A Companion to Catullus

CHAPTER ONE

Introduction

Marilyn B. Skinner

Catullus, as William Fitzgerald acutely observes, is a poet whom we have taken rather too much to our hearts (1995: 235). For a considerable part of the nineteenth and twentieth centuries, both lay and academic audiences reacted to the lyric voice in the Catullan collection as that of a friend and contemporary, whose grief over a brother’s death and anger at betrayals of trust struck us as candid, universally human responses to circumstance. Yet treating Catullus sympathetically as one of ourselves greatly impeded efforts to appreciate his literary achievement as a whole and to locate his poetry within its particular cultural and historical milieu. New Criticism finally taught readers to value the longer works of the learned Alexandrian Catullus and even to relish displays of erudition in the love poetry, but only at the price of dismissing his barbed invective and his coarsely funny occasional pieces as material supposedly displaying a lower level of intent (Quinn 1959: 27–43). Appreciation of the Catullan corpus, obscenity and all, in its entirety and within its proper context had to wait for the rise of New Historicism in the 1980s and the subsequent impact of the cultural studies movement on the humanities.¹

It is just since the 1980s, then, that wide-ranging research has succeeded in grounding Catullus firmly in the socio-historical world around him – by investigating his provincial North Italian background, his family connections, and his dealings with the Roman elite; by observing his interactions with fellow provincials seeking advancement; by teasing out references to matters of everyday life in his poems; by studying, lastly, the circumstances under which his works were produced and disseminated and what they might have conveyed to the audiences at which they were aimed. This historicizing approach has proved unusually fruitful; since Wiseman’s Catullus and His World (1985), influential articles and entire monographs on Catullus have appeared with increasing frequency. Such recent critical studies have employed a variety of incisive tools, including those of anthropology, cultural studies, gender theory, Lacanian psychology, performance theory, reader-response theory, and sociolinguistics, to delineate the basic cultural and rhetorical frameworks within which the poetry operates. They have given us a more nuanced grasp of Catullus’ language and poetics and his standing among his contemporaries.

Unfortunately, this ferment in present critical discourse seldom trickles down to high-school or even undergraduate college classrooms, although on both levels of Latin instruction Catullus is now one of the three ancient authors most commonly encountered. As Ancona and Hallett demonstrate in this volume, his current pedagogical popularity is likewise a nascent phenomenon. Within the living memory of many North American teachers, Catullus was a text assigned only on the college level, and then with some trepidation: despite their relatively easy syntax and their immediate emotive appeal, the poems were deemed simply too racy for the young. Incorporation into the Advanced Placement syllabus (for examinations usually taken in the senior year of high school, approximately age 17) gradually furthered Catullus’ secondary-school canonicity, though he was not finally accepted as a core AP author until 1994. Consequently, although annotated teaching texts and materials on the poet have proliferated over the past few years, and good general introductions, such as those of Martin (1992) and Hurley (2004), are available, students and teachers looking for more detailed summaries of current scholarly opinion find nothing really suitable in English. Hence the Blackwell Companion to Catullus appears to be a timely project. Containing essays on a range of topics by recognized and emerging authorities and drawing together two decades’ worth of research into a collection adaptable for classroom use, this volume is intended to present C. Valerius Catullus to a wider public as a writer who was very much a man of his time and a perceptive eyewitness to the last troubled decade of the Roman Republic.

Unlike most studies of literary figures that attempt to reach out to non-specialist readers, the Blackwell Companion to Catullus does not begin with a chapter on the author’s life, for the very good reason that we know almost nothing about it. Texts, translations, surveys, and entries in reference works dating from earlier periods do contain short biographies of Catullus. Most have been based, directly or indirectly, upon Ludwig Schwabe’s 1862 reconstruction of his career, known to those of us in the field as the Catullroman (Catullus novel). As that term of art hints, Schwabe’s account is quite speculative, and prior biographies that leaned on it wove the scant data into highly imaginative scenarios. They focused on Catullus’ affair with the pseudonymous Lesbia, generally assumed to be Clodia, wife of Metellus Celer (cos. 60 BC) and sister of the radical demagogue P. Clodius Pulcher (tr. pop. 58). Drawing heavily on the first-person statements in the poetry, and treating artistic utterances as confessional pronouncements, they represented their subject as the disillusioned lover of a corrupt and degenerate noblewoman and attributed his purported early death to the suffering caused by that experience (or, alternatively, to tuberculosis, on no evidence whatsoever).

Here, too, the new socio-historical approach results in a changed emphasis. We can still start with the few external facts. Following earlier authorities, the late-antique chronicler Jerome reports Catullus’ birth at Verona in 87 BC (Chron. 150 H.) and assigns to 58–57 BC his death at Rome during his thirtieth year (XXX aetatis anno, Chron. 154 H.). The latter date is demonstrably incorrect: all the poems in the collection to which dates can be ascribed were written during the period 56–54 BC, though we find no unambiguous reference to events subsequent to 54. Most scholars, accordingly, have treated Catullus’ life-span of 29 years as fixed and moved the date of birth down to 84; there has been a recent tendency to shift the death-date as well, down to 52 or even 51 (Granarolo 1982: 19–30; Wiseman 1985: 191; Thomson 1977: 3–4). But there is a possibility that the number XXX could be a scribal error; might Catullus have instead lived almost to the age of forty (XXXX) and thus seen the outbreak of civil war? Cornelius Nepos, to whom he dedicated his libellus, confirms that by 32 BC he was dead (Att. 12.4), but we have no idea how long before that he died, or what, if anything, he might have been doing after 54 BC.²

In his Life of the Deified Julius (73), the biographer Suetonius records: Valerium Catullum, a quo sibi uersiculis de Mamurra perpetua stigmata imposita non dissimulauerat, satis facientem eadem die adhibuit cenae hospitioque patris eius, sicut consuerat, uti perseuerauit ([Caesar] had not denied that Valerius Catullus had put a lasting mark of shame against his name by his lampoons concerning Mamurra, but, on the same day Catullus apologized, Caesar invited him to dinner and continued to accept the hospitality of Catullus’ father, just as he had been accustomed to do). In this volume, T. P. Wiseman unpacks what this sentence tells us about the social standing of Catullus’ family, and David Konstan explores its implications for Catullus’ view of politics. I have elsewhere noted (Skinner 2003: xxi) that, with a father still alive, Catullus would have been a filiusfamilias, or son subject to paternal authority (potestas), legally unable to own property and dependent upon others for his living expenses in Rome. That would make his vitriolic personal attacks upon his father’s guest, no less a personage than the military governor of Cisalpine Gaul, all the harder to explain. In the absence of extenuating circumstances, about which we know nothing, one wonders how on earth Catullus thought he could get away with embarrassing the family so blatantly.

The last bit of information contained in other sources is Apuleius’ testimony (Apol. 10) that Lesbia was a cover name for a woman named Clodia. That statement is corroborated by internal evidence, for in poem 79 Catullus informs us that Lesbius (who, in accordance with Roman nomenclature, must be some paternal relation of Lesbia) is Pulcher, a broad hint at the notorious Clodius Pulcher. As Dyson Hejduk explains (below, pp. 254–5), the identification of Clodia Metelli as Catullus’ mistress is not wholly certain, but there is a reasonable probability that it is correct, given her own social and political visibility. These days, though, historians are less interested in the details of the affair (if it was real) and more concerned with their implications for Catullus’ contemporary Roman audience. In the poems, a married woman associated with a powerful aristocratic clan is not only adulterously involved with the speaker, a young Transpadane, but accused of indiscriminate relations with named and unnamed others and figuratively branded in cc. 37 and 58 a common prostitute. Few today would accept this as a realistic picture of a noblewoman’s life. The cruel beloved is a standard generic component of ancient erotic verse (Dixon 2001: 137–40), and libelous charges of sexual immorality were part of the orator’s and the politician’s rhetorical gear, unscrupulously deployed against female as well as male opponents. Is the construction of Lesbia in the corpus just an assemblage of literary topoi, though, or does it also pass a harsh judgment upon the social scene in which she moved? There would be little point to the poet’s dramatic revelation that Lesbia was the aristocratic Clodia if the world of Roman politics were not somehow relevant to her literary and symbolic function. W. Jeffrey Tatum in this volume consequently finds a telling parallel between her lack of personal integrity and the high-handed way in which the nobility, in Catullus’ eyes, was exploiting the municipal equestrian class, and Konstan provocatively analyzes her insatiable promiscuity in c. 11 as a trope for Rome’s wars of imperial expansion and plunder.

From the poems themselves we learn a few additional facts: that Catullus served for a year in Bithynia on the personal staff of the propraetor C. Memmius, probably in 57–56 BC (cc. 10, 28, 46); that the loss of an elder brother, who died and was buried in the Troad, was a devastating blow (cc. 65, 68a–b, 101); that his family owned property on the peninsula of Sirmio, near Verona (c. 31), and also an estate (most likely a working farm) somewhere between upscale Tibur and the rustic Sabine district (c. 44); that he formed close ties at Rome with numerous other poets and intellectuals (Cinna, Cornificius, his great friend Licinius Calvus, the brothers Asinii, Nepos, probably Valerius Cato) and was acquainted with several distinguished Roman senators, members of the nobility, and key players, including Cicero, Gellius Publicola, Hortensius Hortalus, Manlius Torquatus, and Cicero’s influential ally P. Sestius. For a young unknown provincial, Catullus must have climbed the social ladder in Rome very quickly. Did he simply make the most of good connections, or were other talents brought to bear?

More and more Catullan scholarship is embracing a theory of performativity: that many of Catullus’ poems were originally scripts for live recital by their author, most likely at banquets to which he had been invited, and that in those scripts the speaker fashions a self-image that will further his goals and ambitions. Critics emphasize various and sundry elements implicated in Catullan performance: Selden (1992) considers it a form of rhetorical, and Krostenko (2001a) a mode of linguistic, critique; Fitzgerald (1995) studies it as a tool for controlling and manipulating audience response; Wray (2001) analyzes it as a display of competitive masculinity; more pragmatically, I have suggested (1993a, 2001) that live performance was a tactic allowing a talented outsider to curry favor with those able to help him advance socially, economically, and perhaps politically.³ Several chapters in this volume acknowledge the likelihood of convivial recitation, but it is Elena Theodorakopoulos’ reading of poem 68 in light of that assumption that reveals how postulating a back story of performance on private occasions may clarify old Catullan questions. Consequently, imagining the presence of the poet as a guest, a well-known artist and entertainer, in the dining rooms of leading Roman personages allows us to view him as someone not only having access to privileged information about the workings of power but also very much concerned about its concrete use and abuse.

Contributors to this volume examine current developments in traditional, as well as new, areas of Catullan research. In part I, The Text and the Collection, J. L. Butrica reviews the transmission of the Catullan text from antiquity to the present day, while I myself offer an account of the debate over the vexed question of authorial arrangement (a chore I hesitated to impose on any colleague). Part II, Contexts of Production, then introduces us to the numerous ways in which Catullus’ poetry can be regarded as reflective of its times. T. P. Wiseman, who pioneered investigation of the poet’s family and its later fortunes (Wiseman 1987), provides a history of the Valerii Catulli and their presence in Northern Italy. David Konstan examines the contemporary political scene in Rome, offers an explanation for Catullus’ direct attacks on Caesar and Mamurra, and, most interestingly, finds political reverberations in other ostensibly non-political poems. Andrew Feldherr locates Catullus’ studied appeal to a learned coterie in the context of larger intellectual debates over Hellenization and shows how he and his fellow provincials employed learning to their advantage as they jockeyed for status within the circles of the Roman nobility. Elizabeth Manwell provides an overview of research on gender and masculinity and then analyzes contradictory paradigms of masculinity in Catullus, a matter that has received considerable attention in recent years.

Later generations habitually characterized Catullus as doctus, learned, in tribute to his impressive acquaintance with the earlier poetic tradition. Although numerous predecessors exercised influence on his work, he himself recognizes Sappho and Callimachus as his primary poetic models. In part III, Influences, Ellen Greene shows how Catullus’ appropriation of the Sapphic voice enables him to express his private erotic subjectivity – yet, by disrupting conventional gender polarities, likewise destabilizes his own sense of male identity. Peter E. Knox provides a concise introduction to Callimachus, including a review of his most important works and an explanation of the innovative features of Callimachean poetics; Knox then surveys the far-reaching effects of Callimacheanism on the Roman poetic tradition, from Ennius through Catullus and his fellow neoterics, down to the Augustan Age.

Catullan language and style are distinctive. In part IV, Stylistics, three authorities investigate those formal aspects of the poetry. We still speak of the Catullan revolution as an abrupt break with previous artistic techniques. W. R. Johnson wittily elucidates Cicero’s grumpy reactions toward the poets he christened the neoterics and considers possible reasons why Catullus and his colleagues might have developed their innovative poetics. George A. Sheets analyzes the elements of Catullan style—diction, rhythm and meter, pragmatics—that endow it with its characteristic flavor, while Brian A. Krostenko shows that Catullus’ deployment of the vocabulary that connotes elegance (or the reverse) plays upon ambivalent cultural attitudes toward displays of aestheticism in the political arena.

The Catullan corpus is by no means homogeneous – indeed, no other Latin poetic collection manifests such diversity in genre, meter, tone, and subject matter. Critics therefore frequently treat thematically related groups like the Lesbia poems as coherent elements of the collection and approach some of the longer poems, cc. 64 and 68 in particular, as independent compositions worthy of monographs. In part V, Poems and Groups of Poems, we find studies of thematic categories, as well as in-depth readings of those two major works. William W. Batstone considers a set of poems commonly labeled programmatic pieces and boldly inquires what the label means and whether it can justifiably be applied: what makes verses programmatic, and is the program in the author’s eye or the eye of the reader? Julia T. Dyson Hejduk examines the large body of poems thought to relate to the poet’s affair with Lesbia, finding, intriguingly enough, not one but three distinct Lesbias, with contrasting poetic functions. Vassiliki Panoussi rereads the wedding compositions, 61 and 62, from an anthropological perspective. As re-enactments of ritual activity, each examines weighty cultural issues: tensions between male and female, conflict of personal desires and societal demands, continuation of the family line, sexual fidelity – all topics privately meaningful to the Catullan speaker as well.

Current work on Catullus 64, the short epic known today as The Wedding of Peleus and Thetis, concentrates upon its intertextual relations with predecessors and uncovers the implications of allusions to earlier Greek and Latin masterpieces. Jeri Blair DeBrohun’s chapter on this epyllion specifically analyzes its use of Apollonius of Rhodes’ Argonautica. This Hellenistic poem, she concludes, underlies Catullus’ text in unsuspected ways: it determines the essential structure of the narrative and, through ominous reflections of the suppressed tale of Jason and Medea, tropes the poet’s indebtedness to the past as intergenerational conflict. Elena Theodorakopoulos carefully walks the novice through the massive array of textual and interpretive problems associated with Catullus 68, which, for her, becomes an exceptional attempt to achieve permanence by overcoming the limitations of time and mortality. Finally, W. Jeffrey Tatum considers the function of Catullan invective: beginning with a consideration of the role of polemic in Roman political debate, he examines the conventions of political abuse as they are reflected in Catullus’ poetry and analyzes the hidden messages in Catullan obscenity, showing that the concerns expressed are of a piece with the ethical stance of the speaker throughout the corpus. Despite the apparent diversity of the collection, then, certain leitmotifs emerge that provide an overall impression of engaged social commentary.

How did Catullus’ subsequent readers view his poems, and how have their reactions to the author shaped the ways in which we read him? Reception theory – which studies how later perceptions, products of their own time, are mapped onto the original poem and become part of the text we confront – is represented in this volume by the series of chapters grouped under the rubric of part VI, Reception. Four of these essays deal with responses to Catullus in antiquity. For Randall L. B. McNeill, the great problem is Horace’s apparent dismissal of Catullus as a precursor and model: was the later lyric poet really as ungenerous as he seems? Vergil, on the other hand, makes sophisticated and often poignant reference to certain poems; going beyond a mere listing of passages, Christopher Nappa’s chapter seeks to envision Catullus as Vergil might have perceived him. In Paul Allen Miller’s view, Catullus, not Gallus, is the real inventor of Latin love elegy and poem 68 the single text that gave birth to it; Miller’s reading of 68b complements and complicates Theodorakopoulos’s in taking it as the expression of a polarized subjectivity. Martial, according to Sven Lorenz, redefines Catullus as primarily a composer of iambics and invokes his practices to justify the use of aggressive obscenity, meanwhile insisting that his own joking verses do no harm. This section concludes with two studies of Catullus’ reception in later periods. Julia Haig Gaisser tells of the rediscovery of the text at the beginning of the Renaissance and the slow process of purging its most egregious errors; her account spells out the debt Catullus owes to his earliest editors and commentators. Brian Arkins surveys his assimilation by Romantic, Victorian, and twentieth-century poets and critics, who together created a sentimental image of Catullus still lingering as a ghostly presence in our classrooms.

We come then to the question of how Catullus is to be presented to students, as explored by veteran instructors in part VII, Pedagogy. Ronnie Ancona and Judith P. Hallett discuss problems stemming from the relatively recent adoption of Catullus as a high-school author. Given the short tradition of teaching Catullus in the United States, they find that Latin instructors are less advantaged than their colleagues in the United Kingdom, where his poetry has been on the syllabus for decades. Ancona and Hallett also discover that British and American pedagogical treatments of Catullus differ considerably: in Britain, the biographical approach to the poet is still in vogue, while in the American classroom that method is no longer popular. Acquainting students with the sexually explicit poems is still a controversial matter; teachers may benefit from the authors’ suggestions on that point. Ancona and Hallett’s chapter is followed by that of Daniel H. Garrison, who offers practical strategies for teaching Catullus in college. This juxtaposition of chapters reveals that articulation between levels of instruction is a major educational problem. The poems of Catullus that AP students have read in the high-school classroom were, in Garrison’s words, a thoughtfully chosen subset of his work that was tailored to their youth rather than the complexity of Catullus’ actual oeuvre (p. 516), and their experience of him in college will consequently involve learning to read him in a more sophisticated way. The question Ancona and Hallett pose – Whose Catullus? – is therefore a pertinent one: is he the intellectual property of scholars, kindergarten through twelfth-grade teachers, college teachers, or their respective students? Each category of readers, it seems, views him from a distinct perspective not easy to reconcile with those of the others.

Lastly, there is the Catullus many readers confront only through the medium of an English translation. In part VIII, Translation, Elizabeth Vandiver explains just how difficult rendering Catullus into another language can be. It is hard to find equivalents for both meter and vocabulary, and obscene words pose their own particular difficulties, for Roman cultural assumptions are not the same as ours. Some poems depend on an equivalence of sound and meaning, and others, the longer Alexandrian poems, derive weight from the learned obscurity of their mythological references; how can these effects be replicated in an idiom and a poetics as alien as those of English? As Vandiver finally shows, Catullus’ own ventures in translating from Greek to Latin opt for free adaptation rather than strict fidelity to the language and meaning of the original. Perhaps there is solace in knowing that the poet had at least some inkling of his modern translator’s dilemma.

Although the Companion to Catullus was intended as a reference work, authors were encouraged to go beyond summarizing received critical attitudes and urged to supply the reader with original insights into their subject matter. These chapters can therefore be regarded as innovative contributions to the field. Some break new methodological ground when attempting to offer solutions to long-standing problems. Others frankly acknowledge the controversies that swirl around an author whose surviving text is so lacunose and problematic and whose life is very much a mystery; while they do not reach a definite conclusion on a particular topic, then, they seek to present a balanced survey of all the evidence bearing upon it. Researchers may find it expedient to refer to such essays for capsule accounts of the state of a given question. Students and teachers, for their part, should feel confident that this volume contains the most reliable and up-to-date opinion on Catullus and his unique place in Roman intellectual and literary history. Finally, each of our contributors is at pains to demonstrate that the poet’s artistry, despite its embeddedness in its own cultural milieu, will perpetually speak to the current generation in the form of a lepidus nouus libellus, as a fresh new voice.

NOTES

1 For an assessment of how these two approaches have affected present critical investigations of literature, see Klein (2005: 83–106).

2 In Skinner (2003: 181–3), I tentatively advanced the idea that the historical Catullus might have married and continued the family line. Wiseman, thinking in similar fashion, now calls attention to the small fragment of fresco recovered from the imperial-age villa at Sirmio that depicts a young man holding a scroll. As the scroll could indicate someone distinguished in the literary realm, it may, he suggests, represent Catullus himself (below, pp. 65–6). One other clue to the figure’s identity is his barefoot state, for the analogous bare feet of the Prima Porta Augustus are a symbol of heroization (Müller 1941: 496–7; Galinsky 1996: 161). Deceased ancestors were objects of familial cult: in a letter dubiously attributed to Cornelia, mother of the Gracchae, the writer envisions her son C. Gracchus paying her posthumous rites: ubi mortua ero, parentabis mihi et inuocabis deum parentem (when I am dead, you will make ritual offerings to me and call upon your parent as a divinity, ap. Nep. fr. 2). As a recipient of cult, the young man must be a recognizable and not a generic individual, conceivably the ancestor of the person responsible for the décor of the villa. That is not conclusive proof, of course, but perhaps it is evidence enough to permit serious consideration of the hypothesis.

3 Although some still adhere to the older view that Catullus rejected politics to devote himself to a life of art and enjoyment (e.g., Miller 1994: 134–6), we now see increasing consideration of his use of poetry to negotiate his cultural identity and his provincial status among members of the Roman elite (Fitzgerald 1995: 185–211; Habinek 1998: 94–6) and to critique Roman society from that perspective (W. J. Tatum 1997; Nappa 2001). Because employment on a provincial governor’s staff was one recognized way to launch a political career, Catullus’ term of service abroad with Memmius may also have been undertaken for motives beyond his (ironically) professed hope for self-enrichment.

WORKS CITED

Dixon, S. 2001. Reading Roman Women: Sources, Genres, and Real Life. London.

Fitzgerald, W. 1995. Catullan Provocations: Lyric Poetry and the Drama of Position. Berkeley, Los Angeles, and London.

Galinsky, K. 1996. Augustan Culture. Princeton, NJ.

Granarolo, J. 1982. Catulle, ce vivant. Paris.

Habinek, T. N. 1998. The Politics of Latin Literature: Writing, Identity, and Empire in Ancient Rome. Princeton, NJ.

Hurley, A. K. 2004. Catullus. London.

Klein, J. T. 2005. Humanities, Culture, and Interdisciplinarity: The Changing American Academy. Albany, NY.

Krostenko, B. A. 2001a. Cicero, Catullus, and the Language of Social Performance. Chicago and London.

Martin, C. 1992. Catullus. New Haven, CT, and London.

Miller, P. A. 1994. Lyric Texts and Lyric Consciousness: The Birth of a Genre from Archaic Greece to Augustan Rome. London and New York.

Müller, V. 1941. The Date of the Augustus from Prima Porta. American Journal of Philology 62.4: 496–9.

Nappa, C. 2001. Aspects of Catullus’ Social Fiction. Frankfurt.

Quinn, K. 1959. The Catullan Revolution. Melbourne. Rpt. Cambridge 1969; Ann Arbor, MI, 1971. 2nd edn. London 1999.

Schwabe, L. 1862. Quaestiones Catullianae. Vol. I. Giessen.

Selden, D. L. 1992. "Ceveat lector: Catullus and the Rhetoric of Performance." In R. Hexter and D. Selden, eds., Innovations of Antiquity. New York and London. 461–512.

Skinner, M. B. 1993a. Catullus in Performance. Classical Journal 89: 61–8.

Skinner, M. B. 2001. Among Those Present: Catullus 44 and 10. Helios 28: 57–73.

Skinner, M. B. 2003. Catullus in Verona: A Reading of the Elegiac Libellus, Poems 65–116. Columbus, OH.

Tatum, W. J. 1997. Friendship, Politics, and Literature in Catullus: Poems 1, 65 and 66, 116. Classical Quarterly n.s. 47: 482–500.

Thomson, D. F. S., ed. 1997. Catullus. Edited with a Textual and Interpretative Commentary. Toronto.

Wiseman, T. P. 1985. Catullus and His World: A Reappraisal. Cambridge.

Wiseman, T. P. 1987. Roman Studies Literary and Historical. Liverpool.

Wray, D. 2001. Catullus and the Poetics of Roman Manhood. Cambridge.

PART I

The Text and the Collection

CHAPTER TWO

History and Transmission of the Text

J. L. Butrica

Every work of classical literature extant today has survived through its own unique textual tradition, usually involving copies on parchment or paper of various dates from late Antiquity to the Renaissance, sometimes even papyrus copies from Antiquity itself. The poetry of Catullus is one text of all too many whose survival into the modern world depended upon a single copy – a fact with significant and often unfortunate consequences for our understanding of this author and his work.

Tom Stoppard’s play The Invention of Love deals in part with A. E. Housman’s work on the textual tradition and textual criticism of another Roman poet, Propertius. To clarify the scholarly basis of that work for his audience, and classical textual criticism in general, Stoppard used Benjamin Jowett, the famous translator of Plato, as an unlikely mouthpiece for a speech dealing with the transmission of Catullus:

This morning I had cause to have typewritten an autograph letter I wrote to the father of a certain undergraduate. The copy as I received it asserted that the Master of Balliol had a solemn duty to stamp out unnatural mice. In other words, anyone with a secretary knows that what Catullus really wrote was already corrupt by the time it was copied twice, which was about the time of the first Roman invasion of Britain: and the earliest copy that has come down to us was written about 1,500 years after that. Think of all those secretaries! – corruption breeding corruption from papyrus to papyrus, and from the last disintegrating scrolls to the first new-fangled parchment books, with a thousand years of copying-out still to come, running the gauntlet of changing forms of script and spelling, and absence of punctuation – not to mention mildew and rats and fire and flood and Christian disapproval to the brink of extinction as what Catullus really wrote passed from scribe to scribe, this one drunk, that one sleepy, another without scruple, and of those sober, wide-awake and scrupulous, some ignorant of Latin and some, even worse, fancying themselves better Latinists than Catullus – until! – finally and at long last – mangled and tattered like a dog that has found its way home, there falls across the threshold of the Italian Renaissance the sole surviving witness to thirty generations of carelessness and stupidity: the Verona Codex of Catullus; which was almost immediately lost again, but not before being copied with one last opportunity for error. And there you have the foundation of the poems of Catullus as they went to the printer for the first time, in Venice 400 years ago. (Stoppard 1997: 24–5)

Jowett can be criticized on some minor points: he neglects causes of corruption other than scribal error, ignores the secondary tradition, elides the considerable scholarly activity that intervened between rediscovery and first publication, and in general downplays the element of sheer uncertainty that surrounds the whole enterprise of recovering an ancient text: not to mention that, if Christian disapproval was ever a factor, it seems to have left some of Antiquity’s most flagrantly obscene poetry unmolested. On the whole, however, though the last century and a quarter of scholarship allows us to refine this picture to a considerable extent, it is accurate in its essence: about the transition in Antiquity from papyrus roll to codex, about certain causes of scribal corruption such as unfamiliar scripts, about interpolation, about the alteration of archaic forms, about that single manuscript at Verona, now lost, known as the Verona codex or Veronensis (V), from which all the complete copies of our Catullan corpus derive, and about the vast temporal gulf between Catullus himself and the earliest extant complete text – though 1,400 years would be more accurate than 1,500.

The Text in Antiquity

To understand how our extant MSS of Catullus reflect what Catullus wrote, how he might have arranged his works, and how they circulated in Antiquity, we begin of course by studying the text as attested in Antiquity itself. It might seem an obvious starting point to say that the history of the text begins with the author himself and his own publication of his work, but publication is still a nebulous concept at this point in the Roman Republic, a generation before the first booksellers known by name (see, in general, Starr 1987). It is hard to believe, however, that a literary culture like that of the neoterics (the newer or modern poets, comprising Catullus, Licinius Calvus, Helvius Cinna, and others) – a culture that was, like that of the Greek Alexandrians whom they followed in so many respects, dependent upon both literary and scholarly texts – relied solely upon books made by the poets’ own slaves from copies that their friends happened to own. In fact, Catullus himself refers to running to the booksellers’ cases in search of bad poetry with which to avenge himself upon his friend Calvus (c. 14.17–18, ad librariorum/curram scrinia). Whatever the involvement of the book trade, however, the publication of short or occasional works in particular (the majority of Catullus’ poems) is still likely to have been largely a matter of distributing handwritten copies to friends or to influential grammatici like Valerius Cato (the closest thing then to literary critics), or allowing others to have their own scribes make copies. But at some point, with or without the cooperation of the authors, booksellers were apparently able to offer the (presumably) collected Epigrammata or Poemata of the neoterics; and Catullus himself became a standard text in the book trade, not only as a part of the common literary heritage of the educated Roman who read for pleasure but – fittingly for a neoteric scholar-poet – as a learned authority whose practice mattered in establishing correct Latin usage.

Since there are no ancient copies of Catullus extant (not even among the graffiti of Pompeii), our knowledge of the text in Antiquity depends upon the evidence of what is called the secondary tradition, references and quotations in ancient writers and glossaries as opposed to the primary tradition represented by extant manuscripts of the text itself. The relatively extensive secondary tradition of Catullus sometimes reveals or corrects errors of the primary tradition and casts light upon how the text has been corrupted since Antiquity; it also supplies evidence bearing on the all-important question of how the poet himself arranged his works for circulation, though it comports some awkward problems as well. (The very full collection of testimonia in Wiseman 1985: 246–62 can be supplemented occasionally from Manzo 1967, on the glossarial tradition.)

An excellent starting point for understanding how the reliance on handwritten copies affected the condition of the text and how ancient copies of Catullus relate to the text of our MS tradition is the passage where the second-century AD author Aulus Gellius discusses the text of Catullus 27.4 (NA 6.20.6). He quotes the poem’s first four lines, with the last reading (in modern editions of Gellius like Marshall’s OCT) ebria acina ebriosioris (drunkener than the drunken grape). Gellius maintains that this is the correct reading: "though [Catullus] could have said ebrio, and could have employed the more customary neuter form acinum, nevertheless he said ebria out of a fondness for the sweetness of that Homeric hiatus, because of the harmony with the following letter a. Moreover, those who think that Catullus said ebriosa or ebrioso – for this too is found written by accident – have of course encountered books written from corrupted originals. Gellius’ argument, then, is that some writers like the aesthetic effect of what is called hiatus" (in particular when the same vowel ends one word and begins the next without the first being elided and thus, in effect, eliminated; he has already praised its use in Homer), and that Catullus employed it here, writing ebria acina, with hiatus and without the normal elision.

At best, Gellius only implies that the reading is attested in copies of the poem; one could be forgiven for thinking that he was perversely trying to justify a corrupt reading in a copy he had seen or even for suspecting that it was in fact a conjecture of his own. In any case, he notes that two other readings, ebriosa (acina) and ebrioso (acino), are also found in contemporary copies: disconcerting though it is to find three different readings attested within only two centuries of composition, the situation is not surprising in a world of handwritten books, sometimes produced without the quality control of correction from the exemplar or from other copies, and with an almost inevitable proliferation of fresh errors in each new version. It is also disconcerting that the archetype of the Catullan tradition transmitted yet another corruption, ebriose acino. Few editors nowadays accept Gellius’ argument, and Fordyce (1961: 158) has explained why Gellius’ reading should be rejected: hiatus does not certainly occur elsewhere in Catullus’ hendecasyllabics, while the use of an adjective in both its base and comparative forms within a short span (here ebriosus and ebriosior) appears with some frequency. Of course the claim that ebrioso (acino) is found in copies from corrupt exemplars should be taken with the grain of salt it deserves, since it is presumably determined solely by the desire to champion ebria acina.

One obvious difficulty in dealing with the secondary tradition is the fact that, like primary sources, secondary sources also survive through textual histories in which accidental corruption may occur; in fact, in the passage just discussed, the reading ebria acina that Gellius’ discussion presupposes had to be restored by the nineteenth-century scholar Moritz Haupt from the corrupted version given by the MSS of Gellius’ work, which have ebriose ac in or ebriose ac me – and those MSS also corrupted every word that Gellius subsequently quoted here as known to him from copies of Catullus! (Again the corrections printed in modern editions are due to Haupt.) Incidentally, there just might be evidence here for interaction between primary and secondary sources. Certainly it is curious that both the corrupt archetype of the Catullan tradition and the corrupt text of Gellius agree in the clearly erroneous reading ebriose; this could be coincidental, of course, since the -??e may have been influenced by the endings of Postumiae…magistrae in the previous line, but it could also be evidence that someone in the Middle Ages compared his text of Gellius with a text of Catullus or vice versa (there are no extant MSS of Book 6 of Gellius made before the twelfth century).

Another difficulty of the secondary tradition is uncertainty over whether an author has cited from a text or simply from memory. For example, when Pliny the Younger cites Catullus 16.5–8 at Ep. 4.14.5, he reads et instead of ac in both 7 and 8, which could be a fault of his memory, a fault of his copy of Catullus, or a corruption in the MSS of Pliny. Similar considerations apply when Quintilian (Inst. 9.3.16) quotes Catullus 62.45 with innupta instead of intacta. Responsible editors resolve such cases not by blindly following the primary or secondary tradition in preference to the other but by judging which reading is likely to be right on criteria of sense, style, and content, though a decision might not always be easy.

This is only the beginning of the challenges presented by the secondary tradition. It is unexciting but comforting when primary and secondary sources agree. Disagreements are more interesting, though not of course the ones where the secondary tradition is easily dismissed as erroneous (even if it is not possible to identify whether the erroneous reading is what the ancient writer found in his text or a corruption introduced in the transmission of his work). Some places where a secondary source offers a clearly erroneous reading include: Gellius’ discussion (NA 7.16) of Catullus 92, where the MSS have apparently corrupted dispeream (92.2) to dis spereat (sic) and sunt mea (92.3) to sin ea (as it happens, however, while one branch of the Catullan tradition has omitted nearly the whole of 92.2–4, the other, like the MSS of Gellius, reads ea, and mea is a conjecture of Isaac Vossius); Macrobius’ citation of Catullus 64.327 and 171–2 (Sat. 6.1.41–2) with the corruptions ducenti subtemine [vel sim.] for ducentes subtegmina and non for ne respectively; and Priscian (Inst. 5.77, 7.22), where the MSS read Celtiberosae [Celtiberiae] in Catullus 37.18 instead of cuniculosae. Of course the secondary tradition is of real value when it offers a clearly correct reading where the primary tradition is in error, even if the corruption is one easily healed. Whether or not Gellius was right about ebria acina, he correctly cited 27.2 with inger as its first word, while the Catullan archetype gave the morphologically normal but metrically impossible form ingere. The MSS of Pliny the Younger, which wrongly read et for ac in 16.7–8, correctly give sunt in 8, where the MSS of Catullus give sint, and Quintilian, though he mistakenly quotes Catullus 62.45 with innupta instead of intacta, correctly reads dum [cara], corrupted in the archetype to tum. Unfortunately, not all cases are decided so easily. Priscian (Inst. 1.22) quotes 2b.3 in the form quod [sc. malum] zonam soluit diu ligatam ([the apple] that undid the sash long tied), while the authoritative MSS read negatam (the sash long denied); most editors, but not all, trust the MSS of Priscian over those of Catullus. In 97.6 modern editions of Catullus print the Celtic word for carriage-body that he uses as ploxenum, the form apparently used by Quintilian when he noted (at Inst. 1.5.8) that Catullus found the word in the region of Padua. Festus (p. 260 Lindsay DVS) also cites the line, but gives the word as ploxinum. Since the principal MSS of Catullus have the corruptions ploxnio (O) and ploxonio (GR), it may be that Festus, not Quintilian, was right about Catullus’ spelling.

The secondary tradition is particularly awkward when it attests something not found in the primary tradition. For example, Catullus 64.23–4 was long printed in the form given by the Veronensis:

heroes, saluete, deum genus! o bona mater!

uos ego saepe, meo uos carmine compellabo.

Then the publication in the early nineteenth century of the so-called Verona scholia to Vergil’s Aeneid yielded the following defective quotation in a note on Aen. 5.80: Catullus: Saluete deum gens o bona matrum progenies saluete iter.… While editors do not follow Madvig and the scholion in reading gens rather than genus, they do accept the remainder of the citation as genuine, and progenies saluete iter<um> is now printed as a defective line, numbered 23b to avoid disturbing the traditional numeration. (This case also reveals the presence of interpolation [a kind of mistaken correction] in the Catullan archetype, since it can be presumed that someone consciously altered matrum to mater once progenies was no longer present to justify the genitive grammatically.)

Many editors trust the secondary tradition against the primary tradition at 1.2. According to Servius (author of a fifth-century commentary on the works of Vergil) in a note on Aen. 12.587, Catullus treated pumex as a feminine rather than masculine noun, but our Catullan corpus contains no example of feminine pumex, though it does offer one of apparently masculine pumex in 1.2 (arido…pumice). Editors who emend to arida do so despite the fact that, elsewhere in the secondary tradition, this line is quoted no fewer than six times, always with arido (except in a single MS of Isidore of Seville’s Etymologiae), and despite the possibility that Servius refers to a lost line (or was simply mistaken).

In other cases where a secondary source mentions a grammatical form, a line, or a whole poem absent from the MSS, editors often include these among the fragments of Catullus. Earlier editors were more generous in this regard than recent ones; in 1889 Ellis printed 13 fragments and Postgate 10, but Mynors (1958) gives 5, and Thomson (1997) only 3.

Mynors’s frr. 1–3 are the three printed by Thomson. Fr. 1 comprises four lines addressed to Priapus in the meter called priapean, quoted from Catullus by the metrical writer Terentianus Maurus (the first is cited by six other ancient authorities as well, and Terentianus comments that we know that Catullus wrote more such lines this way). Fr. 2, de me<r>o ligurrire libido est, is another line apparently quoted from a priapeum, this time by the grammarian Nonius Marcellus exemplifying the rare verb ligurrire, a synonym of degustare. Fr. 3, at non effugies meos iambos, is a hendecasyllabic line quoted by Pomponius Porphyrio, a commentator on Horace, in a note on Carm. 1.16.3. If all these quotations are genuine, they imply that as many as three poems of Catullus (conceivably even an entire collection of Priapea) have been lost from the corpus that we have (or perhaps were never there, whether by accident or by deliberate omission).

The two fragments printed by Mynors but not by Thomson are really "testimonia," i.e., statements about Catullus’ poetry rather than quotations from it. Fr. 4 is the remark by Pliny the Elder (HN 28.19) to the effect that Catullus wrote an erotic imitation of spells such as is found in Theocritus and Vergil. This has been dismissed (cf. Gaisser 1993: 278 n. 19), but Pliny knew Catullus far too well for us to assume an erroneous reference: since he twice appeals to poems of Catullus as evidence for historical persons, and begins his encyclopedia not only quoting Catullus 1.3–4 but explicitly softening the meter in line with contemporary practice, he surely deserves to be taken seriously as a witness to the existence of a Pharmaceutria, presumably in hexameters like Pliny’s other examples, Theocritus Id. 2 and Vergil Ecl. 8. Fr. 5 is the statement by Servius (on Vergil G. 2.95) to the effect that Catullus criticized the Rhaetian grape as useless for any purpose and wondered why Cato had praised it; he even claims that Vergil wrote G. 2.95 in awareness of both Cato’s praise and Catullus’ criticism. This too has been dismissed, but caution is suggested by Martial 14.100, an epigram on a panaca (presumably a vessel for storing wine) that connects Catullus with Rhaetian wine. These testimonia give us two more poems of Catullus absent from our corpus, one of them perhaps substantial (Theoc. 2 and Ecl. 8 comprise 166 and 109 lines respectively).

As to the citations or testimonia ignored by recent editors, Servius asserts in a note on Aen. 4.409 that Catullus treated cauere as a third- rather than second-conjugation verb; many scholars (including Mynors) have thought that this reflects the scansion of the imperative cave with a short -e at 50.18 (or possibly 61.145), but such commonplace shortenings do not necessarily prove that Catullus conjugated the verb cauo, cauis, cauit, etc., as Servius suggests. Servius also asserts (on Aen. 5.610) that Catullus used arcus as a feminine noun and (on Aen. 7.378) that he used the neuter noun turben instead of its masculine equivalent turbo; but is it not illogical to dismiss these while accepting the claim about feminine pumex simply because pumex does occur in our text of Catullus while arcus and turben do not? (Masculine turbo is certainly found at 64.107; it is impossible to say with certainty whether turbine in 64.149 and 314 and 68.63 is from turbo or turben, but 64.107 favors the former.) A defective passage of the metrical writer Caesius Bassus cites Catullus in anacreonteo (GLK 6.262.19); the Catullan archetype has transmitted no poem in anacreontics, but Bassus is in general a credible source. If these further references are reliable, they could bring the total of lost poems to nine.

Only one alleged citation can be dismissed with complete confidence, the phrase et Lario imminens Comum, supposedly cited from Catullus in Vibius Sequester’s compilation of bodies of water named in literature; it occurs, however, only as an addition to a fifteenth-century MS of the work (British Library Add. 16,986) and is therefore likely to be a forgery based on Catullus 35.3–4, Noui…/ Comi moenia Lariumque litus.

Two final examples will further illustrate the degree of uncertainty that the secondary tradition can entail. The Veronensis gave Catullus 64.65 as non tereti strophio lactentis uincta papillas. Isidore of Seville, in defining strophium, says "de quo ait Cinna strofio lactantes cincta papillas " (Etym. 19.33.3). This is normally assumed to be an erroneous citation of Catullus under Cinna’s name, though some scholars think that both deviations from our text of Catullus (lactantes for lactentis, cincta for uincta) are in fact what Catullus wrote (cf. Manzo 1967: 155). On the other hand, given the personal association and stylistic affinities of Catullus and Cinna, as well as the tendency of Roman poets to imitate each other very closely, there is at least a chance that Isidore is citing a line of Cinna that Catullus imitated – or one in which Cinna imitated Catullus. Similar considerations apply in the case of Cinna fr. 2, attributed to Cinna by Isidore (Etym. 19.2.9) in the form lucida confulgent alti carchesia mali but usually edited with cum fulgent, as quoted in a scholion to Lucan 5.418 (which, by the way, reads summi for alti). The same or a similar line is quoted by Nonius Marcellus (p. 546 Lindsay DCD) from Catullus Veronensis in the defective form lucida qua splendet carchesia mali. Again it is uncertain whether misquotation or imitation is involved; but it was on Nonius’ authority that, well into the nineteenth century, lucida qua splendent summi carchesia mali sometimes appeared in editions of Catullus as 64.235b. Of course, simple misattribution is a possible explanation.

In addition to being witnesses to the text of Catullus, for good or for ill, ancient secondary sources also serve as witnesses to the arrangement of Catullus’ poetry as it circulated in Antiquity; indeed, they are virtually the only witnesses to this, since (as will be argued below) only a single ancient title survived into the Middle Ages, in a single authoritative MS and in corrupted form. How Catullus himself arranged his works has been discussed extensively in recent years, with scholars divided between what might be called the one-roll theory – that Catullus designed all the extant poetry to stand as a single unified collection – and the three-roll theory – that our corpus represents a combination of three different ancient rolls that respectively contained poems 1–60, 61–8, and 69–113. The ancient citations, however, suggest something different.

Unfortunately for our purposes, standard practice in the literary tradition and often in the grammatical tradition as well was to cite by author’s name alone, with works rarely specified; nevertheless, some clues remain. We can come within a century or so of Catullus’ own lifetime with the Controversiae of the Elder Seneca (ca. 50 BC – ca. AD 40), who says at 7.4(19).7 that Catullus in Hendecasyllabis called his friend Calvus salaputium disertum (= c. 53.5). This surely implies knowledge of a collection called Catulli Hendecasyllabi: when something more specific than an author’s name appears with a citation, it identifies a work, not a meter. (On the other hand, when Priscian notes [Inst. 1.22] that Catullus inter hendecasyllabos Phalaecios posuit the line quod zonam soluit diu ligatam [c. 2b.13], he is referring simply to the meter, since the scansion is his evidence for the treatment of soluit as a trisyllable; perhaps the presence of Phalaecios is the clue that shows that the hendecasyllabic meter is meant and not a collection of poems.) Some decades later, Quintilian (Inst. 1.5.20) refers to the poem on Arrius (c. 84) as a nobile epigramma; it is an epigram, of course, and the existence of the collected Catulli Epigrammata is not necessarily implied, though of course it is not excluded. But when he cites c. 62.45 from Catullus in Epithalamio (Inst. 9.3.16), this surely is evidence for the independent circulation of Catullus 62 under that title, since there is no parallel for citing a poem within a collection by a title; moreover, the ninth-century copy of this poem discussed below confirms the title by calling it Epithalamium Catulli. Later still, Aulus Gellius refers to c. 92, which we would call an epigram, as a carmen (NA 7.16); this must surely be a generic reference of some sort, since it would be otiose to observe that Catullus had written something in a poem. The only possible title in the grammatical tradition occurs in Charisius, who quotes Asinius Pollio Against Valerius, Book I as saying that the masculine noun pugillares is the correct form of this word for writing-tablets, then adds "but Catullus in Hendecasyllabis quite often uses the neuter pugillaria" (GLK 1.97.12–13; in fact it occurs only once, at 42.5, and it is not clear whether this is further evidence of lost poems). If Charisius’ statement derives (as some believe) from an invective against Catullus by Asinius Pollio, it is, like Seneca’s reference to c. 53, a near-contemporary witness to the collected Hendecasyllabi. (I do not regard fr. 3, at non effugies meos iambos, as evidence for a collection of Iambi; the line, which is a hendecasyllabic, refers rather to the poet’s hendecasyllabics, which were regarded as a form of iambic verse.)

Ancient authorities, then, acknowledge cc. 42 and 53 as part of the collected Hendecasyllabi and c. 62 as an autonomous Epithalamium. A collection of Epigrammata seems possible, though far from certain, and the same could be said of a collection of Carmina containing c. 92. Fortunately, the fragments of Catullus’ friends Calvus and Cinna (collected in Courtney 1993: 201–24) provide valuable comparanda for interpreting this meager evidence; they are cited by title more frequently, and with little ambiguity; and, like Catullus, they too wrote short poems in hendecasyllabics and other meters, longer poems like epyllia or epithalamia, mostly in hexameters, and epigrams.

Nine ancient citations of or references to Calvus’ poetry specify a source. Gellius (NA 9.12.10) quotes one hendecasyllabic as representing the usage of C. Caluus in poematis (fr. 2). (Asconius cites fr. 1 as hendecasyllabus Calui elegans, but this does not constitute evidence for Hendecasyllabi as a book title.) A fragment which must come from an epithalamium in glyconics and pherecrateans comparable to c. 61 (Courtney 1993: 203) is cited by Charisius from Licinius Caluus in poemate (fr. 4). Priscian cites a portion of a hexameter from Caluus in epithalamio (fr. 5). Two authorities (pseudo-Probus and Servius Danielis [a term designating material added to certain MSS of Servius’ commentary on Vergil from a second ancient source, generally thought to be another ancient Vergilian commentator named Aelius Donatus]) cite hexameters from Caluus in Io (frr. 9–10, 12–13). A partial dactylic hexameter plus a complete pentameter are cited by Charisius from Caluus in carminibus (fr. 15); Nonius Marcellus cites the partial hexameter in a slightly different form without naming a work. Finally, a scholion on Juvenal 9.133 cites an epigramma of Martial against Pompey, but a partial citation of the poem in Seneca the Elder (Controv. 7.4.7) allows it to be attributed correctly to Calvus (fr. 18). For Calvus, then, we have collected Poemata that include one in hendecasyllabics and one in glyconics and pherecrateans; an Epithalamium in hexameters; the epyllion Io; a collection of Carmina containing elegiac couplets; and a possible collection of Epigrammata.

In the case of Cinna, there are eight such references. In Charisius we have four hexameters cited from Cinna in Propemptico Pollionis (fr. 1) and a reference to Hyginus’ commentary in Cinnae Propemptico (fr. 4). Servius Danielis cites two hexameters from Cinna in Smyrna (fr. 6); Priscian cites the same source for masculine aluus (fr. 7), Charisius for the genitive form tabis (fr. 8). Gellius cites two hendecasyllabics to show that the word nani can be found in poematis Helui Cinnae (fr. 9), and cites a choliambic from Cinna in poematis (fr. 10). Finally, Nonius quotes the end of a dactylic hexameter from Cinna in epigrammatis (fr. 12). For Cinna, then, we have collected Poemata that include one poem in hendecasyllabics and one in limping iambics; the Propempticon for Pollio; the epyllion Smyrna; and a collection of Epigrammata.

Thus, apart from the Carmina attested for Calvus alone, the output of both poets follows the same pattern: a collection of hendecasyllabics and other lyric verses called Poemata, long poems in hexameters called by individual titles (epyllia, epithalamia, a propempticon), and a collection of epigrams. And it is impossible not to be struck by how closely this pattern resembles what we find within our Catullan corpus: a grouping of hendecasyllabics, choliambics, and other meters; a series of long poems, mostly in hexameters, including both an epithalamium and an epyllion; and poems in elegiac couplets, first long ones, then epigrams. There is one clear discrepancy, of course: Catullus’ lyrics are called his Hendecasyllabi, while those of Calvus and Cinna are called Poemata, but the difference is unlikely to be important. Discussing his own light poetry in emulation of Catullus, Pliny the Younger reveals the diversity of titulature possible for such collections: "The one thing that seems to need stating in advance is that I’m thinking of inscribing these trifles of mine Hendecasyllabi, a title limited by meter alone. So, if you prefer to call them Epigrammata or Idyllia or Eclogae or, as many do, Poematia, you may call them that; I offer only Hendecasyllabi " (Ep. 4.14.8). The Greek diminutive poematia is a close enough equivalent of poemata to show that the Hendecasyllabi of Catullus could have been called his Poemat(i)a, just as the Poemata of Calvus and Cinna could presumably have been called their Hendecasyllabi.

However, while the titles of the longer poems (which need titles to distinguish them) are certainly authorial, titles like Hendecasyllabi/Poemata or Carmina need not be. Indeed, I would be reluctant to attribute them to the authors rather than to booksellers, or perhaps to grammatici who prepared editions that were then offered for sale by librarii. There are no explicit aesthetic signs of unity or structure in Catullus’ collections at least (which is not to say, of course, that scholars have not attempted to find them – see Skinner, chapter 3 in this volume), and Catullus may well have died before he could even have contemplated a definitive compilation of his lyrics or elegiacs. Moreover, the purely generic nature of the titles Hendecasyllabi and Carmina suggests an origin in commerce or in literary classification – the intervention of an editor or a bookseller, in other words. If that is the case, then it is impossible to feel certain that standard collections existed, or that every copy of Poemata or Carmina contained the same selection of poemata or carmina, or in the same order.

On the basis of the ancient citations of Catullus and this comparison with Calvus and Cinna, I would suggest that Roman readers of the first century AD knew at least five or six separate works of Catullus and/or collections of his verse, each originally occupying its own papyrus roll (perhaps even more if the Pharmaceutria existed and if there was a separate collection of Priapea). One of these is the Hendecasyllabi, presumably coinciding more or less with what are now called the polymetrics (certainly as far as c. 53), though I would conjecture that it comprised 1–61 rather than 1–60: the work that Calvus composed in the same meter as 61 circulated among his Poemata, not on its own (Jocelyn 1999 makes the same suggestion, without reference to Calvus). The three long works that followed in the Veronensis – the Epithalamium (62), the Attis (63), and the epyllion often identified now as the Wedding of Peleus and Thetis but perhaps called the Ariadne in Antiquity, to judge by the analogy with the Io and the Smyrna (64) – would each have occupied its own roll. (Some scholars, such as Clausen 1976, hypothesize a libellus containing 61–4; it may well be that some papyrus rolls in Antiquity contained 61–4 [or rather 62–4], but there is no reason to think that Catullus published them together, or that whoever compiled the ancestor of the Veronensis found them already united in a single roll or volume.) The only substantial uncertainty is whether the remaining poems represent a collection of Carmina in elegiac couplets (65–8) and a collection of Epigrammata (69–116) or simply a single collection of Carmina (65–116). The latter is more likely if we can argue by analogy with Charisius’ Licinius Caluus in poemate, which evidently means "in one of the Poemata," that Gellius’ phrase in Catulli carmine (i.e., c. 92) means "in one of the Carmina"; but if these collections were indeed created by grammatici or librarii, then perhaps both formats coexisted in Antiquity.

Modern scholars have largely neglected these ancient citations (the exception is Giardina 1974, which E. A. Schmidt 1979 tries to refute) while pursuing instead the