Scholia Reviews ns 7 (1998) 16.

Jacob Rabinowitz, The Rotting Goddess: The Origin of the Witch in Classical Antiquity. New York: Autonomedia, 1998. Pp. 153, incl. 15 black and white illustrations, and an appendix. ISBN 1-57027-035-X. US$12.00

Derek Collins
The University of Texas at Austin

This book presents classicists interested in ancient magic and witchcraft with serious methodological problems, some of which are insurmountable, while others are revealing for what they say about Rabinowitz' neo-pagan agenda. Although Rabinowitz makes pretensions toward scholarship by citing classical scholars and much good evidence, his line of argumentation and use of comparative material from other cultures is reminiscent of James George Frazer, and, most importantly I believe, Margaret Alice Murray, who was a folklorist and intellectual heir of Frazer and whose theories on the origin of European witchcraft have long been dismissed by more careful scholars.[[1]] Neither Frazer nor Murray are acknowledged by Rabinowitz, though Frazer is mentioned in a different connection in a footnote (p. 132, n. 1), but then Rabinowitz does not state forthrightly that he is a neo-pagan either: the reader can discern this for himself when he reads the Dedication to Hekate (pp. 11- 12), in which among other things Rabinowitz asks Hekate to bring his book to birth and to adopt it as her child.

The Rotting Goddess is divided into four parts: Hekate (pp. 15-69), The Witch (pp. 71- 111), Conclusion (pp. 113-22), and Appendices (pp. 123-47). The central argument of the book concerns the supposed transformation of the goddess Hekate from a fertility goddess and Great Mother figure in Asia Minor during the 8th century B.C. (p. 17) into a demonic, 'rotting' goddess who becomes an 'eater of filth' borborophorba (p. 62) in the Papyri Magicae Graecae (e.g. at 4.1402 and 1406), which are dated to the 2nd-4th centuries A.D.[[2]] More specifically, Rabinowitz argues in chapter 1, 'She is a Tree of Life: Hekate and the Cosmic Axis' (pp. 17- 33), that Hekate was first syncretized (Rabinowitz's term) with Artemis, as attested for example, by Aeschylus (Suppliants 676), because Artemis was also 'originally' a Great Mother figure in Arcadia, and because Hekate 'relieved Artemis of her outgrown traits, sparing Apollo's unmarried sister an embarrassing maternity' (p. 19). Much effort is given to discovering the 'original' character of one goddess or another, and Rabinowitz's detective-work can often be, as in the sentence just quoted, as tendentious as Frazer's attempts in The Golden Bough to discover corn spirits behind many dying gods. The fallacy here is that meaning tends to be found only in origins, an error that is avoided by more careful scholars of ancient religion.[[3]]

The next important phase of Rabinowitz's argument (chapter 3, 'Underneath the Moon: Hekate and Luna', pp. 43-58) attempts to explain how Hekate came to possess lunar associations in the Roman period by way of her association with Juno and Diana. Several missteps, however, are made along the way. First, we are told that Luna is 'the central fact of Roman religion' (p. 43), which is thinly supported by the handful of quotations from Varro, Horace, Cicero, and Apuleius that Rabinowitz gives. The theme in all of the quotations given is the moon's role in determining the Roman agricultural cycle, which Rabinowitz contrasts with Hesiod (p. 46), who instead uses primarily astral cues.[[4]] However, this in no way prepares us for what comes next, which is the claim that Juno 'is the most logical representative' for the moon because her epithet, Lucina means 'light' (p. 47), and because the Calends were sacred to her (p. 48). This line of reasoning is not only faulty, it betrays Rabinowitz's unscholarly willingness to view the evidence in whatever manner he sees fit.

In the next portion of Rabinowitz's argument, the same bias reappears in a more pernicious form: because Hekate is triplex, Juno must be too, and in support of this Rabinowitz cites four inscriptions from Latium, IUNONE SEISPITI MATRI REGINAE,[[5]] which we are told are found 'particularly' in Lanuvium. In a Dumezilian flourish, Rabinowitz claims that these inscriptions support his argument for a tripartite character for Juno (Juno as preserver, mother, and queen) which in turn he will relate to a Republican denarius apparently depicting Diana Triplex (pp. 48, 58 n. 17, and fig. 11). However, what we are not told is that these inscriptions are only found in Lanuvium,[[6]] which means that they cannot be taken, except by force, to represent Juno's 'archaic threefold character' (p. 48) for all of Rome. Rabinowitz has willfully misrepresented the evidence here to suit his argument. Furthermore, the 'parallelism' between the supposed triple character of Juno and Diana Triplex on the Roman denarius is progressively transformed by Rabinowitz from hypothesis to fact. On p. 48 the two are 'parallel'; on p. 49 we read that Diana now forms a 'couplement' with Juno; and finally, on the same page, after citing an ode by Horace to Diana (3.22), we hear of the 'absorption [of Diana] into triple Juno.' Once Rabinowitz has connected Hekate to Juno and Diana, the stage is set for Virgil (Aeneid 4.511), who represents a triplex Hekate as an aspect of Diana.

With this association fancifully complete, Rabinowitz returns to the mother goddess, lunar, and triplex qualities of Hekate as we find her in, for example, Aristophanes (Lysistrata 63-4), Theokritos (2.33- 36),[[7]] and PGM.[[8]] What remains is to explain the 'demonization' (Rabinowitz's term, p. 118) of Hekate in Greece--in other words, her association with death, filth, and maleficent magic generally in PGM.[[9]] Rabinowitz's thesis here is that Hekate- worship, originally a fertility cult, was reevaluated negatively as witchcraft because of 'the generally recognized schizophrenia of Greek religion' (p. 116). The first part of this claim sounds remarkably similar to that made by Margaret Murray in her now infamous books, which claimed that witchcraft in 16th and 17th century Europe was a vestige of earlier pagan fertility cults that were debased by Christian fathers.[[10]] Yet there is no evidence of any debasement of Hekate in Greece as Rabinowitz claims.[[11]] The second part of Rabinowitz's claim, that Hekate is debased because of the division between 'chthonic popular worship' and 'aristocratic-ouranian cultus' derives, as Rabinowitz admits, from 19th century classical scholarship (p. 116),[[12]] and not unlike much of the other argumentation in this book, is not worthy of serious consideration. In summary, The Rotting Goddess has to be read too carefully because of its numerous methodological errors to reward significantly the student of ancient witchcraft.


[[1]] Frazer's classic study of religion and magic is The Golden Bough (London 1911- 36, 3rd edition). For the influence of Frazer's theories on classicists in particular, see Fritz Graf, 'Prayer in Magic and Religious Ritual,' in Christopher A. Faraone and Dirk Obbink (edd.), Magika Hiera: Ancient Greek Magic and Religion (Oxford 1991) 188. Margaret Murray's thesis is elaborated in two books, The Witch-Cult in Western Europe (Oxford 1921) 24, and The God of the Witches (London 1931) 13-22. Her thesis has been refuted by, among others, Keith Thomas,Religion and the Decline of Magic (New York 1971) 514-16, Norman Cohn, Europe's Inner Demons (London 1975) 107- 25, and most recently by Robin Briggs, Witches and Neighbors: The Social and Cultural Context of European Witchcraft (New York 1996) 37f.

[[2]] K. Preisendanz (ed.),Papyri Magicae Graecae: Die Griechischen Zauberpapyri. 2 vols. (Leipzig 1928). Henceforth PGM.

[[3]] The importance of avoiding this kind of error is pointed out by, e.g. Denis Feeney, Literature and Religion at Rome: Cultures, Contexts, and Beliefs (Cambridge 1998) 76- 80 and 115. Cf. H.S. Versnel, Inconsistencies in Greek and Roman Religion 2: Transition and Reversal in Myth and Ritual (Leiden 1993) 242 (cited in Feeney, ibid., p. 115).

[[4]] The best exposition of the Hesiodic calendar is in M. West (ed. with prolegomena and commentary), Hesiod. Works and Days (Oxford 1978) 252f. Although Rabinowitz is correct to deemphasize the role of the lunar cycle in Hesiod's farming calendar, his claim that Hesiod virtually 'exclu[des] the moon except as an overall time-frame' (p. 44) downplays considerably the superstitions and proverbs in the 'Days' portion of the Works and Days (lines 765-828), which for example can directly impact the processing of grain (805-7).

[[5]] Hermann Dessau (ed.), Corpus Inscriptionum Latinarum (Berlin 1887) 14: 2091, 2088, 2089, 2121. Henceforth CIL.

[[6]] At CIL 14: 2121, we read tituli Iunonis Sospitis Matris Reginae extra Lanuvium adhuc nulli inventi sunt. Other inscriptions to Juno in volume 14 alone of CIL refer only to IUNO (e.g. 4097 from Albano and 4105a from Praeneste) or to IUNONI (e.g. 1792 from Ostia).

[[7]] Rabinowitz curiously assumes that Theokritos's second Idyll represents 'contemporary' (p. 76) or 'actual' (p. 116) witchcraft, as opposed, presumably, to the preponderance of literary examples from both Greece and Rome that he cites throughout. The assumption underlying Rabinowitz's claim is never made explicit, but it appears to depend on a similarity between the representation of magic (and of Hekate) in Theokritos and in the PGM (which I infer from pp. 61 and 116). However, Theokritos cannot be relied upon as an index of Hellenistic magic, as emphasized most recently by Fritz Graf, Gottesnähe und Schadenzauber: Die Magie in der griechisch- römischen Antike (Munich 1996) 166f. In general, Rabinowitz relies more heavily on literary rather than historical sources for his discussion of witchcraft, and this may reflect the unbalanced state of the evidence. But I find it unpardonable that no mention is made, for example, of Theoris of Lemnos, an historically attested Greek witch (pharmakis) who was prosecuted for witchcraft by Demosthenes and sentenced to death by an Athenian jury (Plutarch, Dem., 14.4; pseudo-Demosthenes, Against Aristogeiton, 1.79ff.; Harpocration, s.v. QEWRI/S).

[[8]] Hekate's association with the crossroads, and her connection with Enodia (Latin Trivia) is dealt with by Rabinowitz in part I, chapter 2, `Queen of Heaven, Queen of Hell: Hekate-Enodia' (pp. 35-42).

[[9]] The terms witch and witchcraft are never defined by Rabinowitz, but he seems to assume that witchcraft refers to one who practices maleficent magic. I infer this definition based throughout and on a statement made concerning the figure of Simaitha in Theokritos's second Idyll: 'Simaitha is a silly love-struck maiden, and one who is even more painfully subject to her desires than her witch predecessors. For them magic was an expedient, for her it is the last resort' (p. 76).

[[10]] See above, n. 1.

[[11]] Again the fallacy here is to assume that 'originally' Hekate was something 'good,' which Rabinowitz then contrasts falsely with her 'negative' image in PGM. For that matter, Greek magic can have both good and ill influences, as demonstrated repeatedly in Christopher A. Faraone and Dirk Obbink (edd.), above, n. 1, especially the articles in this volume by Fritz Graf and John Scarborough, so that a witch as Rabinowitz seems to define it cannot simply be reduced to a practitioner of harmful magic.

[[12]] Despite Rabinowitz's quotation of Burkert's claim (Greek Religion, tr. John Raffan (Harvard 1985) 201 [misquoted by Rabinowitz as p. 210]) that such a division between gods associated with heaven and earth is too common to be of special significance for Greece (p. 117), Rabinowitz would have us believe that this framework nevertheless led to the demonization of Hekate. Walter Burkert's arguments in my opinion remain strong.