BMCR 2005.12.02

Catullus’ Poem on Attis. Texts and Contexts

, , Catullus' poem on Attis : text and contexts. Catullus, Gaius Valerius. Carmen 63.. Leiden: Brill, 2005. 157 pages ; 25 cm. ISBN 9789004141322 $92.00.

[Authors and titles are listed at the end of the review.]

The focus of this volume is that delirious poem about ritual self-castration, in ninety-three exotic galliambics, which W. Y. Sellar famously called ‘the most remarkable poetic creation in the Latin language’.1 Four of these five papers were originally delivered at a so- called ‘Text-in-Context’ day at the University of Groningen in 2003; the fifth is a lecture given by Stephen J. Harrison at Groningen in 2002. The papers are accompanied by a new text and translation of the poem (Harrison) and a technical appendix that reexamines the evidence for Hellenistic galliambic poems (Nauta). The volume appeared earlier as a special issue of Mnemosyne (57.5, 2004).

According to the editors, the aim of the ‘Text-in-Context’ colloquia at the Dutch National Graduate School in Classical Studies is to illustrate how various ‘disciplines construct and approach their subject and how they may cooperate towards a better understanding of culture’. The disciplines represented are Greek and Roman literature, history of religion, and linguistics; and all five contributors approach their subject through the study of texts: four use the methods of classical philology (Harrison, Bremmer, Harder, and Nauta); the fifth (Kroon) takes a formalist approach, albeit one not typically seen in classical scholarship (text linguistic analysis). Two of the papers attempt to recover the cultural and religious contexts that the poem’s fictions presuppose (Bremmer and Nauta), but the evidence they marshal is almost entirely textual. For the most part, the contributors do indeed choose to ‘cooperate’ rather than to engage in other, less harmonious but possibly also more exciting modes of exchange. In particular, most of them share the widely-held view that although the poem is Hellenistic in character, it is not, as Wilamowitz thought, based on a Hellenistic model.2 For a different opinion one has to wait till the appendix, where Nauta reexamines and reinstates the textual and metrical evidence for the view that the poem may in fact have had a Hellenistic model, possibly even Callimachus (‘Hephaestion and Catullus 63 Again’). Ironically, then, the most subversive piece, which also promises to be one of the most cited, is the one that most convincingly reestablishes the wissenschaftliche status quo.

Stephen Harrison, whose lecture at Groningen inspired the theme of this particular ‘Text-in-Context’ day, examines three potentially hot topics through a traditional exercise in intertextuality (‘Altering Attis: Ethnicity, Gender, and Genre in Catullus 63′). On the theme of ethnicity, he shows that Attis’ voyage from one culture (Greece) to another (Asia Minor) reverses the journey of Dionysus’ devotees in the Bacchae. Regarding gender, he argues that the poet constructs Attis as ‘female’ not only through the use of feminine adjectives and pronouns, but also through ‘references’ to Euripides’ Medea and to Apollonius’ Argonautica, as well as to Euripides’ Bacchae, making the hero a variation on the ‘lamenting abandoned literary heroine’ so vividly portrayed in Cat. 64. Much of this, of course, is familiar.3 Like others, Harrison concludes that the poem has affinities not only with Euripidean tragedy, but also with Hellenistic epigram, epyllion, and literary hymn. By incorporating tragic motifs and episodes into a poem that possesses hymnic features (but see Nauta’s challenge on 100-101), Catullus reenacts the genre-mixing strategies of Theocritus 26; like the Hellenistic poets, he is engaged in a Kreuzung der Gattungen that anticipates Augustan generic experimentation.

Annette Harder surveys the “Hellenistic background” for Cat. 63 in an attempt to see ‘how prominent the Hellenistic element in Catullus’ poem was and whether it served some kind of specific purpose’ (‘Catullus 63: A “Hellenistic Poem”?’) (71). Taking what she calls ‘a systematic look at possible Hellenistic elements from various angles’ (72), she proposes five loosely-defined criteria that enable her to conclude that although the poem lacks many features typical of Hellenistic poetry (such as explicit programmatic remarks, an interest in cultic aetiology, and learned play with the mythological and literary tradition), it nevertheless makes ‘a selective, but careful and creative use of Hellenistic elements, with a certain focus on the means of allusion and Kreuzung der Gattungen‘ (65). Harder’s investigation produces results that closely resemble those of Harrison, with one novel observation: four of the poem’s admittedly infrequent ‘Hellenistic’ gestures occur in the episode in which Attis wakes up from his divinely inspired delusion and longs to return home to his Hellenistic Greek town, and the fifth occurs when the narrator ‘wakes up’ from the poem and expresses skepticism about the goddess. Harder suggests that the poet may be setting up an opposition between poetic inspiration and technical refinement; in his moments of lucidity, Attis ‘awakens’ to a memory of the tradition from which he comes.

In a knotty investigation that does little to make its conclusions accessible to the general reader, Jan Bremmer proposes to investigate the myth and cult of Attis as it was manifested in three cultures (‘Attis: A Greek God in Anatolian Pessinous and Catullan Rome’). As his title suggests, the Near-Eastern god Attis is known to us only through Greek sources, but Bremmer begins by arguing that one of the most famous of these must be dismissed. The story at Hdt. 1.34-45 is about a figure named Atys, not Attis, and although this Lydian prince is killed in a boar hunt, his story has nothing to do with the Hellenistic elegy by Hermesianax (summarized by Pausanias) in which Attis was said to have traveled from Phrygia to Lydia only to be killed by a boar. Neither text offers any evidence for an authentic old Lydian tradition about Attis. Bremmer then turns to Greece to challenge the conventional wisdom about the date of Attis’ arrival in the West. Examining the earliest literary mentions of the god, he dismisses an ironic reference to ‘that Attis of yours’ in a fragment of old comedy (Theopompus fr. 28) on the grounds that it refers to a human lover, although surely the point is that in his effeminacy the human lover resembles the mythical one. Turning instead to the anecdote in which Demosthenes uses the god’s name to tarnish Aeschines’ reputation at On the Crown 18.260, Bremmer ingeniously if implausibly shifts the date of the reference to the god from Aeschines’ youth to c. 330 BC, combining it with an Athenian votive stele dedicated to Attis to show that the god was introduced into Athens later than previously supposed.

In his third and longest section Bremmer extricates elements of original Phrygian myth and ritual from the four late Greek and Roman sources that preserve four earlier accounts of the god. Disentangling early from late details, he concludes that the four earlier accounts ‘cleaned up’ features of the Phrygian myth and/or ritual for consumption by a Greek audience. Moving on to sketch an ‘integral’ picture of the myth and ritual at Pessinus in ‘broad strokes’ (41), he attends to onomastic and other minutiae: clearing up some lingering confusion about the names of the protagonists, he shows, among other things, that the mysterious pine cone of Roman ritual, absent from the Greek sources, is an originally Anatolian detail.

Bremmer compresses his discussion of Attis in Rome into just two short paragraphs.4 As one might expect, however, he provides useful commentary on the religious details of the poem itself, particularly its fusion of Metroac and Dionysiac motifs. Noting that the connection between the cults of Cybele and Dionysus was a phenomenon already observed by Strabo, he is not surprised that so many scholars have noticed the poem’s debt to Dionysiac literature, esp. Bacchae and Theocritus 26. But although in the end he evidently agrees with Fordyce’s declaration that Catullus’ Attis bears ‘no resemblance’ to the ‘Attis of myth’, we never find out whether he also agrees that he bears no resemblance to the ‘Attis of ritual’, since he never takes a stand on whether the poem’s cultic details are Dionysiac or Metroac (presumably they are both, but exactly how is what one would like to know).

Building on an earlier article in which he discusses the promise and perils of reception theory,5 Ruurd Nauta, in his ‘Catullus 63 in a Roman Context’, proposes to reconstruct the horizon of expectation of the original audience, showing that the poem would have unsettled readers’ expectations about ‘marriage’, ‘masculinity’, pietas, and Roman-ness in general. Examining the protagonist’s sexual and cultural identity from the perspective of a contemporary Roman reader, he points out that Attis’ behaviour stood in sharp conflict with conventional norms and values. By converting himself into a woman, he has disrupted the normal passage to heterosexual adulthood from homosexual youth ( ego iuvenis, ego adulescens, ego ephebus, ego puer, 63); his present condition thus stands in striking contrast to the more conventional state of marriage celebrated in the other carmina maiora. Attis also abandons the social and economic order: a study of the allegorization of the Magna Mater myth in other Roman authors, particularly Lucretius, shows that a contemporary audience would have seen the self-castration as displaying a lack of pietas towards parents and homeland. Nauta speculates that Catullus’ audience might have read his poem as an implicit discourse about national identity. By introducing a figure associated with the Romans’ place of origin who nevertheless challenges conventional norms of Roman behaviour, the poet may be making Phrygian orgiastic effeminacy a component of Trojan and hence Roman identity. Nauta ends by declaring that the poem contains no Roman elements whatsoever (116) — although surely the leaping steps ( tripudiis 26) with which the ‘un-Roman’ galli mimic the dances of the Roman Salii could be used to support the argument that the poet is incorporating Phrygian effeminacy into Roman identity — or vice versa. Not surprisingly, this rich and sharply focused ‘reconstruction of the mental picture that Romans of Catullus’ time had of galli’ (85) shows no interest in the admittedly scanty (although not unsuggestive) evidence for Roman attitudes towards galli in the time before Catullus.6

Any methodological alarm caused by Caroline Kroon’s title (‘The Effect of the Echo. A Text Linguistic Approach to Catullus carmen 63′) is diminished by her conservative definition of text (‘a hierarchical structure of interrelated utterances… a single whole with a specific communicative aim’ (121)). Explaining that text linguistic analyses are usually performed on narrative prose texts, she observes that ‘there does not seem to be an a priori reason to exclude poetry from such linguistic analyses’ (122). Given that her method has roots in the work of Roman Jakobson, one wonders why she is defensive about applying it to a poem, and why she should be surprised to find that the poem differs in striking ways from a narrative prose text. Noting that narrative prose texts typically achieve ‘coherence’ through their ‘event structure’, she argues that in Cat. 63 coherence is achieved through a ‘theme centered structure’ involving lexical, syntactic, phonetic, rhythmical and (especially) semantic repetition. Repetition, she speculates, may be one of the poem’s central themes: it contains more instances of repetition than Catullus’ other poems, along with an unusually high number of verbs prefixed by re-. The poem’s semantic, syntactic, and ‘sound-rhythm’ repetitions may be related, she concludes, to its primary aim of depicting the protagonist’s oscillation between two opposed modes of existence and their corresponding mental states. Although Kroon effectively explains the aims and methods of text linguistics for the uninitiated, her attempts to link the formal features of the poem to the fictional world inhabited by the protagonist and the real world inhabited by the reader can be baffling, as when she declares that the poem’s repetitions ‘form the outlines of a highly schematized and sketchy “reality”‘ (140) and that the poem evokes a ‘second reality’ in which the device of repetition has become an end in itself (141). Perhaps she means that the poem’s repetitions conspire to enchant the reader just as Attis himself has been enchanted, so that the reader’s reality comes to resemble the fictional world of the protagonist. Perhaps not. As with several of the other articles, some of the more interesting implications are left to the reader to extract.

Kroon’s formalist analysis makes a useful contribution to the volume by recovering the poem itself (as opposed to its intertexts and contexts) as an object of study. In the end, however, Catullus’ Attis eludes his pursuers. Nevertheless, even if this collection sometimes paints a clearer picture of its contributors’ scholarly preoccupations than of the poem’s dazzling ritual poetics, it is a valuable resource that provides a provocative starting point for future work.

The articles are accompanied by individual bibliographies ranging in length from half a page (Kroon) to five pages (Bremmer); there is a general index and an index locorum. The rest of the volume shows signs of hasty production. One wishes there had been time to clean up some of the more jarring infelicities of English style: at least two of the articles are marred by odd locutions, run-on sentences, repetitiousness, comma errors, and footnote-ese. There are a number of minor typographical errors; ‘rejoycings’ should read ‘rejoicings’ (46); ‘derwishes’ > ‘dervishes’ (54 n. 145); ‘Hellenistic model’ > ‘a Hellenistic model’ (59); ‘Ziegler, K’. > ‘Ziegler, K. 1969’ (64); ‘had’ > ‘has’ (bottom of 68); ‘Harrison, 21’ > ‘Harrison, 20’ (100, n. 51); ‘form’ > ‘from’ (108); ‘have got’ > ‘have gotten’ or ‘get’ (bottom of 125); ‘aimlessly’ > ‘aimless’ (135), etc. More disruptive are the cut-and-paste errors on p. 19, where chaos reigns in footnotes 21 and 22; and on pp. 70-71, where several sentences (and sentence fragments!) in the body of the text should be relegated to footnotes, especially the ones beginning ‘E.g. Quinn (1970, 283) speaks about …’ and ‘Thus e.g. Syndikus (1990, 80) found that …’. On p. 78, footnote 20 refers to a note housed on p. 77.


Stephen Harrison, critical text and translation of Catullus 63

Stephen Harrison, ‘Altering Attis: Ethnicity, Gender and Genre in Catullus 63’

Jan N. Bremmer, ‘Attis: A Greek God in Anatolian Pessinous and Catullan Rome’

Annette Harder, ‘Catullus 63: A “Hellenistic Poem”?’

Ruurd R. Nauta, ‘Catullus 63 in a Roman Context’

Caroline Kroon, ‘The Effect of the Echo. A Text Linguistic Approach to Catullus carmen 63′

Ruurd R. Nauta, ‘Appendix: Hephaestion and Catullus 63 Again’.


1. W. Y. Sellar, Roman Poets of the Republic, 3rd ed. (Oxford, 1889) 461.

2. U. von Wilamowitz, Hermes 14 (1879)194-201 = Kleine Schriften II, 1-8; Hellenistische Dichtung in der Zeit des Kallimachos II (Berlin, 1924) 291-95. J. P. Elder considered this suggestion ‘unsupportable’ as early as 1947: see AJP 68 (1947), 394-403. For a full statement of the skeptical view, see D. Mulroy, Phoenix 30 (1976) 61-72.

3. See e.g. G. O. Hutchinson, Hellenistic Poetry (Oxford, 1988) 310-14, and M. Fantuzzi and R. Hunter, Tradition and Innovation in Hellenistic Poetry (Cambridge, 2004) 477-85, originally published as Muse e modelli: la poesia ellenistica da Alessandro Magno ad Augusto (Rome, 2002).

4. Surely even the relative silence of the sources deserves more comment; see e.g. G. Thomas, ‘Magna Mater and Attis’, ANRW II.17.3 (1984), esp. 1508-12, and F. Bmer, ‘Kybele in Rom. Die Geschichte ihres Kults als politisches Phänomen’, Röm. Mitt. 71 (1964), 130-51. Bremmer also brushes aside 94 terracotta images of Attis found on the Palatine and dating to the period from 191-111 BC; for which see L. E. Roller, In Search of God the Mother: The Cult of Anatolian Cybele (Berkeley-Los Angeles, 1999) 271-80.

5. R. R. Nauta, ‘Historicizing Reading: the Aesthetics of Reception and Horace’s “Soracte Ode”’ in I. de Jong and J. P. Sullivan, eds. Modern Critical Theory and Classical Literature. Mnemosyne supplement 130 (Leiden, 1994) 207-30.

6. For Roman attitudes towards Attis and galli in the century and a half before Catullus, see n. 4 above; esp. suggestive for Cat. 63 is the incident in which a slave of Q. Servilius Caepio (cos. 106) castrated himself in the service of the Magna Mater and was trans mare exportatus ne umquam Romae reverteretur (101 BC) (Iulius Obsequens, Prodigiorum liber 44).