Introduction: the problem of theurgy

CANASE AN OSBrER maid llinca Tanaseanu-Dobler, Theurgy in Late Antiquity val. \ en Contents Foreword and Acknowledgements . . .… 2.2222 nonurnnnn 7 1. Introduction: The Problem of Theurgy ,....,.......,., 9 2. Theurgy and the Chaldean Oracles. ............,..... 2.1 Preliminaries ...,.,..,......,.....,.,....... 2.2 Sacrifice? .......,..............,.,..... 2.3 Initiation and Ascent. .................,...... 2.4 Epiphanies and Visions .......,....,.......,.. 2.5 Results: Rituals in the Oracles . ............,..... 21: 21 26 29 34 38 2.6 Greal Absents . ..,..,..............,.,..... 39 3, First Debates about Theurgy: Plotinus, Porphyry, lamblichus . . . . 45 3.1 Theurgy and the Circle of Plotinus .........,...... 3.2 Chaldean Oracles and Theurgy in Porphyry ......,.... 3.2.1 The Philosophy from Oracles ...........,..., 3.2.2 The Letter to Anebo .....,............... 45 56 57 74 3.2.3 De regressu animae .....,............... 83 3.3 Theurgy in lamblichus . ..............,....,... 95 3.3.1 De mysteriis . ..,...,..,......,,...,.. 95 3.3.2 The Pythagorising wrilings . ................ 111 3.3.3 Other works and the anecdotes of Eunapius . ....,., 130 4. Theurgy in the Later Fourth and Early Fifth Century ........ 4.1 The Emperor Julian: Theurgy, Philosophy and the Priestly Life. 136 136 42 Salutius ..,..,..,....,....,.,............. 149 4.3 Eunapius ofSardes ......,......,............. 152 4.4 Synesius of Cyrene . ,...,.....,.,.......,..,.. 162 4.5 Hierocles ..........,...,.........,...... 175 5. The Later Fifth Century: Theurgy and Demiurgy in Proclus. . ... 5.1 Preliminaries ...,..,...................... 186 188 5.2 Terminology ...........,........,,,.,... 5.3 Theurgic rituals and themes in philosophical contexts . . . .. 190 199 5.3.1 Divine Visions: The Commentary on the Republic .... 199 5.3.2 Theurgy and Myth. ,...,................ 204 5.3.3 Theurgic Ascent and Immortalisation in the Commentary on the Republic and the Platonic Theology. ........ 207

llinca Tanaseanu-Dobler Theurgy in Late Antiquity The Invention of a Ritual Tradition Beiträge zur Europäischen Religionsgeschichte (BERG) =book Vandenhoeck & Ruprecht Volume 1

Beiträge zur Europäischen Religionsgeschichte (BERG) Edited by Christoph Auffarth, Marvin Döbler, Ilinca Tanaseanu-Döbler Volume 1 Vandenhoeck & Ruprecht

Ilinca Tanaseanu-Döbler Theurgy in Late Antiquity The Invention of a Ritual Tradition Vandenhoeck & Ruprecht

Umschlagabbildung: Bayerische Staatsbibliothek München, Cod. graec. 304, fol. 35r Amention of theurgic burial in Proclus’ Theologia platonica IV,9 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data available online: http://dnb.d-nb.de. ISBN 978-3-525-54020-6 ISBN 978-3-647-54020-7 (E-Book)  2013, Vandenhoeck & Ruprecht GmbH & Co. KG, Göttingen/ Vandenhoeck & Ruprecht LLC, Bristol, CT, U.S.A. www.v-r.de All rights reserved. No part of this work may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without prior written permission from the publisher. Printed in Germany. Typesetting by Konrad Triltsch Print und digitale Medien GmbH, Ochsenfurt Printed and bound in Germany by Hubert & Co, Göttingen Printed on non-aging paper.

Contents Foreword and Acknowledgements . . . . . . . . . . . . . . . . . . . . . 7 1. Introduction: The Problem of Theurgy . . . . . . . . . . . . . . . . 9 2. Theurgy and the Chaldean Oracles . . . . . . . . . . . . . . . . . . . 2.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Sacrifice? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Initiation and Ascent . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Epiphanies and Visions . . . . . . . . . . . . . . . . . . . . . . 2.5 Results: Rituals in the Oracles . . . . . . . . . . . . . . . . . . . 2.6 Great Absents . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 21 26 29 34 38 39 3. First Debates about Theurgy : Plotinus, Porphyry, Iamblichus . . . . 45 3.1 Theurgy and the Circle of Plotinus . . . . . . . . . . . . . . . . 45 3.2 Chaldean Oracles and Theurgy in Porphyry . . . . . . . . . . . 56 3.2.1 The Philosophy from Oracles . . . . . . . . . . . . . . . . 57 3.2.2 The Letter to Anebo . . . . . . . . . . . . . . . . . . . . . 74 3.2.3 De regressu animae . . . . . . . . . . . . . . . . . . . . . 83 3.3 Theurgy in Iamblichus . . . . . . . . . . . . . . . . . . . . . . . 95 3.3.1 De mysteriis . . . . . . . . . . . . . . . . . . . . . . . . . 95 3.3.2 The Pythagorising writings . . . . . . . . . . . . . . . . . 111 3.3.3 Other works and the anecdotes of Eunapius . . . . . . . . 130 4. Theurgy in the Later Fourth and Early Fifth Century . . . . . . . . 4.1 The Emperor Julian: Theurgy, Philosophy and the Priestly Life . 4.2 Salutius . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Eunapius of Sardes . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Synesius of Cyrene . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Hierocles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 136 149 152 162 175 5. The Later Fifth Century : Theurgy and Demiurgy in Proclus . . . . . 5.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Theurgic rituals and themes in philosophical contexts . . . . . 5.3.1 Divine Visions: The Commentary on the Republic . . . . 5.3.2 Theurgy and Myth . . . . . . . . . . . . . . . . . . . . . . 5.3.3 Theurgic Ascent and Immortalisation in the Commentary on the Republic and the Platonic Theology . . . . . . . . . 186 188 190 199 199 204

Contents 5.3.4 Philosophy and Theurgy : The Commentary on the Parmenides and the Platonic Theology . . . . . . . . . . . 5.3.5 Theurgy and Divine Names: Proclus on the Cratylus . . . 5.3.6 Theurgy and Demiurgy : Proclus on the Timaeus . . . . . 5.3.7 Intellectual and Verbal Songs of the Soul: the Eclogae and Proclus’ Hymns . . . . . . . . . . . . . . . . . . . . . . . 5.4 Theurgy in Proclus . . . . . . . . . . . . . . . . . . . . . . . . . 215 237 243 251 255 6. Other Late Neoplatonists on Theurgy and Rituals . . . . . . . . . . 258 7. Conclusion: Theurgy – Writing about Rituals . . . . . . . . . . . . 278 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lexica and databases . . . . . . . . . . . . . . . . . . . . . . . . . . Works Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287 287 291 291 Index nominum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315 Index rerum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319

Foreword and Acknowledgements This volume is a revised version of my Habilitation thesis which was submitted and defended in 2011 at the University of Bremen. The book was begun during a two-year scholarship at the Greek and Latin Department and the Center for the Study of Religion at the Ohio State University, funded by the Deutsche Forschungsgemeinschaft (DFG), from 2008 – 2009. The manuscript was completed during my current work at the Courant Research Centre EDRIS at the University of Göttingen, also funded by the German Research Foundation (DFG) as part of the German Initiative for Excellence. The DFG has generously funded the printing of this book. It is my great pleasure to express my gratitude to all who have accompanied this project from idea to book. At Ohio State I had the privilege to work with Sarah Iles Johnston. I thank her for our many conversations on theurgy and late antique religion, for generously sharing many of her articles with me before they appeared in print, as well as for her patient reading of the manuscript and her helpful suggestions. Thanks are also due to the other two readers, Christoph Auffarth (Bremen) and Heinz-Günther Nesselrath (Göttingen) for their careful reading of the final draft and their valuable suggestions, as well as for our stimulating exchange on late antique religion, philosophy and the history of religions. Besides Sarah Iles Johnston, the conversations with Fritz Graf, Anthony Kaldellis and Michael Swartz allowed me to participate in a vibrant scholarly dialogue on magic, rituals and religion in Late Antiquity, Byzantium and Judaism at the Ohio State University. Without Christopher Brown’s friendship, his warm welcome and helpfulness, and the wonderful discussions on anything ranging from ancient culture to liturgical reform and ritual change, life in Columbus would simply not have been the same. I also profited greatly from discussions with Helmut Seng (Frankfurt), whose indefatigable organisation of conferences on key topics linked with the Chaldean Oracles has furthermore provided an invaluable forum for research on theurgy and related subjects during the last years. Göttingen, November 2012 Ilinca Tanaseanu-Döbler

1. Introduction: The Problem of Theurgy Theurgy is commonly taken to denote a complex of rites which are based on the so-called Chaldean Oracles, a collection of oracles in hexameters, which were probably composed during the late second century AD. These rituals are mostly known through Neoplatonic sources, who engage in a passionate debate about their relevance to the salvation of the soul and thus to the philosopher’s ultimate goal. The Chaldean Oracles and the rituals connected with them are transmitted (at least on a textual level) in Byzantine times and are later discovered by the Renaissance Florentine Platonists. The study of theurgy has been dominated by the question of irrationality versus rationality. Theurgy has been viewed as the intrusion of extra-rational elements coming from a separate religion into the rational pursuit of late antique philosophy, thus affecting its rational character.1 Such a perspective is ultimately indebted to the grand tales of late antique decay.2 It postulates a purely intellectual Plotinus, whose philosophy is combined by a wavering and psychologically complicated Porphyry with the Chaldean Oracles and theurgy, which then gain increasing importance in the thought of later Neoplatonists like Iamblichus or Proclus. This position was challenged in the last decade of the 20th century, when a resurgence of interest in theurgy led many scholars to focus on the ideas underlying theurgic ritual in the Chaldean Oracles and especially in Iamblichus, producing in-depth discussions of the role of theurgy in Iamblichus’ philosophical system and attempting to vindicate his position on this issue as governed by a rationality of its own.3 Due to these studies we are now in a much better position to understand why and how philosophers such as Iamblichus or Proclus pleaded for the importance of such rituals, so that S. Knipe is right to conclude his survey of scholarly approaches to theurgy from 1963 to 2011 by perceiving these studies as a “fatal blow” dealt to the interpretation of theurgy as “an escapist fall into the subrational realm of the ‘occult’”.4 Whether opting for the irrational or the rational character of theurgy, 1 Cf. the telling title of Saffrey’s article “La thurgie comme pntration d’lments extra-rationnels dans la philosophie grecque tardive” (Saffrey (1990b)). This discussion has focused much on the early figures of Neoplatonism, Plotinus being viewed as the exponent of ratio, Porphyry as taking an intermediary stance and finally Iamblichus as surrendering to irrationality and religion (Saffrey (1990 a–c), Smith (1974)). 2 See e. g. Geffcken (1920), Dodds (1951), 236 – 255. 3 See Nasemann (1991), Shaw (1993), (1995) or (2000), Stäcker (1995), Taormina (1999), 133 – 158, Smith (2000) reading Iamblichus as a philosopher of religion, Johnston (2004a). 4 Knipe (2011), 170.

Introduction scholars tend to regard it as an entity originally separate from Neoplatonism and later adopted into its systems,5 and as such as something already determined in its basic structure and features, lending itself to philosophical reinterpretations. This idea, itself based on the picture of theurgy which the writings of late Neoplatonists such as Proclus give us, leads to a synchronic use of the material, that is, to the reconstruction of “theurgy” by combining sources from various historical periods, from the Chaldean Oracles to late antique sources and sometimes even up to Byzantine authors.6 This synchronic approach has yielded valuable collections of material – most notably the monumental work of Lewy7. It has also achieved important insights into the workings and contexts of late antique rituals and their interconnections. This is especially visible in the work of Johnston, who in her numerous contributions draws on various late antique sources to piece together from the fragmentary evidence a coherent picture of theurgy ;8 other contributions to this endeavour are those of Majercik9, Luck10 or Shaw11. However, a harmonising approach downplays and neglects the historical dynamics at work behind theurgy itself, and recent studies have cast doubts on the viability of this perspective. Thus, Athanassiadi proposed to exclude Psellus as a source for the reconstruction of the Chaldean system and only concentrate on Proclus and Damascius.12 But even in these authors theurgy may have acquired new meanings compared with earlier material, so that a rigorous chronological analysis of the sources each in their own right, without later projections, is necessary. At least in part, van Liefferinge offered an 5 This is apparent e. g. in Dodds (1951), Lewy (1978), the works of Saffrey mentioned above, van Liefferinge (1999), Nasemann (1991), Johnston (1997), Athanassiadi (1999), (2006) or (2010). These scholars mostly localise theurgy in an Oriental, ‘Chaldean’ setting, Saffrey (e. g. (1990a), 78 f) and Athanassiadi pointing to Syria. Another, albeit marginal, approach that also views theurgy as a practice inherited by the Neoplatonists casts it as the direct heir of actual Pythagorean and Parmenidean practices: this is the position of Kingsley (1995) and (2001), followed by Bergemann (2006), 296 – 410 and 2010. I need not go into a detailed critique of Kingsley’s position (for that see the review of his 1995 book by O’Brien (1998)): postulating such chains of continuity and using the sparse historical material as a basis for a complete reconstruction of Pythagorean religiosity goes beyond scholarship into (highly readable) fiction and theosophy. But his approach has one positive side: it makes us more sensitive to the Greek background in which the discourse on theurgy developed – as something new, based on a new text – in the Roman Empire. 6 The classical example for this is Lewy (1978), see also Luck (1989) or Majercik (1989), who bases her discussion of theurgic ritual on Lewy (24 – 46). 7 Lewy (1978), originally published 1956. 8 E.g. Johnston (1990), (1992), (1997), (2004a–b), (2008a); her programmatic approach of “imposing a unity” on the disparate sources is succintly stated in Johnston (2011), 127; see also (2008a), 450 f, where she follows Fowden to extend the category of “theurgy” not only to Platonists and the Chaldean Oracles but also to the Hermetic milieu. 9 Majercik (1989). 10 Luck (1989). 11 Shaw (1995). 12 Athanassiadi (1999a), 158 – 159.

The Problem of Theurgy attempt at such a reconstruction, pointing out the varying semantic content of the term “theurgy” in different authors, from the Chaldean theurgists over Iamblichus up to Proclus.13 But even the idea of an independent Chaldean religious group, an idea that has found its classical scholarly expression in the work of Lewy and still dominates research on theurgy, appears misleading.14 “Chaldean” or “theurgy” rather appear to be labels used freely to denote different textual and ritual traditions, related by the respective authors to the Chaldean Oracles. At the extreme of the scholarly spectrum opposite to Lewy, Majercik or Johnston we find the remarks of Janowitz, who in her review of Shaw (1995) concludes that there is no such thing as a “theurgic ritual” and that the scholarly efforts to reconstruct it are misguided. In another work, she discusses theurgy on very general lines, viewing it as the ritual equivalent for theology : a label for acceptable rituals as opposed to the negative label of magic.15 However, her radical dismissal of theurgy does not do justice to the sources and does not take into account that in later, post-Iamblichean authors like Proclus, theurgy is understood as a distinctive ritual tradition. This book aims to disentangle the different threads that run together under the label of “theurgy”, focusing on its actual ritual dimension and trying to analyse its historical development in pagan sources from the Chaldean Oracles to the sixth century AD. That is, this varying complex of practices and the discourse about them shall be analysed on their own, not as a side issue of Neoplatonic philosophy. Much-debated topics such as the exact metaphysical system of the Chaldean Oracles or the relationship of ‘irrational’ theurgy to ‘rational’ philosophy shall be secondary and only considered insofar as they are relevant to the understanding of the rituals. The basic questions will be: What rituals appear in the respective sources as theurgic? Where do they come from; what parallels can be found for them? How are these rituals interpreted and what goals are ascribed to them? The endeavour is to bring a rigorously historical and diachronic perspective to the study of theurgic ritual which so far has been mostly approached synchronically. On the other hand, this perspective will not simply dismiss ‘theurgy’ but has the advantage of taking the sources seriously and not throwing the baby out with the bath, sketching how ‘theurgy’ developed and eventually came to be perceived as a recognisable ritual tradition with a specific profile. Such a comprehensive study of the dynamics of theurgy as a ritual tradition is acutely needed. In spite of her enticing title La thurgie des Oracles 13 Van Liefferinge (1999). 14 Cf. Tanaseanu-Döbler (2008), 34 and 53, and 2010a. The issue resembles the case of the Hermetica: were they just literary products or rooted and embedded in a cultic and instructional context? But whereas this question has been variously debated for the Hermetica (see the summary of Löhr (1997), 285 – 297 or Fowden (1993)) no similar discussion has so far arisen for the Chaldean Oracles. 15 Janowitz (1997), 664; (2002), 1 – 18, esp. 17 f.

Introduction Chaldaiques  Proclus, van Liefferinge16 focuses mostly on semantic and philosophical issues, trying to present the Neoplatonic ideals of ‘real’ theurgy and its spiritual value as opposed to magic – for a historian of religions a highly problematic undertaking on account of its inevitably normative character. The actual practices and their embeddedness in the late antique Mediterranean religious context are neglected. This is a consequence of her approach: she reads theurgy in relationship to De mysteriis as the key text, and assumes that theurgy is there equated by Iamblichus with the whole of pagan cult which is thus spiritualised and endowed with a deeper meaning. But this view is too seducingly simple, given the varied spectrum of different rituals discussed by the Neoplatonists, in which theurgy remains something distinct, related to an esoteric knowledge and to private rituals effecting contact between the individual and the divine. Another recent book on theurgy with a promising title but bound to disappoint is Albanese/Mander, La teurgia nel mondo antico,17 which consists mainly of a commented Italian translation of the Oracles and a briefly commented paraphrase of De mysteriis, with two additional contributions that highlight comparable Babylonian and Egyptian practices. A systematic discussion of theurgy is missing, and the commentary is very sparsely documented, relying largely on outdated scholarship, mostly Lewy’s work. The research questions sketched above open the way to consider theurgy as an example of an ‘artificial’ ritual tradition, composed from already existing elements to create something new. Theurgy offers the great opportunity to look at such a tradition from its beginning up to its end and to see how the mechanisms of inventing and reinventing such a ritual tradition work. Here I take up the concept of the “invention of tradition”, which made its triumphant entry into scholarship in the title of Hobsbawm/Ranger (1983), especially with Hobsbawm’s introduction ‘Inventing Traditions’ (even if Hobsbawm did not invent the phrase, as e. g. Sarot underlines18). Their focus lay not so much on religion but rather on political traditions concerning the representation of authority or the creation of national identities, but nevertheless their perspective proved fruitful for religious studies, being taken up e. g. by Bell for the study of ritual.19 Hobsbawm has the merit to have refreshed the emphasis on how traditions with their halo of invariance and link to a distant past are actually designed, produced, and changed, a reminder of a perspective that is basic to the study of history and culture. As Sarot notes, the concept of the “invention of tradition” is a valuable heuristic tool precisely because of its 16 Van Liefferinge (1999). 17 Albanese/Mander (2011). 18 Sarot (2001), 21. 19 She incorporated it especially in her endeavour to highlight the power relations and forms of control governing ritual practice (see Bell (1992), 118 – 124).

The Problem of Theurgy fruitful ambiguity which enables it to work as an “eye-opener”.20 I use it here as such. To study theurgy as a ritual tradition subject to constant reinvention, I propose to limit my material to the late antique, mostly Neoplatonic sources, a historically delimited corpus stemming from a milieu that is coherent in both its philosophical and in its religious background, which in the period under scrutiny undergoes tremendous transformations. Neoplatonism is at first an innovative part of a publicly accepted religious scene with its public rituals which it can adopt or criticise; progressive marginalisation and gradual repression of pagan traditions turns the philosophical schools into religiously isolated close-knit networks of intellectuals trying to cope with these transformations and to uphold and intellectually support paganism as they understand and to a large extent create it. This move from public to private has repercussions on the nature of rituals practiced in late antique paganism, and we shall see how for the Neoplatonists theurgy comes in handy to enable this transformation. The history of theurgy certainly does not end together with the Neoplatonic schools but lives on into Byzantine and Renaissance times. The reception of the theurgical vocabulary by Ps.-Dionysius and its adaption for Christian theology and ritual is a crucial and well-studied point of development,21 as it makes both the term and the discourse connected with it palatable to Christians, so that the term surfaces in theological contexts where it is least expected, such as the Hesychast controversy.22 Beside the Christianisation of the term, another, more direct, mode of transmission can be seen at work in Byzantine scholars actively engaging with the traditions of antiquity and trying to make sense of Chaldean metaphysics and theurgy, such as Psellus or Pletho.23 But fascinating as these paths of transmission are, they would open the scope too wide to handle the study of the making of a ritual tradition. How, by whom and why is such a ritual tradition assembled? What are the advantages over existing alternatives, and how is theurgy positioned with regard to other similar phenomena? What strategies are used in order to legitimize the new brand of esoteric rituals and to deal with the problem of its actual newness? How is its esoteric aura constructed, and which are its functions? Finally : how is such an ‘artificial’ tradition transmitted and to what extent can it be kept alive by its practitioners? In order to achieve our aim by answering these questions, some preliminary methodological considerations 20 Sarot (2001), 38 – 40. 21 See already Rorem (1984) and (1993), Louth (1986), Shaw (1999), Struck (2001), Dillon/Klitenic Wear (2007), 85 – 115, Stock (2008), 152 – 232. For an alternative approach arguing along Straussian lines that Ps.-Dionysius might be imagined as a “crypto-pagan” endeavouring to hide Neoplatonic truths in Christian garb so as to pass them on to posterity, see Lankila (2011); as he himself admits, most of his argumentation is hypothetical. 22 Tanaseanu-Döbler (2010b). 23 See des Places (1988), Athanassiadi (2002), Tambrun-Krasker (1995), Tardieu (1980) or (1987), Woodhouse (1986), 48 – 61.

Introduction are necessary, concerning (1) the terminology of theurgy and (2) the notion of ritual. (1) Based on the tacit assumption that there is something like a “Chaldean” group with its distinctive religiosity, with the Chaldean Oracles as their revelatory literature and with specific rites based upon it, “theurgy” is often used in scholarly discourse to denote both the reference to the Chaldean Oracles and to certain esoteric rites connected with them. Thus, allusions to the Chaldean Oracles or quotations from them are often taken to signal ‘theurgy’. But as will become apparent in the course of our study, the text and the rituals need nevertheless not go together ;24 certain authors such as Porphyry may draw upon the Oracles without necessarily assigning to theurgic rituals any greater importance. Discussing the development of theurgy requires us to differentiate between the reception of the Chaldean Oracles and their philosophical system on the one hand, and actual ritual practice on the other. Both are taken up in a variety of ways by the Neoplatonists, who produce different theories about theurgy as a varying complex of rituals on the one hand, as well as a polyphonic discourse of interpretation of the Chaldean Oracles on the other hand. These aspects need to be separated as far as possible in the present enquiry. We will on the one hand ask what rituals were understood as theurgic at a given time, i. e. in a given source; this approach questions the idea of theurgy as substantially comprising the same rituals in Iamblichus as in the writings of Proclus or Damascius. On the other hand, we will consider whether the Chaldean Oracles as a text are present in the discussions of such rituals labelled as theurgy, and to what extent they influence their interpretation. Beside the distinction between text and philosophy, the delimitation of our subject matter is complicated by the fact that the actors in the Neoplatonic discourse on rituals use a variety of synonyms or closely related terms to denote the rituals they mean, such as Reqatij^, tekestij^, Reqouqc_a, lustacyc_a or heiasl|r. Unlike heouqc_a which seems to have been coined by the Chaldean Oracles, all other words are established terms of Greek ritual language by the second century AD. How do we determine what they exactly refer to in a given context, that is, when can we be sure that our sources are discussing theurgy and not some other ritual, such as Egyptian practices or the Eleusinian mysteries? Some scholars tend to employ ‘theurgy’ for 24 Cf. also Majercik (1989), 21 f, who however argues for caution on completely different grounds, as she views the younger Julian on the basis of the communis opinio of earlier scholarship as the first who used the specific term “theurgy”, while his father supposedly did not – on this line of reasoning all the traditions coming from the elder Julian, the “Chaldean” would have been perceived as “Chaldean”, but not necessarily as “theurgic”. For the problems of authorship of the Oracles see P. Hadot (1978), 703 – 706. However, for the present analysis, the Juliani, surrounded as they are by a web of legends, are at best secondary ; what concerns us are the sources on theurgy and the relationship between the foundational text and the ritual discourse which it generates.

The Problem of Theurgy Neoplatonic references to ritual, regardless of the exact terms used in the text. This leads them e. g. to translate Reqatij^, tekestij^ or the other ritual-related terms enumerated above as ‘theurgy’ or ‘theurgic’, obscuring the actual vocabulary in the text.25 A related problem is that ‘theurgy’ is used as a term on the meta-level of scholarly language, without however noting and reflecting on the difference between object and meta-level. Thus, Fowden uses the term “theurgy” very vaguely, extending it also to Hermetic writings, based on the close parallels between the Neoplatonic and the Hermetic anthropology and soteriology ;26 for him theurgy is a “potent combination of cult, magic and philosophy”.27 Following explicitly in his footsteps, Johnston also proposes an inclusive understanding of theurgy, reading e. g. Porphyry’s work on cultic statues or his Philosophy from Oracles as sources for theurgic ritual,28 or employing the Mithras Liturgy as a parallel text which is so close to the Chaldean Oracles that it can be used to fill out the gaps in the ascent practices hinted at in the Oracles – which basically makes the Liturgy a theurgic text.29 This inclusive approach highlights the religious and more precisely ritual koine of the late antique Mediterranean and certainly sharpens our understanding of late antique private rituals. Yet, it obscures the fact that the sources literally speak a slightly different language: the Hermetica and the two works of Porphyry mentioned do not use the term ‘theurgy’ or its cognates, nor is there any trace of the specific vocabulary clustering around the term ‘theurgy’ in Neoplatonic texts, nor, finally, any mention of the Chaldean Oracles as a reference text. It is certainly possible and helpful to use a meta-language term to group together comparable phenomena at the object level; indeed, most key terms in the history of religions stem from a specific religious tradition. Keeping the distinction between object level and meta-level clear, ‘theurgy’ can then be abstracted from its specific context of origin to be used as a scholarly tool as defined by the researcher. This happens most notably in the study of Jewish mysticism and magic,30 but also in studies of medieval magic,31 studies of esotericism32 or even modern artists, when studies inquire into the ‘theurgy’ of Skrijabin.33 Whereas in these examples the distance between the 25 E.g. Athanassiadi (1999), 79, 88 or 327, Clarke/Dillon/Hershbell (2004), 47, 265, 275, 277, 278 or 324. 26 Fowden (1993), 126 – 153. 27 Ibid. 126. 28 (2008a); reference to Fowden 450 f. 29 Johnston (1997); see 183, n. 54. 30 To give two recent examples, cf. Flatto (2010), 210 – 227 or the succint definition of Fishbane (2010), 125: “Theurgy – the power of human action and intention to affect the divine realm – is one of the main defining components of medieval Kabbalah […].” 31 See e. g. Vronse (2007), 25 or 28, who employs the term to describe the Ars notoria, a medieval magical text which promises the instant mastery of all sciences; he explicitly draws the line to Neoplatonic theurgy. 32 E.g. Faivre (1994), 34. 33 E.g. Lobanova (2004).

Introduction original historical context of the term and the phenomena which it is applied to is so significant as to indicate clearly that the term is used on the meta-level of theory and scholarly description, a similar meta-level use of the term for the study of late antique theurgy has its pitfalls. It may result in the blurring of these two levels and thus in the assumption that what the scholar describes as theurgy is actually out there as such, already grouped together as such by the sources. And this may lead to lumping things together which the late antique writers on theurgy see as distinct, though in some cases related. As we are interested in reconstructing the growth and development of theurgy as the Neoplatonists saw it, we must distinguish sharply between the two levels and focus on the object level. At this level, we must note the semantic connection of ‘theurgy’ with a specific vocabulary and with specific synonyms or related terms designating rituals; the term ‘theurgy’ is thus at the centre of a web of terms that emerge as related in the source texts. For each author, we will have to outline clearly what these synonyms and terms are and how exactly their relationship is determined. Beside this semantic aspect, intertextuality plays an important role. Theurgy after all begins its life as a coinage of the Chaldean Oracles and is there paired with certain specific terms and notions: the symbols or tokens (s}lboka or sumh^lata) that are divinely provided in the cosmos and can be ritually manipulated to effect the ascent of the soul, the mention of fire as the divine substance spreading through cosmic channels (aweto_), a specific role of the pneumatic vehicle of the soul which needs to be purified and become warm, dry and light for the heavenly ascent, the ascent to the highest principle cast as the Father – all features which we will outline in detail when discussing the Chaldean Oracles. We will thus include into our study texts where rituals are either explicitly labelled as ‘theurgy’, and/or where the relationship to the Chaldean Oracles is present either through the vocabulary used, or through direct references to their authors, the “theurgists”, or, thirdly, through direct quotations. That means that e. g. tekestij^ will not automatically be taken to signal theurgy, but only when it is paired with either the term ‘theurgy’ and/or with vocabulary that is elsewhere used for theurgy and shows a link to the Chaldean Oracles. Furthermore, the translations will keep the various terms used for ritual apart, so as to convey the semantic complexity of the texts. (2) Regarding ritual, the last years have witnessed an outburst of scholarly studies from various disciplines on the topic. A survey of the various heterogeneous approaches would be out of place and irrelevant for my subject; I refer the interested reader to the theoretical compendium and annotated bibliography of Kreinath/Snoek/Stausberg34 and to the monumental proceedings of the 2008 Heidelberg conference on ritual dynamics35 and will only 34 Kreinath/Snoek/Stausberg (2007). 35 Michaels et al. (2010 – 2011).

The Problem of Theurgy point out my understanding of ritual and a few aspects that are of interest for the present study. It is universally agreed that there can be no single generally valid definition or theory of ritual; instead, ritual is seen as a complex phenomenon that can be studied only from a multitude of different angles. New attempts at definition have drawn on post-classical logic, viewing ritual as a “fuzzy set”, and particularly a “polythetic class”, where various characteristics are shared by the members of the class without any of them possessing all characteristics. That is, modern definitions of ritual tend to single out certain salient characteristics of their object, without claiming that every ritual must possess all and fully acknowledging that the list might be continued. This approach stands in the tradition of Wittgenstein’s philosophy of language, especially his theory of family resemblances.36 For the present study, ritual shall include only religious cases, that is, rituals that work on the assumption that there is a level of reality transcending the common everyday experience, however that level may be called. However challenging and productive the comparison of religious and non-religious rituals may be for uncovering mechanisms of individual psychology or collective behaviour and social cohesion,37 theurgy aims at integrating the practitioner into the Neoplatonic hierarchy of divine beings, which includes heroes and daimones, gods and abstract highest hypostases such as the World Soul, the Intellect, or the One. So we will view ritual as a set of practices, of embodied actions, that are set apart and distinguished from common behaviour, and are directed at bringing about a connection of the practitioner with the realm of the divine and thus at opening an avenue of communication with it. The practices are distinguished from other forms of religious practice, e. g. fasting or almsgiving, by their performance in a special setting, distinguished from everyday life.38 The means to create this separateness and to frame the ritual activity as such vary : it can be achieved either through a 36 See Snoek (2006), 4 – 6 and 10 – 14, cf. also Crossley (2004), 32 and 38. 37 Cf. e. g. the funny but insightful comparison of academic and Christian rituals by Theißen (2007), or Rao (2006), 158, who acknowledges that “however, more often than not, the term ‘ritual’ is associated with the religious domain” or Grimes (2006), who blends theatrical and ritual performances together. See also Boudewijnse (2006), 129 – 138 for psychological approaches to ritual. While a wider understanding of ritual, going beyond religious practices, is bound to unearth structural similarities between different rituals that would otherwise go unnoticed, one cannot and should not divorce ritual from its respective context: one can study ritual only in its sociological, psychological, religious or theatrical context; there is no such thing out there as ‘pure ritual’. Therefore, the stance of Kreinath (2005), who argues that “[t]he study of ritual is a field of its own and not simply a continuation of the study of religion”, is misleading: ‘ritual studies’ cannot be a discipline of its own, but only a focal point of an interdisciplinary endeavour. 38 This separatedness and difference is emphasised by all ritual theorists, see e. g. J. Z. Smith (1987), 103 – 110, Snoek (2006), 13, Bell (1992) and (1997), 81, Turner (2006), 235. However, this should not be taken to mean that rituals are isolated from other social processes (cf. Rao (2006),

Introduction special location, or through temporarily creating a sacred space destined for the performance of the corresponding actions, through designating certain times, e. g. feast days; other possibilites would be specific religious roles not pertinent to everyday life but only activated in ritual, such as the role of medium, the use of special robes, of incense, the use of purificatory actions to integrate the participants into the sphere of ritual and mark the transition from the sphere of everyday behaviour, or simply the withdrawal into solitude and the adoption of a special body posture for meditation or prayer.39 The performance of ritual actions can be framed as the repetition and scrupulous re-enactment of an authoritative tradition, set off as formal and strictly rulegoverned behaviour against more informal behaviour,40 employing basic dichotomies such as male-female or light-darkness41 to enforce the special setting. In this approach, the actual practices involved in rituals can vary from meditation or prayer to public sacrifice; they can be either public and collective or private. As C. Bell has noted, following J. Z. Smith, potentially any human action can be marked off as belonging to this extra-ordinary context and thus be “ritualised”.42 The key aspect of rituals is the central role which the bodily aspect of the human person plays in it. This should not be seen in a Cartesian manner as body versus mind, or action versus thought, but rather as a complex experience integrating both sensory and cognitive processes. Following this path, ritual can be seen as a means to open up a new mode of experiencing the world, impacting not only on the actions and the body but also on the beliefs and worldview of the participant: the holistic experience can reinforce, transform or more deeply ground these beliefs in the person’s self. This may be especially important in groups holding complex systems of abstract beliefs, which can be enacted and experienced in ritual performance in a different way, endowing them with a sense and aura of reality and interaction.43 Along these 39 Cf. Bell (1997), 167. 40 Cf. Bell (1997), 138 – 169, singling out formalism, invariance, traditionalism, performance, rulegovernance, and sacral symbolism as the mechanisms of ritualisation and the characteristics of ritual-like activities. However, this refers mostly to the perception of the practitioners and the way they present their behaviour. Rituals do change and are constantly re-invented to suit the needs of the practitioners; we shall see how innovation plays an important role in the growth of theurgic ritual in Late Antiquity. Humphrey/Laidlaw (2006), 275 see the stipulated non-intentionality as the main characteristic of ritualisation: actions are rendered “non-intentional, stipulated, and elemental or archetypal” ((2006), 278). For a more balanced view on such and related theories of the meaninglessness of ritual see Michaels (2006) or Thomas (2006), 342: “‘Meaninglessness’ as the absence of pragmatics is still an ascribed meaning, attributed by the researcher as participant-observer”. 41 Bell (1992). 42 Bell (1992), (1997), 91; cf. also Harth (2006), 27. He suggests a gradual approach assuming a “graded spectrum that ranges from strong to weak in ritual character” (32). 43 For the importance of the body for ritual experience see e. g. Jennings (1982), 115, Bell (1992),

The Problem of Theurgy lines, it has been argued that it is possible to see ritual as productive of knowledge or, as recently Schildbrack has put it, as a mode of metaphysical enquiry, as a noetic tool towards an apprehension of the “most general contexts of human existence”44 : “A ritual is a metaphysical inquiry, then, to the extent that it aims at increased knowledge of being in the world authentically, that is, being in the world in the way authorized by the very nature of things”.45 Or, with Williams and Boyd (1993), “The different aspects of the ritual speak to different senses and affect different portions of the psyche, so that their unification brings about a unification in the perceiver […] the participant comes to know, in ways not reducible to propositional expression, that he is cooperatively engaged with the creatures of the good creation. He experiences a felt unity, compelling to both heart and mind.”46 Certainly, this cannot be generalised: not all practitioners will experience or use the ritual as a noetic tool, but this perspective renders us sensitive to the possibility of practitioners who do so. The peculiar framing of ritual as a distinctive sphere of action does not mean that it takes place outside of society. The function of rituals to provide and enhance group cohesion stands at the centre of many sociological approaches to ritual, classically expressed by Durkheim. Even if the Neoplatonic philosophers formed networks of highly individualistic intellectuals, this integrative function must also be kept in mind for our analysis of theurgy. Rituals do not only enhance belief but also turn “membership into belonging”, functioning as one focal point of identity construction.47 94 – 117, (1997), 160 or (2006), or Theißen (2007), 32 – 34, who (from a Christian perspective) speaks of a “‘somatische Semantik’ oder eine ‘semantische Somatik’” (33 f). The role of ritual in enhancing and deepening belief along with creating a sense of belonging is emphasised by Marshall (2002) from a Durkheimian perspective. See also Schilbrack (2004), 131: “[…] rituals as a process through which a religion makes their most abstract teachings concrete, giving facticity to their ideology. […] The goal is to have ritual participants perceive metaphysical truths ‘in the flesh’”, or Crossley (2004), 44 f. Another accent is brought into this line of thought by Raposa (2004), who analyses the role of ritual in channelling attention and thus increase awareness; in this sense, ritual can be seen as a “type of inquiry, a kind of thinking embodied in conduct, behavior that can be conceived as a deliberate process of semiosis” (123). 44 Schilbrack (2004), 137. 45 Schilbrack (2004), 139. In this ‘noetic’ approach to ritual, two positions can be distinguished: Jennings (1982) argues for a dynamic relationship, where the growth of knowledge through ritual also requires and triggers the change of ritual (113 – 116), whereas Williams/Boyd draw on Zoroastrian examples to show that the ritual itself, which comes to be viewed as an “artistic ‘masterpiece’” does not change, although it serves as a source of practical wisdom for the participant (1993 and 2006, 294 – 296). They advocate a more balanced view including different modes of gaining “wisdom” through ritual and draw on aesthetic theory and metaphor theory to describe in what ways rituals viewed as “masterpieces” surrounded with an aura of necessity and irreplaceability can both convey meaning and also create a fecund space in which the participants produce meaning which they attach to ritual (1993, 84 – 100 and 118 – 130). 46 Williams/Boyd (1993), 57. 47 See Marshall (2002) for a recent development of Durkheimian theory ; quoted phrase on p. 361.

Introduction The focus on the integration of belief, knowledge and action in ritual and on a distinctive experience of that which is believed through bodily enactment, symbolic concretisation and ‘fleshing out’, an experience which may create a sense of safety and belonging, opens up a promising avenue of enquiry for analysing rituals designed by philosophers for philosophers and thus for understanding the development and the importance of a discourse about theurgy and its role in the face of the transformations of the pagan religious panorama. We will now proceed to trace the various voices which constitute this discourse.

2. Theurgy and the Chaldean Oracles 2.1 Preliminaries The earliest secure attestation of the term “theurgy” probably occurs in the Chaldean Oracles. Later writers place their composition in the reign of Marcus Aurelius, in the second half of the second century ; their authors, the two Juliani, are surrounded by legend.1 Another alleged mention of theurgists (heouqco_) in the Excerpta de musica written by the Neopythagorean Nicomachus of Gerasa, which would attest the term by the 160s, rests only on a conjecture within a corrupt text passage.2 If the conjecture be correct, then it would show theurgists at work making all sorts of strange sounds, hissing and clicking their tongues in a manner highly reminiscent of the magical papyri or Christian exorcisms.3 Interestingly, the term appears as one word among many to designate cultic personnel or ritual specialists in the Onomastikon of Julius Pollux,4 which was written between 166 and 176.5 Considering the aim of the Onomastikon to be a source and tool for the wqe_a t/r vym/r, Julius Pollux must have regarded this term as a proper Greek term fit for rhetorical use; that makes it unlikely that he should have come across it in a magical context, as magic, dubbed mostly cogte_a, appears in the lexicon always in the context of pejorative terms. If he found the term in the Oracles, that might be interpreted as a further indication of the essentially Greek character of the Oracles, which has been noted by Johnston or Hadot.6 1 On the two Juliani see Saffrey (1990a), 64 – 69 and 78 f, and most notably Athanassiadi ((1999), 149 – 156; (2006), 38 – 70; (2010)). But see also the cautious remark of P. Hadot (1978), 705, that it might be safer to follow des Places and “garder aux Oracles l’anonymat”. 2 Nicomachus of Gerasa, Excerpta 6, 277, 7 Jan. 3 See e. g. the vowels and consonants scattered in the prayer found in the Mithras Liturgy, PGM IV 488 – 495 or ibid. 561 f: 5peita s}qisom lajq¹m suqicl|m, 5peita p|ppusom. For Christian practice see e. g. Julian, ep. 79 Bidez (suq_tteim pq¹r to»r da_lomar). However, Proclus mentions in his scholia to the Cratylus that theurgists do make inarticulate noises: LXXI 31 Pasquali: toiaOta d’ 1st·m t± jako}lema s}lboka t_m he_m7 lomoeid/ l³m 1m to?r rxgkot]qoir emta diaj|sloir, pokueid/ d’ 1m to?r jatadeest]qoir7 $ ja· B heouqc_a liloul]mg di’ 1jvym^seym l]m, !diaqhq~tym d], aqt± pqov]qetai; see van den Berg (2008), 165. 4 Julius Pollux, Onomastikon I 14, 4 Bethe. 5 For the date of the Onomastikon see Bethe (1900), V. It survives only in a 9th century epitome, with additions from later Byzantine authors (Bethe (1900), VI–XVII). 6 See Johnston (1997), 170, Hadot (1978), 706 f. A text like the Oracles, presenting itself as ‘Oriental’ in a fundamentally Greek manner with some unusual touches of vocabulary would have fitted Pollux’ “moderate Atticism, which is open to Hellenistic and even Iranic or Latin words” and his “purpose of reflecting the chronological and geographical variety of the Hellenophone provinces

Theurgy and the Chaldean Oracles The Oracles themselves expound a worldview marked by a basically Middle Platonic cosmology and anthropology ; they have been included by Dillon in the “underworld of Platonism”7. Carefully preserving the appearances of Greek oracles down to the hexametric form which dominates the collection, the extant fragments are concerned with cosmogony, with the descent of the souls into matter and their return back to their divine home.8 The divine substance is described as fire, which descends from the highest principles through various channels (aweto_) down to the material world.9 The highest realm is dominated by a triad formed by the Father or Paternal Abyss, who is the first Monad, and his Power and Nous, who emanate from him.10 The Middle Platonic pattern of the hyperessential intelligible principle which brings forth the intellect as the major demiurgic entity11 is enriched with elements from other traditions: the conception of the divine substance as fire reminds us of Heraclitus or Stoic cosmology, and can also be found in other religious traditions of the time, e. g. Gnosticism12. A peculiar aspect of the Oracles’ metaphysics is the important role assigned to Hecate, who seems to function as a mediator between different entities or realms, most conspicuous in her role as the World Soul, the mediator between the intelligible realms and the material cosmos;13 she also seems to be the principal divine speaker of the Oracles. in the Roman Empire” (Zecchini (2007), 26), which fit well into and replicate Commodus’ stance in cultural politics (ibid. 25). 7 Dillon (1996), 392 – 396. Majercik joins him in describing the Oracles along with the Gnostic and Hermetic texts as having a “rather murky quality” (1989, 4). This metaphor of the underworld is highly misleading, considering the affinity of the Oracles with the traditional hexametric form of oracles and their display of Hellenism. Also, their reception by people like Pollux shows that they could be regarded as respectable literary products in their time. 8 Brisson (2000a), 332 – 338 shows the close dependence of the Oracles’ cosmology on the Timaeus presenting them as a theological interpretation of the Platonic dialogue with a soteriological thrust. He also notes other Platonic works whose influence can be detected in the Oracles (333). He reiterates his reading of the Oracles also in Brisson (2003). 9 Here only a brief summary of the Chaldean Oracles’ worldview can be given; for a more detailed analysis see Majercik (1989), 5 – 21. Seng (2009a) and (2009b) has brilliantly shed light on some details of the Chaldean hierarchy of beings. For the triadic structure of the highest entities see e. g. Majercik (2001). A good concise overview of the Chaldean system can be found in Turner (2008), 40 – 47. 10 See frg. 3 – 5; 11 – 12; 18 – 21; 26 – 28 des Places. As Majercik (2001), e. g. 296, points out, the exact nature of the Chaldean triad cannot be reconstructed, because we only have it inextricably interwoven with the various Neoplatonic interpretations. 11 For variations of this metaphysical pattern see e. g. Numenius, frg. 11 – 13 und 15 – 17 des Places or Alcinous, Didascalicus X 2 – 3 and XII–XIV. Cf. also Origenes, De princ. I 2 or Clemens, Strom. IV 156 f for Christian adaptations of this scheme. 12 See e. g. the description of the system of Simon Magus by Hippolytus, Refutatio VI 9, 3 – 10, 4. 13 Frg. 6: Hecate as moeq¹r rl^m (spiritual membrane) between the first fire and the other fire; frg. 32: the life-generating womb of Hecate, receiving the fire of life from above (at the level of the synocheis, the entities in charge of keeping the cosmos together ; frg. 35: the womb of fatherborn Hecate; frg. 50: the centre of Hecate is floating between the two Fathers; frg. 51 – 52: Hecate

Preliminaries The continuous exchange between the milieu of the Oracles and other contemporary strands of thought is shown by the appearance of a similar triadic structure of the highest principles in Sethian Gnostic texts. Their dating varies in accordance to the relationship each scholar establishes between them and Neoplatonic texts, ranging from the late second century to the second half of the third century AD.14 It seems, however, more plausible to assume that such texts like the Apocalypse of Zostrianos were written around the end of the 2nd century AD or at any rate in the first half of the third century AD15 Then one can consider either the possibility of an influence of the Oracles on the Sethian authors,16 or a shared background of triadic speculations, as both the Oracles and the Sethian Gnostic texts are close to Numenius17 and also to Neopythagorean arithmetic speculations. Thus, an account of Pythagorean number theology falsely ascribed to Iamblichus18 points to the role assigned to the triad by Nicomachus of Gerasa as a structuring principle of the universe, quoting the Homeric formula tqiwh± c±q p\mta d]dastai, which in texts from the second and third century can be applied not to sacrifice, as in Homer, but to cosmic tripartition.19 This triad is closely connected or identified with Hecate, thus coming very close to the role of Hecate as ensouler and structurer of the cosmos in the Chaldean Oracles. This triad projects forth the hexad, which is a further principle of cosmic order, associated with the distances between the planets and the harmony of their whirring sound (No_fgsir).20 as the source of the first soul, who vivifies the worlds, and also the source of virtue, 54: she carries nature on her back. For the role of Hecate in the Oracles and its roots in Greek religion see Johnston (1990). Hecate as a connecting force, influencing different parts of the cosmos, is already present in Hesiod’s Theogony 411 – 452, where she is allotted influence in the chthonic as well as the heavenly realms and is presented as a mighty goddess with great influence. This idea appears again in the late antique Orphic hymn to Hecate, where she is hailed as oqqam_am, whom_am te ja· eQmak_am (v.2) and as pamt¹r j|slou jkgidoOwom !m\ssam (v. 7). The Orphic hymns cannot be dated with precision, having been written somewhere between the 2nd and the 5th century AD (Morand (2001), 304); that is, the same period when theurgy develops. Earlier, Plutarch also presented her as both chthonic and heavenly at the same time and thus explains her association with the moon (De defectu orac. 13, 416). 14 See Majercik (1992), 476, n. 6. She herself pleads for a very late date, after the death of Plotinus in 268 (488). An example for an early dating would be the work of Turner (1992), 439 – 455; a middle position is occupied by Brisson (1996). 15 See Turner (1992), 439 – 455 and Brisson’s discussion and criticism of Majercik (1992) (1996, 179). Brisson points out that the Platonic influences in the Apocalypse of Zostrianos are not Porphyrian but Middle Platonic, traceable to Numenius, who was in turn influenced by Middle Platonic speculation embodied in texts like the Second Letter of Plato. 16 That is the tendency of Turner (1992), 439. 17 Turner (1992), 454 – 455. 18 Dillon (2000), 835. 19 The Homeric quotation appears e. g. in Hippolytus of Rome, Refutatio V 8,3 and V 20,8 Marcovich in his account of Naassene and Sethian doctrine. 20 Ps.-Iambl. Theolog. arithm. 49: 2jatgbek]tim d³ aqtμm ja· tqiod?tim ja· diwqom_am pq¹r to}toir 1j\koum7 2jatgbek]tim l³m !p¹ toO tμm tqi\da, Dm :j\tgm owsam paqeik^valem, bok^sasam ja· oXom 1pisumtehe?sam !pocemm/sai aqt^m7 tqiod?tim d³ t\wa l³m paq± tμm t/r heoO v}sim, eQj¹r d],

Theurgy and the Chaldean Oracles Here we find some elements attributed to earlier Pythagorean lore, which appear also in the Chaldean Oracles: the triad, Hecate as a cosmic principle, the whirring sound of the planets.21 Sethian Gnosticism also shows itself conversant with Neopythagorean ideas: Turner (2008) points to Moderatus of Gades as a parallel.22 A comparison with the Sethian texts and with Numenius helps to situate the Oracles better in their context. Numenius’ extant fragments show an interest in religious matters, especially in mystery cults like Eleusis or possibly also the Mithras mysteries23, but nevertheless, his metaphysical fragments are formulated in a highly abstract Platonic terminology without any mythological underpinnings.24 His highest entities remain Platonic. The Sethian texts, building on the same metaphysical structure, employ different names, going from abstractions to concrete proper names as they go down the ontological scale: the first principle is called the Invisible Spirit or the Unknowable One, then comes the Triple Power or Eternal Life, strongly reminiscent of the Chaldean power emanating from the Paternal Abyss and mediating between him and the Intellect. The role of the Intellect in the Chaldean system is occupied by the aion of Barbelo, comprising different aspects: Kalyptos, Protophanes, the Triple Male and Autogenes.25 Intellect is now substituted by a goddess bearing a distinctively barbarian name, who reigns over subordinate personal entities bearing abstracta as names.26 In contrast, the Chaldean Oracles also personify the highly abstract metaphysical structure of Middle Platonism and introduce peculiar entities like the sumowe?r or teket\qwai27, who seem to remain abstract classes of divine beings, without proper names, or the ûpan and d·r 1p]jeima.28 The same penchant for divinised abstractions is at work in the personifications of various aspects of Time (Chronos) and Eternity (Aion), which are mentioned especially by Proclus as fti ja· B 2n±r t± tq_a t_m diast\seym jim^lata pq~tg 5kawe, diw|hem pepeqasl]ma !lvot]qair jah’ 6jastom peqist\sesi. This view of the hexad is taken from Anatolius of Laodikeia, who refers to earlier authors. Dillon (1996), 358 ascribes it to Nicomachus of Gerasa. The importance of the hexad which sums up the distances between the planets and thus contributes to the harmony of their sounds is found in Theolog. arithm. 48. For the triad see his summary of the views of Nicomachus of Gerasa ap. Ps.-Iambl. Theolog. arithm. 17 – 18. 21 For the No?for of the planets see e. g. frg. 146. 22 Turner (2008), 50 with n. 41, 55. 23 See frg. 53 – 58 des Places. The Mithras cult is prominent in Porphyry’s De antro nympharum, which draws heavily on Numenius; see Turcan (1975), 62 – 89. 24 That must certainly remain a hypothetical conclusion, based on the extant fragments. 25 For a reconstruction of the Sethian system see e. g. Turner (2008), 49 – 55. 26 Cf. Turner (2001), 635 on the difference between Sethian and Platonic ascents of the soul: “it seems to me that the basic difference between the two lies in a preference either for myth and dramatic personification, or for conceptual analysis and distinction as a vehicle for explaining the same human problematic”. 27 See frg. 32 and 86 des Places. 28 On these mysterious entities see Seng (2009b) (I thank him very much for sharing his text with me before publication).

Preliminaries deities of the Chaldean pantheon,29 and in the classes of astral entities.30 But instead of barbarian names, the only known mythical figures appearing in them are those of the Greek pantheon: Hecate, perhaps Rhea,31 and probably Athena.32 Eros may have played a part as well, although it is difficult to deduce from the extant fragments whether he was presented as a divine person in his own right or rather as a name for a cosmic power.33 This shows that unlike the Gnostic texts, the Oracles develop their worldview in close connection with the Greek religious tradition and thus aim at a Hellenised educated public (given also the mostly hexametric form and the preference of Ionic word variations).34 For their triadic speculation and conception of Hecate as a cosmic force one should thus rather look to Neopythagorean parallels like Numenius or Nicomachus of Gerasa, whose conception of Hecate we noted above. The fragmentary state of the text makes it hard to decide whether it was composed as one whole oracular poem or a collection of oracular utterances; the reference to it in later Neoplatonic works by the term t± k|cia in plural makes the latter seem more plausible,35 as do the extant Oracles collections and commentaries of Psellus36 and Pletho37, which present the Oracles as a collection of distinct oracles of varying length. Proclus’ references to them as utterances of the gods point out that the oracles were ascribed to different 29 See Lanzi (2006), 37 – 43 and infra, 5.3.1. 30 Recently analysed by Seng (2009a). Just as vague and abstract are the “Fathers overseeing magical operations”, oR 1p· lacei_m pat]qer mentioned by Damascius, De princ. III 31 Westerink/Combs. However, Damascius is the only late antique source who mentions this group of deities, and he is very late – we cannot know if they were part of the Chaldean Oracles in the second century. 31 Frg. 56 des Places. But cf. the alternative view by Majercik (2001), 292 – 294, who sums up the debate about the role of Rhea in the Chaldean pantheon and views the mention of Rhea as a late substitution for the original Hecate in the course of late Neoplatonic attempts to harmonise Chaldean and Orphic theology (294). 32 Frg. 72 according to the context in Proclus, Theol. plat. V 35, 130 Saffrey/Westerink (des Places (1971), 85 reads it as a reference to Hecate, without giving a reason). For a more general structural comparison between the Gnostic, Hermetic and Chaldean systems see also Majercik 1989, 4. 33 See Lanzi (2006), 44, who comes to very reserved conclusions, showing that the fragments do not allow for a clear pronouncement on the issue for Eros. Also, her analysis of the Chaldean mentions of Hades shows him to be not so much a god but rather a cosmic space with all its dangerous powers (2006, 44 – 47). 34 See also Johnston (1992), 316, or (1997), 170, where she speaks accordingly of the “aggressive Hellenism” of the Oracles. 35 Lewy (1978), 36 f. Des Places (1971), 10 also views it as a collection of oracles. 36 See the edition in des Places (1971), 162 – 186 and now the more recent edition of O’Meara (1989). 37 See the edition of Tambrun-Krasker (1995), collection 1 – 4 (numbered oracles) and the commentary of Pletho 4 – 19. Pletho’s text is a revised, restructured and abridged version of Psellus’ collection (Tambrun-Krasker (1995), xi and 37 f, 44 f, 47).