawing down the moon. Magic in the Ancient Greco-Roman world (Princeton and Oxford, 2019); A. Mastrocinque et al. (eds.), Ancient Magic. Then and now (Stuttgart, 2020). A good bibliography can be found in J.L. Calvo Martínez, ‘Cien años de investigación sobre la magia antigua’, MHNH 1 (2001) 7-60. 7 On this concept, see J.N. Bremmer, ‘Preface: The Materiality of Magic’, in Boschung and Bremmer, The Materiality of Magic, 7-19 at 9, n. 8.

X

preface

tives and lines of research.8 Studies, such as those mentioned throughout the footnotes, have drastically changed our understanding of magic in ancient societies. These studies have drawn new conclusions about the meaning, function and use of magical objects, proposed new approaches to understanding how magical objects acquire their agency, proffered a better understanding of the technologies applied to objects, and analyzed the dissemination, transformation, and transmission of magical knowledge in Antiquity. New technologies as well as greater access to objects and photographs in our increasingly globalized world with fewer barriers (at least when it comes to research and collaboration between researchers) have helped us to move away from the old, nearly exclusive attention to texts. As a result, we now possess a more panoramic and global vision of the entire phenomenon. One of the main lessons to be gleaned from recent scholarship is that acts of ritual power are an indissoluble amalgam of diverse components, all of which give the object its power. Ignoring one of these components leads us to understand the phenomenon, the ritual, and the object only in partial terms. The inclusion of non-textual elements in magic books and “activated objects” suddenly infused magical objects and Greek magic books from the second century CE onwards and are especially pervasive in Coptic magic. In fact, they are sometimes the only thing preserved on a magical item; in such cases, the study of imagery, along with the material and – whenever possible – the archaeological data, is all that is available when trying to understand the object and its meaning. It is true that sometimes we cannot properly unravel the full significance that a given image had for its author, user, or observer. The chronological distance, the loss of references, the (sometimes)

8 This new trend of research has produced magnificent studies of which I will list only a few: W. Wilburn, Materia Magica. The Archaeology of Magic in Roman Egypt, Cyprus, and Spain (Ann Arbor, 2012) emphasizes the archaeological contexts of magical practices in Roman Egypt, Cyprus and Spain. In the same year the proceedings of the conference “Contesti magici,” held in Rome in 2009, came out: M. Piranomonte and F. Marco Simón (eds.), Contesti magici. Contextos mágicos (Rome, 2012). The book presents several articles dealing with materiality, archaeological context and non-textual elements of ritual practices. In 2015 two volumes with similar titles appeared. Boschung and Bremmer, The Materiality of Magic focuses on the material aspects of magic, such as gems, rings, drawings, grimoires, amulets, and figurines and focuses its analysis on the Ancient World. C. Houlbrook and N. Armitage (eds.), The Materiality of Magic: An artifactual investigation into ritual practices and popular beliefs (Oxford, 2015) presents a wider geographical and chronological scope than the previous work. Although the subject matters differ greatly among the contributions, all the articles focus on how notions of magic are manifested materially. Finally, the latest general publication on magic, Frankfurter, Guide to the Study of Ancient Magic dedicates a whole section (Part 3) to studies on archaeology and materiality.

preface

XI

drastic shift in iconographic codes and other factors make interpretation difficult. That said, comparative studies, statistical analysis, interconnections between images and texts, and other analytical possibilities will continue to allow us to reach a greater understanding of not only the object but also the accompanying text itself. In short, new approaches to the field are helping us to explain better the complexity of magic rites in Antiquity. The significance of non-textual elements is especially important in amulets and gemstones in particular. Studies of gem amulets have always taken this into account.9 However, the study of figurae magicae and symbols, like the popular charaktēres, found on other objects that are manifestations of ritual power in Antiquity have not been as systematic or dogged. Especially significant is the paucity of attention paid to images in the Graeco-Egyptian magic formularies and curses in the past, although this trend has begun to change in recent years.10 Texts have always been at the center of the debate and images have been given secondary importance, if not treated merely as an afterthought. However, as is evident from the studies in this volume, magical images are of paramount importance in providing agency to a given ritual object. As R. Gordon 2002 showed in his typology of the purpose of images in the magic texts, they can be evocative (they invoke the divine or demonic agent to come to the ritual) or performative (they show the result 9 See, e.g., C. Bonner, Studies in Magical Amulets. Chiefly Graeco-Egyptian (Ann Arbor, 1950); A. Delatte and Ph. Derchain, Les intailles magiques Graeco-egyptiennes (Paris, 1964); S. Michel, Die Magischen Gemmen im Britischen Museum (London, 2001); A. Mastrocinque, Syllogue Gemarum Gnosticarum vols. I and II (Roma, 2003 and 2007); C. Entwistle and N. Adams (eds.), Gems of Heaven: Recent Research on Engraved Gemstones in Late Antiquity c. AD 200–600 (London, 2011); C.A. Faraone, The Transformation of Greek Amulets in Roman Imperial Times (Philadelphia, 2018). 10 See e.g. U. Horak, ‘Christliches und Christlich-Magisches auf Illuminierten Papyri, Pergamenten, Papieren und Ostraka’, Mitteilungen zur Christlichen Archäologie 1 (1995) 27-48; Johnston, ‘Prolegomena to Considering Drawings of Spirit-Beings’; M. Bailliot, Magie et sortilèges dans l’Antiquité romaine: archéologie des rituels et des images (Paris, 2010); R. Gordon, ‘Signa nova et inaudita: The Theory and Practice of Invented Signs (charaktêres) in Graeco-Egyptian Magical Texts’, MHNH 11 (2011) 15-44; R. Martín Hernández, ‘Reading Magical Drawings in the Greek Magical Papyri’, in P. Schubert (ed.), Actes du 26e Congrès international de papyrologie (Genève, 2012) 491-498; G. Németh, Supplementum Audollentianum, Hungarian Polis Studies 20 (Zaragoza, Budapest and Debrecen, 2013); J. Dijkstra, ‘The Interplay Between Image and Text. On Greek Amulets Containing Christian Elements from Late Antique Egypt’, in Boschung and Bremmer, The Materiality of Magic, 271-92; K. Dosoo, ‘Zōdion and Praxis: An Illustrated Coptic Magical Papyrus in the Macquarie University Collection’, Journal of Coptic Studies 19 (2018) 11-56; R. Gordon, ‘Getting it Right. Performative Images in Graeco-Egyptian Magical Practice’, in M. Arnhold et al. (eds.), Seeing the God. Image, Space, Performance, and Vision in the Religion of the Roman Empire (Tübingen, 2018) 101-23; A.T. Wilburn, ‘Figurines, Images, and Representations Used in Ritual Practices’, in Frankfurter, Guide to the Study of Ancient Magic, 456-506; and K. Dosoo and J.-C.  Coulon (eds.), Magikon Zōon: Animal et magie | The Animal in Magic (forthcoming).

XII

preface

that will be achieved by the completion of the rite). Although these two functions are the main ones, we should not overlook their symbolic value, because some images and symbols – in particular the charaktēres in GraecoRoman magic – clearly serve to classify the object on which they are written or engraved as magical; therefore, they would also have been understood as powerful symbols by any observer who shared these visual codes, even if he/ she does not really understand their meaning. The first chapter, by Rita Lucarelli (University of California, Berkeley), pays attention to Egyptian texts, in particular to the so-called “vignettes” of the funerary texts, in order to interpret several of the designs produced by the scribes of the Graeco-Egyptian magical formularies. Lucarelli focuses on highlighting the evidence of the ritual continuum that exists between the texts of the PGM and their Egyptian models, thus demonstrating that the Egyptian Bildzauber was still effective in this new cultural context. Neither the magical drawings in the PGM nor the vignettes in the Book of the Dead are mere illustrations. Both should be understood as visual magical elements that complement the text and that, in line with the last chapter, interact and operate in a shared space of codes that are understood both by those who made the designs and those who used or believed them to be effective. Among the figurative motifs analyzed by Lucarelli are the beetle, the embalming scene, and the stele with its possible connection to the mode of representation in vignettes from the Book of the Dead. The second chapter of the volume, by Francisco Marco Simón and Celia Sánchez Natalías (Universidad de Zaragoza), focuses on the variations of iconography or visual representations of binding in its metaphorical and literal sense in curse tablets. Their contribution examines everything that could be “tied up” in a curse: people, body parts, animals, and even physical spaces. The figurative representations of tied victims in a curse have a performative value, in the sense mentioned above, since they constitute a kind of didascalia, illustrate the aim of a given curse, and offer a metaphorical representation of what the object’s creators hope to see accomplished in the real world. Text and image compose a holistic unit that enhances the effectiveness and power of the ritual operation and is also the realization of the principle of persuasive analogy. The article offers a complete overview of the possible coercive power that a curse could muster through the image of binding. The third chapter, by Raquel Martín Hernández (Universidad Complutense de Madrid), focuses on the identification of the equine-headed god that appears in several of the defixiones found at Porta San Sebastiano in Rome. After a review of the different interpretations that this figure has

preface

XIII

received, Martín Hernández defends the possibility that it is, in fact, a reinterpretation of the canonical image of the god Seth. Although it is possible that the image of this god, who is traditionally depicted with the head of a donkey, has been manipulated to reinforce the persuasive analogy between the invoked god and the victims (the charioteers and the horses of the chariot races), one must also consider the constant fluctuation in the representation of some Egyptian gods, especially Seth and Anubis. The chapter reviews other drawings of the equine-headed god that appear in different magical texts, fundamentally associated with the so-called “ιωερβηθ logos,” to point out the variety of the rites in which this representation appears. In his article, György Németh (Eötvös Loránd University) reviews one of the iconographic motifs featured in the curse tablets found in Hadrumetum, intended to curse racehorses in the circus: the representation of a demonic figure on what looks like a raft. The scholar reviews the interpretations offered by previous analyses, before advocating for an orientalizing interpretation of the motif and recognizing in the figure the daemon Baitmo found on the tablets as a version of Lamashtu. The motif of the hungry and thirsty animal is studied by Attilio Mastrocinque (Università di Verona) as an operative ritual resource for the realization of amulets for various purposes. Mastrocinque explores the possibilities in which text and images (in particular that of the wolf) interact to create agency in the object and turn it into a powerful amulet through various persuasive analogies, which he reviews and studies in the article. He highlights the ambiguous role of some threatening animals in gems and verbal incantations. The threatening and protective facets of some animals could be exploited for different magical purposes. The correct interpretation of its symbolism helps to interpret the piece in its multiple dimensions. Christopher A. Faraone (University of Chicago) explores how circular designs characteristic of gems are transmitted and made in non-circular media such as papyrus and metal foils. In his article he addresses the change in technology that occurred in the production of amulets, which leads him to reflect on the importance of a deeper study of the materiality of magical objects and their uses. Gems allowed users to make a drawing or design on each side, which is impossible in other media because of their material peculiarities or their traditional use (amulets in papyri used to be written only on one side). This problem is solved in different ways for other materials, like integrating the two parts of the design into a single drawing. Faraone also examines the transmission of this technology and different solutions, while advocating for the trend of copying iconography directly from gems for use in other media.

XIV

preface

Walter Shandruk (University of Chicago) applies statistical studies to reach a better understanding of the popularity of the image of the anguiped god. Computer science applied to the study of ancient documents, texts, iconography, and materiality can offer interesting results to further understand the popularity of particular designs, colors and materials for different purposes. Furthermore, such an approach can help answer whether certain combinations are significant or not. The article bases its hypothesis on the application of statistical operations to different components (textual and non-textual) that appear in gems where the anguiped god is represented in order to provide arguments that weigh in on the debate over who is the divinity represented, as well as to investigate why these representations are mainly associated with a single medium: gemstones. The study reveals that the relationship of the analyzed components should largely be related to the Hebrew god, and that the relationships with solar elements are statistically less significant. The eighth chapter, by Korshi Dosoo (Universität Würzburg), focuses on the study of images in Coptic magic books, in particular figurative images whose iconicity is recognizable. Dosoo studies the Graeco-Egyptian precedents of the images he deals with and their symbolism as well as the revival they enjoyed in Coptic magic. He refers to the evocative or performative value of the images, as studied by Gordon, but, according to Dosoo, such a distribution serves to explain their function from an external point of view, though it is not possible to know if users understood them in this way. Dosoo investigates the relationship between the image and the represented supernatural being in the ritual. As he points out, “in most cases Coptic magical images diverge from Christian art stylistically but not iconographically.” In Coptic magic, forces are conjured up by their name, image, and symbols. These undoubtedly lend much importance to figurative images and representations of powerful beings that were manifested through the kinds of stereotyped images that are repeated. Dosoo analyses images of angelic figures, focusing on the representation of elements that identify them as such (especially wings). The study of this particular trait leads him to identify the representations of other “angelical” figures. He also reviews the iconography of the decans, providing a compelling comparative study of their representations and their association with the three Hebrews. The chapter ends with an analysis of the figure of “the Ethiopian” and his relationship with beings associated with the lack of light and the infernal punishments. The last chapter, by Sabina Crippa (Università Ca’Foscari), adresses a series of fundamental questions for a study such as the present volume. After reviewing the dearth of studies on images in their non-sacred aspect

preface

XV

in the field of religious studies, she inquires about the interconnection that exists in Graeco-Egyptian magical papyri between the material medium, the ritual art, and the transcription of the ritual words; she also delves into the complex relationship between the written word, image, and voice, all of which are elements that co-exist in the acts of ritual power in an interconnected way. Crippa advocates for an understanding of text and image in the cultural context in which they were developed and, above all, understood, because one must never forget that magic also had its audience: it was not, after all, only the creator of an image who recognized it as sacred or powerful, since others could also perceive these images in such terms, based on shared understandings and exchanges of patterns operative within a given culture. The copying and reworking of traditional elements together with the inclusion of their own native elements would be the basis for the creation of images of magic in Graeco-Roman Antiquity. Ant yet these new creations could only be meaningful if they were understood by the cultural community for which they were meant to be valuable. Crippa distinguishes between two kinds of magical images: those in which the text generates the image (the body of the victim, the divinity, the daemon) and those in which the text becomes the image (the charaktēres or magic stelae or patterns). Accordingly, she shows the textual capacity of the images and their communicative value. In this sense, she demonstrates how all the elements converge in magic rituals in order to become communicative acts aimed at appealing to supernatural powers. The present collection of essays that we present is, therefore, only a sample of the multiple possibilities that research on the non-textual components of ritual acts offers for the study of magic. Greater attention to this aspect, in conjunction with the study of other agents involved in the elaboration of a given rite, allows us to address new questions, reach new conclusions, offer new hypotheses, discard old interpretations, and, in short, to obtain a greater knowledge of magic in Antiquity.

NOTES ON THE CONTRIBUTORS

Sabina Crippa: Associate Professor in the Department of Humanities at the Ca’Foscari Univesity (Venice, Italy). She holds a PhD in History of Religions from the École pratique des Hautes Études and is the author of books and articles on the relationship between writing, voice and image in religious and magical rituals: ‘Images et écritures dans les rituels magiques (PGM)’, Studi e Materiali di Storia delle Religioni 76 (2010) 117-38, ‘Figure del divino. Riflessioni storico-religiose’, in Federica Fontana and Emanuela Murgia (eds.), Sacrum facere (Trieste, 2018), 15-31. Korshi Dosoo: Junior research group leader of the project ‘The Coptic Magical Papyri: Vernacular Religion in Late Antique and Early Islamic Egypt’ at the Julius Maximilian University of Würzburg (Germany). Formerly Lecturer at the University of Strasbourg and post-doctoral researcher on the Labex RESMED project ‘Les mots de la paix’. His research focuses on magic and lived religion in Egypt from the Ptolemaic to Mamluk periods as revealed by papyrological and epigraphic sources. He is author of several articles on Coptic magic: ‘Healing Traditions in Coptic Magical Texts’, Trends in Classics 13.1 (2021) 44-94. Christopher A. Faraone: Frank Curtis Springer and Gertrude Melcher Springer Professor in the Humanities and Professor in the Department of Classics at the University of Chicago (USA). He is author of many books and articles on magic in Ancient Greece: Vanishing Acts: Deletio Morbi as Speech Act and Visual Design on Ancient Greek Amulets, BICS Supplements 115 (London, 2012), The Transformation of Greek Amulets in Roman Imperial Times (Philadelphia, 2018). Rita Lucarelli: Associate Professor of Egyptology in the Department of Middle Eastern Languages and Cultures and Curator of Egyptology at the Phoebe A. Herst Museum of Anthropology of the University of California, Berkeley (USA). She has written books and articles on Egyptian demonology: The Book of the Dead of Gatseshen: Ancient Egyptian Funerary Religion

XVIII

notes on the contributors

in the 10th Century BC. Egyptologische Uitgaven 21 (Leiden, 2006), ‘Gods, Spirits, Demons of the Book of the Dead’, in Foy Scalf (ed.), Book of the Dead: Becoming God in Ancient Egypt (Chicago, 2017) 127-36. Francisco Marco Simón: Emeritus Professor in the Department of Ancient History at the University of Zaragoza (Spain). His research activity has focused especially on religious systems in the Ancient world as well as on magic-religious practices in Roman Times: with R. Gordon (ed.), Magical Practice in the Latin West, RGRW 168 (Leiden and Boston, 2010), ‘Early Hispanic Curse Tablets: Greek, Latin -and Iberian?’, in Christopher A. Faraone and Richard L. Gordon (eds.), Curses in Context I: Curse-Tablets of Italy and the Western Provinces (Tübingen, 2020). Raquel Martín Hernández: Associate Professor in the Department of Classics at the Universidad Complutense of Madrid (Spain). She has published books and articles on the relationship between magic and mysteries and on the material dimension of Greek magical texts and handbooks in Roman Times: Orfeo y los magos. La literature órfica, la magia y los misterios (Madrid, 2010), ‘A Coherent Division of a Magical Handbook. Using Lectional Signs in PGM VII’, Segno e Testo 13 (2015) 147-64. Attilio Mastrocinque: Professor in the Department of Scienze dell’antichità at the Univesity of Verona (Italy) and coordinator of the international doctorate in Arts and Archaeology (Ghent-Verona). He researches in Roman history and religions of the Roman Empire. He is specialized in reading magical inscriptions on gems and metal lamellae: Syllogue Gemmarum Gnosticarum vols. I and II (Rome, 2003 and 2007), Les intailles magiques du Départment des monnaies médailles et antiques (Paris, 2014). György Németh: Professor in the Department of Ancient History at the Eötvös Loránd University (Hungary). He has published several books and articles on Greek and Roman magical texts and artifacts with a focus on cultural relationships: Supplementum Audollentianum (Budapest, 2013), ‘Voodoo dolls in the Classical World’, in Eduard Németh (ed.), Violence in Prehistory and Antiquity (Kaiserslautern and Mehlingen, 2018) 179-94. Celia Sánchez Natalías: Postdoctoral Researcher in the Department of Ancient History at the University of Zaragoza (Spain). Her research focuses on Latin curse tablets from the West Roman Empire with a focus on manipulation of script and supplementary practices to reinforce the agency of

notes on the contributors

XIX

defixiones: ‘Seth in the Fountain of Anna Perenna? A new interpretation of the magical container 475549’, in Attilio Mastrocinque et al. (eds.), Ancient Magic. Then and Now (Stuttgart, 2020) 113-22, and Sylloge of Defixiones from the Roman West. A Comprehensive Collection of Curse Tablets from the fourth Century BCE to the fifth Century CE. Volume I and II (Oxford, 2022). Walter Shandruk: PhD in Classics and Classical Languages from the Univesity of Chicago. He holds a position as statistical analysis project manager at the Univesrsity of Chicago. His research focuses on statistical studies applied to magical gemstones and the Christian use of magic: ‘The Interchange of ι and η in Spelling χριστ- in Documentary Papyri’, Bulletin of the American Society of Papyrologists 47 (2010) 205-19, ‘Christian Use of Magic in Late Antique Egypt’, Journal of Early Christian Studies 20.1 (2012) 31-57.

I. THE “VIGNETTES” OF THE GREEK MAGICAL PAPYRI. VISUAL ELEMENTS OF THE PHARAONIC MAGICAL TRADITION AND THE USE OF BILDZAUBER IN THE PGM Rita Lucarelli

In magical compositions of the Pharaonic period – in particular in the Book of the Dead – copied especially on papyrus but also attested on tomb walls and in other mortuary contexts, the texts are very often accompanied by “vignettes”, namely depictions functioning as magical complement of the written incantations and as a sort of visual summary of each text’s topic and title; in certain cases, the vignettes may even replace the text of a spell.1 These images, an expression of the so-called Bildzauber, “figurative magic” or “magic with images”, enhanced the amuletic function of the magical papyri or objects where the incantations were copied on.2 The custom of adding iconic elements to texts on papyrus continued in the magical books of the Graeco and Roman Periods and throughout Late Antiquity in the Greek magical papyri (PGM). The study of the visual elements of the PGM has only recently been focused on,3 after having been neglected for a long time due to the generally unclear or complex relationship of these drawings to the texts copied in their vicinity, as well as because of their poor esthetic quality.4 1 For a general overview on the vignettes in the Book of the Dead, see H. Milde, ‘Reading vignettes: an approach to illustrations in the Book of the Dead’, Jaarbericht van het Vooraziatisch-Egyptisch Genootschap Ex Oriente Lux 43 (2011) 43-56. 2 For a comprehensive study of Bildzauber in ancient Egypt and with particular focus on the interplay of text and image, see P. Eschweiler, Bildzauber im alten Ägypten: die Verwendung von Bildern und Gegenständen in magischen Handlungen nach den Texten des Mittleren und Neuen Reiches (Göttingen, 1994). 3 See R. Martín Hernández, ‘Reading magical drawings in the Greek Magical Papyri’, in P. Schubert (ed.), Actes du 26e Congrès international de papyrologie (Genève, 2012) 491-98. 4 Weitzmann’s central study on the relationship between texts and images in antiquity  also mentions the sketches of the PGM, defining them “of extreme crudeness”

2

rita lucarelli

Some of these images are, however, of clear ancient Egyptian origin and seem to derive from models taken from the Book of the Dead and other pharaonic funerary compositions. Such an iconic transmission shows that the ancient Egyptian Bildzauber was still considered effective in the new socio-cultural context where the PGM were composed and used. In order to define an “illustration” or an iconographic element within the ancient Egyptian manuscripts production at large, one has to analyze the relationship between image and text. The entry “Illustration” in the Lexikon der Ägyptologie,5 lists three main types of illustrations; only the first type is an “image without a text” (Bilder ohne Text), while the other two are connected to a text, either because the latter describes and explains the illustration or because the illustration is dependent and secondary to the text.6 All these three types of illustration, as described in the Lexicon’s entry, do occur in magical manuscripts and therefore they need to be interpreted according to their interplay with the text. When looking instead to the ancient Egyptian language, in order to establish an emic view on images as mentioned in magical and religious contexts, we can note different terms employed in sacred writings, such as twt, “image, statue” employed in magical instructions of spells when referring to statuettes and figurines; jrw.w, “creation, form,” used in funerary books for cultic communication; ḫpr.w, “form, mode of being, transformation”, referring to periodic transformations of the sun god, of Osiris as well as of the deceased assimilated to them and, “statue, image, likeness”, which refers to potential epiphanies of the gods.7 This rich semantic spectrum of terms referring to religious images shows how important the latter were considered within the ancient Egyptian ritual and magical practices. In the funerary papyri of the Pharaonic Period and in particular of those of the Book of the Dead “genre”, the pictographic elements are called “vignettes” by egyptologists.8 This is a Middle French word (vignette) (K.  Weitzmann, Illustrations in Roll and Codex: A study of the origin and method of text Illustration (Princeton, 1947) 51). 5 K. Zibelius, s.v. ‘Illustration’, in W. Helk et al. (eds.), Lexikon Der Ägyptologie III (Wiesbaden, 1980) 137-40. 6 On the vignettes in the Book of the Dead in comparison with the Lexikon’s entry, see also R. Lucarelli, The Book of the Dead of Gatseshen: ancient Egyptian funerary religion in the 10th century BC, Egyptologische Uitgaven 21 (Leiden, 2006) 231-32. 7 For an analysis of these terms in the Book of the Dead and Amduat, see Eschweiler, Bildzauber im alten Ägypten, 73-196. More recently, R. Nyord has been discussing the function of different kind of images in ancient Egypt: R. Nyord, Seeing Perfection. Ancient Egyptian Images Beyond Representation (Cambridge, 2020). 8 M. Heerma van Voss, s.v. ‘Vignette’, in W. Helk et al. (eds.), Lexikon Der Ägyptologie VI (Wiesbaden, 1986) 1043-44.

the “vignettes” of the greek magical papyri

3

describing: “a running ornament or small decorative design or picture put on or just before a title page or at the beginning or end of a chapter”.9 Such an association originated by the fact that the first scholars of the Book of the Dead were looking at the papyri produced during the Ptolemaic period, many of which are indeed characterized by displaying small illustrations on the top or bottom register of the scroll, as if they were a sort of decorative element. Instead, when looking at different variants of vignettes in the papyri of the early New Kingdom, it is clear that the illustrations of the Book of the Dead cannot just be indicated as “running ornament or small decorative design”; their dimension varies and large-scale vignettes can even occupy a full “page” of a papyrus, such as the vignettes of Spell 110, with the depiction of the Field of Offerings, and of Spell 125, with the scene of the final judgment of the deceased in the Hall of the Two Truths. Moreover, they can be defined as visual magical elements complementing the text and they are not inserted merely to embellish the scroll as ornamental “vignettes” do. Therefore, the vignettes of the Book of the Dead can and need to be compared to other images occurring in magical texts and books according to a broader perspective, which would consider their fluid symbolism and patterns of adaptation to the specific context of the spell or part of the text (such as the magical instructions), which they accompany. Similarly to the vignette of the Book of the Dead, the sketches of the Greek magical papyri fully interact with the texts that they accompany thanks to a common mental code of religious knowledge, which was probably meant to be shared by both the manuscript’s producer and user. Among the scattered sketches of the PGM, it is possible to individuate the ancient Egyptian origin of some of them. These new, much more stylized images are not used randomly as mere pictorial decorations of the manuscripts but they also relate to the new charms in proximity of which they have been copied. What follows are three significant examples of Book of the Dead motives re-used in the PGM. 1. The Scarab A very explicative example of magical drawings of Pharaonic origin, which have been transported into the Greek magical papyri, is the figure of 9 Merriam-Webster Collegiate Dictionary: https://www.merriam-webster.com/dictionary/vignette (09/06/2021). See also R. Lucarelli, The Book of the Dead of Gatseshen: Ancient Egyptian Funerary Religion in the 10th Century BC (Leiden, 2006) 230.

4

rita lucarelli

a scarab occurring in PGM II (Fig. 1). Scarabs are amuletic images par excellence, widely spread in the ancient Egyptian religious imagery as well as in the whole ancient Mediterranean. They occur on magical objects such as the so-called “scarabs of the heart” amulets, or as decorative elements of jewellery and as protective symbols alluding to the concept of rebirth on a wide series of funerary objects, from papyri to coffins.10 When occurring in the papyri of the Book of the Dead, the scarab mainly appears in variants of the so-called Spell 30B, on the protection of the heart. This popular incantation was also engraved or painted on the already mentioned heart-scarab amulets produced from the New Kingdom onwards. In the heart-scarabs, text and image are embedded into each other and can be considered as complementary parts of the same medium for amuletic magic and as an example of the “materiality” of magic.11 The scarab in question occurs in PGM II (P.Berlin inv. 5026), which is a magical handbook dated recently to the second/third century,12 with spells and instructions for receiving direct oracular visions and including charaktēres and a very popular sketch of the akephalos, a figure that has been discussed widely by scholars.13 The drawing of the scarab is part of an inscription for a doorpost of the bedchamber: “Below the door, [inscribe] the scarab, as it stands/here, having anointed it with the blood of a goat, outside your bedchamber”.14 Here the scarab is part of the instructions for the rite and we know that actually images as illustrative representations for the ritual and magical 10 On the amuletic function of scarabs, see C. Andrews, Amulets of Ancient Egypt (Austin, 1994) 50-60. 11 On the materiality of magic see D. Boschung and J.N. Bremmer (eds.), The Materiality of Magic (Paderborn, 2015) and C. Houlbrook and N. Armitage (eds.), The Materiality of Magic: An Artifactual Investigation into Ritual Practices and Popular Beliefs (Oxford, 2015). 12 See E. Chronopoulou, ‘Edition of the Greek Magical Papyri (PGM) I and VI+II: Introduction, text and commentary’, Doctoral Dissertation, University Pompeu Fabra (Barcelona, 2017) 27, available in http://hdl.handle.net/10803/460896 (13/03/2021). A new edition of all the Graeco-Egyptian magical formularies with English translation and a full revision of their material features (including new datings) is been carried out by the project “Transmission of magical knowledge: The papyrus magical handbooks”: https://papyrusmagicalhandbook.wordpress.com/ (13/03/2021). 13 See Martín Hernández, ‘Reading Magical Drawings’, 492 with bibliography. 14 The translations of the texts of the corpus of PGM are from GMPT. The mention of anointing the place with the blood of a goat in these instructions brings up an interesting comparison with the Israelite Passover protection ritual originated in Exodus 12, which has been compared to ancient Egyptian rituals implying the anointing of unguents by T.  Schneider, ‘Modern Scholarship Versus the Demon of Passover: An Outlook on Exodus Research and Egyptology through the Lens of Exodus 12’, in T. Levy et al. (eds.), Israel’s Exodus in Transdisciplinary Perspective (Cham, 2015) 537-53, in particular p. 547. I wish to thank Thomas Schneider for pointing out such intriguing comparative evidence to me.

the “vignettes” of the greek magical papyri

5

Figure 1. Right, detail of the scarab in PGM II © Staatliche Museen zu Berlin - Ägyptisches Museum und Papyrussammlung, Foto: Berliner Papyrusdatenbank, P 5026 Left, a B&W design made from the photo (R. Martín Hernández)

instructions, to be sketched by the user of the charm, were very common in the PGM. In general, this kind of sketches are rather simple and schematic and do not require artistic skills, differently from the highly elaborated vignettes of the illustrated funerary papyri of the Pharaonic Period. It is interesting to note that, if comparing the scarab of the PGM to those occurring in the magical papyri of the Pharaonic period, in particular in the scrolls of the Book of the Dead containing the already mentioned Spell 30B on the protection of the heart, a similar use of the motive of the scarab occurs as the central and only element of the vignette. Spell 30B contains ritual instructions as well, as we can read in the version from the papyrus of Nu, which is one of the longest and best preserved papyri of the Eighteenth Dynasty: “See, a scarab of nemehef-stone is made, outlined and purified with gold, placed within the heart of a man, and the Opening of the Mouth (ritual) is performed for him, anointed with scented oil.”15 A caption follows: ḏdt ḥr=f m ḥkȝw, “Spoken over him as magic.16 In these instructions, the scarab is strictly related to the ritual of the so-called “Opening of the mouth”, which serves to vivify the deceased’s physical powers. Therefore, the illustration of the scarab, similar to its later occurrence in PGM II, functions at the same time as an amulet and as a complementary object for the ritual. 15 Translation from S. Quirke, BD, Going Out in Daylight – prt m hrw the Ancient Egyptian Book of the Dead translation, sources, meanings (London, 2013) 99. 16 In Quirke’s translation, ḥkȝw is translated as “word power”; I opt here for the more comprehensive term “magic” since the instructions refer also to the ritual power of the performance and to the scarab amulet, not only to the words of the spell.

6

rita lucarelli

Figure 2. Scarab within the sun disk in Atum scene from Niwinski, 2015

In other magical papyri of the Pharaonic period, where images outnumber the text, the same iconographic element can actually be used within a much more intricate web of symbols in pictorial compositions symbolising the triumph of the deceased over death, where the scarab gains an additional function as cosmological symbol of the process of transformation and syncretism of the sun god between night and day. A good example is the image occurring in a funerary papyrus of the twenty-first Dynasty where Atum, in the form of a winged snake, has been depicted with a scarab in a sun disk in between the wings (Fig. 2).17 One cannot find this level of complexity in the sketches of the PGM but we can assume that the authors of some of these images knew them and found inspiration in it. And probably they were also still aware of the complex magical symbology imbedded in the figure of the scarab; from a lexical point of view, the scarab triliteral sign ḫpr, has a double meaning, indicating the verb for “to become” and the substantive “scarab” as well. The scarab is also a substitute symbol of the heart in magic and as such is placed very often on the chest of the mummy as an amulet. Finally, a principle of materia magica is 17 A. Niwinski, ‘Magic in the iconography of twenty-first Dynasty coffins and papyri’, in G. Bakowska-Czerner et al. (eds.), The Wisdom of Thot. Magical Texts in Ancient Mediterranean Civilizations (Oxford, 2015) 63 and fig. 6, p. 65. In the papyrus under discussion (BM EA 10018), the short and peculiar caption to the image runs as following: “Death, the great god who made the gods and the people”.

the “vignettes” of the greek magical papyri

7

involved as well in the magical instructions of the spells mentioning the scarab, when it is said that the amulet has to be made of a certain stone and colour.18 2. The Embalming Scene An example of a composite Book of the Dead vignette transmitted into the Greek magical papyri is the sketch of the embalming scene occurring on the verso of a demotic papyrus, PDM xii, namely Papyrus Leiden I 384, which is part of the so-called “Leiden-London magical papyrus” (Fig. 3).19

Figure 3. PDM xii, col. I. Drawing of the embalming scene by R. Martín Hernández.

This scene, which is also attested on magical gems in Late Antiquity,20 occurs widely in the Pharaonic mortuary papyri as a vignette of Spell 151

18

Eschweiler, Bildzauber, 76 ff. See J. Dieleman, Priests, tongues, and rites: the London-Leiden magical manuscripts and translation in Egyptian ritual (100-300 CE) (Leiden, 2005). 20 A. Mastrocinque, Sylloge Gemmarum Gnosticarum, vol. I (Rome, 2003) nos. 109-10, figure of N. 110; see also C. Sfameni, ‘‘Magicians’ instruments in PGM and the archaeological evidence: some examples’, in Bakowska-Czerner, The Wisdom of Thot, 101. On the occurrence on the embalming scene with Anubis on gems and its comparison with Sethfigures, see N. West, ‘Gods on small things: Egyptian monumental iconography on late antique magical gems and the Greek and Demotic magical papyri’, Pallas 86 (2011) 135-66 at 141 ff. See also the similar iconography on certain mummy labels, which have been recently compared with the faience amulets: R. Martín Hernández, ‘Fayence Mummy Labels Written in Greek’, ZPE 208 (2018) 193-202 at 198-200. 19

8

rita lucarelli

of the Book of the Dead (Fig. 4)21 but also on the walls of tomb-chapels of the New Kingdom.22 In Graeco-Roman Egypt the same image occurs on sarcophagi, mummy masks and wrappings as well; one of the last examples is in the tomb of Petosiris in Qarat el Muzzawaqa in the Dakhlah Oasis (second century BCE) and in the burial chamber of the Kom el-Shuqafa catacomb in Alexandria.23 The sketch in PDM xii proposes again the motive of Anubis and the mummy lying on a lion couch and, similar to the scarab of PGM II, seems to illustrate the instructions for a ritual; the attached text is a love spell of attraction written in Greek and Demotic, which has unfortunately numerous lacunae but seems to describe the same scene:24 “…you bring a sealed…of copper…this lion, this mummy (?), and this Anubis…while they seek…black scarab(?)…put…: …AIDIO ORICH THAMBITO, Abraham who at…PLANOIEGCHIBIOTH MOU ROU and the whole sould for her, HH [whom NN bore]…the female body of her, NN [whom NN bore], I conjure by the…[and] to inflame her, NN whom [NN bore]”. [Write these] words together with this picture on a new papyrus”.

It is interesting that the text instructs to sketch this picture on a “new” papyrus; similar formulations occur also in the funerary instructions of many incantations with amuletic function of the Pharaonic and Ptolemaic Egypt, including those of the Book of the Dead, according to which it seems that the power of the magical texts and images would work better when the writing material is “fresh, new papyrus”.25 We may assume therefore that we are facing, in the instructions of the demotic papyrus, another example of the inheritance of Pharaonic magical practices into Late Antiquity. Similarly to the case of the motive of the scarab, the quality of the sketch of the PGM is much inferior to that of the vignettes of the Book of the Dead with Anubis and the mummy on the funerary bed. However, simple sketches were not uncommon also in the papyri of the Pharaonic period used for daily magic, so that we can’t just assume that the quality of the drawings 21 See B. Luescher, Untersuchungen zu Totenbuch Spruch 151, Studien zum Altägyptischen Totenbuch 2 (Wiesbaden, 1998). 22 See references in Quirke, Going Out in Daylight, 375. 23 For the occurrences of Pharaonic motives in the Graeco-Roman tombs, including the Anubis scene, see M.S. Venit, Visualizing the afterlife in the tombs of Graeco-Roman Egypt (Maryland, 2016) 173, in particular for the tomb of Petosiris in the Dakhla Oasis. 24 PDM xii, 135-46 in GMPT 171, also PGM XII 474-79. 25 See for instance the so-called Spell 167 (Pleyte number) of the Book of the Dead: Quirke, Going out in Daylight, 535: “…and write it out on a new papyrus sheet, to be made as a book and place at his neck”.

the “vignettes” of the greek magical papyri

9

would have been influencing the efficacy of the spell. Moreover, the sketch in the demotic papyrus seems to be a mirror-image of the earlier representations, since Anubis and the mummy head are facing left instead of right.26 3. The “stele” A very intriguing composite image occurs in PGM VII 940-68 (P.Lond. I 121, dated to the fourth century, Fig. 5). The sketch is attached to “a charm to restrain anger and a charm to subject,”27 where it is said: “On clean papyrus write with pure myrrh ink these names together with the ‘stele’”.

Figure 5. Drawing of the “stele” in PGM VII by R. Martín Hernández

The probable image of what is indicated as a “stele” in the spell looks like a square with two crossed lines in the inside and two dog- or donkey/Sethlike heads coming out from the “stele”; a snake lies underneath the square. This is not the image of a traditional Egyptian stele but we do know that 26 In the Pharaonic period, only in the tomb of Sennedjem (TT 1) we can find a similar layout of this scene, as noted by Dielemann, Priests, Tongues and Rites, 34, n. 38. The scene of Anubis and the mummy, with the same orientation towards left as in the PGM, occurs also at least on the obverse of one gem kept at the University College in London and on a amulet found on mummy and dated to the Ptolemaic Period, where however only the mummy on the funerary bed is depicted, facing left as well, but not Anubis; see West, ‘Gods on small things’, figs. 3a and 8a. Tomb scenes, mummy cartonnages and shrouds of the Graeco-Roman period follow instead the Pharaonic tradition, with Anubis facing right. I wish to thank Lorelei Corcoran for discussing the scene orientation and providing examples from portrait mummies of Roman Egypt. 27 Translation by H.D. Betz in GMPT 143.

10

rita lucarelli

“stele”, in the magical papyri, means either “inscribed charm” or “drawing with or without text”.28 The latter – a drawing without text – is clearly the case for the sketch under discussion.29 The text copied below the sketch recites: “Come to me, you who are in the everlasting air, you who are invisible, almighty, creator of the gods. Come to me, you who are the unconquerable daimon. Come to me, you who are never grieved for your own brother, Seth. Come to me, you fire-bright spirit. Come to me, you god who are not to be despised, you daimon, and put to silence, subordinate, enslave him, NN, to him, NN, and cause him to come under my feet.”

The main interpretation, according to the text, is to see the squared box or “stele” as a representation of the coffin of Osiris while the two (donkey?-) heads on its top would represent Seth – who according to a different myth was protecting the solar boat from Apopis; the latter would have inspired the snake under the stele, as a symbol of the subjugation of the enemies as in the earlier Pharaonic rituals.30 According to a different interpretation, which also seeks inspiration from Pharaonic mythological motives,31 this sketch would resemble one scene occurring in the so-called Amduat, a funerary composition from the New Kingdom, used to decorate royal tombs and later on also copied on papyrus, describing the journey of the sun in the underworld, in order to unite with Osiris, during the twelf hours of the night.32 In the depiction of the seventh hour in particular, four coffins are represented, each of them 28 See J. Dijkstra, ‘The Interplay between Image and Text. On Greek Amulets Containing Christian Elements from Late Antique Egypt’, in Boschung and Bremmer, The Materiality of Magic, 271-93 at 278. On the use of the Greek term στήλη in the PGM as “text” or “magical instructions”, which is derived from the inscribed magical stone stelas of the Pharaonic Period, see R. Gordon, ‘Memory and Authority in the Magical Papyri’, in B. Dignas and R.R. Smith (eds.), Historical and Religious Memory in the Ancient World (Oxford, 2012) 14580 at 163-64. See also the entry στήλη in the Léxico de magia y religión en los papiros mágicos griegos available on-line: http://dge.cchs.csic.es/lmpg (14/03/2021). 29 In the same papyrus, on ll. 215-17, a stele with a text inscribed inside occurs instead. 30 See R. Gordon, ‘Shaping the Text: Innovation and Authority in Graeco-Egyptian Malign Magic’, in: H.F.J. Horstmanshoff et al. (eds.), Kykeon. Studies in Honour of H.S. Versnel (Leiden, 2015) 69-111 at 101. See also R. Martín Hernández, ‘More than a logos. The ιωερβηθ logos in context’, in C. Sánchez Natalias (ed.), Litterae Magicae. Studies in Honour of Roger Tomlin (Zaragoza, 2019) 187-209 and Chapter III in this volume. 31 I. Grumach, ‘On the History of a Coptic Figura Magica’, in D.H. Samuel and A.M.  Hakkert (eds.), Proceedings of the Twelfth International Congress of Papyrology, Ann Arbor, 13-17 August 1968 (Toronto, 1970) 169-81, at 170-71. 32 On the Amduat see E. Hornung, The Ancient Egyptian Books of the Afterlife (Ithaca, 1999) (trans. David Lorton); J. Darnell and C. Darnell, Ancient Egyptian Netherworld Books (Atlanta, 2018).

the “vignettes” of the greek magical papyri

11

presenting two human heads on the corners of the lid. As noted already by West however,33 the heads in the Amduat scene are human and they face each other, while the animal heads of the magical sketch in PGM VII face outwards. Moreover, the burial places of the Amduat present knives on their top and the serpent Apopis does not lie under them but it is depicted in a different area of the register.34 When looking instead for parallels into the earlier vignettes of the ancient Egyptian Book of the Dead, the peculiar design of the squared box, with the two crossed lines in the interior recall a couple of images occurring on papyrus too. One of them is a scene from the so-called Spell 149 of the corpus.35 The spell describes, in images and texts, the 14 jȝw.t, namely the “mounds” of the netherworld, together with its inhabitants. In particular, the tenth mound is depicted as a rectangular or oval shape with crossing diagonals in his interior (Fig. 6).36 The rectangular shape suggests the ancient Egyptian way to represent certain doorways and gates of the netherworld, as we can see in the vignettes of Spell 145 of the Book of the Dead, which often presents one or two diagonal lines within the rectangular shape of the door guarded by a demonic guardian (Fig. 7).37 33

West, ‘Gods on Small Things’, 143, n. 45. West also proposes a comparison with vaguely similar sketches occurring on a few magical gems, which show either a mummiform deity with two jackal heads or a jackal’s head coming out from a rectangular and oval object: West, ‘Gods on Small Things’, 143 and figs. 13. Such a parallel is certainly interesting since the gems are produced in the same period as the PGM; however, I would not consider the gems as a source of inspiration for the sketch on papyrus but rather consider this kind of iconographic motives – on papyrus and on gems – as similar re-interpretations of magical illustrations and “vignettes” of the Pharaonic period. An extensive comparison among the iconography of the magical gems and the Greek magical papyri, on the model of that one proposed by West but also extended to the magical papyri of the Pharaonic Period, would certainly be useful for a better understanding of the ancient Egyptian Bildzauber and of its transmission through Late Antiquity. 35 For a translation of Spell 149 and facsimiles of its vignette, see Quirke, Going Out in Daylight, 357-64. 36 Quirke, Going Out in Daylight, 357. I wish to thank Malcom Mosher for having discussed with me the occurrence of the crossed lines in this peculiar vignette in the Theban pre-Ptolemaic and Ptolemaic traditions and for pointing out that the later versions present many more variants of the rectangular/oval shape. In particular, a Theban early Ptolemaic papyrus depicts such a shape as a gate, similar to those of Spell 145; on this papyrus and its Theban Book of the Dead tradition, see M. Mosher JR, ‘An intriguing Theban Book of the Dead Tradition in the Late Period’, British Museum Studies in Ancient Egypt and Sudan 15 (2010) 123-72. 37 Cf. the Ptolemaic papyrus published by I. Munro, Der Totenbuch-Papyrus des Hor aus der frühen Ptolemäerzeit (pCologny  Bodmer-Stiftung CV + pCincinnati Art Museum 1947.369 + pDenver Art Museum 1954.61) (Wiesbaden, 2006) Pl. 20. It is interesting to note how in this papyrus, the rectangular shape of the tenth hill contains only one instead of two 34

12

rita lucarelli

Figure 6. Vignette of Spell 149. Sketch by Quirke 2013, p. 358

In the vignette of the tenth mound, the caption occurring near the rectangular or oval shape says: ntt ḥr r w῾r.t, “which is at the entrance of the district”, suggesting again that we are probably dealing with a gate. The vignette also shows a human figure holding knives and a snake facing the mound. Part of the spell also mentions a snake: ḏsr.tw n=j wȝ.t n῾w pw, kȝ Nwt Nhb-kȝw, “may be cleared for me the path of the naw-snake, the bull of Nut, Nehebkau”. We may therefore assume that the human figure with knives is the guardian demon protecting the mound while the snake depicted is Nehebkau, which according also to other magical and religious texts, was considered a benevolent and protective divine snake.38 The vignette of Spell 149 was very popular in the Ptolemaic Period and attested as much as its text since, as all the other illustrations of the Book of diagonal lines (Pl. 24), similar to other versions of the portals of Spell 145 (see for instance T. Allen, The Egyptian Book of the Dead: Documents in the Oriental Institute Museum at the University of Chicago [Chicago, 1960]). The rectangular shape with diagonal lines is occasionally used also to depict the Lake of Fire of Spell 125c, like for instance in pGatseshen: see R. Lucarelli, The Book of the Dead of Gatseshen, 221. The design of the portal with one diagonal slash across represents the ‘ḥ hieroglyph representing a palace or shrine . The latter appears already in the Old Kingdom and the diagonal slash represents maybe a curtain for the shrines brought on processional barques; I wish to thank Ogden Goelet for pointing to me the similarity of the rectangular figures with diagonal lines with the ‘ḥ buildings. 38 On Nehebkau, see the seminal study of A. Shorter, ‘The god Neḥebkau’, Journal of Egyptian Archaeology 21.1 (1935) 41-48. More recently, see M. Massiera, ‘The so-called statue of Nehebkau: a comparative study’, Journal of Intercultural and Interdisciplinary Archaeology 2 (2015) 25-33, where the numerous occurrences of Nehebkau as a snake more than as an anthropomorphic god are discussed.

the “vignettes” of the greek magical papyri

13

Figure 7. Vignette of Spell 145 from Munro 2006, pl. 20

the Dead, had a magical and ritual value on its own. The guardian-demons depicted in the spell, for instance, also occur in the second Osirian chapel of the temple of Hathor at Dendera, where an abridged version of the spell occurs in relation to the mysteries of death and rebirth of Osiris and where the mounds become part of Osiris’ domain.39 Although there is no clear evidence of a direct relationship of the vignette of the tenth mound of the netherworld from Spell 149 of the Book of the Dead with the sketch of the Greek magical papyrus, we cannot deny, at the same time, a certain stylistic similarity among the two. Such a similarity is certainly not less convincing than the one found in the previously mentioned comparisons with the gems’ sketches or with the mythological motives connected to Osiris, Seth and Apopis. In conclusion, I think that these kind of sketches deserve more attention and need to be compared with earlier Egyptian motives from magical objects. Besides the evidence given by the magical papyri such as those of the Book of the Dead genre discussed in this study, a comparison with the iconographic elements occurring on other kind of magical sources and 39 See S. Cauville, Le temple de Dendara: les chapelles osiriennes, (3 vols.) (Le Caire, 1997) at vol. 1, 363-64 and 372-73; commentary in vol. 2, 175.

14

rita lucarelli

monuments (from tombs to statues and statuettes, coffins, amuletic objects as well as graffiti) produced during the Pharaonic Period would help to disclose the essential role that the sketches of the PGM play as textillustrations. Similarly to when studying a richly illustrated papyrus of the Book of the Dead of the Pharaonic period, also in our study of the PGM, one needs to pay more attention to the images and to their interplay with the text, rather than considering the texts as the principal vehicle of that magical knowledge, for which the illustrations would be only a complement. In the autobiographical pages of his book The Torch in My Ear, the Nobel Price for Literature Elias Canetti wrote: “Pictures are nets; what appears in them is the holdable catch. Some things slip through the meshes and some go rotten. But you keep on, you carry the nets around with you, cast them out, and they grow stronger from their catches.”40 If one attempts to transport this idea from the introspective, autobiographic experience of Canetti to our scholarly study of the ancient Egyptian magical text-illustrations on papyrus, one may experience something similar. We could attempt to interpret these images as nets of earlier motives and symbols in need to be caught and hold in order to keep their magical power intact and stronger.

40 E. Canetti, The Torch in my Ear (New York, 1982) 113 (translated from German by F. Strauss).

II.

IMAGES OF TIED VICTIMS IN MAGICAL TEXTS Francisco Marco Simón and Celia Sánchez Natalías1

1. Knots and the bound victim motif One of the central concepts in magical practices is the knot. It is the very essence of magical power. Furthermore, the knot enables the long chain of connections that express the universal sympatheia, an element that ensures the permanency of the cosmos.2 Indeed, it is the almighty and supreme god who has the power to tie and untie, a power that Christ gave to Peter according to Matthew 16.18-19 (“I shall give you the keys of heaven: what you bind in the earth will be binded in heaven, and whatever you loose on earth will be loosed in heaven”). The knot’s centrality to magic is clearly demonstrated by the usage of the verb ‘to bind’ or ‘to tie’ which can designate magical practices.3 This verb is one of the essential words of power uttered or written by the magician, as it is made clear through the Greek term for the defixiones, katadesmoi  (from the verb deo, ‘to bind tightly’, and hence ‘to enchant’). The Latin terms fascinum, ligature and even defixio have similar meanings.

1 University of Zaragoza, Research Group Hiberus. This paper comes out of the research project HAR2014-57067-P, funded by the Spanish Ministerio de Economía y Competividad. 2 J. Annequin, Recherches sur l’action magique et ses répresentations (Paris, 1973). 3 Accordingly, this trend is already attested in the earliest periods of Egyptian history. It is certainly seen with HkA, a magical power that characterizes divine action and is morally ambiguous, given that it can be either generative or destructive (R.K. Ritner, The Mechanics of Ancient Egyptian Magical Practice [Chicago, 1993] 20-25). Likewise, the power of constraining or tying (Tz, Tzt) can either be positive or negative, protective or threatening; this is especially prevalent in the Late Egyptian and Graeco-Roman periods (W.  Wendrich, ‘Entangled, connected or protected? The power of knots and knotting in ancient Egypt’, in K.  Szpakowska (ed.), Through a Glass Darkly: Magic, Dreams and Prophesy in Ancient Egypt [Swansea, 2006] 243-69).

16

francisco marco simÓn and celia sÁnchez natalÍas

Furthermore, all these meanings point to death as the ultimate state of immobility or being bound.4 This paper aims to approach the diverse variants of the iconography of bound victims in magical texts, and specifically in curse tablets. This includes things buried or enclosed in containers, a subset of objects that, as we shall see, is of great interest to sites like the sanctuary of Anna Perenna. References to the tying or binding of targets crops up in a variety of places and periods. The term katadesmoi is already used by Plato (Resp. 364C) in the fourth century BCE, and it also appears in a contemporary defixio from the Kerameikos cemetery in Athens (NGCT 9), in which the text’s commissioner makes a magical bond against different people, to paralyze their minds, tongues, souls, works, words and possessions before a judicial process. Other defixiones from Attica tie the victims to infernal gods such as Hermes or Hecate (DTA 106 and 107). In another magical text from Metapontum, dated to the middle of the third century BCE, a magical bond is directed against a group of physicians (SGD 124). Another tabella from Olbia, dated to the third century BCE, includes the same act of binding (DT 89).5 This very trope is also documented among Latin texts, though curiously the verb defigo does not appear in many tablets. When it does crop up, it tends to be used in rather simple formulas that precede or follow the name(s) of a curse’s victim(s). The first known usage of the verb in a magical context can be traced to a first-century BCE defixio discovered in a necropolis at Mentana, where the following formula is repeated: “Malcio Nicones (…) defigo in (h)as tabel(l)as…”.6 To this first example, we can add several others from the first and second centuries CE discovered in Italy, Pannonia and Britain.7 Despite the fact that the verb defigo is almost a rarity in the corpus of Latin curse tablets, the idea of binding a victim is attested in the numerous examples of verbs like ligo, obligo and implico. The

4 For the ritual and performative value of writing-bonds, see M. Carastro, ‘Les liens de l’écriture. Katadesmoi et instances d’enchaînement’, in M. Cartry et al. (eds.), Architecturer l’invisible. Autels, ligatures, écritures (Paris, 2009) 263-91. 5 From the time of the Homeric epics, the knots, δεσμοί (desmoi), were already used by divine agents, such as Zeus who used them to subdue the Titans (Il. XV, 18-24; Hes. Th. 501-02, 521-22 and 718). Other examples include Hephaestus (Od. VIII, 266-81), Athena and Poseidon (Od., IV, 380 and 469; V, 383-85; VII, 272; XIV, 61). 6 Vid. DT 134 = SD 50. 7 For Italia, vid. the pieces from Cremona (AE 1975, 449 = SD 105) and Verona (AE 2016, 537= SD 108). The Tablet from Petronell (AE 1929, 228 = SD 530) comes from Panonia, while Britain offers three more examples, two from London (AE 1930, 112 and 1936, 87 = SD 338 and 339 respectively) and one from Clothall (RIB I, 221 = SD 345).

images of tied victims in magical texts

17

term defixio, in the ablative, is first attested in a Latin curse tablet dated to the first century CE from Abusina/Eining.8 The agonistic defixiones from Carthage, datable to the second or the third century CE, mention the magical binding of horses participating in races, as documented in the following formula: “Tie their race, their legs, their spirit and their speed, their strength…” (DT 241). Another interesting example from Carthage, also dated to the second or third century, curses the venator Gallicus, “to bind his feet, his limbs, his senses, his marrow…” (DT 247). This would have left Gallicus paralyzed during the spectacle and, as a result, he would have been devoured by the beasts that he himself was supposed to slaughter. In other texts with the same provenance, the Falernaro bathhouse is also bound in order that no one may bathe there (SGD 140-141). This demonstrates that not only animals and people, but even spaces could be bound. By expanding this survey to Late Antiquity, we could easily multiply the examples of the bound-victim motif in the corpus of defixiones. And indeed, two of the most remarkable texts come from Late Antique Egypt. The first one (SGD 157), from Oxyrhynchus, demands Eulamon “to bind down ([κα]τάδησον) the muscles, the members, the mind, the wits, the understanding, the 365 members and muscles” of the two runners. In another text, from an unknown Egyptian provenance (SGD 161), the defigens binds Teodotis, daughter of Etis, “to the tail of the serpent, the mouth of the crocodile, the horns of the ram, the venom of the asp, the hairs of the cat, the penis of the god” in order that she might not have any sexual relations with any other man besides the defigens. 2. Iconographical variants of the immobilized victim Besides textual references to binding, some defixiones also offer pictorial evidence for the knot’s centrality to many spells. When approaching these magical images, we must begin with the realization that they serve as didascalia, that is, they seek to explain or illustrate a curse’s aim. There are, after all, many curses with no images, but no images without texts (at least in the shape of a tablet: magical figurines would be the exception to this rule). That said, the text is always included after the image was already drawn; accordingly, when an image is present, inscribing the text forms the 8 The text has been edited by J. Blänsdorf, ‘Die Verfluchungstäfelchen aus dem Kohortenkastell Abusina/Eining: Defixionum tabellae Abusinenses (DTAbusina)’, Bayerische Vorgeschichtsblätter 84 (2019) 229-42 at 233; DTAbusina 1, l. 5, which reads “cum defictcsione”.

18

francisco marco simÓn and celia sÁnchez natalÍas

second phase of the piece’s manufacture. This inclusion of a picture was aimed to increase the power of a given spell. As a result, text and image come to compose a holistic unity that would maximize the efficacy and power of the operation.9 Fundamental to magical rituals is the sympathy principle. Through a similia-similibus formula, the symmetrical relationship between the manipulation or ritual distortion of the text was thought to have an effect on the victim.10 These are rituals which traditional historiography has termed as “sympathetic” or “homeopathic” magic, but which more precisely fall into Tambiah’s category of “persuasive analogy”.11 Bucking against dominant theories that consider homeopathic or sympathetic magic as based on the (poor) observation of empirical analogies, Tambiah differentiates “empirical analogy” used in scientific research to predict a future action from “persuasive analogy” used in ritual to stimulate future action. The key to this type of ritual does not lay in the poverty of science or in the failure to analyse scientific data, but rather in a strong belief in the persuasive power of certain types of formulaic language, in this case, a performative language.12 While the majority of defixiones bear no image, those that do contain iconography (approximately 11% of the pieces from the Roman West)13 largely come from the Mediterranean area, and specifically from Italia, Africa Proconsularis, Byzacena, and from cities like Carthage, Hadrumetum and Rome. Nevertheless, there are a few interesting examples that hail from different regions, such as the ensemble of curse tablets from Treveris.14 9 On this topic in general, see A. Viglione, ‘Le immagine figurate nei documenti magici’, in G. Bevilacqua (ed.), Scrittura e magia. Un repertorio di oggetti iscritti della magia GraecoRomana (Rome, 2010) 119-32, esp. 119-24; C. Sánchez Natalías, ‘Paragraphics an Iconography’, in R. Gordon et al. (eds.), Choosing Magic. Contexts, Objects, meanings. The Archaeology of Instrumental Religion in the Latin West (Roma, 2020) 101-18 at 106-17 and R.  Martin Hernández, ‘The Figural Representation of Victims on Agonistic Late-Antique Curse Tablets’, Religion in the Roman Empire 7.1 (2021) 96-124. 10 F. Graf, Magic in the Ancient World (Cambridge MA and London, 1997) 209-13. 11 S.J. Tambiah, ‘Form and Meaning of Magical Acts: A Point of view’, in R. Horton et al. (eds.), Modes of Thought (London, 1973) 99-229. 12 C.A. Faraone, ‘The Agonistic Context of Early Greek Binding Spells’, in C.A. Faraone and D. Obbink (eds.), Magika Hiera. Ancient Greek Magic and Religion (New York and Oxford, 1991) 3-32, esp. 8. 13 Sánchez Natalías, ‘Paragraphics and iconography’, 106. 14 G. Németh, Supplementum Audollentianum (Debrecen, 2013) provides an indispensable complement to A. Audollent (DT), since he collects all unedited drawings that Audollent did not include in his volume. For the collection from Treveris, see R. Wünsch, ‘Die Laminae litteratae des Trierer Amphitheaters’, Bonner Jahrbücher 119 (1910) 1-12.

images of tied victims in magical texts

19

As is well known, the use of defixiones in the Mediterranean is attested from the sixth century BCE onwards. Although in origin the practice developed mainly on the oral and textual level, throughout the High Empire new aspects, such as charaktēres, voces magicae, and iconography, were increasingly incorporated into the genre. As Daniel Miller has well documented, paradoxically “the immaterial” totally relies on “the material”:15 in the same way as sanctuaries contain many objects meant to give “the divine” material form, magical practices also require artifacts and techniques “intimately linked to claims to (social) power and (pragmatic) efficacy”, as Richard Gordon highlights.16 It is precisely in this realm that the use of iconography is found. In chronological terms, the first texts with images come from North Africa and date to the third century CE, while they are attested at Rome from the end of this century onwards, with a strong showing in the fourth and fifth centuries, as the pieces from the Anna Perenna sanctuary and the so-called “Sethian collection” demonstrate. As Richard Gordon has stated: “it was precisely in malign magic, generally considered the most difficult to perform, that special efforts were felt necessary to increase the force of the text”.17 When it comes to what is depicted on the tablets, images portray either the divine powers being invoked or the victim(s) of the curse. There are even several examples that show both divinities and victims in what appears to be a narrative structure. In these cases, the creativity of the defigens ‒or the expert assisting him‒ plays an essential role. At the beginning of the 1970s, an extraordinary cache of magical items was discovered in Egypt and then acquired by the Louvre. It contained a clay jug that held a lead tablet and a nude female figurine. On the tablet was written an amatory spell of subjugation that was supposedly dedicated to a deified Antinous and several infernal gods; the figurine is depicted on her knees with her hands bound behind her back. She was pierced with 13 needles. Scholarly investigation quickly linked this find to a recipe from the PGM (IV 335-406) that calls from the production of two figurines, one male in the form of armed Ares and the other female, depicted on her 15

D. Miller, ‘Introduction’, in D. Miller, Materiality (London, 2005) 1-50, esp. 27-29. R. Gordon, ‘From substances to texts: three materialities of ‘magic’ in the Roman Imperial Period’, in D. Boschung and J.N. Bremmer (eds.), The Materiality of Magic (Paderborn, 2015) 133-76 at 138. 17 R. Gordon, ‘Shaping the text: Innovation and authority in Graeco-Egyptian malign magic’, in H.F.J. Horstmanshoff et al. (eds.), Kykeon. Studies in honour of H. S. Versnel (Leiden, 2002) 69-111 at 107. 16

20

francisco marco simÓn and celia sÁnchez natalÍas

knees with hands bound behind her back. The papyrus goes on to explain that parts of the female figurine must be inscribed with magical words and pierced with 13 bronze needles. Next, both figures are to be attached to a lead tablet and then deposited in the tomb of someone who died either prematurely or violently. And indeed, the parallels between the Louvre ensemble, dated to the third or fourth century CE, and the ritual described in the papyrus are truly jawdropping. While stressing the Egyptian provenance of the pieces in the French museum, their editor nevertheless argued that the pierced female figurine is better understood in the context of the development of Greek defixiones.18 The iconography of the immobilized victim, however, is much older than the images found on any Greek curse tablet, given that similar images were already being produced in Pharaonic Egypt. Ritner has shown that the motif of the bound prisoner is attested in a royal predynastic tomb at Hierakonpolis. This motif would go on to be repeated in small figurines of enemies that represented human victims pierced with nails or needles, which were also accompanied by curses. The figurines would then be burned or buried. The best evidence for this long tradition is found at the fortress of Mirgissa in Nubia alongside the second waterfall, all of which was part of an official act meant to protect the community.19 In Mirgissa, the element that ties the whole together is the burial, an idea which signals ritualized death more strongly than breaking and piercing do. Following from this type of ritual, we find the occasional use of clay sarcophagoi that held cursed figurines. Furthermore, the enclosing of divine, political or personal opponents in a coffin appears in the myth of Osiris in the Book of Apophis. Burial would continue to enjoy a dominant place in magical rituals throughout the Graeco-Roman and Coptic periods, as is documented in Demotic and Greek spells (often of an erotic nature) as well as in various rituals which were adapted from temple rituals aimed at enemies of the state or averse demons to the private realm.20 18 P. Du Bourguet, ‘Une ancêtre des figurines d’envoûtement percées d’aiguilles, avec ses compléments magiques, au Musée du Louvre’, Mémoires publiés par les membres de l’Institut français d’archéologie orientale du Caire 104 (1980) 225-38. 19 Ritner, The Mechanics, 153 ff. Among the aspects of this ritual, which included the remains of human sacrifice, appeared three full limestone figurines which depicted prisoners with tied hands as well as the head of a fourth figurine. 20 R.K. Ritner, ‘Egyptian Magical Practice under the Roman Empire: the Demotic Spells and their Religious Contexts’, Aufstieg und Niedergang der römischen Welt II, 18.5 (Berlin and New York, 1995) 3333-79. The figurine Mnesimachos, which was discovered in a coffin from the necropolis from the Kerameikos, provides a great example of the “total binding” of a victim, by symbolically representing his burial with a container and figurine. This

images of tied victims in magical texts

21

Fig. 1. Ursus, apud Wünsch 1910, Tab. II, fig. 1.

Turning our attention back to defixiones, we can point to several different strategies for representing victims on curse tablets. Sometimes, they are shown in bust-like portraits (see fig. 1), a practice which could allude to the conventions of elite portraiture or, as R. Martín Hernández has recently suggested, as the severed head of the victim.21 More frequently, however, we find the victim’s entire body depicted. Showing the target from head to foot can be seen as a way of making the curse more expansive so as to include each and every limb and organ: whatever is pictorially shown is vulnerable. This type of representation should be taken as an iconographic analogue of the long anatomical sequences found in some defixiones, which perform a sort of ritual dismembering of the victim.22 Among the full-figure representations of the victim, the first attested example hails from Carthage and has been dated to the third century CE. The curse’s target, an auriga, is shown with hands and feet crossed and figure has hands bound behind his back and the victim’s name is inscribed on the figurine’s left leg (J.G. Gager, Curse Tablets and Binding Spells from the Ancient World [New York and Oxford, 1992] no 41). For the containers from Anna Perenna, see below. 21 Such as the example of Ursus (see fig. 1), who was perhaps a gladiator and was represented on a Late-Antique tablet from the amphitheatre of Augusta Treverorum. Vid. Wünsch, ‘Die Laminae litteratae’, no 18 = SD 179. See Martin Hernández, ‘The Figural Representation’. 22 Vid., for instance: DT 135 = SD 51, DT 190 = SD 56, etc. A wonderful example is the defixio against Caecilia Prima: for this piece, vid. AE 2007, 260 = SD 48. Also, see the classic article by H. Versnel, ‘An Essay on Anatomical Curses’, in F. Graf (ed.), Ansichten griechischer Rituale (Stuttgart and Leipzig, 1998) 217-67, and R. Gordon, ‘Diagnosing the signs: The body as subject in Latin defixiones’, in Gordon et al., Choosing Magic, 93-99.

22

francisco marco simÓn and celia sÁnchez natalÍas

Fig. 2. The aurigae Artemius, Eutimius and Eugenius immobilized, apud SV 20.

tied.23 Charioteers also strike similar poses on several of the superb Sethian texts from Rome. This is certainly the case with Artemius, Eutimius and Eugenius. The immobilization of these three targets is further stressed by the representation of the god Osiris within a coffin (SV 20, see fig. 2). Similarly, another defixio (SV 27) exhibits the drivers Eutimius, Artemius and Gregorius with their hands and feet bound. Another tabella from the same collection (SV 28) asks for Restutus, the driver for the blue team, to be chained up. Next to this driver, we can also make out a staircase with 23 D.R. Jordan, ‘New defixiones from Carthage’, in J.H. Humphrey (ed.), The Circus and Byzantine Cemetery at Carthage I (Ann Arbour, 1988), 117-34, esp. 130, fig. 2.

images of tied victims in magical texts

23

Fig. 3. The charioteer Artemius, apud SV 29.

curved upper steps, which seems like a metaphorical means of provoking the fall of this charioteer. In the Sethian collection there is another defixio against Artemius, who is represented standing between two daimones who lock him up in chains in an attempt to immobilize this charioteer and keep him from attaining victory in the race (see fig. 3).24 As Faraone has argued, these images have an “analogous” function to ancient “voodoo dolls”, which also seek to constrict symbolically the victims of the curse.25 In line with the similarities between the tablets and 24

Vid. SV 29. C.A. Faraone, ‘Binding and burying the forces of evil: The defensive use of “voodoo-dolls” in Ancient Greece’, Classical Archaeology 2 (1991) 165-220 at 220. The lists compiled by Faraone have been updated by A. López Jimeno, ‘Una figurita de plomo hallada en Paros y otras figuritas de magia maléfica’, MHNH 10 (2010) 101-18 at 107-11, M. Bailliot, ‘Roman Magic Figurines from the Western Provinces of the Roman Empire: an Archaeological Survey’, Britannia 46 (2015) 1-18 at 6-14, and G. Németh, ‘Voodoo Dolls in the Classical World’, in E. Németh (ed.), Violence in Prehistory and Antiquity (2018) 179-94. In addition to the pieces treated by these authors, we must add two examples from Córdoba and Masatrigo: cf. A. Ventura Villanueva, ‘Magia y superstición: tabellae defixionum’, in D.  Vaquerizo (ed.) Funus cordubensium. Costumbres funerarias en la Córdoba romana 25

24

francisco marco simÓn and celia sÁnchez natalÍas

figures, it is possible to draw a line of influence between the figurines and the full-figure representations on defixiones.26 Given that these figurines can be dated much earlier, they could well have served as a model for the sort of iconography found on later tablets, since the analogies between the poses are markedly similar. There are four other striking parallels for these Carthaginean and Roman charioteers with bound hands. The first is the extraordinary Etruscan figurines from Sovana. Discovered in a necropolis and dated to the fourth century BCE, these two figures are depicted in a standing position with their hands tied behind their backs and their names inscribed on their legs.27 The second is a small male figurine from Keos dated to the fourth century BCE, which is now located in the Staatliche Museen in Berlin. In addition to having its hands tied at the back, there is a list of seven names inscribed on the piece28, a fact that implies that a single figurine could also be used for cursing multiple victims, as is the case with lead tablets. The third example is a figure from Brixia/Brescia, datable to the imperial period, that represents a bearded man standing with his legs together and whose arms have been crossed behind his back. What is noteworthy here is that the figurine was intentionally burned, an action that led to the loss of the feet and likely damaged the genital area. Though scholars once believed that the figure represented Silenus, it has now been reinterpreted as a magical figurine, due to the pose and its leaden material.29 Finally, we should mention a lead figurine discovered during the excavation of a necropolis in Puteoli/Pozzuoli (dated to the first century CE). The figure, deposited on top of tomb 101, represents a man in the same position as the aforementioned one from Brixia/Brescia, with the difference that his right eye was pierced with a nail that goes through the back

(Córdoba, 2001) 194-95, and A. Ventura Villanueva, ‘Magia en la Córdoba romana’, Anales de Arqueología Cordobesa 7 (1996) 141-62 at 149-50. 26 The lead piece from Carystos, dated to the fourth century, provides a sort of border case that straddles the line between curse tablets and lead figurines, since the piece is anthropomorphic in shape and is covered by an inscription which reads: “I inscribe Isiade, the daughter of Autocleia, together with Hermes Katochos. Hold her below with you. I tie Isiade together with Hermes Katochos. Her hands, her feet, her entire body.” This piece serves as a sort of sôma-sêma, to use a Platonic image, cf. Carastro, ‘Les liens de l’écriture’, 272-78, 286 and fig. 3. 27 Faraone, ‘Binding and burying’, no 14. Vid. as well the new edition of these figurines in R. Massarelli, I testi etruschi su piombo (Pisa, 2014) 214-17. 28 J. Curbera and S. Giannobile, ‘A ‘voodoo doll’ from Keos in Berlin’s Antikensammlung’, in Boschung and Bremmer, The Materiality of Magic, 125-32. 29 Vid. M. Bolla, ‘Un caso di magia a Brixia’, Numismatica e antichità classiche: quaderni ticinesi 46 (2017) 119-30.

images of tied victims in magical texts

25

of the skull.30 Indeed, it’s quite similar to the previously mentioned pieces and it fits in nicely with the two-dimensional representations of victims portrayed on curse tablets. Turning back to the Carthaginean and Roman charioteers depicted in curse tablets, it is worth noting that the position of the aurigae and the figurines is not identical: the charioteers’ hands are represented in front of their bodies instead of behind. Such a difference in representation, however, is based on the media used: the two dimensionality of drawing on the defixiones precludes representing the bound hands behind the back, but nevertheless serves the needs of the spell, since clearly showing the act of binding is central. Amongst the imagines magicae, it is worth turning to the ‘Sethian’ defixiones, since they add two new motifs to the mix: the inclusion of snakes and victims that are partially or totally mummified. Let’s start with serpents. Though it is beyond dispute that reptiles play a large role in Roman art, frequently appearing in murals, on ceramics, jewelry, etc., their debut on the imagines magicae is owed to a new perception of these animals in Late Antiquity. Traditionally, Roman culture associated the serpent with the ἀγαθὸς δαίμων (agathos daimon), the ancestral tutelary power and domestic cults with an apotropaic character.31 That said, the serpent was also associated with something more violent and dangerous through the diffusion of myths and stories such as Medusa or the priest Laocoon, who along with his children was enveloped and killed by snakes. Given this background, this animal was ripe for reconceptualization in the realm of aggressive magic in Late Antiquity. In this period, the snake was associated with demonic forces and within the context of curses it became a synonym for constriction and danger. Accordingly, when found in curse tablets of this period, the snake is always represented as approaching or threatening its victim, who is shown in a helpless position. Let’s look at a few pieces. An amatory tabella from the sanctuary of Anna Perenna against somebody “quem pereo fantasia” provides a great example. Dated to the fourth or fifth century CE, it shows a male figure tied up in many ropes. This poor fellow is presumably the victim who is “the love object of a woman who wanted to bind him to her forever”.32 The 30 Vid. C. Gaianella, ‘Una nuova figurina in piombo da Puteoli’, Bollettino di Archeologia 39-40 (1996) 165-68. 31 F. Marco Simón, “El culto a Draco(nes) en el Occidente latino”, in J.F. Martos Montiel et al. (eds.), Plutarco, entre dioses y astros, vol. II (Zaragoza, 2019) 899-914. 32 M. Piranomonte, ‘The discovery of the fountain of Anna Perenna and its influence on the study of ancient magic’, in G. Bąkowska-Czerner et al. (eds.), The Wisdom of Thoth: Magical Texts in Ancient Mediterranean Civilisations (Oxford, 2015) 71-85, esp. 81.

26

francisco marco simÓn and celia sÁnchez natalÍas

figure is flanked by two snakes and is depicted standing. At his feet, we can find something difficult to interpret (maybe his stomach and legs?) and the inscription. Another tantalizing fragment belongs to the “Bologna collection”, which is dated between the fourth and fifth centuries. Here we find an unparalleled scene.33 In the lower portion, we can see a snake with a tripartite tongue, scales and pointed ears. It is coiled up, with head erect, and ready to attack. The reptile looks towards the upper register of the tablet where an auriga, undoubtedly the curse’s victim, is depicted with his limbs stuck to his sides and a frightened expression. Perhaps the most striking aspect of the composition is a “string” that connects the snake to the victim: extending from the reptile’s body this string rises up and confines the target in a sort of coccoon. Though without a clear parallel, this almond-shaped structure is a clear representation of the constriction and immobilization of the curse’s victim. While the examples just examined show snakes closing in on their prey, there are other tablets in which the snake is already in contact with its target. The first example of this motif dates to the second or third century CE in a curse tablet from Carthage. On the tablet we can see a victim ensnared in a rectangular object (perhaps a sarcophagus?) from which only his head emerges. A large snake with scales is coiled at his side. Though the snake’s head is not preserved, we can still glimpse its tongue, which brushes the face of the terrified victim just moments before the reptile devours its prey.34 In addition, in seven of the abovementioned “Sethian” defixiones, other human figures are depicted with a snake wrapped around them.35 One of these tablets targets Kardelos, the son of Fulgentia, and contains a quite noteworthy composition. In the upper register of the tablet we can find the invoked numina: Osiris emerging from a sarcophagus, a daemon with a horse’s head (of disputed identification)36 and two πάρεδροι (paredroi = 33 The exact provenance of the piece is unknown and was found alongside two other tablets in the Museo Archeologico Civico di Bologna. The fragment discussed here is currently being studied by C. Sánchez Natalías. For a preliminary report, see C. Sánchez Natalías, ‘Magia et circenses’, in J. López Vilar (ed.), Tarraco Biennal. III Congreso Internacional de Arqueología y Mundo Antiguo. La gloria del circo. Carreras de carros y competiciones circenses (Tarragona, 2017) 49-54 at 52-53. 34 See DT 245 (editio princeps) and also Németh, Supplementa, 101 and 166, which includes the drawing of the curse tablet. 35 See specifically, SV 16-18, 34, 36, 43 and 45. The target is mummified in SV 16-18, 34 and 43. 36 See Chapter III in this volume.

images of tied victims in magical texts

27

Fig. 4. Adeodatus, apud SV 17.

assistants). In the lower register and perpendicular to the divinities above, we see a mummified figure surrounded by two large snakes which are about to bite the victim’s face (which is now lost). Another Sethian tablet, which takes aim at the auriga Adeodatus, presents a strikingly similar iconographical layout, with the sole difference being that in the lower register there is only a single snake of great size that is about to devour a mummified figure (see fig. 4).37 This figure has been interpreted in multiple ways: for the text’s editor, it is a representation of Osiris surrounded by a snake, whereas for other scholars it is simply a mummy.38 That said, the figure’s position in the larger composition (depicted in a supine posture in the lower part of the tablet) as well as parallels of analogous date lead us to believe that this figure is most likely the victim of the curse.39

37

Vid. SV 17. And also Audollent in DT 156, 157 (where he describes: “Osiris «momie» duobus serpentibus implicata”), etc. Other authors like D. Ogden, Magic, Witchcraft, and Ghosts in the Greek and Roman Worlds (Oxford, 20092) 214 and G. Bevilacqua, ‘Defixio contro Kardelos dalla Via Appia’, in R. Friggeri et al. (eds.), Terme di Diocleziano. La collezione epigrafica (Rome, 2012) 607-10 at 607, identify this figure as a mummy. 39 Viglione, ‘Le immagini figurate’, 121, where she states that the figure is “verosimilmente la vittima dell’azione magica”. In a similar vein, see C. Sánchez Natalías, ‘Magical Poppets in the Western Roman Empire: a case study from the Fountain of Anna Perenna’, in T. Minniyakhmetova et al. (eds.), The Ritual Year 10. Magic in Rituals and Rituals in Magic (Innsbruck and Tartu, 2015) 194-202 at 200. Of the same opinion is Martin Hernández, ‘The Figural Representation’. 38

28

francisco marco simÓn and celia sÁnchez natalÍas

Fig. 5. Magical figurine of Seberinus surrounded by a snake. Museo Nazionale Romano delle Terme di Diocleziano. Dipartimento Epigrafico (inv. no. SAR 475550). Su concessione del Ministero per i beni e le attività culturali e per il turismo –Museo Nazionale Romano.

Among the parallels for use of serpents, we must underline an exceptional figurine discovered in the fountain of Anna Perenna at Rome. The relevant piece was discovered within a series of three leaden containers and represents the victim of the curse surrounded by a giant snake, which is poised to devour the victim’s head (see fig. 5).40 Furthermore, the snakevictim pair is attached to a tablet. The similarity between this threedimensional figurine and the representation on the defixio is striking and goes to demonstrate that curse tablets and magical figurines are indeed two parallel ways to attack a victim.41 Another intriguing parallel can be seen in the first of the Bologna curse tablets (see fig. 6).42A figure is represented standing barefoot in the foreground, with hands bound at waist-level. On his chest a magical sign and 40 Vid. M. Piranomonte, ‘Anna Perenna. Un contesto mágico straordinario’, in M. Piranomonte et al. (eds.), Contextos mágicos. Contesti magici (Rome, 2015) 159-74, at 172-73, fig. 35. 41 Faraone, “Binding and burying”, 220. 42 Vid. C. Sánchez Natalías, ‘The Bologna defixio(nes) Revisited’, ZPE 179 (2011) 201-17.

images of tied victims in magical texts

29

Fig. 6. Curse tablet against Porcellus the veterinarian. Museo Archeologico Civico di Bologna (inv. no. 2238-2241). Image created by C. Sánchez Natalías.

some onomata barbarika have been written, while an eight point star is depicted in the genital area. Three sinuous snakes emerge from the right side of his/her crowned head, and what remains of two others emerging from the left (we can infer that there was originally a third snake on the left as well, not only to respect symmetry, but also judging from comparisons with the iconography of the second defixio from Bologna). Several factors point to an identification of the standing figure as the infernal power that was invoked by the defigens, perhaps Hecate-Selene.43 43 Vid. Sánchez Natalías, ‘The Bologna defixio(nes)’, and C. Sánchez Natalías, ‘Las defixiones de Bolonia: ¿un nuevo retrato de Hekate-Selene?’, in E. Suárez de la Torre et al. (eds.), Mito y Magia en Grecia y Roma (Barcelona, 2013) 273-81 at 276-80.

30

francisco marco simÓn and celia sÁnchez natalÍas

One factor that supports this identification is the contrasting position of the mummified victim Porcellus, who is shown supine below. In our opinion, the agency of the goddess is clearly represented through the snakes which emerge from her head, which markedly differs from the examples of snakes moving towards the head of the victim. Furthermore, the deity’s head is crowned, which is not the case of the victim Porcellus. An iconographical analogy of binding is established between deity and victim. In a quite sophisticated narrative that is almost cinematic, the agency of the goddess is made manifest: the departure of the snakes emerging from her head is meant to signal, through “persuasive analogy”, the constriction of Porcellus, he is seen as a passive and suffering figure already tied up and lying immobile on the ground.44 In this case, the identification of the victim Porcellus is beyond dispute, since his name appears written on both arms. Though the mummies represented on the above mentioned Sethian tablets do not include their victims’ names written on their bodies, but rather they are found off to the side;45 this leaves little room to doubt that these are indeed representations of the curses’ victims. But what remains to analyse is exactly what mummification represents in this context. The Sethian texts against Kardelos and Adeodatos ask the invoked divinities to punish the victims in the “bed of torture” (κατὰ κράβατον) with an awful death. It cannot be determined if the mummification is a pictorial representation of this “bed of torture” which the curses invoke. What can be said clearly is that mummification seems to be the ultimate symbolic representation of the curse’s paralyzing power to constrict and bind the victim: as a mummy is bound and cannot move, so the victim of the curse cannot escape his/her fate. This logic would make perfect sense in line with the magical principle of similia similibus mentioned above.46 In all likelihood and as the texts’ editor has noted, the image of the mummy takes as its model the image of Osiris wrapped in snakes. Good examples 44 This icon of the divine agency represented through the first figure of this Bologna defixio corresponds exactly to the iconography of the second tabella with the same provenance, which is also datable to the fourth or fifth centuries CE (on this, see C. Sánchez Natalías, ‘Fistus difloiscat languat… Re-reading of defixio Bologna 2’, ZPE 181 [2012] 14048). As in the previous example, the crowned figure has hands tied at waist-level, an eight point star on over her genitals; six snakes with mortiferous tongues emanate from his/her head and flank the divinity. Like in the case of Bologna 1, onomata barbarika appear on the chest of the figure and surround her. In this case, the victim is Fistus the senator, and the deity has been summoned to kill him and dissolve his members. 45 Vid. SV 16, ll. 34-37 and SV 17, l. 38. 46 C. Sánchez Natalías, ‘Le defixiones della Tarda Antichità e la loro iconografia’, Chaos et Kosmos 14 (2013) 1-19 at 10.

images of tied victims in magical texts

31

of this motif can be found on magical gems, where Osiris can be seen covered in snakes.47 Indeed during Late Antiquity when new motifs were imported, adopted and repurposed in the Roman West, the influence of Egypt on Graeco-Roman magic is unmistakable.48 Perhaps the clearest expression of this idea of confining and tying a curse’s victim is found in the rituals of aggressive magic carried out in the above-mentioned sanctuary of Anna Perenna at Rome.49 At this site, seven series of lead containers, which probably represent miniature cinerary urns, were unearthed.50 For the most part, this series of artifacts is composed of sets of three nested containers (somewhat resembling Russian dolls). Though the outermost containers do not seem to bear any inscription, the inner two vessels tend to have inscriptions and/or drawings. Several depict a cock-headed figure, identified as Abraxas,51 whose chest contains the inscription “ΙΧΝΟΠ/ ΧΝΚΘ/ΘΘ” or a slight variation of the same formula. G. Néméth has convincingly interpreted these Greek letters as the acronym “Ἰησοῦς Χριστὸς Ναζωραῖος ὁ παῖς, Χριστὸς Ναζωραῖος καὶ θεὸς, θεὸς, θεὸς”, that is: “Jesus Christ the Nazarene the child, Jesus Christ the Nazarene and God, God, God”.52

47 Amongst others, A. Mastrocinque, Les intailles magiques du Département des monnaies médailles et antiques (Paris, 2014) no. 73 and 166-verso, where the mummy or Osiris is engulfed by the snake in the same way as the magical figurine from Anna Perenna. Though iconographically similar, the so-called “idol of Gianicolo” has a quite different meaning. The small bronze statue depicts a figure with covered head and a body surrounded by a snake, whose head points towards the figure’s face. The statuette was discovered in a rectangular receptacle within the triangular altar in the Roman sanctuary for the Syrian gods and nearby votive offerings such as eggs, flowers and seeds, all of which is thought to belong to the fourth century. The overall character of this offering stresses fertility and annual renewal, all of which is further reinforced by the syncretic nature of the figure who is surrounded seven times by the snake. The number seven alludes to the planets and eternal circularity personified by Aion, cf. C. Gorgognoni, ‘Idolo’, in R. Friggeri et al. (eds.) Terme di Diocleziano. La collezione epigrafica (Rome, 2012) 666-67, for further references. 48 On this, see Gordon, ‘Shaping the text’. 49 An interesting parallel of similar date can be found in the Reims magical figurine which has been published by Bailliot (vid. Bailliot, ‘Roman Magic Figurines’, 12, F10). 50 C. Sánchez Natalías, ‘Escribiendo una defixio: los textos de maldición a través de sus soportes’, Acta classica Universitatis Scientiarum Debreceniensis 47 (2011) 79-93 at 88-89. 51 J. Blänsdorf, ‘Contenitore magico iscritto’, in Friggeri, Terme di Diocleziano, esp. 61920, 622 and 624-25, Piranomonte, ‘Anna Perenna. Un contesto’, 167-71 and M. Piranomonte, ‘IX, 49. La Fontana di Anna Perenna’ and ‘Contenitore magico iscritto’, in Friggeri, Terme di Diocleziano, 618-19 at 618 and 625, with the preceding bibliography. For another interpretation of this representation, see Sánchez Natalías, ‘Paragraphics and Iconography’, 109. 52 G. Németh, ‘Jesus in ancient pagan magic. The Anna Perenna drawings’, in G. Bąkowska-Czerner et al. (eds.), The Wisdom of Thoth: Magical Texts in Ancient Mediterranean Civilisations (Oxford, 2015) 43-48.

32

francisco marco simÓn and celia sÁnchez natalÍas

A constant that runs throughout the entire series is that the innermost container holds a magical figurine that undoubtedly represents the victim of the curse. A common feature that links all of these figurines together is how they were placed in the containers: they were put upside down with their heads touching the bottom of the vessel. This placement is far from meaningless, but rather is one further means of eschewing the victim’s chance of escape from the set of containers.53 Trapped within this tripartite cinerary urn which was hermetically sealed with resin, the victims were deposited into the fountain of Anna Perenna, where they remained dedicated to the sanctuary’s goddess and her nymphs. This type of ritual is a new riff on the old motif of “complete constriction” and gives evidence for a curse in which the victim was symbolically and inescapably buried. 3. Mary, Untier of Knots In the church of St. Peter am Perlach in Augsburg, Germany, there is a picture of Johann Georg Melchior Schmidtner (1625-1705) dated to 16991700. In typical Baroque fashion, the painting represents Madonna standing on a crescent moon and accompanied by an apocalyptic description (12.1): “A great sign appeared in heaven: a woman dressed of sun, with the moon at her feet and a twelve star crown on her head”. In the painting the Virgin is represented in heaven where she unties the knots of a long white ribbon, which an angel to her left offers her; having untied the knots, she then passes the ribbon to another angel to her right. Popular devotion has termed this Madonna as “Untier of Knots”. Although this worship began in Germany at the beginning of the seventeenth century, it is actually in Latin America where it has made its greatest impact, due to the influence of cardinal Jorge Bergoglio (now Pope Francis), who as a student of theology, had visited the church in Augsburg many years ago where she made an indelible impression on the young Bergoglio. The image celebrates a miracle that took place on 25 September, 1615, when the knots that were endangering the matrimonial bond between Wolfgang Langenmantel and Sophie Imhoff were untied thanks to the intercession of the Virgin.

53 Archaeologically, we can compare this to the decubito prono (“face down”) burials of some restless dead, whose posture was meant to curtail the spirit from exiting the tomb. On this topic, vid. S. Alfayé Villa, ‘Sit tibi terra gravis: Magical-religious practices against restless dead in the ancient world’, in F. Marco Simón et al. (eds.), Formae Mortis: el tránsito entre la vida y la muerte en las sociedades antiguas (Barcelona, 2009) 185-215 at 210.

images of tied victims in magical texts

33

This image of Mary untying the knots recalls the passage of Irenaeus of Lyon (died in 202), who in his treatise Adversus haereses (3.22.4) writes the following: “The knot of Eve’s disobedience was untied by Mary’s obedience. What Eve had tied with her incredulity, was untied by Mary through her faith”. Although there is no direct evidence that the painter knew this patristic text and hence that there was a direct link between Irenaeus’ passage and the painting, the context of the picture undoubtedly alludes to the symbolic representation of the triumph of the good over the evil, of the Virgin Mary over Eve, who is assimilated to the crescent moon, which also evokes the Ottoman empire (and perhaps the protestant heresy) in the visual language of the Counter Reformation. But we want to call attention to a further text that plays upon the same ideas and imagery as Irenaeus’ text and the German painting of our Lady Untier of Knots. It is a Coptic text called “The Magical Book of Mary and the Angels”, a codex of ritual power that contains a hodgepodge of recipes and spells, which has been studied by Marvin Meyer.54 According to the version from the Heidelberg and the London collections, “I am Mary, I am Mariham, I am the mother of the life (of) the whole world (…) Let the rock split before me today, let the iron dissolve before me today, let the demons withdraw before me today, let the powers of the light appear to me, let the angels and the archangels appear to me today, let the doors that are bolted and closed for me, at once and quickly, so that your name may become my helper and life, whether in all the day or in all the night” (2, 6-3, 11).55 Versions of this prayer are known from Coptic, Ethiopic, Syrian and Arabic sources, as well as in a fuller narrative which tells that Mary is said to offer her prayer in order to deliver Matthias (the replacement for Judas Iscariot according to Acts 1, 26) from prison. It is therefore a prayer to dissolve the iron fetters and to open the prison doors (ibid. 411). 4. Conclusion Beginning with the consideration that the act of binding or tying is central to many magical practices like the use of καταδεσμοί and defixiones, we have sought to examine the variations and themes in the iconography of tied victims, focussing especially on the corpus of curse tablets. These

54 M. Meyer, ‘The Prayer of Mary who dissolve the chains in Coptic Magic and Religion’, in P. Mirecki and M. Meyer (eds.), Magic and Ritual in the Ancient World (Leiden, 2002) 407-15. 55 Meyer ‘The Prayer of Mary’, 410.

34

francisco marco simÓn and celia sÁnchez natalÍas

images illustrate and complete the ideas central to the texts that they accompany and hence are meant to increase the power of the magical act. Such images are not attested on defixiones before the Roman imperial period. That said, the motif of the tied victim is much older and is already found in artefacts from Pharaonic Egypt. We have suggested that it is quite plausible that these images found on curse tablets took the conventions of older magical figurines as a source and inspiration. The so-called “voodoo dolls” from places like Kos and Sovana attest to the centrality of the binding motif to Graeco-Roman magic since at least the fourth century BCE. Among the different versions of tied victims that arose in Antiquity, those that include snakes eating or attacking their victims from Carthage, Rome and Bologna stand out as particularly intriguing. Likewise, the finds from the sanctuary of Anna Perenna that emphasize binding and burial by enclosing figures in lead containers deserve special attention. We conclude with several later texts and the peculiar iconography of Our Lady Untier of Knots to suggest that this nexus of imagery may draw on some conceptions of loosening knots as part of counter-magical spell. If this be the case, this is just one piece of evidence for the longevity, influence and afterlife of the many magical practices described in this paper.

III. EULAMO VS. SETH. ON THE EQUINEHEADED DEMON REPRESENTED IN THE DEFIXIONES FROM PORTA S. SEBASTIANO ROME Raquel Martín Hernández

The study of the iconography and visual elements acting in magic greatly contributes to our comprehension of the whole ritual and leads us to think differently about the processes of communication between human beings and divinities, which of course are not only based in oral communication and its written version.1 The most common visual elements used in Graeco-Roman magical books and artifacts can be individualized in four types: 1) figural representations, 2) diagrams like stelae or seals 3) sentences, words and vowels written forming different shapes like triangles and squares and 4) the popular charaktēres.2 All of these elements, of course, can appear in magical texts and artifacts combined in multiple ways. Non-textual elements are particularly well attested in aggressive rituals, but figural representations of demons and targets of the curses are, of course, the most striking elements. They usually have a specific function and, if we want to make a correlation with spoken or written language, some of them would correspond to the ritual function of the hymn or the invocation. Others would be more related, possibly, with the historiolae, 1 The anthropologist Clifford Geertz applied the notion of “text” to interpret phenomena of the non-linguistic world. He wondered in his works what counts as a text, and encouraged scholars in his field and others to read practices and actions as they were “texts”. Cl. Geertz, Negara: The Theatre-State in Nineteenth-Century Bali (Princeton, 1980) 13 says: “arguments, melodies, formulas, maps, and pictures are not so much idealities to be stared at but texts to be read”. A study of his ideas can be read in E.A. Clark, History, Theory, Text. Historians and the Linguistic Turn (Harvard, 2004) 145-55. 2 A very interesting analysis of non-textual elements in Graeco-Roman magic is R.  Gordon, ‘Shaping the text: Innovation and authority in Graeco-Egyptian malign magic’, in H.F.J. Horstmanshoff et al. (eds.), Kykeon. Studies in honour of H. S. Versnel (Leiden, 2002) 69-111 at 85-107.

36

raquel martÍn hernÁndez

insofar as they represent a mythical episode that is intended to be reupdated with the force of the enchantment. Others, however, have a volitional function, that is, they represent the aim of the curse, the goal that will be got by the force of the rite, which are the most popular. Figural representations tend largely to perpetuate themselves over time and an interesting continuity in their iconography and the constitutive elements can be observed. However, it is in curses, as Richard Gordon asserts, where the ritual practitioner allows himself a greater originality and novelties, especially in places located outside Egypt.3 Joseph Sanzo, on the other hand, has maintained that the practitioners elaborate complex images based on existing iconography, but which appears updated to be effective in the precise magical use for which they have been designed.4 However, such originality is difficult to completely detach from the operative mechanisms of the ritual and its traditional religious background. Within this framework, I would like to re-open the discussion about the identification and meaning of the representations featured in the so-called “Sethian curses” edited by Wünsch in 1898 (SV). My interest will be focused on the main figure of the group, the one that lead Wünsch to classify them as “Sethian”, that is, the equine-headed demon that governs the whole. 1. The so-called Sethian curses The so-called “Sethian curses” were found in a columbarium in the vicinity of Porta San Sebastiano, at the beginning of the Via Apia in Rome, back in 1850.5 The reports of the discovery are very imprecise, and the studies of these tablets are always based on the paleography and the

3 R. Gordon, ‘Getting it Right: Performative Images in Graeco-Egyptian Magical Practice’, in M. Arnhold et al. (eds.), Seeing the God. Image, Space, Performance, and Vision in the Religion of the Roman Empire (Tübingen, 2018) 101-23. “Especially in the context of ‘applied’ aggressive texts by individual practitioners outside Egypt, however, we find an increasing tendency to appeal to a private imaginaire that drew upon different iconographic resources suggested by the precise nature of the task envisaged. (…) The immediate demands of formulating an effective magical praxis encouraged such practitioners to go beyond ‘ready-mades’ and construct their own personal spirit-images”. 4 J. Sanzo, ‘The Innovative Use of Biblical Traditions for Ritual Power: The Crucifixion of Jesus on a Coptic Exorcistic Spell (Brit. Lib. Or. 6796[4], 6796) as a Test Case’, Archiv für Religionsgeschichte 16.1 (2015) 67-98. 5 The “Sethian” curses have been profusely studied since their discovery. See e.g. SV; DT nos 140-87; J. Tremel, Magica Agonistica: Fluchtafeln im antiken Sport (Hildesheim, 2004) 181-216; A. Kropp, Defixiones: ein aktuelles corpus lateinischer Fluchtafeln: dfx (Speyer, 2008) §1.1.4, and A. Kropp, Magische Sprachverwendung in vulgärlateinischen Fluchtafeln (defixiones) (Tübingen, 2008).

eulamo vs. seth

37

structure of the texts. However, the corpus is quite homogeneous, and this problem is not an insurmountable obstacle to suggest hypotheses and establish conclusions such as those that have been and will be established. There are fourty eight lamellae edited by Wünsch, although he considers the existence of a total of 56 tablets. Their state of preservation at the time of his edition was not very good, and the tablets suffered considerably in the processes of unfolding and reading. The majority of the preserved texts are written in Greek and only five feature Latin sentences. The targets of most of these curses are charioteers and horses of circus races. The area where the tablets were found is not far away from the circus of Maxentius, so its location could have been motivated for two main reasons: 1) cemeteries are traditionally prescribed to bury curses because of their evident relation with the deities of the Hereafter and the “reastless dead” that usually are invoked for such rituals, 2) a desire to affect the victims by contagion can be hypothesized. It is possible that the parade that led the charioteers to the circus passed through the Via Apia and, therefore, close to where the curses were buried. A quite interesting number of the defixiones discovered in this cemetery preserve designs of demons and victims as well as charaktēres and magical names (mainly Eulamo) and vowels written in columns. Most of the texts, and especially the texts in which I am interested, are based on a model extracted, very likely, from a formulary or a single recipe, very similar to the recipes preserved in the well-known Graeco-Egyptian magical formularies. 2. The equine-headed demon. Who is he? Since the first studies of these curses, the identification of the different figural representations drawn on the lamellae has been one of the most interesting controversies and, especially, the identification of the equineheaded demon (Fig. 1).6 The main interpretations offered so far by the scholars are: 1) the god is Typhon-Seth and the “religious” context is a lower form of Gnosticism7 2) it is a horse-like-headed demon associated with Eulamo-Osiris8 3) it is Typhon-Seth, the Egyptian god, and the religious 6

On the representation of victims in these curse tablets see Chapter II in this volume. SV 103-10 and DT xxiv, xciii and 211. In p. 198 he describes the image as “de deo Thyphone-Seth, asino capite insigni, flagellum (?) dextra, orbe sinistra tenente”. 8 K. Preisendanz, Akephalos, der kopflose Gott (Leipzig, 1927) 35-7. In p. 37 he states: “Die Ausführung des Zaubers im Ganzen, gewissermaßen die Oberaufsicht, liegt demnach offenbar in der Macht des Sonnengottes Eulamo-Osiris; die Vollführung des Fluches im Einzelnen bleibt den niederen Dämonen überlassen, dem Pferdeköpfigen — falls er nicht 7

38

raquel martÍn hernÁndez

background has nothing to do with Gnosticism9 and 4) it is an anonymous horse-headed demon especially operative in defixiones to manipulate amphitheatre or circus games.10 The interpretation of the demon as a representation of the Gnostic Seth was refuted with very solid arguments by Preisendanz which has been accepted by all the scholars. Preisendanz carefully analyzed the study by Wünsch and other authors and provided a different identification for the demon. He defends the idea that the drawing we are dealing with must be identified with a representation of a helper demon of Eulamo-Osiris. The epiclesis Eulamo appears very often on the “Sethian” curses in two main ways: 1) in a magical formation, mainly a square, and 2) in a binding formula: Εὐλάμων κατέχει Οὔσιρι.11 Eulamo is clearly connected with Osiris and, in later texts, with “solar” figures and symbols, especially on gemstones.12 However, his role is mainly active in defixiones,13 and his

doch Anubis ist — und seinen zwei Beisitzern (…)”. Regarding the iconography of the demon, he thinks that the representation of its head is closer to a horse than to a donkey, and he debates the possibility that Anubis was the model to represent this particular demon (p. 29-30). Gordon, ‘Getting it Right’, 110 identifies the demon as “horse-headed Eulamo”. 9 A. Mastrocinque, ‘Le defixiones di Porta San Sebastiano’, MHNH 5 (2005) 45-60, esp. 51. 10 G. Nemeth, ‘The horse head demon’, Sylloge Epigraphica Barcinonensis 11 (2013) 153-62. 11 The name Eulamo appears in SV nos. 1, 9, 12, 13 (14), 16-24, 26-31, 34, 36, 39 and 49. This name appears with different spellings at the end (see a survey in F. Maltomini, ‘I papyri Greci’, Studi Classici e Orientali 29 (1979) 55-122, esp. 71, n. 11 with bibliography). It has been interpreted mainly as west-Semitic adjective meaning “eternal” since the study by R.  Ganschinietz, ‘Eulamo’, Archive für Religionswissenschaft 7 (1914) 343. Preisendanz, Akephalos, 25-41, cites earlier attempts at derivation. On the other hand, the name has been interpreted as a Greek palindrome = σῶμα λῦε. See K. Preisendanz, s.v. Palindrom, in RE XVIII,3 (1949) 137. 12 Cf. SMA p. 187 and Maltomini, ‘I papiri greci’, 71, n. 11. However, a search of the word ‘Eulamo’ in The Campbell Bonner Magical Gems Database, offers 10 results. Some of them preserve texts or images of aggressive nature. CBd-158 is a thymokatochon, CBd-854 preserves a representation of an armed Ares (or Mars Ultor), CBd-1950 has a representation of the ram-headed Khnum clad in a short skirt, holding a kind of spear in his right hand and the hieroglyph ankh in his left hand in a posture very similar with that of representations of Seth, CBd-2243 preserves an Eros in a menacing posture according to the interpretation by A. Mastrocinque, Syllogue Gemmarum Gnosticarum II (Roma, 2007) 138, Ro 21; “…con cinghia dalla spalla d. attraverso il petto, la d. levata in atto di colpire…”. 13 The name Eulamo appears in the studied “Sethian” defixiones aiming to ruin chariot races and other agonistic purposes (Suppl.Mag. 53), but it also appears in defixiones for other purposes as a wrath-restrainer spell (Suppl.Mag. 57), a sexual spell (DT Addendum 304 and Suppl.Mag. 44), a defixio against a doctor (SEG 14, 615), and in magical formularies for general aggressive purposes (see e.g. PGM VII 401 and PGM IX 8), where it is associated with spells for silencing and subjecting a victim. The name appears in gemstones linked to solar deities. Particularly interesting is CBd-158 in which the sentence κατάσχες τοὺς θυμοὺς

eulamo vs. seth

39

connection with Seth is also quite significant. The name Eulamo appears alongside names of the ιωερβηθ logos (also known as Typhonicos logos) beginning with the earliest examples.14 This connection links Eulamo with both divine brothers, and the texts show that the name is understood as a magical word or name operating in this Egyptian background. Attilio Mastrocinque considers that we are facing, in fact, a representation of Seth, and it is not necessary to think of another operative demon other than the Egyptian god. He asserts: “Seth certamente è il dio raffigurato al centro della maggior parte delle lamine”,15 and interprets the image as Seth and all the religious background of the text as Egyptian. But leaving aside the nature of the name and the divinities invoked and moving on to the study of the image, which is the topic of this book, most of the arguments used by scholars not to see Seth under the representation of the equine-headed demon in the corpus of “Sethian” defixiones are based, mostly, on the difficulties of interpreting the head of the demon with the one of a donkey, the traditional representation of Seth since Ptolemaic times. The problem is basically the length of the ears: No long ears = No donkey = No Seth. I will present some thoughts to suggest the idea that the images could be understood as variations of the representation of Seth, even if they do not have the “normative” donkey ears. 1) The evident Egyptian religious background of the recurring formulae and the related figures that appear in the lamellae must be highlighted. As Richard Gordon asserts in an article of 2002 dealing with figural representations in magic: “the images are almost invariably those of divinities or powers invoked in the spell or rite”.16 This assertion must be taken into account to offer any interpretation and/or identification of the equineheaded demon. The “Osirian” elements can be easily identified in the texts and are pervasive in the images. There is no discussion among the scholars on this particular point. The expression εὐλάμων κατέχι Οὐσίρι17 acts as the operative binding formula. The name Osiris appears everywhere, in Τασοι = “Restrain the wrath of Tasos” was engraved. The gemstone must be considered a θυμοκάτοχος. 14 DT 252, 253 and SGD no 164. See R. Martín Hernández and S. Torallas Tovar, ‘A Magical Spell on an Ostracon at the Abbey of Montserrat’, ZPE 189 (2014) 175-84, and R. Martín Hernández, ‘More than a Logos. The Ιωερβηθ Logos in Context’, in C. Sánchez Natalías (ed.), Litterae Magicae. Studies in Honour of Roger S. O. Tomlin (Zaragoza, 2019) 187-209. 15 Mastrocinque, ‘Le defixiones di Porta San Sebastiano’, 51. 16 Gordon, ‘Shaping the text’, 97. 17 The text could be read as κατέχει, 3rd person of the indicative, meaning “Eulamo restrains Osiris”, οr as κατεχῃ, 3rd person of the subjunctive, meaning “May Eulamo restrain Osiris”.

40

raquel martÍn hernÁndez

different formulations, and it is invoked with other traditional names as Apis and Mnevi.18 The appearance of these names and the Osirian myth in aggressive spells are not without parallels in magical handbooks. Mnevi, the Hellenized version of the Egyptian Mr-wr, the Heliopolitan bull, appears in two texts of aggressive nature: PGM VII 44, a restraining procedure whose title says: “A restraining [rite] for anything, works even on chariots”, and PGM XIXa 6, a long litany of voces magicae to call up a daimon from the dead. To these parallels from the Graeco-egyptian formularies we should add a very interesting defixio to curse the blue team from Oxyrhynchus dated to the fourth-fifth century CE. This is the only agonistic spell preserved on papyrus so far.19 The context and the invoked divinities are Egyptian as well as Jewish, which is a very common mixture in Egyptian magic from the fourth cent. CE onwards. The defixio is addressed to a nekydaimon and the gods and divine powers invoked are Horus and Osiris (Ousirapis Ousor Mnevi), the “God of Gods”, and the archangels Gabriel, Raphael, Michael, Bouel. Osiris appears invoked in other aggresive spells mainly in relation with Seth. In PGM VII 643-51 the name Osiris appears as part of a cup spell formula as well as voces of the ιωερβηθ logos, a formula associated to Typhon-Seth that will be studied below. The same is true for PGM XII 36575, a procedure to cause separation in which, along with the ιωερβηθ logos, a reference to the enemity of Seth against Osiris is mentioned to produce a persuasive analogy.20 Regarding the images of the “Sethian” curses, a common agreement exists to identify the box-like square with a head on the top as the representation of Osiris in his sarcophagus.21 The mummy that appears in some tablets22 has been interpreted as the target.23 However, being the victim represented as a mummified corpse, it would not be illogical to think 18 For a revision of the Egyptian religious background of the tablets see Mastrocinque, ‘Le defixiones di Porta San Sebastiano’, 49-51. 19 See, H. Amirav et al., ‘5205. Spell for the Chariot Race’, The Oxyrhynchus Papyri LXXIX (London, 2014) 149-54. 20 Other references: PGM XXXVI 134-60, PGM LXXVIII 1-14, and PGM CXXVIa 1-21. 21 See Gordon, ‘Seeing the God’, 110. 22 SV 16, 17, 43 and probably 34. 23 The identification of the victim with the mummy is clear in other tablets that feature the name of the victim close to the image, e.g. Porcellus’ defixio (Bologna 4). See C. Sánchez Natalías, ‘The Bologna Defixio(nes) Revisited’, ZPE 179 (2011) 201-17 and Chapter II in this volume. In the case of the ‘Sethian’ curses the name of the victim does not appear close to the mummified corpses represented. On mummies as representation of the victim see Chapter II in this volume.

eulamo vs. seth

41

of a relationship of the victim and the mummy per excellence, Osiris. To support this statement, it must be added that all representations of the mummified body are accompanied by two snakes that constrain it and whose heads appear next to the human head with their mouth open.24 It is logical to think that the images that appear in the defixiones might be reinterpretations of an already known or familiar representation updated to be effective in the precise magical use for which they have been designed. The image of Osiris Chronocrator, and specially the image of Osiris found in the sanctuary of the Janiculum hill in Rome,25 would constitute a wellknown benchmark for comparison. That said, it does not seem too risky to understand the images of these tablets in a close relation to the imaginary of Egyptian mythology. The representation of the victim as a mummified body ritually accomplishes two different missions: 1) it identifies the target with Osiris defeated by Seth (a kind of graphic historiola) and 2) it shows the victim in the most immobilizing possible way, not only rolled up by linen, and inmobilized by serpents, but dead.26 So the representation of the target in this way is metaphorically doubly cursed. If the images of the defixiones must be understood by means of Egyptian mythology and Osiris seems to be quite well represented on them, both in the text and the drawings, the equine-headed demon should be associated to this divinity. Two are the main gods related with Osiris and his mummy that are the perfect candidates: Seth and Anubis, gods which their iconography sometimes fluctuates dificulting the distintion between them especially in Roman Times.27 24 The poppet from the Anna Perenna sanctuary must be added to these representations. It represents a man attacked by a snake which is coiling itself around his body. According to the interpretation of Natalías, “it seems likely that this scenario is a plastic (threedimensional) version of some images depicted on some of the roughly contemporary Roman tablets found in a ‘columbarium’ (or mausoleum) outside the Porta S. Sebastiano”, cf. C.  Sánchez Natalías, ‘Magical Poppets in the Western Roman Empire: a Case Study from the Fountain of Anna Perenna’, in T. Minniyakhmetova and K. Velkoborská (eds.), The Ritual Year, 10: Magic in Rituals and Rituals in Magic (Innsbruck, 2015) 194-202. 25 On this image and the sanctuary see C.J. Goddard, ‘Nuove osservazioni sul Santuario cosidetto ‘Siriaco’ al Gianicolo’, in B. Palma Venetucci (ed.), Culti Orientali tra Scavo e Collezionismo (Rome, 2008) 165-73 and Chapter II in this volume. Another interesting image for comparison is a lamp from Antioquia studied by E. Laflı et al., ‘A New Osiriform Lamp from Antioch in the Hatay Archaeological Museum’, Greek, Roman, and Byzantine Studies 52 (2012) 421-39. 26 On the representations of tied victims in aggressive spells, see Chapter II in this volume. 27 The problems for distinguishing Seth and Anubis in magical gemstones are studied by N. West, ‘Gods on small things: Egyptian monumental iconography on late antique magical

42

raquel martÍn hernÁndez

2) The second argument relates to two tablets found in the same site in which the relation with Seth is more evident. The first one is SV 25 (DT 163). In this tablet Seth-Typhon is clearly invoked several times. This tablet was edited by Audollent along with SV 24, both addressed to the same victim. In fact, Wünsch says that they were made from the same piece of lead: “Aus demselben Stücke ist nr. 25 geschnitten, der rechte Rand von 24 und der linke von 25 passen genau aneinander”. SV 24 follows the model of other “Sethian” defixiones and a representation of the equine-headed demon appears. Then, if both tablets are connected, as Audollent says, the connection between Seth and the image is much clearer than in the other items. The second tablet to consider is SV 12 (DT 151, Fig. 2). Even if the defixio is not very well preserved, it is clear that it belongs to the same group of defixiones. On the one hand, the invocation formula εὐλάμων κατέχι Οὐσίρι and Osiris’ associated name Mnevi can be read in the preserved text. On the other hand, the equine-headed demon, and the so-called paredroi (even the schematic snake) are represented. However, the representation of the demon is slightly different. Actually, it is very similar to the representation of Seth preserved in a papyrus of the Palau Ribes collection in Barcelona (Fig. 3).28 A donkey-headed human being, with short skirt, and holding an arch and an arrow. In the papyrus, the god is clearly identified by his name in the acrostic. Finally, another similar representation, this one less rough, can be added: SV 1 (DT 140, see Fig. 4), one of the “Sethian” curses written in Latin. The object that is held by the demon can be interpreted as a bow, like in the other representations. 3) My third argument concerns the representation of the god Seth in other aggressive spells, especially activated texts containing the ιωερβηθ logos.29 Spells using this logos are strongly linked to the god Seth and are operative mainly for aggressive purposes.30 It is very common to find gems and the Greek and Demotic magical papyri’, Pallas 86 (2011) 135-66 at §3 ‘The god Seth in Late Antique magic’. The article is accessible on-line: https://journals.openedition. org/pallas/2130 (12/03/2021). 28 On this papyrus see G. Michailides, ‘Papyrus contenant un dessin du dieu Seth à tête d`âne’, Aegyptus 32 (1952) 45-53 (ed. pr.); S. Bartina, ‘Set o el horrendo y gruñidor jabalí verrugoso’, Studia Papyrologica 6 (1967) 109-21, and Suppl.Mag. 69. See http://dvctvs.upf. edu/catalogue/p-palaurib-inv-3/ (27/11/2021). 29 Preisendanz considered the absence of the ιωερβηθ logos in these lamellae as evidence not to identified Seth with the equine-headed demon. See Preisendanz, Akephalos, 27. 30 On this logos see Martín Hernández and Torallas Tovar, ‘A Magical Spell on an Ostracon at the Abbey of Montserrat’ and Martín Hernández, ‘More than a logos’.

eulamo vs. seth

43

representations of Typhon-Seth in the activated texts containing this formula. The first attestation of this iconography in Graeco-Egyptian magical texts containing the logos appears in PDM xii (= PGM XII 449-52, Fig. 5) dated to the second-third century CE.31 The identification of this image is clear both because the whole text is an invocation to Seth containing the ιωερβηθ logos and, mainly, because his name is actually written on his chest. Of course the length of the ears is a useful input to identify the head as the one of a donkey.32 However, it is not always easy to identify the god drawn in procedures from formularies and in activated text containing the ιωερβηθ logos by attending only to the shape of the head and, in particular, the length of his ears. Some examples to compare are PGM XLVI (Fig. 6), P.Duke inv. 230 = SB 26.16650, P.Oxy. 68.4673 (Fig. 7) and Petrie, Amulets 135g (Fig. 8).33 In all of these examples there is no doubts among the scholars to identify the divinity with Seth based on the context and the aim of the magical texts, even if the length of the ears is not according to donkey standards. On the other hand, the identification of the image with Seth is maintained for other features: the short skirt, the bare chest, sometimes with drawings on it, the standing position with his head in profile, chest from the front, and objects, mainly weapons, in the two hands. An intriguing case for comparison, however, is the famous representation on PGM III integrated in a spell containing the ιωερβηθ logos (Fig. 9). This representation is especially interesting for my argument. In fact, one of the main purposes of the ritual, as it is said in the title, is to restrain charioteers in races.34 A quick glance at the drawing makes it clear that the head of the main demon is closer to the one of a horse, especially because of the ears and the long mane that has been drawn. The image wears a tunic and a cloak. His outfit is closer to representations of Hermanubis, being Seth usually represented with a short skirt and showing a bare-chested, as it has been already

31 The similarity between this image and the representation of Seth in the Sethian defixiones was already pointed out in SV 102. 32 West, ‘Gods on small things’ notices a problem with the identification of a gemstone from the British Museum (BM 47 in; CBd-752). The drawing is very close to the one featured in PGM XII 449-52 but the god in the gemstone was identified as Anubis by Michel, Die Magische Gemmen. West, ‘Gods on small things’ highlights the problem of identifying Seth and Anubis in some representations. 33 For a more detailed comment on these pieces see Martín Hernández, ‘More than a logos’, 200-01. 34 PGM III 162. “This is the ritual of the cat, [suitable] for every ritual purpose: A charm to restrain charioteers in a race, a charm for sending dreams, a binding love charm, and a charm to cause separation and enmity”, translation by J. M. Dillon in GMPT 22.

44

raquel martÍn hernÁndez

pointed out. Regarding the weapons, he holds a whip in the right hand and a stick on the left in a relaxed posture. The whip appears in other representations of Seth, even in the “sethian” defixiones we are discussing. The object in the left hand, however, cannot be identified with certainty because the papyrus is broken, and the final part lacks. Therefore, we cannot be sure if it is a spear or a caduceus. Consequently, the identification is difficult. However, the invocations of the text in the ritual35 and the associated logos should tip the balance in favor of identifying him with Seth and remind us of the problems for distinguishing Seth and Anubis in Roman Times by terms of iconography (n. 32). But what intrigues me the most are the other two figures that accompany the equine-headed demon. They are two human beings, standing and naked. The texts identify them as “skeleton of the left” and, very likely, “skeleton of the right” (σκελετὸς δεξιός / ]τερος). They seem to wear bands crossed on their chests, and their sex may be represented (more visible in the image of the right). Both of them hold a whip in the right hand. The identification of the drawing on the top is difficult to identify and I could not find a satisfactory explanation. These two human beings (or skeletons) have not been identified – as far as I am aware. Maybe it is a little bit risky, but I think a comparison between these figures and the paredroi, the companions of the equine-headed demon mentioned in DT 155 and represented in different designs in the defixiones of Porta San Sebastiano, could help us to their identification and, again, it would be an argument to bring the whole closer to its Egyptian backgrounds. We should highlight that the human beings are identified, as it has been said, as the skeleton of the right and the one on the left, and the paredroi represented in the Roman curses use to be located to the left and right side of the equine-headed demon. 4) Finally, a brief comment on the horse head and the purpose of the curses. It has been argued, with very good arguments, that the shape of the demon is especially operative for curses against horses and charioteers. However, although I do not think it is an invalid argument (in fact, I think

35 The invocation above the representation is addressed to Seth-Typhon, Sabaoth, Adonai, Iaeo, Abrasax, Damnameneu. The relation of the ιωερβηθ logos with the name of Hebrew deities (especially Iao) and other associated powerful names and logoi is common. See formularies: PDM xiv 675-94 (Ereschigal and Aberamenthoou logos), PGM VII 643-51 (Iao, Sabaoth Adonai, Abrasax, Ablathanalba and Achramachamari), PGM III 70-97 (Iao, Sabaoth Adonai, Abrasax, Iaeo logos, ablathanalba palindrome), PGM IV 154-285 (Aberamenthoou logos and other palindromes), PGM IV 3255-74 (Iao, Sabaoth, Abrasax and Aberamenthou logos), PGM XXXVI 1-34, 69-101 and PGM CXXVIa (Aberamenthoou logos).

eulamo vs. seth

45

it could be conscious), it should be noted that not all the tablets in which this representation appears are made against charioteers and, on the other hand, as we have seen in the examples displayed in this article, the “ears” of Seth differ on length in other spells in which its identification is not questioned. A god/demon with a horse head does not have an Egyptian reference, so we should understand it as a non-standard modification of a “ready-made” model,36 but the question is still there: who was the model? 3. In a way of conclusion Given the arguments put forward in this article, it is more plausible to understand the demonic figure of the curses of Porta San Sebastiano as Seth, given the “Osirian” context of the rest of images, the texts and, above all, because of the iconographic varieties that Seth suffers in other similar execration texts dated more or less to the same period. While it is true that the head may have been modified either in the interest of establishing a ritual analogy between the image and the aim of the curses, it is also true that it is something that we cannot affirm with certainty and that, in any case, it does not discard the identification of the equine-headed demon with Seth. We should not think that the distinctions between the operative gods in enchantments were crystal clear in Roman Times, not even for professionals. We have already warned of the iconographic problems between Anubis and Seth in magical gemstones. However, it is clear that these curse texts tend to accumulate iconographic elements that reinforce the aggressive action, congruent with the intention of the text and the act of ritual power. A constraining god would be ritually more operative in magical terms than another divinity with more protective traits, so Seth would be a perfect candidate. Finally, a question to be solved… Are we sure the users (or even the practitioners) were aware about who were the gods represented in defixiones like these ones, copied out probably from formularies of Egyptian origin? Even if the model was clearly representing Seth (or another god) the practitioner could have modified the drawings to adapt them with other motivations, feelings and beliefs, but mantaining the “Egyptian flavour” of the theriomorphic operative demons.37 Again, we see the problem of innovation vs. authority in aggressive magic. 36 37

I am using the terminology created by Gordon, ‘Seeing the God’. See Gordon, ‘Seeing the God’.

46

raquel martÍn hernÁndez

Figure 1. SV 16, apud R. Wünsch, Sethianische Verfluchungstafeln aus Rom, 1898 modified.

Fig. 2. SV 12, apud R. Wünsch, Sethianische Verfluchungstafeln aus Rom, 1898 modified.

eulamo vs. seth

Fig. 3. P.PalauRib.Lit. I 39 ©Palau Ribes Collection. Used with persmission.

Fig. 4. SV 1. apud R. Wünsch, Sethianische Verfluchungstafeln aus Rom, 1898 modified.

47

48

raquel martÍn hernÁndez

Fig. 5. Image of Seth. PGM XII 449-52. Image created by R. Martín Hernández.

Fig. 6. Nationalbibliothek G 332 + Vienna, Nationalbibliothek P.Rainer 12 = PGM XLVI © Österreichischen Nationalbibliothek, Papyrussammlung.

eulamo vs. seth

Fig. 7. P.Oxy. 68.4673. Image courtesy of the Egypt Exploration Society and the University of Oxford Imaging Papyri Project.

Fig. 8 Amulets 135g.

49

50

raquel martÍn hernÁndez

Fig. 9. PGM III. Image created by R. Martín Hernández.

IV.

BAITMO AND LAMASHTU György Németh1

In 1892, G. Hannezo and captain L. Choppard found a series of curse tablets in the cemetery of ancient Hadrumetum2 that all depicted a peculiar demon: it is standing on a boat with a chandelier in its left hand, while its right hand holds an eared hemispheric vessel identified either as a hydria in the catalogue of the Museum Alaoui or as a simple vessel by R. Gordon.3 On every item, an inscription of two words on the demon’s chest indicates the figure’s name, but there are no two identical name forms in the series of tablets. The form used in the title of this paper can be read on another tablet found by G. Hannezo in 1910, and this is declared to be the correct form by Gordon without further explanation.4 In other cases, the initial letter of the name is not ‘B’ but ‘A’ or (in one instance) ‘D’. In two tablets the drawing of the figure’s chest is lost, thus we cannot tell what name the practitioner attributed to the demon. Therefore we can find five different name forms in seven curse tablets. Baitmo Arbitto5 Antmo Araitto6

1 ELTE University, ORCID 000-0001-878-8102. This study forms part of OTKA [Hungarian Scientific Research Fund] programme no. K 134319 (Corpus of the Curse Tables of Clermont-Ferrand), and the Zaragoza Project titled Procesos de aculturación religiosa en el mundo antiguo y en la América colonial: un análisis comparativo de la retórica y la construcción de la alteridad. 2015- (HAR2014-57067-P). 2 Today: Sousse, Tunisia. 3 R.-M. de Coudray de Blanchère and P. Gauckler, Catalogue du Musée Alaoui (Paris 1897) 127; R. Gordon, ‘Competence and ‘Felicity Conditions’ in two Sets of North African Curse-Tablets (DTAud nos. 275-85; 286-98)’, MHNH 5 (2005) 61-86 at 69. Perhaps it is not unjustified to identify the vessel as a funerary urn, cf. J. Tremel, Magica agonistica. Fluchtafeln im antiken Sport (Hildesheim, 2004) 132. 4 Gordon, ‘Competence’, 69. 5 Tremel, Magica agonistica, 149; G. Németh, Supplementum Audollentianum (Zaragoza, Budapest and Debrecen, 2013) no 230. 6 DT 286.

52

györgy németh Antmo Araito7 Daitmo Arditto8 Aitmo Arpitto9 Unknown10

Although this demon representation is unique in the corpus of curse tablets and magical papyri, Audollent did not publish the drawings of more than two tablets (the drawings were produced by Antoine Héron de Villefosse),11 therefore it is difficult to establish the similarities or possible differences in the demon’s depictions. In 2009 I found in Clermon-Ferrand all of Audollent’s drawings he prepared (though never published) during the composition of his monograph (1904).12 The photo he published in 1910 is presented here in my own drawing.13

DT 286. Image created by A. Villefosse 7

DT 396; Németh, Supplementum, no 286; DT 402 = Németh, Supplementum no 289. The reading is uncertain, and it is ommitted by Audollent. The drawing provides little help to determine what letters may have been written here. Gordon, ‘Competence’, 69 reads Daitmo Arditto without giving further references. Németh, Supplementum, no 288. Coudray La Blanchère, Catalogue, 127 identifies the figure as “Le Dieu Set”. 9 DT 404; Németh, Supplementum, no 291. Coudray La Blanchère, Catalogue, Tab. 31 provides a photo image of the demon confirming Audollent’s reading. 10 The drawings are fragmentary, and the chest of the demon is not visible. Németh, Supplementum, no 287; 290. 11 DT 397; 399. A. Héron de Villefosse, ‘Tabellae’, Bulletin de la Société nationale des antiquaires de France (1901) 326-34. 12 For the details of finding and publishing the drawings, see Németh, Supplementum, 15–16. For the drawings, see Németh, Supplementum, 200-06. 13 Németh, Supplementum, 230. 8

baitmo and lamashtu

53

DT 287. Image created by A. Villefosse.

14

DT 288. Drawing by A. Audollent.14

14 The paper has slightly darkened with the years, thus the pencil drawings are difficult to see.

54

györgy németh

DT 289. Drawing by A. Audollent.

baitmo and lamashtu

DT 290. Drawing by A. Audollent.

55

56

györgy németh

DT 291. Drawing by A. Audollent.

baitmo and lamashtu

57

Németh, Supplementum, 230. Image created by G. Németh

The drawings were clearly prepared by the same sorcerer to gain influence over the Hadrumetum chariot races.15 The demon standing on the boat is either simply called a demon (Adiuro te demon),16 or addressed indeterminately (Adiuro te demon cuiqunque es),17 or the sorcerer referred to a powerful deity superior to the demon, who commands the seas and the air, and who deprived the nekydaimōn of its life.18 Therefore the demon is linked to the sea not only by the boat on which it always stands. Richard Gordon investigated the interpretations of the demon standing on the boat and came to the conclusion that the hull (without steering, oars, and sails) rushes forward unstopably like inexorable fate.19 As for the 15

Tremel, Magica agonistica, 36-137. DT 286, 290 and 291. 17 DT 291. 18 ạḍịụ[ro] te per eum qui te re[solvi]ṭ vite temporibus de[um pelagicum] aerium aḷtis[simum], see DT 290. The reconstruction of the text is certain, because the same words are read in the ‘Ocuria anoχ’ series also found in Hadrumetum, see Tremel, Magica agonistica, 138-41. 19 Gordon, ‘Competence’, 72: “a condensed, evocative image of the fate he is wishing upon the victims of his curses, caught irresistibly in the power of a force they cannot comprehend, as helpless as one is to halt a fastmoving ship.” 16

58

györgy németh

origins of the iconography, he referred on the one hand to the triumphal coinage of the Roman Republic and the early Empire, on the other hand to the so-called Isis Pelagia representations. However, he also emphasized that there is a decisive difference between Isis Pelagia representations and the ship of Baitmo: Isis Pelagia is holding a sail in her hands, so her ship is not out of control. Isis Pelagia20 (also known as Isis Pharia or Isis Euploia), developed in Alexandria, was considered the divinity who invented sailing and navigation. The LIMC entry lists all Alexandrian coins with her image from the age of Domitian. Thus the Hadrumetum magician could easily come across such a coin, since the North-African town of Hadrumetum had close relationships with Alexandria. Nevertheless, the presence of the sail makes it unlikely that the iconography of Baitmo standing on a ship should be traced back to the Isis Pelagia representation. There is a representation of Isis standing on a boat without a sail, a tiller, or an oar, but this is found on a Pompeii fresco that could hardly have had any influence on the Hadrumetum magician, since the wall-painting had long been covered by volcanic ashes by the third cent. CE.21 There is also a seal impression from third cent. BCE Seleukeia depicting Isis on a ship similar to Baitmo’s (i.e. without sail or oar), however, a wide span of space and time separates this image from the demon representation found in Hadrumetum.22 Magic gems, prepared mostly in Egypt, were probably well-known to contemporary practitioners.23 These gems present a number of deities standing on a boat. Standing representations of Osiris depict the god as a mummy in a frontal position, thus we can rule him out as a possible model of Baitmo.24 Harpokratēs and Hōrus stand or sit mostly on a lotus flower instead of a boat.25 A more appropriate prefiguration can be the Anguipes with human legs, standing on a ship, facing right.26 Even more so, because Isis or Hōrus are not as easily connected with cursing as the Anguipes. Obviously, the most probable model could be a markedly harmful demon or deity standing in an ark, facing right. 20

Tran Tam Tinh, ‘Isis’, LIMC V (Zürich, 1990) 782. Tran Tam Tinh, ‘Isis’, 768. There is also another boat in the scene with a swallow in a box representing the soul of the dead Osiris. Naples, Museo Nazionale inv. no 8929. 22 A. Invernizzi, ‘Isis’, LIMC Supplementum I (Düsseldorf, 2009) 299. “Navigium Isidis. La barca diretta a d. trasporta due esili figurine con lunga veste, manto e corona sul capo, che tendono avanti le braccia”. 23 C. Bonner, Studies in Magical Amulets. Chiefly Graeco-Egyptian (Ann Arbor, 1950) (SMA). 24 DMG 312, 376. 25 DMG 279, 283, 293 and 340. 26 DMG 248. 21

baitmo and lamashtu

59

There is a Mesopotamian goddess called Dimme in Sumerian, Lamashtu in Akkadian.27 This deity abducted children from their parents to destroy them. Lamashtu was also capable of causing various diseases, so she was meant to be averted by using amulets that spread from Mesopotamia to Asia Minor, from Egypt to South Italy over the centuries.28 One of the most elaborate items is a Neo-Assyrian bronze amulet now kept in the Louvre (Paris), which depicts the goddess standing on a ship, facing right, while the boat (without sail and oar) is rushing towards a bog. In each hand the figure holds two coiled snakes. Lamashtu is fleeing. She is pursued by a deity called Pazuzu, who wants to stop her from making anyone ill. Above the demon, the representation shows the infected person lying in his bedchamber, then a field with apotropaic figures, and on the top of the image we can see divine symbols.29 1. The bronze amulet kept in Paris An incantation against Lamashtu sums up her characteristics features:30 “Great is the daughter of Heaven who tortures babies Her hand is a net, her embrace is death She is cruel, raging, angry, predatory A runner, a thief is the daughter of Heaven She touches the bellies of women in labor She pulls out the pregnant women’s baby The daughter of Heaven is one of the Gods, her brothers With no child of her own. Her head is a lion’s head Her body is a donkey’s body She roars like a lion She constantly howls like a demon-dog.”31

27 F. Wiggermann, ‘Lamaštu, Daughter of Anu. A Profile’, in M. Stol (ed.), Birth in Babylonia (Winona Lake, 2000) 217-49. 28 E. Götting, ‘Exportschlager Dämon? Zur Verbreitng altorientalischer ‘Amulette’’, in J. Göbel and T. Zech (eds.), Exportschlager: Kultureller Austausch, wirtschaftliche Beziehungen und transnationale Entwicklungen in der antiken Welt (Berlin, 2011) 437-56. A. Bácskay, ‘The Lamaštu amulet BM 120388’, Cuneiform Digital Library (2016), online https://cdli-dev. mpiwg-berlin.mpg.de/articles/cdln/361. 29 Wiggermann, ‘Lamaštu, Daughter of Anu’, 244. 30 W. Farber, Lamaštu: An Edition of the Canonical Series of Lamaštu Incantations and Rituals and Related Texts from the Second and First Millennia b.C. (Winona Lake, 2014). 31 T. Bane, Encyclopedia of Demons in World Religions and Cultures (Jefferson NC and London, 2012) 199.

60

györgy németh

The question is, of course, how well the Greeks and the Egyptians could have known Lamashtu. Walter Burkert claims they knew her rather well, and they even adopted her in several forms.32 Lamashtu lies behind the Greek child-eating demon Lamia, and the kidnapping demon Gellō (perpetuated by Sappho), who is also of eastern origin, as her name reveals: Gallu is an Akkadian name for an evil spirit.33 Burkert comes to the conclusion: “There is no gap between Babylon and Greece.”34 The above mentioned scholars of Assyriology suspected that the name Lamia is derived from Lamashtu. This suggestion is confirmed by the scholium on The Peace of Aristophanes (line 758): “They say that Lamia is the daughter of Bēlos and Libyē”.35 Thus we can see how long the memory of Lamia’s eastern origin survived in the Greek world. Perhaps it is not a coincidence that the third cent. BCE representation of Isis Pelagia standing on a ship without sail or oar was discovered in Seleukeia, the area where most Lamashtu-amulets come from.36 It is true, however, that Greek and Latin magical texts contain no reference to the name of Lamashtu. Nevertheless, a sorcerer may borrow an iconographical type without knowing the original name of the figure in it. A good example is a second cent. CE mosaic doorstep from Hadrumetum with the image of ying and yang.37 The artist who prepared the mosaic must have known the pattern itself (perhaps from an amulet that had travelled a very long road), but probably not its name. There is, however, a GraecoEgyptian love spell from the third-fourth cent. CE (i.e. roughly contemporary with the Baitmo representation) that mentions a certain Lamach, who turns up, interestingly, in the company of another Mesopotamian goddess, Ereshkigal.38 ζαμενηθ ζαταρατα κυφαρταννα αννε Ερεϲχιγαλ επλανγαρβωθιθοηαλιθαθθα διαδ[αξ] ϲωθαρα ϲιερϲειρ ϲυμμυθα φρεννωβαθα ωαη . . λειχοιρετακεϲτρευ ιωαξειαρνευ κορυνευκν[υορο] αλιϲ ϲωθεωθ δωδεκακιϲτη ἀκρουροβόρε ϲωκ . . ρουμε ϲουχιαρ ανοχ ανοχ βριττανδρα ϲκυλμ 32 W. Burkert, The Orientalizing Revolution. Near Eastern Influence on Greek Culture in the Early Archaic Age (Cambridge MA, 1992) 82-87. 33 Sappho fr. 178 Lobel–Page: “Γελλὼ παιδοφιλωτέρα.” 34 Burkert, The Orientalizing Revolution, 83. 35 Schol. Pax. 758: λέγεται ἡ Λάμια Βήλου καὶ Λιβύης θυγάτηρ. 36 Invernizzi, ‘Isis’, 299. 37 A. Rebourg, Sousse Museum (Tunis, 1995) 40. 38 G. Németh, ‘Ereschigal – Ereškigal. Migrations of a goddess’, MHNH 10 (2010) 199-205.

baitmo and lamashtu

61

αχαλ βαθραηλ εμαβριμα χρημλα αοϲτραχιν . αμου ϲαληναϲαυ τατ χολαϲ ϲωρϲανγαρ μαδου[ρε] βοαϲαραουλ ϲαρουνα ϲιϲιϲρω ζαχαρρω ιβιβι βαρβαλ ϲοβουχ Ωϲιρ ουωαι Αζηλ αβαδαωτ[. .] ιωβαδαων βερβαιϲω χιω υ υ υ φθωβαλ λαμαχ χαμαρχωθ βαϲαρα βαθαραρ νεαιπεϲχιωθ[. .]39 “zamenēth zatarata kyphartanna anne Ereschigal eplangarbōthithoēalithaththa diadax sōthara sierseir symmytha phrennōbatha ōaē[..] leichoiretaestreu iōaxeiarneu koryneuknyoro alis sōtheōth dōdekakistē, swallowing the tip of the tail, sōk roume souchiar anoch anoch brittandra skylm achal bathraēl emabrima chrēmla aostrachin amou salénasau tat cholas sōrsangar madoure boasaraoul saroucha sisisrō zacharrō ibibi barbal sobouch Ōsir ouōai Azēl abadaōt .. iōbadaōn berbaisō chiō y y y phthōbal lamach chamarchōth basara batharar neaipeschiōth”40 Ερεϲχιγαλ: Babylonian underworld goddess.41 ανοχ: the Coptic personal pronoun.42 αμου: probably Coptic ‘come’.43 βαρβαλ: probably from Aramaic bar-ba’al ‘son of Bal’.44 Ωϲιρ: Osiris.45 Αζηλ: perhaps Azaēl. αβαδαωτ: perhaps from Hebrew bdwt ‘destructions’.46 λαμαχ: perhaps from Akkadian Lamastu.

The above magical text, as we can see, contains elements from Egyptian, Babylonian, Aramaic, and Hebrew origins. In this context, the emergence of Lamashtu is perhaps not entirely unexpected. If Lamashtu really had any direct or indirect influence on the formation of Baitmo, then the figure of the demon standing on a ship must have undergone an unconscious reinterpretation, too. Lamashtu was indeed evil in the Lamashtu amulets, but the practitioner’s intention was to drive her away with the help of Pazuzu. The ship without a sail, an oar, and a tiller is moved by the will of the goddess to rush back to the morass from where it came to infect and kidnap children. Baitmo, on the other hand, is a harmful

39 40 41 42 43 44 45 46

Suppl.Mag. 42, pp. 132-53. Suppl.Mag. 42, p. 138. Suppl.Mag. 42, p. 148; Németh, ‘Ereschigal’, 199-205. Suppl.Mag. 42, p. 148. Suppl.Mag. 42, p. 148. Suppl.Mag. 42, p. 148. Suppl.Mag. 42, p. 149. Suppl.Mag. 42, pp. 148-49.

62

györgy németh

demon, the spirit of a drowned man, a nekydaimōn forced by the magician to bring harm to certain charioteers. Lamashtu’s figure may have come to North Africa via Greek intermediaries, or alternatively, it might have reached the territory directly as well. Phoenicians built the port of Hadrumetum in the ninth century BCE, when numerous Lamashtu amulets were produced.47 The relationship between Phoenicia and Hadrumetum was continuous in the Neo-Assyrian period (911–612 BCE), when most Lamashtu amulets circulated and when Phoenicia was part of the Assyrian Empire. The Carthaginians may have preserved either such an amulet or at least the image of a deity standing on a ship without sail and oar even in Hadrumetum. This image was exploited by an (otherwise not very well-educated) sorcerer in Hadrumetum when he created Baitmo’s demonic figure charging towards charioteers like insuperable destiny.

47 J. DeGrado and M. Richey, ‘Aramaic-Inscribed Lamaštu Amulet from Zincirli’, Bulletin of the American Schools of Oriental Research 377 (2017) 107-33.

V.

THE HUNGRY WOLF Attilio Mastrocinque

1. Thirsty plants We know of many characters who were supposed to be exceedingly thirsty, and their thirst was the means to perform a magic enactement often concerning some liquid substances. A well known case is that of Tantalus and Ares (Fig. 1), respectively mentioned and depicted on several hematite gems, aimed at controlling the menstrual cycle, and not only at stopping superabundant menses but also at producing menses themselves, when necessary.1 In this contribution I will speak of two other thirsty characters: the wolf and the ibis. Amulets with their images and the verb πίνω, “I drink”, or better π(ε)ινῶ,2 “I am in need”, are known. Their images depict them biting an obnoxious beast such as a crocodile or a snake.3 What is the underlying logic which inspired such puzzling magic images and acts? The mentioned amulets are mostly small bronze plates of the early Byzantine period (ca sixth cent. CE), but recently a more ancient, mid-imperial age hematite amulet has been published, which mentions the thirsty wolf (Fig. 2a and 2b).4 It was in the Lippert collection of castings but it remained 1 Cf. C.A.  Faraone, ‘Does Tantalus drink the Blood, or not?: an Enigmatic Series of Inscribed Hematite Gemstones’, in U.  Dill and C. Walde (eds.), Antike Mythen. Medien, Transformationen und Konstruktionen (Mélanges Fritz Graf), (Berlin and New York, 2009) 203-28; LIM 132. 2 H. Seyrig, ‘Invidiae medici. 1. La faim de l’ibis et la soif de Tantale’, Berytus 1 (1934) 1-5 at 1-2; cf. SMA p. 213. For the reading πίνω: A.A.  Barb, ‘Magica Varia’, Syria, 49  3.4 (1972) 343-70, esp. 357-62. 3 Wolf: J. Spier, ‘An antique magical book used for making sixth-century Byzantine Amulets?’, in V. Dasen and J.-M. Spieser (eds.), Les savoirs magiques et leur transmission de l’Antiquité à la Renaissance (Florence, 2014) 43-66 at 54-61 and 63-65. I am grateful to J. Spier for the images of the bronze amulets which are published here. Ibis: Seyrig, ‘Invidiae medici’. 4 A. Łatjar et al., ‘A Squeeze of a Magical Gem in the Dactyliotheca of Philipp Daniel Lippert on the Basis of a Copy kept in the Collection of the Nieborów Palace’, in V. Dobrowolski (ed.), Et in Arcadia ego. Studia memoriae professoris Thomae Mikocki dicata (Warsaw, 2013) 371-80.

64

attilio mastrocinque

almost unknown until its recent publication. On the obverse side one can see the lion-headed god holding three ears of wheat in the left hand and, in the right hand, the lion-headed snake Chnoubis, who was supposed to control every liquid substance.5 On the reverse, van der Vliet, Łatjar, and Płóciennik have been able to read this puzzling text: ƶƶƶ Λειμόδωρος λειμοδώρῳ τάφον ἔσπειρεν· λύκος ἤμπρευε ναὶ θυῖα κατέφαγεν.

(Chnoubis sign) Leimodoros (Famine’s gift) sowed the tomb with broomrape [leimodoron]. A wolf dragged out (the plant); it verily devoured the sap.6

Broomrape or orobanche is a parasite plant which has no chlorophyll and grows by parasitizing other plants and feeding on their sap. As broomrape was growing on a tomb, the spell probably alludes to its feeding on a corpse and on blood. The editors of this gem correctly recognized that it was a means to control hemorrhages and thus a plant and an animal fed on blood and meat make sense in an amulet for controlling the flux of blood. Theophrastus calls this plant either leimodoron (neuter)7 or haimodoron8 i.e. “a gift of blood”. The Greek tradition knows of “plants of tombs” or “plants of the dead”, such as the phlomos, also known as nekya, “the plant of the dead”,9 which was supposed to stop epistaxis.10 Another probably similar plant was the thapsia, “the plant of the tomb”, or “the deadly plant”.11 Several plants were known as a means to connect the dead with this world and also to convey the blood of the deceased from their roots up to their branches, as Pythagoras thought of the broad bean or Virgil 5

LIM 93-95. The translation is my own. The ναὶ in line 7 is clearly readable but it is possible that it was a καὶ in the original spell, because ναὶ is never used in magical texts, as Christopher Faraone pointed out to me. 7 Thphr. HP 5.15.5; on it and its lifestyle: Thphr. HP 8.8.4; Gal. De aliment. facult. 6, p. 552 Kühn. 8 Thphr. HP 8.8.5. 9 Cyranides 1.13: Νεκύα βοτάνη ἐστὶ ἡ λεγομένη φλόμος. See D. Ogden, Magic, Witchcraft, and Ghosts in the Greek and Roman Worlds: A Source Book (Oxford, 2002) 268. 10 Dsc. Eup. 1.200. τὰς δὲ ἐκ μυκτήρων αἱμορραγίας στέλλει ὀξύσχοινος… ὁμοίως φλόμος… “the great sea-rush stops hemorrhages from the nostrils… and the phlomos has the same effect”. 11 Dsc. 4.153; Gal. De comp. medicam. 12, p. 387 Kühn. 6

the hungry wolf

65

of Polydorus’ tree. Human blood instead of vegetal sap was the proof of a connection between a human being and a plant. In the world of magic arts the paeonia was also called haimagogon, “conveying blood”.12 Rooting out such plants was a dangerous act also because certain birds were ready to punish the uprooter (rhizotomos) who, in case of several dangerous plants, such as paeonia, recurred to a dog, which eradicated the dangerous plant by taking the risk on himself.13 The Carmen de herbis (a poetic work probably from the second cent. CE) connects the name of the kynospastos plant with the rising of the Dog star,14 but its name signifies “rooted out by a dog”. Dioscourides15 and Galen16 deal with the orobanche-broomrape plant by calling it kynomoron or kynomorion, which means “lethal to a dog”, and Galen put it in comparison with the lykoktonon plant, i.e. “killer of wolves”. Aelianus17 knows of a lykoktonos plant which grows near the Nile and kills the wolves which uproot it. The wolf of the spell was thirsty because wolves were commonly supposed to be so,18 as were werewolves.19 I think that another spell on a fourth century CE silver lamella from Poitiers (the ancient Limonium)20 was similarly conceived:

12 Dsc. De paeonia 3.140 (Catalogus codicum astrologorum Graecorum 11.2, ed. Zuretti, Brussels, 1934). 13 Ael. NA 14.27; cf. R. Gordon, ‘Aelian’s Peony’, Comparative Criticism 9 (1987) 59-95. 14 A. Delatte, Herbarius: recherches sur le cérémonial usité chez les anciens pour la cueillette des simples et des plantes magiques (Bruxelles, 19613) 36. The author of this work explained the name so as if it would have been possible to eradicate the plant when the Dog was rising. He was describing the plant aglaophotis whereas Ael, NA 14.27 spoke of the sea aglaophotis, also known as pankynion, whose fruits were supposed to kill sharks (thalattioi kynes). 15 Dsc. 2.142. 16 Gal. De simplicium 11, p. 834. 17 Ael. NA 9.17; (the same thing in Aristophanes, Historiae animalium epitome subjunctis Aeliani Timothei aliorumque 2.223 (S.P. Lambros, Commentaria in Aristotelem Graeca, Suppl. 1.1 [Berlin, 1885]). According to Dsc. 4.77, the lykoctonon is the aconite, otherwise known as wolfsbane. 18 Septuaginta, Prov. 28.15.1: λέων πεινῶν καὶ λύκος διψῶν. 19 Marcellus Sidetes, De Kynanthropia (W.H. Roscher, Das von der “Kynanthropie” handelnde Fragment des Marcellus von Side [Leipzig, 1896] 79-80); Anonymi Medici, De lycanthropia 1, Ideler (ed.), Physici et medici Graeci minores 2 [Berlin, 1842]); Oribasius. De lycanthropia 8.7.1 (J. Raeder, Oribasii synopsis ad Eustathium et libri ad Eunapium, Corpus medicorum Graecorum 6.3 [Leipzig, 1926]). Cf. G. Piccaluga, Lykaon. Un tema mitico (Roma, 1968) 58, 63 and 145-46. 20 GMA no 8. Kotansky (pp. 33-34) thinks that the spell borrowed from a botanical recipe.

66

attilio mastrocinque

Bis gontaurion analabis bis gontaurio suce analabis bis gontaurios catalases vim c anima vim s paternam. asta Margiasse. tutate Iustina quem peperit Sarra.

(Twice). You should pick the centaury plant. (Twice). O juice of the centaury plant! You should pick the centaury plant. (Twice). You will lay hold of strengh, life, paternal strengh. Assist, o Magiarssus! Protect (pl.) Justina whom Sarra bore.21

Also the centaury plant was supposed to have magical and medical properties, its sap was thought of as similar to blood, and its uprooting was allegedly dangerous because an aggressive bird protected it.22 2. Thirsty and hungry animals Some diseases were supposedly caused by demons who either craved or loathed liquids, as the spell on a magical gem in Athens23 proves: φύγε, δαίμων Ὑδροφόβα, ἀπὸ τοῦ φοροῦντος τοῦτο τὸ φυλακτήρι(ο)ν: “flee, hydrophobic demon, from him who wears this phylactery”. The spell of the wolf was used in the Imperial Age, as it is the case with the mentioned gem, but it recurs mostly on Byzantine amulets, which have been recently gathered and studied by Jeffrey Spier.24 Among these amulets I would like to mention the following ones because they offer a clue to understand the magic role of this animal by mentioning it or some parts of the above quoted spell: 1) British Museum inv. no. 56324. C. Bonner, ‘Amulets Chiefly in the British Museum’, Hesperia 20 (1951) 334-35, no 51; SMA pp. 216-217 (505, D 315 Newell); A. Barb, ‘Magica varia’, Syria 49 3-4 (1972) 343-30, esp. 35052; J. Spier, An antique magical Book, 54-58, fig. 8 (Fig. 3a and 3b): (side a) Λιμός σε ἔσπιρεν, ἀὴρ ἐθέρισεν, φλέψ ε κατέφαγεν. τί ὡς λύκος μασᾶσε; τί ὡς κορκόδυλλος καταπίννις; τί ὡς λέων ορώχις; τί ὡς ταῦρος κερατίζις; τί ὡς δράκων 21

Hunger sowed you, air harvested you, vein devoured you. Why do you munch like a wolf, why do you devour like a crocodile, why do you bite (or roar?) like a lion, why do you

Reading and translation by Kotansky. Thphr. HP 9.8.6 and Plin. HN 25.69. 23 A.  Delatte, ‘Études sur la magie grecque, III-IV’, Le Musée Belge 18 (1914) 5-96 and 70-75. 24 Spier, An antique magical Book. 22

the hungry wolf εἱλίσσι; τί ὡς παρᾶος κυμᾶσε; (side b) ἵππος, μῦλος, ἶβις, ἐθυῖα κόλε ἀνδρός, στρουθοκάμηλος, Ἀπόλλω .ιοσοτοιδηυς δουλιασουσατερ παυλινας παυιο ωσωεη…ηο.ηι

67

gore like a bull, why do you coil like a serpent; why do you lie down like a tame creature? Horse, mule, ibis, phallus, ostrich, Apollo… Thy slave…Paulina… (reading by Bonner)

I propose a slightly different reading, especially thanks to the edition by Jeffrey Spier,25 who recognized the name of Apollonius (of Tyana): (side a) Λιμός σε ἔσπιρεν, ⟦α⟧ἀὴρ ἐθέρισεν, φλέψ ε κατέφαγεν. τί ὡς λύκος μασᾶσε; τί ὡς κορκόδυλλος καταπίννις; τί ὡς λέων βρώχις; τί ὡς ταῦρος κερατίζις; τί ὡς δράκων εἱλίσσι; τί ὡς παρᾶος κυμᾶσε; (side b) ἵππος, Nῦλος εἶβις (= Νιλόσειβις), ἐθυῖα κόλε ἀνδρός, στρουθοκάμηλος Ἀπoλλώνιος ὁ Τοιανεύς. δουλία σοῦ σατερ Παυλίνας παυιο ωσωεη…ηο.ηι

Hunger sowed you, air harvested you, vein devoured you. Why do you munch like a wolf, why do you devour like a crocodile, why do you roar like a lion, why do you gore like a bull coil like a serpent; why do you lie down like a tame creature? Horse (=vulva?), Nilotic ibis, phallus of a man, ostrich, Apollonius of Tyana. Of your slave… Paulina…

The beginning is similar to the spell on the Lippert gem, whereas the following questions (Why do you munch like a wolf…?) are like these on many amulets for the womb.26 In order to understand why the vein devoured something, it will be useful to quote another Byzantine amulet27 bearing this spell: στόμαχε, ἀντιστόμαχε, ὡς αἷμα ἔφαγε ὡς αἷμα ἐπίωκες οὕτω καταδῶ σε “Stomach, mouth of the stomach, as you have eaten blood, as you have drunk blood, so I bind you”.

Instead of the disease, a part of the human body is mentioned as a cause of the disease itself. The vein and the stomach were thus greedy for blood and consequently deprived the body of it.

25

Spier, An antique magical Book, 59. A.A. Barb, ‘Diva Matrix. A Faked Gnostic Intaglio in the Possession of P. P. Rubens and the Iconology of a Symbol’, Journal of the Warburg and Courtauld Institutes 16 (1953) 193-238. The recurring spell has, instead, a final sentence saying: “why do not you lie down like a lamb?” 27 R. Mouterde, ‘Objets magiques. Recueil S. Ayvaz’, Mélanges de l’Université St. Joseph 25 (1942-1943) 105-28, esp. 124, pl. 9; SMA p. 217; Spier, An antique magical Book 58, fig. 9. 26

68

attilio mastrocinque

Hippos was a metapher for pudenda muliebra as noticed by Barb.28 Instead of reading the Latin word mulus, written with Greek letters, as it appears below in the specimen no 4 (the Greek name of the half-ass is ἡμίονος, not μῦλος), I follow P. Perdrizet29 in reading Νειλόσειβις. In a fragmentary amulet30 the word is written as Νοῖλος, i.e. Νῖλος. κόλε was interpreted by Bonner as κωλῆ, “thighbone with its flesh”, but it signifies “phallus” in Aristophanes’ Nubes 989, 1019. I do not rule out a vocative case of κόλον (neuter) and εὐθεῖα (feminine) which signifies “straight”, i.e. colon rectum. Phallus, vulva, cacator, and ibis recur on many amulets against the evil eye, where they are attacking it. 2) SMA pp. 216-217, no 315: λύκος πινῶν ⟦ων⟧ ἐβώσκετο πίνω ὕδωρ διψῶ ἄρτον φάω i.e. λύκος πινῶν ⟦ων⟧ ἐβόσκετο, πίνω ὕδωρ διψῶ ἄρτον φάω. “A wolf, hungry, was fed. I drink water, I am thirsty; I eat bread.”

Roy Kotansky31 proposes a better reading and a better translation: λύκος πινῶν ⟦ων⟧ ἐβώσκετο. πίνω ὕδωρ διψῶ, ἄρτον φάω “The hungry wolf ate, I drink water, being thirsty, being hungry, I eat bread!”.

3) Spier, An Antique Magical Book, 62-63, fig. 12 (Fig. 4a and 4b): (Side a) All around a wolf and a crocodile: λύκος πινoν ἐπὶ λιμoξερo τάφῳ πι πίνω.

Spier understands: λύκος πινῶν ἐπὶ λιμoξήρῳ τάφῳ ’πιπίνω. “a hungry wolf, drink at the ditch (or grave?) dried by famine”.

I suppose that the πι at the start of πίνω is the result of dittography, and thus I read: λύκος πινῶν ἐπὶ λιμoξήρῳ τάφῳ. ⟦πι⟧π(ε)ινῶ. “a hungry wolf on a tomb dried by famine. I am in need.” (Side b) + M(arcus) Nonius Licumens DSPDIC (de sua pecunia dicavit)

28

Barb, ‘Magica Varia’, 352. P. Perdrizet, Negotium perambulans in tenebris (Strasbourg, 1922) 32. 30 H. Seyrig, ‘Invidiae medici. 1. La faim de l’ibis et la soif de Tantale – 2. Saint Sisinnius et les petits enfants – 3. La souplesse du moissonneur’, Berytus 1 (1934) 1-11. 31 Kotansky in Spier, An antique magical Book, 61 29

the hungry wolf

69

4) Spier, An Antique Magical Book, 63-64, fig. 13 (Fig. 5a and 5b): (Side a) Over the image of a saint riding on horseback and piercing a demon: Φεῦε φεῦε Ἀβίσιον, ἔνθα γὰρ κατοικῖ Σισίνις καὶ Σισίνια καὶ ὁ λάβραξ ΚΟΙΩΝ (= κύων). Τιτιάνος. (all around the rider): Εἷς Θεὸς ὁ νικῶν τὰ κα(κά).

Flee, flee, Abyssus, Sisinius and Sisinia and the voracious dog dwell here. (Protection of) Titianus. Unique is the God who conquers the evil!

(Side b) A wolf over a crocodile and an ibis menacing a snake. Above: ἵππος, μῦλος, εἶβις, ἐθυεῖα κόλε ἀνδρός, στρουθοκάμηλος, Ἀπoλλώνιος ὁ Τοιανεύς. (all around the wolf and the crocodile): λύκος πειναω ἐβώσκεν λιμωξήρo τάφῳ ⟦τάφῳ⟧ i.e. λύκος πειν ἐβώσκεν λιμoξήρ τάφῳ

Horse, mule, (or Nile?) ibis, phallus of a man, ostrich, Apollonius of Tyana.

The hungry wolf ate on a tomb dried by famine.

The “horse” logos recurs on three Syrian amulets as well.32 5a) Barb, ‘Magica varia’, 345; Spier, An Antique Magical Book, 60, fig. 10 (Fig. 6). (Side a) A saint subdues a demon by using a whip and thanks to the inscription: ἵππος, Μιχαήλ, Γαβριήλ, Οὐριήλ, βοήθ[ι] [λιμ]ός σε ἔσπ(ε)ιρεν ἄλιψ. (= φλέψ) ἐθέρισε. ἄλευε σφραγὶς Θεοῦ μεθ᾽ἡμῶν γενοῦ. ἐγὼ εἰμὶ Νοσκαμ. καταφάγε.

Horse, Michael, Gabriel, Ouriel, help (me). Hunger sowed you, a vein harvested you. Flee! Seal of God be with us! I am Noscam. Devour!

32 See P. Perdrizet, Negotium perambulans in tenebris (Strasbourg, 1922) 31-32 and Barb, ‘Magica Varia’, 344-46, figs. 1-2.

70

attilio mastrocinque (Side b) The evil eye harmed by obnoxious animals and:

Φεῦε φεῦε Ἀβύζου, ἔνθα γὰρ κατοικεῖ Σισίνις καὶ Σισίνια καὶ ὁ λάβραξ ΚΟΙΩΝ (=κύων). φύλαξον Κονσταντινον ὃν ἔτηκεν Θ Χριστίνα. Εἷς Θεός.

Flee, flee, Abyssus. For Sisinius, Sisinia and the voracious dog dwell here. Protect Constantinus, whom the deceased Christina bore. Unique God.

Barb read ἄλευε at the end of l. 3, which is the Homeric verb ἀλέομαι, “flee from one’s life”. The Θ is the well-known abbreviation of thanatos, meaning that the mother Christina was dead. 5b) Barb, ‘Magica varia’, 345; Spier, An Antique Magical Book, 60, fig. 10 (Fig. 7). This is a fragment of a replica of the previous amulet, and probably they were made for two brothers: ἵππος, Μιχαήλ, Γαβριήλ, Οὐριήλ, βοήθ[ι] [λιμός σ]ε ἔσπ(ε)ιρεν αλὶψ (= φλέψ) ἐθέ[ρισε. ἄλευε. σφραγὶς Θ]εοῦ μεθ᾽ ἡμ[ῶν]

Horse, Michael, Gabriel, Ouriel, help (me). (Hunger) sowed (you, a vein) harvested you. (Flee! Seal of G)od be with us!

Φεῦε φεῦε Ἀβίζιον, ἔνθα γὰρ κατοικῖ Σισίνις καὶ Σισίννια [καὶ] ὁ λάβραξ ΚΟΙΩΝ (=κύων). φύ[λαξο]ν Θεόδωρον ὃν ἔτη[κεν Θ Χ]ριστίνα. Εἷς [Θεό]ς. ΤΛ

Flee, flee, Abyssus. For Sisinius, Sisinia and the voracious dog dwell here. Protect Theodoros, whom the decease Christina bore. Unique God.

In line 3 the text is, in both amulets, αλιψ, whereas the inscription no 1 has φλέψ ε. It is possible that the original spell was corrupted because the reading of the models was difficult. The syntax runs better with three verbs referring to seeds. The only verb related to wheat production could be ἀλέω, “I mill”, whose aorist would have been ἤλησε, and not ἄληψε. 6) Spier, An Antique Magical Book, 64, fig. 14 (Fig. 8). Bronze ring bezel with a wolf and a crocodile. Above: ΠΙΝΟ = πίνω, “I drink”, or π(ε)ινῶ, “I am in need”.

the hungry wolf

71

The spell of the wolf is known also thanks to the late antique treatise of medicine by Marcellus Empiricus:33 Cum te in lecto posueris, ventrem tuum perfricans dices ter: “Lupus ibat per viam, per semitam, cruda vorabat, liquida bibebat”. physicum hoc ad digerendum de experimento satis utile. “When you will be lying on a bed, rub your belly and say three times: ‘A wolf went along a road, along a path. Raw food it devoured, liquids he drank’. After having tested it, this remedy proves natural and quite useful to digest”.

Pliny the Elder, on the other hand, reports a recipe against some insects, the cantharides, which were supposed to cause a fungal disease, the inpetigo / impetigo, i.e. ringworm or tetter. As this disease reddens the skin, it was conceived as depending on blood, and therefore the wolf, craving blood, was to frighten the obnoxious insects. Lapis vulgaris iuxta flumina fert muscum siccum, canum. hic fricatur altero lapide addita hominis saliva; illo lapide tangitur inpetigo. qui tangit, dicit: φεύγετε κανθαρίδες, λύκος ἄγριος αἷμα διώκει. “There grows, near running streams, a dry, white moss, upon ordinary stones. One of these stones, with the addition of human saliva, is rubbed against another; after which the first stone is used for touching impetigo, the party so doing uttering these words: ‘Cantharides be gone, a wild wolf seeks your blood’.”34

3. Thirsty and hungry people What is the logic of a dangerous animal craving liquids and food, which is invited, in some occurrences, to devour or to drink? The first clue resides in the description of some terrifying beings which crave both liquid and solid food, and for this reason the generic inscription is π(ε)ινῶ, concerning both needs. In the first paragrapher we have dealt with the amulets with Tantalos and Ares. Ares fought, wounded enemies, and made their blood flow; Tantalus eternally craved water and food, and in the same manner the wolf was empowered in its traditional hunger by 33

Marc. Emp. 20.78. Plin. HN 27.100, translation by Bostock and Riley. It is possible that a similar belief inspired a papyrus amulet where a white wolf is urged to heal a certain Joseph suffering from malaria (λύκος λευκός, λύκος λευκός, λύκος λευκός, θεραπευσάτω τὸ ῥιγοπυρετὸν Ἰωσῆφ): D. Wortmann, ‘Der weisse Wolf. Ein christliches Fieberamulett der Kölner Papyrussammlung’, Philologus 107 (1963) 157-62. 34

72

attilio mastrocinque

eating a sap produced by a hungry plant sowed by famine. In the short inscription the simple verb π(ε)ινῶ with the image of the hungry and thirsty animal sufficed to create agency for a magic act. A ferocious god such as Ares was used to make blood flow, and, on the other hand, the savage wolf was useful in some cases to fight other obnoxious beasts such as the crocodile, and the ibis was useful because it devoured snakes. The ibis was depicted as eating a snake and the inscription described it as being in need of liquid and/or food. Tantalos is described as being thirsty, διψὰϲ, and is invited to drink blood: Τάνταλε αἷμα πίε. Evidently blood was at his disposal. The topic of a hungry and emaciated being is recurrent not only in magical spells but also in some humoristic Greek literature. Limoxeros was the name of a character in the Philogelos, a fourth-fifth century CE collection of funny stories,35 wherein he is an exceedingly poor, greedy, and avaricious man. The same word limoxeros was used in the Life of Aesop, whose early written version is currently dated to the first-second century CE36. In the version G of the Life (§ 18) Aesop is depicted as a hungry and poor slave who was carrying a basket filled with bread. The opposition between the hungry and emaciated man and the basket of bread recurs in the Philogelos, as well, where Limoxeros provides his daughter with a house located in front of a bakery. The Lippert gem presents a similar opposition between its two sides: the lion-headed god holds three ears of grain in one hand and, in the other, the serpent-god Chnoubis. Ears represent abundance of food and Chnoubis was the god of liquid substances and was sometimes accompanied by the hieroglyphic sign for water.37 As in the case of the Tantalos gems, also this gem could have been used in two different ways and for two different aims, either to produce food and liquids or to deprive of them. The Tantalos gems are engraved, on one side, with an image of Ares and the spell of Tantalos and, on the other side, with the womb shaped as a cup, and this womb-cup was the mythical cup of Tantalos, which produced liquid automatically for the gods when they had drunk.38 As Faraone39 has proposed, these gems were used in two ways, for either stopping the flow of blood or producing it. As in a battery, this kind of gems generated two opposite poles. 35 A. Thierfelder, Philogelos sive Facetiae (sub auctoribus Hierocle et Philagrio). Philogelos der Lachfreund von Hierokles und Philagrios (Munich, 1968) 219-26. 36 See M. Andreassi, ‘Il λιμόξηρος nella Vita Aesopi e nel Philogelos’, ZPE 158 (2006) 95-103. 37 Mastrocinque, Les intailles, 93. 38 See Mastrocinque, Les intailles, 132. 39 Faraone, ‘Does Tantalus Drink the Blood, or not?’, 248-73.

the hungry wolf

73

4. Thirsty and hungry characters as magical agents Both the wolf and the dog recur on some ancient spells where they are summoned to flee, evidently because they were supposed to be obnoxious. A late fourth or early third century BCE amulet from Crete, consisting in a lead lamella, bears the following inscription: Ἔπαφος, Ἔπαφος, Ἔπαφος, φεῦ[γ᾽], ἅμα φεῦγε, λύκαινα, φεῦγε, κύων ἅμα συ, καὶ Πρόκλοπος ἅτε σύνοικος. μαινόμενοι δράντων πρὸς δώματα αὑτοῦ ἕκαστος. “Attacker, Attacker, Attacker, flee, together flee, she-wolf, flee dog, together with you, also. Thieving One, as you are their associate. Raving let them run, each to his own home.”40

The same spell is repeated on a lead amulet from Phalasarna41 and a spell on the so-called “Philinna papyrus” reports a similar text: Φιλίννης Θε[σσ]αλῆς ἐπαοιδὴ π[ρὸς] κεφαλῆς π[ό]νον. φεῦγ’ ὀδύν[η κ]εφαλῆς, φεῦγε Ἰδαί[ας] ὑπὸ πέτ[ρα]ς· φεύγουσιν δὲ [λύ]κοι, φεύγ[ουσι] δὲ μώνυχες [ἵπ]ποι … “The incantation of Philinna from Thessaly for headache: Flee headache, flee beneath Idaean crags! Wolves flee and single-hoofed horses flee…”42

In these spells both the wolf and the dog were dangerous beings, and also on some magical gems a dog is depicted within an ouroboros snake which surrounds it and other obnoxious beings.43 However, these beasts were also used in order to avert even worse dangerous beings such as demons. Let us read again the spell of Sisinius: ἔνθα γὰρ κατοικεῖ Σισίνις καὶ Σισίνια καὶ ὁ λάβραξ κύων. “Sisinius and Sisinia and the voracious dog dwell here!”

The menacing hungry dog or wolf was thus thought of as protection for a household. The hungry wolf was, in such cases, a useful character. The 40 M. Guarducci, Inscriptiones Creticae II (Rome, 1939) no 7; translation by C.A. Faraone, ‘A Collection of Curses against Kilns (Homeric Epigrams 13.7-23)’, in A. Yarbro Collins and M.M. Mitchell (eds.), Antiquity and Humanity. Essays on ancient Religion and Philosophy presented to H.D. Betz (Tübingen, 2001) 435-49 at 439. 41 Faraone, ‘A Collection of Curses against Kilns’, 444. 42 PGM XX 15-20; translation by Faraone, ‘A Collection of Curses against Kilns’, 440. 43 See e.g. A. Mastrocinque, ‘Studi sulle gemme gnostiche XI. Amuleto per il respiro; attributi di Persephone; gemma contro i ladri e significato della testa di toro’, Thetis 10 (2003) 89-95.

74

attilio mastrocinque

use of either protective animals killed and buried under a threshold or statuettes representing them and buried in the same way was of old in the ancient Near East, in Greece, and Italy.44 Their souls were supposed to remain close to their burial place to threaten and harm uninvited comers. Thanks to the above mentioned inscriptions we understand the use of the wolf and the ferocious dog: it was a sort of bogeyman which protected a house or a person by scaring evil beings. The order καταφάγε: “munch!” was addressed to the hungry beast which was summoned to eat an obnoxious being. This is the case with the cantharides which risked to be eaten by the wolf, and with the snakes which were devoured by the hungry ibis. Therefore hunger and thirst were used to put an animal or a hero (such as Tantalos) at one’s disposal either to eat/drink another obnoxious animal or a substance or to threaten it. These animals and heroes were totally fantastic, possibly issued from a ritual, activated and implemented by means of some spells. The manipulation of the soul of animals by killing and treating their corpse was recurring in antiquity. For example, thanks to a recipe from a judaizing book, the Sepher ha-Razim (the Book of Mysteries),45 written in the Late Antiquity, we know that the head of a dog, probably a puppy or a fetus,46 was used for magical rituals aimed at sending an obnoxious ghost to torment an enemy during his/her sleep. In this way we know that ghosts of dogs were somehow created and activated, and the wolf recurring in our amulets was a similar product of imagination implemented by magical enactements. We know of another meaning of the feeding of dangerous and obnoxious agents causing diseases. In fact, sometimes the wolf and other thirsty and hungry characters were images of the obnoxious agent which caused those diseases. They were summoned to give up to the consumption of human flesh and blood and drink or eat, instead, something else which was offered to them. A clue for this interpretation is provided by the following recipe in the Anecdota Atheniensia:

44 C.A. Faraone, Talismans and Trojan Horses: Guardian Statues in Ancient Greek Myth and Ritual (Oxford, 1992). 45 M.A. Morgan, Sepher Ha-Razim (The Book of Mysteries) (Chico CA, 1983); B. Rebiger and P. Schäfer (eds.), Sefer ha-Razim I und II: Das Buch der Geheimnisse I und II, 2 vols. (Tübingen, 2009). 46 See A. Bellusci, ‘Oneiric Aggressive Magic: Sleep Disorders in Late Antique Jewish Tradition’, in S. Bhayro and C. Rider (eds.), Demons and Illness from Late Antiquity to the Early Modern Period, Series of Studies in Ancient Medicine (Leiden, 2015).

the hungry wolf

75

ἀστέρα (= ὑστέρα) μελάνη μεμελανωμένη, αἷμα τρώγεις, αἷμα πίνεις, ᾽ς τὸ αἷμα συντελείεσαι. φεῦγε, ῥίγος, ἀπὸ τὸν δοῦλον τοῦ Θεοῦ ὁ δεῖνα καὶ τάξω σοι πέντε πίνακα μέλι καὶ πέντε πίνακα γάλα νὰ τρώγῃς καὶ νὰ πίνῃς, φεῦγε, ῥίγος, ἀπὸ τὸν δοῦλον τοῦ Θεοῦ ὁ δεῖνα· ἀμήν.47 “Black and blackened womb,48 you munch blood, you drink blood, contribute to blood! Flee, chill, from the servant of God NN., and I put at your disposal five dishes of honey and five of milk for you to munch and drink. Flee, chill, from the servant of God NN. Amen.”

This precious recipe opens a window towards another, hitherto unknown world, that of rituals which accompanied the use of amulets. When a spell says “The hungry wolf ate, I drink water, being thirsty, being hungry, I eat bread!”, the wolf is summoned to drink and to eat something which was offered to it. Only in this case the wolf ate and the first person “I drink water, I eat bread” is to be compared to the first person πεινῶ, “I am in need”. It is possible that water and bread were offered to the wolf instead of human blood. The images, inscriptions, and related rituals dealt with in this paper were means to control dangerous situations. Amulets and probably also the related remains of ritual acts (killing of a dog? use of some roots? offering of food and drink?) were used to prevent obnoxious beings from attacking and to make some liquids of the human body either flow or stop flowing. Creating a personality such as that of the wolf was a good means to control a dangerous situation. The recognition of the enemy was vital for controlling him. The hunger and thirst of these fantastic animals were used to threaten the agents of diseases but, in other cases, when they were supposed to be the cause of diseases, they were quenched with some food and drink and the animals were supposed to eat and drink them instead of human flesh and blood.

47

A. Delatte, Anecdota Atheniensia I, 141. The beginning of the spell is typical of spells for the womb, ὑστέρα, but γαστέρα can be taken into account as well. 48

76

attilio mastrocinque

Fig. 1a

Fig. 1b Image from LIM 132

Fig. 2a Image from Łatjar et al., ‘A Squeeze of a Magical Gem’. Used with the kind permission of Paweł Starzyński

the hungry wolf

Fig. 2b Image from Spier, ‘An antique magical Book’. Used with the kind permission of Paweł Starzyński

Fig. 3a Fig. 3b Image from Spier, ‘An antique magical Book’. Used with the kind permission of J. Spier

77

78

attilio mastrocinque

Fig. 4a Fig. 4b Image from Spier, ‘An antique magical Book’. Used with the kind permission of J.Spier

Fig. 5a Fig. 5b Image from Spier, ‘An antique magical Book’. Used with the kind permission of J. Spier

the hungry wolf

Fig. 6 Image from Spier, ‘An antique magical Book’. Used with the kind permission of J. Spier

Fig. 7 Image from Spier, ‘An antique magical Book’. Used with the kind permission of J. Spier

79

80

attilio mastrocinque

Fig. 8 Image from Spier, ‘An antique magical Book’. Used with the kind permission of J.Spier

VI. THE LATEANTIQUE TRANSFER OF CIRCULAR GEMDESIGNS TO PAPYRI AND FOIL: THE OUROBORUS AND SOLOMON’S SEAL Christopher A. Faraone

A gold lamella of early second-century CE date was found in the ruins of a Roman villa in Latium (Fig. 1).1 In the center we see a shorthaired man, who stands naked in a frontal pose and holds three poppies in his left hand and a knife in his right, while a snake entwines itself upwards around his body. Scholars, captivated by the various oddities that this figure presents, have proposed various explanations for his identity, for example, that he is a version of the lion-headed Mithraic god or the magical god Time (Chronos),2 but such debates will not detain us, because we will ask a much simpler question: although the gold lamella itself is roughly rectangular in shape, as are most foil or papyrus amulets, why is the figure framed by a carefully inscribed oval along whose inner edge are inscribed three wellknown magical words?3 This image, I will argue, is not just a representation of a man and a snake, but rather it is a representation of an oval object, indeed a gem, engraved with a man and a snake. One can assert this, because the oval or elliptical shape and the placement of inscriptions around a central figure are both features that we often find on the magical gems of the Roman period, as we see for instance in Fig. 2. It is not, then, difficult to imagine that the image on the gold foil from Latium, as well as the words written around it, were originally designed for a gem amulet, but that in the early second-century CE they were for some reason transferred to a different medium. As we shall see in what follows, there is indeed much evidence for such a transfer, both in the extant recipes for foil and papyrus amulets and in the texts of the amulets 1

GMA 29; drawing by Kotansky and used with permission. See the excellent discussion of Kotansky apud GMA 29. 3 The words ablanathanalba, Iaō and akramachamari run around the inside of the oval border; another, simelsam, is written horizontally in three segments on either side of his hips and knees. 2

82

christopher a. faraone

themselves, where we can see from time to time that a scribe has mistakenly copied handbook instructions onto the amulet or has made other such illustrative mistakes. Some of the best examples of these transfers from gem-carvings to drawings involve the ouroboros, an Egyptian design of a snake devouring its own tail, which in Roman times was often used on oval and elliptical magical gems as a border, in the midst of which we find powerful words or symbols associated with the famous Seal of Solomon. But before we begin, a few words about the so-called “magical gems”, which number in the thousands and appear in a wide variety of designs.4 Usually dated to the Roman imperial period, they seem to disappear by the mid fourth century and the most popular designs on them –for example, Heracles fighting the lion– never appear in the late-antique magical handbooks from Egypt, despite the fact that these papyri contain numerous recipes for many other kinds of amulets.5 One explanation, that I have addressed elsewhere, is the fact that most of these longer handbooks were found in Upper Egypt, far from the eastern Mediterranean basin where the magical gems were most popular.6 Here, however, I will argue for an additional and important technological change: that in Late Antiquity, when most of the longer papyrus handbooks were copied down, the paucity of gem recipes reflects the loss of the skills necessary for cutting and carving gems. As a result, some recipes originally designed for producing gem amulets were rewritten for much simpler production either to be inscribed on a soft metal foil with a stylus or drawn in ink on a small square or rectangle of papyrus.

4 In the nineteenth century they were labeled “Gnostic”, because some of the names and images found on them were thought to be similar to those used by the Gnostics, early dissident Christians, who were eventually deemed heretical and driven underground. Although there are rare cases where these gems clearly do preserve Gnostic images or prayers the adjective has been abandoned in favor of the “magical”, for example, in the titles of the important studies of SMA and D&D. For recent reviews of the modern history of these terms, see, e.g., R. Gordon, review of S. Michel, Die magischen Gemmen im Britischen Museum, in Journal of Roman Archaeology 15 (2002a) 666-70 or A.M. Nagy, ‘Magical Gems and Classical Archaeology’, in C. Entwistle and N. Adams (eds.), ‘Gems of Heaven’. Recent Research on Engraved Gemstones in Late Antiquity c. AD 200–600 (London, 2011) 75-81. 5 C.A. Faraone, The Transformation of Greek Amulets in Roman Imperial Times (Philadelphia, 2018) 9-10. 6 C.A. Faraone, ‘The Problem of Dense Concentrations of Data for Cartographers (and Chronographers) of Ancient Mediterranean Magic: Some Illustrative Case Studies from the East’, in M. Piranomante et al. (eds.), Contextos mágicos/Contesti magici (Rome, 2012) 107-10.

the late-antique transfer of circular gem-designs

83

1. The Ouroboros and the “Seal of Solomon” We can get an idea of the kind of recipe that may have produced the gold foil from Latium by looking at a late-antique handbook that probably came from Upper Egypt near Thebes (PGM VII 579-90):7 “Phylactery: A bodyguard against demons, against phantoms, against every sickness and suffering. To be written on a leaf of gold, silver, tin or on hieratic papyrus. It is carried like a seal, for it is the name of power of the great god and [his] seal and is as follows [long string of magic words]. Those are the names. The symbol is like this: let the snake be biting his tail and let the names be inside the snake and the symbols, too, in this way, as follows: [7 symbols]. The whole figure is drawn thus, as given below, with “thoroughly protect my body [and] the entire soul of me, NN.” And when you have consecrated [it] wear [it].”

We should note in passing that the use of the third-person imperative ἔστω in the instructions is a fairly common way to describe designs in the magical handbooks: “let the snake be biting his tail.” (ἔστω ὁ δράκων οὐροβόρος). As promised in the recipe, a drawing of this design is provided in the next column of the papyrus (Fig. 3).8 This recipe, then, includes a detailed written description of the design and also a full diagram, although we should note that they are not identical: the diagram, for example, has more than twice the number of symbols as those given in the description and it has letters inscribed in a circle outside of the ouroboros-ring that are not mentioned in the recipe. Although the recipe suggests either metal foil or papyrus as the medium for this phylactery, Vitellozzi points out that such designs with the snake-enclosed texts and symbols appear by the dozens on magical gemstones, like those, for example, in Figs. 4 and 5 (BM 536 and 537).9 This leads him to prefer a neglected emendation of Preisendanz to the text of PGM VII. The papyrus reads φορούμενον στρατιωτικῶς ἐστιν, which the Betz translation renders as “When worn, it works mightily.” The adverb στρατιωτικῶς, however, would be best translated as “in the manner of a soldier”, which presumably would refer to some kind of military badge. Preisendanz, however, 7 I follow the translation in GMPT with a few minor changes and one major one: instead of the reading on the papyrus “like a soldier” (στρατιωτικῶς) I use Preisendanz’ emendation “like a seal” (σφραγιστικῶς), as explained below. 8 This excellent drawing is by Raquel Martín Hernández and is used with her permission. 9 P. Vitellozzi, ‘Relations between Magical Texts and Magical Gems. Recent Perspectives’, in S. Kiyanrad et al. (eds.), Bild und Schrift auf ‚magischen‘ Artefakten (Berlin and Boston, 2018) 181-253.

84

christopher a. faraone

suggested plausibly that the word στρατιωτικῶς was an error for another, similarly spelled adverb, σφραγιστικῶς, which could be rendered as “it is worn like a seal.” Vitellozzi points out that this adverb recalls the idea of a gemstone set in a signet-ring and he notes how the emendation makes much better sense of the explanation that follows in the recipe: “It is to be worn like a seal (σφραγιστικῶς), because it is a name of power of the great god and his seal (σφραγίς).” As we shall see, in Roman times the “Seal of Solomon” was indeed sometimes thought to be the stone of a signet ring engraved with an ouroboros, that encircles symbols or nonsense words thought to be the name of the Jewish god. Vitellozzi is surely correct in his reading and I would only add two thoughts: (i) that this recipe was probably first created in the Roman period for the production of an engraved gemstone like the many extant examples, but that by the early fourth century CE –the date of the PGM VII– someone had adapted the instructions to a different technology: writing on metal foil or papyrus; and (ii) that this change in media made little sense of the original instruction “wear it like a seal” and eventually the word became corrupted or reinterpreted. This interpretation of the back history of the recipe also helps explain the ring of letters in the diagram (fig. 3) outside of the ouroboros, which were perhaps to be placed on the bezel of the gem.10 Luckily, we need not rely on our imaginations to reconstruct what such earlier recipes looked like, thanks to the sloppy artisan who produced a green gem that is now lost, but whose design was copied in the early seventeenth century by Nicolas-Claude Fabri de Peiresc, a friend of the painter Peter Paul Rubens (Fig. 6). As we can see once again, a lion-headed ouroboros serpent ran along the edge of the oval gem, just as it does in the drawing in the PGM recipe above in Fig. 3. Twelve lines, almost all of it Greek text, were inscribed within the serpentine border:11

10 These letters seem to be a corrupt version of the Arponchnouphi-logos, for which Bonner SMA pp. 204-205 and DMG 483 and 494. The radiating lines dividing up the letters do not correspond to the usual division of the individual words of the logos and were meant to indicate the slanting edge of the bezel. 11 R.D. Kotansky and J. Spier, ‘The ‘Horned Hunter’ on a Lost Gnostic Gem’, Harvard Theological Review 88 (1995) 315-37, whose text and translation I use. For other examples of such scribal confusion, see D.R. Jordan, ‘Il testo Graeco di una gemma magica dall’ Afghanistan (?) nel Museo Pushkin, Mosca’, in A. Mastrocinque (ed.), Atti dell’ incontro di studio: Gemme gnostiche e cultura ellenistica, Verona 22-23 Ottobre 1999 (Bologna, 2002) 61-68 for a gem in Moscow, whose text begins: “This is the logos: Abaichôrmuid” or LIM 286, where around the shield of the anguipede the scribe wrote the words “the vowels” instead of inscribing the seven vowels individually.

the late-antique transfer of circular gem-designs

85

ὁ Ἰοαηουαυη, ὁ Βακαξιχυχ, ὁ Κερατάγρας, οὗτος ἐστιν ὁ πρωτοπάτωρ, ὁ τοῦ σώματος μόνος ὤν, διὰ πάντων πορευ⟨ό⟩μενος. Σολομῶντος {ι}σφραγίς * ΖΖΖ (fig.) ὁ δρακὼν ἔστω λεοντοκέφαλος “O Ioaêouauê, O Bakaxichuch, O Keratagras. This is the First-Father, the one being single of substance, passing through all (things). The Seal of Solomon * ΖΖΖ (fig.). Let the snake be lion-headed.”

The space marked “(fig.)” in the modern transcription above refers to a tiny drawing of a serpent in line 9 of Peiresc’s drawing in Fig. 6 and the words that follow the drawing –“let the snake be lion-headed”– refer to this same drawing. As the editors suggest, both seem to have been copied mistakenly from a handbook recipe,12 which presumably recommended carving the familiar image of the lion-headed snake-god Chnoubis (with his characteristically single-looped tail) that we so often find as the main design on magical gems, such as we see in Figs. 7-8. But here, thanks to the error of the gem-cutter we catch a rare glimpse of a much earlier handbook recipe – perhaps of second cent. CE date – and one that probably came from outside of Egypt,13 because neither the image of the Chnoubis snake nor his name and symbol ever appear in the Greek magical papyri from Egypt. This handbook, moreover, like the recipe in PGM VII discussed above, apparently included within the text of the recipe a drawing of the Chnoubis snake, as well as words of instruction that use the same form of a third-person imperative ἔστω that we saw earlier: “Let the snake be biting its tail!” (PGM VII 586: ἔστω ὁ δράκων οὐροβόρος) and “Let the snake be lion-headed!” (Peiresc Gem: ὁ δράκων ἔστω λεοντοκέφαλος). There are, however, some further, undetected problems with the text on this lost gemstone sketched by Ruben’s friend. We know that the triple 12

Kotansky and Spier, ‘Horned-Hunter’, 323-24. The image, name and symbol of Chnoubis are curiously absent from the papyrus magical handbooks found in Egypt, but all three do appear in one lapidary handbook that survives in Greek manuscripts and the image is also described (and banned) by the rabbis in the Talmud. For discussion and bibliography, see Faraone, Transformation, 154. 13

86

christopher a. faraone

*ZZZ symbol seems to have been a version of the Jewish tetragramaton – a kind of secret name of Jahweh that is often equated, as it is here, with the Seal of Solomon.14 On gem amulets (Figs. 9-10), moreover, these symbols are often surrounded by an ouroboros, as they are on the lost gem. The ouroboros and these four symbols, however, never appear on Chnoubis gems, although as we saw in Figs. 7-8 the SSS and the ZZZ do appear, albeit never with the asterisk.15 I suspect, therefore, that this gem cutter (or the handbook he was using) mistakenly confounded two separate and probably contiguous recipes, the first for a gem carved with a lion-headed ouroboros surrounding an invocation that ended with the *ZZZ symbol and the second for a gem design showing a lion-headed Chnoubis and his SSS symbol, a mistake arising no doubt from similarity between the two leonine snakes and between Solomon’s *ZZZ and Chnoubis’ SSS. The first recipe, equated or labelled in the handbook as the Seal of Solomon, comprised a list of divine names followed by the tetragrammaton all surrounded by an ouroborus, while the second had only an image of Chnoubis and his symbol, but no ouroboros – indeed, despite the great number of extant Chnoubis gems the god is almost never encircled by the serpent. 2. More Scribal Errors with Evidence for Earlier Designs Other examples of scribal confusion have much to teach us about the contents of earlier handbooks and their re-use of gem illustrations. Take for example, a silver foil from Egypt of third- or fourth-century CE date (Figure 11):16 14

Kotansky and Spier, ‘Horned-Hunter’, 323-24. Ouroboros: Of the 38 Chnoubis gems in the Cabinet des Médailles (LIM 232-269) and of the 34 in the British Museum (BM 304-333 and 335-338) none have an ouroboros. LIM 334 has an ouroboros surrounding four symbols on one side and a magical name on the back that begins with Chnouphi, suggesting that the creator of the Peiresc gem was not the only gem-cutter to confuse the two lion-headed snakes. *ZZZ on Chnoubis gems: Of the gems in the British Museum none has a single asterisk with the ZZZ; two have the ZZZ alone (BM 305 and 315); one has ZZZ repeated three times (BM 306); two have *SSS (BM 313 and 317); three have a triple asterisk with the SSS (BM 310, 312*, 322). Of the gems in the Cabinet des Médailles, two have a triple asterisk with the ZZZ (LIM 243 and 244) and two have the ZZZ alone (LIM 255 and 263). One should note that often the curved lines of the ZZZ symbol suggest that it is simply the SSS in reverse. 16 D.R. Jordan and R.D. Kotansky, ‘A Solomonic Exorcism’, in B. Kramer et al. (eds.), Kölner Papyri, Abhandlungen der Nordrhein-Westfälischen Akademie der Wissenschaften: Sonderreihe, Papyrologica Coloniensia 7.8 (Opladen, 1997) 53-67 (P.Köln 8). I use their translation, except in the final line, where I translate ta daimonia not as “the demons”, but rather as “the divine things”, a reference to the four symbols that we do indeed find within the ouroboros. See further discussion below. 15

the late-antique transfer of circular gem-designs

87

ἐξορκίζω πᾶν πνεῦμα πονερον καὶ κακὸν…. ἐξελθεῖν ἀπὸ Ἀλλοῦτος, ἣν ἔτεκεν Ἄννις, τῆς ἔχούσης τὴν σφραγῖδα τοῦ Σολομῶνος … … Ιαω Σηθ Ιμααμ Πηρει Πιμακεραυιανιηρ Βιμμεσον Κιομμα Πανιαμι Βαρφαραλαξ Οζομο. Τί θλείβεις τὴν Ἀλλοῦν μηδέν Fig. σοι ἄδικον ποιήσασαν; σφραγῖδος Σολομῶνος ἐν ⟨μ⟩έσῳ τὰ δαιμόνια. “I adjure every wicked and evil demon …. To come out of Allous, whom Annis bore, who holds the Seal of Solomon … Iaô, Sêth, Imaam, Perêi Pimakerauianiêr, Bimmeson Kiomma, Paniami, Barfaralax, Ozomo. Why are you troubling Allous, who has Fig. done you no wrong? In the middle of the Seal of Solomon the divine things.”

In the bottom left corner of Fig. 11, we see a roughly drawn ouroboros, in which we find another common rendition of the tetragrammaton. On this amulet, however, the small diagram looks like the kind of seal one might indeed stamp in wax at the end of an official document, except for the fact that – judging from the appearance of this same design on gems – the design on this lamella has not been reversed as one would expect in the case of an impression. It is, rather, a representation of the face of the ring-stone itself and was probably used originally as a diagram for carving the obverse of the gemstone of a protective ring. The rest of the text on this lamella is an exorcism designed to force “every wicked and evil demon” out of a woman named Allous, “who holds the Seal of Solomon,” a reference to the circular design placed in the bottom corner of the silver foil. This seal, of course, is an appropriate device to inscribe on an amulet used to ward off evil demons, because according to popular belief Solomon had used his own signet-ring to imprison scores of evil demons in the foundations of the temple in Jerusalem.17 At the end of the text, moreover, we find next to the diagram yet another example of handbook language mistakenly copied onto an amulet: “In the middle of 17

Kotansky and Spier, ‘Horned-Hunter’, 321-24.

88

christopher a. faraone

the seal of Solomon (i.e. put) the daimonic things.” Since, like the *ZZZ, the four symbols inside the ouroboros are apparently another popular version of the tetragrammaton,18 the term daimonia here must mean something like “powerful symbols,” rather than “the demons” to be warded off. There are, in fact, a number of other lamellae engraved with such circular designs that seem to belong to this same Solomonic tradition. Take, for example, a papyrus amulet in Berlin, also of third or fourt century CE date (Fig. 12) that was inscribed with a variety of designs, magical names and (along the bottom) a prayer to “protect Touthous, whom Sara bore, from every shivering fit and fever.…”19 The plural imperative used in this invocation suggests that it addresses all of the names and symbols that precede it, including those in the tabula ansata on the far right and five angel names on the far left. What should interest us most, however, are the two oval shapes close to the center: on the left another ouroboros surrounds the magical name Semesilam and the seven vowels and on the right an oval containing yet another popular magical name – Salamaza Bamea Iacha – has been placed above and below the by-now-familiar ZZZ. At this point, we should ask again why these last mentioned words and symbols are enclosed in these oval frames. I suggest, in fact, that they, too, were originally designs for the obverse and reverse of a popular type of magical gem, of which we find numerous examples in cornelian, as we can see in Figs. 13-14: they are inscribed on one side with an ouroboros encircling a magical name or two rows of symbols and on the other a series of magical names that almost always includes part or all of the four magical words Salamaza Bamaiaza Aianachba Amorachthi, which was also thought to be a secret name of the Jewish god.20 The invocation also ends, sometimes, with a prayer for protection, using the same Greek verb phylassein that is employed on the papyrus amulet under discussion, for example, a cornelian gem in Venice, whose invocation ends with “Protect from every demonic thing Sabina, whom Calpurnia bore!” or another in Paris: “Protect (Bar?)boulos, whom Akairea bore!”21 18

Kotansky and Spier, ‘Horned-Hunter’, 321-24. Suppl.Mag. 10 = PGM CVI in GMPT and originally published by W. Brashear, ‘Vier Berliner Zaubertexte’, ZPE 17 (1975) 27-30. 20 For discussion of the ouroboros and this Aianagba logos or Salamaza Bamaiza logos with earlier bibliography, see DMG 134-137 and A. Mastrocinque, Les intailles magiques du département des Monnaies, Médailles et Antiques (Paris, 2014) 236-39, who suggests that they are another representation of the tetragrammaton. 21 SGG 2 VeC 17 and LIM 686, with Mastrocinque’s comments in both places. Shorter prayers include “Protect!” (LIM 684) and “Soumarta!” (BM 513), the latter a Semitic word with the same meaning that is sometimes used on Greek amulets; see Faraone, 19

the late-antique transfer of circular gem-designs

89

The papyrus amulet in Berlin, then, incorporates a number of different protective texts and designs. I suggest, however, that the two oval designs and the prayer inscribed beneath the second oval were copied from a handbook recipe designed originally for a carnelian gem. The two protective prayers from Venice and Paris quoted above were inscribed on the bezel of the cornelian gems, a fact which suggests that the original handbook contained oval drawings of the obverse and reverse of the gem and then the text of the prayer itself along with instructions that it be inscribed along the bezel of the gem. We should note, in fact, that the size and shape of the ovals on this papyri is typical of the size of the gems that carry similar formulas. The placement within the second oval, moreover, of a version of the Aianachba-logos above and below the ZZZ symbol simply affirms that we have here another version of the Seal of Solomon. A foil amulet that appeared on the antiquities market in Munich in the early years of this millennium (Fig. 15) shows on the same side a similar pair of designs linked to a similar set of protective prayers: the first “Thoroughly protect Nymphikos, whom Sosike bore, from every demon and every evil!” and the second, which just repeats the beginning: “Thoroughly protect Nymphikos!”22 The larger design is encircled by the by-now-familiar ouroboros – one can barely make out its rather squashed head on the top edge – surrounding a series of magical words, mainly vowels, laid out in two columns, beneath which we find the prayer for Nymphikos, which thus takes the same position at the bottom of the ouroboros, that we found in the PGM VII drawing and the Berlin papyrus (Figs. 3 and 12). The poorly drawn oval design on the upper half of the Munich lamella also resembles the shape – and (in this case) even the dimensions – of a typical gemstone. It contains eight symbols and written along the edge in Greek letters the word κυθρεισιυμα, which the editor resolves as Κυ(ριε) Θρεισιυμα. The biblical quotation at the start of the tablet and the repeated references in the surrounding text to “the name of the god of Abraham, of the god of Isaac, etc.” all suggest that these symbols, too, were thought to be the name of the

Transformation, 185. See also SMA 271, a large carnelian (3.8 × 3.1 cm) with an ouroboros on the obverse encircling five lines of symbols, then the magical words salamaza balaiza amatropatheir and then: “Iaō Iaō, save the man who wears this phylakterion!” If amatropatheir (αματροπαθειρ) means something like “motherless father” and serves as an epithet for Yahweh, then we have further evidence that in the world of magic Salamaza Bamaiaza was thought to be yet another name for the Jewish god. 22 The drawing is after G. Manganaro Perrone, ‘Magia ‘benefica’ nella Sicilia tardoantica’, Epigraphica 69 (2007) 263-86, no 6, fig. 17.

90

christopher a. faraone

Jewish god.23 There is no handbook language on this lamellae, but the parallels with the preceding amulets suggest that the scribe, who produced this foil, copied out the designs and texts for at least two different protective amulets, presumably, as in the case of the papyrus amulet just discussed, in hopes of increasing the efficacy of the amulet by piling up different designs. Two other silver lamellae bear similar circular designs in their otherwise rectilinear layouts. The simpler version was found on the Esquiline hill in Rome (Figure 16) and has letters running along the top that seem, once again, to be borrowed from a handbook, in this case a rubric: “For those who suffer the moon”, a reference, as the editor shows, to “lunatics” or “epileptics”.24 The second was found in the Lycian city of Xanthus (Fig. 16):25 above the circle we find four lines of symbols and letters and below we find a prayer to the “holy symbols” asking them to “drive away every wicked demon, occurrence, happening, encounter or evil eye, and to drive and chase them away from John, whom Theoktista bore!” The Seal of Solomon was, as was discussed earlier, famously effective against demons, epilepsy and madness, so it may well be that all of these circular designs with the ouroboros were used to protect their owner from demons and cure the diseases that they cause. But regardless of their aims, it seems clear that they were originally invented by gem cutters for carnelian and other gems and were only later translated to drawings on papyrus and foil. 3. Conclusion We have seen, then, several examples of amulet designs encompassed by oval frames or ouroboroi:

23 G. Bohak and C.A. Faraone, ‘‘Pay heed, o heaven, and I will speak’ (Deut. 32:1): A Greek Amulet with Biblical and Angelic Names’, MHNH 18 (2018) 6-14. 24 G. Bevilacqua, ‘Magica varia dall’ Antiquarium Comunale’, Bollettino dei Musei Comunali di Roma (1999) 18-30. 25 D.R. Jordan and R.D. Kotansky, ‘Two Phylacteries from Xanthus’, Revue archéologique (1996) 161-74.

the late-antique transfer of circular gem-designs

medium

Frame

image

GMA 29 (2nd CE)

gold foil

PGM VII (4th CE)

foil or papyrus

inscribed oval Ouroboros

man snake none

Peiresc (Roman)

gem

91

magical names

symbols

prayer

four

none

none

many

many

protect against demons none

lion-headed “seal of none ouroboros Solomon” Egypt (3rd-4th CE) silver foil lion-headed “seal of many ouroboros Solomon” Suppl.Mag. 10 papyrus ouroboros none Aianachba(3rd-4th CE) and logos inscribed oval Munich (Roman) metal foil ouroboros none one + and vowels inscribed oval th th Rome (4 -5 CE) silver Ouroboros none many Xanthos silver foil Ouroboros none many (3rd-4th CE)

*ZZZ many ZZZ

exorcize demons protect against fever

eight

protect against demons

many many

lunacy drive off demons

In the cases of the gold-foil amulet from Latium (GMA 29), the papyrusamulet in Berlin (Suppl.Mag. 10) and the metal foil from Munich, the ovals inscribed on rectangular media provide us with an initial answer to the question posed at the start of this essay: why did the scribe encircle the image in this way? The answer is that he was probably using a recipe, probably with a drawing, designed initially for making a magical gem or perhaps he was using a gem in his possession as a model. The ouroboros does not, however, always encircle text. A recipe from a bilingual handbook in Leiden (PGM XII), for example, provide recipes for a ring that will bring the wearer success and favor. The rubric for the first reads “A little finger-ring for every operation and for success” and the recipe runs as follows (lines 201-11): “Kings and governors [try to get it?]. Very effective. Take an air-colored jasper, engrave on it a snake in a circle with its tail in its mouth and also in the middle of the snake a moon(?) having two stars on its two horns and above these the sun(?), beside whom “Abrasax” should be inscribed. And on the opposite side of this inscription (i.e. inscribe) the same name “Abrasax” and around the border you will write the great and holy and

92

christopher a. faraone omni-competent(?), the name Iaô Sabaôth. And when you have consecrated the stone wear it in a gold ring, when you need it, provided you are pure [at the time], and you will succeed in everything you may wish… Likewise, when engraved in gold, these designs also have the same effectiveness.”

The last line, as often happens in these recipes, seems to be a later addition and could refer to an all-golden ring, but, if this were true, one would expect more detail. More importantly there are to my knowledge no examples of a plain gold ring carved with an ouroboros or any of the other designs one finds on magical gems. Given the parallels collected above, however, of such designs on metal foil, it seems far more likely that the simple phrase “in gold” refers to a gold lamella, rather than a golden ring. We have, then, traced some important changes in amulet-making, a shift from carving gems on both sides with texts and images to a practice of inscribing metal foils and papyri with similar designs, but keeping all the images and texts together on a single side. The reason why image and text appear on the same side of the foil amulets can, of course, be easily explained for practical reasons: to remain legible the soft metal foils were usually inscribed on only one side, because the writing shows through in reverse on the other side. This is not the case, however, with papyrus, which can be inscribed on both sides, although papyrus amulets rarely carry text on their reverse sides. There was, then, a rich tradition of handbooks (both magical and lapidary), that originally included either detailed descriptions of gem-designs or drawings of them – or sometimes both, as we saw in the PGM VII recipe. Over time, however, as technologies and styles changed, these gem images, especially those surrounded by ouroboroi, were transferred to other media like metal foil and papyri that were more easily worked, especially by those trained as scribes, rather than gem-cutters. Nonetheless traces of the original patterns linger, especially in the oval and circular shapes of the ouroboroi and texts that surround them.

the late-antique transfer of circular gem-designs

Fig. 1. Drawing of a gold foil amulet from Latium after GMA 29, used with the permission of the editor and the Institut für Altertumskunde, University of Cologne.

2a. 2b. Fig. 2a. Dark green jasper gem in the British Museum depicting Aphrodite surrounded by her magical name (BM 76). Photo of the author reproduced by courtesy of the Trustees of the British Museum. Fig. 2b. Dark green jasper gem in the British Museum depicting the rooster-headed anguipede surrounded by magical names (BM 181). Photo of the author reproduced by courtesy of the Trustees of the British Museum.

93

94

christopher a. faraone

Fig. 3. Drawing of an amulet design from a papyrus handbook (PGM VII). Image created by R. Martín Hernández.

Fig. 4. Carnelian gem in the British Museum with ZZZ and other symbols and letters, all surrounded by an ouroboros; (BM 536). Photo of the author reproduced by courtesy of the Trustees of the British Museum.

the late-antique transfer of circular gem-designs

Fig. 5. Whitish-brown chalcedony gem in the British Museum with symbols and vowels surrounded by an ouroboros; (BM 537). Photo of the author reproduced by courtesy of the Trustees of the British Museum.

Fig. 6. Peiresc drawing of a lost chalcedony gem after photograph in Kotansky and Spier, ‘Horned-Hunter’.

95

96

christopher a. faraone

Fig. 7. Green jasper chalcedony gem in the British Museum with the lion-headed serpent Chnoubis on the obverse and his name and SSS on the reverse (BM 305). Drawing reproduced by courtesy of the Trustees of the British Museum.

Fig. 8. Whitish-brown chalcedony gem in the British Museum with the lion-headed serpent Chnoubis on the obverse and SSS on the reverse, both sides encircled by magical names (BM 304). Drawing reproduced by courtesy of the Trustees of the British Museum.

the late-antique transfer of circular gem-designs

Fig. 9. Carnelian gem in the British Museum with * ZZ surrounded by an ouroboros; (BM 534). Drawing reproduced by courtesy of the Trustees of the British Museum.

Fig. 10. Carnelian gem in the British Museum with symbols surrounded by an ouroboros; (BM 517). Photo of the author reproduced by courtesy of the Trustees of the British Museum.

97

98

christopher a. faraone

Fig. 11. Gold foil amulet from Egypt. Drawing after Jordan and Kotansky, ‘A Solomonic Exorcism’

Fig. 12. Papyrus amulet from Egypt. Drawing after Suppl.Mag. 10 used with the permission of the editors and the Institut für Altertumskunde, University of Cologne.

the late-antique transfer of circular gem-designs

Fig. 13. Carnelian gem in the British Museum with the Salamaza Bamaiaza logos surrounded by an ouroboros; (BM 519). Photo of the author reproduced by courtesy of the Trustees of the British Museum.

Fig. 14. Drawing after SGG 422 of a carnelian gem in Kassel with the Aianachba-logos on the obverse and letters and symbols on the reverse surrounded by an ouroboros.

99

100

christopher a. faraone

Fig. 15. Foil amulet in Munich. Drawing after G. Manganaro Perrone, ‘Magia “benefica” nella Sicilia tardoantica’.

Fig. 16. Silver foil amulet from the Esquiline hill in Rome. Drawing after Bevilacqua, ‘Magica varia dall’Antiquarium Comunale’.

the late-antique transfer of circular gem-designs

Fig. 17. Silver foil amulet from Xanthos. Drawing after Jordan and Kotansky, ‘Two Phylacteries from Xanthus’.

101

VII. THE ANGUIPEDE, ITS ORIGINS AND MARKET DIFFUSION Walter Shandruk

Estimates for the total number of inscribed “magical” gemstones being held in public and private collections range from 3,500 to 5,000.1 By 2004, Simone Michel’s magisterial monograph2 was able to include 2,600 gems from numerous collections and more have been published since then.3 The focus of this paper will be on a gem-type inscribed with a most remarkable figure, the rooster-headed snake-legged creature, the so-called Anguipede. It is perhaps the most famous iconographic element on magical gems, 258 being known to me.4 It is typically composed of five key features: snake legs, head of a rooster, shield in left hand, whip in right hand and armored torso. I would like to address two questions concerning the Anguipede: (1) how can we least problematically understand the bizarre visual design of the Anguipede and secondly, (2) is there a way to understand why the figure has been largely confined to one medium, that of inscribed magical gemstones. The first question has been debated at length for many years. The latter question has received less attention. But, before I continue, a quick remark concerning statistics and probability. I will on occasion in this paper make mention of p-values and 1 R. Gordon, ‘The Power of Stones: Graeco-Egyptian Magical Amulets’, Journal of Roman Archaeology 21 (2008) 713-18, esp. 713, n. 3; S. Michel, ‘(Re)interpreting Magical Gems, Ancient and Modern’, in S. Shaked (ed.), Officina Magica: Essays on the Practice of Magic in Antiquity (Leiden, 2005) 141-70, esp. 141. 2 S. Michel, Die magischen Gemmen: Zu Bildern und Zauberformeln auf geschnittenen Steinen der Antike und Neuzeit (London, 2004). 3 In 2007 part two of Sylloge Gemmarum Gnosticarum (Rome, 2007) was published by Attilio Mastrocinque and he most recently republished the Paris collection (Les intailles magiques du department des Monnaies, Médailles et Antiques [Paris, 2014]), re-edited and expanded by several hundred gems to a total of 693. 4 For general discussion see E.R. Goodenough, Jewish Symbols in the Graeco-Roman Period, vol. 2 (New York, 2016) 245-58; C. Bonner, Studies in Magical Amulets: Chiefly GraecoEgyptian (Ann Arbor, 1950) 123-39. For bibliography prior to A. Delatte, see A.M. Nagy, ‘Figuring out the Anguipede (‘snake-legged god’) and his relation to Judaism’, Journal of Roman Archaeology 15 (2002) 159-72, esp. 161, n. 12.

104

walter shandruk Table 1. Statistically Significant Links to the Anguipede

Anguipede (58%)

0.000148

Jasper (19%)

0.000022

149

Anguipede (57%)

0.000000

Iaô (30%)

0.000000

147

Anguipede (53%)

0.000002

Color Green (23%)

0.000000

136

Anguipede (24%)

0.000012

Abraxas (32%)

0.000019

62

Anguipede (13%)

0.002376

Sabaôth (26%)

0.003393

34

Anguipede (9%)

0.012317

Ablanathanalba (27%)

0.017652

23

Anguipede (8%)

0.014425

Heliotrope (28%)

0.020992

21

Anguipede (5%)

0.034826

Semesilam (26%)

0.0471

14

statistical significance. Part of my research has consisted of running a statistical test on every pair of inscriptional and iconographic elements on the nearly three thousand gems I have studied when writing my dissertation5. The goal has been to determine whether the collocation of any two attributes on a magical gem are statistically significant or not. Or, said differently, whether they occur together significantly more often than what would be expected due to the frequency with which they occur in the corpus in general. Whenever I mention a p-value then I am referring exactly to this; the smaller the p-value then the more statistically significant a correlation between any two attributes on a gem. All of these p-values have been calculated using Fisher’s exact test. 1. Origins Scholars have long argued that the Anguipede was primarily a solar deity;6 of Iranian extraction;7 a cosmic deity combining solar, chthonic and 5 W. Shandruk, ‘A Computational Approach to the Study of Magical Gems’, (PhD diss. unpublished, University of Chicago, 2016). 6 Y. Koenig, ‘Des «trigrammes panthéistes» ramessides aux gemmes magiques de l’Antiquité tardive: le cas de’Abraxas, continuité et rupture’, Bulletin de l’Institut Français d’Archéologie Orientale du Caire 109 (2010) 311-25; J. Coleman Darnell, The Enigmatic Netherworld Books of the Solar-Osirian Unity (Fribourg, 2004) 385-89; DMG 108-09, briefly entertains the Jewish/Iaô hypothesis (Goodenough, et al.) but ultimately sides with the solar interpretation. See also P. Post, ‘Le génie anguipède alectorocéphale: une divinité magique solaire’, Bijdragen: Tijdschrift voor Filosofie en Theologie 40 (1979) 173-210 at 202-04 ; Bonner, Studies in Magical Amulets, 127; A. Delatte and P. Derchain, Les intailles magiques gréco-égyptiennes (Paris, 1964) 30-33. 7 A. Alföldi, ‘Der iranische Weltriese auf archäologischen Denkmälern’, Jahrbuch der Schweizerischen Gesellschaft für Urgeschichte 40 (1949/50) 17-34 at 25-28.

the anguipede, its origins and market diffusion

105

earthly elements;8 a representation of “some one or other” of the powers of Yahweh with some solar associations;9 a visual representation of Yahweh as a Sun-God;10 and a visual epithet of Yahweh.11 As will be seen, the statistics strongly point to the Yahweh-centric hypotheses. About 57% of the time the word Iaô, the Hellenized form of the Hebrew God’s name, Yahweh, accompanies the Anguipede on gems and 30% of the time Iaô is found specifically inside the creature’s shield (on 14 examples other text is found accompanying Iaô inside the shield and on 21 examples other text has replaced Iaô). Twenty-four percent (62 gems, see Table 1) of the time the name Abraxas accompanies the Anguipede. This has led in earlier centuries to call the creature “Abraxas” and attribute to it a Gnostic pedigree; and, although Campbell Bonner decisively put this notion to rest,12 the appellative has lingered on, if only weakly, in some of the secondary literature.13 While the correlation between Abraxas and the Anguipede is statistically significant it is, in fact, only because of the even stronger link between Abraxas and Iaô. Although all of the p-values concerned are extremely small, only 32% of instances of the name Abraxas share an Anguipede, while 60% coincide with Iaô. Furthermore, in 46% of cases where Iaô is present with the Anguipede the name Abraxas is missing and only 14% of Anguipede gems have 8 M.P. Nilsson, ‘The Anguipede of the Magical Amulets’, Harvard Theological Review 44.1 (1951) 61-64; cf. J. Ferguson, The Religions of the Roman Empire (Ithaca, 1985) 167-68. 9 Goodenough, Jewish Symbols in the Graeco-Roman Period, 2, 252. 10 E. Zwierlein-Diehl, Antike Gemmen und ihr Nachleben (Berlin, 2007) 220 ; E. Zwierlein-Diehl, Magische Amulette und andere Gemmen des Instituts für Altertumskunde der Universität zu Köln (Opladen, 1992) 29-35; M. Philonenko, ‘L’anguipède alectorocéphale et le dieu Iaô’, Comptes-rendus des séances de l’Académie des Inscriptions et Belles-Lettres 123.2 (1979) 297-304. Outright identifying the creature as Iaô is also found in authors not specifically concerned with the Anguipede; see M. Piranomonte, ‘Religion and Magic at Rome: The Fountain of Anna Perenna’, in R. Gordon and F. Marco Simón (eds.), Magical Practice in the Latin West (Leiden, 2005) 191-213, esp. 209; Z. Pleše, Poetics of the Gnostic Universe: Narrative and Cosmology in the Apocryphon of John (Leiden, 2006) 188. I can in no way discern how Pleše arrives at the conclusion that Iaô is seven-headed or that he has lunar connections in addition to solar. 11 Nagy, ‘Figuring out the Anguipede’, 159-72. 12 SMA pp. 133-34. 13 J.G. Gager, Curse Tablets and Binding Spells from the Ancient World (New York, 1992) 65; A. Walker, ‘A Reconsideration of Early Byzantine Marriage Rings’, in S.R. Asirvatham et al. (eds.), Between Magic and Religion: Interdisciplinary Studies in Ancient Mediterranean Religion and Society (New York, 2001) 149-64 at 161; F. Marco Simón, ‘Abraxas. Magia y religión en la Hispania tardoantigua’, in J. Alvar et al. (eds.), Héroes, semidioses y daimones (Madrid, 1992) 485-510 at 486-501, regularly refers to the Anguipede as “Abraxas”; G. Quispel, ‘Hermann Hesse and Gnosis’, in J. van Oort (ed.), Gnostica, Judaica, Catholica: Collected Essays of Gilles Quispel (Leiden, 2008) 243-61 at 251, originally published in Gnosis. Festschrift für Hans Jonas (Göttingen, 1978) 492-507.

106

walter shandruk

both Iaô and Abraxas. Even more importantly, only 8 gems in my corpus attest to Abraxas where Iaô is missing. It is, therefore, clear that Iaô is the essential factor that accounts for the high incidence of Abraxas with the Anguipede. The claim that the Anguipede primarily has solar connections has been based on two types of analyses: an analytic breakdown of the components of the Anguipede and an associative argument, that the Anguipede is typically paired with other solar divinities. In the analytical approach, Bonner and Martin Nilsson observe that the rooster has strong solar associations,14 and Bonner (as Michel much later) sees the whip as that of Helios the charioteer.15 Although, Bonner and Nilsson pause at the snake legs and see primarily chthonic associations,16 Erwin Goodenough points out the serpent’s strong solar connections both in Hellenistic and Egyptian contexts.17 I will not deny that the Anguipede can be understood in this analytic manner as a solar deity, but it is striking that the strongest correlations with this figure on gemstones are with elements associated with Iaô. For instance, the Anguipede has a high probabilistic correlation with Sabaôth (p = 0.003393), the second word in the Hebrew bound construction Yahweh Sabaôth, meaning “Lord of Hosts.” When we come to the associative argument, the statistics are even clearer. Bonner’s observation that the Anguipede is often combined with other solar figures has been widely followed,18 but, in fact, none of those associations are statistically significant: Chnoubis (p = 0.999), Cynocephalus (p = 0.813), Harpocrates (p = 0.976), Heliorus (p = 0.501), and Helios (p = 0.685) are all very weakly correlated. With the above result, the solar hypothesis—that the Anguipede should be primarily understood as a solar divinity—is difficult to defend. The solar association is certainly secondary and is best expressed in a few gems where the engraver has deviated from its standard visual depiction. In 15 cases the head is that of another animal (once that of a human in Michigan 93), of which 8 are a lion’s head. The lion’s solar connections have long been 14 Nilsson, ‘The Anguipede of the Magical Amulets’, 62; SMA p. 127. See also Goodenough, Jewish Symbols in the Graeco-Roman Period, 2, 247. 15 SMA p. 128; BM 116. 16 Nilsson, ‘The Anguipede of the Magical Amulets’, 63; SMA p. 128. The chthonic associations are, of course, due to the depictions of the mythical giants as snake-legged from the 4rd century BCE onward. See D. Ogden, Drakon: Dragon Myth and Serpent Cult in the Greek and Roman Worlds (Oxford, 2013) 83-86. Cf. Quispel, ‘Hermann Hesse and Gnosis’, 251. 17 Goodenough, Jewish Symbols in the Graeco-Roman Period, 2, 127-248. 18 SMA pp. 132-33, 282. See Pleše, Poetics of the Gnostic Universe, 188, n. 42; A. Consentino, ‘Il dio anguipede dalla testa di gallo’, in Atti XI Congresso Internazionale di Studi Classici (Athens, 2004) 588-96, esp. 589-90; Nagy, ‘Figuring out the Anguipede’, 169.

the anguipede, its origins and market diffusion

107

recognized19—especially in Egypt, where the animal was closely associated with Horus20—and in our gem corpus it is significantly correlated with Helios (p = 0.0309). There is no doubt that in these few examples, the Anguipede’s solar connections are being emphasized as much as their Yahwistic ones, if not more. But, the leonine alloform of the Anguipede is rare, on only 3% of cases. This leaves the vast majority of Anguipedes emphasizing their Yahwistic association and compels me to focus on the hypotheses positing a Jewish origin or understanding of the icon’s design, namely, the works of Goodenough, Philonenko, and Arpad Nagy. Goodenough’s position is circumspect enough, arguing that the Anguipede—and every other image with which the name Iaô is paired—shows not Yahweh himself, but “some one or other of his powers”.21 He suggests that the Anguipede’s sudden popularity with “no ancestry or development” could only be the result of the figure being invented and given currency in a group that “was most probably a large and important one… whose central interest was in such names as Iao, Iao Sabaoth, Iao Sabaoth Adonai, Michael, and other similar ones, since… these are the names which appear in overwhelming frequency with the anguipede”.22 This group, he argues, must have been Jewish.23 Of the names he cites, Iaô and Sabaôth are, indeed, significantly correlated with the Anguipede, and the suggestion that the image’s sudden popularity is due to its wide currency among Jews, from whose religion magicians so frequently borrowed in antiquity, is tempting but can only be confirmed archaeologically by identifying Anguipede gems in Jewish contexts—something which has not yet happened.24 Arpad Nagy argues that rather than representing an image of the Hebrew God, the Anguipede is the result of plays on the Hebrew root ‫( גבר‬GBR) and

19

SMA pp. 150-51. H.M. Jackson, The Lion Becomes Man: The Gnostic Leontomorphic Creator and the Platonic Tradition (Atlanta, 1985) 112-13. 21 Goodenough, Jewish Symbols in the Graeco-Roman Period, 2, 252. 22 Goodenough, Jewish Symbols in the Graeco-Roman Period, 2, 250. 23 Goodenough, Jewish Symbols in the Graeco-Roman Period, 2, 250. 24 S. Agady et al., ‘Byzantine Shops in the Street of the Monuments at Bet Shean (Scythopolis)’, in L.V. Rutgers (ed.), What Athens has to do with Jerusalem: Essays on Classical, Jewish, and Early Christian Art and Archaeology in Honor of Gideon Foerster (Leuven, 2002) 423-533. In pp. 496-97 they argue that sixth century glass weights found in a complex of five shops in the Street of the Monuments in Beth Shean (Scythopolis) point to a Jewish community, despite the Christian character of some finds; among the finds was an Anguipede gem (501-04). Unfortunately, it was found in shop 3 whereas the weights were found in shop 4. 20

108

walter shandruk

constitutes an image of a epithet of God.25 In Deut. 10:17 one finds ‫( גִ בֹּר‬gibbor), an adjective meaning “strong” or “mighty”, as an epithet of Yahweh. ָ ֽ‫( ג‬gibburah) too is based on this root, meaning The Hebrew noun ‫בוּרה‬ “strength” or “might”, and is often more specifically the “mighty deeds of God” (Ps. 21:14). In Ps. 19:6 (= LXX 18:6) one finds ‫ גִ בּוֹר‬translated by the LXX with the word γίγας (= “giant”, e.g. Is. 3:2, 49:25; Ez. 39:20). The connection here between the Hellenistic depiction of giants with serpent legs, as on the second century BCE Pergamon altar, and the word ‫ גִ בּוֹר‬had already been suggested by Bonner,26 but when one ventures outside scripture to Middle Hebrew1 a tantalizing usage is found, that of ‫ גֶ ֶבר‬meaning “rooster”.27 When this is combined with another typical LXX translation of ‫ גִ בּוֹר‬as μαχατής (= “warrior”, e.g. Jer. 26:12, Zech. 9:13),28 one finds in the root GBR (‫ )גבר‬the three principal elements that distinguish the anguipede: rooster head, the armored torso of a warrior and snake legs—the Anguipede is a visual pun based on an epithet of Yahweh. Nagy’s solution is compelling both for its simplicity (most visual aspects of the Anguipede can be accounted for by a triple pun on a single Hebrew root, GBR) and its consistency with the strong statistical correlations with elements associated with Iaô on magical gems—just like GBR, many of the attributes correlated to Iaô on magical gems can be understood as epithets of Yahweh. If Nagy’s analysis is correct and the Anguipede is simply another (visual) epithet of Yahweh, his line of argumentation offers an interesting 25 Nagy, ‘Figuring out the Anguipede’, 159-72. In two later Jewish mystical texts, the Merkavah Rabbah (8th-10th cent. CE) and Ma’aseh Merkavah (7th-9th cent. CE), “name” and “might” (‫בוּרה‬ ָ ֽ‫ = ג‬gibburah) are equated: “‘He [is] ‘is his name, and his name is ‘[he is]’” (Merkavah Rabbah §655); “His countenance [is] his name, and his name [is] his countenance, and the utterances of his lips [are] his name” (Ma’aseh Merkavah §588); see P. Schäfer, The Hidden and Manifest God: Some Major Themes in Early Jewish Mysticism, trans. Aubrey Pomerance (New York, 1992) 77-81, 97-99. For recensions and dating see J.R. Davila, Hekhalot Literature in Translation: Major Texts of Merkhabah Mysticism (Leiden, 2013) 24748, 302-05. 26 SMA 128. 27 Middle Hebrew is the Hebrew used in the Mishnah and Talmud, further subdivided as Middle Hebrew1 and Middle Hebrew2, with an example of the former being the Hebrew of the Mishnah, i.e., “Mishnaic Hebrew.” Whereas the former was a spoken vernacular, the latter was a written, academic tongue. See W.D. Davies and L. Finkelstein (eds.), The Cambridge History of Judaism: The Hellenistic Age, vol. 2 (Cambridge, 1990) 81-98 and W.D.  Davies and L. Finkelstein (eds.), The Cambridge History of Judaism: The Late Roman-Rabbinic Period, vol. 4 (Cambridge, 2006) 81-98. See also M. Pérez Fernández, An Introductory Grammar of Rabbinic Hebrew (Leiden, 1992) 1-15. 28 H. Kosmala, ‘The Term Geber in the Old Testament and in the Scrolls’, in Congress Volume Rome 1968 (Leiden, 1968) 159-69. In 160 he states: “…a geber is a man who distinguishes himself from others by his strength, or courage, or uprightness, or some other quality”. The extension of this to specifically mean “warrior” is an easy step.

the anguipede, its origins and market diffusion

109

opportunity, and one not taken by him.29 Unlike ‫ גִ בּוֹר‬/ γίγας and ‫ גִ בּוֹר‬/ μαχατής, both rooted in Biblical Hebrew, ‫ גֶ ֶבר‬/ “rooster” entails a lexical development from the Tannaitic period, namely, the first two centuries CE.30 This presents the possibility of positing a terminus post quem for the development of the Anguipede iconography. In the Mishnah, Yoma (hereafter M. Yoma) 1.8, one reads that the ashes from the altar at the temple in Jerusalem were removed at the period of ‫קריאת הגבר‬, “the call of ‫”גֶ ֶבר‬, which can either mean “call of the man/crier” or “rooster crow”. ‫ גֶ ֶבר‬traditionally means “young, strong man” and is cognate with the Aramaic ‫גוברא‬, “man”.31 The word’s development into meaning “rooster” is unsurprising and no doubt a euphemism for the membrum virile,32 a phenomenon not restricted to Hebrew, as seen in the English usage of the term “cock”. Confusion between the two usages is made evident in the debates recounted in the corresponding section in the Babylonian Talmud, Yoma (hereafter T. Yoma) 20b—i.e. whether ‫ הגבר‬refers to “man” or “rooster”. That this confusion should arise in the Babylonian Talmud as to the meaning of ‫ גֶ ֶבר‬suggests that the commentators were too far removed in time and place from the original Sitz im Leben of the Mishnaic phrase ‫ קריאת הגבר‬found in M. Yoma 1.8. A strong hint that the phrase should, in fact, mean “rooster crow” comes from an unlikely place, the Gospel of Mark 13:35, where we find ἀλεκτοροφωνίας (“rooster crow”) as the third night watch. If the watches in M. Yoma 1.8 (= T. Yoma 20b) correspond to the those in Mark 13:35, which, in turn, correspond to the first three of the widely used four-watch system of the Roman period—as seems to be the case33—then ‫ גֶ ֶבר‬in the Mishnaic phrase must mean “rooster” and Jewish practice must have referred to the third 29 He mentions simply that ‫ גֶ ֶבר‬as “rooster” first occurs in a Talmudic context (Nagy, ‘Figuring out the Anguipede’, 166, n. 43), but this could potentially disqualify the meaning from his analysis of the Anguipede since the redactions of the Jerusalem and Babylonian Talmuds reached completion no earlier than the 5th cent. CE. One must demonstrate that the meaning of ‫ גֶ ֶבר‬as “rooster” dates to the earliest stratum of the Talmud, i.e., the Mishnah. 30 Tannaitic refers to the period of the Tannaim, the first two centuries CE during which the traditions of the Mishnah were compiled and redacted around 220 CE by R. Yehudah HaNasi. Mishnaic Hebrew largely reflects the Hebrew spoken at the time. See M.H. Segal, Mišnaic Hebrew and its Relation to Biblical Hebrew and to Aramaic, a Grammatical Study (Oxford, 1908). Hebrew finally died as the vernacular in Palestine around 200 CE. See Ph.S.  Alexander, ‘How did the Rabbis Learn Hebrew?’, in W. Horbury (ed.), From Ezra to Ben Yehudah (Edinburgh, 1999) 71-89 at 75. 31 Goodenough, Jewish Symbols in the Graeco-Roman Period, 2, 252. 32 Kosmala, ‘The Term Geber in the Old Testament and in the Scrolls’, 159; J.P. Peters, ‘The Cock’, Journal of the American Oriental Society 33 (1913) 363-93 at 366. 33 M. Rose, Jahwe: Zum Streit um den alttestamentlichen Gottesnamen (Zürich, 1978) 685-701 at 694-96.

110

walter shandruk

watch as “rooster crow” at least as far back as composition of Mark in the second half of the first century CE.34 Since the tractate M. Yoma is concerned with the maintenance of the temple cult in Jerusalem, it constitutes one of the earliest strata of the Mishnah and may date well before the temple was destroyed in 70 CE.35 However, there are reasons to suspect that the exact formulations of M. Yoma do not date before 70 CE,36 therefore, I will maintain a conservative estimate and posit the end of the first century as the terminus post quem for the meaning of ‫ גֶ ֶבר‬as “rooster” and, consequently, the possibility for the clever punning Nagy’s hypothesis requires. A late first century terminus post quem is further substantiated by the fact that not a single Anguipede gem has been securely dated to the first century or earlier (although a number are given a date range of first to second century CE, which simply reflects the imprecision that plagues the dating of these objects), although this does not preclude the possibility that the earliest Anguipede gems may date a bit before the end of the first century CE. 2. Market Diffusion For a figure that came into existence only at the beginning of the second century CE, or perhaps a bit earlier, it is a fascinating development how quickly it spread on the medium of inscribed gemstones. The Anguipede is the third most common icon on magical gems, showing up on 10% of them, but when the other two more common visual elements are compared (the star and snake), the Anguipede is more complex and the most common visual element that usually consists of the main visual device on a gem (258). The most common icon is that of the star, and the second most common that of the snake, whether in the form of the Ouroboros or some other configuration. Both the star and the snake are typically supplementary

34 The debate concerning the exact date of the Gospel of Mark is too complex to engage here and unnecessary. General consensus agrees that the gospel likely dates sometime between 65 and 75 CE, but opinions are split roughly half and half whether the gospel should be dated before the destruction of the temple (70 CE) or after. See A. Winn, The Purpose of Mark’s Gospel: An Early Christian Response to Roman Imperial Propaganda (Tübingen, 2008) 43-91 at 56, nn. 32-33 for bibliography on pre/post-70 CE debate. 35 H. Danby, The Mishnah: Translated from the Hebrew with Introduction and Brief Explanatory Notes (Oxford, 2012) xxii. 36 D. Stökl Ben Ezra, The Impact of Yom Kippur on Early Christianity: The Day of Atonement from Second Temple Judaism to the Fifth Century (Tübingen, 2003) 19-28; L. Ginzberg, ‘Tamid: The Oldest Treatise of the Mishnah’, Journal of Jewish Lore and Philosophy 1.1 (1919) 265-95 at 266, n. 66.

the anguipede, its origins and market diffusion

111

visual elements surrounding or accompanying a more central figure or set of figures. This is not so of the Anguipede. When the Anguipede appears on gems it is typically the only icon on that side of the gem, usually on the obverse. From that perspective, it is the most common central visual device found on inscribed magical gemstones. Given its popularity on gems, the figure is depicted with startling rarity on other media. Andrew T. Wilburn identifies the figure drawn in PGM XXXVI 231-55 as “a rooster-headed, cuirassed divinity, the anguipede”,37 although it is missing the very feature that would literally make it an “anguipede”, i.e. snake legs, instead having human legs.38 Of course, he is not alone in identifying a rooster-headed, cuirassed figure with human legs as an “Anguipede”—Michel does the very same with BM 230-232. The variant is rare, if it is to be considered as such.39 A more tantalizing figure is found in the Roman-German museum at Mainz, a bronze figurine 14.6 cm tall, depicting the standard Anguipede, originally from the early nineteenth century private collection of Louis Levade, claimed to have been found in the region of ancient Aventicum.40 As for its authenticity, Leibundgut warns caution due to the otherwise absence of three-dimensional Anguipedes and the suspect quality of Levade’s collection, but he also makes clear that the patina and style are not suspicious and there is no overt evidence of it being an early modern forgery.41 One more example was discovered in the Santa Prisca mithraeum, a partial base of a snakelegged statue, and included in Vermaseren’s magisterial corpus of Mithraic monuments and inscriptions. The august scholar labeled it as a Giant, but 37 A.T. Wilburn, Materia Magica: The Archaeology of Magic in Roman Egypt, Cyprus and Spain (Ann Arbor, 2005) 148. 38 Piranomonte, ‘Religion and Magic at Rome’, 209, doubts that these images are related to the Anguipede. 39 It is unclear why Wilburn doesn’t make a similar identification for the figure in PGM XXXVI 69-101 since there one finds a human figure with rooster head, whip (something missing in the previous figure) and what too may be a cuirass (although, the figure is too stylized to be certain). 40 A. Leibundgut, Die Römischen Bronzen der Schweiz II (Mainz, 1976) 38-39, plates 21-22. It is the only bronze Anguipede listed in M. Le Glay, s.v. ‘Abraxas’, in LIMC I (Zürich, 1981) 3. 41 “Da die Patina und der Stil nicht verdächtig sind und auch keine eindeutigen Beweise gegen die Echtheit vorgebracht werden können, muß die Frage vorläufig offen bleiben”, Leibundgut, Die Römischen Bronzen der Schweiz II, 39. Leibundgut laments that this figure has never been cited in all of the rich literature on “Abraxas”. Since Leibundgut’s publication in 1976, the figurine has been cited in a number of publications. See, for example, Consentino, ‘Il dio anguipede dalla testa di gallo’, 590; Marco Simón, ‘Abraxas. Magia y religión en la Hispania tardoantigua’, 486; M. Henig and A. King (eds.), Pagan Gods and Shrines of the Roman Empire (Oxford, 1986) 165.

112

walter shandruk

in fact it seems to be that of our Anguipede.42 In addition to snake legs, a short tunic and the trace of a shield are present. Equally interesting is an Anguipede reported by James Wiseman inscribed on a lead defixio (one of four), found in the “Fountain of the Lamps” in the gymnasium area at Corinth.43 He describes it as having the head and torso of a human and holding a sword in its right hand and snake-wrapped staff in its left. According to Wiseman, the bath became a cult center after the ceiling collapsed in the fourth century, at which point, and until the sixth century, thousands of lamps were deposited.44 Since by this period the chamber was flooded it is possible that they had been deposited before the ceiling collapse, although, there is no way to be certain. But this figure also deviates from the canonical depiction of the Anguipede by having a human head instead of a rooster head. This is a rare variant found on only one gem example that I know of (SMA 180). The few examples of the Anguipede outside of magical gems cited above—and I am not fully convinced the figure on PGM XXXVI should be understood as an Anguipede—it is, at the very least, far from the canonical depiction—these examples are the exceptions that prove the rule: the most popular central device on magical gems, especially in its canonical form, is found virtually nowhere else in Graeco-Roman material culture. Why might this be the case? This problem is accompanied by the related fact that the Anguipede, as far as I know, is nowhere described in any ancient literary source. It does not appear in any of the ancient Greek lapidaries—texts concerned with the magico-medicinal uses of stones—that survive in varying states of completeness (The Orphic Lapidarium, the Orphic Kerigmata, Socrates-Dionysos, Damigeron-Evax, etc). It may be that in this lies a clue as to the use of Anguipede gems, that they had no medicinal valence to them but were rather used as amulets protecting against people or 42 M.J. Vermaseren, Corpus Inscriptionum et Monumentorum Religionis Mithriacae, vol. 1 (The Hague, 1956) 200.491 and pl. 142. Only the snake legs, short tunic and trace of shield on his left arm remain. Vermaseren identifies him as a “Gigant”, but tunic and shield are not typical accoutrements of Giants in Hellenistic and Imperial art. J.R. Harris, ‘Iconography and Context: ab oriente ad occidentem’, in Henig and King, Pagan Gods and Shrines of the Roman Empire, 171-78 at 176, n. 34, is convinced that this must be an anguipede. 43 J. Wiseman, ‘The Gymnasium Area at Corinth, 1969-1970’, Hesperia 41.1 (1972) 1-42 at 33. See brief discussions in R. Cline, Ancient Angels: Conceptualizing Angeloi in the Roman Empire (Leiden, 2011) 121 and A. Karivieri, ‘Magic and Syncretic Religious Culture in the East’, in D.M. Gwynn and S. Bangert (eds.), Religious Diversity in Late Antiquity (Leiden, 2010) 399-434 at 419. 44 Wiseman, ‘The Gymnasium Area at Corinth’, 27. The defixio in question, MF-69-114, was found “between the basins and the bench, but below the level of the bench seat”. Wiseman, ‘The Gymnasium Area at Corinth’, 33.

the anguipede, its origins and market diffusion

113

supernatural forces outside the context of illness. It is nowhere described in magical handbooks from Late Antiquity, PGM or otherwise. In thinking about this problem, I wondered if treating Anguipede gems as commercial products might help offer an angle of analysis that may point at a possible solution. They were, after all, precisely this. The gem itself had to be inscribed by a professional gem engraver, which had its own cost, and the design of the gem in terms of icon placement, accompanying inscriptions, charaktēres, and so on, was likely the product of a professional sorcerer, whose services would also not be free. Perhaps, we can understand the near universal restriction of the Anguipede figure to gemstones as a result of an anomaly in its target market. One of the most successful marketing models was published in 1969 by Frank Bass and the equation he published is known as the Bass Diffusion Model.45 It typically is depicted as a figure-S curve (a cumulative density function), showing the cumulative adoption of a new commercial product (Illustration 2). The S figure is distinctive in that it shows that the majority of commercial adoption happens over a shorter period of time, between initial adoption and full market saturation. That is to say, there is a narrow window in which word-of-mouth and advertisement show rapid gains, but this only after a longer period of initial adoption. The Bass Model equation is as follows:

There are four variables in the Bass equation. The first is t, time. Next is p, which is based on the influence of the original innovators or, perhaps, the influence of initial advertising, called the coefficient of innovation. This p should not be confused with the “p-values” used earlier. The third variable, q, is the rate at which word-of-mouth spreads in favor of the product, or the coefficient of imitation.46 The final variable is M, the size of the potential market, which is to say, the final number of product adopters. A successful product will usually have a rate for word-of-mouth greater than the influence of initial advertising or the influence of the initial innovators/ adopters of the product.

45 F.M. Bass, ‘A New Product Growth for Model Durables’, Management Science 15.5 (1969) 215-27. See also F.M. Bass, ‘Comments on “A New Product Growth for Model Durables”: The Bass Model’, Management Science 50.12 (2004) 1833-40. 46 Bass, ‘A New Product Growth for Model Durables’, 217.

114

walter shandruk

Illustration 2. Cumulative Bass Diffusion.

Illustration 3. Distribution of Gems.

Illustration 4. Percentage Anguipede.

Illustration 5. Cumulative Percentage.

the anguipede, its origins and market diffusion

115

In order to make an informed guess as to where in the adoption process the Anguipede may have been, we need to normalize the number of Anguipede gems in each century by the total number of known gems from that century. Illustration 3 shows the gem frequency by century and Illustration 4 plots the number of Anguipede gems divided by the total number of gems. What we observe is that the percentage of gems having the Anguipede peaks in the third century and then gradually declines. During the rapid decline in gem production after the third century, no doubt due to the rapid Christianization of the Roman Empire starting in the fourth century, the percentage of gems found with the Anguipede also falls rapidly. Had interest continued in this particular gem-type one would expect that its percentage would not decline much or, perhaps, even continue to grow. Rather, it seems that the Anguipede was near the end of its adoption process. Illustration 5 shows the cumulative percentage for Anguipede gems. The curve resembles the middle and upper parts of the S-figure seen in Illustration 2, suggesting we are at the tail end of market penetration. But, this does not account for why during the initial upsurge in adoption the Anguipede figure did not, for the most part, escape the world of inscribed gemstones. I would like to suggest that the Anguipede figure was designed by someone closely associated with the design or production of magical gems, perhaps a Hellenistic Jew or a magician with a strong interest in Jewish religious motifs and language, and first had this figure appear on magical gems. It increased in popularity through the third century, but unlike nearly every other iconographic element on gems, it did not have some antecedent in Graeco-Roman or Egyptian visual culture. This would have made the diffusion of the Anguipede into other visual media more difficult. In addition, if the greatest interest in this gem-type were among those with a curiosity for Jewish magic then this potential market would be significantly smaller than the market for Graeco-Egyptian gems in general and would saturate more quickly. Early saturation within a niche market could effectively type-cast the figure of the Anguipede in popular perception. Finally, this association with Judaism, whether perceived or real, may have become a liability as the Roman Empire became Christianized, thus further constraining the market for these gems and the spread of the Anguipede to other visual media.

VIII.

HEATHEN SERPENTS AND WINGLESS ANGELS? SOME NOTES ON IMAGES IN COPTIC MAGICAL TEXTS1 Korshi Dosoo

The corpus of Coptic magical texts forms an important part of the history of religion in Egypt, linking the private ritual practices of the Greek and Demotic magical papyri with the later Arabic and Hebrew manuscripts of the Cairo Genizah. They are not simply ‘pagan survivals’, but rather represent a creative tradition which sought to negotiate a Christianised cosmos, at times drawing upon older ritual practices and deities, at times drawing upon and extending the logic of Christian prayers and sacraments. Coptic magical texts should properly be considered alongside contemporary Greek and Arabic material, although the Coptic material does seem to be the most abundant in the period under consideration – broadly the fifth to eleventh centuries CE – and the most consistent in its use of images. For this reason, I will focus in this discussion on magical images in Coptic manuscripts, considering material in other languages, and indeed from outside Egypt, only where relevant to illuminating or contextualising these central examples.

1 I would like to offer my sincere thanks to Raquel Martín Hernández for inviting me to contribute to this volume, and giving me the opportunity to reflect upon the fascinating topic of magical images; to Edward O.D. Love, whose correspondence and comments helped me to formulate many of the ideas expressed here; and to Andrea Jördens and Elke Fuchs of the University of Heidelberg Papyrussammlung for giving me access to images from their collection and permission to reproduce them. This article was produced as part of the project The Coptic Magical Papyri: Vernacular Religion in Late Roman and Early Islamic Egypt at the University of Würzburg. Throughout this article, the initial mention of each manuscript will be accompanied by its reference number in the Trismegistos database (www.trismegistos.org = TM#) and its likely date; where no Trismegistos number exists, the primary publication, or image source, will be given in its place. Citations from Greek texts, except where otherwise noted, refer to the editions published online at the Thesaurus Linguae Graecae (stephanus.tlg.uci.edu). Magical gems will be referred to where possible by their CBd number.

118

korshi dosoo

Images are found in many magical manuscripts, but their presence is inconsistent; some important bodies of material – these include the magical archive held at the University of Michigan2 and the large codex known as Anastasy 93 – do not to contain any, while others – such as the manuscripts of the Heidelberg Library4 – offer particularly rich figural material. As with the images of the Graeco-Egyptian magical tradition, those from Christian Egypt are yet to be fully studied, although they are often briefly discussed in the original editions of the texts to which they belong, and more extensive studies have begun to appear with increasing frequency.5 These studies have generally not been systematic, however, considering only narrow categories of texts – those from single collections, or with particular figural content – and so much still remains to be done. For this reason, I will not attempt to offer here a comprehensive overview or typology of images, but rather a series of observations, corrections, and tentative 2 The Michigan archive is the so-called “Wizard’s Hoard” (5th-7th cent. CE), consisting of P.Mich. inv. 593-603 and 1294. Although it has not yet been fully published, images of its constituent texts may be found online on Advanced Papyrological Information System (APIS), and a description and list of the texts may be found at quod.lib.umich.edu/a/ apis/x-3648/1 (12/04/2021). For existing publications of P.Mich. inv. 593 (TM 100021), see W.H. Worrell, ‘A Coptic Wizard’s Hoard’, The Journal of Semitic Languages and Literatures 46.4 (1930) 239-62; P. Mirecki, ‘The Coptic Wizard’s Hoard’, Harvard Theological Review 87 (1994) 435-60; M. Meyer and R. Smith, Ancient Christian Magic: Coptic Texts of Ritual Power (San Francisco, 1994) 293-310. Small tableaux containing charaktēres appear in P.Mich. inv. 595 verso and inv. 600 recto (KYP M15), but there does not seem to be any figural content. 3 TM 100023, 6th cent. CE. 4 TM Archive ID 421, 10 th-11th cent. CE. 5 A. Kropp, Ausgewählte koptische Zaubertexte (Brussels, 1931) vol.3, 208-16; I. Grumach, ‘On the History of a Coptic figura magica’, in D.H. Samuel (ed.), Proceedings of the Twelfth International Congress of Papyrology (Toronto, 1970) 169-81; U. Horak, ‘Christliches und Christlich-Magisches auf illuminierten Papyri, Pergamenten, Papieren und Ostraka’, Mitteilungen zur christlichen Archäologie 1 (1995) 27-48; J. Johnston, ‘Prolegomena to Considering Drawings of Spirit-Beings in Mandaean, Gnostic and Ancient Magical Texts’, ARAM 22 (2010) 573-82; T. Mößner and C. Nauerth, ‘Koptische Texte und ihre Bilder’, in A. Jördens (ed.), Ägyptische Magie und ihre Umwelt (Wiesbaden, 2015) 302-73; K. Dosoo, ‘Zōdion and Praxis: An Illustrated Coptic Magical Papyrus in the Macquarie University Collection’, Journal of Coptic Studies 19 (2018) 11-56; S. Kiyanrad et al. (eds.), Bild und Schrift auf ‘magischen’ Artefakten (Berlin and Boston, 2018); I. Gardner and J. Johnston, ‘“I, Deacon Iohannes, Servant Of Michael”: A New Look at P. Heid. Inv. Kopt. 682 and a Possible Context for the Heidelberg Magical Archive’, Journal of Coptic Studies 21 (2019) 29-61 at 51-53; E.O. Love, ‘Apa Baula and The Destroyer: The Embedding of Efficacy in Figural Amulets from Roman and Early Islamic Period Egypt’, in K. Dosoo and J.-C. Coulon (eds.), Magikon Zōon: Animal et magie | The Animal in Magic (forthcoming); E.O. Love, ‘“Crum’s Chicken”: Demonised Donkeys in the Context of the Demotic and Greek Magical Papyri and the Tradition of Coptic Magic’, in Dosoo and Coulon, Magikon Zōon (forthcoming); J. Johnston et al. (eds.), Drawing Spirit: The Role of Image and Design in the Magical Practice of Late Antiquity (forthcoming).

heathen serpents and wingless angels

119

conclusions, based on a study of several published and unpublished Coptic magical texts.6 This discussion will not be chronological; while it might be desirable to provide a historical narrative of the evolution of particular image types, this does not yet seem possible – we often have few examples of particular image constructions, and these samples are often imperfectly dated due to the well-known problems of Coptic palaeography. But indeed, even if dating was well-established, it is not clear that later images would necessarily demonstrate more derived forms; the transmission of magical material is likely to have been highly informal and haphazard, resulting in more and less consistent lines of transmission. There are many questions that we can ask when confronted with an image in a magical text: how was it produced? What does it depict? What does it do (ritually, socially, metaphysically, phenomenologically)? What is its relationship to the text to which it belongs, and to the wider figural culture? How does it relate spatially to its support and to other elements of the manuscript of which it is part? I will try to address some of these questions here, although they will not be answerable or relevant in every case. I will not generally comment on the quality of the images, although I will occasionally note instances where the copyist appears to have incorrectly reproduced an image or element, or seems to have failed to produce their desired outcome. Such judgements are, of course, problematic – can a modern historian claim to understand an image better than an ancient practitioner of magic? – but there are a few instances where it seems clear that copyists were unhappy with their own capabilities. These include P.Würzburg inv. 42,7 where an initial attempt to produce a complex tableau was abandoned, and the sheet was flipped to allow them to start again, while on the tableau depicting the archangel Michael on fol.  9 of P.Heid.Inv.Kopt. 686 (Fig. 2),8 the initial outline of the left wing9

6 For the most complete list to date of Coptic magical texts, see R. Bélanger Sarrazin, ‘Catalogue des textes magiques coptes’, Archiv für Papyrusforschung 63.2 (2017) 367-408. 7 TM 99598, 10 th cent. CE. For an image of both sides of this manuscript, see the plate accompanying its initial publication in W. Brunsch, ‘Ein koptischer Bindezauber’, Enchoria 8 (1978) 151-57. Note that almost identical tableaux may be found on P.Stras.Inv.Kopt. 550 (KYP M106, 10th-11th cent. CE) accompanying the inverted recto text, edited in D. Tibet, ‘A Magical Request for Revelation’, in A. Boud’hors et al. (eds.), Coptica Argentoratensia (Paris, 2014) 133-41 and on Berlin P. 20911 (TM 107307, 6th-7th CE), albeit very faded. 8 TM 100022, 10 th cent. CE. 9 References to right and left here will be made from the perspective of the figures within tableaux (proper right and left), so they must be reversed when looking at such compositions.

120

Fig. 1. Image of an angel (?) from P.Strasbourg K 204 + 205 (205 frag. C). Image created by K. Dosoo.10

korshi dosoo

Fig. 2. P.Heid.Inv.Kopt. 686, fol. 9r; tableau depicting the Archangel Michael accompanied by two Powers (ⲧⲉⲛⲁⲙⲓⲥ = δύναμις), trampling two demonic figures. Image created by K. Dosoo.

10

was partially erased and redrawn, perhaps to make space for a caption.11 These copying errors are as important as the irregular orthographies and mistakes in the accompanying text, giving us information about the training and experience of those who produced these documents. These copying errors should not be confused with the stylistic tendencies of the images, which are often highly geometric and even, to a modern eye, abstract. As Jay Johnston has noted, the aim of these images was not merely representational; beings such as angels and demons were known to be 10 Published in K. Hevesi, ‘P. Stras. K 204 and K 205 An Unpublished Coptic Magical Collection from the Bibliothèque nationale et universitaire de Strasbourg’, in S. Kiyanrad et al. (eds.), Bild und Schrift auf ‘magischen’ Artefakten (Berlin and Boston, 2018) 49-118. See p. 32 for a brief discussion of the image. 11 Mößner and Nauerth, ‘Koptische Texte und ihre Bilder’, 349, wrongly I think, interpret the second outline of the left wing as “a sort of tubing” (“eine Art von Schläuchen”) which connects the Power on the right to Michael. This seems an unnecessarily strange interpretation; as discussed below, wings are the standard attributes of angels, and it seems clear from the (imperfectly) mirrored image of the Power on the right that the copyist intended simply to have the Power overlapping the wing of Michael; the “tube” (discussed below as the covert of the wing) only passes behind the Power’s head by an accident of positioning.

heathen serpents and wingless angels

121

bodiless and invisible, and thus the use of stylisation is not simply a mark of poor technical skill, but rather an indicator which signals the “otherworldly status” and “alterity” of the spirits caught in the lines of ink.12 Figural images in Coptic manuscripts usually occur alongside nonfigural elements – text, which may be either semantically transparent, or the usually incomprehensible voces magicae (magical names), and charaktēres, that is, symbols, often resembling letters or simple geometric shapes, usually decorated with small circles at their terminal points.13 Together, figural and textual elements and charaktēres may be understood as forming tableaux, which may in turn be interpreted as particular “image constructions”, with specific ritual meanings and functions. The use of charaktēres in Coptic magical practice is particularly innovative; in earlier practice they seem to be understood as representing a divine or sacred language whose meaning is not immediately accessible14 – that is, from an etic point of view they serve a symbolic rather than representational function, indicating the presence of divine or demonic power without containing a specific meaning.15 The symbolic function of charaktēres means that it is usually impossible to know how they were interpreted within tableaux, and so I will only discuss them in detail in a few instances. In Coptic texts, however, we sometimes see charaktēres with the forms of letters used to spell

12 See Johnston, ‘Prolegomena to Considering Drawings of Spirit-Beings’, 574. For a discussion of the concept of angels as bodiless (ἀσώματος) and immaterial (ἄϋλος) beings formed from spirit (πνεῦμα) and fire (πῦρ), see G. Peers, Subtle Bodies: Representing Angels in Byzantium (Berkeley, Los Angeles and London, 2001), esp. 1-2, 16-19, 165. The description of angels as being formed from spirit (or wind) and fire derives from Psalm 104.4 and Hebrews 1.7. 13 For discussions of the phenomena of charaktēres, see R. Gordon, ‘Signa nova et inaudita: The Theory and Practice of Invented Signs (charaktêres) in Graeco-Egyptian Magical Texts’, MHNH 11 (2011) 15-44; K. Dzwiza, Schriftverwendung in antiker Ritualpraxis (unpublished doctoral dissertation, Erfurt, Heidelberg, 2013). As Gordon notes, charaktēres usually seem to be derived from graphemes (Greek, Latin or Coptic), or from simple geometric shapes. 14 See, for example, PDM xiv 1074; P.Heid.Inv.Kopt. 682 ll. 1-2 (TM 99576, 967 CE), in which charaktēres are referred to as “names” (rn / ⲣⲁⲛ). Gardner and Johnston, (‘I, Deacon Iohannes, Servant Of Michael’, 40-43) suggest that the charaktēres beside ll. 1-5 form a “magic square” used in parallel with the charaktēres below l. 42 to generate names by a process known from early modern European magic, but it seems to me more parsimonious to assume that, as in the earlier parallel, the charaktēres themselves are understood to be names written in a divine writing system. 15 There are exceptions to this general rule; certain shapes which may be classed as charaktēres do have a recoverable meaning, such as the symbol of the decan Knm.t (SSS), and others derived from hieroglyphs. We cannot be certain, of course, that those who used these symbols were aware of their meaning, or that other signs did not have conventional or contextual meanings which cannot be recovered.

122

korshi dosoo

out names, words, and even phrases,16 as if they were simply a variant of normal script, while figural elements – staffs and other weapons, facial features, and even entire human and animal bodies – may be transformed into charaktēres through stylisation and the addition of the characteristic terminal circles. The symbolic function of charaktēres remains, but exists alongside a new function, that of adding a symbolic charge to linguistically or visually meaningful content.17 This stylisation creates a problem for interpretation; our ability to “read” images depends to a great extent upon their iconicity, their resemblance (within culturally determined representational norms) to the object they depict, and highly stylised images, with their rejection of iconicity, are less easily read. For this reason, most of the images I discuss here will be those which tend more towards straightforward representation; reliable identification of the more stylised images will depend upon careful work which draws upon the visual language of more certain examples. 1. The Survival of Images from Graeco-Egyptian Magic It is important not to overstate the dependence of Christian magic on the preceding Graeco-Egyptian tradition; while there are certainly elements which carry over from the predominantly Greek and Demotic tradition of the magical papyri of the first to fourth centuries into the predominantly Coptic and Greek manuscripts of the Christian period (understood here as, roughly, the fifth to eleventh centuries), there are numerous differences. In terms of content and style, the earlier Graeco-Egyptian material is extremely diverse, displaying relatively little repetition in surviving images;18 the Coptic material is, in comparison, more homogeneous. Of the image constructions which seem to be transferred from the earlier tradition, the clearest is the ouroboros, the snake biting its tail. This symbol is first attested as a depiction of the serpent Mehen protecting the 16 A dramatic example of this may be found in P.Heid.Inv.Kopt. 520 (TM 102081, 9 th-11th cent. CE), where the sequence ϫⲱ‹ⲉⲓⲥ› ⲡⲛ‹ⲟⲩ›ⲧⲉ ⲉ{ⲩ}‹ⲛⲁⲛ›ⲟ‹ⲩϥ› ⲕⲱ ‹ⲛ›ⲁⲓ ⲉⲃⲟⲗ (“Lord God who is good, forgive me”, a phrase which appears in l.1 of the same text) may be read into the groups in ll. 5-6. 17 Compare the comments of J. Sanzo, ‘The Innovative Use of Biblical Traditions for Ritual Power: The Crucifixion of Jesus on a Coptic Exorcistic Spell (Brit. Lib. Or. 6796[4], 6796) as a Test Case’, Archiv für Religionsgeschichte 16.1 (2014) 79-82. 18 Among the relatively rare attested examples of repeating tableaux in Greek magical texts are the wedjat eye which appears in PGM III 415 (TM 64511, 3rd cent. CE) and PGM V 81 (TM 64368, 4th cent. CE), as well as the repetitive, though not identical, figures which appear in PGM XXXVI (TM 64479, 4th cent. CE).

heathen serpents and wingless angels

123

Fig. 3. PGM VII 593; lion-headed ouroboros used in a body-amulet. Image created by R. Martín Hernández.

unified Re-Osiris on the Second Shrine of Tutankhamun,19 but most early examples are found in the Roman period, where it was associated with magical and alchemical material.20 Within the Graeco-Egyptian magical material, the ouroboros is found in the fourth century PGM VII,21 intended to be copied onto a smaller (metal or papyrus) support to create an amulet to protect the body22 (φυλακτήριον σωματοφύλαξ, ll. 579-90). The energetic 19 D.M. Reemes, The Egyptian Ouroboros: An Iconological and Theological Study (unpublished doctoral dissertation, University of California, Los Angeles, 2015) 137-43; J.C. Darnell, The Enigmatic Netherworld Books of the Solar-Osirian Unity (Fribourg, 2004) 380-83. As Reemes points out (p. 118), an earlier image of Many-of-Faces (ꜥšꜣ-ḥr.w), an analogue of Mḥn, in an ouroboros-like (“ouroboroid”) form may be found in a depiction from the Amduat from the tomb of Tuthmosis I (p. 117). For an example of a XXth Dynasty amulet bearing an ouroboros see Y. Koenig, ‘Le contre-envoûtement de Ta-i.di-Imen. Pap. Deir el-Medina 44’, Bulletin de l’Institut français d’archéologie orientale 99 (1999) 259-81. 20 While the ouroboros is not attested in the few surviving alchemical papyri, it appears on fol. 188v of MS St. Mark 299 (10th-11th cent. CE), and on fol. 196r and 279r of MS Parisinus graecus 2327 (gallica.bnf.fr/ark:/12148/btv1b10723905w; 15th cent. CE). The first of these appears within Cleopatra’s Chrysopeia (1st-3rd cent. CE), while the text accompanying the other depictions cannot be attributed to a known author. The ouroboros is mentioned by (pseudo-)Olympiodorus (Εἰς τὸ κατ’ ἐνέργειαν ζωσίμου ὅσα ἀπὸ Ἑρμοῦ καὶ τῶν φιλοσόφων ἦσαν εἰρημένα (= De arte sacra) vol.2 p. 80 ll. 1-12; 5th/6th CE?) and Theophrastus (Περὶ τῆς θείας τέχνης, 144; 7th cent. CE?). All of this suggests that the ouroboros may have been used as a symbol within alchemy at a period contemporaneous with its appearance in the Greek magical tradition. 21 TM 60204, 4th cent. CE. See also the image of an ouroboros on the applied fever amulet PGM CVI (TM 64270, 3rd-4th cent. CE). On the image of the ouroboros in magical formularies and gemstones see Chapter VI in this volume. 22 For a different interpretation of the reading see Chapter VI in this volume.

124

korshi dosoo

style of drawing, using short, thin lines, is typical of many of those in the Greek magical texts, and there is an attempt to capture some naturalistic detail (Fig. 3): the snake is black, except for its belly, whose long scales are accurately depicted at right-angles to its body, and there is even an (inconsistent) attempt to suggest its rounded form by curving some of the scales. Its head seems to be that of a lion – the mane depicted as a series of dashed lines, its nose and upper lip distinguished as two separate protrusions, and a third indicating the lower jaw. The face is squashed – perhaps inadvertently – so that the ear and eye are far closer together than we might expect, but all of the key elements are there, and the style corresponds broadly with the more expert image of a lion head on the bier depicted on the slightly earlier PDM xii.23 The depiction of the serpent as lion-headed identifies it as an instance of the deity Knephis,24 whose name appears as the first of a series of voces magicae within its body. While this is the only known example of an ouroboros on Greek magical papyrus, numerous instances are known from the contemporary magical gems.25 The examples from the Coptic tradition are later, being dated to between the sixth and eleventh centuries, but it is notable that they seem to retain their amuletic function. PGM XLVIII is a defensive amulet,26 while P.Berlin inv. 15990 is an amulet asking for favour for its wearer.27 The function of P.Moen. 3 (Fig. 4),28 part of a small formulary, is uncertain; it may relate to one of the other texts (a business-procedure and an invocation), or may rather be a depiction of an independent amulet. A fourth example is thus far unpublished.29 Each of these examples shares commonalities with the earlier example – the insides of their coils are filled with voces magicae and charaktēres, which often also fill the space around them, while P.Berlin inv. 15990 even 23

TM 55954, 3rd cent. CE. The depiction of the lion bier is found in col.1. For a discussion of this deity, see D. Klotz, Caesar in the City of Amun: Egyptian Temple Construction and Theology in Roman Thebes (Turnhout, 2012) 133-42; H.J. Thissen, ‘Κμηφ: Ein verkannter Gott’, ZPE 112 (1996) 153-60. While Thissen is certainly right that Κνηφ(ις), the demiurge Km-ꜣ.t⸗f, and Χνουμις, the decan Knm.t, are usually distinguished from one another, the fact that the representation of Knephis in PGM VII has a lion rather than falcon head suggests that it has been influenced by the iconography of the decan, that of a lion-headed serpent. 25 See for example the instances described in A. Delatte and P. Derchain, Les intailles magiques gréco-égyptiennes (Paris, 1964), esp. 48-49, and chap. 8 in this volume. 26 TM 89309, 10 th-11th cent. CE. 27 TM 108896, 6th-8th cent. CE. 28 TM 100007, 9 th-11th cent. CE. 29 P.Monts.Roca inv. 1231 (part of TM 126174 = KYP M30, 7th-9 th CE), a formulary on a rotulus formed from stitching together pages from a codex containing II Samuel. 24

heathen serpents and wingless angels

125

Fig. 4. Images of ouroboroi from Coptic magical texts; PGM XLVIII (top left); P.Berlin 15990 (top right); P.Moen 3 (bottom). Images created by K. Dosoo.

has a smaller, erect serpent within its body. Stylistically, though, they are rather different from that of PGM VII: they all have snake heads, although P.Berlin inv. 15990 has small horns, and PGM XLVIII seems to have something like a crest and beard. They are less naturalistic, too, with the entire bodies of P.Berlin inv. 15990 and PGM XLVIII departing from naturalism by extending the striations which were restricted to the belly of the example from PGM VII to the whole body, and P.Moen. 3 presenting a much simpler, undecorated form. Less clear, if still probable, is the transmission of depictions of SethTyphon. This god appears on a number of early magical texts from Egypt, as well as several curse tablets from outside it, notably Rome, as an anthropomorphic figure with an equine (classically donkey) head, and wielding in one hand a weapon with which he menaces other figures in his tableau.30 30 The clearest depiction of the donkey-headed Seth-Typhon on a magical text is in PDM xii col.4 (TM 55954, 3rd cent. CE), in which he has the name ⲥⲏⲑ written on his chest, and a clearly asinine head and mane. The identification of the animal-headed figures in PGM III 67 and the Sethian curse tablets (SV 1, 6, 9, 12, 16, 20 A & B, 29) has proven controversial, principally due to Preisendanz’s preference for understanding the figure as a horse-demon (Akephalos: der kopflose Gott [Leipzig, 1926]). Horse-demons are not well-attested in magical literature however, and the equine-headed figure is regularly accompanied by the name

126

korshi dosoo

Fig. 5. Image from PGM III 67 depicting Seth-Typhon with the head of an equine, wielding a whip. Image created by R. Martín Hernández.

Similar figures with animal heads, often wielding weapons, are found on a few Coptic texts, although damage to their support often makes it difficult to be certain of their full original form; in a few cases identification is made more likely by the presence of the typhonikos logos, a series of voces magicae associated with Seth-Typhon.31 As might be expected of this ambivalent, violent god, his appearances in the earlier magical tradition are and epithets of Seth. For a fuller discussion of this question, see Chapter III in this volume and Love, ‘Crum’s Chicken’. 31 On the typhonikos logos and images of Seth-Typhon, see R. Martín Hernández, ‘More than a logos. The ιωερβηθ logos in context’, in C. Sánchez Natalias (ed.), Litterae Magicae. Studies in Honour of Roger Tomlin (Zaragoza, 2019) 187-209; Love, ‘Crum’s Chicken’. Apart from P.Berlin 8323 (TM 108884, 7th-10th CE) and Crum 23.2.2 (published in Love, ‘Crum’s Chicken’; KYP M570, 9th-10th CE), discussed below, I consider the most certain depictions to be P.Heid.Inv.Kopt. 473 (TM 102083, 9th-11th CE) which contains the typhonikos logos and a fragment of an animal head, and Naqlun N. 44/95 (KYP M65, 5th-6th CE; discussed in J. van der Vliet, ‘Les Anges du Soleil’, Études Coptes 7 [2000] 319-37), probably a separation spell, which contains a fragmentary, animal-headed figure. P.Monts.Roca inv. 1472 (KYP M22, 6th-9th CE) contains the typhonikos logos associated with Seth, and the morphology of the two animal figures may be influenced by the depiction of Seth, but the overall construction implies to me that the image is intended to represent two individuals being separated; see Dosoo, ‘Zōdion and Praxis’. Cf. the editio princeps in R. Martín Hernández and S. Torallas Tovar, ‘A Magical Spell on an Ostracon at the Abbey of Montserrat’, ZPE 189 (2014) 17584 and Chapter III in this volume.

heathen serpents and wingless angels

127

Fig. 6. Figures resembling the donkey-headed Seth in Coptic magical texts; left P.Berlin inv. 8323; right Crum 23.2.2. Images created by K. Dosoo.

all in the context of malign magic – curses, love and love spells – in which his unruly strength is enlisted to attack or compel their target.32 While the knife-like form of one of the early, though fragmentary, Seth-images, suggests a separation spell,33 the clearest examples of Seth-like figures from the later Coptic tradition (Fig. 6)34 come from unclear ritual contexts, consisting of small sheets containing only charaktēres and voces magicae. We would, nonetheless, expect them to be used in malign magic: even absent knowledge of the identity of Seth-Typhon, an animal-headed figure holding a weapon is readily understood as a dangerous supernatural being. As Edward Love discusses in his forthcoming study of Crum 23.2.2 (Fig. 6),35 this picture is complicated by the apparent labelling of the figure on this manuscript as Alpak, likely an alternative form of Harpax, a demon known from the Testament of Solomon,36 described as the form of the demon of Shamelessness in the Confession of Cyprian,37 and who appears in 32 For a discussion of Seth’s role in magical texts, see D. Fahre, ‘De Seth à Typhon et vice versa’, Egypte Afrique and Orient 22 (2001) 47-52. 33 P.Heid.Inv.Kopt. 473 (see n. 31). 34 P.Berlin 8323 and Crum 23.2.2 (see n. 31). 35 Love, ‘Crum’s Chicken’. 36 See Testament of Solomon XVIII.32 MS P in C.C. McCown, The Testament of Solomon (Leipzig, 1922) 57. 37 O.E. von Lemm, Sahidische Bruchstücke der Legende von Cyprian von Antiochien (St. Petersburg, 1899) 8, col.1.18-28; cf. the version in Pierpont Morgan M609 (F. Bilabel and A. Grohmann, Griechische, koptische und arabische Texte zur Religion und religiösen Literatur in Ägyptens Spätzeit [Heidelberg, 1934] 75. col.2.26-29), in which the equivalent passage says that the bird whose shape the demon takes is not “the one called Harpax” (ϩⲁⲣⲡⲁⲝ), but rather “the bird they call harpē” (Ⲙⲡⲉⲥⲙⲟⲧ ⲙⲫⲁⲗⲏⲧ ⲉⲧⲉϣⲁⲩⲙⲟⲩⲧⲉ ⲉⲣⲟϥ ϫⲉ

128

korshi dosoo

a different form in another illustrated Coptic text.38 In the Confession of Cyprian, Harpax is described as a kind of bird, and so the sharply pointed snout of the “Seth-figure” identified as Alpak may represent not merely a variant on the animal head, but a deliberate attempt to produce an avian shape, although the points which emerge from his head – understandable either as ears or a crest – complicate this picture. If the Seth-Typhon figure provided the model for this image, it seems to have undergone a process either of deliberate reinterpretation, being re-used as a model to depict Harpax, or of accidental misinterpretation, with the donkey head being understood as that of a bird, and identified with the Solomonic demon.39 2. The Function of Images in Coptic Magic In the previous examples we have seen examples of the two uses of magical images which Richard Gordon has described as “evocative” and “performative”.40 “Evocative” images depict the spirit-beings who are intended to serve as the agents of the ritual, providing them with a material form and hence presence. “Performative” images depict the action desired by the ritualist, manifesting and prefiguring its outcome, and providing instructions to the spirit-beings for what they are to do. In practice, the two often overlap – images of Seth wielding a weapon not only evoke the god, but ϩⲁⲣⲡⲏ), presumably referring to the Greek ἅρπη, perhaps the lammergeier or bearded vulture (Gypaetus barbatus; W.G. Arnott, Birds in the Ancient World from A to Z [London and New York, 2007] 63-64). This seems to have originated from a misreading of the Greek text, in which the demon is said to have a body “sharp like a sickle” (ὀξὺ ὅλον τὸ σῶμα ἔχον κατὰ τὴν ἄρπην, 42.7; R. Bailey, The Confession of Cyprian of Antioch: Introduction, Text and Translation [unpublished master’s dissertation, McGill University, 2009]). 38 P.Heid.Inv.Kopt. 408 (KYP M43) in S. Beck, ‘4. Ein Rezept und zwei Beschwörungen’, in Α. Boud’hours et al. (eds.), Coptica Palatina. Koptische Texte aus der Heidelberger Papyrussammlung (Heidelberg, 2019) 43-50. The figure of Arpakh is extremely stylised and has a cross on his head, perhaps indicating his status as a servant of King Solomon. Compare the figure on p. 10 of P.Heid.Inv.Kopt. 685 (TM 102074, 10th cent. CE) – like the Arpakh of P.Heid.Inv.Kopt. 408 this figure is bird-like, has a cross on its head, and seems to represent the being ⲛⲁⲥⲥⲕⲗⲏⲛ, described as the “the one who protects and guards (ⲣⲁⲉⲓⲥ ⲁⲩⲱ ϩⲁⲣⲉ) the body of King Solomon”, ll. 2-5). As Meyer notes, this Nassklēn may be related to Ὀνοσκελίς (“Donkey-legged”), a demon mentioned in the Testament of Solomon IV.2, 4; see M. Meyer, The Magical Book of Mary and the Angels (P.Heid.Inv.Kopt. 685) (Heidelberg, 1996) 82. It may be that both of these texts draw upon a standard iconography – a bird whose head is surmounted by a cross – to indicate a demon tamed and used as a servant by Solomon; compare the images of the Guardians discussed below. 39 For a fuller discussion, see Love, ‘Crum’s Chicken’. 40 R. Gordon, ‘Shaping the Text: Innovation and Authority in Graeco-Egyptian Malign Magic’, in H.F.J. Horstmanshoff et al. (ed.), Kykeon: Studies in Honour of H. S. Versnel (Leiden, 2002) 69-111, esp. 98-103.

heathen serpents and wingless angels

129

perform his act of threatening the spell’s victim, while the ouroboros simultaneously evokes a serpent spirit, and depicts him coiled in a protective circuit. Likewise, there are other conceivable functions of magical images: amulets depicting scorpions are common in Egypt from the fifth through to eleventh centuries, with examples known with text in Greek, Coptic, Arabic and Hebrew,41 and we know from the text which accompanies the most explicit examples that these are intended to ward off the beings they depict, an example of like repelling like, and therefore the opposite of the evocative function of images. While these categories help us to understand how images function from an etic perspective, it is unclear if the users of Coptic magical texts would have understood them in this way. The regular depiction of Jesus and the angels, for example, creates clear parallels between the role of images in magic and the use of icons in Christian worship, the site of considerable theological discussion. Drawing on Neoplatonic concepts of the anagogic role of images, Christian theologians tended to deny power to man-made representations themselves, stressing their role as an aid to the contemplation of the invisible divine powers that lay behind them.42 Nonetheless, the frequent attribution of miraculous power to icons, as well as less theoretical devotion to them, demonstrates that many Christians believed in the ability of images to mediate the power of the beings they depicted. An interesting Egyptian example of this occurs in the Encomium of Eustathius on the Archangel Michael, preserved in five copies, the earliest of which dates to the eighth century.43 This story tells of how the dying governor of the island of Trake has an icon (εἰκών) of the archangel painted for his wife Euphemia, praying to it to serve as her guardian, and she continues to pray and offer incense daily before the icon after his death. When the 41 For a discussion of these, see L. Berkes, ‘An Arabic Scorpion-Amulet on Paper from the 10th-11th c. and its Coptic and Hebrew Parallels’, Chronique d’Égypte 94 (2019) 213-15 and K. Dosoo, ‘Suffering Doe and Sleeping Serpent: Animals in Christian Magical Texts from Late Roman and Early Islamic Egypt’, in Dosoo and Coulon, Magikon Zōon (forthcoming). Coptic examples include P.Hermitage Copt. 66 (TM 109329, 10th cent. CE), Vienna K 7076 (TM 91404, 10th-11th cent. CE), Vienna K 7110 (TM 91413, 10th-11th cent. CE). 42 For discussions of the concept of anagogy through the contemplation of symbolic images, see W.-M. Stock, ‘Theurgy and Aesthetics in Dionysios the Areopagite’, in S. Mariev and W.-M. Stock, Aesthetics and Theurgy in Byzantium (Boston and Berlin, 2013) 13-30, as well as the other articles in this collection; Peers, Subtle Bodies, 90-95, 100-02. 43 For this text, see A. Campagnano, ‘L’Encomio di Michele Arcangelo di Eustazio di Tracia’, in A. Campagnano et al. (eds.), Quattro Omelie Copte (Modena, 1977) 107-72 (text and Italian translation); E.A.W. Budge, Saint Michael the Archangel: Three Encomiums (London, 1894) xxvii-xxxvii, 74*-108*, 93-135 (Bohairic text and English translation); E. Amélineau, Contes et romans de l’Égypte Chrétienne (Paris, 1888) 21-48 (French translation).

130

korshi dosoo

Devil attempts to lead her astray by persuading her to remarry and cease her almsgiving, his identity is revealed by his inability to enter the room where the icon is, or to look upon it – vividly illustrated at one point in the story in which Euphemia chases the devil away with the icon in her hands. Nonetheless, the icon and the angel are clearly distinguished, and when Michael is away praying for the rise of the Nile, the Devil is able to attack her despite the presence of the icon, and he is only stopped from killing her by Michael’s last-minute arrival. The icon is nonetheless possessed of a certain power, flying through the air to a church when Euphemia dies, and subsequently performing miraculous healings, and sprouting fresh leaves and fruit from its olive-wood support once a month. Like the icon of Michael, the images of supernatural beings in Coptic magic do not seem to necessarily be inherently efficacious. Rather, the creation of the activated texts of which they are part necessitates their writing on special surfaces – Aswan ostraca or gazelle vellum, for example44 – with special inks – such as human or animal blood45 – at special times – usually the full or waxing moon46 – while burning specific incense and speaking particular formulae. Even this act does not seem to empower the image in a mechanical fashion, rather the more explicit spells call upon the invoked beings – usually angels – to come, and to descend upon their image and their phylacteries, or to stand by the ritualist or to bless the oil and water used in anointing patients.47 Sometimes the images are in fact washed off, and the liquid used is then ingested or used to anoint the victim, patient, or

44 See for example the use of Aswan ostraca in P.Heid.Inv. 500 + 501 (TM 102087, 6th8th cent. CE?), ll. 1, 117 and gazelle vellum in the same text, ll. 27, 46, 105. 45 See, for example, the use of the blood of a white dove in P.Heid.Inv. 500 + 501 ll. 46, 106 and London Ms. Or. 6796 [2, 3] verso, ll. 112-115 (TM 100019, 6th-7th CE); the blood of a vulture in P.Heid.Inv.Kopt. 686, 254 (KYP M166); human menstrual blood in P.Heid.Inv. Kopt. 681, l. 48 (TM 99609, 10th cent. CE). 46 The image of Jesus-Bes is to be copied during the waxing moon in P.Heid.Inv. 500 + 501, l. 107, as is the image of Jesus in P.Heid.Inv.Kopt. 685 on p. 17. 47 See for example London Ms. Or. 5525 (TM 98056, 4th-9 th cent. CE) ll. 9-11: “…until you come down upon your image and your phylacteries of salvation” (ⲡⲉⲕⲍⲟⲧⲓⲟⲛ ⲙⲛ ⲛⲉⲕⲫⲩⲗⲁⲕⲧⲏⲣⲓⲟⲛ ⲛⲟⲩϫⲉⲓ); P.Heid.Inv.Kopt. 684 (TM 98064, 11th cent. CE) ll. 159-190: “…until you come upon your image, at once…”; P.Yale 1791 (TM 100011, 6th-7th cent. CE) first text, ll. 49-51: “You shall come to me today, I, NN, and you shall stand upon this cup that is placed before me and you shall fill it for me with a voice that is sweet and pleasant”; P.Heid.Inv.Kopt. 685 p. 8, ll. 18-25: “…send me our holy Mother of God, Saint Mary, (the) holy virgin… that she seal (ⲥⲫⲁⲣⲁⲅⲓⲁⲍ = σφραγίζειν) (the) water and the oil, so that the moment that I pour the water upon NN, he may become strong and healthy and completely well”.

heathen serpents and wingless angels

131

client.48 At other times, as in the case of amulets, the images are carried by the user. Again, like the icon of Michael, then, the images can project power, but they require the presence of the being they depict, a presence brought about through the acts of invocation and offering. Nonetheless, the images do seem to have a real relationship with the beings they depict; again, we may relate this to theological ideas concerning iconography. While angels are inherently bodiless beings, their images are not arbitrary; rather, their depictions draw upon the forms they take during apparitions to the faithful, which in turn communicate something of their essential nature, even if they do not straightforwardly disclose their true appearance – the wings of angels, for example, describe their rapid movement between heaven and earth.49 This resemblance is meaningful, and in another miracle story from seventh-century Constantinople, a different Euphemia, a young girl on her deathbed, is able to identify the angels and saint who visited her in a dream by their resemblance to their icons in the local church.50 This significance of the relationship between spirit-being and image is particularly important in Coptic magic; the act of invoking or adjuring a being often relies upon a horkōmotos,51 a thing by whose authority the being is adjured. In Coptic text, lists of horkōmotoi are often stereotyped, consisting of variations on the phrase “I adjure you by your name, your image, your phylacteries, and your dwelling place”.52 The term used for 48 For the washing-off of images, see e.g. London Ms. Or. 6794 l. 57; P.Heid.Inv. 500 + 501 ro ll. 1-5, vo ll. 116-124; P.Heid.Inv.Kopt. 682 ll. 7-8; P.Heid.Inv.Kopt. 685 p. 16 l. 10; P.Heid.Inv.Kopt. 686, 260. 49 See, for example, the statement of John Chrysostom, De incomprehensibili dei natura 3.318-322: “[Angel’s] wings reveal the loftiness of their nature (τὸ ὕψος ἐμφαίνουσι τῆς φύσεως). Gabriel is shown as flying not because angels have wings, but so that you may know that he comes down to human beings from places which are lofty (ἐκ τῶν ὑψηλοτάτων) and from a way of life which is spent on high”; translation from P.W. Harris, St. John Chrysostom, On the Incomprehensible Nature of God (Washington D.C., 1984) 107-08; cf. In Isaiam 6.3.1-4. For a fuller discussion, see T. Martin, ‘The Development of Winged Angels in Early Christian Art’, Espacio, tiempo y forma. Serie VII, Historia del arte, 14 (2001) 11-29 at 18-19; Peers, Subtle Bodies, 110-20. 50 V.S. Crisafulli and J.W. Nesbitt, The Miracles of St. Artemios: A Collection of Miracle Stories by an Anonymous Author of Seventh-Century Byzantium (Leiden, 1997) 178-81. 51 An excellent discussion of the act of adjuration, from which I have taken the term horkōmotos, may be found in S. Shauf, Theology as History, History as Theology: Paul in Ephesus in Acts 19 (Berlin and New York, 2005) 202-18. 52 A full discussion of this formula is beyond the scope of this paper; a particularly extensive example is found in London Ms. Or. 5525 (TM 98056, 4th-9th cent. CE) ll. 1-7: “I adjure you by your name and your power and your image and your phylacteries of salvation and the places in which you dwell and your staff of light which is in your right hand and the chariot (wheel?) of light that is in your left hand and your great (?) Powers that

132

korshi dosoo

image (zōdion) in these adjurations is the same as that used for the accompanying images of angels and other beings. One variant both adjures the being by its “image and… phylacteries of salvation” (ⲛⲫⲩⲗⲁⲕⲧⲏⲣⲓⲟⲛ ⲛⲟⲩϫⲁⲓ),53 and adjures it to come to these same “image and phylacteries of salvation”54 strongly implying that the “image” that they are adjured by is the (prototype of) that depiction which appears in the papyrus. These images, thus, represent as inherent a part of the depicted spirits’ identities as their names and their celestial abodes. The reference to “phylacteries” is particularly interesting. In the case above, and in many of the parallel texts, the “image” is singular while the phylacteries are listed in the plural. In another text the user is instructed to copy the “images and phylacteries”.55 This implies that “phylactery” is a technical term in Coptic magical texts for images, or specific parts of them. The same text also describes phylacteries as being written upon the breast of the Father, while elsewhere in the same and in other texts Greek letters may be written on the Father’s chest.56 It seems therefore to me that the term “phylacteries” may refer to the charaktēres, which are not clearly captured by the Greek word zōdion, with its implication of a figural image. Charaktēres have a textual quality,57 which would explain why they may also be referred to as letters, and usually occur in groups, justifying the references to them in the plural. 3. Characteristic Elements of Coptic Magical Images As I have already noted, Coptic magical images display significant stylistic regularities. This is particularly apparent in their treatment of human stand before you” (ⲧⲓⲧⲁⲗⲕⲁ ⲙⲁⲕ ⲉⲡⲉⲕⲗⲉⲛ ⲙⲛ ⲧⲉⲕϭⲁⲙ ⲙⲛ ⲡⲉⲕⲍⲟⲧⲓⲟⲛ ⲙⲛ ⲛⲉⲕⲫⲩⲗⲁⲕⲧⲏⲣⲓⲱⲛ ⲛⲟⲩϫⲉⲓ ⲙⲛ ⲛⲉⲧⲱⲡⲱⲥ ⲉⲕϣⲁⲁⲡ ⲉⲛϩⲏⲧⲟⲩ ⲙⲛ ⲡⲉⲕϩⲣⲁϥⲧⲟⲥ ⲛⲟⲩⲁⲉⲓⲛ ⲉⲃϩⲛ ⲧⲉⲕⲧϭⲓϫ ⲛⲓⲱⲛⲁⲙ ⲙⲛ ⲛ̣ⲉⲕϩⲁⲣⲙⲁ ⲛⲟⲩⲁⲉⲓⲛ ⲉϥϩⲉⲛ ⲧⲉⲕϭⲓϫ ϭⲁϫⲉ ⲙⲛ ⲛⲉⲕⲛⲁϭⲁⲙ ⲉⲟⲩⲁϩⲓⲣⲁⲧⲟⲩ ⲉⲗⲁⲕ). 53 London Ms. Or. 5525 ll. 2-3, quoted in n. 52 above. 54 London Ms. Or. 5525, ll. 9-11. 55 Rossi’s Tractate 11.16-18: “…the twenty-four letters which are on the phylactery of the Father” (ⲡⲓϫⲟⲩⲧⲁϥⲧⲉ Ⲛⲥϩⲁⲓ ⲉⲧϩⲘ ⲡⲉⲓⲫⲏⲗⲁⲕⲧⲏⲣⲓⲟⲛ Ⲙⲡⲓⲱⲧ); London Ms. Or. 6794 ll. 40-42 (TM 100017, 6th-7th cent. CE): “…the seven vowels which are branded on the chest of the Father…” (ⲡⲓⲍ ⲛⲥⲧⲓⲭⲋ Ⲛⲥϩⲁⲓ ⲛⲁⲓ ⲉⲧϫⲉⲕϫⲱⲕ ⲉϫⲚ ⲧⲙⲉⲥⲧⲚϩⲏⲧ Ⲙⲡⲉⲓⲱⲧ ⲉⲧⲉ ⲛⲁⲓ ⲛⲉ {ⲁ}ⲁⲍ ⲉⲍ ⲏⲍ ⲓⲍ ⲟⲍ ⲩⲍ ⲱⲍ). Both of these may suggest that the phylacteries, like charaktēres, have a text-like quality. Compare the often text-like charaktēres that often appear on the bodies of figures in magical papyri, eg. Figs. 2, 6, 7, 8, 9, 11, 13, 14, 16, 19, 21, 22, 24, 26. 56 Rossi’s Tractate 14.11-13 (TM 98062, 6th-9 th cent. CE?): “…these phylacteries that are written upon the breast of the father” (Ⲛⲉⲓⲫⲩⲗⲁⲕⲧⲏⲣⲓⲟⲛ ⲉⲧⲥⲏϩ ⲉⲧⲙⲏⲥⲧ Ⲛϩⲏⲧ Ⲙⲡⲓⲱⲧ). 57 Compare the comments at note 16.

heathen serpents and wingless angels

133

faces, almost always drawn frontally, and consisting of an oval outline, eyes indicated with flattened circles, and eyebrows and nose depicted with a single line. Variants on this theme include the addition of pupils to the eye, or the addition of a mouth, drawn as one or two strokes, with the lower stroke marking the shadow below the lower lip, or as a circle. Hair, not always indicated, may take several forms, a few of which will be discussed below. Animal heads are usually drawn in profile, with a single visible eye; it is usually difficult to determine which animal is intended, or even if we are looking at the snout of a mammal or reptile or the beak of a bird. Even the addition of “ears” to figures is rarely useful in narrowing this down – earlike crests may be given to snakes, such as the ouroboros in P.Berlin inv. 15990 (Fig. 4), and to figures which are almost certainly birds, such as those in Plate 1 below. Copyists probably used these protrusions, and similar ones resembling horns, antennae or the crests of peacocks and hoopoes, as a way of distinguishing individual beings, rather than of clearly delineating species. The clothing of human figures is usually not clearly marked; the figures generally consist of geometric forms, such as circles and rectangles, although in a few cases additional details suggest that more specific garments were intended, even if they are not readily identified. More obvious is the fact that most figures are depicted in orant position, with their arms extended straight out from their bodies. In many cases, this is probably because the figures we are dealing with represent angels, or perhaps the practitioner or other human beings, in the act of prayer. This is in keeping with common depictions of human figures in other contexts from late antique Egypt, such as funerary stelae and murals,58 but may also indicate the understanding of magical formulae as prayers, an interpretation strengthened by the regular use of the words for prayer in their titles.59 Yet in other cases, such as P.Würzburg 42,60 in which orant figures represent victims of curses in performative tableaux, we should probably not understand them as praying. Rather the orant pose seems to have simply been the default manner of depicting human figures in this genre of text, one which

58 For a few representative examples see T.K. Thomas, Late Antique Egyptian Funerary Sculpture: Images for This World and the Next (Princeton, 2000) figures 30, 31, 86, 90, 95 and 96. 59 See, for example, Vienna K 8302 II l. 1 (KYP M364, 6th-8th), P.Ryl.Copt. 104 sec. IV l. 1 (TM 98059, 11th-12th cent. CE), where invocations are called “prayers” (ϣⲗⲏⲗ), and P.Heid.Inv.Kopt. 684 l. 249 where the Greek word προσευχή, with the same meaning, is used. 60 See n. 7.

134

korshi dosoo

Fig. 7. London Ms. Or. 5525; figure depicting the archangel Michael (?). Image created by K. Dosoo. The conventional reconstruction of this image (see Kropp, Ausgewählte Koptische Zaubertexte, vol. 3, pl. V) is probably misleading in making the wings resemble curved horns.

had the advantage of allowing the arms to be clearly visible, and therefore the identity of the figure as a human to be clear. A third characteristic of Coptic magical images is the thick borders found on many figures, consisting of a “tube” created from parallel outlines, often shaded through hatching to create a striated appearance. In their analysis of the images in the Heidelberg Library, Tamara Mößner and Claudia Nauerth refer to this as a “scaly border” (schuppenartige Umrandung),61 understanding it as an indication of a demonic nature or influence. It is true that this striation resembles that found on, for example, the ouroboroi discussed above, with the earliest example among them depicting the belly-scales of a serpent, and although similar striation is rare in the corpus of Graeco-Egyptian magical images, it can be found, for example, on the legs of the scarab in PGM II,62 the arms of some of the depictions of the headless-god in PGM XXXVI, and the snake carried by one of his 61 62

Mößner and Nauerth, ‘Koptische Texte und ihre Bilder’, 316, 321. TM 88397, 3rd cent. CE. The scarab is depicted at l. 159.

heathen serpents and wingless angels

135

Fig. 8. Image of an angel (?) from Rossi’s Tractate, after M.W. Meyer, Rossi’s “Gnostic” Tractate (Claremont, 1988). Image created by K. Dosoo.

depictions in this same papyrus.63 Such striation is far more common in Coptic magical images, however, being used to mark not only the bodies of serpents, but also the arms and bodies of anthropomorphic and therianthropomorphic beings, and in instances where the being is not only demonic, but also angelic, and in which it represents the human victim of a spell. In an analysis which parallels that of Mößner and Nauerth, Jay Johnston has suggested that these striated borders are indicative of supernatural power more broadly, marking both supernatural beings, and humans being influenced by them, which would seem to agree with the depictions.64 The origins of this striation are more obscure; despite the earlier instances in the Graeco-Egyptian magical corpus, it does not become systematic until the later Christian manuscripts. A promising suggestion seems to me that of Angelicus Kropp,65 who proposed that, in the case of the angels, it represented rays of light emanating from the body, and indeed, the brightness of angelic manifestations is a key feature of angels in many texts; in the Encomium of Eustathius Michael is described as shining brighter than the sun, with feet burning with flames.66 Striations are used in Roman63

The image is found below col. 2 on the recto, and relates to the spell in ll. 35-68. J. Johnstone, ‘On “Being” a Ritual Knife: Materiality, Form and Ritual Agency in Paris, Louvre E14250’, Paper presented at Magic in the Ancient World: Textuality and Contextuality, Macquarie University, Sydney, 2014. 65 Kropp, Ausgewählte Koptische Zaubertexte, vol. 3, 211-12. 66 158.7-8, 164.5-7 (Campagnano, ‘L’Encomio di Michele Arcangelo di Eustazio di Tracia’; the Bohairic version is more explicit than the Sahidic). 64

136

korshi dosoo

period art in Egypt, including in the context of magical gems, to represent rays of light, and I would tentatively suggest that striated borders be understood principally as indicating light or fire emanating from a figure, a visual manifestation of the power inherent in or acting upon them.67 Alongside these typical visual elements, Coptic magical images may rely heavily on the iconography of the larger culture, although this is inconsistent. One clear instance can be found in London Ms. Or. 6796,68 in which an exorcism procedure which uses a historiola giving an unconventional account of Jesus’ crucifixion and descent to the underworld is illustrated by Jesus crucified between the two thieves, an image construction well attested in non-magical contexts, known examples of which the drawing here follows closely.69 In other cases, however, depictions of the central figures of Christianity are far more generic. In Vienna 1992,70 an orant figure (Fig. 10) is drawn on the reverse side of an amulet containing Psalm LXX 90.2, a psalm attributed to David. The figure on the reverse, labelled David (Ⲇ ⳰Ⲁ ⳰Ⲇ), must therefore represent the prayer’s composer and first user, but the image is otherwise an apparently generic orant figure. His identity, therefore, is determined solely by the caption attached to him. This principle is not unique to magical texts; as Glen Peers has discussed, the supporters of icons understood text and image as equally representations of sacred figures, which acted together to determine the identity of particular figures, and a similar reliance on captions can be seen in a depiction, from the 67 Rays of light are regularly represented by dense groups of short lines in art from Roman Egypt – see, for example, the depictions of solar deities on magical gems (e.g. depictions of Chnoubis, such as CBd-6), and the haloes of gods in icons and statues (e.g. the halo of Sobek on CG 17569 [2nd cent. CE]; T.F. Mathews and N.E. Muller, The Dawn of Christian Art in Panel Paintings and Icons [Los Angeles, 2016] 126). While Christian haloes are not normally marked with rays, they too may have offered an iconographic forerunner of the “striations of light” found in magical texts – the cruciform halo of Jesus, or the garlanded borders of regular haloes (see, for example, the halo of the tetramorphic angel from the Church of Saint Macarius [7th cent. CE]; Massimo Capuani, L’Égypte copte [Paris, 1999] 63) – may have inspired a striated halo in at least one Christian magical gem (J. Spier, Late Antique and Early Christian Gems, [Wiesbaden, 2007] no. 638a, plate 87). 68 A reproduction of the image may be found in A. Kropp, Ausgewählte Koptische Zaubertexte, vol. 3, Tafel 1. On this text, see now Sanzo, ‘The Innovative Use of Biblical Traditions for Ritual Power’. 69 A. M. Kropp, ‘Die Kreuzigungsgruppe des koptischen Papyrus Brit. Mus. Ms. Or. 6796’, Oriens Christianus 25-26 (1930), 64-68; Kropp, Ausgewählte Koptische Zaubertexte, vol. 3, 215. Note particularly the way in which the thieves are shown as crucified with their arms tied behind their back in the ‘oriental’ form, common to many early depictions; see E. Balicka-Witakowska, ‘The Crucified Thieves in Ethiopian Art: Literary and Iconographic Sources’, Oriens Christianus 82 (1998) 204-56 at 209-11. 70 TM 113765, 8th cent. CE.

heathen serpents and wingless angels

Fig. 9. P.Köln Kopt. 3; figure of angel (?) in orant position. Image created by K. Dosoo.

137

Fig. 10. Vienna 1992 verso; figure of King David in orant position. Image created by K. Dosoo.

ninth century Church of the Dormition in Nicaea, of four orders of angels – a Dominion (κυριότης), an Authority (ἐξουσία), a Ruler (ἀρχή), and a Power (δύναμις) – who are iconographically identical, but distinguished by their accompanying text.71 Two other examples complicate this, however. On p. 9 of P.Heid.Inv. Kopt. 685, a long prayer for exorcism attributed to Mary ends with the image of a haloed orant figure with striated arms holding two long objects, understood by Marvin Meyer, the original editor, as “cosmic wands” (Plate 1).72 While similar shapes, often labelled as “wand, staff” (ῥάβδος) are held by some of the angels discussed below,73 the specific form of this object resembles the palm-branches held by religious figures in Roman and late-antique art, albeit subjected to the stylisation in the direction of a charaktēr described above.74 The figure is labelled twice, both as Mary (ⲙⲁⲣⲓⲁ) 71

Peer, Subtle Bodies, 120. Meyer, The Magical Book of Mary, 80. 73 See, for example, P.Heid.Inv.Kopt. 686, 2; London Ms. Or. 5525 l. 4. 74 While palm branches are often carried by martyrs, they may also denote victory (see for example the winged victory which carries one in the Barberini diptych from the Louvre [6th cent. CE]; louvre.fr/en/oeuvre-notices/leaf-diptych-emperor-triumphant [08/03/2021]), and carrying palm-branches in other contexts in Mediterranean religions is a sign of participation in religious festivities (see W. Burkert, Structure and History in Greek Mythology and Ritual [Berkeley, 1979] 43-44; compare the account of the entry of Jesus into Jerusalem [Matthew 21.8, Mark 11.8-9, John 12.12-13] and its depictions in Christian art). Ritualists are regularly instructed to hold branches while carrying out magical rituals: myrtle branches in P.Macq. I 1 (TM 113926, 8th cent. CE) ll. 20-21 and Rossi’s Tractate ll. 3-4, and they may also be described as attributes of spirits – the luminary Dauithe is described as carrying a “golden palm branch” in P.Macq. I 1 4.10, 17 and London Ms. Or. 5987 (TM 98061, 6th-7th cent. CE) l. 72. 72

138

korshi dosoo

and as Jesus Christ (Ⲓ ⳰Ⲥ Ⲭ ⳰Ⲥ); the fact that the latter letters are written as charaktēres should not disqualify them as a caption, since similar letters are used to label the depiction of the crucified Jesus in London Ms. Or. 6796. Iconographically, the figure could be either – the long robe could be that of Mary or Jesus, and is not clearly distinct from that worn by a more straightforward depiction of Jesus on p. 17 of the same document. While the border around the head could be Mary’s palla (shawl) it is not clearly distinct from the haloes found regularly on male figures. Finally, the two serpent-like creatures found beneath the figure’s right foot, presumably representing the forces of evil over which the spell promises triumph, echo Jesus’ role as the one who tramples on serpent and dragon, but represents a motif attributed to many saints.75 Here, then, the identity of the figure is not clear iconographically, and the captions are ambiguous; while the identity of the figure is likely Mary, given her prominence in the practice to which the image is appended, and the other forms of her name which appear in the tableau, this instance should alert us to the danger of assuming that accompanying text always represents a caption. The ambiguity of the figure might even be deliberate, perhaps representing the practitioner, who speaks in the person of Mary in the formula – “I am Mary” (ⲁⲛⲁⲕ ⲙⲁⲣⲓⲁ)76 – while Jesus’ name might indicate not his identity, but rather his presence. The final example, from London Ms. Or. 6795 (Fig. 11),77 is taken from a spell for fishing, whose formula invokes Jesus to send the angel Raphael to collect fish for the practitioner. The accompanying image, apparently to be copied onto a glass vessel, consists of an orant figure wearing a long gown, whose right hand holds a fishing rod which a fish, perhaps an oxyrhynchus, has bitten. The name Jesus (Ⲓ ⳰Ⲥ) appears three times, once on the human 75 The motif of the saint trampling a demon is a common one in Christian art, while LXX Psalm 90.13 was generally understood to describe Jesus as trampling evil underfoot in the form of the lion and the dragon which appear in this text. In Western Christianity, images of Mary often depict her trampling the Devil, who may take the form of a dragon. This seems to derive from the Latin translation of Genesis 3.15, in which the passage describing the male child of Eve bruising the head of the Serpent was translated with the feminine pronoun ipsa, resulting in an interpretation in Western (but not Eastern) Christianity of the passage as referring to Mary; see J.P. Lewis, ‘The Woman’s Seed (Gen 3:15)’, Journal of the Evangelical Theological Society 34.3 (1991) 310-13; A. Reed, ‘Blessing the Serpent and Treading on its Head: Marian Typology in the San Marco Creation Cupola’, Gesta 46.1 (2007) 41-58. 76 P.Heid.Inv.Kopt. 685 2.20-22. 77 TM 100018, 6th-7th cent. CE. An interesting parallel is found on P.Bas. II Pahl. 1 (618629 CE), a Pahlavi administrative text on whose back two fishing figures have been drawn, one in a boat, the other apparently holding a bird and a chalice; see S. Schmidt, ‘P.Bas. II 70 and 71. A Look Behind the Text’, Archiv für Papyrusforschung 64.2 (2018) 324-42.

heathen serpents and wingless angels

139

Fig. 11. London Ms. Or. 6795; tableau depicting orant figure catching a fish. Image created by K. Dosoo.

figure’s body, once between human and fish, and once on the other side of the fish. Again, the texts cannot be taken straightforwardly as labels; while Jesus is invoked in the spell, he is asked to send Raphael to carry out the task, not to fulfil it himself, and while he appears in a historiola based upon the Miraculous Draught of Fish,78 in this story it is his apostles, and not he himself, who catch the fish. If we take the name as a label, then we must assume, with Horak, that the fish too is Jesus,79 and while this is perhaps not altogether impossible – the fish is a well-known Christian symbol – it seems out of place in a practice which aims at catching real fish to kill and eat. Again, I suggest that the simplest interpretation is to understand Jesus as being depicted solely through his name, his presence surrounding the fish and filling the figure, who perhaps represents the practitioner. 4. Angels in Coptic Magical Texts As we might expect, angels represent one of the most important categories of beings in Coptic magical texts; they are the agents most often summoned by spoken formulae, and sent by God and Jesus to carry out magical miracles. Depictions of angels are particularly abundant in magical texts, but their iconography has posed challenges to previous studies, some of which may be resolved through a synoptic approach. Images of angels seem particularly associated with exorcistic practices, recalling perhaps, the idea that the Devil could not enter a room containing an image of the Archangel Michael which we encountered in the Encomium of Eustathius, but they may be found in almost any context, from favour 78 79

John 21.1-14; Luke 5.1-11. Horak, ‘Christliches und Christlich-Magisches’, 35.

140

korshi dosoo

Fig. 12. LACMA 80.202.214; remains of figure of angel (?). Image created by K. Dosoo.

Fig. 13. P.Heid.Inv.Kopt. 681; figure of angel (?). Image created by K. Dosoo.

Fig. 14. Coptic Museum 4959 (upper part); figure of angel (?). Image created by K. Dosoo.

heathen serpents and wingless angels

141

spells to love spells and in the long formularies containing multiple recipes using a single formula and image. Although the earliest Christian images of angels depict them as indistinguishable from humans, their iconography as winged youths became standard by the fourth century,80 and as early as the late second century Tertullian could claim that “every spirit is winged; so it is with angels, so it is with demons”.81 Given this, it is strange that Mößner and Nauerth’s study of images in the Heidelberg Library concludes that angels in the magical texts rarely have recognisable wings;82 this would suggest that magical images differed from those produced by the broader culture not only stylistically, but also in their basic iconography. Similarly, Horak notes in her description of the archangel Michael from London Ms. Or. 5525 (Fig. 7) that he has not wings, but rather “hatched serpents” which curl from his shoulders.83 Similar forms are found on the shoulders of P.Köln Kopt. 3 (Fig. 9),84 which Horak identifies as Jesus equipped with “snake-shaped neck extensions”,85 rejecting the suggestion of the original editor that they are in fact the wings of an angel.86 Here, she follows the analysis of Kropp,

80 See Martin, ‘The Development of Winged Angels in Early Christian Art’, 18-23; Peers, Subtle Bodies, 23-38. 81 Apology XII.8: Omnis spiritus ales est. Hoc angeli et daemones. Text and translation in T.R. Glover and G.H. Rendall, Tertullian: Apology. De Spectaculis. Minucius Felix: Octavius (Cambridge MA, 1931) 120-21. 82 Mößner and Nauerth, ‘Koptische Texte und ihre Bilder’, 313: “Selten tragen sie deutlich erkennbare Flügel”. 83 Horak, ‘Christliches und Christlich-Magisches auf Illuminierten Papyri’, 42: “schraffierte Schlangen”. 84 TM 101249, 6th cent. CE. 85 Horak, ‘Christliches und Christlich-Magisches’, 38: “…schlangenartige Halsfortsätze”. Horak’s identification of the figure as Jesus rests on the similarity of these “snakes” to the shapes emerging from the head in P.Berlin 8503 (TM 99586, 8th-9th cent. CE), which has shapes emerging from its head resembling the cross of a cruciform halo, and which she therefore identified as “Jesus-Ebbaēl”. While it is true that the cruciform halo is generally restricted to Jesus, there is nothing else in the image suggestive of the iconography of Jesus (certainly not ‘snakes’ emerging from the shoulders), so it is unclear if this single element should be taken as decisive. The strangeness of this particular image places it, for me, in the category of those whose analysis should await further research, but based on its similarity to the possible cherubim/seraphim in P.Heid.Inv.Kopt. 679, discussed below, I suspect it should rather be understood as an angel depicted as a winged head. Support for understanding the figure as an angel rather than Jesus is found in the label (written as charaktēres) Ebbaēl, a name which also appears as the primary invoked being in the curse; the termination -ēl is typical of angel names, and there is no clear reason to think he is identified with Jesus. 86 M. Weber in E. Lüddeckens et al. (eds.), Demotische und Koptische Texte (Papyrologica Coloniensia II) (Opladen, 1968) 89-90. Interestingly, as Weber notes, the invocation is addressed to Jesus; he suggests that the image has been added to a text to which it is not

142

korshi dosoo

who understood such forms, perhaps due to their striation, as scaled serpents, derived either from the miscopying of figures being attacked by snakes found in curse tablets, or else perhaps (given that they curled upwards) the horns of the moon-goddess.87 In fact, it is clear to me that these shapes are stylised wings, as we might expect in depictions of angels. Bird wings (Fig. 15) can be understood as consisting of two parts: coverts, which are short feathers covering the muscle and bones of the wing, and remiges, long flight feathers.88 Particularly on the inner wing, the part visible on front-facing angels, these are often distinguished by colour or pattern. Depictions of bird wings often ignore these details of their anatomy, and in these cases we typically find wings drawn in outline, and filled with undifferentiated semicircles representing the overlapping tips of feathers; such depictions may be found, for example, in LACMA 80.202.214 (Fig. 12),89 the upper image of P.Heid.Inv.Kopt. 681 (Fig. 13),90 and that of Coptic Museum 4956B (Fig. 25).91 Depictions which pay more attention to wing anatomy, however, may distinguish the two parts, and this is the more common approach in early Christian art from Egypt.92 Again, though, stylisation may reduce the coverts to a single form delimiting the upper edge of the wing, occasionally marked with striations or crosses representing feathers, while the remiges may either be marked with feathers, or left empty, so that in crowded scenes, such as that of the depiction of Christ Pantocrator surrounded by angels in the central sanctuary dome of the Monastery of Saint Anthony (Plate 2), the coverts are depicted as curved tubes, while the remiges are present only as the colour that fills them, strictly related. While this is possible, it seems equally likely that the angel is here (as elsewhere) the manifestation of the power which Jesus is to send. 87 Kropp, Ausgewählte Koptische Zaubertexte, vol.3, 212-13. 88 This description is slightly simplified, since both coverts and remiges may be divided into different categories (primary and secondary, lesser and greater). Note also that images often depict the outermost primary of the remiges as part of what I here call the covert, which in effect becomes the entire upper outline of the wing. 89 TM 642006, 4th-8th cent. CE?. 90 TM 99609, 10 th cent. CE. 91 This text remains unpublished, but is described in M.W. Meyer and R. Smith, ‘Invoking Aknator the Ethiopian’, Bulletin of the Institute for Antiquity and Christianity 18.3 (1991) 10-12; a parallel text, containing a very similar, though fragmentary, image, may be found in Coptic Museum 4959 (TM 100008, 9th-11th cent. CE; [Fig. 28]). Both images are discussed below. 92 While I focus here on Christian art, examples of earlier, non-Christian art which depict wings similarly (with tube-like upper coverts and remiges added as long feathers) can easily be found; see for example the depiction of Nike from Yale University Art Gallery 1929.228, figures 1.16-17 in Mathews and Muller, The Dawn of Christian Art.

heathen serpents and wingless angels

143

Fig. 15. Diagram of a bird-wing; the coverts are marked grey while the remiges are white. Edited version by K. Dosoo of tracing by Kjoonlee from drawing by Muriel Gorreop, image in the public domain (commons.wikimedia.org/wiki/File:BirdWingFeatherSketch.svg).

Fig. 16. Freer F1908.45.13; figure of angel (?). Image created by K. Dosoo.

distinct from the blue of the surrounding sky, with a dark lower border, again rather tube-like. A very similar approach may also be seen in a ninthcentury illustration of Gabriel from one of the Hamouli codices.93 This, then, is why the angels in magical texts have “hoses” emerging from their shoulders – they are their wings, reduced to coverts, and striated to represent either the light they emanate, or, more likely, their feathers. This is clearer in instances where attempts to depict the remiges are present in the form of strokes radiating outwards from the coverts, although these are usually added both above and below the main form of the wings, indicating a common misunderstanding of wing anatomy – examples of this may be 93 Pierpont Morgan Ms. M583 fol.8r (TM 44376, 848 CE); image available online at http://ica.themorgan.org/manuscript/page/6/214164 (13/03/2021). Compare also Benaki Museum ΓΕ 6988 (6th cent. CE), an embroidery from Egypt in which an angel is depicted with similarly tube-like wings.

144

korshi dosoo

found in the upper part of Coptic Museum 4959 (Fig. 14)94 and Freer F1908.45.13 (Fig. 16)95 and a clear non-magical example in a drawing from Pierpont Morgan Ms. M611 fol. 18v (9th cent. CE).96 It is worth noting that in some cases, such as P.Gieben Copt 1 (Fig. 19),97 these wings are feathered at their ends, making them difficult to distinguish from the equally striated arms ending in fingers, although, given the iconographic parallels for similar wings, it seems better to understand them as such rather than as additional pairs of arms. An interesting variant on the winged angel may be found on fol. 9r of P.Heid.Inv.Kopt. 686 (Fig. 2).98 The central figure, Michael, is depicted with both wings visible, the coverts represented as hollow arches, and the space within them filled by letters and vertical lines, which presumably stand in for the remiges. As we might expect for the martial angel, he is trampling two figures, presumably representing the Devil, and/or Death. He is accompanied by two Powers (δύναμις), who hover, each labelled, at either side of him. These should not, I think, be understood as vague manifestations of Michael’s power, however, but as members of the order of angels of the same name;99 there is, indeed, mention of two Powers with flaming swords in the text,100 a typical description of warrior angels. But, as Peers has 94 Interestingly, this manuscript contains depictions of three angels, whose wings are drawn quite differently; the first and third of these (Figs. 14, 20) are fairly conventional, with tube-like coverts, decorated in the case of the first with feathering above and below, while the second (Fig. 28) is a depiction of an “Ethiopian” (infernal) angel, whose wings (coverts and remiges) are drawn as outlines filled with dense waved lines representing dark feathers. 95 KYP M490, 8th-9 th CE; this text is unpublished, but an image is available online at https://asia.si.edu/object/F1908.45.13/ (12/03/2021). 96 This image is visible online at http://ica.themorgan.org/manuscript/page/4/77456 (13/03/2021). 97 TM 140205, 9 th cent. CE. See also the feathering on the end of the right wing of the angel on the top of P.Yale 1791 (Fig. 21), and the wings of the third angel of Coptic Museum 4959 (Fig. 20), which give the impression of terminating in three-fingered hands. 98 For a more detailed study of this image, see Gardner and Johnston, ‘I, Deacon Iohannes, Servant Of Michael’, 51-53. 99 The idea of Powers as an order of angels comes from Ephesians 15.21, which described God raising Christ “above all Principalities, Authorities, Powers, and Dominions” (ὑπεράνω πάσης ἀρχῆς καὶ ἐξουσίας καὶ δυνάμεως καὶ κυριότητος); cf. Pseudo-Dionysius, De caelesti hierarchia 33.8-16 (VIII.1). A similar scheme is suggested by the mention in P.Yale 1791 of Archangels, Principalities, Authorities, Forces on High, Powers, Cherubim and Seraphim (ll. 32-37: ⲁⲣⲭⲏⲁⲅⲅⲉⲗⲟⲥ, ⲁⲣⲭⲏ, ⲏⲕⲝⲟⲩⲥⲓⲁ, ϭⲟⲙ ⲉⲧϩⲉⲙ ⲡϫⲓⲥⲉ, ⲧⲩⲛⲁⲙⲉⲥ, ⲭⲉⲣⲟⲩⲃⲉⲛ, ⲥⲩⲣⲁⲡⲫⲉⲛ), and in P.Heid.Inv.Kopt. 686, ll. 38-42 of Angels, Archangels, Cherubim, Seraphim, Forces, Powers, Bodiless Ones and Authorities (ⲁⲛⲅⲉⲗⲟⲥ, ⲁⲣⲭⲁⲛⲅⲉⲗⲟⲥ, ⲭⲁⲣⲱⲃⲓ(ⲛ), ⲥⲉⲣⲁⲫⲓⲛ, ϭⲁⲙ, ⲧⲉⲛⲁⲙⲓⲥ, ⲁⲥⲱⲙⲁⲧⲟⲥ, ⲉⲝⲟⲩⲥⲓⲁ). 100 P.Heid.Inv.Kopt. 686, §109-111: “I adjure (ⲱⲣⲕ) you today by your great Power(s?) (ⲛⲧⲉⲕⲛⲟϭ ⲛⲧⲉⲛⲁⲙⲓⲥ) … one on your right, another on your left, with swords of fire in their hands”.

heathen serpents and wingless angels

145

Fig. 17. Drawing from P.Heid.Inv.Kopt. 679 verso, perhaps depicting an angel; the complex charaktēres which cover the rest of the sheet are omitted here. Image created by K. Dosoo.

noted, there was no standard iconography for such orders of angels, so that they instead tended to borrow the generic iconography of the winged youth.101 Here, however, they draw instead upon the six-winged iconography of the cherubim and seraphim: bodiless heads surrounded by three pairs of wings – one above, one below, one to the side – as is standard in depictions in, for example, Church murals.102 As in the example from the Monastery of Saint Antony (Plate 2), the wings of these cherubim-powers are treated differently from those of the angel with a single pair of wings, consisting of thin, saw-like forms, which are internally divided into coverts and remiges. As in the depictions of the cherubim and seraphim in other contexts, their arms are also shown, creating a fourth pair of limbs, with their right arms holding flaming swords (depicted along very similar lines to their wings), and their left holding an object which is difficult to identify, though perhaps an orb. This six-winged variant on the angel may help to explain the mysterious figures identified by Mößner and Nauerth as the “snake-woman” (Schlangenfrau);103 this being appears most often in contexts where angels are called upon, and bearing in mind the iconographic confusion between wings and arms, I suggest that the twisting-arms of the “snake-woman” should instead 101 102 103

Peers, Subtle Bodies, 19, 41-46, 120. Peers, Subtle Bodies, 46-48. Mößner and Nauerth, ‘Koptische Texte und ihre Bilder’, 307 and 328.

146

korshi dosoo

Fig. 18. Figure from lower part of P.Heid.Inv.Kopt. 518, perhaps depicting a six-winged angel. Image created by K. Dosoo.

Fig. 19. Image from lower part of P.Gieben Copt 1, perhaps depicting an angel. Image created by K. Dosoo.

heathen serpents and wingless angels

147

be understood as wings. In P.Heid.Inv.Kopt. 518 (Fig. 18),104 for example, the striated “arms” follow the pattern of the wings of seraphim – one pair below the body, one above, one alongside – and the figure has a distinct pair of unstriated arms folded across its body. Similarly, the figures on the front and back of P.Heid.Inv.Kopt. 679 (Fig. 17)105 have two pairs of arms emerging from their heads, one curled straight down, one curled closer to their body; the third pair is perhaps lost, or miniaturised as the “tubes” going into their crowns. The accompanying spells would lead us to assume that the depicted figure of the first of these is the Archangel Michael106 – a conclusion also reached by Horak, who cites the “snakes” as additional evidence for this107 – while the second adjures “mighty angels” to carry out the curse.108 Alongside their distinctive wings, angels often have short curly hair (Figs. 8, 9, 19, 20 (?)) and/or haloes (Figs. 1, 2, 8, 21; cases of apparent multiple haloes, as in Figs. 1 and 2, are probably schematic depictions of haloes surrounding hair)109 or crowns (Figs. 13, 17, 18),110 again features drawn from their standard iconography, although there are variants – Freer F1908.45.13 (Fig. 15) appears to have an animal-head, perhaps suggested by the images of the animal-headed cherubim (the zōa or “living creatures”) which regularly accompany Christ Pantocrator.111 More regular are the 104

TM 99553, 9th-11th cent. CE. TM 102079, 11th cent. CE. 106 Michael is the only being directly adjured in the surviving parts of P.Heid.Inv.Kopt. 518, l. 44; in ll. 4-5 another being (presumably God) seems to be adjured in a fragmentary section to send both Michael and Gabriel, and several other beings are mentioned, but Michael is clearly the most prominent figure, and apparently the spell’s intended agent. 107 Horak, ‘Christliches und Christlich-Magisches’, 43. 108 P.Heid.Inv.Kopt. 679, ll. 4-5: “I adjure you today, oh mighty angels (ⲛⲓⲁⲛⲅⲉⲗⲟⲥ : ⲛϫⲟⲣⲓ) who stand before him (God?) at all times”. 109 Compare the comments of Gardner and Johnston, ‘I, Deacon Iohannes, Servant of Michael’, 52. 110 Compare the crowns worn by Gabriel in Pierpont Morgan Ms. M.583 fol.8r (see n. 93) and Michael in Ms. M607 fol.1v (895 CE; https://www.themorgan.org/manuscript/214169, 13/03/2021) which resemble triangles or trapezoids with their narrowest points facing downwards. A similar crown, albeit with points like those of the angels in magical texts, is worn by King David in a drawing on M.583 fol. 38v (http://ica.themorgan.org/manuscript/ page/22/214164, 13/03/2021). More detailed examples of such crowns may be seen on Michael and King David in a painting from the Church of the Archangel Michael from Bawit; see H. Rochard, ‘Le culte des archanges en Égypte byzantine et au début de l’époque arabe: Le témoignage des peintures de Baouît’, in A. Boud’hors and C. Louis (eds.), Études coptes XV. Dix-septième journée d’études (Paris, 2018) 117-35, esp. 120, fig. 2. 111 M. de Grooth and P. van Moorsel, ‘The Lion, the Calf, the Man and the Eagle in Early Christian and Coptic Art’, BABESCH 52-53 (1977-1978) 233-45; K. Stevenson, ‘Animal Rites: The Four Living Creatures in Patristic Exegesis and Liturgy’, in M.F. Wiles and E.J. Yarnold (eds.), Studia Patristica 34 (Leuven, 2001) 470-92. 105

148

korshi dosoo

attributes carried by angels; in late-antique iconography, angels are generally depicted with sceptres topped with crosses in their right hands and orbs in their left.112 These may have been misinterpreted at times by artists – as Elizabeth S. Bolman has noted, in the examples from the Monastery of St. Antony at the Red Sea the orb resembles the round eucharistic loaf,113 while in the Encomium of Eustathius the author describes the orb in the hand of Michael as being “a wheel like (that of) a chariot”, suggesting they were unsure of what exactly it was.114 In most cases, angels in magical texts follow this iconography, with staffs in their right-hands (Figs. 2 (?),115 7, 8, 9, 12, 16, 19,116 20, 27 (?)), usually topped by a cross, though often marked by horizontal bars or other forms whose identity is unclear; the designation in one case of the object as a “staff of light” may suggest they were originally rays of light.117 Their right hand often holds a circular object (Figs. 1, 7, 9, 16, 20), and, given the constancy of the orb in iconography, I suggest 112 Examples include Louvre E 16921 (6th cent. CE; http://www.louvre.fr/oeuvre-notices/ pilier-dit-de-baouit 18/03/2021]); the depiction of St. Michael from the Monastery of Abu Makar in the Wadi Natrun (reproduced as Met 48.105.4; the MMA website gives the date as 2nd-6th cent. CE, but a date towards the later end of this range, if not still later, seems likely; https://www.metmuseum.org/art/collection/search/479681 [18/03/2021]); the angels from Cell 1727 of the Monastery of Apa Jeremiah (6th-7th cent. CE; J.E. Quibell and S. Thompson, Excavations at Saqqara (1908-9, 1909-10): The Monastery of Apa Jeremias [Cairo, 1912], plate XXV; the published image is unclear, but the cleaned mural is on display at the Coptic Museum in Cairo); the depiction of Gabriel in Pierpont Morgan Ms. M.583 fol.8r (see n.  92); and that of the same angel from the Bema arch of the Hagia Sophia (9th cent. CE; figure 12 in Peers, Subtle Bodies, 54); the steatite icon from 11th cent. CE Constantinople, no. 60 in E. Coche de la Ferté, L’antiquité chrétienne (Paris, 1958); the depiction of Michael from the Staatliche Museen zu Berlin Prussischer und Byzantinische Kunst inv. 2429A (12th  cent. CE; Bolman and Godeau, Monastic Visions, 134, fig. 8.13); and the image of Archangel Michael from the Monastery of Saint Antony (13th cent. CE; E.S. Bolman and P. Godeau, Monastic Visions: Wall Paintings in the Monastery of St. Antony at the Red Sea [New Haven, 2002] p. x, fig. 4). On the interpretation of the orb, see Peers, Subtle Bodies, 203-04, nn. 13-14. 113 Bolman and Godeau, Monastic Visions, 134. 114 ⲟⲩⲧⲣⲟⲭⲟⲥ ⲙⲡⲧⲩⲡⲟⲥ ⲛⲟⲩϩⲁⲣⲙⲁ (Campagnano, ‘L’Encomio di Michele Arcangelo’, 164.138-139). 115 The image of the central angel is ambiguous; it appears that he has two arms emerging from the left side of his body, but a long cross sceptre can be seen emerging from his right wing, beginning in a small tubular shape which mirrors in position that of his left arm. It seems likely that the copyist altered their original plan – to have one arm at either side of the body holding a cross and sceptre – by adding an extra arm on the left, and filling the shape created for the right arm with omicrons rather than the striations and fingers drawn on the left arm. 116 Note that in P.Gieben Copt 1 the angel’s hand does not actually touch the object which is understood here as a staff, so that it might appear to be part of a separate, larger object. Comparison with other images, in particular the cherubim from Coptic Museum 4959 (Fig. 20), however, suggests that it should be understood as a staff. 117 London Ms. Or. 5525 l. 4: ⲡⲉⲕϩⲣⲁϥⲧⲟⲥ ⲛⲟⲩⲁⲉⲓⲛ.

heathen serpents and wingless angels

149

Fig. 20. Image labelled “Sabaōth the Great Cherubim” from Coptic Museum 4959 (verso). Image created by K. Dosoo.

that this represents its origin, although in many cases it resembles more closely a wreath. Again, in P.London Ms. Or. 5525 it is designated as an “harma of light”,118 interpreted by its editor as a “shield”, deriving from the Latin arma, “military equipment”. Given the relative rarity of Latin loanwords in Coptic, and the regular occurrence of the homophonous harma (“chariot”) in contexts relating to angels,119 I suspect that the creator of this image, like the author of the Encomium, understood the shape, as being (like) the wheel of a chariot.120 In other cases, the angel may be equipped with other implements – in Rossi’s Tractate the angel holds something which might be a feather or a palmleaf, although the accompanying text suggests that it is in fact a sword (Fig.  8).121 Likewise, the angel on the upper part of P.Yale 1791 (Fig. 21)122 118 London Ms. Or. 5525 l. 5-6: ⲛⲉⲕϩⲁⲣⲙⲁ ⲛⲟⲩⲁⲉⲓⲛ ⲉϥϩⲉⲛ ⲧⲉⲕϭⲓϫ; the fact that the object is subsequently referred to with a singular pronoun suggests that it is singular rather than plural, as the possible nu would imply. 119 See, for example, P.Heid.Inv.Kopt. 686, 3: “He (Michael) set his chariot behind him” (ⲁϥⲕⲁ ⲡⲉⲡϩⲁⲣⲙⲁ : ϩⲓ ⲡⲁϩⲟⲩ ⲙⲁⲃ). 120 We might imagine that the original text had something like “the wheel of your chariot of light”, the first part of which could have dropped out in the copying process. 121 Rossi’s Tractate 9.8-12: “You must send to me Gabriel, the angel of righteousness (ⲇⲓⲕⲁⲓⲟⲥⲏⲛⲏ), with his sword unsheathed (ⲃⲏϣ) in his right hand”. Note, however, that the text specifies that the sword be in the angel’s right hand, although the image depicts it in his left. 122 TM 100011, 6th-7th cent. CE.

150

korshi dosoo

holds a trumpet to his mouth, which seems to be blowing out charaktēres; this is drawn from his role in the accompanying formula (and similar texts) as the one who gathers the angels by blowing his trumpet,123 and who is invoked to give the user a good singing voice.124 Interestingly, angels are not usually labelled; unlike the orant figures examined above, their identity is usually indicated iconographically rather than through explicit captions. One exception to this is in P.Heid.Inv.Kopt. 684 (Fig. 22),125 in which, at the end of a love spell calling upon the Gabriel, the archangel is depicted holding the two lovers in his hands like tiny puppets being manipulated.126

123 P.Yale 1791, ll. 44-45: “Harmoseēl the great ruler (ⲁⲣⲭⲱⲛ), who gathers (ⲥⲱⲩϩⲉ) those of the heaven and those of the earth”; cf. London Hay 10122 (TM 99566, 6th-7th cent. CE) recto ll. 14-16: Hōrmosiēl the angel, the one in whose hands is the trumpet (ⲍⲁⲗⲡⲓⲝ), who gathers the angels to the greeting (ⲁⲥⲡⲁⲥⲙⲟⲥ) of the Lord, of the whole council of the Lord”; Investiture of Gabriel (892-3 CE): “I am Hormosiēl, the trumpeter of the aeons of light (Ⲡⲥⲁⲗⲡⲓⲥⲧⲏⲥ Ⲛⲛⲁⲓⲱⲛ ⲘⲡⲟⲩⲞⲉⲓⲛ). I play the trumpet (ⲥⲁⲗⲡⲓⲍⲉ) before the righteous as they go into the city of the beloved one (ⲙⲉⲣⲓⲧ) and inherit its goodness (Ⲁⲅⲁⲑⲟⲛ)” (C.D.G.Müller, Die Bücher der Einsetzung der Erzengel Michael und Gabriel [Louvain, 1962] 67, ll. 19-20); Egyptian Museum 49547 (TM 102068, 9th-11th cent. CE): l. 23-24 “Hail Hōrmosiēl, the one who plays (ⲥⲁⲗⲉ = ψάλλειν) (the trumpet) before the veil (ⲕⲁⲧⲁⲡⲉⲧⲁⲥⲙⲁ) of the Father”; Book of the Resurrection of Jesus Christ (10th cent. CE?): “Harmosiēl, the one with the spirit (Ⲡ ⳰Ⲛⲁ) trumpet”; M. Westerhoff, Auferstehung und Jenseits im koptischen ‘Buch der Auferstehung Jesu Christi, unseres Herrn’ (Wiesbaden, 1999) 136, l. 10). Harmosiel is probably a variant of Harmozel, well known as one of the Four Luminaries of Sethianism, although his appearance here may represent a shared knowledge of the angelic name rather than a direct dependence upon Sethian tradition; see D.M. Burns, ‘Magical, Coptic, Christian: The Great Angel Eleleth and the “Four Luminaries” in Egyptian Literature of the First Millenium CE’, in H. Lundhaug and L. Jenott (eds.), The Nag Hammadi Codices and Late Antique Egypt, (Tübingen, 2018) 141-62. A very similar description is given of the luminary Dauithe in London Ms. Or. 6794, ll. 8-10, although he uses a lute and bell rather than a trumpet to gather the divine council. 124 Compare the description of Michael’s appearance with a trumpet in the Encomium of Eustathius (Campagnano, ‘L’Encomio di Michele Arcangelo di Eustazio di Tracia’, 164.6). 125 Other exceptions include the image of “Sabaōth the Great Cherubim” from Coptic Museum 4959 (verso); Michael on P.Heid.Inv.Kopt. 686 fol. 9r has the names of five of the archangels written below him (ⲙⲓⲭⲁⲏⲗ, ⲅⲁⲃⲣⲓⲏⲗ, ϩⲣⲁⲫⲁⲏⲗ, ⲥⲟⲩⲣⲓⲏⲗ, ⲥⲉⲧ(ⲉⲕⲓⲏⲗ)). While the fact that Rossi’s Tractate has been lost makes it difficult to be sure of the structure of the tableau, the image of the angel appears to be surrounded by various voces magicae, including the names ⲅⲁϥⲣⲓⲏⲗ, ⲙ[ⲓ]ⲭⲁⲏⲗ, ⲭⲁⲙⲁⲣⲓⲏⲗ and ⲁⲃⲣⲁⲥⲁⲝ. The fact that, in two of these instances, the angels seem to be labelled with several names perhaps implies that the single image stands for, or contains the power of, multiple angelic beings, or we might understand that an identical image is to be copied several times to represent each angel. A separate, though perhaps related, phenomenon, which I do not have time to discuss here, is that of writing letters (particularly vowels) on the bodies of the angels; a particularly interesting example is that of London Ms. Or. 5525, which seems to have the name of the spell’s patient (ⲥⲟⲩⲣⲁ) written on its body; cf. n. 55. 126 For a fuller discussion see Dosoo, ‘Zōdion and Praxis’, 23-51.

heathen serpents and wingless angels

151

Wingless, his identity is only clear from the label above his head. Interestingly, his face seems to have become a complex charaktēr, the right side apparently composed of four small circles, the right of series of overlapping lines. This strange decision – to apparently give the angel four eyes – recurs in depictions of the beings known as decans, and, unlike the schematised wings, seems to represent an iconographic habit peculiar to the magical texts.

Fig. 21. Image from the upper part of P.Yale. 1791 depicting an angel blowing a trumpet. Image created by K. Dosoo.

Fig. 22. Figure from P.Heid.Inv.Kopt. 684, apparently depicting the angel Gabriel holding two figures representing human lovers. Image created by K. Dosoo.

152

korshi dosoo

5. Decans and the Three Hebrews The figures I refer to here as decans appear in two manuscripts, Coptic Museum 4960127 and P.CtYBR inv. 1793,128 neither of which is fully published. The text of the first of these is largely complete, and consists of an invocation to an unnamed being to provide any service which the user demands, described in a series of oppositions: “whether good or [evil], whether favour or disgrace, [whether] binding or loosing, whether destruction or giving life, whether gathering-in or scattering, whether establishing or overturning”.129 This invocation is accompanied by a tableau containing three figures, labelled Arkhōn, Abrak, and Lamei, repeated three times. Their identification as decans depends on an invocation in P.Cologne 10235130 whose text extensively parallels that of Coptic Museum 4960, but is more explicit about the identity of the being it invokes; while it does not name him, it calls upon him to come with his two decans, named Arkhōn and Lamei, suggesting that he is Abrak (or his equivalent), the central figure of the tableaux. The second manuscript which contains a depiction of the ‘decans’, is P.CtYBR inv. 1793. While its text is more fragmentary and difficult to read, what legible text survives suggests that it may have been intended for similarly broad purposes.131 The iconography of these figures (Fig. 23) is so striking that they certainly share a history of transmission, although stylistic differences make it unlikely they are fragments of a single text. They appear as groups132 of three orant figures with striated arms and legs, wearing tunicae, whose lower parts have stripes resembling clavi or a tabard. The central figure seems to wear some kind of cloak fastened to his left shoulder, perhaps a chlamys. All three wear caps, and, in the Cairo example, have curly hair, surmounted with small, charaktēr-like antennae, probably the points of their crowns, with two prongs each for the smaller outer figures, three in the case of the larger central figure; this perhaps represents the “crown of 127

TM 100009, 6th-8th cent. CE. 5th-7th cent. CE; See findit.library.yale.edu/catalog/digcoll:2759461 (18/03/2021) for an image and brief description. This text is currently being edited by Michael Zellmann-Rohrer. 129 ⲉⲓⲧⲉ ⲡⲉⲧⲛⲁⲛⲟⲩϥ ⲉⲓⲧⲉ ⲡⲉⲧ[ϩⲟⲟⲩ] ⲉⲓⲧⲉ ⲭⲁⲣⲓⲥ ⲉⲓⲧⲉ [ϭ]ⲁⲓⲟ [ⲉⲓⲧⲉ ] ⲙⲟⲩⲣ ⲉⲓⲧⲉ ⲃⲱⲗ ⲉⲓⲧⲉ ϩⲱⲧϥ ⲉⲓⲧⲉ ⲧⲁⲛϩⲟ ⲉⲓⲧⲉ ⲥⲱⲟⲩϩ ⲉϩⲟⲩⲛ ⲉⲓⲧⲉ ϫⲱⲣⲉ ⲉⲃⲟⲗ ⲉⲓⲧⲉ ⲟⲩⲱϩ ⲉⲓⲧⲉ ⲡⲱⲛⲉ (reading my own). 130 TM 99604, 6th cent. CE: “(…and I invoke you) that you come to me with your two decans, who are Arkhōn and Lamei”, ll. 8-9. 131 The text contains mention of “giving grace” (ⲧⲓⲭⲁⲣⲓⲥ) and “giving him life” (ⲛϥⲧⲁⲛϩⲟⲟϥ), both of which phrases appear in Coptic Museum 4960. 132 P.CtYBR inv. 1793 has only a single representation of the three decans, while Coptic Museum 4960 contains three – two on its papyrus section, one on its parchment section. 128

heathen serpents and wingless angels

153

glory” (ⲡⲉⲕⲗⲟⲙ ⲛϣⲟⲩϣⲟⲩ, l.25) mentioned in Cologne 10235. The central and rightmost figures hold branches and wands in their right hands, and wreaths or orbs in their left, while the leftmost figure seems to hold a sword or knife in his right hand, and a club or torch in his other. Their most striking feature, however, is their faces; like Gabriel, they seem to have multiple eyes – three in the case of the right and central figure, four in the case of the leftmost. In each case, the figures are captioned (right to left): Arkhōn, Abrak, and Lamei, while fragmentary charaktēres, apparently consistent across the three examples, are found above their names.

Fig. 23. Images of decans (?) from P.CtYBR inv. 1793. Image created by K. Dosoo.

Originally, decan, from the Greek dekanos, referred to the 36 asterisms used in Egyptian astronomy to divide the ecliptic; each decan marked the hour by its rising, or later transit, for ten days in a cycle which was repeated annually.133 This meant that each sign of the Zodiac had three associated 133 On the decans see H. Behlmer, ‘Zu einigen koptischen Dämonen’, Göttinger Miszellen 82 (1984) 7-23; W. Gundel and S. Schott, Dekane und Dekansternbilder: ein Beitrag zur Geschichte der Sternbilder der Kulturvölker (Glückstadt, 1936); H.G. Gundel, Sternglaube und Sterndeutung: Die Geschichte und das Wesen der Astrologie (Darmstadt, 1966); O. Neugebauer and R.A. Parker, Egyptian Astronomical Texts, vols. 1 and 3 (London, Providence, 19601969); K. Dosoo, ‘Baktiotha: The Origin of a Magical Name in P.Macq. I 1’, in P. Buzi et al.

154

korshi dosoo

decans, so that we might think that the regular appearance of decans in threes should be understood as deriving from this astrological division.134 In Coptic literature, however, decans are more closely related to the concept of death, a role which ultimately derives from their position as astrological entities who rule fate, and as beings who nightly traverse the underworld. In hagiographies and apocryphal texts, decans are among those beings who claim the souls of the wicked at death, and drag them to hell to torment them with terrible weapons.135 The iconography of the decans here does not seem to rely on these earlier conceptions, however; in Graeco-Egyptian astrology decans were usually described as animal-headed humans, and they are described similarly – as beast-headed creatures wearing dark armour – in Christian texts; a depiction of a decan from a mural in Tebtunis follows this description.136 Instead, the decans here draw upon the imagery of three orant figures used to represent the Three Hebrews (Plate 3); interestingly, when decans are regularly mentioned in other Coptic magical texts, their names often seem to be related to those of the Three Hebrews.137 Alongside this basic iconography, their superhuman status is indicated through their angel-like attributes (staff and orb), their crowns, and their multiple eyes. This latter feature is, to the best of my knowledge, almost unparalleled in Christian art, although we should note that multiple eyes are found on the wings of cherubim, following their descriptions in Ezekiel and Revelation as “many eyed”,138 while God the Father is described (eds.), Coptic Society, Literature and Religion from Late Antiquity to Modern Times (Leuven, 2016) 1237-44. 134 Compare the remarks of M. Weber, ‘Ein koptischer Zaubertext aus der Kölner Papyrussammlung’, Enchoria 2 (1972) 55-63, esp. 60, who understands the central figure in P.Cologne 10235 as an archangel associated with a planet, and the decans Archon and Lamei as standing in for the zodiacal signs (the day and night houses) associated with that planet. 135 See Behlmer, ‘Zu einigen koptischen Dämonen’; Dosoo, ‘Baktiotha: The Origin of a Magical Name in P.Macq. I 1’. 136 See C.C. Walters, ‘Christian Paintings from Tebtunis’, Journal of Egyptian Archaeology 75 (1989) 203-04, plate XXIX.1. 137 See, for example London Ms. Or. 6794 (l. 47), where the decans are named as ⲓⲁⲕ ⲙⲉⲓⲁⲕ ⲥⲉⲙⲓⲁⲕ, and London Hay 10391 (TM 100015, 6th-7th cent. CE,) l. 21 where they are called ϣⲁⲕ ⲙⲏϣⲁⲕ ϣⲁⲭⲁ. In both cases, the –(ⲓ)ⲁⲕ terminations are reminiscent of the names of the Hebrew Youths in Daniel 3.22 (ⲥⲉⲇⲣⲁⲕ ⲙⲓⲥⲁⲕ ⲁϥⲧⲉⲛⲁⲕⲱ). For a fuller discussion see Dosoo, ‘Zōdion and Praxis’, 16. 138 See Ezekiel 1.18, 10.12; cf. Revelation 4.6-8, 5.6, where the Living Creatures and the Lamb are described as many- and seven-eyed respectively. Depictions of cherubim covered in eyes are abundant in Christian art from Egypt and elsewhere; see for example the cherub holding Macarius’ wrist and the depiction of the Four Living Creatures from the Deesis Chapel, both from the Monastery of St. Antony (13th cent. CE) in Bolman, Monastic Visions, 50, figure 4.20, 72 figure 4.38. Compare the Investiture of Gabriel, in which the angel Leleth

heathen serpents and wingless angels

155

as having seven eyes in Rossi’s Tractate;139 such models might have inspired this particular indication of numinosity. The specific traits of the individual decans may be derived from their purpose in this spell, which is phrased in terms of positive and negative pairs such as “destroying” and “giving life”. Arkhōn, the decan on the right, holds a palm branch and an orb, attributes appropriate for an angelic being, while Lamei, on the left, seems to hold weapons – a club and a sword; it seems reasonable to understand them as manifestations of the positive and negative power of Abrak, the central, singular figure invoked by each of the spells. A final example is Dauithea, one of the four luminaries of Sethian “gnosticism”,140 who appears on London Ms. Or. 6794,141 a spell for a good singing voice, in which he is described as a decan (Fig. 24).142

Fig. 24. Image of Dauithea from London Ms. Or. 6794. Image created by K. Dosoo.

His round body is particularly distinctive, and Kropp has suggested that this, and the eight-pointed star on his belly, are indications of his astral is described as having a body covered with eyes (Ⲉⲣⲉⲡⲉϥⲥⲱⲙⲁ ⲧⲏⲢ ⳰Ϥ ⲣⲏⲧ Ⲛⲃⲁⲗ; Müller, Die Bücher der Einsetzung der Erzengel Michael und Gabriel 66, ll. 25-27). 139 “I invoke you, Gabriel, by the seven eyes (ⲡⲥⲁϢ ⳰Ϥ ⲛⲚⲃⲁⲗ) of the Father”, Rossi’s Tractate 18.11-13; the following text gives the names of eyes followed by seven ‘amens’. The concept that God has seven eyes probably derives from the image of Jesus as the seven-eyed lamb in Revelation 5.6. 140 See Burns, ‘The Great Angel Eleleth’, 141-162. 141 TM 100017, 6th-7th cent. CE. 142 “…and he (the Father) shall send me that great decan, mighty in his power, who is Dauithea”; London Ms. Or. 6794, ll. 26-28.)

156

korshi dosoo

character as a decan;143 there is no parallel known to me of depictions of decans as round-bodied beings, however – they are more often therianthropomorphs. Like the Three Hebrews in their clearest depiction in a magical text (Plate 3), and a few of the angels, including the Powers who accompany Michael (Fig. 2), Dauithe’s head is surmounted by a cross, a strange depiction of his sanctity, albeit one paralleled by depictions of angels in non-magical contexts.144 These crosses are likely an indication of the figure’s divine power – again, in the Encomium of Eustathius the heroine Euphemia recognises the Devil, disguised as an angel, by the absence of a cross on his staff, which she compares to the seal of an emperor, carried by his servants as proof of their delegated authority.145 In the text, Dauithe is described as carrying in his left hand his spirit-lute (ⲕⲓⲑⲁⲣⲁ Ⲙ ⳰ⲠⲚ ⳰Ⲁ).146 While this depiction of Dauithe as a musician probably derives from his namesake, King David, it shows some divergences from the usual iconography. Christian depictions of David as musician are generally drawn from the older iconography of Orpheus, and thus depict a seated figure with a lyre.147 Although the triangle below the cross on Dauithe’s head may be an example of a stylised Phrygian cap derived from depictions of Orpheus, his instrument is a late antique lute, not a harp, likely based on a contemporary instrument, albeit stylised as a charaktēr. The lute shows not only six tuning pegs atop its neck, but also four symmetrical holes in its soundboard, a relatively faithful depiction of real lutes from this approximate period.148 The text describes him as having a gold bell (ϣⲕⲗⲕⲓⲗ ⲛⲛⲟⲩϥ) in his right hand,149 although the image would seem to resemble a more generic staff or palm-branch rather than a bell.150 143

Kropp, Ausgewählte Koptische Zaubertexte, vol. 3, 213-14. See, for example, the angel depicted in the Vision of Isaiah from the Christian Topography of “Cosmas Indicopleustes” in Biblioteca Apostolica Vaticana, gr. 699, fol. 72v (9th cent. CE), image in Peers, Subtle Bodies, 56, fig.14. 145 Campagnano, ‘L’Encomio di Michele Arcangelo di Eustazio di Tracia’, 154.27-156.17. The Devil tries to deflect Euphemia’s suspicions by claiming that artists simply added crosses to beautify their pictures. 146 London Ms. Or. 6794, l. 8. 147 See, for example, figures 1, 4 (from Cyprus), 5-11 in T.E.A. Dale, ‘The Power of the Anointed: The Life of David on Two Coptic Textiles in the Walters Art Gallery’, Journal of the Walters Art Gallery 51 (1993) 23-42, fig. 1; cf. L. Török, Transfigurations of Hellenism: Aspects of Late Antique Art in Egypt. AD 250–700 (Leiden, 2005) 346-47; J.B. Friedman, Orpheus in the Middle Ages (New York, 2000) 147-55. 148 See, for example, the lute preserved at the Metropolitan Museum of Art (Met 12.182.44, 3rd-5th cent. CE). Image available on-line at metmuseum.org/art/collection/ search/473395 (19/03/2021). 149 London Ms. Or. 6794, l.7. 150 Compare P.Macq. I 1 4.9-10, in which Dauithea is described as having “the golden palm branch” (ⲡⲓϥⲁ ⲛⲟⲩⲃ) in his hand. 144

heathen serpents and wingless angels

157

Fig. 25. Image of guardians from the upper part of P.BM EA 10122. Image created by K. Dosoo.

6. Guardians in Coptic Magical Texts Another group of beings who recur in Coptic magical texts are the Nine Guardians (ϩⲟⲩⲣⲓⲧⲉ), found in several formularies, two published and one unpublished.151 The fullest description of them occurs in P.Heid.Inv.Kopt. 685 p. 12.1-16.5 (Plate 1), a procedure to create an amulet by copying the figures three times – once onto a piece of paper which is to be worn, and once onto a cup whose contents are to be drunk, and once on a “sheet”.152 The Guardians are said to watch over the Father, Son and Holy Spirit, three assigned to each person of the Trinity, explaining their number.153 While their names vary between their three occurrences, we may note the recurrence of names based on the root Beth- perhaps related to the 151 A fourth example may be that of P.Strasbourg K 233, a very faint, unpublished parchment text. 152 The instruction for the ritual are in P.Heid.Inv.Kopt. 685 16.8-15. It is possible that the three instructions represent alternatives – that is, that the amulet may be worn, drink, or used in another way. 153 P.Heid.Inv.Kopt. 685 13.1-9.

158

korshi dosoo

Greek boētheia (“help”). In this first text, all nine figures have crosses on their heads, and carry two cross-topped sceptres – the first six figures holding one in each hand, the last three apparently holding two in a single hand.154 The bodies of these first six are roughly humanoid, but shaped like rectangular stelae, upon which are written various charaktēres and words, including the names of God and the instructions “guard NN” (ⲣⲁⲉⲓⲥ ⳽⳽); two have the name Jesus Christ written above them, probably indicating, as I have suggested above, the presence of Jesus rather than identifying the figures.155 The final three figures are depicted as birds, with backwardfacing faces, two-earlike crests and long bodies formed from twisting, striated tubes; it is tempting to see these as wings, as in the case of the angels, but their wings seem instead to be the shapes to the right of their bodies, overlapping with the hands which hold their double sceptres. Their long bodies may instead derive from the iconography of peacocks in late antique Egypt, which, along with doves and eagles, are regularly depicted in religious, and particularly funerary, contexts as symbols of paradise, distinguished from other birds by their long tails and crests.156 As Mößner and Nauerth have recognised, the forms of the Guardians here probably follow those of the persons of the Trinity to whom they are assigned – the first six, assigned to the Father and Son, are anthropomorphic, the final three, assigned to the Holy Spirit, are bird-shaped, following the Spirit’s conventional depiction as a dove.157 The second published depiction of the Guardians appears in P.BM EA 10122 (Fig. 25),158 at the top of a damaged rotulus; the spell to which they belong seems to have been lost, but presumably their purpose was also amuletic. All nine figures are broadly anthropomorphic, but with avian heads and broad, distinctly duck-like beaks; it seems here that the bird and 154 These figures are somewhat difficult to interpret; they seem to have a single arm and single wing coming out of their bodies on the same side; perhaps the wing and arm together represent a single limb combining the characteristics of both, or perhaps they represent the two, separate, forelimbs of the bird-like figures, each of which holds one of the crosses. Since they are depicted in profile, perhaps we should understand them as being angel-like, with two arms and two separate wings, with only one of each visible. 155 It might be tempting to think that the guardians labelled “Jesus Christ” (Ⲓ ⳰Ⲥ Ⲭ ⳰Ⲥ) are those who protect him, but in fact the accompanying text (13.1-9) tells us that while one of them (ⲅⲁⲣⲛⲃⲓⲟⲩⲑ) does protect Jesus, the other (ⲙⲁⲛⲓⲝ) protects the Father, and neither of the other two guardians who protect Jesus has his name written beside them. 156 For a brief discussion, see L. Evans, ‘Animals in Coptic Art’, Göttinger Miszellen 232 (2012) 63-73, esp. 69. 157 Mößner and Nauerth, ‘Koptische Texte und ihre Bilder’, 344. 158 TM 99566, 6th-7th cent. CE.

heathen serpents and wingless angels

159

Fig. 26. Image of guardians taken from an unpublished parchment in a private collection. Image created by K. Dosoo.

human attributes of the previous example have merged, or perhaps not yet been distributed between the guardians of the different persons of the Trinity. Several of them, like the birds of the first group, look backwards, and the bodies of the first six are marked with various shapes, although in this case they resemble an ecclesiastical garment such as a mandyas (mantle); all of them hold a single cross-topped staff. The final example is from a codex from a private collection, to date unpublished (Fig. 26).159 While the published image only shows two of the guardians, there seems to be little doubt it belongs to the larger construction; a bird-headed figure, wearing a mantle, holds an L-shaped staff, looks backwards, and has “Sabaoth, guard NN” (ⲥⲁⲃⲁⲱⲑ ⲣⲁⲉⲓⲥ Ⲇ ⳰Ⲇ) written on its belly; below it, the bird-like head of a being labelled Betha can be seen, topped with a cross. While the origin of the Guardians is obscure, we can see that their iconography draws upon that of the broader culture – the association of birds with angels, and particularly with the Holy Spirit, and the use of ecclesiastical costumes and crosses to indicate holy power and authority. 159 For an image of part of this text, see D. Minutoli (ed.), Inedita offerti a Rosario Pintaudi per il suo 65° compleanno (P.Pintaudi) (Firenze, 2012) pl. lxv, fig. 2.

160

korshi dosoo

7. ‘Ethiopians’ in Coptic Magical Texts Coptic Museum 4959 and 4956B,160 although imperfectly published, contain one of the most interesting image constructions to be found in the corpus of magical texts, associated with the second text on the recto of the former (Figs. 27 and 28). The manuscripts call upon a figure called Aknator the Ethiopian (ⲁⲕⲛⲁⲧⲱⲣ ⲡⲉϭⲱϣ). While the purpose of the spell is unclear, Aknator is a sinister figure, perhaps invoked in a divinatory procedure; identified “the messenger (of all creation)”,161 he is described as “the one whose head is in heaven while his feet are on the earth, whose forms (sic) are in the Tartarus of Amente”.162 160 Both manuscripts are formularies in the form of rotuli; Coptic Museum 4959 (TM 100008, 9th-11th cent. CE) seems to contain several recipes, written in at least two hands, across its verso and recto, while 4956B seems to be written in a single hand, and to contain only the invocation of Aknator. A translation of 4959 appears in Meyer and Smith, Ancient Christian Magic as no. 119, along with part of the image of Aknator from fragments 4-5; an image of its upper part appears in A.S. Atiya (ed.), The Coptic Encyclopedia vol. 5, 1501. 4956B is entirely unpublished, but it is described in Meyer and Smith, ‘Invoking Aknator the Ethiopian’, and Smith seems to have restored parts of 4959 by reference to 4956B. While this fascinating text certainly deserves a fuller treatment, I have been able to use the published images of 4959 and photographs of 4956B to improve upon the readings of Smith, and to provide a more accurate copy of the image of 4956B; since the texts are currently unpublished, the line numbers provided will be my own, and may vary from those in a future publication. 161 Coptic Museum 4959 fr. 1 ll. 12-13: “I adjure you, Aknator the Ethiopian, the “messenger” (ϯⲡⲁⲣⲁⲕⲁⲗⲉⲓ ⲉ[ⲡ]ⲟⲟⲩ · ⲁⲕⲛⲁⲧⲱⲣ ⲡⲉϭⲱϣ ⲡϥⲁⲉⲓϣⲓⲛⲉ); Coptic Mueum 4956B l. 27: “[Ak]anator the Ethiopian, the messenger of all (creation)” ([ⲁⲕ?]ⲁⲛⲁⲧⲱⲣ ⲡⲉϭⲱϣ ⲡⲃⲁⲓϣⲓⲛⲉ ⲡⲧⲏⲣⲃ). 162 Coptic Museum 4959 fr. 1, ll. 13-14: ⲡⲉⲧⲉⲣⲉϫⲱϥ ⲉⲛⲧⲡⲉ ⲉⲣⲉⲣⲁⲧϥ ⲙⲡⲕⲁϩ ⲉⲣⲉⲛⲉϥⲉⲓⲛⲉ ϩⲚ ⲛⲉⲧⲣⲧⲁⲣⲥ {ⲛ}ⲛⲁⲙⲉⲛⲧ⟦ⲁ⟧ⲉ. My reading of this section diverges considerably from that of Smith (Ancient Christian Magic, 241), but makes more contextual sense, and has numerous parallels in Coptic magical texts. Beings who stretch between heaven and earth or the abyss are invoked with almost identical wording in Crum ST 398 ll. 1-4 (TM 100113, 6th-8th cent. CE), P.Mich. inv. 4932 recto ll. 11-14 (TM 99569, 5th-6th cent. CE) and London Hay 10414 verso ll. 16-18 (TM 99562, 6th-7th cent. CE), Carlsberg 52 ll. 17-22 (TM 65321, 7th cent. CE). Such descriptions may go back to theological developments within the Egypt of the New Kingdom; see J. Assmann, ‘Primat und Transzendenz. Struktur und Genese der ägyptischen Vorstellung eines “Höchsten Wesens”’, in W. Westendorf (ed.), Aspekte der spätägyptischen Religion (Wiesbaden, 1979) 7-42; J. Darnell, The Enigmatic Netherworld Books of the Solar-Osirian Unity: Cryptographic Compositions in the Tombs of Tutankhamun, Ramesses VI and Ramesses IX (Fribourg, 2004) 374-82; cf. M. Stadler, ‘Metatranszendenztheologie im Alten Ägypten Pyramidentextspruch 215 und der ramessidische Weltgott’, in S. Ernst and M. Häusl (eds.), Kulte, Priester, Rituale. Beiträge zu Kult und Kultkritik im Alten Testament und Alten Orient. Festschrift für Theodor Seidl zum 65. Geburtstag (St. Ottilien, 2010) 3-31, who argues that similar ideas may be found already in the Pyramid Texts. Note that while the Pharaonic and Roman examples describe the supreme deity, the Coptic texts seem to more often describe powerful demonic beings. For the concept of the “Tartarus of Amente”, the deepest part of hell, see J. Zandee, Death as an Enemy: According to Ancient Egyptian Conceptions (Leiden, 1980) 319-20.

heathen serpents and wingless angels

161

Fig. 27. Image of ‘Aknator the Ethiopian’ from Coptic Museum 4956B. Image created by K. Dosoo.

The image which appears after the text in each copy, presumably depicting the eponymous figure, is strikingly unusual: a large head with large, curly hair, a body with small wings, and a long gown. His wings, his feet, and part of his gown are covered with dense, waved lines, resembling both feathers and his curly hair, but so dense that his wings, feet, and part of his robe, seem to be black. The darkness of his body (and perhaps, of his robe), and his curled hair are probably indicators of Aknator’s “Ethiopian” origin, dark skin and curled hair being considered typical traits of Ethiopians in the Roman world,163 with the term having a broader meaning than today, referring generally to the inhabitants of Africa south of Egypt.164 While the large, dense mass of hair is not typical of depictions of Ethiopians in Roman art, its employment here may be intended to signal differentiation from the shorter curled hair which denoted angels in art and which many Egyptians presumably possessed; more particularly, it may be intended to depict the characteristic haircut of male Kushites, captured in the more life-like Roman depictions, or simply the style we now call an afro.165 163 See F.M. Snowden, Blacks in Antiquity: Ethiopians in the Graeco-Roman Experience (Cambridge MA, 1970) 1-33; L.A. Thompson, Romans and Blacks (London and Oklahoma, 1989) 62-85. 164 D. Selden, ‘How the Ethiopian Changed His Skin’, Classical Antiquity 32.2 (2013) 322-77. 165 Somewhat realistic Hellenistic and Roman depiction of Kushite hair may be found in Snowden, Blacks in Antiquity figs. 40, 42, 44-46, 52-53, 61-62, 72-74, 111-112, 115, 117, 120;

162

korshi dosoo

Fig. 28. Fragmentary image of ‘Aknator the Ethiopian’ from Coptic Museum 4959. Image created by K. Dosoo.

If the depiction of Aknator is unusual, his identity is clearer; he is an infernal being, one of the angels of hell, a demon, or the Devil. From relatively early in Christian history, the Devil was described as black to denote the absence of light, and hence goodness in him, and this description led to an association of the Devil and his demons with real dark-skinned populations. In the Mediterranean world, this meant that demons could be described as Ethiopians, or (outside Egypt) as Egyptians, and take on some of the other physical traits associated with these groups, such as curly hair; the association is so strong that mentioning that a figure is “black” or “Ethiopian” often suffices to indicate their demonic nature in texts.166 While other depictions of demons in Coptic art use black skin as a marker,167 Aknator is unusual in being depicted iconographically as an angel.168 a particularly clear example is Brooklyn Museum 70.59 (2nd cent. BCE; brooklynmuseum. org/opencollection/objects/3790, 19/03/2021). 166 D. Brakke, ‘Ethiopian Demons: Male Sexuality, the Black-Skinned Other, and the Monastic Self’, Journal of the History of Sexuality 10.3/4 (2001) 501-35; G.L. Byron, Symbolic Blackness and Ethnic Difference in Early Christian Literature (London, 2002) 55-103; P. Frost, ‘Attitudes towards Blacks in the Early Christian Era’, The Second Century 8.1 (1991) 1-11; P. Habermehl, Perpetua und der Ägypter (Berlin, 20042) 161-77; J.N. Bremmer, Maidens, Magic and Martyrs in Early Christianity (Tübingen, 2017) 139 (with further bibliography). 167 See for example the depiction of Pakaou binding a demon named Sofonesar, depicted as a winged, black-skinned figure, from the Monastery of St. Antony (Bolman and Godeau, Monastic Visions 48, fig. 4.17). 168 Compare the gown on the image of the angel on the upper part of Coptic Museum 4959 (Fig. 14) with that of Aknator in 4956B (Fig. 27). Strangely, a shape resembling the cross-sceptre held by many angels may have been drawn to the right of Aknator, although the lack of a connecting arm may indicate that it is rather a cross-shaped charaktēr.

heathen serpents and wingless angels

163

The clue to his precise identity may be found in his title “messenger”. Alongside “decan”, this is one of the names used in Coptic literature for the beings who tear the soul from the body upon death, and who punish the wicked in hell; the fact that Aknator is called the “messenger of all creation” may mean that he is the foremost of these, perhaps Death or the personification of Hell (Amente).169 Coptic descriptions of Hell describe these infernal angels in a manner which shows that they were often conflated with demons; their countenance or raiment is “dark”, and they are themselves “dark”, while the angelic ruler of Amente may be described as a dark giant.170 Even if we were to take Aknator as a purely demonic figure, however, he would not be the only depiction of this kind; two images of the Devil as a black-skinned, white-winged angel may be found in the twelfth century Copto-Arabic BNF Copte 13, fols. 9v and 81v.171 Another depiction of an “Ethiopian” demon is found in the second spell of P.Heid.Inv.Kopt. 683,172 a love spell which invokes Salathiel “the Ethiopian of the Four Powers” to inflame the female victim with lust in the same way that he stripped shame from Eve in the Garden of Eden.173 As is clear from the historiola, Salathiel is the Devil himself, who corrupted Eve, and his name probably results from a confusion of the angel Salathiel with Satanael, in one tradition the name held by Satan when he was still an angel.174 The tableau which accompanies this text contains a fairly extreme instance of the abstraction to which Coptic magical art often tends (Plate 4); 169 For the title “messenger” (ϥⲁⲓϣⲓⲛⲉ) see Behlmer, ‘Zu einigen koptischen Dämonen’, 12. For a discussion of the Angel of Death in Egyptian Christianity, see J.-M. Rosenstiehl, ‘Tartarouchos-Temelouchos’, in Deuxième Journée d’Études Coptes. Strasbourg 25 Mai 1984 (Louvain, 1986) 29-56; Bremmer, Maidens, Magic and Martyrs, 275-77. 170 See texts I.1.5, I.1.6, I.1.10, II.1.2, II.1.11, II.1.18, II.1.19 from V. MacDermot, The Cult of the Seer in the Ancient Middle East (Berkeley and Los Angeles, 1971) 579-23. 171 J. Leroy, Les Manuscrits coptes et coptes-arabes illustrés (Paris, 1974) 119, 129, plates 46, 56. Note that unlike Aknator, the Devil in the first of these images is horned, and in the second has various animals – serpents, elephant heads – drawn upon his body to emphasise his monstrous nature. 172 TM 100000, 10 th-11th cent. CE. 173 P.Heid.Inv.Kopt. 683, ll. 23-28: “I entreat you and I invoke you again today, Salathiel, the Ethiopian of the four powers, whose names are Arax, Barax, Sax, Thax, the one who went to Eve while she was within the enclosure (of Eden); the one who deceived her mind until she ate from the tree (and) she stripped naked in pleasure (ἡδονή?)”. 174 The shift from Satanael to Salathiel probably began with the more obvious, and likewise attested, metathesis to Sanatael, followed by a confusion of /n/ and /l/, two sounds with the same point of alveolar articulation, of the unaspirated and aspirated /t/ and /t h/, and of the vowels /a/ and /i/ in an unstressed position; for a fuller discussion see Love, ‘Apa Baula and The Destroyer’; cf. J. van der Vliet, ‘Satan’s Fall in Coptic Magic’, in M. Meyer and P. Mirecki (eds.), Ancient Magic and Ritual Power (Leiden, 1995) 401-18.

164

korshi dosoo

there are five figures, surrounded by charaktēres, each consisting, apparently, of a head on a black torso. The one on the right, the most anthropomorphic, is probably the one to whom the caption Salathiel refers; he wears an angel-like crown with three points, which has a further two, bauble-less points curling outwards from its sides – these might be understood as stylised curly hair. His face is painted a darker shade than those of the other figures, presumably indicating his identity as an Ethiopian, and two shapes rise from his stunted torso, perhaps representing arms. The other four figures, two avian, two still more abstract, perhaps represent the four Powers who are mentioned, but not described in detail in the text. 8. Conclusions Coptic magical texts use images in diverse ways – to evoke supernatural beings, to performatively prefigure their outcomes, to depict beings to be averted. I have argued that in most cases Coptic magical images diverge from Christian art stylistically, but not iconographically; the most commonly depicted figures, angels, follow their standard iconography, and are drawn as winged youths holding a cross and orb, or more rarely as sixwinged seraphim or cherubim. This dependence upon standard models is often missed due to stylistic features, which reduce wings to snake-like tubes, and alter orbs to resemble wreaths. But these texts also draw upon older magical iconography, as in the example of ouroboroi, and develop innovative iconographies, using animal-human hybridity, multiple eyes, and abstraction to suggest the numinous aspects of beings which cannot be fully captured in two-dimensional line drawings. This series of notes cannot claim to provide a definitive theory or interpretative scheme for Coptic magical images; rather it has been intended to draw attention to particular features, and to demonstrate how reading images in conjunction with other examples, with their accompanying text, and with the visual culture of late antique Christianity more broadly may enable us to understand them more fully.

heathen serpents and wingless angels

Plate 1. Pages 12 (left) and 9 (right) of P.Heid.Inv.Kopt. 685 depicting the nine guardians (left) and Mary (?, right). © Institut für Papyrologie, Universität Heidelberg.

165

166

korshi dosoo

Plate 2. The dome of the sanctuary at the Monastery of Saint Anthony on the Red Sea, depicting Christ Pantocrator surrounded by angels and cherubim. [From P. Godeau and E. Bolman; image reference S12; ADP/SA 1999]

heathen serpents and wingless angels

167

Plate 3. Image of the Three Hebrews in the furnace from the Catacombs of Priscilla in Rome (3rd-4th CE); image taken from Wikimedia Commons, public domain (commons.wikimedia.org/wiki/File:Fiery_furnace_01.jpg).

168

korshi dosoo

Plate 4. Image of ‘Salathiel the Ethiopian’ and the Four Powers (?) from the lower part of P.Heid.Inv.Kopt. 683. © Institut für Papyrologie, Universität Heidelberg.

IX. DRAWING AND WRITING. REFLECTIONS ON THE TRANSMISSION OF RITUAL KNOWLEDGE Sabina Crippa

Among the many rituals described in the PGM, a remarkable aspect stands out: a complex and variable relationship between the voice, the written word, and images. This connection is at play in the majority of procedures described in the corpus and assumes a major role in the ritual.1 When it comes to the relationship between these different modalities, this article is principally focused on images that have not been adequately treated by historians of religion.2 1 Regarding the role of the voice in magical rituals see e.g. P. Smith, ‘Aspects de l’organisation des rites’, in M. Izard and P. Smith (eds.), La fonction symbolique. Essais d’Antropologie (Paris, 1979) 139-170, S. Crippa, ‘Entre vocalité et écritures. Les voix de la Sibylle et les rites vocaux des magiciens’, in C. Batsch et al. (eds.), Zwischen Krise und Alltag. Antike Religionen im Mittelmeerraum (Stuttgart, 1999) 95-110. Regarding written words see e.g. G.R. Cardona, Antropologia della scrittura (Turin, 1981) 163-67, D. Frankfurter, ‘The Magic of Writing and the Writing of Magic: The Power of the Word in Egyptian and Greek Tradition’, Helios 21 (1998) 189-221, R. Gordon, ‘Shaping the Text. Innovation and Authority in Graeco-Egyptian Malign Magic’, in H.F.J. Horstmanshoff et al. (eds.), Kykeon. Studies in Honour of H.S. Versnel (Leiden, 2002) 69-111, S. Crippa, ‘Images et écritures dans les rituels magiques (PGM)’, Studi e Materiali di Storia delle Religioni 76.1 (2010) 117-38 at 117, F. Graf, ‘Magie et écriture: quelques réflexions’, in M. de Haro Sanchez (ed.), Écrire la magie dans l’antiquité (Liège, 2015) 227-37. Regarding images see e.g. A. Delatte, ‘Études sur la magie grecque, AKEPHALOS THEOS’, Bulletin de Correspondance Hellénique 38 (1914) 189-249, T. Hopfner, Griechisch-ägyptischer Offenbarungszauber, 2 vols. (Vienna and Leipzig, 19131921), R. Martín Hernández, ‘Reading magical drawings in the Greek magical papyri’, in P. Schubert (ed.), Actes du 26e Congrès international de papyrologie (Geneva, 2012) 491-98, Gordon, ‘Shaping the Text’, 90-96 and C. Sánchez Natalías, ‘Paragraphics and Iconography’, in R. Gordon et al. (eds.), Choosing Magic. Contexts, Objects, Meanings: The Archaeology of Instrumental Religion in the Latin West (Rome, 2020) 103-123. 2 In fact, the term “representation”, in the sense of “image of the divinity”, is absent in the main French and Italian encyclopedias and dictionaries of the discipline. Only Bonnefoy (Y. Bonnefoy, Dizionario delle mitologie e delle religioni [Milan, 1989]) has “imagination” and “mythology”, but not “representation”. Nor do we find it in Di Nola and Bellinger (G.J.  Bellinger (ed.), Enciclopedia delle religioni [Milan, 1989]; A.M. Di Nola (ed.), Enciclopedia delle religioni [Florence, 1970-1976]). In the recent dictionary by R. Azria and

170

sabina crippa

In recent decades, little attention has been paid to the role of drawings in specific ritual and/or “sacred” contexts. This absence reflects a longstanding and prevalent approach in scholarship that is historically and geographically de-contextualized. However, this fact has recently given rise to a new set of methodologies and research questions that are typical of other scholarly disciplines that have promoted the analysis of so-called “sacred” images.3 This work is invaluable for the study of “sacred” images, because it highlights historic changes in the very concept of “space” and even the “body” conceived of as a vessel for practices and roles: body and space have been linked to a new, dynamic concept of the human being.4 Furthermore, scholars have become increasingly aware of the relevance of context, of its presence (or lack thereof) in the rituals themselves and of the “cultural background” to which drawings belong.5 From the viewpoint of Semiotics or Anthropological linguistics, the representation of aspects of ritual or of divinities in their materiality can be understood as a source for different “agentivities”6 which the representation of an object makes present. In the case of cults, the object made through the act of drawing and/or writing of text becomes the focus of the presentification, of the mise en présence of the invisible. Recent scholarly studies have focused their analysis on the status and the function of drawings in ancient cultural contexts, opening up new and D.  Hervieu-Léger; (Dictionnaire des faits religieux [Paris, 2010]), both terms “image” and “imagination” are missing. In P. Bonte and M. Izard (eds.), Dizionario di antropologia ed etnologia (Turin, 2006) a definition for “a system of representation” appears as a set of concepts and values. Conversely, in G. Filoramo (ed.), Dizionario delle religioni (Turin, 1993) although the items “image” and “representation” are missing, but “iconoclastic” and “icon,” with the meaning of painted religious picture, are included. See also D. Albera et al. (eds.), Dictionnaire de la Méditerranée (Arles, 2016). In a historical-religious context, the theme of iconoclasm frequently surfaces, though it is a much narrower field than we might think. In the anthropological field too, as in U. Fabietti, Materia sacra. Corpi, oggetti, immagini, feticci nella pratica religiosa (Milan, 2014) 189-97: even analysing the role of the sacred images as such, this scholar bases his speculation on monotheistic traditions. Consequently,  the concepts mentioned are “truth of faith”, “devotion” and thus iconoclasm and iconophobia. 3 For a look at new research on images in a sacred context during Antiquity, cf. contributions in F. Fontana (ed.), I culti isiaci nell’Italia settentrionale. 1: Verona, Aquileia, Trieste (Trieste, 2010). 4 See J. Rüpke, ‘Individualization and Individuation’, in id. (ed.), The Individual in the Religions of the Ancient Mediterranean (Oxford, 2013) 3-38. Regarding the use of human images in magical practices see A.T. Wilburn, ‘Figurines, Images, and Representations Used in Ritual Practices’, in D. Frankfurter, Guide to the Study of Ancient Magic (Leiden, 2019) 456-506. 5 Gordon, ‘Shaping the text’, 97-107, and Wilburn, ‘Figurines, Images, and Representations’. 6 Cf. the term agentivity in A. Duranti (ed.), Antropologia del linguaggio (Milan, 2005).

drawing and writing

171

promising areas of future research. Based on a fundamentally interdisciplinary approach (e.g. History, History of Religions, Archaeology, Epigraphy, Iconography), two major perspectives have emerged: the problem of image efficacy, and the problem of making the divine visible (i.e. “putting on image” of the divine).7 These academic perspectives have oriented researchers’ interest towards the potential use of images to establish a dialogue in “sacred” contexts with the participants in a ritual practice. In polytheistic cultures, we may think of “symbols” imbued with intentionality or we might speak of the prerogative of the images, which enter into dialogue with the viewer: during the ritual, the images re-activate themselves and come into communion with their viewer. The “sacred” sight, in fact, is a culturally conditioned event: recognising the sacredness of an image is only possible if it has been previously shared, but also if human beings are able to connect these images to many other aspects linked to both the divine and human world. However, we should not limit ourselves to the concepts of efficacy, symbol, or sacredness. Pivotal to this topic are ritual practices in their historical and cultural transformations; furthermore, we should not only analyse the static images found in sanctuaries or written sources alone. An important question to ask that is highlighted by the role of drawings in rituals (of course, taking into account all the baggage involved in the reconstruction of ancient practices) is that of representation itself used as a tool meant to contextually re-establish, re-enforce, evoke and bring out various elements: divine roles, essential ritual objects and figurative representations of specific languages belonging to a cultural community which possess the knowledge and codes to value and understand them. If we keep this in mind and turn to the PGM in particular, we must point out that many specialists have identified the need to create a corpus of images (bodies, objects, amulets involved in magical rituals). That said, an exhaustive study on this subject still does not exist; rather, there are merely various analyses of individual images.8 Nevertheless, in recent times, some researchers, philologists and historians of religions in particular,

7 As pointed out in Fabietti, Materia sacra, 189, M. Merleau-Ponty, L’Oeil et L’Esprit (Paris, 1964) underlined the importance of “matter and sight” to understand the meeting point between the human subject and image in these cases: the problematic fulchrum of the study on drawings is formed by the deep connection between sight and body. See also D. Freedberg, The Power of Images Studies in the History and Theory of Response (Chicago, 1990), esp. ‘The God in the Image’, 27-40. 8 The only study entirely dedicated to images remains Hopfner, Griechisch-ägyptischer Offenbarungszauber.

172

sabina crippa

have stressed the importance of the study of images in their textual and contextual complexity. 1. Images and drawings in the corpus of PGM As Martín Hernández has pointed out, the few studies of drawings from the PGM have seen these representations as mere ornaments or arcane mumbo jumbo, meant to persuade devotees of the alleged power of the spell and, at the same time, provide iconographic evidence for the weirdness of magical rituals.9 Martín Hernández suggests, instead, analysing different kinds of images from the PGM with the purpose of stressing the necessity of an integrated study of text and image in magical sources. The author, indeed, assumes that a fragmentary papyrus that contains just an image or a vowel sequence in a triangle cannot lead to the identification of the type of text accompanying it or which it is taken from. Images should be integrated in far more extensive research of the magical context, precisely because they can provide a more detailed and credible analysis of these sources. With this purpose in mind and arguing that it is necessary to produce a more exhaustive catalogue and classification of images connected to the text itself so as to better address the myriad problems that arise while editing these sources, Martín Hernández advocates for a four-part classification of magical drawings, depending on the function that they play in the text: 1) images that act as a representation illustrating the instructions given in the text; 2) images that activate the purpose of the specific ritual practice; 3) magical words written in specific formations with a performative value; 4) arcana symbola concerning mystical/mysterious or esoteric knowledge.10 In her opinion, the magical images analysed demonstrate the need for accurate research into their use in order to acquire a deeper understanding of Graeco-Egyptian magic. Furthermore, they inform us about the different kinds of magical books and may help in the interpretation of a particular spell when the written portion defies interpretation. Contextualizing images in relation both to the text and the complexity of traditions underlying the various contents of the PGM, Richard Gordon offered some new hypotheses and orientations in a 2002 article. Above all, he considers that the connection between word and image in the PGM reveals their dependence on Egyptian tradition because they are “ways of 9 10

Martín Hernández, ‘Reading magical drawings’. Martín Hernández, ‘Reading magical drawings’, 491.

drawing and writing

173

reinventing in a Greek medium the irrecoverable ‘total textuality’ of hieroglyphic rites, with their fusion of writing, drawing and showing.”11 In the light of this interpretation, he also suggests that the visual aspect of both sound sequences and images shows that “[t]he great majority are signs invented ad hoc, on the basis of a knowledge of Greek, hieratic and demotic. The practitioners learned how to apply a small number of techniques of ‘Verfremdung’ to the basic thirty shapes, but they were also free to make up their own signs.”12 Starting from the Egyptian conception of semiotic image/text resources, Pascal Vernus provides another relevant reflection on the interpretation of the image-text link in ritual function: he suggests a reading of different types of relationships between text and representation.13 In the Egyptian tradition, the role of the hieroglyphs was paramount, through the iconic affinity of writing and the discursivity of the image. This tradition enriched these two techniques of re-enacting reality through the development of a combination of texts and images, which nevertheless still depended on the conceptions that guide each expressive mode.14 Vernus’ erudite study and proposed typologies for the relationships between the oral (and then written) text and image constitute a first substantial look at the corpus of drawings in the PGM. Beyond the charaktēres and signs,15 the multiplicity and variety of these images are evident. At first, they may appear as a representation of a ritual action or of an object: from depictions of doors or tools (not to mention the animal pictures frequently found on amulets) to the figurative but simple representations of a ritual itself. From Vernus’ analysis of the text/image relationship, it is possible to recognize two main branches in the PGM: not only does the discourse lead to creating an image (of the victim’s body, of a divinity, or a ritual gesture), but writing becomes an image in and of itself. In this second case, the text becomes an image, because the writing takes on the shape of an animal, 11

Gordon, ‘Shaping the Text’, 96. Gordon, ‘Shaping the text’, 90. Most of these images were absent in the first “magical” Egyptian sources. 13 P. Vernus, ‘Des relations entre textes et représentations dans l’Égypte Pharaonique’, in A.M. Christin (ed.), Ecriture II (Paris, 1985) 45-66. 14 Cf. J. Winaud, ‘L’image dans le texte ou le texte dans l’image? Le cas de L’Egypte ancienne’, Visible 2, 2 (2006) 141-58, where the author mentions how, in the Egyptian tradition, an image without text was far more an exception than not, and that “writing games” reached their peak in the Greek/Roman age. 15 On charaktēres and signs, see in particular Frankfurter, ‘The Magic of Writing and the Writing of Magic’. 12

174

sabina crippa

human or object. The most common case is certainly the use of tablets or engraved amulets. These engraved/transcribed voices carry out the goal of the magical act through writing: the amulet is a tool that enables one who seeks protection to enjoy the magical efficiency of the formula without saying it, but by exhibiting it, carrying it, that is, by letting the writing do its own work. The individual personally participates in the magical mechanisms “activated” by the writing. And this is the case with healing texts, such as this amulet used to protect Elena from all diseases:16

Fig. 1. Image apud GMPT (1986)

The association between writing and image can frequently be seen in animal shapes, among which the snake is quite important. In PGM VII 579-90, the writing is arranged into a circle that forms the shape of a serpent containing the phylactery’s text:

16 Amulet dating back to the 3rd century CE, origin unknown: cf. R.W. Daniel, ‘Mich. Inv. 6666: Magie’, ZPE 50 (1983) 147-54.

drawing and writing

175

Fig. 2. PGM VII. Image created by R. Martín Hernández

In PGM VII 300, in contrast, the writing draws a serpent including the figure of the so-called Ibis:

Fig. 3. PGM VII. Image created by R. Martín Hernández

176

sabina crippa

In some other cases, the user is left in the dark when it comes to the text; the communication flow leaves out the subject and the inscription acts autonomously. It could be produced again, applied, touched, and even swallowed. As PGM IV shows, it is necessary for the ritual participant to engrave the sounds and swallow them or at least put them under his/ her tongue before saying a formula, as is made clear in the following instructions: Once the stone is inscribed and consecrated put it under your tongue, turn it over and pronounce the spell. (PGM IV 1745)

In light of the above-mentioned considerations, it is fruitful to recall a study of Andrei Lebedev who, from the analysis of two graffiti from Olbia, suggested that the circular shape reveals the nature of these texts,17 certainly “devotiones”,18 suggesting in this way an association between the shape of the image and the typology of the ritual action inscribed or drawn. Considering this, these studies remain fundamental for the interpretation of individual drawings. The results of this work only go to show that we should be attuned to both the presence of these images and the role that they play in magical ritual praxis. The focus of this work, therefore, is to contribute to the growing body of scholarly research by offering interdisciplinary considerations about the image in relation to the ritual word and related praxis, on the one hand, and in relation to the study of knowledge traditions and their transmission in a ritual context, on the other. First of all, it should be remembered that in PGM it is explicitly stated that the purpose of the expressive modalities described therein is to facilitate communication with the divine, which is the essential step for the successful realization of ritual practice. Therefore, the magical context of these papyri must be taken into consideration: it is a communication process meant to identify the set of shapes, linguistic variety and images that the ritual practitioner disposes of in the PGM corpus. To study it, we must take into account not only the linguistic objects and isolated drawings 17 On the sacralising efficacy of “circle writing”, cf. M. Pastoreau, ‘L’écriture circulaire’, in R. Laufer (ed.), Le texte et son inscription (Paris, 1989) 15-21. In Western tradition, we could speak of true “circle writings” from the first coins dating to the 5th century BCE. There are many evocative texts, among which formulas and/or evocations are present. On the sacralising power of writing/images, cf. P. Vernus, ‘Support d’écriture et fonction sacralisante dans l’Égypte pharaonique’, in Laufer, Le texte et son inscription, 23-34. 18 Cf. A. Lebedev, ‘The Devotio of Xanthippos. Magic and Mystery Cults in Olbia’, ZPE 112 (1996) 279-83. For an in-depth study on this subject, see also S. Crippa. ‘Images et écritures dans les rituels magiques’, Studi e Materiali di Storia delle Religioni (2010) 117-38.

drawing and writing

177

themselves, but also the entire system that includes all the cultural elements required for the communication process to be successful. Far from being homogeneous and original compositions, these books not only attest to a great diffusion of such practices, but also inform us of the date of the collection, its organisation and, in particular, the cementing of a ritual tradition and body of knowledge after a long period of performance. These handbooks – codes used by “professionals” who would swap these volumes between themselves, send them to the kings of Egypt as letters or transmit them secretly from father to son – present themselves in the form of recipes with all the peculiarities that this textual genre implies: a plurality of addressees, authorial anonymity, time and space distance among agents, the possibility of re-activation of the text and so on. Independently of their purpose and composition, all the recipes found in the PGM, as well as in Egyptian magical texts,19 reveal the presence of three essential elements: praxeis, logoi and, very often, a material tool. Even if praxeis and logoi can alternate with one another in the texts, not only does every aspect of the practical ritual suggest its vocal/oral performance20 but, in different cases, drawings or signs and charaktēres of particular relevance are present in the recipes.21 Instead of reducing the components of “magical” rituals to an opposition between legomena/dromena that has given way to the association between PGM and mystery cults22, it seems more plausible and fruitful to see how the recipes’ instructions compromise a communication strategy that employs various code systems: linguistic, gestural23 and, in particular, a whole panoply of phonic and graphic resources to create, as mentioned above, the communio loquendis cum diis.24 PGM recipes are structured on two levels of enunciation: on the one hand, a complex structure of prescriptions and instructions for the correct fulfilment of the magical praxis; on the other, lists of sound sequences and

19 Cf. R.K. Ritner, The Mechanics of Ancient Egyptian Magical Practice (Chicago, 1993) 35-41; cf. Y. Koenig, Magie et Magiciens dans l’Égypte ancienne (Paris, 1994) 82-85. 20 Cf. PGM IV 475-750; PGM VIII 1-52; 52-63. 21 See e.g. PGM XXIIa 9-10; 11-14; 15-17; XIXb; XX or XLII. 22 The first and most often adduced is A.-J. Festugière, ‘La valeur religieuse des papyrus magiques (Excursus E)’, in id., L’idéal religieux des Grecs et l’Évangile (Paris, 1932) 295-96. For an argumentative discussion, cf. J. Harrison, Themis. A Study of the Social Origin of Religion (Cambridge, 19273) 42-45 and 328-31, and the critical response by C. Calame, Thésée et l’imaginaire athénien. Légende et culte en Grèce antique (Lausanne, 19962) 20-27. 23 For example, to stretch hands or bend knees: PGM III 622; IV 1345. Cf. PGM I where more ritual gestures are present, e.g. lines 58 and 118. 24 Apul., Apol. XXVI, 6.

178

sabina crippa

names that the ritual operator (or whoever is addressing the scribe or physician to obtain some therapeutic outcome) must pronounce depending on the purpose of the ritual action. Instructions are not limited to techniques for properly repeating spells in a consistent manner; instead, they form a truly sophisticated mode d’emploi that includes various elements that are deeply interconnected: the recitation of vowel sequences, prescriptions for the use of specific objects, information on the production of tools (e.g., surgical instruments),25 but also hymn intonation, prayers, and the production on papyrus of certain types of images that differ in appearance and function. Among the instructions, we find orders for the editing and subsequent use of the drawings, which, as Vernus, Gordon, Martín Hernández and others26 have already suggested, exist within a tightly connected relationship between the voice and written word. These considerations have led me to formulate a hypothesis about images in a ritual context. I have done so on the basis of fundamental philological and papyrological works and also thanks to the research on other ritual communities in which the image plays a predominant role within a written text intended for oral recitation. This hypothesis is based on the consideration that in this perspective the image may not simply be a functional tool for isolated ritual performances. If we consider, then, that text or image are not isolated entities and interpret the drawings as elements within a broader communication process in which they play different functions, a first argument would be the following: establishing the relationship between image and text in a given document (or object), based on linguistic and anthropological studies,27 helps to understand the relationship between spoken word and image in the context of ritual enunciation. The presence of a strict voice/oral word relationship which is then transcribed on papyrus leads us to think of the image as a “third level of enunciation,” involving the ritual identity of the practitioner, as much as the exercise of the cultural memory that is essential for the transmission of knowledge. The enunciation of the transmission of knowledge of the image

25

Cf. infra, note 35. Vernus, Des relations entre textes et représentations; Gordon, ‘Shaping the text’; Martín Hernández, ‘Reading magical drawings’. 27 Particularly in anthropological studies on pre-Colombian cultures, the communicative procedures between images and memory constitute the privileged label of the ritual word. For a bibliography see C. Severi, Il percorso e la voce (Turin, 2008). On the relationship between discourse and art, see J. Sherzer, Linguaggio e cultura. Il caso dei Kuna (Palermo, 1987), passim. 26

drawing and writing

179

requires a redefinition of the ritual practitioner’s complex identity (whether scribe or oral performer in a praxis) in the light of a communicative process adapted to a particular ritual act. More specifically, this enunciation raises questions about the role of the ritual identity or identities implied in the editing of texts and/or in ritual praxis: that is, to see the image as not only an object of analysis within a textual context, but also as shedding light on the ritual operators who participate in (or are necessary for) an action that is perceived as a magical ritual.28 Such ritual words in relation to drawings play a relevant part in the varied and complex magical reality, the peculiarity of which has been closely observed in an in-depth study by J. Dieleman:29 As “cross cultural texts,” the PGM attest to the existence of a society where scribes, who were interested in different languages and cultures, introduced new procedures by means of using practices of Greek, Roman or other origins in order to endow sources with an appearance that would be appropriate for their specific audience and that would increase the new status of the ritual operator.30 It is probably the new, dynamic codification of ritual and knowledge that is at the origin31 of the specific connections between expressive modalities that constitute the distinctive trait of the PGM’s multiple textual form. Moreover, as a space for the codification of a further level of enunciation, the image takes on value as an encryption of knowledge within the text and then in the ritual praxis of different kinds of knowledge: ritual but also “proto-scientific”, becoming a founding element of mnemotechnical

28 D. Frankfurter, ‘Charismatic Textuality and the Mediation of Christianity in Late Antique Egypt’, in L. Feldt and J.N. Bremmer (eds.), Marginality, Media and Mutations of Religious Authority in the History of Christianity (Leuven, 2019) 47-67. 29 On this definition and analysis, cf. J. Dieleman, Priests, tongues, and rites: the LondonLeiden magical manuscripts and translation in Egyptian ritual (100-300 CE) (Leiden, 2005) 294. For recent studies on the transmission of knowledge in a milieu based on meeting/ conflicts among different cultures, see the articles in S. Crippa and E. Ciampini (eds.), Languages, Objects, and the Transmission of Rituals (Venice, 2017). 30 The books’ peculiarity probably depends on the necessity to oppose the negative conception of magic of Greek origin which is absent in Egypt. Cf. Gordon, ‘Shaping the text’, 71 and especially D. Frankfurther, ‘Dynamics of Ritual Expertise in Antiquity and Beyond: towards a New Taxonomy of “Magicians”’, in P.A. Mirecki and M.W. Meyer, Magic and Ritual in the Ancient World (Leiden, 2002) 159-78. 31 On this dynamic feature characterized by multiple expressive codes in the magical ritual, which is always defined as repetitive and conservative, cf. V. Valeri, s.v. Rito, in Enciclopedia Einaudi (Turin, 1981) 229: “The rite no longer seems a code for transmitting preexistent messages, but a mechanism that enables you to obtain new information. It finally is a conglomerate potentially creating knowledge.”

180

sabina crippa

practice.32 The complexity and heterogeneity of the type and contents of these documents, therefore, help explain the difficulty in making them the object of any accurate scholarly study. Among the different types of knowledge included in the PGM, it is very interesting to remember the series of prescriptions regarding knowledge about nature and the quality of the stones and minerals used for therapeutic purposes. All in all, we can identify a corpus of eighteen stones: generally analysed to underscore the “mystical” value of the drawing inscribed on the stone (whether divine or animal) signifying the presence of the divinity at the scene of the ritual. Thanks to a comparative study among the PGM, the astrological lapidaries and different handbooks of astrological medicine,33 we can uncover the therapeutic value of these stones. The use of these stones, in fact, derives from a base of knowledge including pharmacopoeia, medicine and mineralogy, which attributes a specific series of healing properties to each stone. The instructions in the PGM enable the ritual operator (e.g., the scribe or healer) to grasp the close correspondence between image and support from the therapeutic value that the medical tradition (“iatromathematics”) has identified in every mineral. As far as musical knowledge is concerned, the repetition and the structure of some texts, which are characterized by groups of continuously repeating vowels that follow precise rhythmic stresses, has given rise to musicological interpretations, according to which they can be taken as musical exercises practised in mystical singing academies and consequently integrated into ritual texts. The modulation of some vowel sequences preserved in quite different, yet codified, forms could be, according to Ruelle and Poirée, evidence for the existence of true musical documents.34 From the observation of this knowledge conveyed through the complex textuality of the PGM, the enunciative/mnemonic value of the image begs us to investigate other typologies of ante litteram encyclopaedic books of the time, such as alchemical, astrological, and medical ones.

32 On the relationship between memory and orality and between memory and image in sources mainly characterized by heterogeneity, see G.R. Cardona, ‘Il sapere dello scriba’, in P. Rossi (ed.), La memoria del sapere (Rome and Bari, 1988) 3-28. 33 M.G. Lancellotti, ‘Médecine et religion dans les gemmes magiques’, Revue de l’histoire des religions 218 (2001) 427-56. 34 Ch.E. Ruelle, ‘Le chant des sept voyelles grecques d’après Démétrius et les papyrus de Leyde’, Revue des Études Grecques 2 (1889) 38-44; E. Poirée, ‘Le chant gnostico-musical des sept voyelles grecques. Analyse musicale’, in J. Combarieu (ed.), Congrès International d’histoire de la Musique (Paris, 1901) 28-42.

drawing and writing

181

The PGM recipes, in fact, share many traits with both practical medical books and alchemical sources.35 The prescriptions mentioned above could be more generically connected to a vast and documented religious, pedagogic and, at the same time, scientific literature, as happens for other textual genres, particularly in the fourth century. On the one hand, the same operative strategies are also documented in Galen’s writings dating to the second century: observing the weather, environment, and timing are important part of the practice. The need for an authoritative figure is another shared feature36 as well as the presence of a set of hygienic and precautionary prescriptions. On the other hand, one can often note a textual structure that is commonplace in alchemical texts: (1)  a theological, philosophical, and cosmological prologue, then (2)  the preparation of organic and inorganic ingredients with healing properties and, most of all, (3) a series of images showing the preparation of the tools and their proper use. The most significant case is the massive therapeutic use of minerals which was widespread in ancient pharmacology.37 Furthermore, it is rather interesting to observe the different uses of the representations of objects and explicit tools but also charaktēres, signs and images that cannot be directly linked to a specific function related to the preparation of such objects. The PGM are thus included in a common “practical” knowledge circulating in books with very different uses one from another, far removed from today’s classifications.38

35 On this subject, see I. Andorlini, ‘Prescription and Practice in Greek Magical Papyri from Egypt’, in H. Froschaner and C. Romer (eds.), Zwischen Magic und Wissenschaft. Artz und Heilkunst in den Papyti aus Aegypten (Wien, 2009) 23-34 and R. Thomas, ‘Greek Medicine and Babylonian wisdom: circulation of knowledge and channels of transmission in the archaic and classical periods’, in H.F.J. Horstmanshoff and M. Stol (eds.), Magic and Rationality in Ancient Near Eastern and Graeco-Roman Medicine (Leiden, 2004) 175-85. 36 In the PGM the authority is usually an ancient philosopher (e.g. Democritos or Pythagoras) or an Egyptian priest. In medical treatise, a physician believed to be an inventor of a specific therapy. 37 On the textual form of “notes” of the system of norms and procedures for ancient medical practices, which were widespread but were defined as minor literature because they were not included in the cultural tradition, cf. B. Cavarra, ‘Alchimia e medicina nei testi bizantini’, in C. Crisciani and A. Paravicini Bagliani (eds.), Alchimia e medicina nel medioevo (Firenze, 2003) 1-18; cf. A. Garzya, ‘Testi letterari d’uso strumentale’, Jahrbuch der Österreichischen Byzantinistik 31 (1981) 263-87; cf. C. Crisciani, ‘Il farmaco d’oro. Alcuni testi tra i secoli XIV e XV’, in Crisciani and Paravicini Bagliani, Alchimia, 217-46. 38 An explicit example are the procedures contained in the PGM corpus defined as “iatromagical”, where the names of diseases, healing/medicinal plants, surgical instruments, etc. are given. Cf. M. de Haro Sanchez, ‘Catalogue des papyrus iatromagiques grecs’, Papyrologica Lupiensia 13 (2004) 37-60.

182

sabina crippa

Nevertheless, even if we find references to masters/authoritative figures, philosophical influence and knowledge about the transformation of matter, all of which is common and typical of an age defined as the “first encyclopaedism,” there are still some traits that appear to be distinctive in the ritual practices belonging to the “magic making” in these Graeco-Egyptian sources. Unlike ancient alchemical and medical texts, magical recipes are structured on two levels of enunciation,39 defined by the presence of deictics (now now) informing us of the start of the practice itself, and of super-segmental instructions for the use of the voice.40 As regards the role of the image in particular, sometimes this part of the text is the most important part of the practice itself: quite a lot of papyri exclusively indicate the tabula or the amulet needed in their descriptions of the ritual (via one or more images, cf. PGM VII 589ff.). In this respect, the association might be made with other sources, such as a collection of writings circulating from the second century BCE under the name of Nechepsos, presented twice as a wise man and, at other times, as an astrologer and physician. Like the PGM, this is a corpus put together at different moments throughout the Roman Imperial age and is composed of books that also contain elements of astrological knowledge, medicine, and the so-called “magical” repertory. This corpus also is characterized by a plurality of explicit knowledge and transmitted through a tight relation between the written word, so-called “mystical” signs and astrological images, both objects and bodies.41 It remains quite difficult to offer a synthesis solving the heterogeneity of these traits in order to fully identify an exclusive typology of “magic making” or magical practice from the study of the images. Nevertheless, it is possible to briefly formulate some hypothetical questions. The first question deals with the codification and the transmission of knowledge: is it possible to consider the images in their diversity and purposes as knowledge transmission “procedures” as happened in ancient medicine? Or, seen in the

39 According to the common meaning in Linguistics, i.e. the process of selection of variable elements that accompanies the realization of word performances linking it to the situation, the speakers and the moment: G.R. Cardona, Dizionario di Linguistica (Rome, 1988) 121. 40 Sound sequences are to be uttered just like a psalmody, a murmuring with hardly any audible sound (cf. PGM I 299; IV 745 etc.). 41 For a papyrological and content analysis of the texts, cf. in particular J.L. Fournet, ‘Un fragment de Néchepso P.IFAO inv. 313)’, in H. Melaerts (ed.), Papyri in honorem Johannis Bingen octogenarii (Leuven, 2000) 61-71.

drawing and writing

183

perspective of textual pragmatics42 and in particular that of studies on the enunciation devices of textual “recipes,” would it be possible to think of them as texts dealing with the required construction of the complex relationship between authors and receivers (ritual practitioner, physician, scholar, the one seeking help, the divinity), which constitute an invisible and implicit dialogue43 in these recipes? If these texts confirm the existence of a typology of ritual sources which are “proto-scientific” as well as “magical,” the presence of the image in these texts, even if they have different purposes, suggests the existence of a whole process of transmission of knowledge based on the interaction between image and memory, which often emerges in ritual texts. All of this would further attest to the use of drawing as a link between memory and ritual and/or scientific tradition. A striking example of this oral word/image association that is present in the magical context as an element of enunciation can be seen in PGM XXXVI 1-34. In this image, thanks to the presence of divine names and vowel sequences in relation to the body parts of the represented divinity, we have access to the conceptualization typical of the Egyptian theological

Fig. 5. PGM XXXVI. Image created by R. Martín Hernández

42

Cf. K. Adamzik, Textsorten-Texttypologie (Munster, 1996). An interesting analysis of images as a link between memory and religious faith can be found in Severi, Il percorso. 43

184

sabina crippa

tradition, concerning the physiological nature of the divinity’s sound identity. In this case, by transferring the vocalic/oral sequences of single divinities to the body parts, the drawing allows the transmission of this peculiar ideology of the anatomic/divine field of the Egyptian context.44 2. Conclusion From the questions we have dealt with, we might consider the study of images as a difficult (at least to our contemporary eyes) aspect in the process of grasping the variety, typologies, and modalities of ritual practices in the Ancient Mediterranean. Above all, it singles out a relevant and distinctive trait: the inescapable mixture of elements typical of oral traditions, the transmission and codification of various kinds of knowledge as well as images. These are elements that, in their co-presence, inform the complex communicative process allowing for the identification of a specific ritual praxis. In the magical material from the Ancient Mediterranean (the PGM in particular), it can be seen that ritual praxis builds a communicative process where, besides words and objects, there is a complex use of images, which might be contradictory and not easy to read. These images, though, translate the fundamentals of the oral and proto-scientific tradition. It could be argued, therefore, that not only the word in its ritual context, but also the relationship established between images and words (oral and/or written), allow one to seek and capture the peculiarity of a determined cultural tradition. In this respect, in the specific domain of the comprehension of the ritual tradition of “magic” in Antiquity, different modalities of expression (the oral and written word, image, charaktēres, signs) should be understood as enunciation tools that are equally relevant and important.

44 For an accurate analysis of the concept of the anatomic field and phonic identity of the divine represented in the PGM, see S. Crippa, La voce. Sonorità e pensiero alle origini della cultura europea (Milan, 2015) chapter II, which highlights the role of anatomo-physiological principles which, according to Egyptians, determine language. The application of these principles is possible thanks to the correspondence between body parts and divinities.

LIST OF ABBREVIATIONS

Greek and Coptic papyri are cited following the Checklist of editions of Greek, Latin, Demotic and Coptic Papyri, ostraca and tablets (https:// library.duke.edu/rubenstein/scriptorium/papyrus/texts/clist_papyri.html) Papyri in other languages are cited following the conventions of the speciallity. AE Amulets BM CBd Crum

Crum ST DMG

DT

DTA

GMA

GMPT

L’Année Épigraphique (Paris, 1888–). W.M. Flinders Petrie, Amulets. Illustrated by the Egyptian collection in University College, London (London, 1914). S. Michel, Die magischen Gemmen im Britischen Museum, 2 vols. (London, 2001) 2 vols. The Campbell Bonner Magical Gems Database. http://www2.szepmuveszeti.hu/talismans/visitatori_salutem W.E. Crum, Varia parchment and papyri. Collection in the Griffith Institute. http://archive.griffith.ox.ac.uk/index.php/varia-parchmentand-papyri. W.E. Crum, Short texts from Coptic ostraca and papyri (Oxford, 1921). S. Michel, Die magischen Gemmen. Zu Bildern und Zauberformeln auf geschnittenen Steinen der Antike und Neuzeit (Berlin, 2004). A. Audollent, Defixionum Tabellae. Quotquot innotuerunt tam in Graecis Orientis quam in totius Occidentis partibus praeter Atticas (Paris, 1904). R. Wünsch (ed.), Inscriptiones Atticae Aetatis Romanae, Defixionum Tabellae. Appendix. Inscriptiones Graecae, iii.3 (Berlin, 1897). R. Kotansky, Greek Magical Amulets. The Inscribed Gold, Silver, Copper and Bronze Lamellae. Part I: Published Texts of Known Provenance (Opladen, 1994). H.D. Betz (ed.), Greek Magical Papyri in Translation. Including the Demotic Spells (Chicago, 1986).

186 KYP

list of abbreviations

Kyprianos Database of Ancient Ritual Texts. https://www.coptic-magic.phil.uni-wuerzburg.de/index.php/ manuscripts-search/. LACMA Los Angeles California Museum of Arts. LIM A. Mastrocinque (ed.), Les intailles magiques du department des Monnaies, Médailles et Antiques (Paris, 2014). LIMC Lexicon Iconographicum Mythologiae Classicae. MHNH MHNH. Revista internacional de investigación sobre magia y astrología antiguas. NGCT D.R. Jordan, ‘New Greek Curse Tablets’, Greek, Roman, and Byzantine Studies 41 (2001) 5-46. RE Real-Encyclopädie der classischen Altertumswissenschaft. RIB R.G. Collingwood et al. (eds.), Roman Inscriptions of Britain I: Inscriptions on Stone (Oxford, 1965). SB Sammelbuch griechischer Urkunden aus Aegypten. SD S. Sánchez Natalías, Sylloge of defixiones of the Roman West (Oxford, 2022). SEG Supplementum Epigraphicum Graecum. SGD D.R. Jordan, ‘Survey of Greek Defixiones not included in the Special Corpora’, Greek, Roman, and Byzantine Studies 26 (1985) 151-97. SGG A. Mastrocinque (ed.), Sylloge gemmarum gnosticarum, Bollettino di Numismatica Monografia 8.2.1 and 8.2.2 (Rome 2003 and 2008). SMA C. Bonner, Studies in Magical Amulets chiefly Græco-Egyptian (Ann Arbor, 1950). Supplementum G. Németh, Supplementum Audollentianum (Budapest, 2013). Suppl.Mag. R.W. Daniel and F. Maltomini, Supplementum Magicum. Abhandlungen der Rheinisch-Westfälischen Akademie der Wissenschaften. Sonderreihe. Papyrologica Coloniensia 16.1 and 2 (Opladen, 1990-1992). SV R. Wünsch, Sethianische Verfluchungstafeln aus Rom (Leipzig, 1898). TM Trismegistos Database (www.trismegistos.org). M. Depauw and T. Gheldof, ‘Trismegistos. An Interdisciplinary Platform for Ancient World Texts and Related Information’, in Ł. Bolikowski, V. Casarosa, P. Goodale, N. Houssos, P. Manghi, J. Schirrwagen (eds.), Theory and Practice of Digital Libraries (Cham, 2014) 40–52. ZPE Zeitschrift für Papyrologie und Epigraphik.

INDEXES

Index I: Gods, Demons and Divinities

A Abrak 152, 153, 155 Abrasax/Abraxas 31, 44, 91, 104, 105, 106, 111 Adonai 44, 107 Agathos Daimon 25 Aion 31 Aknator 142, 160, 161, 162, 163 Alpak 127, 128 Anguipede/s XIV, 84, 93, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115 Anna Perenna XIX, 16, 19, 21, 25, 28, 31, 32, 34, 41 Anubis XIII,7, 8, 9, 38, 41, 43, 44, 45 Apis 40 Apopis 10, 11, 13 Archon 154 Ares 19, 38, 63, 71, 72 Arkhon 152, 153, 155 Athena 16 Atum 6 B Baitmo XIII, 51, 58, 60, 61, 62 Bouel 40 Ch Cherubim 141, 144, 145, 147, 148, 149, 150, 154, 164, 166 Chnoubis 64, 72, 85, 86, 96, 106, 136 Christian God 130, 139, 144, 147, 154, 155, 158

C Cynocephalus 106 D Damnameneu 44 Dauithe 137, 150, 156 Dauithea 155, 156 Dimme 59 E Ereshkigal 44, 60, 61 Eros 38 Eulamo 37, 38, 39 Eulamo-Osiris 38 Eve 33, 138, 163 G Gabriel 40, 69, 70, 131, 143, 147, 148, 149, 150, 151, 153, 154, 155 Gallu 60 Gellō 60 H Harpax 127, 128 Harpocrates 58, 106 Hathor 13 Hebrew God XIV, 107 Hecate 16 Hecate-Selene 29 Helios 106, 107 Hephaestus 16 Heracles 82 Hermanubis 43

188

indexes

Hermes 16 Hermes Katochos 24 Horus 40, 58, 107

Ouriel 69, 70 Ousirapis 40 Ousor 40

I Iao 44, 81, 87, 89, 92, 104, 105, 106, 107, 108 Isis 58, 60

Pazuzu 59, 61 Poseidon 16

J Jesus/Christ 31, 129, 130, 136, 137, 138, 139, 141, 142, 144, 147, 155, 158, 166 Jewish god 84, 88, 89, 90 K Khnum 38 Knephis 124 L Lamashtu XIII, 59, 60, 61, 62 Lamei 152, 153, 154, 155 Lamia 60 M Mars Ultor 38 Mary (the virgin) 32, 33, 130, 137, 138, 165 Medusa 25 Mehen 102 Michael 40, 69, 70, 107, 119, 120, 121, 129, 130, 131, 134, 135, 139, 141, 144, 147, 148, 149, 150, 156 Mnevi 40, 42

P

R Raphael 40, 138, 139 Re-Osiris 123 S Sabaoth 44, 92, 104, 106, 107, 149, 150, 159 Salathiel 163, 164 Satan 163 Satanael 163 Seraphim 141, 144, 145, 147, 164 Seth (see also Typhon-Seth) XIII, 7, 9, 10, 13, 38, 39, 40, 41, 42, 43, 44, 45, 48, 87, 126, 127, 128 Sun God 2, 6, 105 T Tantalos 71, 72, 74 the Devil 130, 138, 139, 144, 156, 162, 163 the Guardians 128, 157, 158, 159, 165 the Trinity 157, 158, 159 Titans 14 Typhon-Seth 37, 40, 42, 43, 44, 125, 126, 127, 128 U

N Nehebkau Nut 12

Unique God

70

12

O Osiris 2, 10, 13, 20, 22, 26, 27, 30, 31, 37, 38, 39, 40, 41, 42, 58, 61, 123

Y Yahweh 89, 105, 106, 107, 108 Z Zeus

16

indexes

189

Index II: Index Fontium

A Amulets 135g 43, 40 Anastasy 9 118 B BM 47 BM 76 BM 116 BM 181 BM 230 BM 231 BM 232 BM 304 BM 305 BM 513 BM 517 BM 519 BM 534 BM 536 BM 537

43 93 106 93 91, 111 91, 111 91, 111 96 96 88 97 99 96 83, 94 83, 95

C Carlsberg 52 160 CBd-6 136 CBd-158 38 CBd-854 38 CBd-752 43 CBd-1950 38 CBd-2243 38 Coptic Museum 4956B 142, 160, 161 Coptic Museum 4959 140, 142, 144, 148, 149, 150, 160, 162 Coptic Museum 4960 152 Copto-Arabic BNF Copte 13 163 Crum 23.2.2 126, 127 Crum ST 398 160

DMG 279 58 DMG 283 58 DMG 293 58 DMG 312 58 DMG 340 58 DMG 376 58 DMG 483 84 DMG 494 84 DT 89 16 DT 134 16 DT 135 21 DT 140 42 DT 151 42 DT 155 44 DT 156 27 DT 157 27 DT 163 42 DT 190 21 DT 241 17 DT 245 26 DT 247 17 DT 252 39 DT 253 39 DT 286 51, 52, 57 DT 287 53 DT 288 53 DT 289 54 DT 290 55, 57 DT 291 56, 57 DT 396 52 DT 397 52 DT 399 52 DT 402 52 DT Addendum 304 38 DTAbusina 1 17 DTA 106 16 DTA 107 16

D DMG 174-137 88 DMG 248 58

G GMA 8 65

190

indexes

GMA 29 81, 91, 93 L LACMA 80.202.214 140, 142 LIM 93 64 LIM 94 64 LIM 95 64 LIM 132 63, 76 LIM 232-269 86 LIM 286 84 LIM 334 73 LIM 684 88 LIM 686 88 London Hay 10122 150 London Hay 10391 154 London Hay 10414 160 London Ms. Or. 5525 130, 131, 132, 134, 137, 141, 148, 149, 150 London Ms. Or. 5987 137 London Ms. Or. 6794 131, 132, 150, 154, 155, 156 London Ms. Or. 6795 138, 139 London Ms. Or. 6796 130, 136, 138 N NGCT 9

16

P P.Bas. II Pahl. 1 138 P.Berlin inv. 5026 4, 5 P.Berlin inv. 8323 127 P.Berlin inv. 8503 141 P.Berlin inv. 15990 124, 125, 133 P.Berlin inv. 20911 119 P.BM EA 10018 6 P.BM EA 10122 150, 157, 158 P.Cologne inv. 10235 152, 153, 154 P.CtYBR inv. 1793 152, 153 P.Duke inv. 230 43 P.Heid.Inv. 500 + 501 130, 131 P.Heid.Inv.Kopt. 408 128 P.Heid.Inv.Kopt. 473 126, 127 P.Heid.Inv.Kopt. 518 146, 147

P.Heid.Inv.Kopt. 520 122 P.Heid.Inv.Kopt. 679 141, 145, 147, 148 P.Heid.Inv.Kopt. 681 130, 140, 142 P.Heid.Inv.Kopt. 682 118, 121, 131 P.Heid.Inv.Kopt. 683 163, 168 P.Heid.Inv.Kopt. 684 130, 133, 150, 151 P.Heid.Inv.Kopt. 685 128, 130, 131, 137, 138, 157, 165 P.Heid.Inv.Kopt. 686 119, 120, 130, 131, 137, 144, 149, 150 P.Hermitage Copt. 66 129 P.Köln 8 86 P.Köln Kopt. 3 137, 141 P.Macq. I 1 137, 153, 156 P.Mich. inv. 593 118 P.Mich. inv. 593-603 118 P.Mich. inv. 595v 118 P.Mich. inv. 600r 118 P.Mich. inv. 1294 118 P.Mich. inv. 4932 160 P.Moen. 3 124, 125 P.Monts.Roca inv. 1231 124 P.Monts.Roca inv. 1472 105 P.Oxy. 68.4673 35, 40 P.PalauRib.Lit. I 39 39 P.Rainer 12 48 P.Ryl.Copt. 104 133 P.Stras.Inv.Kopt. 550 119 P.Strasbourg K 204 + 205 120 P.Strasbourg K 233 157 P.Würzburg inv. 42 119, 133 P.Yale 1791 130, 144, 149, 150, 151 PDM xii 7, 8, 43, 124, 125 PDM xiv 44, 121 PGM I 177, 182 PGM II 3, 4, 5, 8, 134 PGM III 43, 44, 50, 122, 125, 126, 177 PGM IV 19, 44, 176, 177, 182 PGM V 122 PGM VII 9, 11, 38, 40, 44, 83, 84, 85, 89, 91, 92, 94, 123, 124, 125, 174, 175, 177, 182 PGM VIII 177

indexes PGM IX 38 PGM XII 8, 40, 43, 48, 91 PGM XIXa 40 PGM XIXb 177 PGM XX 73, 177 PGM XXIIa 177 PGM XXXVI 40, 44, 111, 112, 122, 134, 183 PGM XLII 177 PGM XLVI 43, 48 PGM XLVIII 124, 125 PGM LXXVIII 40 PGM CVI 88, 123 PGM CXXVIa 40, 44 S SB 26.16650 43 SD 48 21 SD 50 16 SD 51 21 SD 56 21 SD 105 16 SD 108 16 SD 179 21 SD 338 16 SD 339 16 SD 345 16 SD 530 16 SGD 124 16 SGD 140 17 SGD 141 17 SGD 157 17 SGD 161 17 SGD 164 39 SMA 180 112 SMA 187 38 SMA 213 89 SMA 271 75 SMA 315 68 Spell 30B 4, 5 Spell 110 3 Spell 125 3 Spell 125c 12

Spell 145 11, 13 Spell 149 11, 12, 13 Spell 151 7 Spell 167 8 Suppl.Mag. 10 88, 91, 98 Suppl.Mag. 42 61 Suppl.Mag. 44 38 Suppl.Mag. 53 38 Suppl.Mag. 57 38 Suppl.Mag. 69 42 Supplementum 15 52 Supplementum 16 52 Supplementum 101 26 Supplementum 166 26 Supplementum 230 51, 52, 57 Supplementum 286 52 Supplementum 287 52, 53 Supplementum 288 52, 53 Supplementum 289 52, 54 Supplementum 290 52, 55 SV 1 28, 42, 47, 125 SV 6 125 SV 9 28, 125 SV 12 28, 42, 46, 125 SV 13 31 SV 14 31 SV 16 26, 30, 38, 40, 46 SV 17 26, 27, 30, 38, 40 SV 18 26, 38 SV 19 31, 38 SV 20 22, 38, 125 SV 21 31, 38 SV 22 31, 38 SV 23 31, 38 SV 24 31, 34, 38, 42 SV 25 34, 42 SV 26 31, 38 SV 27 22, 38 SV 28 22, 38 SV 29 23, 38, 125 SV 30 38 SV 31 38 SV 34 26, 38, 40

191

192 SV 36 SV 39 SV 43 SV 45 SV 49 SV 102 SV 103 SV 104 SV 105 SV 106

indexes 26, 38 38 26, 40 26 38 43 37 37 37 37

SV 107 37 SV 108 37 SV 109 37 SV 110 37 V Vienna K 1992 136 Vienna K 7076 129 Vienna K 7110 129 Vienna K 8302 II 133

indexes

193

Index III : General Index

A agency VIII, X, XI, XIII, 30, 72 amulet/s XI, XIII, 4, 5, 6, 7, 9, 59, 60, 61, 62, 63, 64, 66, 67, 68, 69, 70, 71, 73, 74, 75, 81, 82, 86, 87, 88, 89, 90, 91, 92, 93, 94, 98, 100, 101, 112, 123, 124, 129, 136, 157, 171, 173, 174, 182 angel/s 32, 33, 88, 120, 121, 129, 130, 131, 132, 133, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 154, 156, 158, 159, 161, 162, 163, 164, 166 angelic 14, 135, 150, 155, 163 anthropomorphic 12, 24, 125, 135, 158, 164 archangel/s 33, 40, 119, 129, 134, 141, 150, 154 B binding XII, 16, 17, 20, 25, 30, 33, 34, 38, 39, 43, 152, 162 bird/s 65, 66, 127, 128, 133, 138, 142, 143, 158, 159 blood 4, 64, 65, 66, 67, 71, 72, 74, 75, 130 boat 10, 51, 57, 58, 59, 138 bronze 19, 20, 31, 59, 63, 111 bull 12, 40, 67 burial/s 8, 11, 20, 32, 34, 74 Ch chain/s 15, 23, 33 charaktēr/es XI, XII, XV, 4, 19, 35, 37, 113, 118, 121, 122, 124, 127, 132, 137, 138, 141, 145, 150, 151, 152, 153, 156, 158, 162, 164, 173, 177, 181, 184

charioteer/s XIII, 23, 24, 25, 37, 43, 44, 45, 62, 106 chest 6, 28, 30, 31, 43, 44, 51, 52, 125, 132 Christian art 14, 137, 138, 142, 154, 164 C circus XIII, 37, 38 copper 8 Coptic magical images 14, 132, 134, 135, 136, 164 crocodile 17, 63, 66, 67, 68, 69, 70, 72 crown/s 32, 147, 152, 154, 164 curse tablet/s VIII, XII, XIII, 16, 17, 18, 20, 21, 24, 25, 26, 28, 29, 33, 34, 37, 51, 52, 125, 142 curse/s XI, 12, 16, 17, 19, 20, 21, 23, 25, 26, 27, 28, 30, 31, 32, 35, 36, 37, 38, 40, 42, 44, 45, 57, 127, 133, 141, 147 D defixio/nes XIII, 15, 16, 17, 18, 19, 20, 21, 22, 24, 25, 26, 28, 29, 30, 33, 34, 37, 38, 39, 40, 41, 42, 44, 45, 112 demon/s 12, 20, 33, 35, 36, 37, 38, 39, 41, 42, 43, 44, 45, 51, 52, 57, 58, 59, 60, 61, 62, 66, 69, 73, 83, 86, 87, 88, 89, 90, 91, 120, 125, 127, 128, 138, 141, 162, 163 demonic XII, XIII, 11, 25, 45, 62, 88, 120, 121, 134, 135, 160, 162, 163 diagram/s 35, 83, 84, 87, 143 dog/s 9, 59, 65, 69, 70, 73, 74, 75 donkey head/ed XIII, 10, 42, 125, 127, 128 dragon/s 138 drawing/s VII, X, XII, XIII, 1, 3, 4, 8, 10, 18, 25, 26, 31, 38, 41, 43, 44, 45,

194

indexes 51, 52, 57, 82, 83, 84, 85, 89, 90, 91, 92, 117, 124, 129, 136, 144, 147, 164, 170, 171, 172, 173, 176, 177, 178, 179, 180, 183, 184

evil eye

E 68, 70, 90

F figurine/s VIII, X, 2, 17, 19, 20, 21, 24, 25, 28, 31, 32, 34, 58, 111 G gem/s and gemstone/s X, XI, XIII, XIV, 7, 9, 11, 13, 31, 38, 39, 41, 42, 43, 45, 58, 63, 64, 66, 67, 72, 73, 76, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 99, 103, 104, 105, 106, 107, 108, 110, 111, 112, 113, 114, 115, 124, 136 gnosticism 37, 38, 155 gold 5, 81, 83, 91, 92, 93, 98, 156 H historiola/e 35, 41, 136, 139, 163 horn/s/ed 17, 91, 95, 125, 133, 134, 142, 163 horse/s XIII, 17, 26, 37, 38, 43, 44, 45, 67, 69, 70, 73, 125 horse head/ed 26, 38, 44, 45 I ibis 63, 67, 68, 69, 72, 74, 175 icon/s VII, 30, 107, 110, 111, 113, 129, 130, 131, 136, 148, 170 iconicity XIV, 122 iconography VIII, XII, XIV, 7, 11, 16, 18, 19, 20, 24, 29, 30, 33, 34, 35, 36, 38, 41, 43, 44, 58, 109, 124, 128, 131, 136, 139, 141, 145, 147, 148, 152, 154, 156, 158, 159, 164, 171 illustration/s XII, 1, 2, 3, 5, 11, 12, 14, 86, 143

imagery

VII, X, 4, 33, 34, 154

K katadesmos/oi 15, 16 knot/s 15, 16, 17, 32, 33, 34 L lamella/ae 37, 39, 42, 65, 73, 81, 87, 88, 89, 90, 92 lead X, 19, 20, 24, 27, 31, 34, 42, 73, 112 lead tablet/s 19, 20, 24 lion 8, 59, 66, 67, 82, 106, 124, 138, lion head/ed 64, 72, 81, 84, 85, 86, 91, 96, 123, 124 logos XIII, 39, 40, 42, 43, 44, 69, 84, 88, 89, 91, 99, 126 M magical papyri XV, 1, 3, 5, 6, 7, 10, 11, 13, 42, 52, 85, 117, 122, 124, 132 mummification 30 mummy/ies 6, 8, 9, 27, 30, 31, 40, 41, 58 O orant 133, 136, 137, 138, 139, 150, 152, 154 ouroboros/oi 73, 82, 83, 84, 86, 87, 88, 89, 90, 91, 92, 94, 95, 97, 99, 110, 122, 123, 124, 125, 129, 133, 134, 164 P pattern/s XV, 3, 60, 92, 142, 147 persuasive analogy XII, XIII, 18, 30, 40 phylactery/phylacteries 66, 83, 130, 131, 132, 174 poppet/s 41

rays

R 135, 136, 148

indexes ritual power VIII, X, XI, XV, 5, 33, 45 rooster 106, 108, 109, 110 rooster-headed 93, 103, 108, 111, 112 S sarcophagus/i 8, 20, 26, 40 Seal of Solomon 82, 83, 84, 85, 86, 87, 88, 89, 90, 91 seal/s 69, 70, 83, 84, 87, 130, 156 shield 84, 103, 105, 112, 149 silver 65, 83, 86, 87, 90, 91, 100, 101 snake/s and serpent/s 6, 9, 10, 11, 12, 17, 25, 26, 27, 28, 29, 30, 31, 34, 41, 42, 59, 63, 64, 67, 69, 72, 73, 74, 81, 82, 83, 84, 85, 86, 91, 96, 103, 106, 108, 110, 111, 112, 122, 124, 125, 129, 133, 134, 135, 138, 141, 142, 145, 147, 163, 164, 174, 175 spell/s 1, 2, 3, 4, 5, 7, 8, 9, 11, 12, 13, 17, 18, 19, 20, 25, 33, 34, 38, 39, 40, 41, 42, 43, 45, 60, 64, 65, 66, 67, 70, 71, 72, 73, 74, 75, 126, 127, 129, 130, 135, 138, 139, 141, 147, 150, 155, 158, 160, 163, 172, 176, 178 star 29, 30, 32, 65, 91, 110, 155 statue, statuette 2, 13, 31, 74, 111, 136 stelae XV, 10, 35, 133, 158 sword/s 112, 144, 145, 149, 153, 155

195

symbol/s XI, XII, XIV, 4, 6, 10, 14, 38, 59, 82, 83, 84, 85, 86, 88, 89, 90, 91, 94, 95, 97, 99, 121, 122, 123, 139, 158, 171 T tail 17, 61, 82, 83, 85, 91, 115, 122, 158 tetragrammaton 86, 87, 88 trumpet/s 150, 151 V vessel/s 31, 32, 51, 138, 170 victim/s XII, XIII, XV, 15, 16, 17, 18, 19, 20, 21, 23, 24, 25, 26, 27, 28, 30, 31, 32, 33, 34, 37, 38, 40, 41, 42, 57, 129, 130, 133, 135, 163, 173 vignette/s XII, 1, 2, 3, 5, 7, 8, 11, 12, 13 W weapon/s 43, 44, 122, 125, 126, 127, 128, 154, 155 werewolves 65 whip 44, 69, 103, 106, 111, 126 wing/s XIV, 6, 120, 131, 134, 141, 142, 143, 144, 145, 147, 148, 151, 154, 158, 161, 164 wolf/ves XIII, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75 womb 67, 72, 75

PRINTED ON PERMANENT PAPER

• IMPRIME

SUR PAPIER PERMANENT

N.V. PEETERS S.A., WAROTSTRAAT

• GEDRUKT

OP DUURZAAM PAPIER

50, B-3020 HERENT

- ISO 9706