Bryn Mawr Classical Review 98.3.15

Fritz Graf, Magic in the Ancient World, trans. F. Philip. Cambridge MA: Harvard University Press, 1997. Pp. vi + 314. ISBN 0-674-54151-0.

Reviewed by C. Robert Phillips, III, Department of History, Lehigh University, Bethlehem, PA 18015,

This book arrives anticipated. Widely circulated and discussed in its French version, the English version occasioned a pre-publication panel discussion, including the author, at the annual meeting of the Society of Biblical Literature (November, 1997).1 The reasons for the anticipation are patent. The author has high repute for his work on Greek religion. And hitherto there have been no general books on magic -- non-specialists were obliged to cobble from a plethora of specialist articles. Those articles obviously assume familiarity with a variety of sources obscure even for classical studies. For example, there is the ever-increasing stock of curse tablets (defixiones), often published in out-of-the-way places and often of excruciating linguistic difficulty. Ditto the magical papyri -- and although the fundamental collection (PGM) has been translated, even reading the English requires substantial background. Finally and generally, the term "magic" has occasioned and continues to occasion considerable theoretical discussion, often of polemical tone. Classicists, as a function of their training and their field's self-understanding, are usually ill-equipped to find, let alone to evaluate, such discussion.

Thus one should not hope for more than one book can reasonably deliver. On the evidential side, Graf must be selective -- inevitably someone's favorite item will not appear. On the theoretical side, there are enough theories and nuances thereon for at least a five-foot shelf -- selectivity becomes essential here, too. Moreover, scholarly publication marches on, and thus, inevitably a book on magic can (wrongly) seem dated even as it appears -- thus Graf could not utilize Simon Pullyn's Prayer in Greek Religion (1997) -- although he might have been able to consult the Oxford D.Phil. from which it came. Finally, "magic" inevitably entails what some (wrongly on my view) have considered religion, which means a book on "magic," regardless of its perspective on the concept will (rightly on my view) have to consider "religion." A daunting task even for one of Graf's formidable scholarship. Despite all this, a book on magic does not automatically receive immunity from criticism; some issues of evidence and theory have such overreaching importance that a reader can reasonably expect to find them discussed -- and a reviewer reasonably complain of their omission.  


There are seven chapters: Introduction; Naming the Sorcerer; Portrait of the Magician, Seen from the Outside; How to Become a Magician: The Rites of Initiation; Curse Tablets and Voodoo Dolls; Literary Representation of Magic; Words and Acts. The titles betoken an outlook which will receive discussion throughout this review. First, it is a largely ahistorical and thematic view, despite references to, e.g. the earliest appearance of magic in Greece and Italy. Second, emphasis on the magician and magical artifacts rather than how non-adepts, often but not always the litterati, viewed magic, as if magicians somehow represented a self-defined, perhaps self-aware, group. Third, an idiosyncratic view of categories; "voodoo," e.g., is an ethnographic term, of debatable cross-cultural utility.

Then there are some mechanical matters which affect comprehension. The book began in a French version (vi), then came a "significantly changed German version." Whence the current English translation? Perhaps from the French since 252n58 references Libanius' work not, as elsewhere, Oratio but as Discours -- whether that guess be accurate, a forthright statement of the relationship between the book's three incarnations would help not only readers, but those who must cite the work.

Then there is documentation. Graf has utilized a selective bibliography; readers seeking a quick view of his extensive reading will have to peruse the notes slowly. Sometimes (257n2) a work receives a "social science citation format" reference which does not appear in the bibliography. There is no index of ancient passages discussed, let alone referenced. This is unfortunate, because Graf's generous use of ancient sources will be particularly valuable to the uninitiated. In bygone times, such omissions were understandable; in these contemporary times of word processing, they are not; they would not have added excessive bulk to the book.

Finally, there is a matter of tone. I have just mentioned the unfortunate polemical tone of much work on magic, a tone which often makes the field seem more one of warfare than Wissenschaft, as high praise and low denunciations fly about. There is a hint of that here. Thus appear judgements such as "a fine essay" or "a great step in the right direction" or, contrariwise, "much less innovative." But a specialist reader who does not share Graf's overall view of magic may take the opposite view of any such characterizations; a non-specialist may never realize an opposite view be possible. Finally, at least once Graf references "a still unpublished paper." Unfair! While a letter, e.g. can, in theory, be checked, unpublished papers (as opposed to forthcoming papers) do not always contain what is claimed for them and they are less amenable to checking. Put differently, as did Thomas Hardy's Jude, "it might have been the Ratcatcher's Daughter in double Dutch."


Pars pro toto, the very first sentence (1): "The practice of magic was omnipresent in classical antiquity." Here be theoretical issues, to be discussed shortly. But the sentence coincides with the aforementioned chapter titles and seems to represent Graf's overall conceptualization of magic -- that despite differences of practitioner and method and time and place, there exist enough constants about magic so that a study of those constants will be profitable to the point of skirting historical considerations.2

Now there is a kind of very rough chronology in chapter two, the almost obligatory (for classicists) "origins and etymology" chapter. Fifth-century Athens is differentiated from the earlier and the later. Rome fares rather less well, starting from Cicero, then back to the Twelve Tables, then forward to Pliny the Elder on the Twelve Tables. Then there is "the second, Julio-Claudian stage" (56). The discussion is brief, and non-historians will not appreciate the out-of-place reference to the Lex Cornelia (ibid., without reference to Sulla). Is there a third stage? Graf does not say; he might plausibly claim the fourth century. But his apparently total omission of magic cases from the pages of Ammianus Marcellinus, absent a disclaimer for same, constitutes an enormous flaw since the major political and historical evidence for magic in the fourth century thus vanishes. It is true that Graf utilizes evidence from that period, especially Christian authors and the Theodosian Code, in neither case with much regard for chronology, as if a legal enactment has the same timeless quality he implicitly assigns to magic throughout the book.

Why does history matter? Magic, whatever the merits of Graf's view of its timelessness, appears in the then-and-now of history; virtually the only time it has not came in the last century's ethnography, when the non-Western "primitives" were deemed to have no history in the Western sense. Religious ferment and charges of religious irregularity fly thick and fast in times of socio-political stress. Even if one rejects the possibility of, e.g. Diopeithes' law against "new gods" in Athens of the 430s, certainly some Greeks of the following century found it credible enough -- and some moderns can find it credible in light of Athenian politics of the 430s. Or is it mere coincidence that Sulla's singular rise to power and subsequent overhaul of the traditional power arrangements of the senate, tribunes, and popular assemblies also involves a law which, explicitly or implicitly, included phenomena moderns would consider magic? All Graf says (59) is that alleged magic clauses in the Lex Cornelia "met a need." What of the collocation of magic charges with adultery and treason, familiar from Tacitus and Ammianus Marcellinus?

Even when Graf does use literary evidence, there appears scant indication of the historical circumstances, apart from some early Roman material (discussed infra). Graf might reasonably argue that dating and provenance issues of his defixiones and papyri preclude historical details. But the reviewer must counter that, by choosing evidence of that ilk, important though it may be, which fits his ahistorical outlook, Graf has seriously skewed his overall view. Put differently, if Graf had included modest historical consideration, his whole thesis might collapse. Differently again, by making magic timeless and universal, Graf runs the risk of trivializing it. Differently for a third and final time, it is very conventional and nineteenth century to separate out magic -- interesting in light of Graf's aforementioned willingness to castigate what he considers others' conventionality. And this leads to another area.


Let me be clear on one point at the outset. Creating theory is one thing, using theory is another thing. Few scholars do both, nor should they -- and this is especially true of classical studies, of a strong empirico-positivist tradition now being expanded and rethought by utilization of theory. Consequently, I do not imply Graf should have theorized. But I do assert he could have done better with his use of theory.

The first chapter, a part of which (8-19) surveys the history of the study of magic disappoints. Graf knows Wissenschaftsgeschichte well, and his chapter on it in his Greek Mythology (1987; English trans. 1993) was a masterpiece as he astutely deployed history of scholarship acutely to illuminate countless issues. But now, names follow names in almost dizzying procession and not always correctly; it is absolutely incorrect to link (19) Kurt Latte with H.J. Rose as "inveterate Frazerians".3 In general, the reader who thinks history of scholarship is a history of what went wrong will have no reason to think differently. True, matters are complex, but where does Graf stand? At the very least, he could have clarified by reference to how the last century relentlessly plied the trichotomy magic/religion/science -- and that it arose from an enormous argument about the relation of religion and science, an argument which still continues today.

Too, Graf seems to avoid a fundamental issue for moderns, regardless of theory: the Judaeo-Christian tradition. It is too easy, and too common, for us today to look at paganism from the a priori of monotheism and the a priori of canonical doctrine. The result is mere lip-service to the radically different systems of polytheism. A view, or legal charge, of something as magic can, put simplistically, arise when one person does not like another's ritual, or new divinity. And if that other person seems reprehensible on other grounds, it is all too plausible to invoke claims of religious irregularity. And if the person making the charges be one with secular coercive authority, the case is closed, to the detriment of the now-magician. Or, one can view it with Peter Brown. Classical antiquity granted the existence of individuals with powers beyond those associated with day-to-day reality. The cause for debate centered on the nature of the theological entities which provided those powers, and the uses to which the human recipients put them.

That is, there is a relative quality to charges of magic, both on theological grounds, but also socio-political grounds. Pagans accuse Christians of magic; Christians accuse Pagans, and Jews, of magic. Are any of them "right?" Once again, Graf's evidential emphasis on defixiones and papyri, along with his ahistorical outlook, precludes answers. Surely understanding magicians' self-understanding matters, but the context of non-magicians' views matters too.4 Societies and their religious activities operate in the dialectic, not in hermetically sealed categories.

The other relative quality is that of the vexed issued of the relation of magic to religion. Whatever moderns may think, ancients thought otherwise. Absent the theological fixity of the varieties of Judaeo-Christian monotheism, "religion" in that modern sense simply did not exist in antiquity. There was a plethora of divinities multiplied by a plethora of theologies and rituals, further multiplied by the regional, and still further multiplied by the haphazard way by which religious information was transmitted. Anything could, in theory, be considered magical (in the modern sense) and anything could, in theory, be considered religious (also in the modern sense). Who thought what and why? That is where theory and history intersect -- unfortunately largely outside of the confines of Graf's study.


What do the Romans have to do following such methodological considerations?5 The book's purview is classical antiquity. And yet too often the Roman material comes up short, to say the least. To say the most, it is either ignored or, when it does appear, it is misused.

For example, consider the lengthy section "The Roman World" (36-60) in chapter two. Fairly enough, Graf begins with the Twelve Tables. But then the problems arise. He does not differentiate the possible problems of their quotation in different sources (Elder Pliny, Seneca). Since we rely precisely on quotation, and the inherent possibility of modernization of archaic language and thought, such consideration is crucial. Again, Graf rightly insists (42) that such early magic legislation "punishes the violation of the right to property." But his ensuing comment "damage to fields and harvests can rather quickly call into question the status of the landowner and thus harm the social equilibrium" will not do. Theoretically, it betokens functionalism and all the ahistorical problems, among others, of that now-discredited (outside of classical studies) position. Practically, a small landowner would not have any recourse, since the Twelve Tables omitted the Leges Actionis, reserving that for the socio-economic elite -- put simply, a small landowner would not know how to proceed with an action and thus could not succeed. A large landowner would, of course, know how to frame the action but might well not need to, since the propertied of archaic Rome often utilized self-help rather than law. That is, the Twelve Tables may well present legal theorizing rather than reflecting what society did.

Consider now another larger area, that of the defixiones' concern with binding and nailing. Graf aptly demonstrates that connection (chapter 5 passim, especially 121-6, 135-51) and more than once impressively observes nail holes in physical evidence. But then there is archaic Rome where nails in the temple of Juppiter Capitolinus were employed against a plague (Liv. 7.3.3). Too, nailing was connected with worship of the goddess Nortia (OCD, 3rd ed., s.v.) who, interestingly, never came to Rome (Varro, ARD 33a Cardauns = Tertul. Nat. 2.8.6). Is such nailing magical, or religious, or both? A possibly bogus question if we remove nineteenth-century scientism's opposition of magic and religion.

I have chosen two examples almost at random; they could be multiplied. I cannot explain them, although I note the discontinuity between the relatively scant, sometimes inaccurate, material on the Romans and Graf's academic position (Ordinarius for Latin Philology at the University of Basel) at the time of composing the book; further in this connexion I recall the absence of material from Ammianus Marcellinus. That is, whatever Graf's current academic interests and circumstances, de facto he seems to operate here as a Hellenist studying a Greco-Roman phenomenon. Now that is no cause for censure -- his publications on Greek religion are fundamental. But doing the Greco-Roman means doing both Greeks and Romans. In this he has much in common with Walter Burkert, whose Creation of the Sacred (1996), allegedly covered all of classical antiquity but proffered a Hellenocentric outlook.6 The implication? That the scholar who understands the Greeks understands the Romans. Although it is no longer, as it was in the previous century, a case of simply dismissing the Roman, the mythology dies hard and is the more insidious now because it creeps in under the ecumenical guise of considering the Greco-Roman.


This is a skewed book. It privileges the Hellenic evidence and the material evidence while downgrading the Roman evidence and the historical evidence; a tacit unity of ancient magic appears in opposition to ancient religion. Those versed in the religions of classical antiquity and the methods for their study may quarrel with it, as I have done. Thus, as a specialist work it contributes to the ongoing debate about magic -- as an example of a book which, apart from the increase in evidence, could easily have been written a hundred years earlier. But the field remains open for a balanced introduction for non-specialists.  


1. At the time of writing this review, it appears that the panel material will be published in Numen.  

2. Interestingly, in an earlier essay Graf takes precisely this position, while in the very first note there directing the reader to the French edition of the current book for more details; "Excluding the Charming: The Development of the Greek Concept of Magic," in M. Meyer & P. Mirecki, eds., Ancient Magic and Ritual Power (Leiden, 1995) 29-42.  

3. The linkage seems not a slip; cf. Graf's "Der Mythos bei den Römern. Forschungs- und Problemgeschichte," in id., ed. Mythos in mythenloser Gesellschaft (Stuttgart & Leipzig, 1993) 33.  

4. For readability I use the term "magician" although I do not agree with it since it relies on questionable modern categories; see my "Nullum Crimen Sine Lege: SocioReligious Sanctions on Magic," in C. Faraone & D. Obbink, eds., Magika Hiera. Ancient Greek Magic and Religion (New York, 1991) 260-76 and Jonathan Smith, "Trading Places" in Meyer et al. (above, n.2) 12-27.  

5. I am grateful to Andreas Bendlin (Braesnose College, Oxford) for helpful discussion of this section's ideas.  

6. See my "Walter Burkert In Partibus Infidelium: A Classicist Appraisal of Creation of the Sacred" forthcoming in Method and Theory in the Study of Religion, and "Walter Burkert In Partibus Romanorum" forthcoming in Religion.