Catullus: Poem 64 1107018595, 9781107018594, 9781139088213

Citation preview

C AMBRIDGE CLASSIC AL TEXTS AND COMMENTARIES editors J. DIGGLE  M. HATZIMICHALI  G. KELLY  S. P. OAKLEY J. G . F. P OWELL  L. P R AUS C ELLO   M. D. RE E V E  D . N. SE D L E Y   R . J . TA R R A N T   A . J . W OODMAN

66

CATULLUS POEM 64

C ATU L LU S POEM 64 EDITED WITH AN INTRODUCTION, COMMENTAR Y, AND EPILOGUE BY

GAI L TRI M BL E University of Oxford

Shaftesbury Road, Cambridge CB2 8EA, United Kingdom One Liberty Plaza, 20th Floor, New York, NY 10006, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 103 Penang Road, #05-06/07, Visioncrest Commercial, Singapore 238467 Cambridge University Press is part of Cambridge University Press & Assessment, a department of the University of Cambridge. We share the University’s mission to contribute to society through the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107018594 doi : 10.1017/9781139088213 © Cambridge University Press & Assessment 2025 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press & Assessment. When citing this work, please include a reference to the doi 10.1017/9781139088213 First published 2025 Printed in the United Kingdom by CPI Group Ltd, Croydon CR0 4YY 2025 A catalogue record for this publication is available from the British Library A Cataloging-in-Publication data record for this book is available from the Library of Congress ISBN 978-1-107-01859-4 Hardback Cambridge University Press & Assessment has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

TO CLARE ELTIS AND RUTH IREDALE

C ONT E NT S

Acknowledgements page ix INTRODUCTION  1 Contexts for Catullus 64 (a) Catullus (b) The Neoterics and ‘Epyllion’ (c) Lucretius (d) Date and Political Context (e) Cultural Context (f) Visual Art 2 Aspects of Catullus 64 in Context (a) Greek Model(s)? (b) Myths (c) Forms (d) Style 3 Reception of Catullus 64 (a) Ancient Literature (b) Modern Scholarship 4 Text of Catullus 64 (a) Publication and Transmission (b) Manuscripts (c) Edition and Apparatus 5 Note on the Commentary

1 1 1 5 10 12 16 18 21 21 25 37 53 67 68 88 94 94 97 102 109

SIGLA 

112

TEXT AND CRITIC AL APPARATUS 

113

COMMENTAR Y 

133

EPILOGUE 

731 vii

CONT ENTS

Bibliography 1 Abbreviations 2 Editions of and Commentaries on Catullus 3 Other Sources of Conjectures Included in the Apparatus 4 Editions of and Commentaries on Other Texts 5 Works Cited by Author and Date Indexes 1 Latin Words 2 Greek Words 3 Subjects 4 Passages

760 763 766 772 781 831 831 835 836 868

viii

ACK NOWL ED G EMENTS

This commentary began in a very traditional way. In my final year as an undergraduate I was earnestly discussing with my tutor, Stephen Harrison, what I might take on for a doctoral thesis, and I mentioned (after a while) that my favourite poem was ­Catullus 64. ‘Needs an edition,’ he said. I started work on the commentary in the spring of 2006, as a Master’s student. I continued it for four years as a doctoral student (for two of which I was also an increasingly full-time college lecturer), for one year as a Junior Research Fellow, and then as a Tutorial Fellow and Associate Professor for a further twelve years (for two of which I was on research leave, and for two, on maternity leave). For support during those years I am extremely grateful to Corpus Christi College, Oxford, to Trinity College, Cambridge, and to Trinity College, Oxford – especially its successive Senior Tutors, Valerie Worth and Rebecca Bullard – to the Faculties of Classics at Cambridge and particularly at Oxford, and to my colleagues and students. I am also grateful to the Arts and Humanities Research Council, which funded my graduate studies and later awarded me a Research Fellowship (grant number AH/K008145/1) to pursue the project. And I am grateful to the family who have lived and borne with me: my parents Mary and Michael, my brother Hugh, my husband Tom, and my children Roger, Jonathan, and Felicity. I owe a great deal to the many colleagues who read and commented on sections of my work, gave advice, or discussed ideas with me. Philip Hardie and Stephen Heyworth, with extraordinary generosity, kept on supervising my work for much longer than they were formally my doctoral supervisors, and read drafts of very nearly everything: Steve, in particular, on the spot in Oxford, has always had my back. Stephen Hinds and Gregory Hutchinson, who examined my doctoral thesis on the first half ix

ACKNOWLEDGEM ENT S

of the poem, left me with much to think about, and the late E. J. Kenney, as one of the readers of the dissertation with which I applied for the Fellowship at Trinity, Cambridge, gave me unforgettable encouragement. Orange Series editors past and present have given me the benefit of their expertise and patience: the late Neil Hopkinson, Gavin Kelly, Stephen Oakley, Jonathan Powell, Michael Reeve, Richard Tarrant, and Tony Woodman. Others who have provided me with feedback and suggestions on a less formal basis include David Butterfield, Gian Biagio Conte, Lyndsay Coo, Jaś Elsner, Stephen Harrison (again), the late Nicholas Horsfall, Boris Kayachev, Dániel Kiss, Maxine Lewis, Fiachra Mac Góráin, David McKie, Llewelyn Morgan, Glenn Most, and Damien Nelis, as well as the participants in commentary workshops at Georgetown, Cornell, and Oxford, and the audiences of seminars and conferences at Cambridge, St Andrews, VU Amsterdam, Emory, Oxford, Bristol, and Royal Holloway. I would also like to thank everyone who helped with my visits to Paris, Rome, and a temporary Bodleian reading room to consult the three primary manuscripts. Abigail Buglass and Basil Nelis checked references and saved me from many other errors; Jane Burkowski was a wonderfully thorough and sympathetic copyeditor. Michael Sharp never gave up on me. I have also felt particularly supported by two fellow Cambridge University Press commentators, Monica Gale and Oliver Thomas. All of which is to say that as with Jarndyce and Jarndyce in Bleak House, children have been born into the cause, young people have married into it, old people have died out of it. I am completely responsible for all of the book’s remaining defects. I have carried on with it, at a pace I was not fully in control of, and with many halts, while also trying to do my best with many other things: teaching, professional duties, other research projects; house moves, pregnancies, lockdowns – and dozens of weddings, including my own. For helping me to keep going, I would especially like to thank Matt and Taz Hosty, Lucy Matheson, and Katherine Wodehouse. None of this would have been imaginable without Tom West. I dedicate the book, however, to two women I have known for even longer than I’ve known this poem. Clare Eltis read Catullus x

ACK NOWLEDGEM ENTS

with me, suggested what it might be like to be an Oxford classicist, and ran extra classes on the parts of this poem that weren’t on the A-level syllabus. Ruth Iredale taught me a great deal of Latin grammar, read Homer and Virgil with me, and can speak Latin hexameters more memorably than anyone I ever heard win a declamation prize. G. C. T. Oxford 29 September 2023

xi

I NT RODUC T I ON

This Introduction is intended to be truly introductory, highlighting some of the issues surrounding Catullus 64 that it is useful for a reader to have considered before detailed engagement with the text. Accordingly, it explores ways in which the poem can be situated within some of its most important contexts (Section 1), and discusses some key overarching aspects of the poem with particular reference to its literary predecessors ­(Section 2). The Introduction then gives a brief general guide to the reception of Catullus 64 in ancient literature and in modern scholarship (Section 3), before describing its transmission from antiquity to the Renaissance, and explaining the methodology behind the text and apparatus presented in this edition (Section 4). A final section provides a user’s guide to the lemmatised commentary (Section 5). For an essay offering an interpretation of the poem as a whole, the reader should turn to the Epilogue, which is intended to model by its position at the end of the book the way in which its ideas emerge from a thorough consideration of the entire text: it is in effect a conclusion to the lines of interpretation that are advanced throughout the commentary.

1  CONTEXTS FOR C ATULLUS 64 (a) Catullus The poem known to modern readers as Catullus’ 64th is his longest single work, and makes up (by word count) almost one fifth of his transmitted corpus.1 It is important to understand it as a poem by this author and as part of this oeuvre, but s­ everal 1

On this corpus and its transmission see 4a below.

1

INT RODUCTION

factors make situating it within this most basic context more problematic than is the case for many works of classical literature. Firstly, there is nothing in the poem itself that gives firm evidence for identifying it as an ‘early’ or ‘late’ work of Catullus, whether we accept the traditional dates for his birth and death (84–54 bc ) and for his active work as a poet, or explore the possibility that he might have kept writing into the 40s (see 1d below). If it is comparable to the Smyrna of his friend Cinna (see 1b below), perhaps it took years to write, as Smyrna, according to poem 95, took nine; different parts of it might have been written or revised at the same time as various other poems were being written, and as various other events in Catullus’ life and the wider world were taking place. Secondly, while there are good reasons – not least its length – for believing that 64 might originally have been ‘published’ alone (see 4a below), we cannot say, even if it was not first circulated in a book-roll along with other poems of Catullus, whether or not there were particular poems that he would have wanted it to be read alongside. Thirdly, there is the question of the ‘voice’ in which 64 is uttered. From his oeuvre as a whole, Catullus comes into view strongly both as a poet (improvising verses on tablets, explaining his reasons for translating Callimachus, presenting a p ­ olished libellus of his work) and as a person (pursuing relationships with lovers and friends, meeting people on the streets of Rome, grieving for his brother); the ‘Catullus’ who emerges in these roles may be a mask, a persona (or a set of personas, varying from poem to poem), but he is called ‘Catullus’ (the name appears a total of twenty-five times), and as a young man in the Rome (and Italy) of the mid first century bc he fits everything we know from external sources that allows us to situate Catullus’ life and work.2 In 64, however, the narrator, as typically in epic, presents himself for the most part only as a poet, telling and commenting on his tale(s); only in his comments about his own time in the poem’s epilogue (382–408) does he explicitly suggest that he

2

See 1d below. For ancient testimonia to Catullus and his poetry see Wiseman (1985) 246–62, Luque Moreno (2020) 79–106.

2

1  CONTEXT S FOR C ATULLUS 64

might also be a person with an existence outside the text. He does not name himself. To read 64 as a poem by Catullus in the context of his other works, we as modern readers must therefore simply observe what we can about its connections to other poems in the corpus, whether or not we might imagine Catullus writing those poems at the same time or presenting them to his first readers ‘together’ in any sense. Firstly, more technical observations may be made which draw on the fact that we can read this poem together with other works by the same person: we can meaningfully comment on Catullus’ diction and versification, for instance, and how they compare to his choices in other poems (see 2d below). Then, we may interpret the more general connections of both form and content that we can see between 64 and some groups of Catullus’ other poems in particular. The most obvious such group consists of Catullus’ other ‘long poems’ or carmina maiora. As well as its relative scale,3 64 shares at least some of the following features with some or all of poems 61–3 and 65–8: dactylic metre, elevated style, mythical content, narrative, the mimesis of speech or song, the use of complex structural principles (especially ring-composition), and the theme of a wedding (or of some kind of failed or negated wedding). An awareness of these wider links has an influence on the ways in which I suggest interpretations of particular similarities between 64 and the other long poems throughout the commentary.4 A second relevant group of poems within Catullus’ corpus – not, in contrast to the long poems, a group with fixed boundaries – is that of Catullus’ ‘personal’ poems (long or short): that is, those in which ‘Catullus’ seems to express his own feelings. Poem 64 contains, in the voices of both the narrator and Ariadne, m ­ oments of emotional Although at 409 hexameters (in my text) it is still a much more substantial work than Catullus’ next longest poems, 61 (228 surviving lyric lines) and 68 (a total of 160 lines in elegiac couplets; for those who prefer to separate 68a from 68b more completely than I would, the latter still holds this position at perhaps originally 120 lines). 4 This commentary naturally privileges the interpretation of 64 itself in the light of these connections. Readings of the long poems as a group in one sense or another include Wiseman (1969) 20–5, Most (1981), Trimble (2018), ­Geisthardt–Gildenhard (2019) 248–51. 3

3

INT RODUCTION

expression and moral judgement which are comparable to similar moments in personal poems of Catullus, and as I note these, I again suggest ways in which our knowledge of the ­poems’ shared authorship may invite us to interpret 64.5 However, this brings us back to the question of how we might interpret the primary voice of 64 in the context of our knowledge that we are reading a poem by Catullus. In this commentary I tend to follow the contemporary practice of using the term ‘the narrator’ to refer to the persona that emerges from the narrator-text of this poem, hymning the heroes, exclaiming over Ariadne, asserting that the gods no longer visit ‘us’,6 rather than following Lygdamus in saying that it is Catullus who does these things.7 I tend to use the name ‘Catullus’ rather with the aim of referring to the human being who wrote this poem. But the boundaries between ‘the narrator’ and ‘Catullus’ are extremely permeable. The lack of information that the narrator gives about himself means that he is not characterised as being anyone other than Catullus. He is a poet, and he knows a great deal from some kind of inherited tradition (2 dicuntur n.) about the Greek mythical past, and so it is not particularly clear whether we should say that the narrator or Catullus is the one choosing to tell the story of Ariadne in a particular order, for instance, or to allude to specific earlier texts which (we deduce from the text in front of us) Catullus certainly knew. Moreover, the majority of what we claim to know about the human being Catullus is deduced from his poetry, and it is not clear either where we might draw the line between, for instance, deducing that Catullus had read a considerable amount of Callimachus (something we tend My approach thus follows that of scholars who interpret the poems of Catullus generally as ‘resonating’ with each other (cf. e.g. Gaisser (2009)) rather than e.g. identifying ‘cycles’ or trying to put the ‘Lesbia poems’ into a chronological order. 6 Although many contemporary scholars make frequent use of the term when writing about 64, the figure of the narrator is particularly important for Schmale (2004) and, in a more complex way, for Fernandelli (2012): see further 3b below. 7 Lygd. 6.41–2 sic cecinit pro te doctus, Minoi, Catullus, | ingrati referens impia facta uiri. For recent work on how ancient readers may have understood the way in which the author ‘speaks’ in the text see esp. Whitmarsh (2013b), Grethlein (2021). 5

4

1  CONTEXT S FOR C ATULLUS 64

to attribute to the human being), and deducing that he could be passionately judgemental (something we might be inclined to attribute to a persona, ‘Catullus’ if in poem 30, ‘the narrator’ if in poem 64).8 In the commentary I aim to explore with an open mind the signals that the text might be sending us at any given moment about the ways in which the narrator and the poet Catullus may coincide or threaten to separate (for further reflections on this, see the Epilogue, Section 3). One small way in which I acknowledge their interpenetration is by referring to the narrator in the masculine throughout the commentary, as Catullus (outside poem 64) refers to himself. (b)  The Neoterics and ‘Epyllion’ Situating Catullus 64 in the context of other high-genre hexameter poetry being written at Rome in the first half of the first century bc is hampered by the lack of extant texts.9 We can observe how Catullus responds to the style of Ennius’ Annales (2d below) and, in one place, how he alludes to a particular passage (7n.), but, without more textual evidence, not whether there might have been any interaction between 64 and late Republican historical or panegyrical epic more subtle than the complete contrast between Cinna’s Smyrna and Volusius’ Annales that Catullus claims in poem 95 (cf. poem 36). The fragments of Matius, Ninius Crassus, and Naevius (not Cn. Naevius) give evidence for Latin translations of the Iliad and the Cyclic Cypria, but we cannot tell whether Catullus’ account of the wedding of Peleus and Thetis drew on Naevius’ in any sense. Cicero’s poetry offers some longer fragments whose style can be compared to that of Catullus’ hexameters (2d below), and his Aratea in particular shows us Cf. Trimble (2020) 142–5 on Catullus inside and outside his poetry. For a survey of the application of persona theory to Catullus since the 1980s, and challenges to it, see Skinner (2015) 160–5; add esp. McCarthy (2019). 9 For the Latin fragments that do survive see Courtney (20032), Hollis (2007). For Roman epic between Ennius and Lucretius, with some reference to Greek epic in the same period (including that being produced in a Roman context by, at least, Cicero’s teacher Archias), see Nethercut (2020), who casts doubt on the traditional assertion of a monolithic ‘Ennian’ tradition; cf. also Knox (2011) 199–200. 8

5

INT RODUCTION

that Hellenistic hexameter works were also being translated into Latin;10 but little can be concluded about whether other early works of his such as the Alcyones or Pontius Glaucus might have been mythological narratives with any resemblance to Catullus 64.11 The translation of Apollonius’ Argonautica by Varro Atacinus is conventionally dated to the mid 40s, but it might have preceded Catullus 64;12 the extant lines (about fifteen) offer one very probable case of imitation one way or the other (119n.), but few other clues as to whether the Argonautic element of 64 might have made use of Varro as well as Apollonius.13 It has been traditional, however, to situate Catullus’ work as a whole within that of an identifiable group of ‘new poets’ or ­‘neoterics’. Such an identification is based on three suggestive passages of Cicero,14 together with some of the names 10

11

12

13

14

The Aratea was written in Cicero’s youth (ND 2.104 admodum adulescentulo), perhaps therefore the 80s bc , and may well have been read by Catullus, although most of the verbal similarities between it and 64 collected by Luck (1976) 233–4 seem as likely to be the result of a shared hexameter tradition as of imitation strictly speaking: see 14, 125, 194, 251, 284, 383nn. and 2d below, and cf. 1c below on Lucretius. See esp. Knox (2011) for scepticism of views that the Alcyones and Pontius Glaucus ‘must’ have been Callimachean or proto-neoteric ‘epyllia’. This suggestion does not depend on dating Catullus 64 to the 40s (see 1d below). Varro’s chronology is based on Jerome’s statements that he was born in 82 bc, but Graecas litteras cum summo studio didicit only in his thirty-fifth year (Var. At. fr. 105a Hollis). If both statements are accurate, most of Varro’s poems have to be dated after 48 or 47 bc and before his death in about 35 bc (see Courtney (20032) 235–7, Hollis (2007) 177–9); but his Bellum Sequanicum – which, on the conventional understanding, must have been written without the knowledge of Graecae litterae – is usually dated not long after the campaign it commemorates, which took place in 58 bc. Perhaps some of Jerome’s details are wrong and the Argonautae belongs to the 50s too. Hollis (2007) 197 dates it soon after 40 bc on the grounds that ‘[i]t has left no trace in Virgil’s Eclogues’, not a strong argument considering the small extent of the extant fragments. For an argument placing the Argonautae in the political context of the mid 40s, see however Arcellaschi (1990) 212–19, Braund (1993), Nelis (2001) 398. For other small-scale similarities with Varro which may or may not be allusions see 6, 9, 258nn. Att. 7.2.1 (50 bc ), suggesting that Atticus might sell hunc σπονδειάζοντα to cui … τῶν νεωτέρων; Orat. 161 (46 bc ), remarking that the ecthlipsis of final s considered politius by Ennius and his contemporaries is regarded as subrusticum by poetae noui ; Tusc. 3.45 (45 bc ), after quoting some lines of ­Ennian tragedy, exclaiming that this poeta egregius is disparaged by his cantoribus ­Euphorionis.

6

1  CONTEXT S FOR C ATULLUS 64

­ entioned by Catullus himself as fellow poets and/or friends, m and ­attestations (including fragments of their work) of some of the same poets in other authors; some accounts, following in particular Clausen (1964), also include an interpretation of the available evidence about Parthenius (including his extant works) which makes him the founding father of a neoteric ‘school’ and a crucial importer to Rome both of ‘Callimacheanism’ and of the interest in transgressive mythological love stories shown in his own Erotica pathemata.15 Countering scepticism of various kinds has been expressed,16 and it seems clear that we should steer clear of describing the neoterics as a ‘school’ with firm boundaries, and especially of drawing a line between ‘new poets’ and Ennian ‘traditionalists’ as confidently as Cicero appeared to do on those particular occasions.17 Yet the evidence remains that at least two of the poets traditionally named as neoterics, Catullus’ friends Calvus and Cinna, wrote poetry in a range of metres and genres similar to Catullus’,18 and, in particular, produced poems which to some extent resembled Catullus 64. The surviving fragments and testimonia show that Cinna’s Smyrna was a hexameter poem about the incestuous Myrrha, while Calvus’ Io, also in hexameters, treated (at least) Io’s wanderings: both contained lines in which someone, very plausibly the narrator, addressed the heroine, in Cinna’s case with te opening the line (Cinna fr. 6 Courtney = 10 Hollis; cf. 253n.), in Calvus’ with the particle a (Calvus fr. 9 Courtney = 20 Hollis; cf. 71n.). Some of the surviving hexameters and elegiacs by both Calvus and Cinna suggest further stylistic elements shared with Catullus 64 (see 2d below) such as the positioning of a noun at the end of the line

15

16

17 18

See also esp. Rostagni (1932–3), Alfonsi (1945), Deichgräber (1971), Traglia (19742), Wiseman (1974) 50–6, Lyne (1978a), Johnson (2007). On Parthenius’ Erotica pathemata and its date cf. also 403–4n. See esp. Crowther (1970), Courtney (20032) 189–91, Lightfoot (1999) 50– 72. For a balanced recent assessment see Woodman (2021). For the hendecasyllables, scazons, and elegiacs of Calvus’ and Cinna’s poetry, and its erotic, invective, and literary-critical content, see Courtney (20032) 201–24, Hollis (2007) 11–86; Calvus also wrote both lyric and hexameter ‘epithalamia’. Calvus in particular is paired with Catullus in many ancient sources.

7

INT RODUCTION

in agreement with an adjective before the caesura, Greek words, and diminutives. And it is as composers of hexameters with an identifiable flavour, to which such features contribute, that such poets have the best claim to be called ‘neoterics’: when Cicero produces the line flauit ab Epiro lenissimus Onchesmites, he tells Atticus that he is welcome to sell it to one of the neoteroi.19 On this basis, I use the word ‘neoteric’ throughout the commentary to refer to contemporaries of Catullus who were writing poems which in style and content were probably comparable to his, and in particular to describe characteristics that seem likely to have been shared between 64 and similar poems by these contemporaries. The other term traditionally used for setting 64 in its contemporary literary context is ‘epyllion’. I have argued elsewhere that the usage of this word in classical scholarship since the first half of the nineteenth century has in large part been built up around attempts to explain what makes Catullus 64 so much itself, and therefore that it is easy to run into difficulties when trying to use it to characterise any poem other than, precisely, 64.20 However, the position adopted in this commentary is that if any poems other than 64 itself can fairly be described as ‘epyllia’, it is these neoteric works which were written in hexameters and had mythical content – at least insofar as we can reasonably conclude that they might have shared other features with 64 such as approximate length (long for a neoteric poem, short for an epic), purportedly narrative mode, a certain hexameter style (see above, this subsection), and elements which cut across style and content such as the subjectively sympathetic presentation of a heroine in distress implied by Calvus’ a uirgo infelix. Io and Smyrna are the most likely possibilities: other contemporary poems which may have been neoteric epyllia in this sense i­nclude

19

20

See above, n. 14. I thus follow Lyne (1978a) 167–9 in reading Cicero’s line as intended to imitate other elements of neoteric style besides the spondeiazon (see also 84n.), despite the objection of Lightfoot (1999) 55–6 that such elements can be found in Cicero’s own early poetry; I do not think that the style needs to be exclusive to the group of poets Cicero means in order to be recognisably characteristic of them. Trimble (2012). Cf. also Masciadri (2012) esp. 11–15.

8

1  CONTEXT S FOR C ATULLUS 64

Cornificius’ Glaucus, a poem on the Magna Mater by C ­ aecilius, 21 and the Diana or Dictynna of Valerius Cato. However, our scanty knowledge of any of these poems makes it very difficult to say much more about a context in this probable ‘genre’ that will help us to interpret Catullus 64 itself.22 On the other hand, we know considerably more – since these poems are either extant in full or much better attested in fragments – about certain Hellenistic short epic narratives which have often been identified as epyllia and used in order to argue that Catullus and the other neoterics wrote their epyllia in the belief that they were imitating an identifiable Hellenistic form.23 But I use the word ‘epyllion’ much more warily to describe even the most plausible Hellenistic candidates such as Callimachus’ Hecale, Moschus’ Europa, and poems of Theocritus such as 13, 18, 22, and 24. It is more productive to see certain key aspects of 64 which have often been connected to so-called Hellenistic ‘epyllia’, such as its interest in more personal episodes in the lives of mythical heroes, or its use of ecphrasis, within the broader context of Hellenistic poetry in a range of genres, and indeed of the traditions that Hellenistic poetry had itself inherited (see 2c below). In particular, it seems to me very doubtful that the ecphrasis in Europa or the crow’s tale of Erichthonius in Hecale, let alone the aetion of Heracles in Callimachus’ Victoria Berenices,24 should be treated as reliable evidence for the idea that a nested narrative structure, with one story told as a ‘digression’ from another via a narrative ‘device’, was in any sense an expected feature of Hellenistic or neoteric epyllion, rather than a much older and more widespread narrative possibility which was developed by

21

22

23 24

See esp. Lyne (1978a) and, on the possible influence exerted on Ciris by ­Cato’s poem, Lyne (1978b) 45. For the two surviving fragments of Glaucus see Courtney (20032) 225–7, Hollis (2007) 149–54; Caecilius’ and Cato’s works are completely lost. We might guess that the Magna Mater poem might have had more in common with Catullus 63 than 64; Diana/Dictynna may (also) have been a poem better described as a narrative hymn. See Trimble (2012) 60–1, 67–70 for a discussion of the problems involved in trying to argue that Catullus 64 is not a ‘typical’ neoteric epyllion because it is not only about the sufferings of a lovesick heroine (cf. esp. Cova (1949)). So e.g. Hollis (20092) 25. See 2a below on Thomas (1983).

9

INT RODUCTION

Catullus into the primary structuring p ­ rinciple of this particular poem. At most, the evidence of Virg. Ecl. 6.45–60 (the story of the ­Proetides inside the story of Pasiphae) and Ov. Met. 1.583– 750 (the story of Syrinx inside the story of Io herself) might suggest that Calvus’ Io also told one story inside another.25 But it is at least as plausible that Virgil and Ovid were commenting on neoteric poetry, or simply on the close relationship between the works of Catullus and Calvus in particular, by using the form of Catullus’ epyllion (as they also do in Georgics 4 and in many places elsewhere in the Metamorphoses : see 3a(i) below) in combination with the content of his friend’s. We have no more direct evidence that other neoteric epyllia made use of such a structure. It is difficult to know what to make of the fact that while Smyrna and Io are cited by their titles (4a below), Catullus 64 is not: ‘what one-word title could possibly have been appropriate?’, asks Most (1981) 113 n. 19, implying that an epyllion with a title such as Smyrna would have told only one story, but that has not stopped 64 acquiring a title along the lines of Peleus and Thetis in its modern reception.26 (c) Lucretius In Lucretius’ De rerum natura we have a much larger body of surviving text on the basis of which we can investigate Catullus’ relationship with one key contemporary.27 Many examples of similar wording can be observed between the two poets (for lists see esp. Munro on Lucr. 3.57, Friedrich (1908) 395–7, ­Bailey (1947) iii.1753–4, Giesecke (2000) 181–2),28 and in most 25

26 27

28

Both passages also allude to Io directly (Ecl. 6.47, 52; Met. 1.632, 714). On Calvus’ poem in the light of these and other stories with a bovine connection (including those of Moschus’ Europa and Catullus’ Ariadne) see Höschele (2012); for the idea that it might have included an ecphrasis, Lyne on Ciris 21ff. See 3b below. For Petrarch’s apparent title ‘Peplon’ see below, 4a n. 388. For Lucretius and Catullus mentioned together as important poets of their time see Ov. Tr. 2.425–30 (listing them between Ennius and other late Republican poets, beginning with Calvus), Nep. Att. 12.4 (see 1d below), Vell. 2.36.2 (with Varro – I think Varro Atacinus – in a trio of auctores carminum). See also Hutchinson (2001) 157 and n. 18, with further bibliography; add Giesecke (2000) 10–30, Biondi (2003).

10

1  CONTEXT S FOR C ATULLUS 64

(but not all)29 of these cases the correspondence is between DRN and Catullus 64. Often a similarity seems best explained simply by the fact that both poets were exploring the possibilities of writing hexameters, and doing so in a high poetic style that was influenced by earlier Latin epic but also by the language of their own time and – in quite different ways – by particular Greek traditions (for Catullus see 2d below). But any given phrase may in fact have been ‘borrowed’ by one poet from the other, and in some cases the rarity of a particular collocation, for instance, or the wider context, makes it particularly likely that we should identify deliberate imitation and, often, interpretable allusion. However, the question of who was imitating whom is a vexed one. On the conventional dating of both poets (for Catullus see 1d below), Catullus’ poems evoke a period (57–54 bc) during which Lucretius is likely to have been writing (or finishing) DRN, which may first have become available in 54. If we wish to allow for the possibility of dating Catullus 64 to the early 40s, we should bear in mind that Hutchinson has argued for a similar date for DRN too.30 The two poets might have read or heard (parts of) each other’s work in preliminary form before ‘publication’; the fact that Lucretius’ addressee Memmius is usually identified with the propraetor of Catullus’ cohors in Bithynia (1d below) makes it particularly tempting to speculate that they may have known each other through him.31 For the purposes of interpreting Catullus 64, therefore, it is historically justifiable to try out each potential allusion in either or both directions; 29

30

31

Scholars who see parallels with Lucretius as confined to 64 (as Munro b ­ elieved) or even to ‘the Ariadne episode’ (not quite Munro’s argument) tend to conclude on this basis alone that ‘Catullus must have been the borrower’ (Frank (1933) 251). For connections with Lucretius outside the story of ­Ariadne see esp. 44n., 362–70n. as well as many similarities of phrasing which may or may not be coincidental. Without citing verbal similarities in the same way, several scholars have also suggested that the satirical finale to DRN 4 responds to Catullus’ love poetry: e.g. Kenney (1970) 388–90, Gale (2007) 69. Hutchinson (2001). Courtney (20032) 171 rather posits a much earlier date, between 65 and 60 bc . See also Morgan–Taylor (2017) for a reassertion of the traditional date, with references to other responses to Hutchinson. For further hypotheses dating Catullus’ interaction with Lucretius specifically to his time in Bithynia in 57–56 bc, see Frank (1933) 251–4, Romano (1995).

11

INT RODUCTION

see also the explicitly ahistorical intertextual readings of Tamás (2016). What emerges is an ‘agonistic’ relationship between the two texts over such issues as luxury, divine intervention, and the reliability of perception:32 as Tamás notes, this is not surprising considering that the argument of DRN so overtly provokes disagreement, but it also helps to highlight some significant themes of Catullus’ poem.33 (d)  Date and Political Context Among the many references in Catullus’ corpus to contemporary people and places, it is possible to identify a few that imply specific dates.34 The earliest datable event is Catullus’ service in Bithynia (poems 31, 46) under the propraetor Memmius (poem 28; cf. poem 10), since we know of a C. Memmius who was praetor in 58 bc and would therefore have been a provincial propraetor in 57–56. The latest is often said to be Calvus’ prosecution of Vatinius (poem 53; cf. 14.3), which took place in 54. Catullus refers a few times to Caesar and Pompey, and specifically to Pompey’s second consulship (poem 113), which was in 55 bc, and to Caesar invading Britain (11.10–12, 29.4, 20; cf. 45.22), which he did twice, in both 55 and 54. But Catullus does not apparently mention the civil war between them which began in January 49. Traditionally, therefore, the dates given in Jerome’s Chronicon of 87 bc for Catullus’ birth and 58 or 57 bc for his death in his thirtieth year are corrected to 84–54, with c. 56–54 as a concentrated period of poetic production.35 However, it is possible that poem 52, with its joke about Vatinius swearing falsely by his consulship (prospective or actual?), could be dated to 47 bc, when Vatinius in fact became suffect ­consul.36 Catullus would then have no specific references to ­ political

32 33 34

35

36

So Tamás (2016) esp. 2; cf. Gale (2007) 70 ‘a degree of antagonism’. See also the comparative studies at Godwin (1995) 7–16. See Skinner (2015) 119–20, Du Quesnay–Woodman (2021) 1–2, with further bibliography. See Burgess (2002) on the Chronicon, esp. 2–4, 31–2 for Jerome’s statements about Catullus. On this controversy see Barrett (1972), Ryan (1995), Hutchinson (2003) 208.

12

1  CONTEXT S FOR C ATULLUS 64

events between 54 and 47: but any argument from silence must be dubious for such a small corpus in which poems with different concerns and in different genres interact with their political contexts in both more and less obvious ways. Broadly political readings of Catullus’ poetry in recent years have investigated the roles played by such concerns as the profits of empire, social mobility, and cultural identity;37 as noted by Skinner (2015) 120, the increased emphasis on political context in this sense seems to accompany a greater willingness to consider how Catullus’ poetry might be read as a product of the 50s and/or early 40s. Jerome aside, there is no firm evidence for the date of Catullus’ death until Nep. Att. 12.4 post Lucreti Catullique mortem, which indicates that he must have been dead by 32 bc at the latest,38 and might have died at about the same time as Lucretius; in addition, Ov. Am. 3.9.61–2, which has Catullus and Calvus meeting the dead Tibullus in the Underworld with ivy around their iuuenalia … tempora, implies that he died as a iuuenis (under the age of about 45?) and might have died at about the same time as Calvus (dead as an adulescens by 46 bc according to Cic. Brut. 279–80). But we cannot go further with any certainty,39 even if a temptingly plausible lifespan of 87–48 or 47 bc can be arrived at by keeping Jerome’s birth date of 87 and making a final guess that his statement that Catullus died in his thirtieth year might contain a corruption of xxxx to xxx.40 Other than Jerome’s date and Ovid’s evidence that he died young, there is also nothing to tell us exactly when Catullus was born or how far in advance of 56 bc he might have started writing.

37

38

39

40

For introductions and guides to recent scholarship see Skinner (2007a) 57– 128, (2015) 171–82, Damon (2021). Atticus died in 32, and Nep. Att. 19.1 states that the earlier chapters of the Life were published while he was still alive. Skinner (2003) 181–3, Wiseman (2007) 59 propose that after his brother’s death he might have married and had children. Wiseman (1985) 189–98 argues that the attested poems which have not survived (cf. 4a below) were written after the extant corpus, and also that the mimes usually attributed to a different, later Catullus should be included among them; see also Wiseman (2023) 42–6. See e.g. Wheeler (1934) 89–90, Du Quesnay–Woodman (2021) 2, Skinner (2007b) 3.

13

INT RODUCTION

Catullus 64, meanwhile, does not date itself in any straightforward way. The poem refers directly to the world outside itself only in its epilogue, and there the only statement that is made unambiguously about the narrator’s present is the very general one of 407–8; the time at which the crimes of 399–404 are said to have taken place is not so clear (397–408n.), even though some have tried to find specific references to late Republican Rome in the details (399–404nn.) Moreover, as noted in 1a above, the relative dating of the poem within Catullus’ corpus is an open question. A common suggestion has been that 64 in its finished form should be assigned to the end of Catullus’ career (combined with the traditional dating of his death and of the corpus as a whole, this gives a date of 54 bc), perhaps on the general grounds that that seems the right place for an ancient poet’s longest work and only epic.41 By contrast, arguments that use small-scale connections of theme or even vocabulary to date 64 against other individual poems do not carry conviction, because they make so much depend on so little, and in most cases are open to counterarguments that would reverse the chronology of the suggested relationships; in general they also fail to allow for the possibility that 64 might have been composed over an extended period of time.42 It is more helpful to ask instead what difference it makes to the interpretation of 64 to consider how it might have been (being) written and first read within the changing political contexts of the whole period from the early 50s to the early 40s bc. For the main body of the poem, this will mean taking into account how the mythical subject matter and the language employed to describe it might have had the potential for political meaning. So the descriptions of Peleus’ palace (see esp. 31–49n.) can be 41

42

On classical literary careers see Hardie–Moore (2010). Cf. also 1a above on the gestation of Smyrna. For example, Thomson (1997) 386 assumes that 64.154–7 ‘rework’ 60, rather than vice versa; Courtney (2000) 49 believes that the rare epithet unigena is better suited to its context at 64.300 (see n.) than at 66.53, and that 64 must therefore predate 66; Biondi (2003) 219–34 argues that Catullus could not have written the positive description of (implicitly) the Trojan War at 64.394–6 after the death of his brother at Troy (68.89–100; cf. 65.7–8), and that 64 must therefore predate 68.

14

1  CONTEXT S FOR C ATULLUS 64

read as a reflection not just of a contemporary luxury that (some of) Catullus’ readers would have recognised, but of contemporary concerns they might have shared about Roman imperial appropriation of the artefacts of the eastern empire, including Greece, and about the wider Hellenisation of Roman culture.43 The voyage of the Argo in search of the golden fleece invites a reading in similar terms.44 It is not necessary to accept in full the allegorical reading of Herrmann (1930), who proposed that Peleus and Thetis stood for Pompey and Julia, with the Fates’ prophecy of Achilles, in particular, to be read as the prediction of their hoped-for son.45 But the poem does contain significant examples of language applied to its mythical protagonists, whether their heroic activities or (especially) their personal relationships, which had or (with an early date for the poem) would soon gain particular resonance in the context of Pompeian or Caesarian politics: see esp. 137, 323, 335nn. on clementia, magnus, foedus.46 Even more strikingly, the setting of the story of Peleus and Thetis in Thessaly, and specifically at Pharsalus, may have been intended by Catullus to evoke the battle of August 48 bc, and would certainly have evoked it after the event, as is shown by the allusions to 64 made by Virgil, Ovid, and Lucan when writing about civil war.47 It seems to me important to ­emphasise that 43

44

45

46

47

See Feldherr (2007) 99–101, Dufallo (2010), (2013) 39–73, connecting this to the poem’s own appropriation of Greek mythology and literary culture. This act of ‘translation’ is the central theme of Young (2015) 24–51. See esp. Nelis (2012), who argues that Catullus’ use of the Argonautic myth should be seen in the light of its political significance for Apollonius and Callimachus in particular, and cf. 1–11n. On Herrmann’s theory see further 323–81n. and Bongi (1942), Konstan (1977) 101–3, (2007) 79–80, Nelis (2012). Herrmann compared the prophecy of the puer in Virg. Ecl. 4, while Ambühl (2016a) §2.13 suggests the prophecy of Ptolemy Philadelphus made by the unborn Apollo in Call. Hymn. 4 as a potential influence on Catullus: for the suggestion of Kuiper (1921) 226–7 that Menelaus and Helen in Theoc. 18 might be meant to reflect Ptolemy Philadelphus and Arsinoe, see also Hunter (1996) 163–4. Konstan (2007) reads 64 as contributing to a widespread concern in Catullus’ poetry with the imperialistic greed of Pompey and Caesar specifically (cf. esp. poem 29). See esp. 35–7, 37 coeunt, 44, 397–408, 397, 399, 405nn. and Newman (1990) 217–24, Nelis (2012) 21–6, Ambühl (2015) 145–77, (2016a), (2016b), and cf. 3a below.

15

INT RODUCTION

the lament for the ‘present’ state of the world at the end of the poem invites readers situated in any time and place to ask themselves whether they agree with the narrator’s judgement (see 382–408n. and the Epilogue to the commentary, Section 3), and the same can also be said of the more general questions raised in the main part of the poem about the stability of amor and the worth of traditional heroic uirtus; I would not want to limit the challenge issued by the poem to all its readers to an ‘indictment’ of Roman values applicable only to the original political context in which the poem was written.48 However, that context – and especially the remaining uncertainty as to exactly what it might have been, depending on the poem’s (final) date – ought to be one key consideration for readers trying to decide whether the explicit and implicit universal judgements made in the poem ought to be revised – or not – as political contexts change.49 (e)  Cultural Context The real-world context in which 64 originated would have extended beyond political circumstances, both narrowly and broadly understood, to the cultural assumptions (and questions about those assumptions) and social and material experience that Catullus would have brought to writing the poem, and his first readers to reading it. One of the aims of this commentary, as of others, is therefore to bring considerations about broad cultural context to bear on the interpretation of the text. However, the task is made complex by the nature of the poem as an epic set in the mythical past, in various locations in Greece, and influenced by traditional ways of presenting the mythical world both in Greek literature – which was itself composed in cultural contexts, including archaic Ionia, fifth-century Athens, and Ptolemaic Alexandria, quite distinct from late Republican Rome – and, as far as we can tell from our limited evidence, in earlier Roman literature too. It is also affected by the limitations to our knowledge

48 49

Cf. Konstan (1977). See Trimble (2013) for an argument that Virg. Ecl. 4 presents such a reading of Catullus 64, exposing its pessimism as out of date by 40 b c .

16

1  CONTEXT S FOR C ATULLUS 64

of the so-called Realien of the ancient world, since quite often the textual or archaeological sources that we look to for information about the objects, practices, and beliefs that would have been familiar to Catullus and his readers survive from points in Greek and/or Roman antiquity that are again at some distance from the Rome of the mid first century bc – although they may still be relevant to it as the best available evidence. As classicists so often must, then, I try to use the evidence that we have, to make connections where possible, but also to indicate the gaps in what we can say with confidence about the ways in which the poem presumably interacts with its cultural context in a wide range of areas. These would include physical settings, objects, and activities, including houses and furniture (e.g. 43–9, 276nn.), gardens (278–93n.), clothing (e.g. 63–5, 307–9nn.), ships and sailing (e.g. 10, 84, 225–7, 234–5nn.), dyeing (e.g. 49, 227nn.), and spinning (esp. 311–14n.). The overlapping category of cultural practices might cover, among other things, farming (38–42nn.), feasting (303–4nn.), weddings (e.g. 31, 118, 329nn.), prayer (103–4n.), and sacrifice (369–70n.). There would also have been relevant cultural assumptions about approved or expected ways of behaving or feeling: particularly important to the poem are patriotic courage and military prowess (e.g. 81–2, 102, 323–4, 339, 348), p ­ ersonal commitments and truthfulness (e.g. 58–9, 123, 132–63, 231–2, 322, 334–6), divine justice (e.g. 134, 188–201, 203, 386, 397–8, 405–6), the family (e.g. 21, 29–30, 117–20, 150, 180–1, 207–45, 298, 376–80, 399–404), romantic love (e.g. 19, 91–100, 253), and sex (e.g. 27, 145, 147, 328–32, 372–80, 401–4). One particular point is worth highlighting here: although it is beyond the scope of this commentary to go into the arguments in detail, I take the position that Catullus and his first readers would have associated a woman’s loss of virginity with blood (see esp. 49, 163, 307–9, 362–70nn.).50 50

I follow D. P. Fowler (1987), who identifies passages reflecting this association in Greek and Roman literature from the Iliad onwards. However, Rosenberg (2018) argues that the pre-Christian classical world lacked the assumption that a woman will bleed at first sex, which he believes was an ancient Jewish idea, particularly associated with Deut. 22:13–21, imported by Christians into Graeco-Roman culture; cf. 376–7n.

17

INT RODUCTION

(f)  Visual Art Visuality is a fundamental aspect of Catullus 64, discussed throughout the commentary: not only is one of the poem’s two stories framed as the ecphrasis of an artwork, but vividly ‘pictorial’ descriptions are found outside the passages which formally describe the uestis (cf. 2c(ii) below and the Epilogue to the commentary, Section 1), while seeing and the response to what is seen are repeatedly thematised. One important area of cultural background to the poem, both for Catullus and for his original readers, would therefore have been their experience of viewing visual art. Again (cf. 1e above) the Greek and Roman art that I mention in the commentary with the aim of shedding some light on that experience comes from a variety of times and places across antiquity, depending on what survives (or is attested in written sources); it also consists of various different kinds of art objects that would have been viewed in different ways, such as vases intended to be used and handled, or domestic wall-paintings that might have been examined (or ignored) by various users of a room.51 My approach is therefore both speculative, so that I often suggest that Catullus or his readers might have seen something like a surviving object, and comparative, so that I point out ways in which the content or form of extant artworks is comparable to Catullus’ text in ways that I think advance the text’s interpretation, even if we cannot be sure whether those features would in fact have formed part of the visual experience of mid first-century Romans.52 There are several points in the poem, then, where Catullus and/or his first readers may particularly have brought to the text their memory of seeing and responding to similar content in visual art. They may have remembered depictions of scenes and characters from the narrated myths themselves (cf. 2b below), especially in those passages where the text is most overtly pictorial: see esp. 12–18n. (the Nereids), 61, 251–64nn.

51

52

For the limited survivals of ancient figural textiles that might be compared to the uestis of the poem, see 51 mira … arte n. Accordingly I occasionally compare modern art as well as ancient: see 62n. on Van Gogh, 251–64n. on Titian.

18

1  CONTEXT S FOR C ATULLUS 64

(­ bacchants and related figures), 265–302n. (the ‘procession’ of divine wedding-guests), and above all 50–75n. (Ariadne). They may also have drawn on their experience of more generic artistic topics, from interiors (cf. 31–49) and gardens (cf. 278–93) to depictions of old age (303–22n.) and the broad theme of the ‘erotic’.53 However, the poem’s use of form can also be compared to artistic techniques.54 The way in which 64 invites its readers to explore an often fantastical mythical world has been likened by Dufallo to the trompe l’œil of contemporary Second Style wall-painting.55 Its presentation of several of its characters as themselves engaged in looking can be compared with the use that art can make of the gazes of depicted figures, often making the viewer consider the role of their own gaze at the artwork (cf. esp. 50–75, 241–5, 267–8nn.).56 Most significantly of all, I think, the poem’s presentation of two separate myths, and of ‘scenes’ within each of them that can often appear discrete from one another, is comparable to the effect of artworks that similarly juxtapose multiple images from mythological stories.57 The art object may encourage its viewers to combine images from the same story into a narrative, and/or to fill in gaps in that narrative;58 it may also invite them to look together at images from the same or from different stories in order to make comparisons and contrasts (whether visual or thematic) between them, and to consider how many gaps in each mythological narrative 53

54

55 56

57

58

Cf. e.g. 63–5, 129nn. on Ariadne, 91–2n. on Theseus, 362–70n. on Polyxena. For some recent approaches to the erotic in Greek and Roman art (both ‘genre’ and mythological) see esp. Vout (2013), Valladares (2021). On the importance (and comparability) of form in narratives and pictures see Grethlein (2017). Dufallo (2013) 48–58; see also 43 recessit n. See esp. Fitzgerald (1995) 140–68, Elsner (2007) 67–109, and cf. Fredrick (1995) on gendered gazes in and at the paintings in a Roman house: all these scholars particularly emphasise Ariadne. Some images of mythological characters, settings, etc. will more obviously imply a mythological narrative than others do: for the idea of a distinction between ‘narrative images’ and non-narrative or descriptive images, see Giuliani (2013) esp. 1–18. Exactly what counts as the same story or a different story might be debatable in any given case. So might what counts as the same image or scene: cf. 251–64, 251 at parte ex alia nn. for whether, on the uestis, Bacchus and his followers are part of the same image as the depicted Ariadne.

19

INT RODUCTION

need to be filled in order for such comparisons and contrasts to work.59 Ancient art had various forms which juxtaposed images in this way, including many vase styles as well as architectural forms such as friezes, altars, or entire temples: among s­ettings which would have juxtaposed what we might be tempted to think of as ‘separate’ works, we may compare Roman houses which deployed mythological paintings in combination with each other or with different kinds of painting, within rooms or within whole houses, perhaps along ‘axes’ reflecting social status, or according to other organisational schemes.60 Catullus 64 has also been connected in particular with the famous François vase, an Attic black-figure volute krater of c. 570 bc,61 which depicts the wedding of Peleus and Thetis (cf. esp. 265–302n.) alongside other scenes from the ‘same’ myth – the Calydonian boar hunt, featuring Peleus, and episodes from the adult life of Achilles, including his pursuit of Troilus (with Polyxena) and his death (cf. 323–81n.) – and from ‘different’ myths, including the battle of the Lapiths and Centaurs, and Theseus and Ariadne.62 The appearance of Theseus and Ariadne on the same vase as Peleus and Thetis is called ‘noticeable’ or ‘suggestive’ by readers of Catullus 64;63 but the early proposal that a lost archaic Greek poem (cf. 2a below) had directly influenced both the first-­ century ­Roman poem and the sixth-century Attic vase has been 59

60

61

62

63

For the ways in which the poem encourages a similar response, see the Epilogue, Sections 1 and 3, and e.g. 246–50, 251–64nn. on the different resolutions to the story of Ariadne, 323–81n. on the Fates’ presentation of ‘scenes’ from Achilles’ life. See esp. Fredrick (1995), drawing on and critiquing the fundamental work of Wallace-Hadrill (1994), and Jones (2019) esp. 197–229. Elsner (2007) 88–9 and Dufallo (2013) 53 compare Catullus 64 with the ‘megalographic’ frieze in oecus 5 of the Villa of the Mysteries in Pompeii (dated to the 50s bc and thus contemporary with Catullus), in which a depiction of Bacchus and Ariadne apparently interrupts a scene which has both Dionysiac and bridal aspects. Florence, Mus. Arch. 4209; the painter is named as Kleitias, the potter as ­Ergotimos. On the François vase see esp. Torelli (2007), Shapiro–Iozzo–­LezziHafter (2013). A labelled frieze depicts Theseus’ ship and crew, the young Athenians promised to the Minotaur, Theseus with a lyre, a nurse, and Ariadne with her ball of thread; the interpretation of the image is controversial. See e.g. von den Hoff (2013), Giuliani (2013) 255–7. Ellis (18892) 280 n. 1, Morwood (1999) 227.

20

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

superseded by a more nuanced understanding of the relationship between the vase’s iconography and archaic poetry.64 However, in their comparable (but far from identical) juxtapositions, both the François vase and Catullus 64 represent particularly sophisticated explorations of a way of treating mythical m ­ aterial which had many other outworkings in ancient visual art.65

2  ASPECT S OF C ATULLUS 64 IN CONTEXT (a)  Greek Model(s)? Before surveying some of the main ways in which the content (2b) and form (2c, 2d) of Catullus 64 can be seen to relate to the poem’s known literary predecessors, it is worth briefly examining the possibility that either content, form, or both might be explained by the theory that Catullus used one or more Greek ­poems, now lost, as a ‘model’. This might seem prima facie viable given that Catullus’ extant poetry includes translations or close adaptations of Sappho (poem 51) and Callimachus (poem 66), and that he is also attested to have written an imitation of Theocritus 2 (327n.), and since the poem does not ­explicitly identify its author as ­Catullus or its context as first-century Rome.66 For ­historical 64

65

66

See Shapiro (2013) 13–14. Soon after the vase’s discovery in 1844, Lenormant (1850) 635–41 proposed a single model, perhaps the Cypria, for both the vase-painter and Catullus. Stewart (1983) suggests that a poem of Stesichorus on the wedding of Peleus and Thetis may have been ‘Kleitias’ starting point’ (p. 67), but not that it would have contained references to all the myths depicted on the vase; he does not mention Catullus 64, but makes a more general comparison to literary technique for the ways in which the vase combines its ‘narratives’. So, in different ways, both Antonelli (1915) and Morwood (1999). One other artwork which may depict the wedding of Peleus and Thetis, may depict Ariadne and Theseus, and certainly juxtaposes two scenes (or parts of the same scene?) in an enigmatic way, is the (probably) Augustan cameo glass vessel known as the Portland vase (London, BM 1945,0927.1). It too has (therefore) been connected to Catullus 64 (including by Lenormant (1850) 636): see esp. Skalsky (1992), (2010), who argues that one of the images ‘illustrates’ 64.26–30, Haynes (19752) 19–20, who suggests that a ‘romantic version’ of the myth of Peleus and Thetis inspired both Catullus and the vase, and, for overviews of the vase’s interpretation, Harrison (1976), Painter–Whitehouse (1990a), (1990b). Cf. 1a, 1d above, and the Epilogue, Section 3. For the same question asked about poem 63 see esp. Mulroy (1976), Harder (2005).

21

INT RODUCTION

accounts of the debate in the nineteenth and early twentieth century over a possible model or models, see Schmale (2004) 18–22, Fernandelli (2012) xviii n. 6, as well as Ellis (18892) 280, Riese (1866) 504–5 for older material, and cf. 3b below. Here I will consider the various versions of this idea more schematically, by means of some key examples, in order to highlight which aspects of the poem it is that they attempt to explain, and (if possible) what sort of dependence on the model(s) they hypothesise. Some scholars have simply proposed that Catullus had a specific lost model for his decision to write a poem on the wedding of Peleus and Thetis. This idea appears early in the modern criticism of 64: Scaliger, citing the possible evidence for a Hesiodic ‘epithalamium’ for Peleus and Thetis (see 2b(i) below and n. 92), claims that Catullus was following Hesiod’s ‘example’;67 Voss, on the other hand, rejects the idea that Catullus ‘imitated’ Hesiod or Agamestor, and advances his own opinion that he ‘imitated’ Sappho, as in his other ‘epithalamia’.68 It is difficult to tell how close an ‘imitation’ such scholars have in mind. More can be said about the later arguments that Catullus depended closely on a single model, usually said to be Hellenistic, which might explain particular details of the poem: so Merkel maintains that the poem was ‘translated’ from a longer Alexandrian original,69 Riese that it is a translation of a poem by Callimachus,70 Reitzenstein that Catullus ‘follows’ a poem written by an unknown P ­ tolemaic court poet to celebrate a royal wedding.71 It is often possible to sense here some discomfort with the way in which Catullus’ poem 67 68

69

70

71

Scaliger (1600) 70 ‘epithalamium Pelei et Thetis [sic ] scripsit exemplo H ­ esiodi’. Voss (1684) 189 ‘ut alibi passim, ita quoque in hoc Epithalamiorum libello, credo imitatum esse Sapphonem’. Merkel (1837) 359–60; ‘e Graeco conversum’. He argues that the promise of line 24 (see n.) would have made better sense in a longer poem which later returned to the Argonauts. Riese (1866), particularly focusing on the imitation of Call. Hec. fr. 165 Hollis = 732 Pfeiffer at line 111 (see n.). Riese suggested other Callimachean fragments as the ‘originals’ of particular lines of 64, and also argued that the relevant lines of Callimachus’ poem had already imitated Theocritus (96n.) and were imitated by Euphorion (30n.). This and his other arguments were so thoroughly called into question by Schulze (1882) 208–24 that Riese himself later abandoned his theory of a translation (Riese (1884) 154). Reitzenstein (1900) 86–102 (‘folgt’ at 88): cf. esp. 31–49, 265–302nn.

22

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

s­upposedly about Peleus and Thetis incorporates the long ‘digression’ on Ariadne: for Merkel this is explained by the theory that he was translating from the sort of longer Alexandrian catalogue poem that typically combined such unrelated stories,72 while Riese simply says that ‘disproportion’ is typically Callimachean,73 and Reitzenstein briefly explains that his posited original would have introduced Ariadne as a contrasting narrative, again in Alexandrian fashion.74 Much later, Thomas, who ­proposes not that 64 is a translation of a lost poem, but that a specific model for its use of ecphrasis to introduce an extended narrative within an ‘epyllion’ is to be found in (his reconstruction of) Callimachus’ fragmentary Victoria Berenices, still gives the sense of wanting to excuse Catullus by attributing what is ‘curious’ about 64 to an ‘experiment’ with importing pure Alexandrianism.75 Another approach to this perceived problem holds that the poem fails as a unified whole (or should not be judged as one) not because it translates only part of a longer original, as Merkel had claimed, but because it combines adaptations of two shorter and originally separate originals. This idea was put forward by Drachmann, by Pascal, who returned to the idea that in the outer story Catullus was imitating the supposed ­Hesiodic poem on Peleus and Thetis, and most famously by Pasquali, who argued that both models were Alexandrian but Catullus combined them by means of the specifically Roman ­puluinar ;76 more recently, Lefèvre has again proposed that Catullus used two lost Alexandrian models, a ‘Peleus-Gedicht’ and a

72

73 74

75 76

Merkel suggests that Catullus’ model was ‘adornatum … Alexandrinorum modo’, and claims that works by Euphorion, Nicander, or Parthenius would have offered a farrago of stories, ‘partim artificiose nexarum, partim prope licentia longa digressionum et emboliorum ambage sibi invicem interpositarum’ – like Ovid’s Metamorphoses. Riese (1866) 508 ‘Ungleichmäßigkeit’. Reitzenstein (1900) 101–2; he also argues that the uniformity of content and style between the two parts of Catullus’ poem means that they must descend from the same original. Thomas (1983) esp. 112–13; cf. 1b above. Drachmann (1887) 74–89, Pascal (1904), Pasquali (1920); cf. esp. 47, 53, 265–302nn. All three scholars explain the surprisingly narrative character of the ecphrasis (2c below) by the supposition that the Ariadne poem that Catullus imitated was not itself ecphrastic.

23

INT RODUCTION

‘Theseus-­Gedicht’.77 Other more recent scholarship, without trying to give such a detailed account of the whole poem in such terms, has also re-emphasised the possibility of a lost Greek poem that might have influenced Catullus’ handling of either one story or the other: so Nuzzo argues that the poem’s treatment of Peleus and his Thessalian kingdom might be owed to an ‘original’ written at the court of a Macedonian ruler in Hellenistic Thessaly,78 while Knox reminds us that a Hellenistic poem on Ariadne is one plausible explanation for some of the similarities between the portrayals of the heroine in 64 and in Nonnus’ Dionysiaca.79 It seems to me helpful to bear in mind that Catullus might well have known and alluded to poems that are completely lost to us, as well as those that are now available only in fragments, and those that are extant parts of our canon;80 much less so to be tempted towards the position that a specific lost work or works might hold the key to all the interpretative challenges of the poem. The more strongly a hypothesis maintains that 64 is a close translation (rather than a freer ‘imitation’) of a poem or poems attributed to a specific time and place other than Catullus’ own, the more objections it faces from arguments pointing out either the poem’s many allusions to earlier Greek, (other) Hellenistic and Roman works, or those aspects of it that seem most ‘Roman’ or ‘Catullan’.81 Accordingly, even scholars who believe in a model or models may say that there are sections or aspects of the poem which do not derive from it or them.82 77 78 79 80

81

82

Lefèvre (2000). Nuzzo (2003) 33–5; cf. 324 Emathiae n. Knox (1998); see 3a(ii) below and n. 331 and cf. 132–201n. See 2b and 2c below, and note especially works which might have influenced the poem in both form and content – such as the third stasimon of Eur. IA, or the story of Medea in Ap. Arg. – without being ‘models’ that could explain all of Catullus’ choices. For the former see e.g. Schulze’s refutation of Riese (n. 70 above); for the latter, important early examples include Ramain (1922), Morpurgo (1927). On these two ways of undermining the translation theory see also Wheeler (1934) 149–52. So e.g. Lefèvre (2000) 196–200 sees Catullan originality in various places, including the whole of lines 1–51 and 397–408, while Pasquali (1920) 22–3 ends by praising the artistry of Catullus’ ‘imitation’, including both his allusions to works other than his main models and his treatment of ‘la vita e l’amore’.

24

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

It is also very noticeable that while those who have proposed a Greek model often write as if excuses need to be made for the poem’s ‘peculiarities’,83 those who have contended most strongly against the idea, from Haupt attacking Merkel, to Perrotta rebutting Pasquali, to Klingner advancing his epoch-making account of the unity of the poem,84 have done so in a spirit of appreciation, defending both the poem and Catullus as its author.85 Like them, as well as most scholars writing both before and after the vogue for the theory of a Greek model, I hold Catullus responsible for the poem that we have. (b) Myths In the following subsections I draw together some observations about how Catullus’ treatment of the poem’s two main myths can be seen against the background of the appearances of those myths elsewhere in classical literature and art. More on individual details and on related stories (e.g. the voyage of the Argo, the career of Achilles, the death of Polyxena, the myth of ages/ races), as well as many more incidentally mentioned parts of the mythical tradition, can be found in the commentary on the relevant lines. (i)  Peleus and Thetis The relationship between the Nereid Thetis and the mortal Peleus, son of Aeacus and participant in the Argonautic expedition and other ‘group ventures’ of the generation before the Trojan War, is apparently always treated as a formal marriage, and their wedding is a well-known event in the earliest Greek literary sources. For discussion of the myth of Peleus and Thetis see especially Gantz 220–32, RE xix.1.284–302 (Lesky), ­Roscher iii.2.1833–9 (Bloch). It combines two elements which, although not strictly incompatible, represent different mythical 83 84

85

Thomas (1983) 112. Haupt (1855) 7–13, Perrotta (1931), Klingner (1956); cf. also Wilamowitz (1924) ii.298–9 and see 3b below. Cf. Trimble (2012) 58–60.

25

INT RODUCTION

story-types and were probably originally separate:86 the prophecy that Thetis would bear a son greater than his father, which caused Zeus (and sometimes also Poseidon) to give up his own interest in her and ensure that she was married to a mortal (27, 294–7nn.),87 and the idea that Peleus had to acquire his bride by force, holding on to her as she attempted to escape by multiple metamorphoses (20n.).88 The tone in which the wedding itself is presented, however, varies significantly according to its context and the role it is playing in a narrative (if any). One strand of the tradition focuses on the glamour of the event as a gathering of the gods: this may be understood as a pictorial version, characterised by vivid representation of the setting (usually Mount Pelion, in particular Chiron’s cave: 31–49, 278–9nn.) and the guests, and associated with appearances of the wedding in visual art (265–302n., LIMC s.v. ‘Peleus’ nos 198–212; cf. 1f above and see also Jouan (1988)) and in ecphrasis (V.Fl. 1.130–9), as well as occasions on which it is used as an idealised point of comparison for a human wedding (Men. Rhet. 2.5.4 Race = 400.11–22 Spengel, [Sen.] Oct. 706–9, Stat. Silu. 1.2.215–17, Claud. Nupt. praef.).89 Such a presentation may be combined, especially in lyric, with an emphasis on Peleus himself as supremely fortunate in being given a divine wife (25n.). This is sometimes extended with a reference to the virtue by which he earned such a blessing (25, 323–6nn.). However, Peleus’ good fortune at his wedding can also be contrasted with his later bad fortune in losing his son Achilles to an early death: this is the point made by the exempla of Peleus and Cadmus at Pind. Pyth. 3.86–106, but it is also a note sounded in the Iliad (24.534–42). More generally, the Iliad sets the pattern for an important alternative way of presenting the wedding, in which it is seen as part of a longer 86 87

88

89

See esp. Reitzenstein (1900), Lesky (1956). For the variant that Zeus made Thetis marry a mortal because she had rejected him out of respect for Hera, see 376n. Here the story-type is that of a human falling in love with a creature from a different world (the sea), with the elemental incompatibility of the couple eventually leading to their separation: for comparative material see also ­Frazer (1921) i i .383–8. On the ironies that can arise from using Peleus and Thetis for this exemplary purpose see Wasdin (2018) 157–70.

26

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

story and looked back on not as an idealised event of the mythical past, but as an event of the personal past which is regretted, specifically, by Thetis herself: Il. 18.429–43 (cf. Il. 18.84–90). The important details here are Thetis’ unwillingness, the couple’s later separation, and especially the sorrow that will come to Thetis from Achilles’ death, not simply because he dies young but, given that she is a goddess, because he is mortal in the first place. This approach to the story is appropriate for tragedy, as when Thetis remembers Apollo’s wedding-song at Aesch. fr. 350 TrGF (see esp. 299–302n.), or appears to Peleus in his old age at Eur. Andr. 1231–72; it also predominates in Apollonius’ later epic version (Arg. 4.790–817, 866–80). Finally, a quite different way of treating the wedding as part of a narrative, this time a world-historical one, again focuses on the divine guests rather than on the couple: this is the story, which was told at the beginning of the Cypria, in which the event is interrupted by Eris, who provokes the contest among Hera, Athena, and Aphrodite that leads to the Judgement of Paris and the Trojan War.90 If, as seems likely, the Cypria contained a relatively lengthy narrative of the wedding, we might suspect that it also incorporated some of the other possible emphases mentioned above, but it is impossible to tell quite how this might have been handled.91 Nor can we say much about the way in which the story was told in the Hesiodic Catalogue of Women (see esp. [Hes.] frr. 207–13 M–W); if March (1987) 3–26 is even partly right in her extended reconstruction, then it will have been a fascinating example of a full narrative of the wedding that described both the origins of the event and some of its unhappy consequences, with a ‘wedding-song’ sounding the praises of Peleus apparently in the voice of human characters unaware of what will happen

90

91

See Jouan (1966) 68–87, West (2013) 69–75. Sophocles wrote an Eris (perhaps a satyr-play) which may have dramatised Eris’ action, but there are no extant sources describing that action in any detail earlier than Apollod. Epit. 3.2 (where, however, the event is not connected with the wedding), Hyg. Fab. 92. It is therefore debated in particular whether the idea that Eris threw an apple for the goddesses to argue over is early or late. The most it seems safe to say is that the divine guests were probably described: see 265–302, 278–9, 299–302nn.

27

INT RODUCTION

next.92 Much less can be guessed about the probably Hellenistic Θέτιδος Ἐπιθαλάμιος of Agamestor from the single surviving fragment.93 For comparative context for Catullus’ version of the story we may, however, also look to other poetic treatments of heroic weddings: our knowledge of the Hesiodic Wedding of Ceyx is again extremely fragmentary,94 but that couple’s story had a (or at least one) tragic ending (e.g. Ov. Met. 11.410–750), and we know of other poems on the weddings of couples who in different ways would have less than happy futures together: the wedding of Hector and Andromache was described by Sappho (fr. 44 Voigt),95 and that of Helen and Menelaus apparently by Stesichorus,96 as well as in the extant Theoc. 18, while Apollonius narrates the wedding of Jason and Medea (Arg. 4.1128–1200; see 31–49, 265–302nn.). It seems that most of these e­ xamples focused on the celebratory atmosphere of the wedding itself, again sometimes by making use of an inset wedding-song (323– 81n.), and perhaps without directly referring to the couples’ subsequent history (though in the extant texts possible allusions to that history can be scrutinised: see e.g. 372–80n.). A particularly important influence on Catullus as a way of treating the wedding of Peleus and Thetis itself, however, was probably the third stasimon of Euripides’ Iphigenia at Aulis. This ode (IA 1036–97) exploits its position as a lyric inset within a tragic context to combine a pictorial account of the wedding, and a macarismos of Peleus, with an inset prophecy of Achilles which inevitably links the wedding with the Trojan War and therefore with the ode’s context in the play, the imminent death of ­Iphigenia, and the 92

93

94 95

96

See 323–81n. on frr. 211 and 212b M–W. Tzetzes, Prol. ad Lycophr. 4.12–15 Scheer cites lines 7 and 10 of fr. 211 as coming from a Hesiodic ἐπιθαλάμιον εἰς Πηλέα καὶ Θέτιν, but this ‘epithalamium’ is now generally thought to have been part of the Catalogue of Women rather than a separate poem (as believed by e.g. Reitzenstein (1900) 78–86; cf. also 2a above). See Hirschberger ad loc. and see also 384–6n. on [Hes.] fr. 1.6–7 M–W. SH 14, four elegiac lines on the etymology of Achilles’ name transmitted by Tzetzes on Lyc. Alex. 178 (89.8–15 Scheer). See frr. 263–9 M–W with Merkelbach–West (1965), and cf. 31–49n. See also Sapph. fr. 141 Voigt, describing libations at a divine wedding, possibly that of Peleus and Thetis. See esp. schol. Theoc. 18 arg. (331.12 Wendel) ἐν αὐτῷ τινα εἴληπται ἐκ τοῦ πρώτου Στησιχόρου Ἑλένης and fr. 88 Davies–Finglass.

28

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

chorus’ own closing reflections on the disappearance of aidoˉs and areteˉ that her sacrifice represents (see further Walsh (1974) and esp. 299–302, 323–81nn.). Catullus 64 similarly makes the most of pictorial and lyric elements in its presentation of the wedding even though its own primary genre is epic narrative (2c(i) below). It too has extended descriptions of the setting and the guests (31–49, 265–322), as well as praise of Peleus for his good fortune in his marriage (25–30, 323–9, 373). However, it goes further in the picture it presents of what a ‘blessed’ marriage might be in having Peleus actually fall in love with Thetis (19n.), while Thetis herself does not ‘disdain’ or ‘refuse’ the marriage (20n.): uniquely, the wedding seems to result from the couple’s own feelings (albeit followed by divine approval: 25–30nn.), while love and concord are emphasised (appropriately for the inset genre) in the Fates’ wedding-song (328–36, 372–4). On the other hand, the poem repeatedly, though never explicitly – sometimes, indeed, in the course of stating the exact opposite – makes the reader think of details from the epic or tragic version of the myth, including the prophecy that guided Jupiter’s decision about Thetis (27, 294–7nn.), her own unwillingness to marry Peleus (20n.), and the couple’s problematic parenthood (379–80n.) and separation (12–18, 334–6nn.). Achilles’ death is unambiguously, though obliquely, present in the Fates’ song (323–81, 362nn.) and may also be evoked elsewhere (290–1, 299–302nn.). The question raised is whether, ‘in this version’, in the story implied by this short epic with its complex incorporation of different genres, the reader is supposed to conclude that some of the less happy but more typical elements of the myth ‘really did’ happen, or not (see esp. 321–2, 372–80nn.). Meanwhile, the world-historical aspect of the wedding is complicated by two further innovations of Catullus’ opening lines: the shifting of the event in time and space to make it a result of the voyage of the Argo (19–21n.), and located not in a wild space presided over by a Centaur but in a palace whose ­luxury, insofar as it is artificial, often seems ‘modern’, even Roman (31–49, 265–302nn.). Both of these changes associate the wedding with the end of the golden age (1–11, 31–49, 294–7nn.), even though the golden age itself is simultaneously evoked (see 29

INT RODUCTION

also esp. 278–93n.). While the Trojan War, the traditional end of the heroic age, is clearly presented as imminent by the Fates’ prophecy of Achilles, the arrival of Eris is not narrated, and hints towards it (see 323–81n.) are more subtle (or debatable) than those towards more traditional and less romantic views of Peleus and Thetis themselves: however, it has been argued that the reader should fill in this event either on the analogy of the poem’s inner story of Ariadne, whose ‘climax’ is not narrated either (Townend (1983); 251–64n.), or from the appearance of a dropped apple in poem 65, on the assumption that Catullus’ long poems are meant to be read in sequence and to supply gaps in each other’s ‘elliptical narrative’ (Ambühl (2014) 118– 23).97 When the poem ends with its own contentious account of a world history starkly divided between past and present (382– 408), the reader is prompted to consider the traditional mythical role of the wedding as a historical turning point, and, again, to question whether the narrator’s judgement on the wedding in this connection – that it was a clear example of the state of things in the blessed past – is the right one (see esp. 384–6n.). (ii) Ariadne Ariadne, daughter of Minos, is associated with both Theseus and Dionysus in our earliest Greek evidence: for discussion of her myth see especially Gantz 114–16, 260–70, RE suppl. x i i i . 1111–41 (Herter), Webster (1966), Brommer (1982) 86–92, and see also n. 103. Although neither partner is mentioned at Il. 18.590–2, the idea there that Daedalus built her a dancing-floor (χορός) in Cnossus invites speculation about some kind of link to the Labyrinth;98 meanwhile, at Hes. Theog. 947–9, Dionysus marries her and Zeus makes her immortal for his sake. Elsewhere in the Hesiodic corpus (and/or in the Aegimius) Theseus was apparently said to have abandoned Ariadne because of his desire for another woman, Aegle (frr. 145, 298 97

98

See below, n. 107, and cf. above, 1a n. 4, for some other approaches which read the long poems as a group. See Edwards ad loc., with further references, and for the association of Ariadne, the Labyrinth, and dance see also Theodorakopoulos (2000) 131–4.

30

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

M–W), and we might hypothesise that, as in later accounts, this abandonment preceded her marriage to Dionysus and immortalisation; on the other hand, a very different way of combining the roles played by hero and god, in the opposite order and with the opposite outcome for Ariadne, appears at Od. 11.321– 5, in which Odysseus sees her alongside other dead heroines, and states that Theseus once tried to bring her from Crete to Athens, but before he could enjoy her, Artemis killed her in Dia on the evidence (μαρτυρίῃσιν) of Dionysus. In the Cypria (arg. 4b West) the story of ‘Theseus and Ariadne’ was one of those told by Nestor to Menelaus after the abduction of Helen, and it is not clear whether – if there is any relevance to Menelaus’ situation – Theseus might be parallel to Menelaus himself or to Paris.99 It seems likely, however, that what becomes the dominant order of events in later sources emerged in an Athenian context around the late sixth or early fifth century:100 though we may speculate about what an epic Theseis might have said,101 it seems clear enough from a detailed summary preserved in schol. Od. 11.322 (i i . 505.25–506.18 Dindorf = 3 F 148 FGrH  ) that Pherecydes gave a version in which, after Ariadne fell in love with Theseus on Crete and helped him to escape the Labyrinth, he took her to Dia,102 but then abandoned her there – on the orders of Athena – after which Dionysus appeared and slept with her (μίσγεται). In the visual art of the same period, vase-paintings in which Ariadne is shown as the bride of Dionysus, and those in which she is the helper of Theseus against the Minotaur, begin to be joined by others showing Theseus leaving her behind at the instigation of a god, while Dionysus approaches to claim her.103 See Webster (1966) 23–4, Gantz 269, West ad loc. See below, this subsection and n. 111, for the Athenian national idealisation of Theseus beginning about this time. 101 Arist. Poet. 1451a19–21 shows that such a poem was in existence by the fourth century, but it is often dated to the late sixth: see also Preller–Robert i i .2.677 and n. 2, Mills (1997) 19–20 and n. 74. 102 On the name of the island see 52n. 103 Cf. also the picture in the Athenian temple of Dionysus Eleuthereus described at Paus. 1.20.3. For the various strands in Ariadne’s iconography see LIMC s.v. ‘Dionysos’ nos 708–79, s.v. ‘Ariadne’, Eisner (1977), Richardson 99

100

31

INT RODUCTION

Later authors and artists, however, remained aware that there were points of friction in this composite story.104 One area of complexity concerns Ariadne herself and exactly how she moved from one partner to the other (in either order). It is hardly surprising that according to Plut. Thes. 20.5 ‘some of the Naxians’ claimed that the wife of Dionysus and the lover of Theseus were simply two different Ariadnes.105 She is often associated with a crown or wreath, which in Pherecydes’ version is given to her on Dia by Dionysus and (at some unspecified point afterwards) catasterised;106 but in archaic Greek art she sometimes has a wreath instead of (or as well as) a ball of thread (113n.) when she is with Theseus on Crete, and this may reflect an alternative tradition in which she has already had some kind of involvement with Dionysus before Theseus arrives, and uses Dionysus’ gift of a shining crown to help Theseus see his way through the Labyrinth – perhaps a ‘betrayal’ of her divine lover that gives a reason for her to be killed, as she is in Od. 11.107 Meanwhile, even in the ‘canonical’ version, Ariadne’s transition from ­Theseus to Dionysus on Dia/Naxos may be presented in alternative ways that make a difference, not to how virtuous or otherwise her actions

(1979), McNally (1985), Gallo (1986), Deroux (1986), Dyer (1994) 231–43; cf. Gantz 116, 265–9, Webster (1966). 104 E.g. Plut. Thes. 20.1 πολλοὶ δὲ λόγοι καὶ περὶ τούτων ἔτι λέγονται καὶ περὶ τῆς Ἀριάδνης, οὐδὲν ὁμολογούμενον ἔχοντες. See also 58 immemor n. on Philostr. Imag. 1.15. 105 Elsewhere too Plutarch’s account suggests that the tradition varied in different places around the Greek world: for these ‘local’ versions see Calame (19962) 105–16. 106 The catasterism is frequently mentioned in Hellenistic and Roman poetry, including by Catullus at 66.59–61 (see n. 107). See Kidd on Arat. Phaen. 71–3, Harder on Call. Aet. fr. 110.59, Armstrong (2006) 312–16. 107 See Gantz 264–8, Webster (1966) 24–6. The only literary sources for the crown shining in the Labyrinth are Erat. Catast. 5 and Hyg. Astr. 2.5.1, but Eratosthenes attributes to Epimenides (3 B 25 D–K) the story that Dionysus gave Ariadne the crown on Crete. There does not seem to be any trace of Ariadne’s crown in Catullus 64, although it is possible that a reader of Catullus’ long poems might use 66.59–61 to complete Ariadne’s story (so Townend (1983) 26, Höschele (2009) 140–1, Ambühl (2014) 121–2; cf. above, p. 30).

32

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

might be understood to be,108 but to the balance of her experience between blessedness and misery (cf. 251–64n.). Theseus typically abandons her while she sleeps,109 but in some versions she awakes to find Dionysus already there, and herself already welcomed into her timeless existence as the wife of a god; in others, she wakes up alone on the island, and has time (though we may wonder how much time) to lament her situation before Dionysus appears. The impression given of events on Dia may also be affected by the question of whether anyone else was present. The first meeting between Dionysus and Ariadne may involve just the two of them, or a large Bacchic retinue, and even where Ariadne wakes up before Dionysus arrives, she is not always alone in her distress: visual representations often include symbolic figures, perhaps representing Love, Sleep, etc. (cf. 56n.), while Pherecydes stated that Aphrodite revealed her destiny to her before the arrival of Dionysus.110 For a discussion of these variant possibilities in wall-paintings from Pompeii and Herculaneum (in which Ariadne is a particularly frequent motif), see Elsner (2007) 91– 109 and cf. 1f above; for the routes explored by post-Catullan writers such as Ovid and Nonnus, e.g. 50–75, 132–201nn. and cf. 3a below. A second difficulty with what became the conventional telling of the myth is the question of why Theseus abandons Ariadne. The problem will have been particularly severe for authors, artists, and audiences influenced by the Athenian moulding of

Herter (1942) 228–37 argues that the Alexandrian Ptolemies in particular, for whom Dionysus was an important god, were concerned that Ariadne should not be understood to have behaved immorally; see however the scepticism of Hunter on Ap. Arg. 3.997–1004. 109 Cf. 56 fallaci … somno n. So far the detail is perfectly realistic, although the idea of a sleeping Ariadne woken by Dionysus (and/or by a dance: n. 98) would suit a nature-goddess being woken in spring, and such an idea may underlie Callimachus’ reference to a Naxian ritual at Aet. fr. 67.13–14 Pfeiffer/ Harder οὐδ’ Ἀριήδης | ἐς χ]ο ̣ρὸν εὑδούσης ἁβρὸν ἔθηκε πόδα. See e.g. Webster (1966) 22–3, McNally (1985) esp. 159, 183. 110 Pherecyd. 3 F 148a FGrH θαρρεῖν αὐτῇ παραινεῖ· Διονύσου γὰρ ἔσεσθαι γυναῖκα καὶ εὐκλεῆ γενήσεσθαι. This is suggestively similar to Venus’ epiphany to Europa at Hor. Carm. 3.27.66–76: see 3a(i) below. 108

33

INT RODUCTION

their city’s primary hero into a morally admirable character:111 such a man might be expected to keep his promises, honour the obligations he has incurred by accepting help, and avoid behaving cruelly towards someone under his protection. As already indicated, this problem can be solved by having Theseus removed by the gods: they might use force, give direct instructions, or induce amnesia. However, the suspicion that Theseus might have deserted a potential wife for his own more personal reasons also seems to have remained active,112 and this is not surprising given the number of women he is involved with in the mythical tradition,113 and the fact that this was the (or a) version given in a source with the authority of ‘Hesiod’.114 Against this background, we might wonder how fifth-century Attic tragedians handled these events and the characters of Theseus and Ariadne. A series of short dramatic fragments which concern Theseus and the Minotaur and include both Ariadne and Theseus as speaking characters (frr. 730a–g TrGF   ) has been doubtfully attributed to Sophocles’ Theseus, but the fragments do not give us any substantial evidence about how the relationship might have been treated in this play.115 For Euripides’ Theseus, as well as a few slightly more substantial fragments, we have most of a papyrus hypothesis (test. iiia TrGF, P.Oxy. 4640), from which it seems that this tragedy put both Ariadne and Theseus onstage, and dramatised events on Crete both before and after Theseus’ adventure in the Labyrinth. The extant text of the hypothesis ends with Theseus sailing to Athens but not marrying Ariadne, and with a reference to Minos and (his) ‘younger daughter’, and it has been suggested that a speech from a god (Athena?)

This may have begun around the late sixth or early fifth century: see in general Walker (1995), Calame (19962), Mills (1997). 112 For the various versions of both possibilities in different sources see 58 immemor n. 113 Cf. esp. Istrus, 334 F 10 FGrH (= Athenaeus 13, 557b), who apparently divided the list into τὰς μὲν … ἐξ ἔρωτος …, τὰς δ᾿ ἐξ ἁρπαγῆς, ἄλλας δ᾿ ἐκ νομίμων γάμων: see Jackson (1999) 153–4. 114 See above, pp. 30–1, on frr. 145, 298 M–W. Plut. Thes. 20.2 tells us that fr. 298 was deleted from the Hesiodic corpus by Pisistratus to please the Athenians, according to Hereas of Megara (486 F 1 FGrH ): see Liddel in BNJ 2 ad loc. 115 See Radt and Lloyd-Jones ad loc.; more optimistically, Mills (1997) 245–52. 111

34

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

at the end of the play may have given predictions of (or instructions for) these events.116 This is not enough, however, to tell us how Ariadne and Theseus were actually characterised during the main part of the play: whether Ariadne was modest and innocent, or shameless like the Phaedra of Euripides’ lost first Hippolytus, or whether Theseus was uncomplicatedly admirable, or manipulative like the Jason of his Medea.117 Perhaps Euripides chose to complicate the presentation of his own ‘national hero’ in this play as much as Virgil did in the story of Aeneas and Dido; perhaps he created a Theseus who looked likely to desert Ariadne even without divine instructions. In Hellenistic poetry, too, we may have lost a portrayal of Ariadne that would have had a significant influence on Catullus and presented Theseus from her point of view.118 In extant Hellenistic references to the myth, the fact that Theocritus’ Simaetha mentions Theseus’ obliviousness of Ariadne as part of a prayer for her own lover to forget his new flame (Theoc. 2.44–6) surely suggests that she thinks the tradition (45 φαντί) allows for ordinary human infidelity, even if the scholia ad loc. tell us the story in which Theseus is afflicted by the gods with literal forgetfulness (see 58, 200–1nn.); and Apollonius’ Jason, offering the helpful Ariadne as an exemplar for Medea (Ap. Arg. 3.997–1004, 1096–1101), clearly thinks it safest to avoid entirely the question of what happened to Theseus after he and Ariadne sailed away from Crete.119 Finally, although our lack of early sources makes it difficult to be confident,120 we might guess that Theseus’ other moment of ‘forgetfulness’, when he causes his father’s death by failing to hoist See Collard–Cropp i .415–17; for Theseus’ marriage to Phaedra see 118n. See esp. Wright (2019) 171–2, complicating the interpretation proposed by Mills (1997) 252–5. The hypothesis does say that Ariadne engaged in successful persuasion of both Minos and Theseus (test. iiia.11 TrGF ἔπ̣ε̣ ι̣ εν, 12 παρεστήσατο). 118 See 2a above and n. 79, 3a(ii) below and n. 331. 119 For this passage and Catullus 64 see esp. 53, 116–31, 117–18nn. In an allusive reversal of Jason’s use of exemplarity, Catullus frequently evokes Jason and Medea as an ominous model for Theseus and Ariadne (and even for Peleus and Thetis): see 1–11, 132–201nn., with further references. 120 Gantz 276. Simonides (550–1 PMG) mentioned Theseus’ coloured sails (227n.) and the death of Aegeus, but we can say very little about how he presented the story. 116 117

35

INT RODUCTION

the correctly coloured sail (202–50n.), might also have been seen as an awkward detail in the career of an idealised hero. Catullus’ response to the questions about Ariadne and Theseus raised in the mythical tradition is bound up with his decision to present the myth in an ecphrasis with narrative digressions – one including a speech (see 2c below). It may seem obvious that the emphasis of this version is on Ariadne as the unhappy victim of Theseus: the narrator initially seems to describe the kind of picture in which she awakes to find herself alone and miserable (52–70), draws attention to her part in many of the earlier events on Crete (86–104), and then – perhaps following the hypothesised Hellenistic Ariadne poem, but in any case making an enormous impact of his own on the subsequent tradition – allows her to impugn Theseus fully in a lengthy speech (132–201). Moreover, even when Bacchus does appear, the description of the god and his followers perhaps questions whether Ariadne will really be made happy by her divine rescue (251–64n.). At the same time, Catullus uses the idea that Ariadne cursed Theseus (apparently his own innovation) to connect together Theseus’ two instances of ‘forgetting’ into a story in which, in the narrator’s judgement as much as in Ariadne’s, Theseus is morally responsible for his poor treatment of her, and is punished by a kind of poetic justice (202–50n.). However, the return to ecphrastic form at the end of this story (see esp. 246–50, 251–64nn.) reminds the reader that the entire presentation of the myth in this poem is this narrator’s response to a picture – or rather, since the picture itself is fictional, this poet’s response to the tradition. The implication in 251–64 that the picture on the uestis (if we could only see it) might have had a very different emphasis from its ecphrasis reminds us that there could be other ways of depicting Ariadne’s experiences on Dia; the shifts in parts of the narrative digressions (esp. 76–85, 212–48) towards a point of view more associated with Theseus or with Athens suggest that there could be other ways of telling the story (note esp. 76 perhibent, 212 ferunt) of Theseus’ expedition to Crete. As in the case of Peleus and Thetis, then (2b(i) above), we are left to consider not only whether Ariadne ‘is’ ­happy, whether Theseus ‘is’ perfidious, but whether this version of the myth is the one we should trust. 36

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

(c) Forms (i)  Narrative and Speech The basic literary form of Catullus 64 is epic narrative, incorporating, as epic narrative usually does, extended direct speech from some of the characters. As is also common in epic, this form also ‘hosts’ or temporarily morphs into other forms:121 most importantly, the ecphrastic description of a work of art (2c(ii) below), an inset song with a refrain (323–81n.), and a discursive conclusion (382–408n.), but also others on a smaller scale, which are discussed in more detail in the commentary on the relevant lines (and in several cases also mentioned later in this subsection). I focus here, however, on the poem’s use of the narrative and speech which are the predominant forms both in the outer story of Peleus and Thetis and – in fact – in the inner story of Ariadne,122 and suggest some of the most significant ways in which its techniques can be seen to relate to (what we know of) the literary traditions to which Catullus was responding, both in epic (broadly understood) and in the possibilities offered by some other genres for ways of telling mythical stories or presenting mythical speech.123 The poem’s narrator-text, then,124 includes passages of fast-moving summary narrative which typically appear at the ­beginning of new sections of the poem and give the background to a part of the story that will be narrated or described more

The terminology of ‘host genre’ and ‘guest genre’ is used by Harrison (2007a). The latter only begins and ends with formally ecphrastic passages (2c(ii) below). 123 The predecessors discussed will be overwhelmingly Greek: the fragmentary state of the evidence for pre-Catullan Roman narratives (see 1b above) makes it particularly difficult to judge whether there was anything in Ennius’ Annales, for example, that might have influenced the aspects of Catullus’ narrative technique that I will draw out in this subsection. The recent turn in Ennian scholarship towards the re-evaluation of evidence for the large-scale architecture of the Annales (esp. Elliott (2013); see Farrell–Damon (2020) esp. 2–4) makes it intriguing to speculate whether Ennius’ narrative on a smaller scale, too, might have been less straightforwardly complete and ordered than has traditionally been assumed. 124 With the two following paragraphs cf. the more strictly schematic discussion of Bartels (2004) 17–60, who analyses the whole poem in terms of the narratological categories of order, tempo, mode, and voice. 121 122

37

INT RODUCTION

slowly: key examples are 1–11, 73–85, 116–23, and it is in this compressed manner that the poem tends to cover some traditionally epic motifs such as sea journeys (featured in all of these passages) and the motivation for heroic action (5–6, 80–3). Other passages at least begin by narrating specific actions at a more moderate pace: so e.g. the Nereids rise from the water (12–15), the Thessalians arrive at Peleus’ palace (31–5) and leave again (267–8, 276–7), Ariadne sees and falls in love with Theseus (86–7, 91–3), or, to turn from human action to the poem’s very brief equivalent of a Homeric divine scene on Olympus, Jupiter hears and responds to Ariadne’s curse (202–6). However, none of these is extended beyond a few lines: one of the longest – and arguably the one which offers the clearest example of narrative resolution in the whole poem – is the five-line account of the death of Aegeus (241–5), which is followed by a sentence containing more moralising comment than narrative (246–8). Other passages of this kind are similarly succeeded or interrupted by passages of a less purely narrative mode. Some introduce a speech (124–31, 212–14). In many places, the narration slows down so dramatically (e.g. 278–93), or narrates so impersonally (e.g. 38–42), or concentrates so much on visual details (e.g. 43–9), or on an ongoing, repeated action (124–9, 305–20), or on someone’s inner emotions described with imagery and simile (87–90, 207–9, 238–40), that it turns to a greater or lesser extent from narration into description, thus bringing many of the narrative sections of the poem close to the mode of the strictly ecphrastic passages.125 Epic similes break up sections of narrative (89–90, 269–77) or even replace the narrative that might have been expected (105–11n.). This unpredictable use both of narrative tempo or duration and of focalisation (some of the passages just discussed, e.g. 86– 93, strongly convey the subjectivity of a character, while others, e.g. 278–93, appear much more objective or distanced) contributes to the fragmentation of the text into a series of sections that

125

Note esp. the smooth transition, helped by the present tenses, from the description of Peleus’ palace in 43–51 to the description of Ariadne that begins at 52 (see further 50–75n.).

38

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

strikes many readers of 64;126 and this fragmentation also relates to the role played by a third narratological category, that of narrative order. The story of Ariadne is told by means of flashbacks and flashforwards from the scene on Dia that is both described as it appears on the uestis (52–70, 249–64) and separately narrated (124–201), while the story of Peleus and Thetis, although it proceeds in a sequential order which is self-consistent, contains innovations to the usual chronology of the mythical tradition (2b(i) above) that create discontinuities between one part of the narrative and the next (see esp. 31 finito tempore n.). Unsurprisingly, then, the poem is full of moments of transition from one section of narrative (or description) to another.127 On the part of the narrator, these can be handled either ‘actively’ or ‘passively’.128 Sometimes the narrator steps forward, either explicitly drawing attention to the narrative possibilities among which he is choosing (116–31n.), or commenting on his subject matter (e.g. 246–8, 301–2), in several cases using apostrophes, exclamations, and questions to present himself as in some degree involved with it (e.g. 68–72, 94–102, 253), and ultimately coming as close as he can to leaving epic narrative behind either for extended discursive ‘commentary’ (384–408) or for lyric speech (22–30). At other transitional moments, the narrator is much less obtrusive, and yet repeated techniques highlight how he is choosing to respond to the mythical tradition (quondam and other adverbs of time: 1, 76, 124, 212, 296, 382; dicuntur and other verbs meaning ‘they say’: 2, 19, 76, 124, 212), or how he apparently sees scenes or events as contrasting with each other (at : 43, 241, 251) or as causally related (nam or namque : 52, 76, 105, 212, and cf. 384); in a less defined way, even a relative pronoun or adjective (12, 31, 80, 205, 249, 267, 278, 303) or a Cf. in particular readings which label individual sections of the poem in order to make an argument about its structure: examples include Thomson (1961) 50, Traill (1981) 233, Warden (1998) 406. On sections of the text and objects within the text, see also the Epilogue, Section 1, and cf. 2d(ii) below. 127 For some of what follows on these transitions I gratefully acknowledge the influence of an unpublished paper by Rodie Risselada, ‘The language of ­Catullus 64’, presented at an OIKOS colloquium on Catullus 64 held at the Vrije Universiteit Amsterdam in December 2013. 128 Cf. DeBrohun (2007) 294. 126

39

INT RODUCTION

part of talis (265, 382) may still be used to imply that the complex use of order and duration in his narrative results from connections made subjectively in his own mind.129 The result of all this is that although the poem tells almost all of the canonical story of Ariadne, and has precedent for its omission of Eris from the wedding of Peleus and Thetis (2b above), its handling of narrative leaves the reader feeling that both stories have been told selectively, with certain aspects given much more attention than others. In fact, this short epic contains very little ‘ordinary’, forward-moving epic narrative in comparison to the amount of space given to (in themselves often unexceptionably epic) special effects: direct speech, descriptions, similes, and more or less overt interventions from the narrator, among which we can include the explanatory or scene-setting summaries of background events. As we shall see, then, although it is not productive to look for a comprehensive ‘model’ for its overall narrative technique either in a single lost poem (see 2a above) or in a codifiable genre (see 1b above on ‘epyllion’), literary predecessors of a comparable length have a particularly significant role to play. The traditional assertion that Catullus 64 draws particularly on Hellenistic narrative forms is essentially true, but in a more complex way than has sometimes been implied. Not all Hellenistic narratives (not even all short ones) are the same, and some of their characteristic techniques can also be traced back to earlier Greek poetry.130 We may start with the way in which a selected mythical episode is told apparently for its own sake in poems such as Callimachus’ Hecale, Theocritus 24, or Moschus’ Europa: however, opening techniques which may draw attention to the narrator-poet’s choice of a starting point, including the indefinite temporal adverb, reference to earlier authority, and summary of the background before the narrative slows down into the main ‘scene’, are used not only in such free-standing Cf. Williams (1968) 228–9, and for the narrator’s subjective reception of the mythical tradition in the poem, see also the Epilogue, Section 3. The Fates use some of the same formulae (349 saepe fatebuntur nn., 353 namque, 366 nam): see 323–81n. for their relationship to the narrator. 130 See esp. Perrotta (1923), Fantuzzi–Hunter (2004) 191–6. 129

40

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

narratives but in those which are part of larger structures (1 quondam, 2 dicuntur nn.). On the one hand, the scene-setting at the start of certain episodes in Apollonius’ Argonautica (e.g. Arg. 1.609–32, on the situation on Lemnos; 1.1211–20, on Hylas’ background) suggests the selectivity of a rhapsode, performing a shorter ‘extract’ from a larger body of epic material: compare how Demodocus begins his two songs in Od. 8, the first with ποτέ (Od. 8.76), the second with a summary of the background (Od. 8.499–504), or the opening of the Hesiodic Shield, ἤ οἵη …, which aligns it with the episodes of the Catalogue of Women.131 On the other, mythical episodes introduced in similar ways may appear in contexts in which they are (ostensibly) told for a purpose, often exemplifying or explanatory. This may be in a personal context (cf. esp. Theoc. 13, often called an epyllion, but in its use of a mythical situation to illustrate a contemporary one more similar to e.g. Call. Ia. 12);132 or in a hymn (compare the narratives beginning at Call. Hymn. 3.4, 4.51, 5.57, 6.24 with those of the Homeric Hymns; Theoc. 22 is also explicitly framed as a hymn, while in Theoc. 26 and probably [Theoc.] 25 the hymnic nature of the poem is revealed at the end, as in Catullus 63); or in aetiological or even didactic poetry (in Call. Aet. cf. e.g. the story of Acontius and Cydippe, frr. 67–75 Pfeiffer/Harder; in Arat. Phaen. esp. the myth of Dike, Phaen. 96–136 (cf. esp. 384–6, 397–408nn.), drawing on the wider variety of mythical narratives in Hes. Op.). Although there is no framing device at the very beginning of 64, Catullus does briefly associate it with the hymnic tradition at the end of the opening sequence (19–30n.); the idea of a hymnic purpose then seems to be forgotten, but the epilogue invites reinterpretation of the poem’s narrative(s) as having aetiological or didactic explanatory value instead (382–408n.). It seems that a genre in which the poet is called upon to select appropriate mythical episodes for the purpose in hand will tend to encourage the foregrounding of that poet’s voice and of the The first fifty-six lines of the Shield, on the background to the birth of Heracles, were also transmitted as part of the Catalogue. 132 For Fernandelli’s identification of Call. Ia. 12 as a potential influence on ­Catullus 64, see 323–81n. 131

41

INT RODUCTION

choices he is making. Another such genre is epinician, and the narrative technique of 64 has been compared to that of P ­ indar at least since Realinus.133 In the epinicians (and hymns) of Pindar and Bacchylides, as in 64, we can see the poet drawing attention to his choice of episodes and to the version he is choosing to tell (cf. esp. Ol. 1.25–53 on Pelops), and even using relative pronouns to introduce a new story as if the connection has just occurred to him.134 In principle, too, there is little need for a poet writing in such a genre to tell mythical stories in chronological order: perhaps the closest comparanda for the story of Ariadne as treated in 64 are Pind. Pyth. 4, in which Medea’s prophecy on Thera is followed by a select narrative of the entire Argonautic expedition, and Bacchyl. 11.40–112 Snell–Maehler, which contains a flashback on Proetus in the middle of the story of his daughters, although neither offers such an ostentatiously non-chronological arrangement of so many episodes from the same myth. The ‘digressions’ put into the mouths of characters in Callimachus’ own aetiological epinician, the Victoria Berenices,135 but also in his Hecale, are more often cited as models for the way in which 64 includes events that are in the past or the future from the point of view of the ‘main narrative’, but there is every difference between hearing a character reminisce or even prophesy (cf. 323–81n.) and listening to the narrator demonstrate his freedom to choose from a body of available mythical material. At this point it is helpful to reintroduce the question of scale. In a narrative much shorter than the Iliad or Odyssey, a speech, ecphrasis, or simile that is roughly the normal Homeric length (or even considerably shorter) will, simply by reason of its length, have a greater prominence in the poem, a prominence that can potentially be used for all kinds of purposes depending on the nature of the text: several of the (relatively) short poems discussed Realinus (1551) 12v, on 64.52: ‘nisi dicat aliquis illum poetas Lyricos esse imitatum, qui … longius euagari possunt: ut Pindarus non saepe modo sed semper fere facit’. 134 On such ‘relative connection’ in Pindar see Des Places (1947) 48–50, Pfeijffer (1999) 37–41. 135 This poem, if it should be thought of as a separate poem, opened book 3 of the Aetia (frr. 54–60 Harder); for further scepticism of the idea that it should be privileged as a model for 64, cf. 1b and 2a above. 133

42

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

above in this subsection, both hexameter and lyric, are therefore precedents for 64 in the proportional space that they give to such elements.136 One way of understanding Catullus’ narrative technique in 64, however, is that he has gone one step further than these predecessors by, as we have seen, building up his poem almost entirely out of speeches, similes, descriptions – and also out of the kind of narratorial interventions, whether emotional apostrophes or scene-setting flashbacks and other digressions, that have precedent in Homer, but are a much more important part of the technique of Apollonius’ Argonautica.137 In Apollonius all such special effects are still ultimately absorbed into the linear narrative of the longer epic: in particular, after a flashback, Apollonius’ narrator always returns to the ongoing story of the Argonauts.138 In contrast, Catullus’ narrator allows what begin as explanatory flashbacks (76 nam perhibent olim, 212 namque ferunt olim: cf. δὴ γάρ at Ap. Arg. 1.611, 1211) to develop into extended passages of narrative, description, exclamation, simile, and speech that in fact carry much of the content of the story – even if he claims at one point (116–23) that he is getting his narrative back on track. Catullus has therefore taken building blocks of Apollonian narrative technique, just as he has responded to many aspects of the content of the Argonautica;139 but his manner So e.g. in [Hes.] Scut. the ecphrasis of Heracles’ shield takes up 182 lines out of 480 (the ecphrasis of Achilles’ shield in Il. 18 is 131 lines), while in Mosch. Eur. the ecphrasis of Europa’s basket occupies 26 lines out of 166; on these ecphrastic poems see also 2c(ii) below. Moschus’ poem contains three speeches, two from Europa (21–7, 135–52) and a prophetic one from the bull (154–61); mythological prophecies are given prominence by being the only direct speech in e.g. Pind. Ol. 8, Isthm. 8. In its 74 lines, Theoc. 13 uses a series of three similes culminating in one describing Heracles’ emotions at the climactic moment (24, 50–2, 62–3). For the wider dynamics of Theocritus’ use of proportion in both narrative and dramatic poems (especially the way in which he balances song against ecphrasis in both 1 and 15 just as he often balances song against song), and the particular relevance of this to 64, see 323–81n. and the Epilogue, Section 1 and n. 27. 137 See Hunter (1993) 101–19. 138 As he does so at Arg. 1.1220, the narrator strongly implies that he could tell a much longer story about Hylas’ father Theiodamas: cf. 116–31n. Catullus essentially takes up this possibility. 139 For the tradition of identifying the Argonautica as a particularly important influence on the form and content of 64 (e.g. Ellis (18892) 283 ‘the best 136

43

INT RODUCTION

of combining them into a unique kind of short epic draws on – but goes further than – the ways in which poets including Callimachus, Theocritus, and even Pindar had put together elements from the epic tradition to create poems on a smaller scale. Many Hellenistic hexameter poems, short or long, explore the interface between epic and tragedy, and Catullus is also the inheritor of this tradition. The speeches of Ariadne and Aegeus (132–201, 215–37) exploit the opportunities offered by their first-person form to make use of elements from other ‘spoken’ genres such as comedy, oratory, and various kinds of personal poetry: see the commentary on both for connections to such poetry within the Catullan corpus (cf. 1a above), to lyric of various kinds, to epigram, and to subsequent Latin elegy. But above all they draw on epic precedents alongside others from tragedy, both Greek and Roman, and in Ariadne’s speech in particular we can see a conflation of allusions to both genres when she echoes both the Apollonian and the Euripidean–Ennian Medea (see esp. 132–201, 132–8, 177–83nn.). Where some Hellenistic poets, however, had brought their entire hexameter poems closer to tragedy by their use of extended character-text – with Lycophron’s Alexandra cast as a single tragic messenger-speech, and [Moschus]’ Megara, with its conversation between Megara and Alcmena, approaching the purely dramatic form of a tragic scene – 64 is in some ways more like Callimachus’ Hecale in evoking tragedy by means of its narrative structure in combination with aspects of its content.140 I once argued for reading the inner story as a ‘tragedy of Ariadne’;141 while Ariadne’s speech

i­llustration of it is to be found in Apollonius’ Argonautica, of which Catullus must have made a considerable study’, Fernandelli (2012) 160–210 ‘il modello principale’, Clare (1996), DeBrohun (2007)) see Calzascia (2015) 50–70. Calzascia’s detailed examination of the similarities and differences between the two poems does not, in my view, succeed in proving her conclusion that the importance of the relationship between them has been overstated. 140 For Hecale as ‘Callimachus’ “tragedy”’ in a similar sense see e.g. Acosta-Hughes– Stephens (2012) 198, who note on the structural side that the poem has ‘a tight temporal frame encompassing one night and one day, one narrative event, anagnorisis and peripeteia, prologue, inserted dialogue, a “messenger” speech, and a “tragic” ending with the death of Hecale commemorated in the foundation of a cult’. 141 Trimble (2009); cf. also Fernandelli (2012) 363–7 and see 267–8n.

44

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

might suggest an alignment with Hellenistic solo performances of tragic excerpts,142 the narrator can be seen to extend it into a complete praxis – with a scene-setting prologue (perhaps all of 52–123), messenger-speech narrating the fulfilment of a curse (202–48), and concluding deus ex machina (251–64) – even as his repeated return to Ariadne on the shore implies the other tragic unities of time and place.143 The outer story too, however, is focused on the events of a single day in a single location,144 incorporating its reference to the future by means of a divine prophecy;145 understood as part of this story, the ecphrasis might be considered equivalent to a choral ode narrating a separate myth that is somehow to be compared or contrasted with the main one being staged,146 while Catullus’ account of Peleus and Thetis begins still more obviously with a tragic prologue in the form of the poem’s opening allusions to the prologues of the Medea tragedies of Euripides and Ennius. Tragedies too can begin with ‘once’, ποτέ (1 quondam n.), and with a summary of background events, because they are another way of presenting a mythical episode. Perhaps it is even possible to read the subjective narrator as some kind of tragic character, even though he began by sounding much less emotionally interested in the sailing of the Argo than was Medea’s tragic Nurse (1–11n.). Wiseman’s theory that Catullus 64 might have been a ‘libretto’ for a fully staged, danced, and sung performance may always remain an outlier;147 but if the poem was ever read by Catullus to (or with) his friends,148 See esp. Gentili (1979), Easterling (1997). It is unity of action which is most clearly stipulated for tragedy in Arist. Poet. (1450b23–6, 1451a16–35), while the traditional unities of time and place owe more to early modern observation of actual tragic practice; for a tragic plot as typically occupying no more than a single day cf. however Poet. 1449b12–13. 144 For the idea that the narrator imagines himself present at the wedding of Peleus and Thetis, see 32 frequentat n. and the Epilogue, Section 3 and n. 49. 145 Cf. e.g. the prophecies made ex machina by Heracles at the end of Soph. Phil. or Dionysus at the end of Eur. Bacch., or the prophecies of the eventual fall of Zeus made by Prometheus throughout [Aesch.] Prom. 146 The third stasimon of Eur. IA (2b(i) above) would be one example of this. 147 Wiseman (2015) 109–10, (2018). 148 This is the first suggestion made by Wiseman (1985) 127–8. Skinner (2001) thought-provokingly imagines how some of Catullus’ shorter poems might have been received in this kind of performance context. 142 143

45

INT RODUCTION

the l­iteral voicing of the narrator and the characters would have had an interesting effect on the audience’s understanding of the relationship between epic narrative and speech.149 (ii) Ecphrasis Catullus 64 not only tells one story inside another, but frames that inner story as an ecphrasis.150 A substantial scholarly interest in ancient ecphrasis has developed in the late twentieth and early twenty-first century: in part this has been a response to the importance that the concept has gained in literary and cultural studies more generally over a similar period, especially as a way of thinking about the relationship between visual and verbal media.151 In antiquity, ἔκφρασις as a critical term, attested especially in scholia and in the progymnasmata of imperial Greek rhetoricians, covered a wide range of descriptive writing and speech: what mattered was not the object of the description but the effect it had on the reader or listener, bringing what was described ‘before the eyes’ (ὑπ’ ὄψιν) by means of ‘vividness’ (ἐνάργεια, a key concept) and ‘clarity’ (σαφήνεια).152 Many classical scholars, however, continue to use ‘ecphrasis’ in its modern sense of the (verbal) description of a work of (visual) art, and this is the practice which I generally follow (although, as we shall see in this subsection, it is necessary to stretch the definition when using the word to discuss Catullus 64). One reason why this usage is justified is the undeniable existence of a strong tradition of such ecphrastic descriptions in ancient literature.153 Two broad categories can be

Also, for instance, on how they might have perceived the separability (or otherwise) of Catullus and the narrator (1a above), or the reference of 406 nobis (see n.). 150 For attempts to explain this central feature of the poem’s narrative structure with reference to the genre of ‘epyllion’ and/or to reconstructed predecessors, see esp. 1b and 2a above. 151 For a general overview and bibliography see esp. Squire (2015); some additional material in Koopman (2018) 2–5. 152 See above all Webb (2009). 153 The key survey of these remains Friedländer (1912) 1–103. For the argument (contra Webb (2009), esp. 1–11) that ancient writers and readers would have recognised the descriptions of works of visual art as ‘a paradigmatic example’ 149

46

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

identified.154 Firstly, freestanding ecphrases, including ecphrastic epigrams155 and the Imagines of Philostratus.156 Secondly, and most relevantly for this poem, ecphrases inset into other literary forms: these are particularly found in epic (and Hellenistic variants of epic), but also in drama (and Hellenistic variants of drama), and later in the novel. Catullus’ creation of an inset ecphrasis should therefore be seen in relation to such ecphrases in earlier works,157 both longer forms (the Homeric epics, especially the description of the shield of Achilles at Il. 18.478– 608; Attic tragedy, e.g. the shields of the Seven in Aesch. Sept., the temple sculptures at Eur. Ion 184–218, or Achilles’ un-­ Homeric shield at Eur. El. 452–86 (cf. 323–81n.); Apollonius, especially Jason’s cloak at Arg. 1.721–68; and more)158 and also shorter ones, in which the ecphrases, although still inset, have a proportionally greater weight (2c(i) above: note especially the description of Heracles’ shield at [Hes.] Scut. 139–320; two Theocritean examples, both occurring in poems whose form is dramatic, the bucolic Theoc. 1 (the goatherd’s ecphrasis of the cup at 27–56), and the urban Theoc. 15 (the descriptions of the Adonis display given by Gorgo, Praxinoa, and the singer at 78–86, 112–30); the descriptions of temple artworks given by the characters of Herodas 4; and the narrator’s description of Europa’s basket at Mosch. Eur. 37–62).159 One fundamental way in which the ecphrasis in 64 resembles these canonical of (the broader ancient understanding of) ecphrasis, see Elsner (2002) 2 and cf. Squire (2008). 154 See esp. Elsner (2002). 155 On this tradition see e.g. Goldhill (1994), Bing–Bruss (2007). 156 Elsner (2002) 13–15 discusses the ways in which the work of Philostratus and his successors also draws on the traditions of the second category. 157 Schmale (2004) 106–28 discusses several of these examples as background to Catullus 64. 158 Dufallo (2013) 14–38 discusses some less frequently cited early Roman examples in Naevius’ Bellum Punicum (an ecphrasis, now extremely fragmentary, of an object depicting the Gigantomachy) and in Plautus’ Menaechmi and Terence’s Eunuchus. Dufallo approaches ecphrases in Roman texts (including that in Catullus 64: Dufallo (2013) 39–73) as a refraction of the wider Roman reception of Greek culture: cf. 1d above. 159 Cf. also 2a above for the argument made by Thomas (1983) for an inset ecphrasis in Callimachus’ Victoria Berenices, with the doubts expressed by Harder on Call. fr. 54.11–19.

47

INT RODUCTION

examples is that the artwork to which it relates is figural, not abstract; the ecphrasis is therefore ‘the verbal representation of visual representation’,160 although, again, the way in which it ‘represents’ the visual representation that seems to be in question stretches the usage of such terminology. Only the opening and closing passages of ‘the ecphrasis’ are strictly identifiable as descriptions of the picture on the uestis at Peleus’ and Thetis’ wedding. See esp. 50–75, 249–50, 251–64nn.: after the introductory statement in 50–1 about the content communicated by the uestis, 52–70 describe Ariadne on the beach, while 71–5 introduce a narrative mode continued by 76 nam perhibent olim. The description of the picture does not resume until 249, and lasts until 264, while 265–6 again make a statement about the uestis as a whole. The ways in which these opening and closing passages relate to the content and techniques of ancient ecphrasis (and visual art) are discussed in the commentary: what emerges particularly strongly, however, is that explicitly they say almost nothing about the artwork itself (except in the introductory and concluding couplets at 50–1 and 265–6, and at 251 at parte ex alia), and nothing at all about its creator (unless we count 51 mira … arte), or about its viewers, although those viewers’ experience is discussed after the formal end of the ecphrasis at 267–8. In the terms of Becker’s analysis of ‘levels of representation’ in ecphrasis,161 the text ignores the opus ipsum, the artifex, and the animadversor.162 Instead, it operates on the level of the res ipsae, essentially describing Ariadne and Bacchus rather than the picture of Ariadne and Bacchus.163 And yet, since both of these passages open by telling or reminding the reader that there is a picture in question, we may still read them both as examples of ecphrasis – but, in Laird’s terms, of ‘disobedient’ ecphrasis,164 which does not ‘limit itself’ to ­describing what a visual medium can depict, but specifically Heffernan (1993) 3. Becker (1995) 42–3. 162 Contrast e.g. the examples from Il. 18, [Hes.] Scut., and Mosch. Eur. given in n. 179 below. 163 Cf. Trimble (2020) 122–3. 164 Laird (1993) 19 and passim: see further esp. 50–75, 261–4nn. 160 161

48

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

explores the possibilities open to visual, descriptive, and narrative art by laying heavy emphasis on aspects of the scene that are easier to convey by verbal than by visual representation, including sound, movement, psychology, and the passing of time. In the same way, even though the central part (the majority) of the poem’s inner story (71–248) consists of narrative, almost all of it formally marked off by ‘they say’ formulae implying that the narrator has a verbal source for his material separate from the picture on the uestis,165 it is still reasonable to refer to the whole inner story as ‘the ecphrasis’, since its formally ecphrastic frame presents it in its entirety as the narrator’s response to the visual stimulus which that picture provides. Moreover, going beyond this question of terminology, many of the issues raised in scholarship on ancient ecphrasis are relevant to the unusual ecphrasis in this poem. The move onto the level of the res ipsae, leading as it does to the narrative digressions,166 might be understood as a particularly effective enactment of ecphrastic enargeia,167 in which, as readers, our attention is so thoroughly absorbed by the depicted scene that we receive it as if it were real (or rather, as if it were any other piece of mythical material being conveyed to us through epic poetry); Krieger has argued that enargeia in fact leads to an empathetic fusion of subjectivities between the describing poet and the reader, since we are ‘called upon to identify ourselves with the poet in participating similarly (or rather identically) in the described e­ xperience’.168 However, in this ecphrasis the presence of Catullus’ narrator, and the choices he is making, are so obtrusive from the very beginning (see esp. 50–75n.) as to make us suspicious of such an identification with

76 perhibent, 124 perhibent, 212 ferunt; contrast however 116–23 (see 116– 31n.) and perhaps 202–11 (see 202n.), 238–50 (see 238–40n.). In my view, this articulation of the inner story should lead the reader to conclude that the uestis ‘actually’ depicts only what is described in 52–70 and 249–64 (insofar as it is fair to draw any such conclusion about this fictional object): see further 50–75, 249–50, 251–64nn. 166 On narration, description, and ecphrasis see Fowler (1991), Koopman (2018). 167 Iff-Noël (2019) argues that Catullus specifically elaborates on the topos found in Hellenistic ecphrastic epigrams that artworks may ‘seem to speak’ and ‘come to life’ (cf. 132–201n.). 168 Krieger (1992) 94, also discussed by Bartsch–Elsner (2007) iii. 165

49

INT RODUCTION

his e­ xperience of viewing the picture of Ariadne even as we are drawn into it: we might compare the treatment of ecphrasis in some of the dramatic examples mentioned above in this subsection, in which the responses of the viewing and describing characters are similarly prominent in the text.169 When we finally discover that the narrator has chosen not to reveal to us until the very end of the ecphrasis that the uestis also depicts Bacchus and his retinue, we are reminded not only that every piece of ecphrastic text is focalised through (at least) the one who utters it, but that that person’s subjective choices can actually hinder enargeia if they confront us with the fact that the picture being described is not in fact ‘before our eyes’ but that we see it only through theirs.170 We may take this further and consider that, like other examples of ecphrases inset into epic,171 this one is fictional, with the poet (perhaps even the narrator) in fact creating the artwork that he is professing to describe.172 It fits into the tradition of ecphrasis as mise en abyme,173 in which, even while it highlights the differences between verbal and visual art,174 the technique also aligns the two media with each other by encouraging the reader to reflect on how an art object or a text is created and (especially) received, and also more generally on the communicative power of visual signs and of words, and on the factors that may complicate or impede such communication (see esp. 267–8n. on For instance, the members of the chorus at Eur. Ion 184–218 direct each other’s attention to different depicted characters and scenes, as Catullus’ narrator directs the reader’s attention at e.g. 58, 60, 251; the exclamations of the observers at e.g. Theoc. 15.80–6, Herodas 4.20–34 might be compared to the narrator’s at 71, or even his apostrophes at 69, 253. 170 For this effect see esp. 50–75, 267–8nn. Cf. also 2b(ii) above on the ecphrasis as the narrator’s interpretation of the myth of Ariadne. 171 Contrast ecphrases that respond to real artworks, such as the epigrams on Myron’s sculpture of a cow (AP 9.713–42, 793–8) or the ecphrases in Pausanias. 172 This is highlighted by the lack of information given about the creator of the uestis (51 mira … arte n.). 173 See e.g. Elsner (2002) 3–9 and on mise en abyme in general Dällenbach (1989). Throughout the commentary I use ‘mise en abyme ’ in its looser sense, to refer to an inset object which resembles in some relevant way the whole of which it is a part, rather than replicating that whole precisely on a smaller scale. 174 Not necessarily implying the superiority of one over the other, although ecphrasis is sometimes read as an intervention in the so-called paragone (‘rivalry’) between text and image: see e.g. Becker (2003).

169

50

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

the experience of the internal viewers of the uestis, 202–50n. on the further mises en abyme within this ecphrasis of Aegeus’ mandata and Theseus’ dyed sail, and the Epilogue, Section 1). This effect is sharpened by the fact that the object of Catullus’ ecphrasis is a textile and activates the traditional metaphorical association between textile and text, thematised elsewhere in the poem in the descriptions of the spinning Fates and even of the Argo.175 The idea of a mise en abyme, however, raises more than formal considerations, since it also invites reflection on the relationship between the content of an ecphrasis (both what is depicted in the artwork, and the further material that its describer may include) and the content of the literary work into which it is inset. This has sometimes been discussed in terms of the distinction between ‘meaningful’ and merely ‘decorative’ ecphrastic episodes within longer works, though we might wonder whether there are any true examples of the latter.176 But, again, C ­ atullus 64 is a special case because its ecphrasis is obviously not a subordinate episode: although the poem has been compared in particular to two other short epics which make ‘meaningful’ use of See 10, 12–18nn. and esp. 303–22n. The classical tradition of textile ecphrases begins in Homer with Helen’s weaving at Il. 3.125–8 and Aphrodite’s girdle at Il. 14.214–23; cf. also esp. Jason’s cloak at Ap. Arg. 1.721–68, the hangings around Adonis at Theoc. 15.78–86, and after Catullus e.g. Cloanthus’ cloak at Virg. Aen. 5.250–7, the weaving of Minerva and Arachne at Ov. Met. 6.61– 128, Proserpina’s abandoned tapestry at Claud. Rapt. 1.246–72, and, at Ciris 21–34, the refusal to include more than a brief ecphrasis of a peplos offered to Minerva (cf. 1b above and n. 25, 3a(ii) below and n. 290). In some of these examples the implied comparison between the textile artwork and the text in which it appears is provocatively combined with a focus on the female creator of the textile (and/or its female viewers); one of the things that Catullus 64 does not tell us about the creator(s) of the uestis is their gender, even though it later shows us female figures producing both a textile and a text. 176 Friedländer (1912) (n. 153 above) argued that archaic ecphrases were decorative while those of the Hellenistic period began to be meaningful. For a thorough theoretical investigation of how we might approach the ways in which meaning might be generated between an ecphrasis and its setting, see Fowler (1991); note also esp. Perutelli (1978). One particular tradition, in which the content of an ancient ecphrasis acts as a prolepsis (sometimes even a prophecy) of events in the world in which the object of the ecphrasis appears, has been analysed by Harrison (2001) (cf. also Harrison (2010), (2013)); Harrison (2001) 84–7 believes that the ecphrasis in Catullus 64 implies an unhappy future for Peleus and Thetis (cf. 251–64n.). 175

51

INT RODUCTION

their prominent ecphrases, the Hesiodic Shield and Moschus’ Europa,177 even simply in length its ecphrasis takes up proportionally greater space than theirs do,178 while the ‘disobedient’ nature of its ecphrasis contrasts with the more traditionally Homeric ecphrastic techniques with which both the Shield and the Europa mark their ecphrases as different in texture from the surrounding narrative.179 In Catullus 64, as we have seen (2c(i) above), the ecphrasis becomes an epic narrative, while the surrounding text is dominated by visuality and description (cf. also 1f above); in its form, the poem therefore takes to extremes the hypervisuality or ‘ecphrastic contagion’ emphasised in some scholarship on ecphrastic texts,180 while also experimenting with what might be called hypernarrativity or ‘narrative contagion’ within its own ecphrasis.181 In combination with the balanced proportions of the poem, this creates a structure in which, in content, too, the ecphrasis has the potential to be ‘meaningful’ for the interpretaThe ecphrasis of Europa’s basket, with its scenes from the myth of Io, suggests multiple connections and contrasts with Europa’s story; see also n. 181 below. For arguments about the meaning of the ecphrasis of the Shield in relation to that poem’s outer narrative of Heracles, see esp. Toohey (1988), Horn (2016). For Catullus 64 and the Shield see esp. Wheeler (1934) 133–5, Konstan (1993), and for reflections on ecphrasis as a mise en abyme of ‘epyllion’ Landolfi (1998), Prioux (2016); cf. also 1b and 2c(i) above. 178 Catullus 64 therefore goes further than these predecessors (even the Shield) in the use of textual proportion discussed in 2c(i) above and n. 136. 179 Most obviously the repeated ἐν (δέ) ‘(and) on it’ ([Hes.] Scut. 144, 154, etc., Mosch. Eur. 43, 44, 50; cf. Il. 18.481, 483, 485, etc.); cf. also other references to the opus ipsum (arrangement, material, colour), to the artifex (Hephaestus in all three cases, evoked both by name and by the use of τεύχω, ποιέω, etc.), and to the animadversor (especially in the theme of wonder: [Hes.] Scut. 140, 165, etc., Mosch. Eur. 38; cf. Il. 18.549). 180 That is, the way in which ecphrastic writing encourages its readers to respond to the textual evocation of visual perception elsewhere (especially elsewhere in the same text): the term ‘ecphrastic contagion’ was coined by Whitmarsh (2002) 111–12. 181 Here Europa is a key predecessor, since its narrative contains highly pictorial sections (see esp. Harden (2011)), while its ecphrasis consists of scenes from a single myth, which invite combination into a narrative (Petrain (2006)). However, the overall effect is different because the ecphrasis itself is restricted in size and in other respects maintains traditional ecphrastic technique (above, this subsection and n. 179). Moschus’ readers do not forget they are reading an ecphrasis, nor do they have difficulty in deciding whether Europa or Io is the main subject of the poem. 177

52

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

tion of the outer story, and the outer story also has the potential to be ‘meaningful’ for the interpretation of the ecphrasis. (d) Style Two contexts are particularly relevant for understanding the style of Catullus 64. One is the history of Latin poetic style in general, but especially in epic and also in other genres that were either considered elevated or written in dactylic hexameters, or both. The other is Catullus’ corpus as a whole, with its diverse range of genres and metres. Throughout the commentary, observations on individual points of the poem’s diction, syntax, versification, and other aspects of style are made paying particular attention to these two contexts, though of course against the wider background too of what we know of the classical Latin (literary) language, and also of (literary) Greek. In this subsection of the Introduction I first outline how the poem’s style can be seen to relate to each of the two key contexts, and then offer a general characterisation of its diction (choice of words and phrases from among available options) and particularly noteworthy aspects of its versification (arrangement of words and larger syntactical structures to create hexameter verse). Details, including further references for some of the features discussed, are to be found in the commentary on individual instances: the commentary also includes further discussion of notable exceptions. Poets of the late Republic inherited a well-established Latin poetic style, especially in the elevated form associated with earlier Roman epic and tragedy. Ennius and others had already been engaged in exploring the artistic possibilities of the Latin language in poems which were fundamentally influenced by the content of Greek literature, and which soon came to be written in Greek quantitative metres.182 Catullus 64 owes much to this inheritance, especially in the use it makes of traditional poetic diction which in many cases can fairly be called ‘archaic’ (2d(i) below). But although our evidence for the earlier first century 182

For an overview of Latin poetic style see Adams–Mayer (1999a), esp. Coleman (1999); on epic at Rome up to and including Cicero see e.g. Goldberg (1995).

53

INT RODUCTION

is limited (see 1b above), we can see from the extant fragments of Cicero’s poetry that he at least was already combining the inherited Latin tradition with stylistic elements drawn more specifically from Hellenistic poetry, especially in a preference for a ‘stricter’ hexameter technique.183 Catullus 64 – and very possibly also comparable poems written by other neoterics (1b above) – introduced more Graecisms of diction and phraseology, some of them identifiably Hellenistic, but also combined Hellenistic interests in certain metrical features and in the careful arrangement of nouns and adjectives (again already of some significance for Cicero) and developed them further in Latin to produce a distinctive result which was clearly recognised as such by later Latin poets, and which was different from the direction taken by Lucretius (cf. 1c above). This Catullan (or neoteric) epic style was adopted most closely in Ciris (whenever that poem was written).184 It was one of the traditions that influenced the style of all three parts of Virgil’s corpus,185 and via Virgil had an important indirect influence on what followed; however, in style as well as in content, later Latin poets also made use of Catullus 64 directly (see 3a below). In style as in many other features (cf. 1a above), Catullus’ corpus offers both wide variety and connections among the poems which are suggestive but hard to pin down.186 Basic similarities and differences of genre and metre are of course important: in many elements of its diction, and of those aspects of versification which can be compared across metres, 64 in its epic elevation is clearly distinct from most of the shorter,

On the style of Cicero’s poetry see Ewbank (1933) 40–71, Traglia (1950), Soubiran (1972) 96–105, Courtney (20032) 149–52. Some stylistic differences can be observed between the youthful Aratea on the one hand (see 1b above) and the fragments of his later hexameter works (and translations from Homer) on the other. 184 On the style of Ciris see Lyne (1978b) 15–32 (who believes that the poem is at least post-Ovidian), Kayachev (2020) 6–15 (who dates it to c. 45–30 b c , still imitating Catullus among others). See also below, 3a(ii) and n. 288. 185 There are particularly close similarities with the style of Eclogue 4 (see Duckworth (1969) 50–2 for a metrical comparison); cf. 3a(i) below. 186 For overviews see Sheets (2007), Chahoud (2021), with further references. 183

54

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

personal poems,187 just as it is close in both respects to the Callimachean 66 and to the mythical but emotional 68b, and as it shares epic vocabulary and some techniques of repetition with 63, while being differentiated from it by the very different pace of its metre.188 The differences between the hexameters of 64 and those of 62 are subtle, and perhaps in many cases to be explained by the amoebean form of the epithalamium.189 Notwithstanding all this, however, there are some Catullan stylistic preferences which can be observed to stretch across the whole corpus, including 64 (e.g. compounds in per-, for which see 320n.). The division of the oeuvre into three groups – polymetrics (1–60), long poems (61–8), epigrams (69–116) – on the grounds of style more broadly as well as metre and length was influentially proposed by Ross (1969); although it is bound up with the controversial question of how the poems might originally have been arranged (see 4a below), and although the stylistic uniformity of the polymetrics in particular has been challenged,190 many of Ross’s observations about certain kinds of words shared by the long poems and the polymetrics but absent (or nearly absent) from the epigrams remain thoughtprovoking, and may particularly affect the way in which, as ­readers of Catullus generally, we understand 64’s use of features such as compound adjectives (52n.) or diminutives (60n.). On the other hand, 64 is of course connected to 69–116 as well as to 65–8 (and 62) by its dactylic metre, and there are elements of specifically dactylic style which are shared across all of Catullus’ dactylic poetry (such as the preference for maintaining ‘Hermann’s Bridge’), as well as o ­ thers that differentiate the epigrams from the longer elegies and the hexameter poems Sheets (2007) 208–9 offers an illuminating stylistic comparison of 64.1–24 with poem 4. 188 Fordyce (1961) 262–3. 189 See Agnesini (2007) 104–24. Poem 62 is even more end-stopped than 64, and contains proportionally many more hexameters with a line-internal pause. Its lack of spondeiazontes is less easily attributable to its amoebean nature. 190 Notably by Jocelyn (see esp. Jocelyn (1999)), arguing that lyric μέλη (including 61) should be distinguished both from iambic poems and from phalaecian hendecasyllables. See also Heyworth (2001) esp. 119, 137. 187

55

INT RODUCTION

(notably a freer approach to elision).191 A few further points of stylistic similarity across the dactylic poetry emerge from the commentary (e.g. 213n. on (g)natus, 305n. on elided initial monosyllable). (i) Diction The strand in the diction of Catullus 64 which associates it with the traditional ‘high style’ in Latin poetry can most obviously be seen in the words, forms, and phrases that can be paralleled in earlier epic, tragedy, etc.: see e.g. 1 prognatae, 23 deum genus, 79 dapem and 304 dape, 394 Mauors nn. Many of these are archaic, that is, no longer part of up-to-date Latin usage when Catullus was writing (as far as we can tell), and we can reasonably conclude that most of the archaisms in the poem that are not themselves attested in the extant fragments of earlier elevated poetry would have belonged to the same tradition and/or contributed to a similar atmosphere, especially if they seem to play a similar role in other poetry of Catullus’ time and later: see e.g. 27 diuum genitor, 174 nauita, 217 senectae nn., and cf. also 73 tempestate n.192 The same is probably true in general, too, of aspects of diction which are not so much archaic as more generally ‘poetic’, that is, where we can see from our evidence for the poetic traditions of Catullus’ time and later that these aspects were typically associated with poetry and not with prose, or with poetry of higher rather than lower genres: see e.g. 25 taedis, 50 priscis hominum … figuris (enallage), 55 necdum etiam, 118

See Hutchinson (2003) 213, with further references, Butterfield (2021) 158–65, and for Hermann’s Bridge see 2d(ii) below and n. 212. Ross (1969) 115–37 overstates the differences between the longer elegies and the epigrams, but helpfully analyses the similarities between the longer elegies and 64 in terms of the marked arrangements of adjective–noun pairs discussed in 2d(ii) below. 192 Not all archaisms, however, belong to the high style, as is clear from a study of archaism across Catullus’ corpus: see generally Heusch (1954) and for examples of apparently more colloquial archaisms in 64 e.g. 182 memet, 218 quandoquidem nn. (see also below, p. 59). 191

56

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

c­ onsanguineae, 225 lintea nn.193 In all cases, the features in question may include ­lexical choices, including categories of words such as compound adjectives (52 fluentisono n.), adjectives in -osus or -eus (12, 15nn.), or simplex pro composito verb forms (35 linquunt n.); certain usages of words not always in themselves archaic or poetic, including particular metonymies or synecdoches (e.g. 2 Neptuni, 6 puppi, 21 iugandum nn.); non-standard morphological forms or syntactical constructions (e.g. 319 custodibant, 157 qui (causal, with the indicative), 53 cum (if following classe is instrumental ablative) nn. – these are often easier to identify as specifically archaic);194 and habits of compound phrasing, typically influenced by the style of Greek poetry, such as periphrases for objects, especially the sea (6, 12, 14, 128, 179, 185), or the addition to nouns of what could potentially be ‘standing’ epithets, in imitation of Homeric formular practice (e.g. 2 liquidas … undas, 294 sollerti corde Prometheus nn.).195 The poem’s occasional use of heavy alliteration, though it can often be seen to have a specific expressive or allusive purpose (e.g. 1, 85, 132–8, 188, 201, 261–4, 278, 340nn.), also gives a flavour of earlier Latin poetry (see esp. 159, 351nn. and cf. further triple alliterations at 53, 92, 262, possibly 287).196

In some cases particular phrases shared across contemporary and later hexameter poets cause (in particular) Kroll in his commentary, Norden (19574) to go further, positing a ‘common source’ in Ennius: see 134, 166, 198, 275nn. on hexameter endings. 194 The state of Catullus’ text as transmitted means that it is difficult to form a view on the extent to which he also used phonological or orthographical archaisms – either forms that would have been archaic to him, or that were still current in his time but would seem archaic to readers of Latin today (see 4c below). It is possible that 64 originally contained more forms comparable to the instances of gnatus that I have left in the text (400 natus n.); however, in most individual cases it does not seem reasonable to ‘restore’ an archaic form by conjecture (see e.g. 104 suscepit, 153 iniecta nn.), and there are sometimes good reasons, too, for concluding that a transmitted archaic form is not what Catullus wrote (278 Peli, 320 hae n.). 195 On the further stylistic uses to which Catullus puts adjective–noun pairs, see 2d(ii) below. 196 Klein (2023) argues that Catullus uses alliteration programmatically to establish the poem’s ambivalent relationship with Ennius in particular. 193

57

INT RODUCTION

The Greek element in the poem’s diction appears most prominently in the large number of Greek proper names which help to establish the Greek mythical setting: these names are often deployed in groups (e.g. 1–3, 35–7, 52–4), and in several cases (Ariadne, Crete and Athens, the Greeks and Trojans) there is careful uariatio of names for the same person, place, or nation. Roman names for gods are in the minority (2 Neptuni, 367 Neptunia, 26 and 171 Iuppiter, 271 Aurora … Solis, 282 Fauoni, 306 and 383 Parcae, 390 Liber, 394 Mauors); more typically, divinities are referred to by Greek names, Greek titles (72 Erycina, 193 Eumenides, 251 Iacchus), or by longer periphrastic expressions, often incorporating more Greek names (e.g. 8, 95–6, 204, 228, 300, 324, 395), and such ‘learned’ periphrases, borrowed from the style of Hellenistic poetry, appear with other references too (e.g. 211, 290–1, 324, 346, 367). Greek names are often, but not always, given Greek morphology,197 and adjectival forms are also used (1 Peliaco n.). Greek words other than names include a few more general terms which were demonstrably established in the Latin poetic register (e.g. 127 and 185 pelagi, 179 ponti), while extensions of this sort of poeticising Graecism are perhaps seen in 154 leaena, 227 carbasus (see nn.). Comparably, the poem’s examples of Greek-inspired usages of Latin words include some which already appear in (extant) older Latin epic (e.g. 67 and 128 salis = ‘sea’, 125 fudisse = ‘utter’), others which do not (e.g. 9 currum = ‘ship’, 18 nutricum = ‘breasts’; cf. also 110 saeuum n.). Most of the Greek words in 64, however, refer in some sense to Greek objects or practices: these may convey the exoticism of Bacchic and other Greek cultic activity (esp. in 251–64, e.g. 197

See esp. 3 Phasidos, 11 Amphitriten, 15 Nereides, 21 Thetidi, 74 Piraei, 120 Thesei, 247 Minoidi nn. In some cases either a Greek or a Latin form is confirmed by metre. Elsewhere I look first at what is offered by the manuscripts, although I also acknowledge that scribal corruption of a Greek ending to a Latin one is in principle much more likely than the reverse (Housman (1972) 817–39); Catullus may have used more Greek forms than I print (cf. n. 194 above on possible archaic orthography). As with any textual decision, in individual cases local factors must also be taken into account (see esp. 336 Thetidi … Pelei n.).

58

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

t­ hiaso, euhoe, bacchantes, thyrsos, orgia, tympana, bombos; see also 287 choreis and cf. 20 and 141 hymenaeos) or of ‘Hellenised’ luxury (46 gaza, 49 conchyli), but also in some cases a much more down-to-earth tone, as everyday objects in this mythical Greek world are named in ordinary Greek (63 mitram, 65 strophio, 319 calathisci nn.).198 A few other features, too, interrupt or modify the generally elevated linguistic atmosphere to which most aspects of both the traditional and Hellenising strands contribute. The best known is the poem’s use of diminutive nouns and adjectives (60 ocellis n.), in which a form taken from everyday speech has become a distinctive part of the neoteric version of epic style. Noticeable also are certain lengthier connective words and phrases, some identifiably archaic (but not elevated), prosaic, or both, which emphatically articulate particular moments of narrative or speech (see esp. 56 utpote … quae, 66 omnia quae, 184 praeterea, 198 quae quoniam, 218 quandoquidem, 290 non sine nn.).199 More generally, although the poem normally avoids diction that is colloquial or prosaic,200 it contains a scattering of individual exceptions: for instance, expressions that help to give vivid immediacy to parts of Ariadne’s speech (132 sicine, 167 prope iam, 175 malus hic nn.), lexical choices associated with lower genres that contribute to the tenderness of her presentation by the narrator (87 suaues, 97 puella nn.), or phrases from less elevated registers that cast the gods in a particularly human light (27 suos … amores, 298 coniuge natisque nn.).201 Finally, the diction of Catullus 64 is characterised by a tendency towards the repetition or ‘iteration’ across the poem of words and phrases of all registers;202 the intratextual connections that these iterations create are discussed in the commentary and more generally in the Epilogue, Section 1. Such usages have more in common with e.g. 63.60 palaestra, stadio, gyminasiis, or even with e.g. 12.13 mnemosynum, 22.5 palimpsesto in poems set in Catullus’ own world. On Greek words in Catullus see esp. Oksala (1982), Sheets (2007) 197–8. 199 On these see also Williams (1968) 703–4. 200 See esp. Adams–Mayer (1999b) for a careful exploration of different kinds of ‘unpoetic’ language. 201 See also above, n. 192, on apparently more colloquial archaisms. 202 The terminology of ‘iteration’ is owed to McKie (2009) 84–92. 198

59

INT RODUCTION

(ii) Versification Most of what is noteworthy about the metrical and prosodic technique of Catullus 64 concerns the individual hexameter line, which is the most important unit in the poem’s construction.203 A large majority of the lines end with a pause in sense,204 and the result of this (or the same thing put in another way) is that very many of them consist either of a single finite clause or a single participial phrase. Many examples of the latter could be removed without affecting the completeness of the sentence that would remain (e.g. 5, 7); others describe the subject or object of the sentence in a complete noun-phrase (e.g. 1, 8, 15) or in a phrase lacking only the noun (e.g. 25, 52), so that the line again has an integrity of its own even though it is not endstopped. Lines with an appreciable internal pause in sense are rare in comparison,205 and many of these involve either a single word enjambed from the previous line ‘in rejet’ (23 heroes, 108 eruit nn.), or a pause at the main caesura (usually penthemimeral: see next paragraph), often dividing the line into two shorter units which are themselves finite clauses (34n.) or complete phrases (e.g. 4, 141, 253, 324). All of this forms a particularly clear contrast with the narrative periods and verse paragraphs of Virg. Aen. and Ov. Met. It is more typical of Cicero’s hexameter practice,206 and may owe something to the influence of late Hellenistic poets such as Bion.207 On Catullus’ hexameter technique in the context of the wider style of poem 64 see in general Cupaiuolo (1965), Michler (1982) 116–46, Coronel Ramos (1999). 204 In what follows I use general characterisations more than specific numbers, because for many of the features discussed numbers will vary according to each reader’s perception of individual cases (as well as the edition of the text that the reader is looking at), and the resulting variations can distract from more important overall conclusions. Statistics supporting the points made in this subsection are given by Ott (1973) for Catullus 64, and more generally for Latin hexameter verse by e.g. Duckworth (1969), Ceccarelli (2008), Geiger (2021). See also Kroll’s introduction to his commentary on the poem (Kroll (19292) 141–2). 205 Kroll (19292) 141 counts 55, Wilkinson (1963) 194 ‘about 40’; see n. 204 above on statistics. 206 See esp. Ewbank (1933) 57. 207 Reed (1997) 48–9. In contrast, Catullus’ elegiacs are heavily enjambed in comparison to Augustan norms (cf. n. 191 above). 203

60

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

By far the most widely used caesura in the poem is the penthemimeral (strong caesura in the third foot);208 here Catullus is following a general tendency of Latin hexameter style (as opposed to Greek) which had already been firmly established by Ennius.209 However, a noticeable minority of lines have a caesura after the ‘third trochee’ (weak caesura in the third foot),210 a rhythm which, by contrast, would have sounded Greek to Catullus’ readers, especially if this caesura was felt as the main caesura in the line;211 less subjectively, we can observe that this ‘third trochaic’ caesura is usually supported by strong caesurae in both the second and fourth feet, not so only in fourteen cases (187, 224, 232, 252, 253, 291 have only the former, 73, 152, 195, 227, 405 only the latter; 115, 141, 206 have neither). The very few lines that have no caesura at all in the third foot (18, 128, 193, 248) all again have a strong caesura in both the second and fourth. There are no lines with a weak caesura in the fourth foot – that is, ‘Hermann’s Bridge’ is maintained throughout.212 This last feature is exceptional in Latin, comparable only to Cicero’s Aratea (which breaks Hermann’s Bridge only twice) and Tibullus’ first book (once); it follows the strict practice of Callimachus and most other Hellenistic poets in their hexameters.213 Probably the best-known metrical feature of Catullus 64 is its use of spondeiazontes, on which see 3n., and 1b above for Cicero’s characterisation of such lines as typical of the

I count 365 lines (89% of 409) where such a caesura is present, together with a further two (79, 384) where it is present but ‘blurred’ by elision and accompanied by a weak caesura in the third foot (these lines are therefore included in the number given in n. 210 below), one (196) where it is blurred and accompanied by strong caesura in the second and fourth feet, and one (224) where it is blurred and where the weak caesura in the third foot is also blurred (see n. 210). 209 Skutsch (1985) 46–8; see also Goldberg (1995) 92–3. 210 I count thirty-four lines where such a caesura is present, and a further four (23b, 187, 224, 252) where it is present but ‘blurred’ by elision. 211 On this caesura in Latin see Gérard (1980); see also Lyne (1978b) 18–20, discussing Catullus 64 in some detail. 212 Hermann’s Bridge is broken once in Catullus’ longer elegies (68.49), and apparently three times in the epigrams: see Hutchinson (2003) 213–14 n. 27, with further references, and cf. above, n. 191. 213 See West (1982) 152–7, Hopkinson (1984) 51–5, Hollis (20092) 15–23. 208

61

INT RODUCTION

neoteroi.214 Most spondeiazontes in the poem end with a tetrasyllabic word, whether this is a Greek name or a verb (in only one case neither, 269 matutino; trisyllabic words, belonging to a wider range of categories, at 44, 74, 96, 252, 297). However, as in the case of caesurae, the poem’s line-endings generally follow a Latin tendency already dominant in Ennius, in which (assuming that the fifth foot is a dactyl) the last word of the line is either disyllabic or trisyllabic.215 Exceptions are discussed at 141n. (Greek tetrasyllables), 114n. (pentasyllables), 315n. (monosyllables); cf. also 298n. (hypermetre). A distinctive feature of the second half of the line is the frequency of diaeresis after a spondee in the fourth foot:216 although avoided by Callimachus and more or less infrequent in other Hellenistic hexameters,217 this had been regularly allowed by Homer and then by Ennius, but is much more common in Catullus 64, as in Cicero, whereas Virgil returns to something more like Ennian proportions.218 In particular, the poem often places a molossic word (or combination) between the penthemimeral caesura and the diaeresis at the end of the fourth foot (the ‘monumental molossus’), sometimes deploying these lines in conspicuous clusters (e.g. 1–4, 38–40, 61–5, 86–90, 186–92, 221–5). If it is correct to understand Latin versification in terms

Apart from his parodic line at Att. 7.2.1, only one spondeiazon by Cicero is extant, at Arat. 3 Soubiran Chelae, tum pectus quod cernitur Orionis, where it imitates Aratus’ own positioning of the Greek name. 215 Skutsch (1985) 49 gives a figure of 75% of such endings in Enn. Ann.; Cicero and Lucretius as well as Catullus only increased this proportion. 216 Authorities differ over whether or not a word-break in this position should be called a bucolic diaeresis, or whether the term should only be applied when the fourth foot is a dactyl and/or when there is pause in sense. In the commentary I follow the looser practice of using the term for any diaeresis at this point in the line; a strong pause in this position appears only at 23, 186, and in the Fates’ refrain, but there are many more examples of the words filling the last two feet going closely together and receiving a certain emphasis from their position after the diaeresis (see esp. 408 lumine claro n.). 217 See above, n. 213. 218 Skutsch (1985) 47 gives figures of 9.5% for Homer, 14% for Ennius, 12% for Virg. Aen. I count 208 such lines in Catullus 64, giving 51% of 409, not counting cases where the fourth foot ends with a preposition which governs the immediately following noun or adjective (14, 53, 66, 72, 112, 121, 125, 133, 151, 154, 241, 244, 271, 275, 278, 298, 329, 349, 350, 392, 398), nor cases where there is elision at this point (44, 109, 180, 184, 248, 322). 214

62

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

of the relationship between metrical ‘ictus’ and word-accent,219 this (as well as appearances of spondaic words or pairs of monosyllables in the fourth foot) creates a high proportion of fourth-foot ‘homodyne’, in which the coincidence of ictus and accent typical of the fifth and sixth feet of the line starts early, in the fourth, making lines of this shape sound noticeably ‘smooth’;220 along with the prevalent end-stopping and the maintenance of Hermann’s Bridge, this contributes to the impression of some readers that the poem’s versification is particularly regular or even ‘monotonous’.221 The overall proportion of spondees is in any case high in comparison to other Latin hexameter poets, and these prominent rhythms involving a fourth-foot spondee, as well as the spondeiazontes and the preference for opening the line with a dactyl followed by three spondees,222 perhaps make a slow regularity seem particularly characteristic of the poem. However, the kind of end-stopped, spondaic, penthemimeral stateliness observable in (for instance) 31–49 is certainly not the typical rhythm of every passage: in particular, the speeches of Ariadne and Aegeus contain a perceptibly higher proportion of enjambments and of third-foot trochaic caesurae than many of the narrative or (especially) descriptive sections, as well as fewer five-word lines (227n.), while the spondeiazontes are concentrated in lines 1–120 (nineteen examples) and 251–302 (nine), with the remaining one in the song of the Fates (358n.) and none in the speeches.223 See esp. 58n., discussing the anomalous line-ending pellit uada remis. The terminology of ‘homodyne’ and ‘heterodyne’ is owed to Jackson Knight (1939); see esp. 36–43 for his emphasis on the fourth foot. For discussion of Catullus 64 and other poems in these terms see esp. Wilkinson (1963) 127–30, Lyne (1978b) 21–3; the tables at Duckworth (1969) table ii, Ott (1973) 13 agree that the poem has a fourth-foot homodyne in about 60% of lines, the highest of any poet in Duckworth’s list except for Calpurnius Siculus (a similar percentage). 221 E.g. Wilkinson (1963) 129, referring particularly to the characteristic molossi; Duckworth (1969) 39–41, drawing attention also to Catullus’ strong preference for using only eight of the sixteen possible arrangements of dactyls and spondees in the first four feet. 222 I count 107 lines (26% of 409). See further Cupaiuolo (1965) 41–6, Duckworth (1969) 37–43, Ott (1973) 37–40, Michler (1982) 117–21. 223 For further comparisons between the metrical and other formal features of different sections of the poem, some more convincingly meaningful than others, see Cupaiuolo (1994). Various more specific expressive uses of metre are discussed in the commentary. 219 220

63

INT RODUCTION

The careful arrangement of words within individual lines, above all of pairs of nouns and adjectives in agreement, is typical of the poem. The various forms of this general technique in hexameter poetry, and its historical development, are discussed by Norden (19574) 391–8, Patzer (1955), Conrad (1965), Pearce (1966); see also Reed (1997) 49–52 on Bion and Ross (1969) 132–7 on Catullus. The separation of noun and adjective, with one element at either the beginning or end of the line and the other either just before or just after the main caesura, was one possible arrangement for Homeric noun–epithet pairs (usually formulae); it was further developed by Hellenistic poets, who reversed Homeric practice by tending to place the adjective (itself usually less formulaic or ordinary than in Homer) first in any pair, and who also began to deploy two adjective–noun pairs in the same line, and to make wider use of the possibility (very rare in Homer) of framing or enclosing the line between an adjective at the beginning and a noun at the end. Ennius mostly used the versions in which an adjective just before or after the main caesura agrees with a noun at the end of the line,224 and these are also the preferred patterns in Lucretius; but Cic. Arat. combines them with more instances both of the enclosing word order and of the double version with two adjectives and two nouns.225 Catullus 64 makes extensive use of all the possibilities so far discussed, as well as other types of hyperbaton (predominantly, still, within a single line). For the adjective and noun enclosing the line see 5n.; for the often rhyming use of an adjective before the main caesura in agreement with a noun at the end, 2n.; for the adjective just after the caesura enclosing the second half of the line in agreement with a noun at the end see e.g. 1, 4, 6, 7, 9, 10;226 and for the adjective before the caesura in This should be seen, however, against the fact that it is far more common in Enn. Ann. for noun and adjective to be simply contiguous: Skutsch (1985) 67 gives figures of 243 contiguities to 88 separations, as opposed to 200 and 207 for a sample of Virg. Aen. 225 The importance of hyperbaton for Latin poets must be linked to their strong preference for avoiding homoeoteleuton between a noun and its juxtaposed adjective: see Shackleton Bailey (1994) and cf. 122n. 226 As Conrad (1965) 206 notes, this basically Homeric pattern, used by Hellenistic poets and imported into Latin by Ennius, ‘remained a standard feature 224

64

2  ASPE CTS OF C ATULLUS 6 4 IN CO N TE XT

agreement with a dactylic noun in the fifth foot, e.g. 40, 86, 92, 133, 189.227 Most noticeable of all is the frequency of a double arrangement, either in the order adjective a, adjective b, noun B, noun A (and thus often also exemplifying the enclosing pattern) or abAB (sometimes called the ‘interlacing’ or ‘interlocking’ option).228 Often such a line also includes a single verb, which in combination with the interlaced order can produce the purest form of the ‘golden line’, ab verb AB (59n.),229 but also a version which becomes just as stereotyped in the poem, a verb bAB (7n.), and other five-word variations such as ab verb BA (e.g. 89, 314, 321).230 Although five-word lines are particularly favoured in the poem, however (1n.; cf. also 15n. on four-word lines), such patterns are by no means confined to them: they also appear in lines connected to what precedes them by a postponed conjunction (for this feature generally see 43n.), allowing the first adjective in an enclosing or interlacing pattern to come first in the line (e.g. 159, 173, 210, 293, 388), and again in lines which simply include conjunctions, prepositions, etc. in the natural place (with -que or -ue, or with a word modifying the second adjective–noun pair; this can produce the same result as postposition, as e.g. at 57, 90, 163, 242, 264, 316, 351; but cf. esp. 39–41, 63–5, and also e.g. 225, 259, 309, 344, 350). All of these techniques can be used to position words in such a way that their semantic ­similarity or contrast is particularly

of Latin epic style’, the most frequent of the possibilities under discussion in Catullus 64 as also in Cic. Arat., Lucretius, and Virg. Aen. 227 See Conrad (1965) 214–18. 228 To the more general works cited in the preceding paragraph add on this Wilkinson (1963) 215–17, Baños Baños (1992), Hoffer (2007), and in particular on the (pre)history of similar forms in Greek, Hopkinson on Call. Hymn. 6.9, with further references. 229 The history of the term ‘golden line’ has been traced back through English-language scholarship and pedagogy by Mayer (2002), (2020), Heikkinen (2015). Mayer argues that the concept itself is postclassical, Heikkinen that the apparent parody at Pers. 1.99 (see 263–4n.) indicates that it was one familiar to Latin poets by Persius’ time. It seems to me primarily important to appreciate that the ‘golden line’ is only one small part of a network of interrelated ways of organising nouns and their adjectives in Latin verse. 230 Wilkinson (1963) 216 christened this the ‘silver line’.

65

INT RODUCTION

highlighted;231 more generally, on their own small scale they exemplify and contribute to the poem’s aesthetic of balance and juxtaposition (see the Epilogue, Sections 1 and 3) as well as to the surely self-conscious refinement of its style. Soubiran connects the hyperbaton and other careful word-ordering of 64 and of Cic. Arat. with the relatively low overall rates of elision in these poems.232 Finally, a brief consideration of how the poem’s individual lines are built up into larger structures. We have seen that they can be connected by unobtrusive conjunctions, sometimes made more unobtrusive by postposition, but also (2d(i) above) occasionally by longer and more emphatic connectives; and by extending a participle into a participial phrase that fills the rest of the line (see further 5n.). Typically, these techniques maintain the endstopped or otherwise unified nature of the lines, and although a freer use of enjambment is possible (e.g. 47–9, 81–3, 178–81, 218–20),233 more often a line that has begun with a single enjambed word is then itself end-stopped or otherwise completed (e.g. 44, 92, 98, 117, 137, 150, 153; exceptions at e.g. 23, 32).234 Among the poem’s various techniques of verbal repetition,235 its Adams–Mayer (1999b) 17 ‘Paradoxically a separation (hyperbaton) may have the effect of juxtaposing (artificially by the norms of prose) two words which enter into some sort of relationship in the meaning of the line.’ Adams–Mayer here draw on the spatial ‘metaphor’ for language – which ceases to be a metaphor when speech is written down – observed by Lakoff–Johnson (1980); they suggest that on the small scale being discussed here ‘some Latin poets had, up to a point, a spatial concept of the structure of their verses’. Cf. my discussion of the ‘spatial form’ of Catullus 64 in the Epilogue, Section 1, and see also Harrison (2023) on how significant verbal juxtapositions in the poem may be ‘vertical’, operating from one line to the next, as well as ­horizontal. 232 Soubiran (1966) 600–4; see also the discussion at Lyne (1978b) 16–18, and cf. Heikkinen (2015) 73, who observes ‘almost an inverse correlation’ between the frequency of elision in the works of Latin hexameter poets and the frequency of ‘golden’ lines. 233 For the last two examples cf. above, p. 63, on the features that distinguish the style of both Ariadne’s and Aegeus’ speeches. 234 At 175–6, 318–19 a double adjective–noun hyperbaton is extended across two lines, with enjambment of the last word (only); in both cases the second line then completes the rather long sentence, with a sense of climax. 235 On repetition in Catullus see Évrard-Gillis (1976), Traina (1978), and Wills (1996) via index. 231

66

3  R ECEPT ION OF C AT ULLUS 64

characteristic epanalepses (27n.) are treated in the same way: the second line expands on the repeated word, but within a single line. Anaphora is often line-initial (19–21n.), connecting consecutive lines (also 28–9, 39–41, 63–5, 257–9; cf. 96–100, 154–6), alternating lines (9 and 11, 121 and 122, 215 and 216, if my textual decisions are right) or larger groups (387–96) into vertical ‘blocks’. Other obvious blocks of text appear in the stanzas of the Fates’ song (323–81n.) and in Ariadne’s ‘epigram’ on male infidelity (143–8n.); there are also ‘couplets’ which pivot from one section of the poem to another (50–1n.). The versification of Catullus 64, then, keeps bringing the text to a halt and inviting admiration,236 reflection, and comparison (see the Epilogue, Section 1); but it is also full of techniques for moving unpredictably on again.237

3  RECEPTION OF C ATULLUS 64 This section covers firstly the reception and influence of C ­ atullus 64 in antiquity, and secondly how the poem has been understood by scholars since the Renaissance and especially since the nineteenth century. These are also the two aspects of the poem’s afterlife with which the commentary itself is concerned. I do not address its creative reception in the modern world by writers, artists, and musicians, although there would be much to say here on material from visual art (e.g. Titian’s Bacchus and Ariadne)238 to opera (e.g. Strauss–Hofmannsthal’s Ariadne auf Naxos)239 to poetry and prose fiction (e.g. Mark Haddon’s ‘The Island’),240 and on translations into modern languages.241 Such a corpus would be dominated by Ariadne (cf. 132–201n.): attention in future research might be directed to those responses Jenkyns (1982) 102 ‘line after line presents itself individually and seems to ask, “Am I not admirable?”’ 237 Cf. on these paradoxical dynamics Williams (1968) 706. 238 See e.g. Bull (2005) 247–9, Freedman (2011) 81–6. 239 See esp. Oade (2024) 194–292. 240 Haddon (2009), republished in Haddon (2016) 19–52. 241 See Stead (2016) 99–152 on ‘Catullus 64 in translation and allusion’ in ­Romantic Britain. 236

67

INT RODUCTION

to the Ariadne myth which demonstrate their engagement with Catullus 64 specifically either in the particular details they emphasise or in the way in which they present Ariadne as an inset within a frame. (a)  Ancient Literature Catullus 64 occupies a crucial place in the Roman epic tradition, and the extent of its impact on later writers in that tradition and beyond is clear from the many discussions of particular details in the commentary. The poem’s hexameter technique can be seen to have influenced the practice of subsequent poets, and its version of epic diction to have contributed probable coinages and linguistic preferences to the development of the Latin poetic ‘high style’ (see 2d above), as well as collocations or quasi-formulaic phrases which went on to enter the repertoire of composers of Latin dactylic verse in particular.242 On a slightly larger scale, the commentary traces the influence of Catullus’ versions of a range of literary topoi and epic building blocks, such as scenes of nymphs rising from the water (12–18), the imagery of words wasted in the wind (59, 142), accusations of inhuman ­parentage (154–7), the ratifying nod of Jupiter 242

For example, certain pairings of nouns with adjectives (e.g. 51 mira … arte, 155 spumantibus … undis), or ‘formulae’ occupying the end of the line after the characteristic bucolic diaeresis (e.g. 1 uertice pinus, 183 gurgite remos, 351 pectora palmis) or associated with the beginning of the hexameter (e.g. 55 necdum etiam, 120 omnibus his, 171 Iuppiter omnipotens, 372 quare agite). As noted above (2d(i) n. 193), there are cases where we can observe the same words or phrases in the poetry of e.g. Cicero or Lucretius and speculate that they had entered the dactylic repertoire considerably earlier, perhaps with Ennius. Where Catullus provides the earliest extant examples, it is of course impossible to be sure that the elements in question did not appear in pre-Catullan poetry or in lost works by his contemporaries; but it seems reasonable to assume that not all of them did. We can also see that many words and phrases became more securely fixed in the tradition because of their use by Augustan poets, but this is not always the case: some do not recur until post-Augustan epic (276 sic tum n.), others not until late antiquity (114 labyrintheis, 243 lintea ueli, 331 languidulos; see also on e.g. 161 famularer, 173 stipendia, 263 raucisonos). This might be attributed to independent creativity (perhaps particularly with compounds and diminutives), or to appearances in lost intermediate parts of the tradition, but perhaps to later poets revisiting Catullus 64.

68

3  R ECEPT ION OF C AT ULLUS 64

(204–6), tree-catalogues (288–91), and the association of textile work with poetic c­ omposition (303–22). Certain sections of the poem had a particularly far-reaching afterlife, above all Ariadne’s speech, the Argonautic proem, and the epilogue. So did key elements of its form (cf. 2c above): its nested structure of a story within a story, its unusual deployment of narrative order, tempo, and focalisation, its foregrounding of the narrator, and some aspects of its prominent use of ecphrasis.243 This subsection complements the commentary by presenting a broadly chronological sketch of some of the most important ways in which later authors responded to Catullus 64 on all of these different levels (though with a greater focus on the larger scale). Two important methodological issues should be remembered throughout. Firstly, some of the central questions faced by every study of intertextuality come into particularly sharp focus when taking an overview of the interactions between one poem and the entire sweep of subsequent ancient literature: how (or even whether) to judge in different circumstances whether a similarity in wording or in the treatment of a theme is accidental or deliberate on the part of the later author, and how to make distinctions among different kinds of intertextual relationships, which might include (among other possibilities) a borrowing for convenience’ sake of Catullus’ way of saying something; a deferential imitation in which an author acknowledges the role played in literary history by Catullus as an important predecessor; and the kind of particularly meaningful allusion that encourages the reader of the receiving text to interpret its concerns in the light of its similarities to and differences from Catullus 64, and perhaps also to reinterpret the concerns of ­Catullus 64 in the light of its reception.244 Secondly, we have to bear in mind the complexity of the relevant literary tradition: how the r­ eception

Though in the last of these respects in particular, Catullus 64 as a whole is an extreme case with no true comparanda (2c(ii) above). 244 For how some of these issues play out when thinking about the relationship between 64 and its own predecessors, see also 1–11n. I make these kinds of distinctions, where possible and appropriate, when considering intertextual relationships in both directions throughout the commentary (cf. also Section 5 below). 243

69

INT RODUCTION

of ­Catullus 64 is mixed up with the reception of other texts. Although this is true for almost any work of classical literature, a number of factors make it particularly significant for this poem. Some of the earliest and most extensive creative receptions of 64 are by Virgil and Ovid, authors whose works immediately took a central place in the Roman literary canon: this means that we can sometimes see how later writers read the poem in combination with the ways in which it had already been received by these two poets in particular,245 but also that we may not always be confident of how much thought those later writers were giving to Catullus’ text in comparison with those of his Augustan successors. At the same time, we should also consider that ancient authors may have been responding to 64 not only alongside Virgil and Ovid, or indeed alongside those texts of (for example) Homer, Euripides, Apollonius, Theocritus, and Lucretius which are still part of our own canon, but also alongside other works now lost to us. In particular, some parts of the survey below might – or perhaps might not – be written differently if we could read and more fully trace the afterlife of poems such as Calvus’ Io or Cinna’s Smyrna (cf. 4a below),246 and thus tell a more general story about the reception of neoteric epyllion. (i)  From Virgil to Ovid The reception of Catullus 64 first becomes clearly visible just before the start of the Augustan period, in Virgil’s Eclogues.247 Virgil’s engagement with the poem throughout his career was exceptionally wide and deep.248 Eclogue 4, with its reprise of the

This may create what can be called a ‘two-tier allusion’ (Hinds (1987a) 151 n. 16) or ‘window reference’ (Thomas (1986) 188–9); in the commentary I sometimes describe an alluding author as ‘looking through’ an intermediate text to Catullus. 246 As made clear above (1b, 2a, 2c(ii), 2d), my view is that we would be more likely to find similarities in hexameter style and in some aspects of content and tone than in structure or in the role played by ecphrasis specifically. 247 For possible earlier responses in works which may or may not predate 64 see above, 1c (Lucretius), 1b (Varro Atacinus). 248 Unsurprisingly given its generic affinities and metre as well as many of its themes and forms, 64 was for Virgil a particularly important part of C ­ atullus’ 245

70

3  R ECEPT ION OF C AT ULLUS 64

refrain sung by Catullus’ Fates (327, 382–3nn.), offers in its prophecy of a future child a sustained response to the view of human history which Catullus’ epilogue on moral decline seems to cast back over the whole poem;249 this is supported by numerous allusions to the myths and diction of 64,250 and by the most consistent adoption in Virgil’s corpus of Catullus’ hexameter style.251 In Eclogue 6, the mythological part of Silenus’ song may allude in its structure, style, and tone to 64 as well as to Calvus’ Io,252 while echoes of Ariadne’s speech appear at moments of heightened erotic emotion elsewhere in the collection too (146, 154, 191nn.). The Eclogues’ exploration of Theocritean forms, especially inset songs, was influenced by Catullus’ Latin version of such a form in the song of the Fates.253 The narrative second half of Georgics 4 constitutes another extended response to 64 and possibly to neoteric epyllion more broadly: into the smallscale but in many ways Homeric narrative of Aristaeus it sets the story of Orpheus and Eurydice, narrating their ill-fated love with neoteric techniques including emotional speech and apostrophe, and evoking the themes and wording of 64 specifically in both inner and outer stories.254 But the Georgics also implicitly revisits the question of the ages of the world (12, 39, 44, 397– 408nn.); as in Eclogue 4, Virgil draws other sections of 64 besides the epilogue into his moralising discourse about the human condition (see also 31–49n.), often recasting them in Roman terms and in particular with reference to Roman civil war.255

oeuvre: see e.g. Westendorp Boerma (1958), Gonnelli (1962), Ferguson (1971), Nappa (2007). 249 On ‘optimistic’ and ‘pessimistic’ interpretations of this response see Trimble (2013) 271 n. 36. 250 For lists see Berg (1974) 208–9 nn.  33–6, Du Quesnay (1977) 97 n.  272, Trimble (2013) 268–9 nn. 26–7 and 271 n. 37. In the last of these I argue that Ecl. 4 reverses the connotations of many of the words it borrows from Catullus. 251 See above, 2d and n. 185. 252 Ecl. 6.41–63: see above, 1b and n. 25. 253 Perhaps also by the imitation of Theocritus 2 that Catullus is attested to have composed: see esp. 326, 327nn. on the refrains of Ecl. 8. 254 Crabbe (1977), Griffiths (1980), Perutelli (1980). See esp. 135, 140nn. 255 See below, this subsection and 3a(ii), for Virgil’s influence here on both Ovid and Lucan, and cf. above, 1d and n. 47.

71

INT RODUCTION

In the Aeneid, Virgil’s debt to Catullus 64 is most obvious in the story of Dido:256 her speeches allude repeatedly to Ariadne’s single speech (132–201n.),257 while the descriptions of her emotions rework elements of how Ariadne is described (e.g. 62, 86–99, 124nn.) and more generally imply a narrator who in his tendency to sympathise and empathise with his heroine is rather like the narrator of 64 (60 maestis, 100 quam tum nn.), perhaps like the narrators of other neoteric epyllia (71 a misera n.).258 One factor contributing to the moral complexity of Virgil’s account of Dido and Aeneas is the way in which these allusions evoke the moral judgements of Ariadne and Theseus made by Ariadne and the narrator in Catullus 64, and presumably by the reader too. Elsewhere, the Aeneid also demonstrates repeated engagement with Catullus’ Aegeus in its depictions of old men, fathers and sons, instructions and laments (see esp. 202–50n. and cf. 139–40, 159, 218nn.); moreover, it sometimes responds not to the characters and subjectivity of 64 but to its passages of more detached heroic narrative and metaliterary imagery (above all in the ecphrasis of the temple doors at Cumae, Aen. 6.20–34, on which see esp. 77, 113nn.; but cf. also e.g. 105–11, 267–8, 353–5nn.). Finally, while the Aeneid contains many examples of Virgil building on Catullus’ treatment of a recurring epic theme (see e.g. 194n. on describing the Furies, 357–60n. on Achilles’ fight with the river), it is notable how often this involves irony or reversal as Catullan motifs are adapted for new

See e.g. Oksala (1962), Kilroy (1969), Fourcade (1979), Clausen (1987) 40–60, Libby (2015). 257 Sometimes also to other parts of the poem (31, 339nn.); for the idea that Dido herself has read Catullus see Trimble (forthcoming) and cf. below, this subsection, on Ovid’s heroines. 258 The ‘subjective style’ in the narration of the Aeneid (Otis (1964) 41–96), although perhaps particularly influenced by the attitude that Catullus’ narrator shows (much of the time) towards Ariadne, is of course not limited to passages involving Dido. Fernandelli (2012) 475–512 ingeniously connects the subjectivity of the Virgilian narrator with the subjectivity of Aeneas, as narrator of the fall of Troy in Aeneid 2 but also as viewer and interpreter of the Trojan pictures in the Carthaginian temple in Aeneid 1, and therefore as himself like Catullus’ narrator in the tendentious way in which he focalises an ecphrasis. 256

72

3  R ECEPT ION OF C AT ULLUS 64

contexts, or transferred to different kinds of characters with contrasting fates (e.g. 156, 241, 392nn.).259 Among the other Augustans there is much less to say about Horace, whose fraught relationship with Catullus’ poetic legacy in general is well known.260 Only occasionally do similarities of phrasing between Horace’s poems and Catullus 64 begin to look like conscious imitations, let alone interpretable allusions (see perhaps 254 lymphata mente, 285 confestim, 339nn.). However, Horace’s apparently suspicious response to Catullan emotionalism includes his parody of Ariadne in the lamenting Europa of Carm. 3.27,261 while in Serm. 2.6 he satirises a luxury described in terms reminiscent of Peleus’ palace (31–49n.); on the other hand, Chiron’s prophecy to Achilles in Epode 13 makes rather more sympathetic use of Catullus’ song of the Fates (323–81, 357–60nn.), and Carm. 4.14 may even employ an echo of the song to undercut the ode’s praise of Tiberius (353–5n.).262 In love elegy, the most substantial allusion to Catullus 64 comes in Propertius 1.3: an opening simile that closely evokes Catullus’ descriptions of Ariadne (53, 133–4, 249nn.) establishes 64 as an essential intertext for the whole poem, in which the interactions between Propertius and Cynthia respond to Ariadne’s status in Catullus’ poem as both the object of a controlling gaze and an angry, speaking subject.263 Propertius also uses Catullus’ accounts of the voyages of both Theseus and the Argo in Cf. Hardie (2012) esp. 218 ‘Virgil can rerun Catullan plots in both direct and inverted forms’, and Gale (2021) 237–8 on how Virgil’s characters, even more than their Catullan predecessors, become ‘slippery’ and ‘layered’ through intertextuality. See 135, 139–40nn. for the same technique in Georgics 4, and on Eclogue 4 cf. esp. n. 250 above. 260 See e.g. Putnam (2006), McNeill (2007). 261 Buscaroli (1937), Lowrie (1997) 308–11, Putnam (2006) 111–15. Thorsen (forthcoming) points out that Europa is Ariadne’s paternal grandmother, and one might wonder if Horace, perhaps with Calvus’ Io in mind too, is commenting on the bovine element in the stories of various heroines, as other poets do elsewhere (cf. above, 1b and n. 25). In including an epiphany of Venus he may even be influenced by a detail in the myth of Ariadne that Catullus had omitted (see above, 2b(ii) and n. 110). 262 For Carm. 1.3, the propempticon that refers to the first ship, see 1–11n. and cf. below, n. 271. 263 See esp. Breed (2003), Robinson (2013), Flaschenriem (2022), with references to numerous earlier studies, and cf. esp. 200–1n. 259

73

INT RODUCTION

a sequence of elegies (3.20, 3.21, 3.22) involving male lovers leaving women (or other things) behind.264 Lygdamus actually names Catullus as the learned and sympathetic singer of Ariadne’s abandonment,265 in a context in which he seems to identify himself with the betrayed heroine, and it has recently been argued that this invites the reader to understand Lygdamus’ entire collection with reference to Catullus’ Lesbia poems as much as to 64.266 It seems plausible, in fact, that the elegiac combination of emotional soliloquy (or solipsism), mythological and literary learning, idealism contrasted with disillusionment in love, foregrounding of women as objects of erotic and aesthetic attention, and exploration of feminine voices (albeit voices frequently appropriated by men) might have some of its roots in the elegists’ appreciation of Catullus’ work as a corpus in which 64 and the personal poetry can be read in combination.267 Tibullus partakes in this, but his clearest specific interactions with Catullus 64 appear in 1.3 and 1.10, in passages about peace, agriculture, and the golden age (31–49n.),268 and in the hymnic 1.7, where the Fates have prophesied Messalla’s triumph.269 Ovid’s reception of Catullus 64 began in his love elegy. The Amores contain conspicuous allusions to a range of Catullus’ poetry,270 among which the opening lines of Am. 2.11 take a confident place: in having the poem begin not with its a­ ddressee, See further Racette-Campbell (2013), Westwood (2022), and Heyworth– Morwood on 3.20.1–2, 17–18, 3.21.17–18, 3.22.12. 265 Lygd. 6.41–2, quoted above, 1a n. 7 ; cf. 58, 132–201, 200nn. On dating Lygdamus see Navarro Antolín (1996) 3–20, Fulkerson (2017) 35–46, Heyworth (2024); although he may have been writing after Ovid, Fulkerson rightly stresses that his elegies are set in the Augustan heyday of the genre (see also Fulkerson on Lygd. 6.41–2). 266 La Bua (2019), (2020). 267 Cf. 1a above; see also Gardner (2007). For this dynamic at work specifically in Prop. 1.3, which alludes to other poems of Catullus as well as to 64, see Zetzel (1996) 83–6 with reference to Ross (1975a) 54–7. For connections between the language of Catullus 64 and the elegiac sermo amatorius see e.g. 59, 132–3 perfide … | perfide, 151 fallaci, 197 cogor, 357 testis erit nn. 268 Vaiopoulos (2019) 154–63 sees wider allusion to Catullus 64 in 1.3, with ­Tibullus casting himself as an Ariadne. 269 See esp. 259–60, 264, 303–22, 357nn. as well as several other instances of shared wording, and see also Gaisser (1971). 270 Hinds (1987b) 6–7. 264

74

3  R ECEPT ION OF C AT ULLUS 64

Corinna, but with the curse on the Argo as the inventor of seafaring, Ovid foregrounds the ways in which this elegiac propempticon engages with Catullus’ epic proem and with its epic and tragic predecessors.271 Heroides 10 is probably the earliest of Ovid’s extended reimaginings of Catullus’ Ariadne:272 as the heroine herself tells the story of her abandonment in her letter to Theseus, Catullus’ ecphrastic and narrative passages about her are all absorbed into a new, epistolary version of her desperate, pleading speech (see esp. 55, 61, 113, 126–9nn. as well as 132– 201n.).273 Critics have disagreed on whether this Ariadne, who does not curse Theseus, is a weaker, perhaps even risible character in comparison to her Catullan model,274 and a typical question for modern readers of the Heroides is raised in particular by the allusions in Her. 10 to both of the ‘endings’ to Ariadne’s story that appear in 64 (see 235n. on Theseus’ sails, 348–51n. on bacchants, and cf. 246–50n.): is Ovid alluding over his heroine’s head, or can we read Ariadne as having a metaleptic awareness of Catullus?275 When Ovid returns to Ariadne’s story in Ars amatoria 1.527–64 (supposedly as an exemplum intended to show how ‘Bacchus’, or wine, can be useful to lovers), he takes charge of narrating what happens next, after the moment depicted on

See 1–11n. and contrast Prop. 3.22, Hor. Carm. 1.3, neither of which begins with its allusions to the voyage of the Argo. 272 Catullus 64 is the main stimulus for this poem, though hardly the only one (see e.g. Jacobson (1974) 213–15, Knox (1998), Vaiopoulos (2019)). For discussion of all three of Ovid’s substantial retellings of the Ariadne myth, see Landolfi (1997), Armstrong (2006) 221–60. 273 Even in Her. 10, however, Ariadne’s actual speech is minimal (Her. 10.35–6, addressing Theseus; 56–8, addressing her bed); her Catullan speech is also drastically abbreviated at both AA 1.536–7 (132–201n.) and Fast. 3.473, 475 (132–3, 143–8nn.), and flippantly summarised in the condensed version of her story at Met. 8.169–82 (170n.). The Met. 8 passage follows almost immediately after the Scylla episode (see below, this paragraph). 274 The reading of Verducci (1985) 244–85 is particularly provocative. Among studies of Her. 10 see also esp. Schmidt (1967), Jacobson (1974) 213–27, Pavlock (1990) 113–46, Battistella (2010). 275 Cf. Barchiesi (1986) esp. 93–102 = Barchiesi (2001) 9–28 esp. 18–25, Barchiesi (1993) esp. 347, Smith (1994) esp. 251–2, Knox (1995) 18–25. Spentzou (2003) also approaches the internal authors of the Heroides as readers of earlier texts, and Fulkerson (2005) considers them as readers of each other’s letters. 271

75

INT RODUCTION

the uestis of 64, as a terrified Ariadne hears the Bacchic noise and meets her divine rescuer (see esp. 261–4n.); on the other hand, the experienced Ariadne of Fasti 3.459–516, who believes she has been abandoned again, this time by Bacchus, certainly seems to remember her previous experiences in Catullus’ poem (132–3n.). Many more Ovidian heroines also show the influence of Catullus’ Ariadne,276 especially those whose myths are similar to hers, or who have some kind of family connection: in the remaining Heroides, Phyllis (see esp. 59, 126–9, 139–40nn. and cf. 132–3n.) and Phaedra (esp. 118, 150–1, 152–3nn.), and in the Metamorphoses, Scylla (esp. 59, 132–8, 137–8, 154–7nn.) and Medea (esp. 86–7, 91–2, 98, 154–7nn.).277 In the exile poetry, Ovid even uses Catullan allusion to take on the role of the abandoned Ariadne himself (57, 70, 142, 154–7nn.).278 Catullus 64 must have been one important source of inspiration for the narrative technique of the Metamorphoses, even if here one can speculate not only about other neoteric epyllia but about the use that Ovid, like Catullus, might have made of lost Hellenistic narratives as well as extant ones.279 Ovid’s epic tells stories inset within other stories,280 but also tells parts of the same story in an unpredictable order and at an unpredictable pace, and deals with narrative space and place as freely as it does with narrative time. It follows Catullus 64 in its frequent, obtrusive use of narrative transitions (see 2c(i) above and cf. e.g. 52 namque n.), and similarly draws attention to the subjectivity of its primary narrator, and also of its various internal narrators. In particular, since Ovid picks up Catullus’ association between textiles and narrative (see esp. 312n. and 2c(ii) above), we could read such figures as the Minyeids (Met. 4.1–415) or Arachne and Minerva Cf. 132–201n. on Ariadne as the exemplary abandoned heroine, because of her appearance in Catullus 64 but also because of the nature of her story. Fulkerson (2005) 137–40 argues that the Ariadne of Her. 10 herself plays an exemplary role for the other women of the Heroides. 277 We can only guess at the use Ovid might have made of Catullus 64 in his lost Medea. 278 See further Trimble (2010). 279 Cf. above, 2a and n. 72. 280 The story of Io with the inset story of Syrinx (Met. 1.583–750) is just one early, perhaps programmatic example (cf. above, 1b and n. 25). 276

76

3  R ECEPT ION OF C AT ULLUS 64

(Met. 6.1–145), all of whom have their own agenda, as responses to the tendentious narrative choices and moral judgements made, not least as he views the uestis and constructs his e­ cphrasis, by the subjective narrator of 64.281 Like 64, too, the Metamorphoses ends by leaving freewheeling mythical narrative behind to focus on the human world as it is now, which in Ovid’s epic is straightforwardly the contemporary Roman world (contrast 399–404n.). At the end of the poem Ovid alludes to Catullus via Virgil in order to associate the wedding of Peleus and Thetis, and the Trojan War it would precipitate, with recent civil wars (37, 324, 368nn.); at the beginning, he had described the iron age in the terms of Catullus’ epilogue (397–408n.), while also connecting the end of the golden age with the sailing of Catullus’ Argo (2, 386nn.).282 This is a particularly large-scale example of Ovid’s noticeable habit of alluding to Catullus 64 in such a way as to indicate a perceived connection between one part of the poem and another: so, for instance, Thetis’ metamorphoses evoke the artistry of her bridal coverlet (50n. on Met. 11.241 uariatis saepe figuris), the bacchants rescuing Ariadne sing a wedding-song (256–64n. on AA 1.563 pars ‘hymenaee’ canunt),283 and Ariadne on her island implausibly has a ‘marriage bed’ of her own (Her. 10.51–8).284 (ii)  After Ovid The first extant author to receive Catullus 64 in combination with both Virgil and Ovid is Manilius. As typically from this point Cf. n. 258 above on subjectivity in the Aeneid as a response to Catullus 64. On the ways in which the complex temporalities of the Metamorphoses reflect those of Catullus 64, see Myers (2012) 249–54; she notes that Ovid has moved his own narrative of Peleus and Thetis (Met. 11.217–65) to the point at which it can most clearly mark the start of the poem’s transition from myth to history. 283 Ovid also makes a connection here across Catullus’ long poems to the refrains of poems 61 and 62; cf. 3a(ii) below on Sen. Med. and Stat. Silu. 1.2. 284 On Ariadne’s metaliterary bed and bedclothes in Her. 10 and how they recall her appearance on the coverlet in Catullus 64 see esp. Verducci (1985) 262–7, Barchiesi (1993) 346–7. If Ovid’s Ariadne is aware of her own literary status here, perhaps she responds to a similar self-awareness on the part of Catullus’ Ariadne (163n.). 281 282

77

INT RODUCTION

on, then, in some cases it seems safest to say that a potentially allusive similarity of expression may look to Catullus as well as to a Virgilian and/or Ovidian model (42, 405nn.); but in others, a direct reference to Catullus looks likely (3, 251nn., both cases where Manilius elegantly translates the myths of 64 to the stars).285 On a larger scale, the story of Andromeda at Manilius 5.538–618 is clearly a response to Virgil’s Aristaeus episode as a narrative digression in the last book of a didactic poem, and it retells a story from the Metamorphoses (Ov. Met. 4.668–739). But it also looks back through Virgil and Ovid to neoteric epyllion, with at least one allusion to Io,286 and with several observable echoes of Catullus 64 which show that Manilius, like Ovid, was reading the poem as a coherent whole. Andromeda is like Ariadne in being vulnerable and scantily clad on the seashore before being unexpectedly rescued: but Catullan wording as she is bound to the rock aligns her with Thetis (332n.), Prometheus (296n.), and the victims of the Minotaur (83n.), and Nereids rise from the sea to wonder at her (12–18n.) before Perseus arrives and falls in love with her at first sight (reminiscent therefore of Peleus as well as Bacchus: cf. esp. 19, 253nn.). The poems of the Appendix Vergiliana which engage significantly with Catullus 64 may also be considered here: Aetna apparently dates from some time in the mid first century ad,287 and both Culex and Ciris are predominantly considered to be post-Ovidian.288 Critical attitudes to Latin pseudepigrapha are changing, and these poems are now often read as self-conscious and sophisticated ‘fakes’ that raise questions about the Augustan (and earlier) canon.289 If this is right, then Ciris looks very much like an imitation of a neoteric epyllion; interestingly, it tells only Biondi (1981) discusses loci similes in Catullus and Manilius, but argues that they are reminiscences without true ‘arte allusiva’ (cf. n. 362 below). 286 Manil. 5.587 infelix uirgo, the narrator’s apostrophe to Andromeda. Cf. 1b above and see Uden (2011). 287 Before the eruption of Vesuvius in ad 79, which it does not mention. 288 The dating of Ciris is the most controversial, however (Kayachev (2016a) 1–7). Among others who argue for an early date, see esp. Kayachev (2020) 5–30, making a case for c. 45–30 bc . 289 See already Ross (1975b) on Culex and Moretum, Bretzigheimer (2005) on Ciris, and esp. Peirano (2012), Franklinos–Fulkerson (2020). 285

78

3  R ECEPT ION OF C AT ULLUS 64

one story, and is essentially focused on its heroine throughout, though we might see an allusion to the key structural device of Catullus 64 in the opening comparison between the poem that the poet would like to write for Messalla and the peplos offered to Minerva at the Panathenaea.290 Ciris is the extant Latin poem that is closest to 64 in style (both diction and versification), and this is a contributing factor to its large number of close verbal similarities to Catullus’ text.291 Some are in parallel contexts to those of Catullus 64, especially the heroine’s lament (132–201n.) and the description of how she fell in love (86, 93nn.). Others appear in different contexts: for instance, the narrator addresses Nisus and metamorphic birds instead of the heroes and Peleus (23–23b, 25nn.), the monstrous Scylla turns pale in fear at herself rather than as a result of desire (100n.).292 On a once prevalent view of Ciris, the latter group in particular could be read as proof of its poet’s derivativeness, even incompetence;293 but several of them do not look very different from some of the ways of alluding to 64 noted above for (in particular) Virgil, Ovid, and Manilius, especially when they suggest the poet’s perception of connections across the poem (see esp. 87–8n. on Scylla’s imagined marriage bed).294 Culex is more of a mock epic in the tradition of the Batrachomyomachia,295 and some of its allusions to 64 have a notably parodic tone (see  esp. 363n.); many of The peplos is equivalent to the temple that Virgil promises to build for the future Augustus in the proem to Georgics 3, and is similarly the subject of a brief ecphrasis (Ciris 21–34); but it is a textile. Cf. 2c(ii) above on how artworks, especially textiles, and also ecphrases of them, can be metaphors (often mises en abyme) for texts. 291 Cf. above, 2d and n. 184. For a list and formal classification see Woytek (2018) 73–89, albeit in the context of an unconvincing argument about the authorship of Ciris. 292 Further examples at e.g. 68, 122, 154, 317, 364nn. 293 See esp. Lyne (1978b) 36–47 on Ciris’ ‘method of composition’, which he describes as ‘approaching that of a cento’, and cf. further attempts to mine Ciris for ‘plagiarised’ fragments of lost neoteric poems, e.g. Thomas (1981), Knox (1983). 294 Defending the allusiveness – and quality – of Ciris, see Kayachev (2016a), Fulkerson (2020). 295 Stat. Silu. 1.praef. sed et Culicem legimus et Batrachomachiam etiam agnoscimus, nec quisquam est illustrium poetarum qui non aliquid operibus suis stilo remissiore ­praeluserit. 290

79

INT RODUCTION

them cluster in an ominous locus amoenus description with a tree-­ catalogue (Culex 108–56; see esp. 288–91, 290–1nn.), full of references to mythology, and marked in particular by imitations of Catullus’ characteristic technique of epanalepsis (132–3, 285–6, 403–4nn.).296 Aetna twice evokes the Ariadne/Theseus section of 64 in a summarising allusion, once near the beginning when listing hackneyed topics of mythological poetry (Aetna 21–2: see esp. 132–201n.), and once near the end when reviewing the mythical and literary wonders of Athens (Aetna 583–4: see 132– 3, 235nn.); 64 seems to be a key example of the type of poetry that the didactic author rejects for h ­ imself.297 Although it has been argued that Catullus came to be remembered in the imperial period primarily for the shorter and lighter poems in his output,298 significant responses to 64 are still visible in Neronian and Flavian works, mostly those of higher genre, even as they also involve themselves with much broader literary traditions. In Seneca’s Medea, the heroine borrows from Ariadne as well as other predecessors (154–7, 178, 181, 195, 197nn.), while the voyages of 64 are among those in the background of the two deeply pessimistic ‘Argonautic odes’, Medea 301–79 and 579–669 (6 ausi sunt, 248 ipse recepit nn.), and Catullus’ wedding-songs – poems 61 and 62 as well as the song of the Fates – are one influence on the chorus’ song for Jason and Creusa at Medea 56–115 (323–81n.).299 Troades also contains traces of the Fates’ prophecy of the Trojan War, such as when Pyrrhus responds to the goddesses’ summary of his father’s career (348–51n.) and when Polyxena reasserts her agency as she dies (370n.). In Phaedra, the heroine and her nurse sometimes echo 64 as they deal with both Hippolytus and Theseus (136–7, 160–3, 246–8nn.), and Hippolytus’ description of the iron age catches the tone of Catullus’ narrator in the epilogue On 64 as an important influence on Culex see also Drew (1925), Salvatore (1994) 221–35. 297 So De Vivo (2019). 298 See esp. Gaisser (1993) 7–15, (2009) 168–74, with a focus on Martial and the younger Pliny. Newlands (2021) argues that the relationship of Flavian poets to Catullus was ‘not casual’, but emphasises his importance as a model for short poems and occasional poetry. 299 See further Allendorf (2017) 122–33, 139–45. 296

80

3  R ECEPT ION OF C AT ULLUS 64

(397–408n.).300 Meanwhile, in a poetic passage of the Apocolocyntosis prophesying the life of Nero, Seneca offers a parody of Catullus in content and hexameter style (303–22n.).301 Turning to Lucan, one might see the overarching influence of Catullus 64 – both directly and via Virgil and Ovid in particular – on the narrative technique of the Bellum ciuile, whether its inconsistency and moral ambiguity,302 or the emotional involvement of its narrator.303 Lucan also follows Virgil and Ovid in looking to the poem when describing civil war: book 1 includes an adaptation of Catullus’ deserted fields (31–49, 35–7nn.) combined with allusions to his epilogue (397n.), while several passages in books 6 and 7 use the two catalogues of arrivals at Peleus’ palace (31–42, 278–302) to explore how Catullus’ mythical Thessaly has become the paradigmatic location for civil war, or all war.304 ­Lucan’s is a particularly dark reading of 64, in which, more clearly than in Ov. Met., the wedding of Peleus and Thetis leads to the horrors of Catullus’ epilogue, and the sailing of the Argo is part of the narrative of decline.305 Elsewhere, Lucan uses a negative list like those in some of Catullus’ descriptions (38–42, 63–5) to negate various details of Catullus’ wedding scene in the passage See also 397–408n. for a similar passage in the pseudo-Senecan Octavia in the voice of ‘Seneca’ himself, conceivably commenting on Catullus by suggesting how a Roman writer might describe his own present in iron-age terms. Octavia 706–7 combines reminiscences of different parts of 64 (14, 373nn.) as a character makes a cheerful reference to the wedding of Peleus and Thetis in an ominous context (cf. above, 2b(i) and n. 89 for Peleus and Thetis as a problematic exemplum). 301 Apoc. 11.6 also quotes Catullus 3.12 (making the line apply to heaven instead of hell). Seneca never mentions Catullus explicitly, nor otherwise quotes from him in his prose works (see Mazzoli (1970) 209–11), but possible allusions have been identified, most plausibly to Catullus 36 at Epist. 93.11; cf. 405n. for the suggestion that Seneca had 64 in mind at Dial. 4.9.2. 302 O’Hara (2007) 131–42. 303 Esposito (2019) esp. 788–9. Cf. 71 a misera n. for an echo of a Catullan phrase in one of Lucan’s many narratorial apostrophes. 304 See esp. Ambühl (2015) 145–77, (2016a), (2016b), discussing in particular the geographical and mythical excursus on Thessaly at Luc. 6.333–412, and consecutive passages in the aftermath of the battle of Pharsalus: Caesar’s tour of the battlefield (7.787–824), the arrival of birds and animals to eat the corpses (7.825–46, on which see also Martina (1991) 192), and the final apostrophe to Thessalia infelix (7.847–72). 305 Luc. 2.715–19, 6.400–1, both cited 1n. 300

81

INT RODUCTION

on the anti-wedding of Cato and Marcia,306 while his Cornelia sometimes sounds like Catullus’ Ariadne,307 but also, addressing Sextus, like Catullus’ Aegeus (231 condita n.).308 Among the Flavian epicists, Valerius Flaccus responds to Catullus as an Argonautic poet, evoking his language particularly in the proem to the Argonautica (1–3, 6nn.) and in other scenes in book 1 as the Argo is conceived, built, and launched (10, 11, 278nn.). The ecphrasis of pictures on the Argo (V.Fl. 1.130–48) begins with the wedding of Peleus and Thetis, albeit presented with visual and emotional details (Thetis’ unwillingness to marry) that are unlike Catullus’; their wedding feast is then connected via Catullan ecphrastic phrases (V.Fl. 1.140 parte alia, 1.148 in mediis; cf. 251, 48nn.) to the violence of the Lapiths and Centaurs, a parallel for the way in which, in Catullus, it leads to the violence of the Trojan War. Catullus 64, then, clearly plays a part in Valerius’ exploration of what the voyage of the first ship means for human history, whether he ultimately sees it as a positive development (reversing Catullus’ epilogue) or a negative one (with Catullan allusions complicating apparent optimism);309 but Valerius’ Argo is also a metapoetic artefact, a means by which the Flavian poet examines his literary relationships with his predecessors, including Catullus.310 The echoes of 64 in Silius Italicus’ Punica are more incidental for the most part, although some of them introduce Catullan overtones that are interesting in their new context in a historical epic (see esp. 12–18n. and perhaps 132, 340nn.). The works of Statius, however, include several extensive engagements with Catullus’ poem which imply reflection on its place in the Roman epic ­tradition. Among the other influences on Siluae 1.2, commemorating the wedding of Stella and Violentilla, the combination of Catullus 64 with Catullus 61 and 62 was surely important in the invention of the kind of Luc. 2.350–80; see Ambühl (2015) 375–6. Nagyillés (2009). Pompey too is at one point described like Ariadne on the shore (168n.). 308 See also Matthews (2011) on echoes of Ariadne alongside other ‘love poetry’ in Caesar’s speech to Antony at Luc. 5.476–97. 309 Davis (1989) and Ripoll (2014) are nuanced readings that ultimately come down on the side of pessimism and optimism respectively. 310 Stover (2010). 306 307

82

3  R ECEPT ION OF C AT ULLUS 64

semi-narrative hexameter epithalamium which would proliferate in late antiquity (see 323–81n. and below, this subsection, on Claudian and Sidonius).311 The poem itself borrows details from the wedding in 64 (31–49, 292–3, 318, 372–4nn.), and perhaps pays homage to Catullus’ poem in describing the bridegroom as more attractive than Peleus (Silu. 1.2.215–17), the bride more than Ariadne (Silu. 1.2.131–3), even while it sets aside the bloodier and more heroic aspects of 64 as the kind of poetry that Stella refuses to write (348n.).312 Stella and Violentilla seem to be elevated into the realm of myth, their wedding apparently set in a time like that of Peleus and Thetis, when the gods could visit mortals;313 but Statius has not forgotten Catullus’ Polyxena (364 niueos … uirginis artus n.). Statius’ Thebaid suggests that 64 can be an appropriate model for an epic episode. The wedding of the daughters of Adrastus in book 2, like Siluae 1.2, evokes the wedding of Peleus and Thetis (32–3, 276nn.); it also incorporates an ecphrasis of a rich but ominous gift (Theb. 2.265–305). The episode of Hypsipyle in book 5 includes a long flashback narrated by Hypsipyle herself, which contains an allusion to Catullus’ account of the launching of the Argo (1, 11nn.) as part of the inset rather than the frame; this heroine, at one point linked to Catullus’ Ariadne by a simile conveying stony disbelief (61n.), is, like her, eventually saved by Bacchus.314 Finally, the Achilleid – at least as we have it – pervasively explores the role played by 64 as an epic predecessor.315 Thetis, ­rising from the sea in reaction to the sailing of Paris from Sparta (12–18n.), Bitto (2016) 157–74 argues that as Silu. 1.2 is an epic epithalamium, the Achilleid (see next paragraph) is a kind of epithalamial epic, with both works responding to the link between the two genres made in Catullus 64. 312 Stella’s poetry is connected to Catullus’ shorter works at Silu. 1.2.102 hic nostrae defleuit fata columbae ; his lament for a dove (and perhaps the book named after it) is compared to Catullus 2 and 3 by Martial (1.7, 7.14). Tanner (1986) 3042–4 suggests that the part played by 64 in Silu. 1.2 constitutes a tribute to Stella’s Catullan interests. Newlands (2021) 253–8 rather emphasises its programmatic importance for Statius’ own poetics in the Siluae, characterised by Catullan doctrina and ecphrasis. 313 Vessey (1972) esp. 183–4. 314 Heslin (2016), showing how Statius alludes to 64 here in combination with Virg. Georg. 4, Culex, and Call. Hec. 315 See Lauletta (1993), Heslin (2005) esp. 69–70, 88–93, 109–14, 142–5, 261–7. 311

83

INT RODUCTION

soon compares this voyage to that of a rather Catullan Argo (1– 11n.); her complaint to Neptune is answered by his prophecy of Achilles’ future, which, up to a certain point, closely follows the prophecy of Catullus’ Fates (see esp. 344–6, 348–51, 357–60, post 361nn.), as if to remind Thetis that Achilles’ destiny is already written in the literary tradition.316 Achilles’ song to Thetis and Chiron (Ach. 1.188–94), which ends with Theseus defeating the Minotaur and then with his own ‘mother’s bed’ (Ach. 1.193 maternos … toros), is a miniature version of Catullus 64 which demonstrates an awareness of Catullus’ own ecphrastic mise en abyme (if Achilles had fully ‘sung his mother’s marriage bed’ as it appeared in Catullus, he would have told the story of Ariadne).317 Statius’ most overt allusion to 64, however, at Ach. 1.960 irrita uentosae rapiebant uerba procellae (59n.), shows Achilles’ words – this time his promises to Deidamia – as being worth no more than those of Catullus’ Theseus in the face of the same inevitable tradition. In the Latin literature of late antiquity it becomes more difficult to identify with confidence examples of direct and creative engagement with Catullus 64. Although Catullus is named and quoted by scholarly writers in the fourth century and later, it is possible that some of them were aware of him only from earlier references and quotations (see 4a below), and the question normally asked by modern scholarship of late antique poetic texts is simply whether the author knew Catullus’ works at first hand. Whereas for earlier periods it tends to be assumed as a starting point that the author would at least have had the opportunity to read Catullus, and the question asked is rather whether any potential interaction is particularly significant,318 for the poetry of late antiquity, scholarship tends rather to start with similarities (usually in wording, rarely on a larger scale) and to consider firstly whether these are more likely to be attributable to direct knowledge of Catullus rather than to a shared language and Heslin (2005) 113. Hinds (1998) 123–9, Heslin (2005) 88–93. Cf. above, 3a(i) and n. 284, on Ariadne’s metaliterary torus in Ov. Her. 10. Kozák (2016) takes the argument further with reference to Achilles’ other appearances as a singer at Stat. Ach. 1.572–83, 2.157–8. 318 See e.g. 3a(i) above on Horace. 316 317

84

3  R ECEPT ION OF C AT ULLUS 64

(where relevant) a shared metre and poetic register, or indeed to the influence of other predecessors (especially more canonical ones); allusive interpretability may then be used as another piece of evidence for familiarity with Catullus’ text.319 I discuss what I consider to be possible allusions to 64 in late antique authors at e.g. 316n. (Ausonius), 372n. (Martianus Capella), 384, 385nn. (Avienius),320 but further citations in the commentary of texts from this period are not intended to imply that I am always confident that the author was thinking of Catullus’ poem.321 There are some authors, however, for whom it is a little easier to build up a convincing picture of engagement with Catullus 64. Claudian shows evidence of small-scale allusion or imitation in several poems (75, 217, 270, 295, 359nn.), while his De raptu Proserpinae is a short (but unfinished after three books) mythological epic which foregrounds female suffering, divine domesticity, and lush visuality, and includes a purple tapestry (itself unfinished) as a mise en abyme.322 The preface to book 1 of the De raptu therefore evokes 64 in its metaphorical description of the first ship in history (not explicitly the Argo) as a model for Claudian’s epic endeavours; and it is plausible that the preface to the epithalamium for Honorius and Maria (Nupt. praef.), by describing the wedding of Peleus and Thetis, similarly acknowledges Catullus’ place in the history of the epic epithalamium.323 When Sidonius Apollinaris writes about Peleus and Thetis in the prefaces to both of his epithalamia (Carm. 10 passim, Carm. 14.26–30), he may have Catullus in mind as well as Claudian: the epithalamium for Polemius and Araneola (Carm. 15) includes a long mythological See e.g. Barwinski (1888) on Dracontius, Weyman (1888), (1889) on Paulinus of Périgueux, Pucci (2009) on Christian authors, Condorelli (2022) on authors of the sixth century. 320 Both Ausonius (Praef. 4) and Martianus (3.229) name Catullus elsewhere in their works as they directly quote his shorter poems. Realistically we can infer very little about which of Catullus’ poems might have been circulating together at different stages of late antiquity (see 4a below). 321 Nuzzo includes many more loci similes from late antique literature: see his full index of passages cited. Cf. also above, n. 242. 322 See above, 2c(ii) and n. 175, and 49n. 323 See above, this subsection, on Stat. Silu. 1.2, and cf. 323–81n. Specific allusions to Catullus 64 in Claud. Nupt. praef. are noted at 344, 357–60nn.; it also mentions Pelion in the first line. 319

85

INT RODUCTION

ecphrasis of the bride’s weaving, and ends with the Fates uniting the ‘life-threads’ of the couple in language reminiscent of Catullus.324 Prudentius, too, develops a familiar theme in the reception of 64, the repurposing of Catullus’ epilogue for a new world view: Cathemerinon 3 associates the evils of the epilogue with the time after the fall of Adam and Eve (405n.), and Cathemerinon 11 has the Christian God reversing them by ‘deigning to visit’ the world in the incarnation of Christ (407n.).325 Finally, there is the possibility that imperial Greek literature also responds to Catullus 64. The position that Greek writers could and did engage in the creative reception of Latin literature, once controversial, is now becoming preponderant,326 and for imperial Greek interest in Catullus specifically, one can cite the anecdote at Aulus Gellius 19.9.7 about ‘some Greeks’ (Graeci plusculi) living in Rome in the second century who deprecate Latin poetry in comparison to Greek and say that no Latin poet has written anything to compare with Anacreon nisi Catullus … forte pauca et Caluus itidem pauca.327 Cases have recently been made for engagement with Catullus’ Ariadne in Alciphron’s Letters of Courtesans 4.19,328 in Chariton’s Callirhoe,329 and in Philostratus’ Imagines 1.15, which has been read as a ‘correction’

Sidon. Carm. 15.201 fuluaque concordes iunxerunt fila sorores, a ‘golden’ line (ignoring -que) which perhaps borrows in particular from 64.335 coniunxit, 336 concordia. On this poem see Onorato (2020) and cf. also 320 pernentes uellera n. 325 See Heinz (2007) 41–6, 49 n. 108 (cf. also Heinz (2007) 148–51 on allusions in Cath. 3 to the invocation of Hymen in Catullus 61), O’Daly (2012) 109, 338, Richardson on Cath. 3 stanza 27. Cf. 168n. on the possible presence of Catullus’ description of Ariadne behind Prudentius’ reference to her at Contra Symmachum 1.135–44. 326 See esp. Jolowicz (2021), Verhelst–Scheijnen (2022), Carvounis–Papaioannou–Scafoglio (2023); more sceptically, Agosti (2020). 327 Jolowicz (2021) 19–20, Jolowicz (forthcoming), expressing due caution especially about what might be implied as to which or how many poems of Catullus these Greeks know. Jolowicz (forthcoming) also summarises what has been suggested elsewhere about Greek awareness of poems of Catullus other than 64. For the ancient testimonia to Catullus see above, 1a and n. 2. 328 Schoess (2018b) 99–102; the letter-writer contrasts herself with Ariadne in some detail. 329 Jolowicz (forthcoming), focusing especially on Callirhoe 1.8, the passage in which the heroine wakes up after having accidentally been buried alive. 324

86

3  R ECEPT ION OF C AT ULLUS 64

of Catullus’ ‘disobedient’ ecphrastic technique.330 In particular, Nonnus’ narrative of Dionysus’ rescue of Ariadne in Dionysiaca 47 contains a number of suggestive similarities with Catullus’ version of the story, especially in Ariadne’s speech (132–201n.), as well as points of difference which would be more interesting if they were responses to Catullus (e.g. 149, 164–70, 185, 261–4nn.). It remains possible in this and other cases that Catullus and the Greek authors are looking to a Hellenistic poem on Ariadne which is no longer extant (cf. 2a and 2b(ii) above);331 even if proposing a lost Hellenistic source for every similarity between Latin and later Greek writers has become a cliché, the possibility can never be entirely excluded, and for small-scale linguistic parallels it may seem the more likely explanation (20 hymenaeos, 28 Nereine nn.). Moreover, once the possibility of Latin influence on Greek writers is admitted, it might be as difficult as in some of the Latin examples discussed earlier in this subsection to disentangle Catullan influence from that of other authors, especially Ovid where Ariadne is concerned.332 However, the commentary offers further opportunities for exploring possible relationships between Catullus 64 and other imperial Greek texts, including several that do not involve Ariadne: Quintus Smyrnaeus’ account of Polyxena (362–70, 364nn.), Musaeus’ opening description of worshippers gathering for the festival of Aphrodite (31–49, 38–42nn.), Colluthus’ version of the wedding of Peleus and Thetis (265–302n.).

Iff-Noël (2021), arguing that Philostratus responds with a particularly anti-­ illusionistic ecphrasis, ‘silencing’ the articulate Ariadne of 64 and thus ­staging a refusal to engage with Latin literature. On Imag. 1.15 cf. 58, 122, 251–64, 261–4nn. 331 This is proposed for Catullus and Nonnus by Maass (1889) 527–9, Castiglioni (1908) 31–52, Barigazzi (1963) 445–54, and supported by Knox (1998) (though for scepticism of Barigazzi’s identification of Euphorion’s Dionysus as the poem in question, see Knox (1998) 77 n. 24, Lightfoot (1999) 59 n. 181). Herrera Montero (1996) argues for the direct influence of Catullus on Nonnus; Schoess (2018a) 153–225 assumes in her discussion of Nonnus’ Ariadne that Nonnus could have made use of Latin texts, including Catullus 64. 332 Schoess (2018b) and Jolowicz (forthcoming) both mention Ov. Her. 10 and Fast. 3 alongside Catullus. D’Ippolito (1964) 115–30 argues that Nonnus knew Catullus’ Ariadne only via Ovid. 330

87

INT RODUCTION

(b)  Modern Scholarship The modern scholarly reception of Catullus 64 began in editions and commentaries. The accounts in Gaisser (1993) and Thomson (1997) 43–56 of scholarship on Catullus in the Renaissance, and in the seventeenth and eighteenth centuries, show how the text of the poems began to be published with ­accompanying notes, and how those notes began to address questions not just of textual criticism but of interpretation.333 ­Catullus 64 offered scholars such as Parthenius, Guarinus, Muretus, Statius, and Scaliger rich opportunities for glossing unusual words, explaining mythical references, and citing parallel passages from Latin and Greek authors, but also for identifying the tone of a passage, describing the effect of a simile, or expressing an opinion on questions of character and plot such as Theseus’ attitude to Ariadne.334 Similar concerns occupied scholars who published commentaries on 64 alone, from Realinus (1551) to Mitscherlich (1786).335 Early modern work on the poem, then, taking the typical form of the period, atomised the text in order to focus on details,336 and many of the questions that it raised have continued to set part of the agenda for commentaries on Catullus 64 ever since, including this one.337 But this approach also chimed with certain ways of reading 64 that the text itself encourages,338 and that can be seen in later scholarship other than commentaries: on the one hand, identifying contradictions or inconsistencies between different parts of the poem, and on the other, appreciating individual passages for their ingenuity The banality and absurdity of the comments found in manuscript O on the first few lines of 64 (e.g. on 1 ‘narrat hic historiam aurei velleris’, on 5 Colchis ‘insula est’; see 4a below) show how badly accurate annotation was needed. 334 See esp. Gaisser (1993) 161–2 on Muretus’ treatment of the poem. 335 For more on Realinus and on two other manuscript commentaries on 64 from the late sixteenth century, see Gaisser (1992) 286–92. Despite its title, Mitscherlich (1786) is for the most part a commentary on Catullus 64, with a few pages of notes on Propertius at the end. 336 On this tendency of the commentary form see esp. Gibson–Kraus (2002), index s.v. ‘segmentation’. 337 For a perceptive characterisation of the key commentaries on Catullus from the later nineteenth and early twentieth century (Ellis, Baehrens, Friedrich, Kroll, Lenchantin de Gubernatis) see Levens (1954) 285–7. 338 Cf. esp. the Epilogue, Section 1. 333

88

3  R ECEPT ION OF C AT ULLUS 64

or their beauty. At the same time, early scholarship was also trying to come to terms with the poem as a whole, especially how it should be classified or what title it should be given: Realinus and Muretus both begin their comments on the poem by arguing against the title Argonautica suggested in the manuscripts (1–11n.) and in favour of a title naming Peleus and Thetis instead,339 while Scaliger and Voss thought it was best described as an epithalamium.340 In nineteenth-century classical scholarship, and especially in the German Altertumswissenschaft of the period, the issue of how to reconcile the various parts of 64 with an understanding of it as a whole was dominated by the question of whether it had had a Greek model or models: see 2a above for the work of Merkel (1837), Riese (1866), and Reitzenstein (1900), and, beyond the German nineteenth century, Pasquali (1920). A new critical concept which might classify the poem had also been introduced by Haupt (1855): the possibility that it should be understood with reference to Hellenistic predecessors and Roman contemporaries as an ‘epyllion’.341 Scholars wondered whether either of these ideas could explain or excuse the poem’s ‘faults’ of logic, or the disproportion of its two stories, or whether it might be more important to forget such worries and appreciate its individual delights.342 As literary criticism became a more overt concern in the early twentieth century, classical scholars expressed opinions about how 64 should be valued in comparison with the rest of Catullus’ poetry, or rather in comparison with their idea of the ‘real’ Catullus himself; the question of the derivativeness or

Realinus (1551) f. 1r (nuptiae Pelei et Thetidis), Muretus (1554) f. 89r (carmen de nuptiis Pelei et Thetidos). 340 See Masciadri (2012) 11–15 and cf. 2a above and 323–81n. 341 Cf. 1b above, and for the use of ‘epyllion’ in the nineteenth century Wolff (1988), Tilg (2012). Tilg (2012) 45 n. 37 calls Haupt (1855) the ‘turning point’ in the application of the word to Catullus 64. 342 See e.g. Haupt (1855) 7–9, arguing that the poem is no less ‘praeclarum atque admirabile’ for the presence of problems such as Theseus’ ship apparently preceding the Argo (53 cum classe n.), and Baehrens’ summing-up at the end of his commentary on the poem (i i .451–2), in which he describes how he thinks readers will have experienced its ‘virtutes’ and ‘vitia’. 339

89

INT RODUCTION

­ riginality of the poem was often seen in personal terms. So for o Pasquali ‘egli in questo carme non è il vero Catullo, quello dei canti d’amore’; for Havelock, valorising Catullus’ ‘lyric genius’, 64 has little ‘significance as a part of his own poetry’ because it ‘is read for the emotional episodes and semi-lyrical passages that it contains … These are strung together with a minimum of hasty narrative into an ill-assorted series.’343 Wheeler devoted one of his Sather Lectures on Catullus to the poem, but he ultimately finds its ‘Graeco-Roman’ learning and complexity a bit much for his taste.344 It was becoming possible, however, to understand 64 too as a personal poem of Catullus. Wilamowitz’s short discussion in his Hellenistische Dichtung ends by reading the epilogue as the dying Catullus’ personal reaction to the political chaos of his times, from which he has briefly been able to escape into a world of Greek art.345 Ramain looks for ‘une idée dominante’ that will explain the whole poem:346 quoting the view of the British philosopher Shadworth Hodgson (incorporated by Ellis into his introduction to 64) that the poem’s ‘theme’ is ‘the glory of marriage’ as demonstrated in both of its stories,347 Ramain argues along the same lines that this ‘idée’ is to be found in the contrast between the conjugal love of Peleus and Thetis and the illicit love of Ariadne and Theseus, and that this moralising contrast was inspired by Catullus’ own experience; ‘qui a été plus que l’amant de Lesbie une victime de la passion?’348 Morpurgo and Perrotta both go further, arguing that Catullus’ originality in this poem is sealed by the way in which he puts his own feelings about Lesbia into Ariadne’s mouth.349 Klingner also

Pasquali (1920) 18 (‘in this poem he is not the real Catullus, the Catullus of the love lyrics’); Havelock (1939) 77. 344 Wheeler (1934) 151 ‘There is in the poem a nescio quid nimii, as Ellis vaguely feels’. 345 Wilamowitz (1924) i i .304. 346 Ramain (1922) 140. 347 Hodgson (1870) I.535-6; Ellis (18892) 280-1 and n. 4. 348 Ramain (1922) 152 (‘who more than the lover of Lesbia was a victim of ­passion?’). 349 Morpurgo (1927) esp. 334–5 (though at p. 337 Ariadne swept off by Bacchus now represents Lesbia herself); Perrotta (1931) 400–9 esp. 407 ‘Catullo, abbandonato da Lesbia, parla in persona di Arianna.’ 343

90

3  R ECEPT ION OF C AT ULLUS 64

thought that Catullus’ life was an important influence on 64,350 but his monograph of 1956, a watershed in the scholarship on the poem, puts more emphasis on the complex formal balance between its two stories; for Klingner this is what guarantees the poem’s unity, and it is against the background of that unity that its variety can be enjoyed.351 Critical literature on Catullus 64 proliferated from the mid twentieth century onwards, and the following overview will highlight only the best-known examples of the most influential trends.352 Commentaries in this period, too, began to take account more explicitly (some more than others) of the work that was being published in journal articles, book chapters, and the few monographs on 64,353 and my own commentary attempts to do this wherever possible. Scholarship continued to address questions of textual criticism, style, metre, and individual points of interpretation, but in the rest of this subsection I will focus on overarching literary discussions of the poem and the contexts from which they emerged; although I think that it is important to examine their arguments against the text of the poem from the smallest scale upwards, I am particularly aware that a lemmatised commentary inevitably fragments such contributions just as it does the primary text. Klingner (19614) 218–38, esp. 228–30, focusing not on Ariadne but on Thetis as a reflection of Catullus’ idealisation of Lesbia as a ‘goddess’ in poem 68b (cf. 27, 334–6nn.). 351 Cf. the Epilogue, n. 2. For further discussion of Klingner (1956) see Schmale (2004) 22–3, Fernandelli (2012) xviii–xxiv. 352 For fuller accounts see Nuzzo (2003) 22–32, Fernandelli (2012) xvii–xxxiv, and esp. Schmale (2004) 17–43, supplemented by Skinner (2015) 282–98 for the period 1985–2015. Harrauer (1979) 81–90 and Holoka (1985) 210– 22 provide unannotated lists for the Renaissance to 1977 and 1880–1983 respectively. Gaisser (2007) is helpful for situating the scholarly reception of 64 in the second half of the twentieth century in the context of broader developments in Catullan criticism and academic culture. 353 See Fordyce, Quinn, Godwin, Thomson, the study or running commentary of Syndikus (1990), and now Fo. Each of these scholars covers 64 as part of their work on all of Catullus’ poetry (or most of it, in the notorious case of Fordyce); Syndikus and Godwin include 64 in a volume on poems 61–8, but publish other volumes on the shorter poems (Syndikus (1984), (1987), Godwin (1999)). Nuzzo is important as a book-length scholarly commentary on 64 alone. 350

91

INT RODUCTION

In the 1960s and 1970s, British and especially American scholars influenced by New Criticism identified the overall message of Catullus 64 as a pessimistic judgement on the state of humanity. Curran and Bramble accepted the seriousness of the epilogue (382–408n.), but saw irony or ambiguity in much of what might seem more positive or celebratory in the poem,354 especially its account of the heroic age, which, they argued, is revealed to have been ‘not so very different’ from the present; ‘it was never any better’.355 As in the earlier twentieth century, some suggested a personal explanation for this pessimism – most famously Putnam, who held that the poem reflects Catullus’ experience not only of the loss of Lesbia but also of the death of his brother.356 On the other hand, Konstan, more like Wilamowitz, made the case for the poem as a political statement on Catullus’ part, his ‘indictment of Rome’.357 These readings provoked two main strands of reaction. Some scholars attacked the pessimistic reading as anachronistic: Giangrande maintained that the poem should be read as an uncomplicated celebration of heroism,358 Dee and Courtney that it lacks any very serious ethical involvement on Catullus’ part.359 Others, notably Jenkyns and later Fitzgerald, returned to the theme of readerly enjoyment in order to argue that what is most important about the poem is precisely its beautiful surface, which ought to be appreciated in a spirit of aesthetic detachment;360 as Jenkyns notes, this was already a line of argument in French criticism.361 For Kinsey (1965), however, the ‘irony’ of his title undermines the poem’s seriousness; on irony see also esp. the Epilogue, n. 55. 355 Bramble (1970) 41; Curran (1969) 192. 356 Putnam (1961). See also esp. Harkins (1959), who uses the term ‘autoallegory’, Thomson (1961), Forsyth (1976). 357 Konstan (1977). 358 Giangrande (1972); his subsequent altercation with Wiseman in the pages of Liverpool Classical Monthly can be followed in Wiseman (1977), Giangrande (1977), Wiseman (1978), and Giangrande’s last word was Giangrande (1981). Cf. esp. 251–64n. 359 Dee (1982), Courtney (1990). 360 Jenkyns (1982) 85–150, Fitzgerald (1995) 140–68; Jenkyns uses the concept of dandyism, Fitzgerald that of the consumer. See however the Epilogue, n. 36, and e.g. 294–7, 362–70nn. for nuances in their approach. 361 Jenkyns (1982) 93 and n.  8 cites Boucher (1956), who in arguing for the poem’s ‘sensibilité esthétique’ (p. 201) is reacting to Ramain (1922) among 354

92

3  R ECEPT ION OF C AT ULLUS 64

Other work of the later twentieth century focused not on the overall purpose or tone of 64 but on its formal aspects, applying – or pioneering – scholarly developments in several key areas. The poem’s intertextual techniques were analysed as ‘allusivitá’ by Traina,362 and as polemical ‘reference’ by Thomas.363 The effects of its relationships with particular predecessors were explored in more detail: the importance of Hellenistic poetry was revisited, especially Apollonius’ Argonautica,364 but other articles argued for the importance of the Iliad,365 the Hesiodic Catalogue of Women,366 and Roman comedy.367 The structure of the poem, a recurring theme in twentieth-century scholarship,368 was analysed in terms of ring-composition or of two complementary ‘movements’,369 while Duban’s discussion of the repeated words and images that help to bind it together pointed towards later studies of its intratextuality.370 O’Connell examined its imagery and ‘pictorialism’ both within and beyond the ecphrasis,371 and its unusual ecphrastic technique was explored by several scholars, Laird being the best known.372 In his monograph on 64 and other Latin ‘epyllia’, Perutelli revisited the question of how meaning arises from the relationships between different parts of the poem, but paid more attention to form and genre: the descriptive mode of the ecphrasis, and the discursive mode of the epilogue.373 others; see also especially Martin (1977), for whom the poem is ‘une œuvre d’art pur, sans moralité ni signification’, ‘a pure work of art, without either morality or meaning’ (p. 93). 362 Traina (1972), in the tradition of Pasquali (1942) on ‘l’arte allusiva’. 363 Thomas (1982): see 1–11n. and for ‘reference’ see esp. Thomas (1986). 364 Clare (1996), DeBrohun (2007); see also above, 2c(i) n. 139. 365 Stoevesandt (1994–5). 366 Pontani (2000). 367 Reitz (2002). 368 Appearing in the titles, and used in the thematic arguments, of Murley (1937), Kinsey (1965), Bramble (1970). 369 Traill (1981), Warden (1998); cf. above, 2c(i) n.  126. Cf. also Cupaiuolo (1994). 370 Duban (1980); see esp. the Epilogue, pp. 740–1. Theodorakopoulos (2000) appears in a collected volume on intratextuality. 371 O’Connell (1977). 372 Laird (1993); cf. esp. Deroux (1986), Dyer (1994), Landolfi (1998). 373 Perutelli (1979), in chapters on ‘l’inversione speculare’ (32–43 = Perutelli (1978)) and ‘il commento separato’ (44–68).

93

INT RODUCTION

At the turn of the twenty-first century, issues of meaning and form were brought together in particular by the investigations of Gaisser and Theodorakopoulos, who both use the metaphor of the Labyrinth to describe 64 as characterised by a thoroughgoing indeterminacy: for them, the poem’s narrative, chronological, and intertextual complexities baffle the reader and call into question the validity of any moral perspective.374 The role played in this confusion by an unreliable narrator is emphasised by O’Hara, by Oliensis (with a Freudian twist), and especially by Schmale in her monograph of 2004;375 the narrator’s limitations are explored with further subtlety by Fernandelli in his 2012 book, an all-encompassing study of Catullus 64 which also makes particular contributions on the poem’s narrative technique and intertextuality with its predecessors, and shows how it aims to be a totalising masterpiece.376 Meanwhile, new strands of criticism are reflecting a more general turn in classical scholarship back towards ethical and political concerns. Gender in the poem can be approached from feminist perspectives of various kinds,377 while its foregrounding of acquisitive travel and indulgent spectatorship can be seen to reflect Roman imperial attitudes to Greek culture, in which not just the characters and the narrator, but also the poet and the reader, might be implicated.378

4  TEXT OF C ATULLUS 64 (a)  Publication and Transmission It seems plausible that 64 might originally have been ‘published’ and circulated by itself, in a single papyrus book-roll.379 There is Gaisser (1995), Theodorakopoulos (2000). O’Hara (2007) 33–54, Oliensis (2009), Schmale (2004); for my response see 1a above and the Epilogue, n. 47. 376 Fernandelli (2012); cf. the Epilogue, p. 737 and n. 21. 377 E.g. Panoussi (2003) and the Kristevan readings of gender and time in 64 by Gardner (2007), Seider (2020). 378 See 1d above on Feldherr (2007), Dufallo (2010), (2013) 39–73, Nelis (2012), Young (2015) 24–51. The ‘confident consumer’ of Fitzgerald (1995) 167–8 is becoming a more anxious one. 379 So also e.g. Butrica (2007) 19–22, Du Quesnay (2021) 174. 374 375

94

4   TEXT OF C ATULLUS 6 4

no direct evidence – no ancient author refers to the poem either by (any) title or as part of a book or collection – but poem 95 implies individual publication for Cinna’s Smyrna, and both Smyrna and Calvus’ Io are named as individual works by ancient commentators and grammarians.380 Quintilian’s citation at Inst. 9.3.16 of 62.45 as Catullus in epithalamio (or Epithalamio ?) suggests that the much shorter 62 circulated as a unit identifiable in itself, while extant Roman poems closer in length to 64 for which independent circulation also seems likely include Horace, Ars poetica (476 lines), Ovid, Tristia 2 (578 lines), Ciris (541 lines), Culex (414 lines). The connections of form and content that tie 64 to the other ‘long poems’ of Catullus (see 1a above) need not mean that it ever formed part of a libellus consisting of, for instance, 61–4 or 61–8, although cases have been made:381 at some point, however, it came to rest in its current position within the single (though not complete)382 collection of Catullus’ poems that was transmitted from the end of antiquity to the beginning of the Renaissance, which is essentially the same collection that we read today. I am sympathetic to Butrica’s theory that this was put together by a single reader in late antiquity, who created a codex-length collection that incorporated traces (or more) of earlier papyrus libelli, probably including some that had been put together by Catullus himself, as well as short individual poems and fragments, and longer, originally independent poems including 64, all thoughtfully arranged as far as the nature of the poems and libelli allowed (no arrangement could be perfect).383 Servius Danielis cites lines from both Caluus in Io (Calv. frr. 9, 13 Courtney = 20, 24 Hollis) and Cinna in Smyrna (Cinn. fr.  6 Courtney = 10 Hollis); the former attribution also appears in [Probus] (Calv. frr. 10, 12 Courtney = 21, 23 Hollis) and the latter in Charisius (Cinn. fr. 8 Courtney = 8 Hollis) and Priscian (Cinn. fr. 7 Courtney = 9 Hollis). See 1b above for these other neoteric ‘epyllia’. 381 See esp. the versions of the ‘three-book theory’ presented by Baehrens (1876–85) ii.57–61, Quinn (1972) 9–20, Schafer (2020). On the role that Catullus himself might have played in the arrangement of his poems as transmitted, or in any part of it, see Skinner (2007c). 382 See Butrica (2007) 17–19, Du Quesnay (2021) 167–8, with further references, and cf. post 235, 327nn. 383 Butrica (2007) 23–4; cf. Du Quesnay (2021), esp. 174–5, 217–18. 380

95

INT RODUCTION

Scholars and grammarians in late antiquity quote 64 among other poems of Catullus: see app. crit., and for more on Nonius, Macrobius, the Verona scholia, Aelius Festus Apthonius, and Isidore of Seville, Zetzel (2018). Some of these citations may perhaps have been at second hand: in Macrobius we can certainly see isolated lines of Catullus’ short epic entering the tradition of commentary on Virgil, just as lines of Io and Smyrna did (above, n.  380; cf. also 114–15n. for Servius on Virg. Aen. 5.591). The few extant indications that anyone was reading Catullus in the medieval period, however, do not include any direct references to 64;384 among potential allusions to Catullus in medieval texts, one or two have been connected to the poem, but they are among the least convincing.385 The surest evidence for knowledge of Catullus in this period is the statement made by the bishop Rather of Verona, in a sermon given in ad 965 or 966, that his sins of time-wasting included reading Catullum numquam antea lectum, but there is no hint as to which particular poems tempted him. Around 1300, however, apparently as a result of the emergence of a manuscript of Catullus in Verona, knowledge of Catullus began to spread among various ‘pre-­ humanists’,386 and 64 was clearly of interest to them. Geremia da Montagnone, compiling a compendium of moralising passages around 1315, excerpted Ariadne’s bitter ‘epigram’ on men’s lust and fickleness (143–8).387 Petrarch echoed the poem in his own works, and noted in the margins of his codex of Virgil and Statius some of their most blatant reworkings of Catullus’ lines

See generally Kiss (2015b) xiii–xvii, (2016) 131–8. Ullman (1960) 1034–5 examines some references to Scylla in texts from twelfth-century England. I am unconvinced by the suggestion made by Cristóbal López (1991) 375 and Arcaz Pozo (2005) 103–4 that the elegiac comedy De tribus puellis, written in France in the twelfth or early thirteenth century, alludes to the ecphrasis of Catullus 64: both mira … arte and ex alia parte (De tribus puellis 229, 235) are typical ‘ecphrastic’ phrases used by Virgil and Ovid among others (51, 251nn.), and the idea of describing a bed painted with scenes of Jupiter ‘deceiving girls’ and of Mars, Venus, and Vulcan may well have occurred to an author familiar with Ovid (as the author of De tribus puellis clearly was) without influence from Catullus. 386 See Gaisser (1993) 19–20, Kiss (2016) 138–9. 387 On Geremia and Catullus see esp. Ullman (19732) 79–112. 384 385

96

4   TEXT OF C ATULLUS 6 4

(59, 141, 171–2, 327).388 With the copying of more manuscripts of Catullus in the fourteenth century and many more in the fifteenth, 64 was read and studied alongside the rest of Catullus’ poems, both by anonymous Renaissance readers who copied and annotated manuscripts, and by identifiable scholars.389 It seems to have been printed separately as early as 1493 (in Leipzig),390 and in 1551 it was separated out again by Bernardus Realinus for individual treatment in a commentary, as occasionally since.391 (b) Manuscripts The manuscripts of Catullus on which modern editions, including this one, are based are known as O, G, and R.392 What follows is a brief account of their nature and relation to each other, intended to explain the assumptions underlying my editing of the text: for more detailed discussion, including demonstrations of the shared errors that make it possible to place each manuscript on the stemma, and accounts of the scholarly discoveries and breakthroughs in the nineteenth and twentieth centuries that gradually allowed this picture to emerge, see especially Oakley (2021), with further bibliography. On Petrarch and Catullus see e.g. Ullman (19732) 191–6, Billanovich (1997), Giazzi (2004) 111–17, Petoletti (2004), and cf. Trimble (forthcoming). The codex with Petrarch’s annotations is in the Biblioteca Ambrosiana in Milan, A 79 inf. (hence ‘Ambrosian Virgil’), and the annotations are edited in Baglio– Nebuloni Testa–Petoletti (2006). Intriguingly, Petrarch attributes two of his four citations of 64 (59 and 141) to ‘Catullus in Peplon’, and uses the phrase again in his annotation to Servius’ prefatory note on the Aeneid; we can only speculate about the origins of this apparent title, which is found nowhere else (Ullman (19732) 193 ‘Petrarch’s own invention’), though we might guess that it has something to do with the uestis in 64, perhaps also with the peplos of Ciris 21–34 (see above, 2c(ii) n. 175, 3a(ii) and n. 290). Billanovich (1997) discusses three potential allusions to 64 in Petrarch’s works, and Thomson and Catullus Online include another two as testimonia (on lines 186, 274): but see 4c below for the policy on testimonia adopted in this edition. 389 See above all Gaisser (1993) and cf. 3b above. 390 Barinus (c. 1493) (1493 is the date of the editor’s prefatory letter); see Kiss (2018) 2154, 2163. Other very early editions of 64 alone include Suchtenius (c. 1505), editio Vindobonensis (1514); on the former see Worstbrock (2013), on the latter Masciadri (2012) 12. 391 Realinus (1551). See 3b above. 392 For T see below, n. 395. 388

97

INT RODUCTION

G (so named, Germanensis, because it was formerly kept at the abbey of Saint-Germain-des-Prés; now in the Bibliothèque nationale in Paris) and R (in the Vatican Library, hence Romanus) are ‘siblings’: both descend from a hypothesised manuscript, no longer extant, which is given the designation X.393 X, in turn, was the sibling of the oldest extant manuscript, O (in the Bodleian Library in Oxford): following the work of McKie,394 most scholars now believe that O and X were copied from another lost manuscript, designated A, which had itself been copied from the single manuscript of Catullus that had emerged in Verona about 1300 (see 4a above), conventionally called V (Veronensis).395 This seems to give the best explanation of the conIn the stemma, although not elsewhere in this edition, the names of manuscripts no longer extant are placed in square brackets. 394 McKie (1977) esp. 38–95. 395 The standard stemma is economical, ignoring the hypothetical possibility that there may have been further lost manuscripts if A was not copied directly from V, or O not directly from A, etc.: but the stemmatic relationship among the extant manuscripts, which determines their value as witnesses to the archetypal text, is well established. T (Thuaneus, after its sixteenth-century 393

98

4   TEXT OF C ATULLUS 6 4

trast ­between the system of poem titles shared by OGR and that attested by G ­ eremia, who appears to divide Catullus’ text into longer ‘capitula’. There are over 120 later manuscripts of Catullus, but although no full stemma for these recentiores exists, none of them (nor any later hand in any of them) has been proven not to descend from OGR.396 Of the three primary manuscripts O is the oldest: it was copied in northern Italy probably around the third quarter of the fourteenth century.397 O is an unfinished manuscript in that it never received the titles and rubricated initials for which the scribe left room at (what he considered to be) the start of each poem.398 If O’s exemplar A contained other interlinear and marginal material, which descended via X to G and R (see next paragraph), then the scribe of O omitted it too: there are very few marginalia in O, almost all found on ff. 1r–1v (poems 1–4) and 21r (the opening of 64), and none of them is shared with GR. It is uncertain whether these marginalia, which include some variant readings, are in the hand of the original scribe (O1):399 in two places (16, 145) where the hand and ink of the marginalia make me doubt it, I indicate this with O1?. O’s text is characterised by a thick scattering of fairly straightforward mistakes not

owner Jacques Auguste de Thou) is an anthology, copied in northern France in the ninth century, which contains of Catullus’ poems only 62: it does not descend from V but is, for poem 62, an independent witness to the ultimate archetype, here called Ω. 396 For lists see Thomson (1997) 72–92, Catullus Online, Kiss (2015a) 173–7, Oakley (2021) 287–90. G. G. Biondi and his colleagues at the Centro Studi Catulliani of the Università di Parma (www.catullus.unipr.it) are more open than other scholars working today to the possible existence of manuscript witnesses to a tradition separate from that represented by OGR: see e.g. Biondi (2015) esp. 48, Agnesini (2020). 397 For descriptions of O see e.g. the unpaginated preface to the facsimile of ­Mynors (1966), Thomson (1997) 28–30, and Catullus Online, with further discussion at Kiss (2020b) 607–12; for digital images see https://digital .bodleian.ox.ac.uk or Catullus Online. 398 A historiated initial was, however, added to the first recto of O in probably the second quarter of the fifteenth century, in Lombard style. 399 Thomson (1997) 81 n. 72 cites the opinion of A. C. de la Mare that ‘there is no reason to attribute anything in O to a second hand’. But see now Kiss (2020b) 609–10, citing Irene Ceccherini, for a more confident view that the annotating hand is distinct from the main hand of O.

99

INT RODUCTION

shared by the X tradition, but this does not mean that textual choices between O on the one hand and GR on the other should be influenced by any simple assumption that in any given case O is more likely to be wrong: there are certainly places where O’s reading might be difficult to correct if we did not have GR (e.g. 332 leuia GR: uenia O), but also those where O preserves a correct reading that GR have lost entirely (e.g. 139 blanda).400 G was copied in Verona in 1375 (the scribe himself gives us the date and refers to the contemporary ruler of Verona on f. 36r). R is slightly younger, having been copied in Florence for Coluccio Salutati (1331–1406), the humanist and chancellor of Florence, between 1375 and c.  1392–4.401 Both G and R contain a few poem titles and variant readings added by the original scribes (indicated by G1, R1), and further additions and corrections by later hands. R was corrected against its exemplar X by Coluccio Salutati himself, whose hand in this manuscript is referred to as R2. As well as the majority of the poem titles (including the two in 64: see app. crit.), Salutati seems to have added further variants that were present in X (whether inherited from A or added in X as conjectures or corrections of the scribe’s own mistakes) but had been omitted by R1; he certainly also added a large number of his own conjectures, some made as corrections in the text, others in the guise of variants. One copy of R, known as m (Marcianus, in the Biblioteca Nazionale Marciana in Venice) and copied in Florence around 1400, is particularly significant because, having been copied from R after the additions of R2 and then further corrected against R + R2 by a second hand, m2,402 it was Reeve (2020) 7–8 points out that theoretically blanda could be a conjecture, but also argues that this is unlikely in O because of its date. On the characterisation of O and X outlined here, and its implications, see further Thomson (1997) 28–30 and esp. McKie (1977) 237–90. 401 For descriptions of G and R see e.g. Thomson (1997) 30–5 and Catullus Online ; on G see also Samaran–Marichal (1974) 359 and pl. c x i v and Kiss (2020b) 612–16, on R De Robertis–Fiesoli (2008) and Kiss (2020b) 616–21. For digital images of G see https://gallica.bnf.fr or Catullus Online, and for digital images of R (recently much improved) https://digi.vatlib.it. 402 See esp. McKie (1986), refuting the theory of Thomson (1973) that ‘early’ and ‘late’ phases of R2’s work can be distinguished using the various readings of m. m2 must also have seen G, as he occasionally added a reading of G to m: see Oakley (2021) 270–1. 400

100

4   TEXT OF C ATULLUS 6 4

used by a corrector of G, designated G2, as a source for a further layer of titles, corrections, and variants added to G.403 This means that it is sometimes difficult to tell what G’s original reading was before correction, and certainly, in many cases, whether any given correction was made by G1 or G2.404 However, the most important consequence of our understanding of this process is that we now know that neither m nor G2 is an independent witness to the text of Catullus. As recommended by e.g. McKie (1986) 67–8, Oakley (2021) 287, I therefore eliminate m and G2 from the apparatus except when they offer readings which do not appear in R + R2 (see further 4c below). I understand such readings of m and G2 to be conjectures, as I also do the many readings that I print in the text or apparatus whose earliest attestation is in one or more of the recentiores. There are also several later hands in R, studied in detail by Thomson,405 who names one R3 and all the others (dated to the early sixteenth century) Ra. Nothing added by R3 is visible in poem 64, but the activity of Ra is concentrated on the first 290 lines of this poem; it seems that all that the Ra hands do is to add conjectural readings found in recentiores or early printed editions, or on one occasion (at line 196) apparently not attested earlier elsewhere. There may also be a few ad-

On the theory that the main hand of R, m2, and G2 were all in fact the same scribe, see Kiss (2020b) 618–21, drawing on De Robertis in De Robertis– Fiesoli (2008) 238–9. 404 Bonnet (1877) 59–62 is recommended by Oakley (2021) 267 n. 17 as still the best guide to the observable differences between G1 and G2, all the more valuable because Bonnet wrote before the emergence of R and m and therefore uninfluenced by any view on how exactly these manuscripts related to G2’s work. Bonnet describes G1 and G2 as distinguishable in terms of how, where, and what they write: G1 as a spindly, slightly angular hand in greyish ink which adds some corrections, variants (all interlinear), and glosses which could all have come either from the exemplar of G or from conjecture, G2 as a larger, firmer, and more rounded hand in brownish ink which is responsible for most of the titles and some metrical notes as well as corrections and variants which (Bonnet realised) have every appearance of deriving from a separate manuscript. However, even Bonnet admits that in the case of most of the corrections made in the text itself (often only a stroke, or one or two letters) it is impossible to identify either G1 or G2 with confidence. 405 See esp. Thomson (1970). 403

101

INT RODUCTION

ditions to G by at least one later hand separate from G1 and G2; I cite one variant reading (line 255) as G2?. (c)  Edition and Apparatus In this edition of poem 64 I aim to present the text that, in my judgement, Catullus is most likely to have written. The work to establish this text necessarily includes the editor’s two tasks of making choices among the readings of the primary manuscripts, and considering conjectural emendations, but it is weighted towards the latter, because the text of Catullus that emerged in V and for which OGR give evidence had already suffered extensive corruption. It is clear that this corruption is wide; I share the opinion of those who believe that often it is also deep.406 I understand this to be a fairly radical edition, more open than its predecessors to accepting conjectures into the text: but readers will judge.407 My confidence that Catullus did indeed write the text that I print of course varies throughout the poem according to all the relevant evidence, discussed as appropriate in the commentary; however, I have made an editorial choice not to use obeli but, in cases such as 287 crebris, to print something that Catullus might have written rather than something that he could not have written. For the same reason, I print conjectural supplements where something is missing from the text, except in the one case where I indicate a lacuna of a whole line (121b). On the other hand, although I aim to recover Catullus’ words, I do not aim at a thoroughgoing reconstruction of how he might have spelled them.408 Instead, this Contra e.g. Goold (1958) 111 ‘the corruption is of a superficial rather than a profound nature’. 407 For the poem I am editing, I follow many of the recommendations of Harrison (2000), who argues in particular that in comparison to existing editions ‘a new text of Catullus’ should include more conjectures in both text and apparatus (but see next paragraph on the difference made to the latter by Catullus Online). It will be clear, however, that this edition is much more conservative than one produced along the same lines as Trappes-Lomax (2007) would be: reviews such as Heyworth (2008), Skinner (2008), Butterfield (2009) express a scepticism, which I share, about some of Trappes-Lomax’s guiding principles. 408 Catullus’ own texts of his poems may well have contained many more examples of ‘archaic’ orthography than are presented in this and other modern 406

102

4   TEXT OF C ATULLUS 6 4

edition aims to use ­orthography and punctuation that will present the text clearly to twenty-first-century readers familiar with contemporary conventions for printing classical Latin literature. My use of paragraphing has the same goal, and accords with the ways in which the poem is divided and subdivided in the commentary. The purpose of the apparatus is to show how textual decisions have been reached from the evidence of the primary manuscripts and the efforts of scholarly conjecture. It is not primarily intended to present evidence for the character of the manuscripts cited or the relationships among the later hands present in them (for which see 4b above), nor to make any new arguments about such issues (as well as my own interests and expertise, this decision reflects the fact that this book is not a new edition of all Catullus’ works). The nature of the apparatus has also been influenced by the scholarly context of this edition, especially the existence of Catullus Online.409 This website, created by Dániel Kiss and centred on his critical edition of the full text of Catullus, aims to include a comprehensive repertory of conjectures, and is also very thorough in reporting corrections and orthographical variations (including those of word division) in OGR and (often) m; it also offers a convenient way of accessing the high-quality digital images of O and G that have been put online by the libraries that own them (nn. 397, 401). The rest of this subsection is intended to explain how these aims and contexts are reflected in my presentation of the text and apparatus. I report fully, then, what OGR present as the text of Catullus (ignoring orthographical variation, as detailed at the end of this subsection). I have chosen to cite only extant manuscripts rather than using A (or the traditional V) to indicate the editions: but the late and corrupt manuscripts give only occasional hints of this, and any edition that aimed to reconstruct Catullus’ spelling throughout would still retain a high degree of uncertainty, while both distracting readers by its unfamiliarity and running into at least as many problems of consistency as a more typical ‘normalising’ approach. For a discussion of orthography in the manuscripts and modern editions of Catullus see Bonvicini (2012) 45–99; for the inconsistencies that remain in this edition, see the end of this subsection. 409 www.catullusonline.org; for a discussion of the aims and nature of the website, see Kiss (2020a).

103

INT RODUCTION

archetype of OGR,410 or X for the hyparchetype of GR: however, the commentary occasionally discusses what readings or variants may have been present in X, A, or V. I also report fully where O1, G1, and R1 correct their original text or add variants, except that I generally omit cases where the reading of the manuscript’s exemplar is clear, and a reference to O1, G1, or R1 would simply mean that a mistake made by O, G, or R was (or could have been) noticed and rectified immediately by the scribe, calamo currente. I report R2’s corrections and variants fully (except when these involve only word division or orthography as defined in the list below), but without explicitly indicating in the apparatus whether I believe that on any particular occasion R2 is transferring to R a reading found in X or is making a conjecture.411 I report G2 only when he is the originator of the correction or variant in question – that is, when it is not found in m + m2 or R + R2;412 I have made a fresh attempt to distinguish between G1 and G2, using the characterisation of the two hands established by Bonnet (n. 404) and my own observation, but I may not always have made the right choice. In citing O1, G1, G2, R1, and R2 I distinguish between corrections which replace the original reading and variants which supplement it (indicated by u.l. = uaria lectio). However, in the first case, I do not indicate whether the correction is written over the original reading – which may or may not have been previously erased – or is written above the line or in the margin with a clear expunging dot or omission mark in the relevant place in the text; in the second case, I do not indicate whether the variant is written above the line or in the margin, or whether it is preceded by the abbreviation for aliter, the abbreviation for uel, or nothing. Beyond corrections and variants, I also report where any of these hands add annotations that indicate their view on McKie (1977) suggests using V only when a quotation in Geremia confirms the reconstructed reading of A. I treat Geremia and other pre-OGR testimonia in a granular way, as I do O, G, and R themselves. 411 Occasionally I express a judgement on this in the commentary. Cf. the excursus on the suggested origins of R2 variants at Thomson (1997) 38–43, which is still influenced, however, by a modified version of Thomson’s original view that R2 worked on R in identifiable phases (Thomson (1997) 37–8: n. 402 above). 412 I make one exception to this rule in the case of the poem’s two titles. 410

104

4   TEXT OF C ATULLUS 6 4

textual matters (this is confined to four appearances of ‘deficit’, indicating awareness that the text is incomplete), but I do not report annotations concerning grammar, diction, etc. (e.g. on 1 prognatae O1 adds ‘i.e. natae’), except in one case where such a comment may actually disguise a variant reading (on 145 apisci).413 Nor do I report annotations on the content of the text (e.g. in the margin near line 1 O offers ‘narrat hic historiam aurei uelleris’), with one exception: titles. I have seen O, G, and R. I report other manuscripts only when they are or seem to be the originators of conjectures that I wish to include. Accordingly, except in the few cases where it helps to explain cited readings of G + G2 (at 282, 288, 315),414 I report m (and m1, m2) only when the reading in question is not found in R + R2 and I think it is a conjecture worth including.415 I have seen microfilm of m (kindly supplied from the Faculty of Classics at the University of Cambridge by Stephen Oakley). I have not examined any other manuscripts (either the originals or images),416 and when I indicate that the earliest attestation of a conjecture is in a manuscript other than O, G, R, or m – which I do with the single siglum ϛ (stigma) – I am normally following Catullus Online, to which I This is possible if the comment was indeed added by the original scribe of O, with access to the exemplar A, but presumably not if the comment is actually owed to a later hand: see above on ‘O1?’. 414 And in the case of the two titles (n. 412). 415 On the relations between R + R2, m + m2, and G2 see 4b above. 416 With two exceptions: firstly Vat. lat. 3269 (MS 110 in the lists given by Thomson, Catullus Online, and Oakley (2021)), for which I have consulted the digital images available at https://digi.vatlib.it). This manuscript was attributed to Pomponio Leto in Fulvio Orsini’s inventory of Vatican manuscripts (in Vat. lat. 7205) and hence in an annotation on f. 2r, but the attribution was categorically rejected by Muzzioli (see Muzzioli (1959) 347); I am grateful to Helen Dixon and Dániel Kiss for pointing me to Muzzioli and sharing with me their agreement that Leto neither wrote nor annotated this manuscript. It contains the following conjectures which Mynors attributes to Leto in his apparatus (cf. Mynors xi n. x ‘ipsius Laeti manu exaratus dicitur Vaticanus lat. 3269’): 104 subscepit, 120 praeoptaret, 122 placido, deuinctam. None of these conjectures appears in Bibl. Casanatense 15 (MS 91), which was written by Leto. (This account of Leto’s relationship to both MS 91 and MS 110 was set out in Muzzioli (1948): but an editorial note on the first page of Muzzioli (1959) describes the earlier piece as having been published in a limited number of copies and ‘rimasto quasi ignorato’, and Mynors seems not to have known it.) My thanks again to Stephen Oakley for sending me an image of MS 91. 413

105

INT RODUCTION

refer readers interested in discovering exactly which manuscript or manuscripts contain the conjecture. I attempt to attribute all conjectures to the earliest person known to have made them. Here I also take Catullus Online as my primary guide; if I have come across an earlier attestation of a conjecture, I cite that instead, but I have not systematically checked all of Catullus Online’s data on this point. I therefore attribute a conjecture to two sources only when priority seems to be genuinely uncertain (this is most often the case where conjectures appear both in recentiores and in early editions). If two scholars made the same conjecture independently but one is known to be later than the other, only the earlier scholar is credited. The arrangement of conjectures in the apparatus aims to be logical rather than chronological: similar ideas are presented together, and connections are implicitly drawn between conjectures that influenced or could have influenced each other. Conjectures other than those found in manuscripts or in a few early editions are simply cited with scholars’ names. Many come from the editions listed in Section 2 of the Bibliography, where they can be found ad loc. either in the text, apparatus, commentary, or textual notes; others come from the publications listed in Section 3, which is clearly annotated. Those where the scholar’s name appears with an asterisk are previously unpublished, and I am grateful to their originators (Stephen Heyworth, Stephen Harrison, Stephen Oakley, Peter Thonemann) for allowing me to include them. I do not normally indicate in the apparatus whether conjectures were made in an editor’s text, apparatus, or commentary, confidently, doubtfully, or even by mistake, or whether they were later withdrawn; such cases are sometimes discussed in the commentary, and much more information is given by Catullus Online. Testimonia that predate OGR are given in the apparatus in brief form (see Catullus Online for fuller versions). The text of Catullus that they give is quoted only if it differs from that of OGR (but not if the differences are only orthographical, as d ­ efined below). For this purpose I count as testimonia only quotations from Catullus, not even the closest of allusions: ancient allusions to Catullus are frequently discussed in the commentary, but Petrarch’s are not.417 417

See 4a above and n. 388.

106

4   TEXT OF C ATULLUS 6 4

As mentioned earlier in this subsection, this edition does not aim to reconstruct Catullus’ orthography. The general principle operating throughout the text and apparatus is that the word itself is much more important than its spelling. I therefore aim to ignore orthographical differences (defined below) among the manuscripts, and to normalise the orthography both in the text itself and when citing the text printed, other manuscript readings, and conjectures in the apparatus. To ‘normalise’ means to print what I think will look normal to twenty-first-century readers of Latin: I tacitly expand scribal abbreviations, and for controversies of spelling I draw on relevant scholarly resources, including OLD, TLL, and higher authorities when appropriate, not least the style notes for this series, which prescribe assimilated prefixes, and third-declension accusative plurals consistently in -es. However, there are still a few cases (including gnatus/a beside natus, subtegmina rather than subtemina, externauit beside exstantes and other words in exs-, admirantes beside alludebant, suscepit beside succumbens) where for particular reasons (whether manuscript evidence, readerly familiarity, or etymology) I have left orthographical inconsistencies in the text; often in such cases there is more on the spelling in the commentary. The following orthographical variations are generally ­ignored: • single/double consonants (e.g. pupis, querellas, flamam, tullit, brachia, spummosa, siccine, falaci, referram, sustolens, digittis), especially in proper names (e.g. pelliaco, thesalie, iachus), and including where one of the two consonants in a pair is indicated by an abbreviation (e.g. neptu ˉ ni, coˉnubia) • assimilation/non-assimilation of prefixes (e.g. ammirantes, adludebant, inmemor), including cases where this coincides with mistaken word division (e.g. impectore, in memori, considere) and single/double consonants (affesso) • i/y (e.g. pynus, nereydes, himeneos, dya, ydalium, limphys, thyaso, sathirorum, sydere, phrigii, polixenia) • ae/e¸/e/oe (e.g. fedans, hec, ere, phetontis, pe¸ne¸, persepe, innupte), including cases where quae is written as que or as the abbreviation for -que • o/u after semi-vowels (iocundo, iocundior, diuolso) 107

INT RODUCTION

• omission or addition of h after c, t (e.g. cholchis, pulcerrima, tesea, minothauro, michi, baccantis, nichil, tirsos, thedas), before vowels at the beginnings of words (e.g. hera, ortata), in the exclamation a (ah, ha) • c/t before i + vowel (e.g. eciam, tientem, racio, supplitium) • t/tt/ct (e.g. prospettans, negleto, cunta, mictam) • d/t (aliquit, reliquid ) • f/ph (e.g. prophani, frigii, nephando) • x/s/xs (e.g. extantes, estremis, sirtix, ansia) • posquam for postquam • c/qu/cqu (e.g. secuntur, necquicquam) • m/n before q (quoscunque, unquam) • mn/mpn (sompno) However, the apparatus does include cases where, on the one hand, a minor mistake that is not purely orthographical produces nonsense (e.g. conpexit for conspexit), and, on the other, an orthographical variation produces something that happens to be a Latin word, even if it is one that makes no sense in the context (e.g. sumo for summo). Word division, however, is a special case: in this area it is so difficult to tell what scribes were thinking when they left or did not leave a gap before que or after in (for instance) that errors, variations, and uncertainties in word division are generally not reported at all. A few exceptions to these general principles may, however, be observed in the apparatus. I sometimes reproduce the manuscripts’ original orthography in order to show how this may have played a role in corruption: this can include reproducing abbreviations or indicating their expansion with brackets.418 I capitalise proper names in the text, but not when citing (other) manuscript readings for these names or variations on them in the apparatus, since it is not always possible to tell whether scribes thought they were writing a proper name or not. Where a change of orthography (including word division or capitalisation) or of punctuation changes or clarifies the meaning of the text to the extent that it seems appropriate to credit the 418

Occasionally I also reproduce an abbreviation when it is not clear to me exactly what the scribe meant it to convey (e.g. in O at 151, 353, R2 at 28).

108

5  NOTE ON T HE COM M ENTA R Y

­ riginator of this change with a conjecture, I do credit that pero son (if their name is known). Finally, it is worth stating explicitly that although the apparatus should give all the essential information itself, it is written to be used in conjunction with the commentary. The commentary, as it explores the interpretation of the text, sometimes examines interpretations that I think are extremely unlikely or even impossible. Since interpretation cannot be separated from textual criticism, this means that occasionally I include conjectures in the apparatus which I think have no chance of being correct, but which are part of a line of interpretation I want to discuss: if this means that the apparatus does not always strike all readers as uere criticus, so be it.

5  NOTE ON THE COMMENTAR Y This book is long, but abundantly cross-referenced, and should be easy to navigate. The lemmatised commentary has a conventional arrangement: introductory notes on sections of the poem are followed (often) by notes on groups of lines within those sections, then by notes on individual lines (or pairs of lines where the syntax runs over a line break), which are usually further divided into sub-notes introduced by lemmata citing the text. Readers looking for information on a point of detail should always consult the notes on each of these levels that cover a word or passage they are interested in, but I often include explicit cross-references up or down within the hierarchical structure in order to encourage this. Cross-references cite notes either by line number (or numbers) alone, or by line number with lemma where this will aid navigation. Material described as being ‘above’ or ‘below’ will be found above or below within the note on a whole line or group of lines, rather than merely within the sub-note on a lemma. The corpus of texts that forms the basis for my comments on the frequency of words and similar phenomena is the ancient Latin writing of pre-Christian Roman culture, from the earliest to the latest extant works of pagan antiquity, but with more weight given to those from the first century bc to the second 109

INT RODUCTION

century ad, which I sometimes refer to as ‘classical’ Latin. In addition, I occasionally refer explicitly to late antique, usually Christian writers. I have drawn evidence for this corpus from TLL and from two databases, PHI 5.3 and LLT,419 supplemented by guidance from OLD. Similarly, my main reference points for Greek are TLG (excluding post-ancient texts) supplemented by LSJ. When quoting or referring to ancient works, and, as far as possible, also when counting examples, I aim to print or assume what I consider most likely to be the correct text, rather than necessarily following the text of a particular edition uniformly. Orthography of other Latin texts, too, normally follows the practice I have adopted for my own text of Catullus 64 (see 4c above). Abbreviations of ancient authors and works generally follow OLD and LSJ; exceptions such as Ap. Arg. are unlikely to cause confusion. I usually cite other commentators on Catullus 64 by name only without adding ‘here’ or ‘on this line’, and the same is true for Trappes-Lomax.420 In my citations of commentators on other texts, it should be easy to see the difference between references to co-authors (‘see Heyworth–Morwood ad loc.’) and references to the authors of independent works (‘see Austin and Horsfall ad loc.’). ‘See X ad loc.’ directs the reader to X’s comments on whatever element of the text I am discussing, even if those comments form part of X’s note on a larger section of the text. A commentary explores what it is like to read a text, and this one refers to various readers. These include myself (‘I’), myself together with my imagined community of modern readers (‘we’), a modern reader or readers in the generalised third person (plural ‘they’ or singular ‘they’), actual past readers, including both ancient writers and modern scholars (‘Virgil’, ‘Haupt’, etc.; see 3 above), and the Roman reader or readers that Catullus may have imagined (plural ‘they’ or, occasionally, singular ‘he’). As I have thought about all of these readers while writing See the Bibliography, Section 1. I have searched PHI 5.3 either via its website https://latin.packhum.org or by using the Diogenes application created by Peter Heslin to search the same database, as formerly supplied by the Packard Humanities Institute on CD-ROM. 420 For my citations of the originators of conjectures see 4c above. 419

110

5  NOTE ON T HE COM M ENTA R Y

the commentary, and not least about the various times, places, and contexts in which Catullus 64 has been and may be read, I have been particularly aware that while it may sometimes be appropriate on the available evidence to close off an interpretation that I consider impossible, or to argue strongly for a particular idea that I believe to be correct, more often there will be something useful to be said for and against each of several competing or overlapping possibilities. Throughout the commentary, I aim above all to be honest about my own level of certainty about the possibilities of meaning that I am discussing.421 I hope that many future readers of the poem will find me a helpful guide as they develop their own interpretations in response to Catullus’ text.

421

Cf. 4c above on my textual decision-making.

111

SI G LA

O G R m

Oxoniensis Bodleianus Canonicianus class. lat. 30, circa 1350–75 scriptus Parisinus lat. 14137, anno 1375 scriptus Vaticanus Ottobonianus lat. 1829, circa 1375–94 scriptus Venetus Marcianus lat. xii.80 (4167), circa 1400–6 scriptus

O1G1R1m1 manus prima uel lectionem primam corrigens uel lectionem pleniorem praebens R2 Ra m2 G2

manus secunda in R (Colucii Salutati) manus recentior in R manus secunda in m manus secunda in G

ϛ

codices recentiores uel unus uel plures

See also the Introduction (Section 4) and Bibliography (Sections 2 and 3).

112

TEXT AND C R I TI C AL AP PARATUS

Peliaco quondam prognatae uertice pinus dicuntur liquidas Neptuni nasse per undas Phasidos ad fluctus et fines Aeeteos, cum lecti iuuenes, Argiuae robora pubis, auratam optantes Colchis auertere pellem ausi sunt uada salsa cita decurrere puppi, caerula uerrentes abiegnis aequora palmis. diua quibus retinens in summis urbibus arces ipsa leui fecit uolitantem flamine currum, pinea coniungens inflexae texta carinae; ipsa rudem cursus prora imbuit Amphitriten.   quae simul ac rostro uentosum proscidit aequor tortaque remigio spumis incanuit unda, emersere freti candenti e gurgite uultus aequoreae monstrum Nereides admirantes. illa, atque alia, uiderunt luce marinas mortales oculi nudato corpore nymphas nutricum tenus exstantes e gurgite cano.   tum Thetidis Peleus incensus fertur amore, tum Thetis humanos non despexit hymenaeos, tum Thetidi pater ipse iugandum Pelea sensit.

5

10

15

20

ante 1 titulum Argonautia praebet R 2 (sic etiam m2G2)  1  citat Aelius Festus Apthonius  (‘Marius Victorinus’ ), GL 6.125.6  3  Phasidos R 2 u.l.: fasidicos O: fascidicos GR  Aeeteos Ald. (Aeetheios in contextu et Aeteios in comm. Parthenius): ceticos O: tetidicos O1 u.l.: oeticos GR  4  pubis ϛ: ­puppis OGR  6 ­decurrere] percurrere Harrison*  7  uerrentes R 2: uerentes OGR  9  uolitantem O1GR: uoluntatem O  10  texta O: testa GR  11  ipsa Baehrens: illa OGR  cursus Heyworth: cursu OGR  prora liber uetustus Mediolanensis teste Voss: post eam O: proram O1 u.l.: primam GR : prima ϛ: prorae Mitscherlich  Amphitriten Bernardus Pisanus: aphitritem O: aphitrite O1: amphitricem R : amphitritem GR 2: amphitrionem R 2 u.l.  12  proscidit R: procidit OG  13  tortaque ϛ, Avancius: totaque OGR: motaque Baehrens  incanuit Ald.: incanduit OGR  14  freti Schrader : feri OGR: fero Schrader   candenti e] candentes Eldick  uultus] ponti Koch  15  aequoreae GR: aequore O  monstrum] monstrorum O1 u.l.   16  illa atque Bergk, ϛ teste Helm: illa atque GR: illa O: illa haudque Sabellicus: illaque Puccius : illa, si qua Lachmann: illa, atque Vahlen, Schulze: illa, ante Friedrich: illa, ante Goold  uiderunt ϛ: uidere OGR   atque illa uidere luce marinas Baehrens  in marg. deficit deas uel aliud simile substantiuum add. O1?  17  oculi ϛ: oculis OGR  19  tum GR: cum O  20  tum m: cum OGR  21  tum Ald.: cum OGR  sensit] sanxit Pontanus, duo uetusta exemplaria teste Voss  

115

C ATULLUS : POEM 64

o nimis optato saeclorum tempore nati heroes, saluete deum genus! o bona matrum progenies, saluete iter uos ego saepe meo uos carmine compellabo. teque adeo eximie taedis felicibus aucte, Thessaliae columen Peleu, cui Iuppiter ipse, ipse suos diuum genitor concessit amores: tene Thetis tenuit pulcherrima Nereine? tene suam Tethys concessit ducere neptem, Oceanusque, mari totum qui amplectitur orbem?   quae simul optatae finito tempore luces aduenere, domum conuentu tota frequentat Thessalia, oppletur laetanti regia coetu: dona ferunt prae se, declarant gaudia uultu. deseritur Scyros; linquunt Pthiotica Tempe Crannonisque domos ac moenia Larisaea; Pharsalum coeunt, Pharsalia tecta frequentant. rura colit nemo: mollescunt colla iuuencis, non humilis curuis purgatur uinea rastris, non glebam prono conuellit uomere taurus, non falx attenuat frondatorum arboris umbram,

23b 25

30

35

40

22  saeclorum R 2: saeculorum OGR  23–23b  scholia Veronensia ad Virg. Aen. 5.80 Catullus: saluete deum gens o bona matrum progenies saluete iter[…]   23  genus] gens scholia Veronensia  matrum scholia Veronensia: mater OGR: matre R 2 u.l.  23b  om. OGR: saluete iter ex scholiis Veronensibus suppl. Orioli  iter  Peerlkamp: iter  Orioli: iter  Fröhlich: iter  Munro: iter aut iter Tucker : iter Bettini: iter Luck  24  meo] memor Traill: mero Palmer : melo uel meli Butterfield: alio Heyworth*  uos (2o)] post Bergk  compellabo] concelebrabo Harrison  25  t(a)edis O: thetis GR   28  Nereine Haupt: nectine OGR: neptine R 2 u.l.: neutu ˉ ne R 2 u.l.: Nerinarum Allen: Nerinaon Trappes-Lomax  29  Tethys ϛ: thetis OGR  31  quae] cui uel quis Müller  optatae ϛ: optato OGR  finito GR: finite O  32  aduenere ϛ: adlenire OGR  33  oppletur GR: opp(u)letur O  35  deseritur] destituunt Nisbet  Scyros ϛ: siros O: syros GR: Cieros Meineke  Tempe] templa Lennep: tecta Fröhlich  36  Crannonisque Victorius: graumonisque O: graiunonisque GR  moenia Larisaea ϛ: nicenis alacrisea OGR (-ss- O)  37  Pharsalum Pontanus: farsaliam OGR  38–42  aliter ordinant alii  : 38, 40, 42, 39, 41 Ramírez de Prado  : 39, 41, 38, 40, 42 Ramiresio male ascripsit Mitscherlich  : 38–9, 41, 40, 42 Ramler : 39, 41, 40, 38, 42 Baehrens  : 38, 41, 39–40, 42 Schmidt  41  falx attenuat frondatorum arboris] falce attenuat frondator roboris Lennep  

116

C AT ULLUS : POEM 64

squalida desertis rubigo infertur aratris. ipsius at sedes, quacumque opulenta recessit regia, fulgenti splendent auro atque argento. candet ebur soliis, collucent pocula mensae, tota domus gaudet regali splendida gaza. puluinar uero diuae geniale locatur sedibus in mediis, Indo quod dente politum tincta tegit roseo conchyli purpura fuco.   haec uestis priscis hominum uariata figuris heroum mira uirtutes indicat arte. namque fluentisono prospectans litore Diae Thesea cedentem celeri cum classe tuetur indomitos in corde gerens Ariadna furores; necdum etiam sese quae uisit uisere credit, utpote fallaci quae tum primum excita somno desertam in sola miseram se cernat harena. immemor at iuuenis fugiens pellit uada remis, irrita uentosae linquens promissa procellae. quem procul ex alga maestis Minois ocellis, saxea ut effigies bacchantis, prospicit, eheu prospicit et magnis curarum fluctuat undis, non flauo retinens subtilem uertice mitram, non contecta leui uelatum pectus amictu, non tereti strophio lactentes uincta papillas, omnia quae toto delapsa e corpore passim

45

50

55

60

65

42  rubigo OG: robigo R  infertur] increscit Nisbet  43 at ϛ: ad OGR   45  mensae] mensis Doering   : mensa Heyworth  47  puluinar GR: pluuinar O   52  fluentisono] fluctisono Mähly (fluctisono ϛ)  litore OG: litora R   Diae ϛ: dia OGR  54 indomitos GR : indomites O  Ariadna ϛ: adriana OGR  55  quae uisit uisere Voss: -que sui tui se OGR (terni R 2 u.l.)  ad finem uersus post  se credit  rasura in  R,  deficit  add. supra R 2  56 tum GR: tunc O  58 fugiens pellit uada remis] pellit uada remis TrappesLomax  59 citat Petrarca in margine codicis Ambrosiani (A  79 inf.), f.  246v, ad Stat.  Ach.  1.960  60  alga] acta Falcoburgius  Minois OGR 2: minoeis R  61  saxea R 2: saxa OGR  eheu Bergk: heue OGR: euoe Ald.: euhoe Lachmann  62 et R 2: cum OGR  64  contecta GR: contenta O  uelatum] praeclarum McKie: nudatum Schwabe: uariatum Nisbet: laniatum Owen: uiolatum nescioquis teste Peiper : uesanum Riese: decoratum Kakridis: bullatum Fea: niueum per Mähly: niueum tum Baehrens  65 Isidorus, Etymologiae 19.33.3 de quo ait Cinna: strophio lactantes cincta papillas  66 delapsa e ϛ: delapse O: delapso G: delapso e R  

117

C ATULLUS : POEM 64

ipsius ante pedes fluctus salis alludebant. sed neque tum mitrae neque tum fluitantis amictus illa uicem curans toto ex te pectore, Theseu, toto animo, tota pendebat perdita mente. a misera, assiduis quam luctibus externauit spinosas Erycina serens in pectore curas, illa tempestate, ferox quo tempore Theseus egressus curuis e litoribus Piraei attigit iniusti regis Gortynia tecta.   nam perhibent olim, crudeli peste coactam Androgeoneae poenas exsoluere caedis, electos iuuenes simul et decus innuptarum Cecropiam solitam esse dapem dare Minotauro. quis angusta malis cum moenia uexarentur, ipse suum Theseus pro caris corpus Athenis proicere optauit potius quam talia Cretam funera Cecropia nec funera portarentur. atque ita naue leui nitens ac lenibus auris magnanimum ad Minoa uenit sedesque superbas. hunc simul ac cupido conspexit lumine uirgo regia, quam suaues exspirans castus odores lectulus in molli complexu matris alebat, quales Eurotae progignunt flumina myrtos auraue distinctos educit uerna colores, non prius ex illo flagrantia declinauit lumina, quam cuncto concepit corpore flammam

70

75

80

85

90

68  sed neque ϛ: sineque OGR: sic neque Rossbach   tum … tum GR: tamen … tamen (tn̄ … tn̄ ) O  69  ex te GR: ex O  70  pendebat OG: perdebat R   71–2  citat Nonius 154 Lindsay s.v. externauit  71  luctibus] fluctibus Baehrens   73  tempestate] tempestate Baehrens  quo tempore ϛ: -que et tempore OGR: quo ex tempore Lachmann: qua robore Fröhlich (quo robore aut quo pectore Schrader, cum robore Ritschl ): qua pectore Peiper : qua corpore Thomson   75  iniusti] inuisi Heinsius  Gortynia Palladius: cortinia OGR  tecta ϛ teste Parthenio: tempta OGR: templa ϛ  77  Androgeoneae Victorius (Androgeaneae Calphurnius): cum androgeane O: cum androgeanee GR  80  angusta] augusta ϛ  moenia R 2: incenia O: inoenia GR  82  proicere] prohicere O: proiicere GR  83  Cecropia Baehrens: cecropiae OGR: Cecropidum Heyworth*  funera (2o) GR: flinera O  85  Minoa] Minon uel Mino Kayachev  86  conspexit GR: conpexit O  89  Eurotae ϛ, 1472: europe OGR  progignunt ϛ: pergignunt OGR: percingunt ϛ: praecingunt Baehrens  myrtos] mirtos O, -tos R 2 u.l.: mirtus GR  92  cuncto] toto ϛ, Ald.  corpore] pectore ϛ

118

C AT ULLUS : POEM 64

funditus atque imis exarsit tota medullis. heu miseros agitans in miti corde furores sancte puer, curis hominum qui gaudia misces, quaeque regis Golgos quaeque Idalium frondosum, qualibus incensam iactastis mente puellam fluctibus, in flauo saepe hospite suspirantem! quantos illa tulit languenti corde timores! quam tum saepe magis fulgore expalluit auri, cum saeuum cupiens contra contendere monstrum aut mortem appeteret Theseus aut praemia laudis! non ingrata tamen frustra munuscula diuis promittens tacito suscepit uota labello. nam uelut in summo quatientem bracchia Tauro quercum aut coniferam sudanti cortice pinum indomitus turbo, contorquens flamine robur, eruit (illa procul radicitus exstirpata prona cadit, late quaeuis cumque obuia frangens), sic domito saeuum prostrauit corpore Theseus nequiquam uanis iactantem cornua uentis. inde pedem sospes multa cum laude reflexit errabunda regens tenui uestigia filo, ne labyrintheis e flexibus egredientem tecti frustraretur inobseruabilis error.

95

100

105

110

115

94  miseros agitans Markland: misere exagitans OGR : miserae exagitans Puteolanus: miser exagitans ϛ  in miti Avancius: immiti OGR: immites Lennep   96  quaeque (1o) ϛ: quod neque O: quique GR  Golgos Politianus: ­cholcos O: colchos GR  100  quam tum Faernus: quanto OGR  102 ­ appeteret O: oppeteret GR  103  frustra] fauste McKie: superis Fea  104 suscepit Statius (subs- ϛ, suc- ϛ teste Statio): succendit OGR  105 uelut GR: uult O   106 ­coniferam Parrhasius: cornigeram OGR: conigeram ϛ  sudanti R 2: fundanti OGR  107 turbo, contorquens flamine robur] subito contorquens flamine turbo Palmer  108 eruit GR : emit O  radicitus] radicibus ϛ   exstirpata Palladius: exturbata OGR  109  quaeuis cumque Ellis (quaecumuis Voss): -que cum eius OGR: quaecumque sibi Statius: -que et cominus uel sim. ϛ: -que ruinis Birt: quaecumque habet Baehrens: quae sunt cumque Trappes-Lomax: quae ­corruit Nuzzo: quae cingunt McKie  obuia O, R 2 u.l.: omnia GR  frangens] frangit ϛ, 1472  110  domito saeuum] domito taurum Suringar (domitum taurum Lennep)  Theseus] taurum ϛ  111 nequiquam] nequicquam GR : ne quidquam O  uanis GR: nauis O  113  errabunda GR: ereabunda O  114 ­ labyrintheis] laberinthis O: laberintheis GR  115 frustraretur O: frustaretur GR  

119

C ATULLUS : POEM 64

  sed quid ego a primo digressus carmine plura commemorem, ut linquens genitoris filia uultum, ut consanguineae complexum, ut denique matris, quae misera in gnata deperdita lamentata est, omnibus his Thesei dulcem praeoptarit amorem; aut ut uecta rati spumosa ad litora Diae aut ut eam deuinctam lumina somno liquerit immemori discedens pectore coniunx? saepe illam perhibent ardenti corde furentem clarisonas imo fudisse e pectore uoces, ac tum praeruptos tristem conscendere montes, unde aciem in pelagi uastos protenderet aestus, tum tremuli salis aduersas procurrere in undas, mollia nudatae tollentem tegmina surae, atque haec extremis maestam dixisse querelis, frigidulos udo singultus ore cientem:   ‘sicine me patriis auectam, perfide, ab aris, perfide, deserto liquisti in litore, Theseu? sicine discedens neglecto numine diuum immemor (a!) deuota domum periuria portas? nullane res potuit crudelis flectere mentis consilium? tibi nulla fuit clementia praesto,

120 121b

125

130

135

116  a ϛ: cum OGR  117  linquens OR: liquens G  119 in gnata OG: ­ignata R  lamentata est Conington (lamentatur Bücheler, lamentetur Murgia): leta OGR: laeta Lachmann: lucta Rossbach  ad finem uersus deficit  add. R 2  120  his GR: hiis O  praeoptarit Statius (-ret ϛ, -uit ϛ): portaret OGR: praeferret ϛ, Calphurnius  121 aut ut GR : aut O  uecta R: necta OG  rati Passerat: ratis OGR  post 121 lacunam unius uersus posuit Baehrens  122  , aut ut eam Lachmann (fugerit Fröhlich)  eam Palmer (placido ϛ, molli ϛ, dulci ϛ, suaui Shumilin): eam OGR: securo Trappes-Lomax: torpenti McKie  deuinctam ϛ: deuincta OGR  in marg. deficit  add. R 2  125  e GR : ex O  126 praeruptos GR: praeruptes O  tristem ϛ: tristes OGR  conscendere GR: confendere O   127  in ϛ: om. OGR  protenderet R: pretenderet OG: per tenderet Baehrens   128  tremuli] tremulam Kokoszkiewicz  salis GR: salus O  130 haec OG1: hoc GR  maestam dixisse OR: dixisse maestam G  132 patriis GR: patris O  auectam R 2: auertam OGR    , R 2 u.l.  aris] oris ϛ, Puteolanus  133 in GR: om. O  134  discedens OR: discendens G  135 a OGR 2: ab R : heu Trappes-Lomax  136  nullane ϛ: nullaue OGR   crudelis … mentis R 2: crudeles … mentes OGR: crudele … mentis Vendel-Heyl  

120

C AT ULLUS : POEM 64

immite ut nostri uellet miserescere pectus? at non haec quondam blanda promissa dedisti uoce mihi, non haec miseram sperare iubebas, sed conubia laeta, sed optatos hymenaeos, quae cuncta aerii discerpunt irrita uenti. nunc iam nulla uiro iuranti femina credat, nulla uiri speret sermones esse fideles; quis dum aliquid cupiens animus praegestit apisci, nil metuunt iurare, nihil promittere parcunt: sed simul ac cupidae mentis satiata libido est, dicta nihil meminere, nihil periuria curant. certe ego te in medio uersantem turbine leti eripui, et potius germanum amittere creui quam tibi fallaci supremo in tempore dessem. pro quo dilaceranda feris dabor alitibusque praeda, neque iniecta tumulabor mortua terra. quaenam te genuit sola sub rupe leaena, quod mare conceptum spumantibus exspuit undis, quae Syrtis, quae Scylla rapax, quae uasta Charybdis, talia qui reddis pro dulci praemia uita? si tibi non cordi fuerant conubia nostra, saeua quod horrebas prisci praecepta parentis, at tamen in uestras potuisti ducere sedes, quae tibi iucundo famularer serua labore, candida permulcens liquidis uestigia lymphis, purpureaue tuum consternens ueste cubile.

140

145

150

155

160

138  uellet OG: uellem R  miserescere ϛ: mirescere O : mitescere GR   139  blanda O: nobis GR  140  non ϛ: nec OGR  miseram ϛ, 1472: misere OGR: misera Baehrens  141  citat Petrarca in margine codicis Ambrosiani (A 79 inf.), f. 108r, ad Virg. Aen. 4.316  142  discerpunt ϛ: disserpunt GR: desserpunt O  143  nunc A.  Guarinus: tum OGR: hinc Fröhlich  143–8  nulla uiro  … nihil periuria curant citat Hieremias, Compendium moralium notabilium 4.5.11   145  praegestit R 2, Hieremias? : postgestit OGR  apisci O 1?R (supra scripto pro adipisci O 1? ), Hieremias: adipisci OG, Hieremias u.l.  148  meminere Czwalina: metuere OGR  149  in medio uersantem] m. u. i. Trappes-Lomax  leti GR: lecti O  151  dessem] deē m O: deessem GR  152  alitibusque OR: altibusque G  153  praeda GR: postea O  iniecta Calphurnius (iniacta Passerat): intacta OGR  156  Scylla] scilla O: silla GR  157  talia qui GR: taliaque O  reddis GR: redis O  159  parentis GR: peremtis O  160  uestras GR: nostras O   163  post 160 transp. O  

121

C ATULLUS : POEM 64

  sed quid ego ignaris nequiquam conqueror auris, externata malo, quae nullis sensibus auctae nec missas audire queunt nec reddere uoces? ille autem prope iam mediis uersatur in undis, nec quisquam apparet uacua mortalis in acta. sic nimis insultans extremo tempore saeua fors etiam nostris inuidit questibus aures. Iuppiter omnipotens, utinam ne tempore primo Cnosia Cecropiae tetigissent litora puppes, indomito nec dira ferens stipendia tauro perfidus in Creta religasset nauita funem, nec malus hic celans dulci crudelia forma consilia in nostris requiesset sedibus hospes! nam quo me referam? quali spe perdita nitor? Idaeosne petam montes? at gurgite lato discernens ponti truculentum diuidit aequor. an patris auxilium sperem? quemne ipsa reliqui respersum iuuenem fraterna caede secuta? coniugis an fido consoler memet amore? quine fugit lentos incuruans gurgite remos? praeterea nullo colitur sola insula tecto, nec patet egressus pelagi cingentibus undis.

165

170

175

180

185

164  sed R 2: si OGR  ignaris] ingratis Wakefield  conqueror ϛ: conquerar OGR  auris R 2: aures OGR  165  externata O: extenuata GR  167  prope iam] properans Pleitner  168  acta Falcoburgius: alga OGR  169  saeua] saeue Nisbet  170  fors GR: fers O  171–2  citant Macrobius, Saturnalia 6.1.42, Petrarca in margine codicis Ambrosiani (A 79 inf.), f. 113v, ad Virg. Aen. 4.657–8 (ex Macrobio?  )  171  ne] non Macrobius, Petrarca  172  Cnosia R: Gnosia OG  174  in Creta O: in Cretam GR: intortum Gratwick  175  hic GR: haec O  176  consilia in ϛ: consilium OG: conscilium R  nostris om. O   requiesset ϛ: requisisset OGR  177  nam] nunc Spengel: iam Peiper    nitor] nitar ϛ  178 Idaeosne Parthenius: idoneos ne OGR: Idmoneos ne R 2: Idmeneos ne 1472 (Idomeneosne uel  Idomeniosne Voss, Idomeneusne Lachmann, Idomeneine Bücheler)  at Puccius: a OGR: an Ellis  179  ponti O: pontum GR  diuidit Puccius: ubi diuidit OGR  180  an R 2: in OGR  quemne OGR: quem ue R 2   181  respersum iuuenem fraterna] f. i. r. Fisch  182 ­ consoler memet GR: consoles me manet O  183  quine O: quiue GR  lentos O: uentos GR   incuruans] curuans in Trappes-Lomax (qui fugit incuruans lentos in g. r. Avancius)   184  colitur Palmer : litus OGR: recipit Baehrens: laeta est Birt   nullo colitur sola insula tecto] nullum sola insula tectum Trappes-­Lomax  185  patet OG: pater R  

122

C AT ULLUS : POEM 64

nulla fugae ratio, nulla spes: omnia muta, omnia sunt deserta, ostentant omnia letum.   non tamen ante mihi languescent lumina morte, nec prius a fesso secedent corpore sensus, quam iustam a diuis exposcam prodita multam caelestumque fidem postrema comprecer hora. quare facta uirum multantes uindice poena Eumenides, quibus anguino redimita capillo frons exspirantes praeportat pectoris iras, huc huc aduentate, meas audite querelas, quas ego uae misera extremis proferre medullis cogor inops, amens, ardenti caeca furore. quae quoniam uerae nascuntur pectore ab imo, uos nolite pati nostrum uanescere luctum, sed quali solam Theseus me mente reliquit, tali mente, deae, funestet seque suosque.’   has postquam maesto profudit pectore uoces, supplicium saeuis exposcens anxia factis, annuit inuicto caelestum numine rector; quo motu tellus atque horrida contremuerunt aequora concussitque micantia sidera mundus. ipse autem caeca mentem caligine Theseus consitus oblito dimisit pectore cuncta quae mandata prius constanti mente tenebat, dulcia nec maesto sustollens signa parenti sospitem Erectheum se ostendit uisere portum.   namque ferunt olim, castae cum moenia diuae

190

195

200

205

210

186  nulla spes] nulla spes Puccius, ϛ teste Statio: nulla spest Trappes-Lomax   187  ostentant] intentant Orelli  190  iustam GR: iusta O  multam Ald.: mulctam GR: muletam O  191  comprecer ϛ: comprecor OGR  192  multantes Ald.: mulctantes OGR  194  praeportat GR: postportat O  195  meas OR 2: et meas GR  196  misera] miserae ϛ: miseram Schmidt  extremis] imis R a u.l.: ex imis ϛ  197  amens, ardenti Nisbet: ardens, amenti OGR  200  quali solam ϛ: qualis sola OGR  201  funestet OR: fimestet G  203  factis] fatis ϛ teste Nuzzo  204  inuicto ϛ: inuito OGR  205  quo motu  Heyse:  quomodo tunc  OGR: quo nutu ϛ: quo nutu et Fea  206  concussitque] concussusque libri ueteres teste Voss   concussitque … mundus] concussique … mundi Pleitner  207  mentem ϛ: mente OGR: mentis ϛ  210  dulcia] lucida Schrader  211  Erectheum Voss (Erectheos … portus Heinsius): ereptum OGR  uisere OR: uiscere G   212  castae Pontanus: classi OGR  moenia R: moenico OGR1  

123

C ATULLUS : POEM 64

linquentem gnatum uentis concrederet Aegeus, talia complexum iuueni mandata dedisse: ‘gnate mihi longa iucundior unice uita, reddite in extrema nuper mihi fine senectae, gnate, ego quem in dubios cogor dimittere casus, quandoquidem fortuna mea ac tua feruida uirtus eripit inuito mihi te, cui languida nondum lumina sunt gnati cara saturata figura, non ego te gaudens laetanti pectore mittam, nec te ferre sinam fortunae signa secundae, sed primum multas expromam mente querelas, canitiem terra atque infuso puluere foedans, inde infecta uago suspendam lintea malo, nostros ut luctus nostraeque incendia mentis carbasus obscurata dicet ferrugine Hibera. quod tibi si sancti concesserit incola Itoni, quae nostrum genus ac sedes defendere Erecthei annuit, ut tauri respergas sanguine dextram, tum uero facito ut memori tibi condita corde haec uigeant mandata, nec ulla oblitteret aetas: ut simul ac nostros inuisent lumina colles, funestam antemnae deponant undique uestem, candidaque intorti sustollant uela rudentes, quam primum cernens ut laeta gaudia mente agnoscam, cum te reducem lux prospera sistet.’   haec mandata prius constanti mente tenentem Thesea ceu pulsae uentorum flamine nubes

215 217 216 218 220

225

230

235

213  concrederet Avancius: cum crederet OGR  Aegeus] egeus R 2: egens OGR  214  complexum] complexo ϛ  215 gnate GR: gnati O  longa] longe Hoeufft  217 ante 216 transp. Baehrens  extrema] extremae Ald.   216  quem GR: quoniam (qm̄ ) O  219  cui OR 2: quem GR  221 laetanti GR: lectanti O  224  infuso R 2: infulso OGR  227 obscurata] obscura R 2   dicet] decet ϛ teste Statio  228  Itoni Barbarus: ithomi O: ythomi GR  229  ac ϛ: has OGR  sedes GR : secles O  Erecthei Voss: freti OGR  231 tum O: tu GR  232  oblitteret OR 2: oblitueret G: obliterat R  233 ac ϛ: haec OGR  234  antemnae] antenne R: antē nene ne O: antenne ne G  post 235 inserendum esse posuerunt Muretus, Faernus lucida qua splendent summi carchesia mali (tribuit Catullo Nonius 876 Lindsay)  237 ­ agnoscam] cognoscam ϛ teste Ellis  lux Baehrens: aetas OGR: fors Avancius: sors ϛ, Ald.  sistet O: sistens G: sistant R: sistent R1  239  ceu R 2: seu OGR  

124

C AT ULLUS : POEM 64

aerium niuei montis liquere cacumen. at pater, ut summa prospectum ex arce petebat, anxia in assiduos absumens lumina fletus, cum primum inflati conspexit lintea ueli, praecipitem sese scopulorum e uertice iecit, amissum credens immiti Thesea fato. sic funesta domus ingressus tecta paternae corde ferox Theseus, qualem Minoidi luctum obtulerat mente immemori, talem ipse recepit. quae tum prospectans cedentem maesta carinam multiplices animo uoluebat saucia curas.   at parte ex alia florens uolitabat Iacchus cum thiaso satyrorum et Nysigenis silenis, te quaerens, Ariadna, tuoque incensus amore. cui famulae passim lymphata mente furebant euhoe bacchantes, euhoe capita inflectentes: harum pars tecta quatiebant cuspide thyrsos, pars e diuulso iactabant membra iuuenco, pars sese tortis serpentibus incingebant, pars obscura cauis celebrabant orgia cistis, orgia quae frustra cupiunt audire profani; plangebant aliae proceris tympana palmis, aut tereti tenues tinnitus aere ciebant;

240

245

250

255

260

240  aerium GR: aereum O  242  fletus] uisus Baehrens  243  inflati] ­infausti Heyse: falsi Slater : infecti Sabellicus  245 fato GR : facto O   246–7 tecta paternae (Ald., ϛ) | corde Trimble: tecta paterna | morte OGR : tecta paternae | Marte Marcilius: morte paterna | tecta Trappes-Lomax  247 ­ minoidi 1472, ϛ: minoi a OGR  249 quae OR : quem G  tum ϛ: tamen OGR (tn̄ OR)  prospectans] a- R 2 u.l.  cedentem GR : credentem O  251 ­ parte ϛ: pater OGR  252 cum GR : tum O  Nysigenis] nisigenis R 2: nisi genis OGR  253 te R 2: et OGR  qu(a)erens OR : querenus G  Ariadna ϛ: adriana OGR  post 253 lacunam posuit Bergk  254 cui famulae McKie: qui tum alacres OGR : cui thyades Skutsch: thyades huic Trappes-Lomax: thyiades at Heyworth: cui tum alacres Kalinka: quae tum alacres Bergk: quam tum alacres Schwabe: quicum alacres Baehrens  lymphata GR : linphata O  post 254 lacunam posuit Lenchantin  255 euhoe … euhoe ϛ (euohe … euohe G2?   ): euche  … euche OGR  bacchantes] bacchantes Pighi  post 255 lacunam posuit Cornish  256 harum] horum ϛ   259 ­ cauis GR1: canis OR  260 del. Koeler  audire] aperire McKie: spectare Oakley*  261 aliae] alii 1472  262 tenues] tenuis GR : tenais O  tinnitus OR 2: tintinitus GR  

125

C ATULLUS : POEM 64

multis raucisonos efflabant cornua bombos barbaraque horribili stridebat tibia cantu.   talibus amplifice uestis decorata figuris puluinar complexa suo uelabat amictu. quae postquam cupide spectando Thessala pubes expleta est, sanctis coepit decedere diuis. hic, qualis flatu placidum mare matutino horrificans Zephyrus procliues incitat undas, Aurora exoriente uagi sub limina Solis, quae tarde primum clementi flamine pulsae procedunt (leuiterque sonant plangore cachinni), post uento crescente magis magis increbrescunt, purpureaque procul nantes ab luce refulgent: sic tum uestibulis linquentes regia tecta ad se quisque uago passim pede discedebant.   quorum post abitum princeps e uertice Peli aduenit Chiron portans siluestria dona: nam quoscumque ferunt campi, quos Thessala magnis montibus ora creat, quos propter fluminis undas aura aperit flores tepidi fecunda Fauoni, hos in distinctis plexos tulit ipse corollis, quo permulsa domus iucundo risit odore.

265

270

275

280

263  multis Puccius: multi OGR: multaque Ald.2: multae Scaliger  raucisonos … bombos] raucisonis … bombis Ald.  efflabant ϛ: efflebant OGR: inflabant Dousa filius  267 Thessala OR 2: thessalia GR  269  hic, qualis GR: haec, qualis O: ac quali Voss: ac qualis Ramler   : ac ueluti aut nam ueluti Heyworth*   270  Zephyrus] cephirus O: çephirus GR  procliues] procliuis O1: procliuit O: procliuas GR  271  sub limina ϛ: sublimia OGR: sub lumina Ald.   272  tarde] rarae ϛ, Latini  273 leuiterque O: leuiter GR : lenique Lennep: leni re- ϛ: leni et re- Voss: lenes re- Schmidt  274 increbrescunt 1472: increbescunt OGR  275  purpurea] purpureae Lennep   nantes ab] uariantes Puccius: uibrantes Mitscherlich : fluitantes Mähly: nascente ab Baehrens  refulgent ϛ: refulgens OGR  276  tum ϛ: tamen OGR (tn̄ OR, tam̄ G): ibi Haupt (tibi m)  uestibulis (uel -o) Schrader  : uestibuli OGR: e uestibulis Ellis: Thessalii Mähly: festini Baehrens: uestiflui Nisbet   tum uestibulis] tempestiui TrappesLomax  linquentes ϛ: linquentis OGR  277 ad ϛ: at OGR  278 e] a Goold, typothetae errore?  Peli] peley O : pelei GR : Pelii Ald.  280 quoscumque Ald.: quodcumque OGR  campi ϛ: campis OGR  Thessala GR: thessalia O  magnis R: magis OG  282  aura R 2: aurea OGR  aperit Housman: pit (perit aut parit) ORm, G ante rasuram? : parit G2  fecunda OGR 2: facunda ˉ R  283  in distinctis G (sic Wakefield ): indistinctis OR  corollis R 2 u.l.: corulis OG: curulis R  284  quo ϛ: quod O: quot GR  

126

C AT ULLUS : POEM 64

confestim Peneos adest, uiridantia Tempe, Tempe, quae siluae cingunt super impendentes, naiasin linquens crebris celebranda choreis, non uacuus: namque ille tulit radicibus altas fagos ac recto proceras stipite laurus, non sine nutanti platano lentaque sorore flammati Phaethontis et aeria cyparisso. haec circum sedes late contexta locauit, uestibulum ut molli uelatum fronde uireret. post hunc consequitur sollerti corde Prometheus, extenuata gerens ueteris uestigia poenae, quam quondam silici restrictus membra catena persoluit pendens e uerticibus praeruptis. inde pater diuum sancta cum coniuge natisque aduenit caelo, te solum, Phoebe, relinquens unigenamque simul cultricem montibus Idri: Pelea nam tecum pariter soror aspernata est, nec Thetidis taedas uoluit celebrare iugales.   qui postquam niueis flexerunt sedibus artus large multiplici constructae sunt dape mensae, cum interea infirmo quatientes corpora motu

285

290

295

300

305

285  Peneos 1472: penies OGR (- os R 2 u.l.): Peneus Puccius: Peneios ϛ  adest ϛ: adest ut OGR  287  del. Astius  naiasin Haupt (naiadum Realinus): minosim OGR: Peneisin Lenchantin: Meliasin Madvig: Limosin Fröhner : Haemonisin Heinsius: Minyasin Scaliger  crebris Lachmann: doris OGR: claris ϛ: sacris Housman: solitis Magnus: uariis Riese: udis Mähly  celebranda] celebrata McKie  alii alia  288  non uacuus B. Guarinus (non uacuos Bergk): non accuos O  : non acuos GR: nonacrios R 2 u.l., m1 u.l.: nonacrias G2 u.l.  ille] inde Marcilius: ipse Baehrens  radicibus Lenz: radicitus OGR  288–9  altas | fagos GR: altas | fages O: actas | fagos Turnebus teste Heinsio: alnos | fractas McKie  290 ­ nutanti R 2: mutanti OGR  sorore ϛ: sororum OGR  291  flammati O: flammanti GR  cyparisso A.  Guarinus: cupressu OGR  292  circum GR: tircum O   contexta R 2: contesta OGR  295  poenae R 2: pena OGR  296  quam ϛ: qua OGR  silici] triplici Baehrens: ultrici Heyworth*  restrictus GR: resittus O   298  diuum m: diui OGR  natis ϛ: gnatis OGR  300  montibus Idri] montis Ithyni Parthenius, ϛ (Ithini Avancius): montis Ithomi Ald.: montis Itonis uel Ithonis Puccius: montis Itoni uel Ithoni uel Ithomi A.  Guarinus: montis Olympi Harrison: montium Abydi Fröhlich: montis agrestis Heyworth*  Idri] ydri OGR: Idae ϛ: Hydrae Statius: Imbri Schrader : Istri uel Histri Baehrens: Iri Ellis: Pindi Hall: udis Marcilius: hirtis Schrader  301  pelea G: palea OR  303  niueis] niueos Statius  304  constructae sunt] sunt exstructae Baehrens  305  cum ­interea] dum uetera Heyworth* (tum u. Garrod )  

127

C ATULLUS : POEM 64

ueridicos Parcae coeperunt edere cantus. his corpus tremulum complectens undique uestis candida purpurea talos incinxerat ora, at roseae niueo residebant uertice uittae, aeternumque manus carpebant rite laborem. laeua colum molli lana retinebat amictum, dextera tum leuiter deducens fila supinis formabat digitis, tum prono pollice torquens libratum tereti uersabat turbine fusum, atque ita decerpens aequabat semper opus dens, laneaque aridulis haerebant morsa labellis, quae prius in leui fuerant exstantia filo; ante pedes autem candentis mollia lanae uellera uirgati custodibant calathisci. hae tum clarisona pernentes uellera uoce talia diuino fuderunt carmine fata, carmine, perfidiae quod post nulla arguet aetas.   ‘o decus eximium magnis uirtutibus augens, Emathiae tutamen, Opis carissime nato, accipe, quod laeta tibi pandunt luce sorores, ueridicum oraclum: sed uos, quae fata sequuntur,   currite ducentes subtegmina, currite, fusi. adueniet tibi iam portans optata maritis

310

315

320

325

306  coeperunt] ceperunt GR: teperunt O  edere GR: eclere O  307 ­ uestis Parthenius, ϛ: questus OGR  308  talos ed. Iuntina (-que talos ϛ): tuos OGR   incinxerat G: intinxerat OR  309  roseae niueo A. Guarinus (roseae ϛ, niueo ϛ): roseo niuee GR: roseo uinee O  uittae] uitte OR: uirte G? : uicte G2  311  colum R 2: collum OGR  312  fila GR: filia O  313  pollice ϛ: in pollice OGR   315  opus dens RmG2: epusdens O: opus dans G ante rasuram?  319  custodibant O: custodiebant GR  calathisci ϛ: calathisti OGR  320  hae ϛ: haec OGR   pernentes Heinsius: pellentes OGR : uellentes Fruterius: pectentes Statius: polientes Heinsius: plectentes Hertzberg: carpentes nescioquis teste Camps  pernentes uellera] pellentes aera Faernus: pellentes aethera Palmer : complentes aera aut complentes atria McKie   ante 323 titulum Epythalamum Thetidis et Pelei praebet R 2 (sic etiam m2, -ium G2)  324  tutamen opis ϛ (tu tamen opis R 2 u.l.): tutum (tutuˉ) opus OGR   uersum ita interpungendum esse docuit Housman  carissime] clarissime Puteolanus  326  oracl(u)m ut uid. OR : oraculum G  fata GR: facta O   327  Macrobius, Saturnalia 6.1.41 Catullus: currite ducenti subtemine currite fusi   Petrarca in margine codicis Ambrosiani (A  79 inf.), f.  8r, ad Virg. Ecl. 4.46 currite dicenti sub tegmine currite fusi: Catullus (ex Macrobio? )  post 327 lacunam quattuor uersuum posuit Pleitner  328  optata R 2: aptata OGR  

128

C AT ULLUS : POEM 64

Hesperus, adueniet fausto cum sidere coniunx, quae tibi flexanimo mentem perfundat amore, languidulosque paret tecum coniungere somnos, leuia substernens robusto bracchia collo.   currite ducentes subtegmina, currite, fusi. nulla domus tales umquam contexit amores, nullus amor tali coniunxit foedere amantes, qualis adest Thetidi, qualis concordia Pelei.   currite ducentes subtegmina, currite, fusi. nascetur uobis expers terroris Achilles, hostibus haud tergo, sed forti pectore notus, qui persaepe uago uictor certamine cursus flammea praeuertet celeris uestigia ceruae.   currite ducentes subtegmina, currite, fusi. non illi quisquam bello se conferet heros, cum Phrygii Teucro manabunt sanguine Troicaque obsidens longinquo moenia bello, periuri Pelopis uastabit tertius heres.   currite ducentes subtegmina, currite, fusi. illius egregias uirtutes claraque facta saepe fatebuntur gnatorum in funere matres, cum incultum cano soluent a uertice crinem, putriaque infirmis uariabunt pectora palmis.   currite ducentes subtegmina, currite, fusi.

330

335

340

345

350

329  Hesperus GR: Hespereus O  330  om. O  tibi] te Lachmann  flexanimo Muretus (flexanima Puccius, flexanimae Avancius): flexo animo GR  mentem … amore Muretus: mentis … amorem GR: mentis … amore ϛ  331  somnos ϛ: sonos OGR  332  post 333 transp. Voss, post 374 Herrmann  leuia GR: uenia O  334  tales umquam ϛ: umquam tales OGR  336  Pelei Ald.: Peleo OGR   post 336 lacunam unius uersus posuit Herrmann  340  uago] uagi Trappes-Lomax   341  praeuertet ϛ: peruertet O: pr(a)euertit GR  344  teucro R: teucto O: teuero G  manabunt GR: manebunt O  campi ϛ: teuen O, R 2 u.l.: tenen GR: Teucri ϛ: trunci ϛ: riui ϛ: muri Statius: acerui Voss: cliui Haupt: fines Fritze: (Phrygiae …) terrae Statius, (Phrygiae … manabit …) tellus Peiper  alii alia  347  subtegmina O: sub tegmine GR  350  incultum uel inculto Baehrens: in ciuos O: in ciuum O1: in ciuium GR: incomptum Nuzzo: lacerum Garrod : incinctum Lenchantin  cano Baehrens: canos OGR  soluent GR: soleunt O  crinem Baehrens: crimen O: crines GR  351  putria Heinsius: putrida OGR  352  subtegmina OGR 2: sub tegmine R  

129

C ATULLUS : POEM 64

namque uelut densas praecerpens messor aristas sole sub ardenti flauentia demetit arua, Troiugenum infesto prosternet corpora ferro.   currite ducentes subtegmina, currite, fusi. testis erit magnis uirtutibus unda Scamandri, quae passim rapido diffunditur Hellesponto, cuius iter densis angustans corporum aceruis alta tepefaciet permixta flumina caede.   currite ducentes subtegmina, currite, fusi. denique testis erit morti quoque reddita praeda, cum teres excelso coaceruatum aggere bustum excipiet niueos perculsae uirginis artus.   currite ducentes subtegmina, currite, fusi. nam simul ac fessis dederit fors copiam Achiuis urbis Dardaniae Neptunia soluere uincla, alta Polyxenia madefient caede sepulcra; quae, uelut ancipiti succumbens uictima ferro, proiciet truncum summisso poplite corpus.   currite ducentes subtegmina, currite, fusi. quare agite optatos ambo coniungite amores: accipiat coniunx felici foedere diuam,

355

360

365

370

353–6  om. R, suppl. R 2: post 347 transp. Peiper  353  densas GR 2: dē psas O   praecerpens ϛ, Statius: praecernens OR 2: praeterriens G: praecidens Baehrens: praetondens Trimble : praesternens Scaliger  messor O: cultor GR 2  post 354 lacunam unius uersus posuit Voss, duorum uersuum McKie: | T. i. prosternens suppl. Fröhlich exempli gratia   355  Troiugenum R 2: tronigenum O: trouigenum G  prosternet OG: prosternens R 2  ferro GR 2: ferrum O  post 355 lacunam indicauit ϛ, probante Rossbach   357  magnis] magni Heinsius  359  densis Nisbet: caesis GR: cessis O: celsis Baehrens  360  flumina OR: lumina G  361  subtegmina OGR 2: sub tegmine R  post 361 lacunam posuit et suppl. Fröhlich exempli gratia : lacunam duorum uersuum posuit McKie, longiorem Müller   362–79  aliter ordinant alii: 362, 366–70, 363–4, 371–9 Hermes: 366–8, 365, 363–4, 369–74, 362, 375–7, 379 Herrmann  364 ­ perculsae] percussae ϛ  post 364 lacunam unius uersus posuit Voss  365  post 366 transp. Rossbach  subtegmina OGR 2: sub tegmine R  366  ac ϛ: hanc OGR  fors GR: fons O  copiam OR: copia G  368  madefient ϛ: madescent OGR  369 ­ succumbens OR: subccubens G, succubens G1: procurrens ϛ: procumbens Pleitner  370  post 371 transp. Voss  proiciet O : proiiciet GR  summisso] succiso Baehrens   371  subtegmina OGR 2: sub tegmine R  post 371 lacunam unius uersus posuit Rossbach  372  ambo McKie: animi GR: an̄ (ante) O : animis Parthenius: auidi Nisbet  373  coniunx] iuuenis Nisbet  

130

C AT ULLUS : POEM 64

dedatur cupido iam dudum nupta marito.   currite ducentes subtegmina, currite, fusi. non illam nutrix orienti luce reuisens hesterno collum poterit circumdare filo, anxia nec mater discordis maesta puellae secubitu caros mittet sperare nepotes.   currite ducentes subtegmina, currite, fusi.’   talia praefantes quondam felicia Pelei carmina diuino cecinerunt pectore Parcae. praesentes namque ante domos inuisere castas heroum et sese mortali ostendere coetu caelicolae nondum spreta pietate solebant. saepe pater diuum templo in fulgente residens, annua cum festis uenissent sacra diebus, conspexit terra centum procumbere tauros. saepe uagus Liber Parnasi uertice summo thyadas effusis euhantes crinibus egit, cum Delphi tota certatim ex urbe ruentes acciperent laeti diuum fumantibus aris. saepe in letifero belli certamine Mauors

375 377 379 380

385

390

375  subtegmina OGR 2: sub tegmine R  377  hesterno ϛ: esterno O: externo GR  378  currite ducentes subtegmina, currite, fusi om. ϛ, del. Voss, del. et alium uersum excidisse censuit Hermes  subtegmina OGR 2: sub tegmine R  378–80 aut 379–81  om. O  379  discordis] discordi Trappes-Lomax  381  currite (2o) O: ducite GR  382  Pelei] Peleo Tollius  383  cecinerunt ϛ: cecinere GR: cernere O: cecinere e Baehrens  384  non ante sed post praesentes interpunxit Earle  385 heroum et Sigicellus: nereus OGR: saepius et Calphurnius: uerius et Meleager  post 386 languidior tenera cui pedens sicula beta (~ 67.21) OGR: del. Parthenius  387  templo in fulgente residens Fisch: t. in f. reuisens OGR: fulgentia templa reuisens Muretus: templa illa in gente reuisens Fröhlich: templa in fulgore reuisens Peiper  388  cum ϛ: dum OGR  uenissent ϛ: uenissem O: uenisset GR  389  terra GR: terram O: terrae Leyser  procumbere] procurrere ϛ: percurrere nescioquis teste Ramler  tauros ϛ: currus OGR  390  summo GR:  sumo O  391  thyadas GR: thiadas O: thyiadas Sillig  euhantes] euantis OGR: ouantis R 2  egit GR: esit O  post 391 lacunam posuit et exempli gratia suppl. Müller  392–3 del. Koeler  392  certatim ϛ: certatum OGR  ruentes ϛ: tuentes OGR  393  acciperent ϛ: acciperet OGR  laeti] leti R 2 u.l.: lacti OGR   laeti diuum] uatem diuum aut laeti Phoebum Peiper  acciperent laeti diuum] acciperent Letonigenam aut Letoidem (uel Letoum) acciperent laeti Heinsius  fumantibus] spumantibus ϛ: fumantibus Baehrens  394  Mauors OR: mauros G  

131

C ATULLUS : POEM 64

aut rapidi Tritonis era aut Rhamnusia uirgo armatas hominum est praesens hortata cateruas. sed postquam tellus scelere est imbuta nefando iustitiamque omnes cupida de mente fugarunt, perfudere manus fraterno sanguine fratres, destitit exstinctos natus lugere parentes, optauit genitor primaeui funera nati, liber uti nuptae poteretur flore nouellae, ignaro mater substernens ilia nato impia non uerita est diuos scelerare penates: omnia fanda nefanda malo permixta furore iustificam nobis mentem auertere deorum. quare nec tales dignantur uisere coetus, nec se contingi patiuntur lumine claro.

395

400

405

395  rapidi] rapida Nisbet  Rhamnusia 1472, ϛ teste Della Corte: ranusia GR: ramu ˉ sia O: Amarunsia Baehrens (Amarynthia Goold )  398 omnes] homines Harrison*  400  post 401 transp. Rossbach  natus O: natos GR  401  optauit] captauit Mitscherlich: patrauit Baehrens  402  uti nuptae Mähly: ut innuptae OGR: ut hinc nuptae Baehrens: ut intactae Heyse: ut iniustae Papanghelis: qui innuptae Dyer  poteretur ϛ: potiretur OGR  nouellae Baehrens: nouercae OGR : iuuencae Nisbet*, Thonemann* : nuriclae Postgate   flore nouercae Courtney exempli gratia  403–4  ilia … | impia] Heyworth/Trimble* (impia … | ilia Reeve): se impia … | impia OGR: se improba … | impia Nisbet  404  penates ϛ: parentes OGR  406  iustificam OR: iusticiam G  mentem GR: mente O  auertere GR: aduertere O

132

CO M M E N TA R Y

C OMME NTA R Y

1–11: THE ARGO The opening of the poem seems to lead in the wrong direction. In the absence of any formal invocation of an inspiring divinity, the content of the first few lines would be expected to announce the subject matter of the whole work (Arist. Rhet. 3, 1415a11–24; cf. 1 quondam n.). Here, the extended opening sentence (seven lines, as at Il. 1.1–7, Virg. Aen. 1.1– 7) focuses on the Argo itself, and thus suggests two possibilities, neither of them the wedding of Peleus and Thetis. On the Argonautic myth and its appearances in Greek and Roman literature see RE ii.1.721–3 s.v. ‘Argo’, 743–87 s.v. ‘Argonautai’ (both Jessen), Gantz 340–73, Zissos (2008) xvii–xxv, and the works cited at Mac Góráin (2015) 233 n. 1. Already at Od. 12.70 Circe refers to Ἀργὼ πᾶσι μέλουσα, which may imply the existence of an epic Argonautica that predated the Odyssey (West (2005)); there were apparently extensive treatments of the myth in at least three lost archaic epics, the Corinthiaca ascribed to Eumelus, the anonymous Naupactia, and a poem by Epimenides, but Apollonius’ would have been the Argonautica best known to Catullus, as to us (perhaps together with the Latin version of Varro Atacinus, if that poem had already been written: see the Introduction, 1b). The first possibility, then, is that this epic will be an Argonautica: note the title ‘Argonautia’ added here by R2, presumably from his exemplar X, and copied by m2 and G2 (cf. also 323–81n.). An epic often identifies its main subject or protagonist in the first word (West on Hes. Theog. 1); here the geographical adjective Peliaco describes the ship’s origin, as Callimachus’ Hecale begins by calling its heroine Ἀκταίη (fr. 1 Hollis = 230 Pfeiffer). Other details personify the Argo as if it were itself the epic hero (1 prognatae, 2 nasse nn.). The lines allude to the opening and several other passages of Ap. Arg. (see esp. 1–3n. and cf. ­continued allusion in 12–18, 19–30). However, the story of Medea had become an increasingly important part of the Argonautic tradition, and was particularly associated with tragedy, above all because of the classic status of Euripides’ Medea (Mastronarde (2002) 44–70), which Ennius in turn had adapted into a Roman classic. This Argonautic opening also reworks the first few lines

135

COM M ENTAR Y: 1–11

of both E ­ uripides’ and Ennius’ Medea tragedies, making the alternative suggestion that this will be a Medea. The conventions of both epic and tragic beginnings are more widely exploited. The narrative movement is both wayward and repetitive: the Argo journeys from Pelion to Colchis (1–3), is en route (4–7), is built (8–10), sets off again (11). This could potentially be either a tragic prologue, which tends to set the scene by narrating past events, or an epic proem, which often summarises what will happen in the subsequent narrative (both sorts of summary being often highly compressed, with many anachronies: see 1n. and cf. e.g. Il. 1.1–7 with Genette (1980) 36–7). The allusions to particular Medeas in 1–7 initially suggest the tragic type, but 8–11 then return to the beginning of the Argonautic story, inviting a reinterpretation of 1–7 as a foreshadowing epic summary (Zetzel (1983) 261, Clare (1996) 62–5). Moreover, epic and tragic openings presuppose very different roles for the speaker: in tragedy, emotionally involved with a current situation and, at least if human, with limits to their knowledge about it; in epic, narrating a past story from a more objective and effectively omniscient viewpoint (cf. Race (1992) 13–16). Here, the first line evokes the wish of Euripides’ and Ennius’ Nurse that the story she narrates had not taken place (cf. 171– 6n.), but dicuntur then signals a different speech act, the distanced indirect statement of a Hellenistic poet-narrator (2 dicuntur n.); yet later there are hints of emotional colouring more appropriate to the tragic Nurse (5 auertere n.). This conflation of epic and tragedy reflects the interrelationships in the evolution of the two genres (see further the ­Introduction, 2c(i)). Scene-setting tragic prologues are where ‘tragedy comes closest to having a “narrator”’ (Segal (1992) 85), and Ennius’ Medea prologue is itself both a ‘dramatic’ and a ‘narrative’ opening: as in Euripides, Ennius’ first words indicate a wish (Eur. Med. 1 εἴθ’ ὤφελ’ ~ Enn. trag. 208 Jocelyn = fr. 89.1 FRL utinam), but subsequently the etymologising quia (212 = fr. 89.5) and temporally problematic nunc (211 = fr. 89.4 nunc nominantur nomine) imply the distanced position of an epic narrator, and a learned one at that (cf. Thomas (2021) 67 ‘Medea’s nurse sounds more like a scholiast’). Accordingly, Catullus might have written this opening to demonstrate his understanding of the debts owed by Hellenistic poetry to Euripides, and the influence of Hellenistic poetry, in turn, on Ennius (Zetzel (1983) 262–3), or indeed on Accius, whose Medea tragedy seems to have drawn on Apollonius (Arcellaschi (1990) 163–95; cf. 15n.). He might have been aiming to acknowledge Ennius as both a tragedian and an epicist, especially by simultaneously alluding

136

COM M ENTAR Y: 1–11

to a scene from the Annales (7 uerrentes n.); or to reconstruct the Argonautic myth that Ennius had demythologised (Biondi (1980) 133–9). He might even have wanted to open his innovative epic with a metapoetic voyage associated with the primal, pre-Odyssean Argonautica (Harrison (2007b) §3; see also 1 Peliaco n.). The relationships between this passage and its earlier intertexts, then, can be read as deliberate, even polemical ‘references’ to other works, intended to display the ­author’s learned understanding of literary history and of the text’s place in a tradition, as emphasised in the seminal article of Thomas (1982). Alongside such an approach, however – and without being as committed as Thomas is to the knowability of the author’s intention, on which see Hinds (1998) 17–51 on ‘philological fundamentalism’, with Thomas’ reply at Thomas (1999) 6–7 n. 10 and later reflections on the debate at Thomas (2021) 47–9 – it is possible to pay more attention to the ways in which the allusions and generic ambiguity here may convey meaning about the poem’s content (cf. Traina (1972) 104–5). The passage can be read as a new kind of Medea prologue in which tragic aspects, including Medea herself, are excluded in order to produce an optimistic introduction to a happier Argonautic love story ­ ­ (Klingner (1956) 6–12, Harmon (1973) 312). Alternatively, the ­presence through allusion of Medea and Jason may cast doubt on that happiness and complicate the contrast between the two ­couples, Peleus/Thetis and Theseus/Ariadne (e.g. Bramble (1970) 37–8, ­ ­Zetzel (1983) 260–2, Clare (1996) 60–5); the appearance later in the poem of the Medea-like figure of Ariadne (cf. esp. 132–201n.) is the strongest invitation to reread this opening as equivalent to a Medea prologue after all. Reading it as the proem to an Argonautica, however, is also productive. The Argonautic expedition is elsewhere understood in political terms by both Greek and Roman authors (Nelis (2012), citing in particular the work of Hunter (1993) on Apollonius’ epic in its Ptolemaic context; see also Mac Góráin (2015)), and Catullus may, as part of this tradition, be associating the Argonauts’ quest for the fleece with Roman imperial acquisitiveness (Feldherr (2007) 99–100), perhaps even with his own appropriation of Greek literary culture in composing this poem (Young (2015) 28, 49); cf. 31–49n. Moreover, Catullus’ Argo is the first ship in history (11n.). This idea is probably no earlier than Eratosthenes, Catasterismi 35 ( Jackson (1997), Dräger (1999), though see also Bär (2012) on the interpretation of Eur. Andr. 865 πρωτόπλοος), but for the Romans (see esp. Feeney (2007) 118–31) it became strongly associated with the originally separate tradition that the golden age had lacked ships (Hes. Op. 236–7, Arat. Phaen. 110–11)

137

COM M ENTAR Y: 1–3

and therefore that seafaring was characteristic of the iron age (see further 6 ausi sunt n.). The sailing of the Argo thus introduces the motif of the myth of ages, a central and particularly complex concern of the poem (see esp. 22, 31–49, 397–408nn.). The proem was influential on subsequent Roman descriptions of the Argo (on intertextuality in Argonautic literature more generally see e.g. Desbordes (1979), Davis (1989)), many of which can be read as responses to the interpretative possibilities just discussed. Prop. 3.22.11– 14 imitates Catullus’ vocabulary (Colchum, Phasin, Peliacaeque, rudis, natat, prorae, pinus) when connecting the Argo to Tullus’ contemporary travels; Ov. Am. 2.11.1–2 prima malas docuit mirantibus aequoris undis | Peliaco pinus uertice caesa uias evokes the passage in the opening to a propempticon in which he wishes that the Argo had never sailed (on the possible response to Catullus in Hor. Carm. 1.3, a propempticon not explicitly mentioning the Argo, see Putnam (2006) 107–11). See also 1–3n. on the proem of Valerius F ­ laccus’ Argonautica, the more incidental references cited at 1n., and esp. Stat. Ach. 1.64–5, in a context alluding to this passage and other parts of Catullus’ poem (12–18n.), where Statius’ Thetis sees the Argo as a symbol of impiety and greed (Ach. 1.65 Iasonia … rapina; cf. 5n.). 1–3  This epic opening, summarising a journey (cf. e.g. Virg. Aen. 1.1–7), signposts an Argonautica by involving itself in an intertextual debate about how much of the Argo’s journey an Argonautica proem should cover (see Clare (2002) 28–32). In Ap. Arg. 1–4 the ship travels through the Pontus towards Colchis with no explicit comment on whether it got there or succeeded in its quest: V.Fl. 1.1–4, by contrast, takes the Argo all the way to the heavens (see further Zissos ad loc.). Here, although 3 ad might mean only ‘towards’, the allusions suggest a similar extension of the Apollonian proem, moving from the Argo’s origin as in the Medea plays (1) to a reworking of lines from the completion of the outward journey at the very end of Ap. Arg. 2 (3n.). The concentration of proper names establishes the atmosphere of Greek myth in general and the world of an Argonautica in particular, with an opposition between Greece and Colchis (cf. 344–6n. for similar effects elsewhere in the poem). Geymonat (1982) analyses their ‘decorative’ effect, but also shows how names are identically distributed in 1–7 here and in Ap. Arg. 1.1–7 (1, 1, 1+1, 1, 1, –, –); the pattern constitutes another invitation to compare Apollonius’ proem. The lines also evoke another narrative opening with a swift voyage and proper names, 63.1–3; cf. 6 cita … puppi n. and the resemblances between Ariadne and Attis (esp. 50–75, 132–201nn.).

138

COM M ENTA R Y: 1

1  Beginning with the pines from which the Argo was built, rather than with its leaving Colchis and passing the Symplegades, Catullus follows Ennius’ ‘correction’ of Euripides’ hysteron-­proteron time sequence (Eur. Med. 1–6, Enn. trag. 208–11 ­Jocelyn = fr. 89.1–4 FRL). This fits an epic narrator better than the Nurse, who first mentions Medea leaving home because that is what she primarily regrets (cf. Rhet. Her. 2.34). Five-word lines are ­particularly favoured in the poem (I count 118 = 29% of 409; see further 7n. and the Introduction, 2d(ii)). The plosive alliteration of the first, third, and fifth words marks allusion to Eur. Med. 3–4 μηδ’ ἐν νάπαισι Πηλίου πεσεῖν ποτε | τμηθεῖσα πεύκη: the same alliteration in the same context at Theoc. 13.17–18, Ov. Am. 2.11.1–2, Her. 12.10, Met. 6.721, Phaedr. 4.7.10–11, Luc. 2.715–19, 6.400–1, Stat. Theb. 5.335–7, Ach. 1.64–5 (cf. also 278n.). Peliaco: Adjective based on a Greek name, as at 35, 36, 37, 75, 172, 178 (places), 3, 77, 267, 367, 368 (persons). Many replace an expected genitive: this common feature of Latin poetic style develops a native Latin idiom (for prose occurrences see K–S i.209–12) in imitation of Greek epic and tragic usage. See Fordyce on 44.10, Austin on Virg. Aen. 2.543, Harrison on Virg. Aen. 10.156–7, Löfstedt (19562) i.107–24, Wackernagel (1920–4) ii.68–75, (1953–79) ii.1346–73. Peliacus is first found here: next at Prop. 3.22.12 Peliacaeque trabis, Ov. Am. 2.11.2 (1–11n.), then in subsequent poetry. The Greek equivalent Πηλιακός only has one, later ancient attestation, at AP 16.110.4 Πηλιακοῦ δούρατος. Weber (1987) 265–6 n.  33 shows that such Greek-derived adjectives in ‑iacus in particular were later felt to have a ‘neoteric aura’, and are often neologisms: cf. Prusiaca at Cinna fr. 11.4 Courtney = 13.4 Hollis. Catullus replaces the adjective Pelius found in the first line of ­Ennius’ Medea, trag. 208 Jocelyn = fr. 89.1 FRL utinam ne in nemore Pelio, and attested in Latin poetry as early as Liv. Andron. trag. fr. 27 TrRF haut ut quem Chiro in Pelio docuit ocri. Mount Pelion is associated with the Argo’s origin almost unanimously (see McKeown on Ov. Am. 2.11.1–2). Since Pelion is an important location for Peleus and Achilles, as the home of Chiron (cf. 278–9nn.), the source of the wood for Achilles’ spear (Il. 16.143–4 = 19.390–1, recurring Πηλιάς μελίη), and even the normal setting of Peleus’ and Thetis’ wedding (31–49n.), the word also hints at the real theme of the poem’s outer story, possibly even punning on Peleus (see Stoevesandt (1994–5) 193–4 and Fo here). The hero is – almost – named in the first word of the poem after all (1–11n.), and perhaps we even hear an echo of Il. 1.1 Πηληιάδεω Ἀχιλῆος, pointing to the literary ambition of Catullus’ epic and to the Achillean and Iliadic themes which will emerge towards the end of the poem (see esp. 323–81n. and cf. Mac Góráin (2015) 247).

139

COM M ENTAR Y: 1

quondam partly works with Peliaco and prognatae to contrast line 1 with lines 2–3 in time and space: the pines, originally born on Pelion, later swam all the way to Colchis. But the main force of the word here is to indicate the narrator’s distance from his subject, set in an age too long past to be specified – ­appropriately enough, in this poem in which chronology is often confusing (Feeney (2007) 268 n. 98; see e.g. 19–30, 31–49nn. and below on Ap. Arg. 3.997). Baehrens, Godwin, Traill (1981) 237, Jenkyns (1982) 98–9, e.g., compare fairy tales and/ or translate ‘once upon a time’: either quondam or olim may appear with this sense at the beginning of any kind of story, anecdote, or exemplum (e.g. 67.4, 68.73, 111, Cic. Ver. 2.4.72 hoc quondam oppidum, Plaut. Stich. 539 fuit olim, quasi ego sum, senex), while the indeterminacy of time could be compared to indeterminacy of place and/or characters’ names at other narrative openings, e.g. Apul. Met. 4.28.1 (Cupid and Psyche) erant in quadam ciuitate rex et regina (see Kenney ad loc.), Petron. 111.1 matrona quaedam Ephesi tam notae erat pudicitiae. However, quondam here clearly draws on the use of ποτέ (κοτέ, ποκά) at the beginning of Hellenistic poetic narratives, whether free-standing, especially short narratives that might be called ‘epyllia’ (Call. Hec. fr.  1 Hollis = 230 Pfeiffer, Theoc. 18.1, 24.1, Mosch. Eur. 1; cf. Call. Ia. 8 fr. 198 Pfeiffer, Theoc. 6.2, [Theoc.] 19.1), or inset within larger structures (Call. Hymn. 5.57, Aet. fr. 75.4 Pfeiffer/Harder, Ia. 4 fr. 194.6 Pfeiffer, [Bion] 2.10, Sotadea fr. inc. 16 Powell). See Bühler and Campbell on Eur. 1, Hollis on Hec. fr. 1, and cf. esp. Ap. Arg. 3.997, where Jason uses δή ποτε at the beginning of the story of Ariadne as he tells it to Medea, thus obfuscating (or highlighting) the tendentious chronology there (53 cum classe, 116–31nn.; DeBrohun (2007) 300). Race (1992) 14 traces the technique back through Euripidean prologues (Andr., El., HF, Bacch., Med. 3 Πηλίου … ποτέ; cf. Soph. Phil. 5) to Bacchyl. 20 Snell–Maehler; see also Nelson (2023), index s.v. ποτέ, cf. esp. Od. 8.76 ποτέ at the beginning of Demodocus’ first song, and see the Introduction, 2c(i). ­Peliaco quondam adopts two p ­ articular characteristics typical in this tradition, the temporal adverb as second word and the conjunction with a name: all the Hellenistic examples above show one or the other, while the closest parallels are Mosch. Eur. 1 Εὐρώπῃ ποτέ and, significantly (1–11n.), Call. Ia. 8 fr. 198 Pfeiffer Ἀργώ κοτ’. Kubiak (1981) argues that Catullus also has the Latin precedent of Cic. Arat. 420 Soubiran uir quondam Orion, opening the story of Orion. quondam is there elided, however, as are many of the ποτέ examples, whereas here it is emphasised by the caesurae preceding and following it; C ­ atullus never elides quondam (eight occurrences), and it often receives similar emphasis for wistful or sarcastic weight when used by a speaker remembering a

140

COM M ENTA R Y: 1

time in his/her own past (e.g. 8.3; cf. Clausen (1982) 192–3). As it has these connotations elsewhere in the poem (139), and when Catullus again uses it in this position in an opening hexameter (72.1), some emotional charge may be felt even in this more ‘antiquarian’ context, preparing for the narrator’s nostalgia for the heroic age (22–30) and implying the contrast with the present that is explicit when quondam returns in ring-­composition at the beginning of the e­ pilogue (382n.). Cf. also olim below (76, 212). prognatae ‘sprung from’. A formal, archaic word first attested in the early third-century epitaphs of the Scipios (CIL i 2.7.3, 10.9), prognatus occurs comparatively often in early epic and tragedy (e.g. Enn. Ann. 36 Skutsch/FRL, trag. 291 Jocelyn = adesp. fr. 55.1 TrRF   ); its fourteen occurrences in Plautus and three in Hor. Serm. are in contexts of either real or parodied solemnity. It is especially connected with socially lofty or divine/heroic ancestry: cf. e.g. Liv. 1.40.3 deo prognatus (Romulus), Cael. ap. Cic. Fam. 8.15.2 Venere prognatus (Caesar, facetiously). Accordingly, despite a few prose examples of its use of plants, animals, and objects (OLD s.v. 1), its effect here is to personify the pines as heroes of noble origin. Cf. the similar metamorphosis from trees to ship at 4.10– 17, and the pine ship proud of its ancestry at Hor. Carm. 1.14.11–12. The rhythm with a molossic word or combination after a strong third-foot caesura is particularly characteristic of the poem (the ‘monumental molossus’): see the Introduction, 2d(ii), and for some effects of clusters of such lines, of which the first occurs in 1–4 here, see also 38–42, 63–5, 91nn. uertice ‘summit’, but among the other anthropomorphic details here perhaps also ‘head’ (cf. 63, 309, 350); this might align the pines’ origin with the miraculous birth of Minerva, anticipating her appearance in 8–11. pinus: Singular pinus for ‘ship’ is later common in poetry (e.g. Virg. Ecl. 4.38, Prop. 3.22.14, Ov. Met. 1.95, Sen. Med. 336), but here the plural is a pivot, still referring to the trees when only line 1 has been read, but then to the ship in line 2; the metonym is made as the ship is made. Cf. Virg. Aen. 10.230–1 nos sumus, Idaeae sacro de uertice pinus, | nunc pelagi nymphae, classis tua with Hardie (1987) 163–4; the line-ending uertice pinus (‑um) is also reprised at Ov. Met. 10.103, Luc. 1.573, Petron. 131.8.3. The avoidance of ordinary nautical words in this opening perhaps focalises the description through the watching Nereids, but this is not obvious before their appearance: by mimicking their ignorance, the narrator paradoxically puts himself on the same level as his characters, despite his concern elsewhere in these lines to stress his distance from the story (2 dicuntur n.). See also 7 abiegnis n.

141

COM M ENTAR Y: 2

2 dicuntur: The subject matter of the poem is introduced as something known to the narrator from an inherited tradition: similarly 19 fertur and, marking narrative digressions in the ecphrasis, 76 perhibent, 124 perhibent, 212 ferunt. Elsewhere in Catullus cf. also 2b.1, 68.109. Here the verb of speech, emphatically placed, works together with the temporal adverb quondam, as at 76, 212, 68.109–11, Call. Aet. fr. 75.4 Pfeiffer/Harder, Sotadea fr. inc. 16 Powell, Cic. Arat. 420–1 Soubiran, Prop. 1.20.17. Such appeals to tradition are found in Homer (φασί at e.g. Il. 2.783, 5.638, 6.100, Od. 3.245, 6.42) and continue throughout the subsequent Greek tradition (Nelson (2023) 73–175); see Nisbet– Hubbard on Hor. Carm. 1.7.23 fertur and esp. Norden on Virg. Aen. 6.14, with one way of classifying their different uses. Modern scholarship has particularly emphasised the cases in which the generalised appeal to oral tradition (‘is/are said’, ‘they say’) masks a reference to the specific literary predecessors known to the well-read poet: Ross (1975a) 78, with reference to the second of Norden’s four categories (‘alexandrinische Manier’), called this the ‘Alexandrian footnote’, a term adopted in particular by Hinds (1987a) 58 and n. 22, (1998) 1–3. On this understanding, the narrator portrays himself as a scholar citing one or more earlier texts, and invites the reader to find, admire, and interpret nearby allusions, as Hellenistic narrators do at e.g. Call. Hymn. 5.56, fr. 612 Pfeiffer, Ap. Arg. 1.18. However, as was already clear from Norden’s treatment, this is only one of the effects that ‘they say’ can have (see Nelson (2023) 6–27 esp. 9–12); it can, for instance, indicate that the subject is generally well known, its details perhaps contested in different traditions or texts (cf. 1–11, 76, 212nn.), or it can draw attention to a narrative innovation (19, 124nn.). It may even produce a distancing or fictionalising effect, as the narrator avoids taking responsibility for the truth of his story: dicuntur here is read thus by Jenkyns (1982) 99, Gaisser (1995) 582. For further discussion of this example see also the bibliography cited by Fernandelli (2012) 20 n. 72. liquidas … undas ‘flowing waters’: in epic style a natural object is given an obvious epithet describing an appealing physical feature. It is less Homeric, more Apollonian to divide epithet from noun rather than making the combination fill a half-line before or after the main caesura: see Parry (1971) 24–9 and the Introduction, 2d(ii). The pattern in which the adjective and noun in agreement are placed immediately before the caesura and at the end of the line is common in the poem (I count 85 examples = 21% of 409 lines, of which only 246, 269, 286, 302, and 402 (a conjecture) have the noun first), and this version, in which the adjective and noun also rhyme, is particularly noticeably so (of the 85 examples I count 47 rhymes = 11.5% of 409; according

142

COM M ENTA R Y: 3

to Thomas (1982) 155 n. 43 this is more than double the proportions found in Enn. Ann. or Virg. Aen.). After Catullus, liquidus becomes a stock poetic epithet for the sea, and liquidas recurs with undas at Virg. Aen. 5.859, Ov. Met. 1.94–5, the latter evoking these lines to describe the end of the golden age: nondum caesa suis … | montibus in liquidas pinus descenderat undas. Neptuni: Already an established metonym for seawater (cf. 31.3 uterque Neptunus and e.g. Enn. Ann. 516 Skutsch/FRL imber Neptuni), this name can be taken as metonymical here, but need not be: it is also the first indication of the closeness of the gods to the physical world in the heroic age and particularly of the sea as a divine element before it suffered human invasion (Syndikus (1990) 121; cf. Curran (1969) 175–6). Cf. 11 Amphitriten n. nasse: Cf. 4.3 natantis … trabis, 66.46 nauit of sailors. Although later quite commonly used of travel by sea (OLD s.v. 2b; an etymological connection with nauis was probably felt, on which see Maltby s.v. no), the word is particularly unexpected and personifying here because of the contrast with the verb used of the trees in the Medea prologues, ‘fall’ (πεσεῖν/accidisset). With the contracted form cf. esp. 174 religasset, 176 requiesset (Catullus never uses the uncontracted form with ‑uiss- in a perfect infinitive or pluperfect subjunctive) as well as 97 iactastis, 120 praeoptarit, 151 dessem, 319 custodibant, 398 fugarunt, and cf. Bailey (1947) i.83–4 on Lucretius’ preference for contracted forms: Catullus’ too may owe something to contemporary usage as well as tractability in verse. 3  The line ostentatiously combines elements of three consecutive lines from the completion of the Argo’s journey in Apollonius: the chiasmus of exotic names echoes Arg. 2.1277–8 Κολχίδα μὲν δὴ γαῖαν ἱκάνομεν ἠδε ῥέεθρα | Φάσιδος, while Aeeteos translates 1279 Αἰήταο (also at spondaic line-end). Phasidos ad fluctus, however, reintroduces Euripidean word order (Eur. Med. 2 Κόλχων ἐς αἶαν). For the elegantly chiastic nounphrase cf. also 36 Crannonisque domos ac moenia Larisaea, 84 naue leui nitens ac lenibus auris, 218 fortuna mea ac tua feruida uirtus, 229 nostrum genus ac sedes defendere Erecthei; Catullus seems to prefer ac in such contexts (Ross (1969) 28–30) but presumably chooses et here after ad. This is the first spondeiazon of 29 in the poem (in my text), the highest concentration in any extant Latin work (7% of 409 lines); see the Introduction, 2d(ii), for their distribution, and see further Salvatore (1965) 257–67. The rhythm was not rare in Homer and Hesiod, but cultivated by Hellenistic poets (Crowther (1970) 322 n.  2, Lightfoot (1999) 56, Hollis (20092) 17–19); it was famously characteristic of νεώτεροι for Cicero (Att. 7.2.1: see the Introduction, 1b and

143

COM M ENTAR Y: 4

cf. 84n.). Both in Greek and in Latin many spondeiazontes end with names, often t­etrasyllabic and usually Greek even in Latin poetry; in others, the rhythm can underline the sense, especially of a verb (e.g. 15 admirantes). The form with spondaic fourth foot is particularly rare (elsewhere in Catullus only 44, 68.87, 116.3; never in Callimachus; see further West (1982) 154). This suggests expressive force even with the name: as a spondeiazon can form a ‘closing cadence to a kind of “paragraph”’ (Thomson here), the rhythm at the end of this line may depict the end of the voyage (see below on fines). Phasidos: R2 restores a reference to the main river of Colchis, flowing from the Caucasus to the Black Sea, which had been known in poetry since Hes. Theog. 340 and is associated with the Argonauts at Pind. Pyth. 4.211–13, repeatedly in Ap. Arg. The Greek genitive, directly from Ap. Arg. 2.1278, contrasts with Latin Thetidis in 19, 302 (if the transmitted readings are correct: see 21 Thetidi n.). Fordyce here lists all Catullus’ uses of Greek inflections (as transmitted or usually restored); scribes regularly corrupt such endings to Latin forms (see the Introduction, 2d(i) and n. 197). According to Var. LL 10.70 the trend to use Greek rather than Latin endings for Greek names began with Accius. There is a likely allusion to this line at Manil. 4.517 Phasidos ad ripas et Colchida tergore uexit, describing the earlier journey of the golden ram (now the constellation Aries); see Biondi (1981) 110–11. fluctus: The periphrastic form ‘[name of river]’ + ‘streams’ is Homeric (Il. 2.461, 2.533, 7.135, etc.), followed by e.g. Ap. Arg. (including 2.1277–8: see above), and in Latin epic since Ennius (Ann. 222 Skutsch/FRL sulphureas … Naris ad undas). fines: The ‘boundaries’ of Aeetes’ territory, but perhaps also the ‘end’ of the Argo’s journey; a teasing word at the beginning of a poem, and in a reference to such an archetypal beginning as the sailing of the Argo. Cf. 31 finito tempore n. Aeeteos: Another early restoration (first Aldine edition, following Parthenius): the adjectival form of the Colchian king’s name appears at Ap. Arg. 2.1279 and is not elsewhere extant in Latin. The orthography ‑eos (not ‑aeos) follows Housman’s principle (Housman (1972) 887–91) that while Greek adjectives formed from first-declension feminine names end in ‑αῖος (Latin ‑aeus), those formed from other names have ‑ειος (‑eus). 4 cum: A new clause is tacked on to a sentence that might have felt finished at the end of line 3. cum here with the indicative (6 ausi sunt) is a loose connective, simply indicating that the action of 4–7 took place

144

COM M ENTA R Y: 5

simultaneously with that conveyed by the indirect statement of 1–3: cf. 344–6n., 350–1, 363–4 and see LHS ii §333c. The temporal clause eases the transition to finite main verbs that is complete at 9 fecit (see n.). lecti iuuenes: Groups of ‘chosen’ warriors are frequent in epic and in military narrative generally (e.g. Livy); cf. esp. the echo of this line at Aen. 8.179 tum lecti iuuenes. For the Argonauts as ‘chosen’ cf. e.g. Pind. Pyth. 4.189, Ap. Arg. 3.347, 4.831, Theoc. 3.17–18 (below), Virg. Ecl. 4.34: lines 4–5 draw closely on Enn. trag. 212–14 Jocelyn = fr. 89.5–7 FRL quia Argiui in ea delecti uiri | uecti petebant pellem inauratam arietis | ­Colchis (contrast Eur. Med. 5 ἀνδρῶν ἀρίστων). The change from E ­ nnius’ delecti (perhaps a case of simplex pro composito, on which see 35 linquunt n.) allows for the possibility of a metapoetic pun on ‘read’ young men, another marker of allusiveness after 2 dicuntur n. (so Young (2015) 48; cf. Prop. 3.22.12 Peliacaeque trabis totum iter ipse legas with Heyworth–Morwood ad loc.). lecti iuuenes … robora pubis may also combine two descriptions of the Argonauts in Theoc. 13 (17–18 ἀριστῆες … προλελεγμένοι, 27–8 θεῖος ἄωτος | ἡρώων); for similar combinations of concrete and abstract cf. Lucr. 1.86 ductores Danaum delecti, prima uirorum (cf. 368n.), Virg. Aen. 5.729 lectos iuuenes, fortissima corda and see 78n. Argiuae: Perhaps a Homerising way of referring to the Greeks in general, but also suggesting Ennius’ derivation of Argo from Argiui uiri (trag. 211–12 Jocelyn = fr.  89.4–5 FRL); cf. Manil. 1.694 Argiuumque ratem. Where Ennius had been explicit, Catullus never mentions the name of the Argo, but alludes to possible etymologies: he excludes Apollonius’ (9n.) but (pace Thomas (1982) 150) does not choose between Ennius’ version and derivation from the Homeric use ἀργός = ‘swift’, evoked by 6 cita (Traina (1972) 99–105; Gaisser (1995) 584 n. 17). robora pubis: Poetic plural version of robur = ‘strongest ­element’ of a population; its literal meaning ‘oaks’ assimilates the men to their wooden ship (cf. 7 palmis). For pubis cf. 267, 68.101, in both cases used with a national adjective, as here and often (e.g. adesp. fr. 46.2 TrRF Attica pubes); it especially denotes the young adult population of military age (OLD s.v. 1). The diction of the whole line is evoked at Virg. Aen. 8.518–19 robora pubis | lecta, Stat. Theb. 1.606 lectis iuuenum, qui robore primi. 5  auratam … pellem: The first example of twenty-six in the poem of a noun and its adjective beginning and ending the line (e.g. 54, 72, 99), always with the adjective first; the pattern suits the poem’s characteristic single-line sense-units, and is similarly frequent in Cicero’s poetry and

145

COM M ENTAR Y: 5

in Ciris and Culex. See May (1910) 2–7, Conrad (1965) 224–9, Pearce (1966), and the Introduction, 2d(ii). The golden fleece is mentioned prominently in both Argonautica and Medea proems (Ap. Arg. 1.4, Eur. Med. 5, Enn. trag. 213 Jocelyn = fr. 89.6 FRL): this earliest extant use of auratus = ‘golden’ rather than artificially ‘gilded’ draws on Ennius’ pellem inauratam. As inauro was a more specific technical term for ‘gild’ (TLL vi i .1.841.25–842.18), Catullus may be ‘improving’ Ennius’ epithet to a more poetic level of diction: later poets describing the fleece use auratus (Ov. Her. 6.2, Manil. 1.263, 2.212, 5.377, Sen. Med. 983, V.Fl. 5.490) or aureus (Prop. 2.26a.6, Ov. Fast. 3.876, Sen. Tro. 1036–7, Med. 361, V.Fl. 5.200–1, etc.), while inauratus recurs of it only at Hyg. Fab. 3.1–2. optantes: Recurring optare associates desire for love and marriage (31, 120, 141, 328, 372; cf. 62.30, 66.79) with desire for heroic actions (also 82) and with the narrator’s desire for the heroic age (22). The addition of a new detail in a p ­ articipial phrase is common in the poem: Norden (19574) 380 contrasts the preponderance in Virgil of finite verbs in parenthesis. ­Norden calls the participial construction prosaic (followed by Kroll and Fordyce here), but Schiesaro (2022) 593–6 shows that the preference for participles, especially deployed in groups, has precedent in older Latin poetry and is paralleled in Lucretius, while Cicero uses participles more frequently in his poetry and his most ‘literary’ prose than in his letters. The influence of Greek syntax is likely (Coleman (1999) 84–5; cf. Coleman (1975) 118–19). Particularly notable in this poem is the pattern seen here, in which the participial phrase follows the main verb, often in fact following what could have been a completed sentence (cf. 4 cum n.), and occupies the whole line: cf. e.g. 7, 10, 18. On this see esp. Kayachev (2020) 10–11, who notes the high density of such lines in Cicero’s poetry, and cf. Pearce (1966) 140–1. In such lines, even more than in participial phrases generally, ‘[t]he distinction between a state described and an action narrated is blurred’ (Coleman (1999) 84), an interesting effect in a poem which insistently explores the relationship between description and narration (see esp. the Epilogue to the c­ ommentary). Colchis: Either ablative of source, ‘from Colchis’ (see OLD s.v. Colchus 2b for the use of Colchi as the place name), or dative of disadvantage in the more literal sense ‘from the Colchians’ (OLD s.v. 2); cf. 392 Delphi n. auertere: The usual word for taking something away as plunder in war, even ‘stealing’; possibly glossing Ennius’ per dolum (trag. 214 Jocelyn = fr. 89.7 FRL; see Traill (1981) 238 n. 13). As it literally means ‘to

146

COM M ENTA R Y: 6

turn aside’ and is sometimes used of such praeda as cattle or horses (e.g. Virg. Aen. 1.472, 8.208, Liv. 1.7.5), it may remind the reader that this object came to Colchis as an animal. Cf. also 406 auertere n. 6  ausi sunt: The heroes are daring not in confronting Aeetes but in confronting the sea, always seen as perilous in the ancient world (e.g. Hes. Op. 618–21, 663–93, Prop. 3.7.29–42). Helping to imply that this was the first voyage in history (11n.), the words may particularly activate the association of the invention of seafaring with the end of the golden age (1–11n.) rather than a heroic-age achievement: see also Nisbet– Hubbard on Hor. Carm. 1.3.12, Smith on Tib. 1.3.37–8, and cf. 22, 31–49nn. audacia is often stressed in such contexts, as Bramble (1970) 35–7 emphasises, arguing that Catullus’ Argonauts are thus characterised as impious here. However, Bramble’s main examples are a propempticon (Hor. Carm. 1.3.25) and a specific tragic situation (Sen. Med. 301, 318, 347), and the tone of the verb is less easy to categorise than that of the noun, especially within epic narrative here and at V.Fl. 1.3. On the audacia of the Argo see further Davis (1989). uada salsa: Again, dignified diction from old Latin poetry: uada (cf. 58, 63.47) = ‘waters’ in general rather than ‘shallows’ first in Accius (trag. 698 Dangel), while Ennius has mare salsum (Ann. 453 Skutsch/ FRL), aequora salsa (praetextae 2 FRL = scen. 367 Vahlen2); see also 67 fluctus salis n. The collocation uada salsa recurs in classical Latin only in the precise repetition of this phrase at Virg. Aen. 5.158 and its echo at Sil. 17.155. This is the second of eight references to the sea in 1–18, none using the everyday word mare largely avoided in the poem (though cf. 30, 155, 269, and 16 marinas). cita … puppi: The synecdoche puppis = ‘ship’ is first attested here. At Cic. Arat. 389 Soubiran, cited by Kroll and Klingner (1956) 8 as an earlier example, it has the standard meaning ‘stern’, since Cic. Arat. 126–38 Soubiran (cf. Arat. Phaen. 342–52) explains that the constellation Argo travels through the sky stern-foremost. This is also probably the first ‘swift ship’ (Homeric νῆα θοήν, Il. 1.308 etc.; cf. Ap. Arg. 1.111) in extant Latin epic, though cf. Enn. trag. 43 Jocelyn = fr. 151a.12 FRL classis cita and 63.1 celeri rate (1–3n.), as well as Var. At. fr. 4 Courtney = 124 Hollis celeris … carinae (cf. 9 currum n. and on the dating of Varro’s Argonautae see the Introduction, 1b; the attribution of this fragment to Varro’s poem is not completely certain). For the etymological force of cita see 4 Argiuae n. decurrere: A problematic transitive use with accusative uada salsa. decurrere is generally intransitive meaning ‘run down’ (e.g. Virg. Ecl. 5.84),

147

COM M ENTAR Y: 7

and while there are parallels for currere used t­ ransitively with an accusative word for ‘sea’ (e.g. Virg. Aen. 3.191 currimus aequor, 5.235 aequora curro), in maritime contexts decurrere is usually constructed with ‘sea’ in the ablative and means ‘come in to land’, equivalent to κατατρέχειν (e.g. Virg. Aen. 5.212), whereas this seems to describe an unspecified point during the course of the voyage. Baehrens and Quinn link the transitive use and the sense of ‘race across’ to chariot-racing language, preparing for 9 currum, but even these uses are not precise parallels (Thomas (1982) 153–4): each example at TLL v. 229.21–6 ‘refers to a course which has been carried completely through to its conclusion’, as does the metaphorical use at Lucr. 4.1196 cited by Nuzzo. It seems then that Catullus’ use here is unparalleled, possibly intended to convey the idea of ‘running down a slope of water’ (Ellis) although not to shore. ­Thomas (1982) 154 suggests the influence of an orthodox use of κατέδραμε at Call. Aet. fr. 108 Pfeiffer/Harder, describing the Argo coming into harbour; Polt (2012) points to Ap. Arg. 1.390 κατόλισθε, of the Argo slipping down into the sea (he also compares V.Fl. 1.186 decurrunt, describing the same moment). Alternatively one could ­consider percurrere (Harrison), of which this would be a more standard use (cf. e.g. Lucr. 1.287–8 altum | murmura percurrunt caelum, Ov. Met. 10.655 segetis canae stantes percurrere aristas); but corruption to decurrere among so many surrounding p s might seem unlikely. 7  The mingling of wood and water is highlighted by the structure of the line, one of several possible arrangements of two adjectives and two nouns separated by one verb (see Kroll here and the Introduction, 2d(ii)), the most famous being the ‘golden line’ (59n.): adjective a, adjective b, verb, noun A, noun B. This variation, a verb bAB, also occurs at 10, 113, 162, 245, 295, 332, 345: in each case the verb is a present participle. See also Conrad (1965) 235–6, who compares Cic. Arat. 103, fr. i i i .1 Soubiran (with finite verbs). caerula: caerul(e)us, already frequent in Ennius, is commonly applied to both sea and (dark, Mediterranean) sky (André (1949) 162–71). uerrentes: The metaphorical verb works with aequora to indicate sweeping the sea’s surface (12 aequor n.); similarly in Lucretius (5.388, 1227, 6.624) and Virgil (Georg. 3.201, Aen. 3.290, 5.778, 8.674). It is first attested of rowing at Enn. Ann. 377 Skutsch/FRL; Thomas (1982) 156–7, 160–1 traces six reminiscences of Ann. 376, 377–8 in 6–13 here (also salsa, caerula, uolitantem, carinae, spumis; see also Perrotta (1931) 183, Avallone (1944) 18–20). Cf. 58, 269nn. for further potential allusions to the same passage. There may be still more at 4.3 impetum, 5

148

COM M ENTA R Y: 8

uolare (Davis (2002) 119–20); for links between Catullus’ ­descriptions of the phaselus and the Argo cf. 1 prognatae, 2 nasse, 7 palmis, 11 imbuit nn. and see further Massaro (2010). Ennius was describing the sailing of a (Roman?) fleet, possibly at the battle of Myonnessus in 190 bc ; we do not know enough about the wider context of the surviving lines to be able to tell whether the allusions here import more than a Roman epic atmosphere (cf. Zetzel (1983) 255, 257–8), though it may be relevant that Virgil evokes the Ennian passage still more closely at Aen. 8.91–3 (Thomas (1982) 160–1). See also 1–11n. on Roman seafaring and the Argo. abiegnis: Yet another kind of wood after 1 pinus (cf. 10 pinea), 4 ­robora. This specificity would suggest a display of learning even to a reader unaware of Catullus’ ‘correction’ of Ennius in reassigning firwood to the Argo’s oars (made of ἐλάτη, the usual equivalent of abies, at Ap. Arg. 1.914 etc.) rather than the ship itself. The Argo is made of fir (abiegna) at Enn. trag. 209 Jocelyn = fr. 89.2 FRL, but consistently of pine elsewhere (see Jocelyn ad loc. and Thomas (1982) 146–8), including at Eur. Med. 4 πεύκη. palmis: Cf. palmulae at 4.4, 17; both palma and its diminutive are attested a few more times of oars (TLL x .1.148.76–9, 155.58–71), with one further appearance in classical poetry (Virg. Aen. 5.163 palmula). Anthropomorphic play continues (cf. 1 prognatae, uertice, 2 nasse, 4 robora), but the image is ambiguous: either rows of oars spread out like the fingers of a hand, or oar-blades as hands on the ends of arms. 8  The line is a periphrasis for a cult title meaning ‘city-guardian’, e.g. Πολιοῦχος, Ἐρυσίπτολις (see below), Πολιάς, Ἀκραία; cf. Liv. 31.30.9 praesidemque arcis Mineruam. For other antonomastic designations of Athena/Minerva cf. 228, 395, of other goddesses 96, 300, 395. quibus: The connecting relative is not prosaic (Axelson 48–9); see further 31 quae simul n. on this recurrent feature. The postponement of the relative pronoun or adjective is also regular in the poem (30 qui, 48 quod, 56 quae, etc.) and is a widespread poetic feature that may go back as far as Ennius (Ann. 469 Skutsch/FRL): see Harrison on Virg. Aen. 10.531, with f­ urther refs, and contrast 43 at n. retinens ‘occupying’, mainly a prosaic, military sense (OLD s.v. 8) but applied at Lucr. 4.412–13 terrarum milia multa | quae uariae retinent gentes to an area’s usual inhabitants. The simple teneo has a wider range of meanings along the lines of ‘occupy’, ‘inhabit’, ‘rule over’ (OLD s.v. 7–9; cf. 67.4 cum sedes ipse senex tenuit) and is elsewhere attested of gods (e.g. Hor. Carm. 3.4.62, 3.26.9, 3.28.14).

149

COM M ENTAR Y: 9

in summis urbibus arces: True plurals, as at Virg. Ecl. 2.61–2 Pallas quas condidit arces | ipse colat: Athena was ‘city-guardian’ in many places besides Athens (e.g. Il. 6.305 Ἐρυσίπτολις, Troy; Hdt. 1.160.3 Πολιοῦχος, Chios). But cf. 241 summa … ex arce, of the Athenian Acropolis: this line may be a veiled programmatic indication that the poem will include Athenian myth. 9  ipsa … fecit: The opening indirect statement (1–3) and temporal clause (4–7) are succeeded here by finite main verbs; for the moment the narrator does not continue to remind us of the source of his story. As a piece of narrative technique this is nothing unusual (cf. e.g. the Homeric narrator handling the opening of Demodocus’ song at Od. 8.266–71), yet worth noting in a poem that draws attention to its narrator’s knowledge and authority: see esp. the Epilogue, Section 3. The narrator adopts an awed tone (cf. ipse at 21, 25–6) as he evokes Apollonius’ description (Arg. 1.111–12 αὐτὴ γὰρ καὶ νῆα θοὴν κάμε, σὺν δέ οἱ Ἄργος | τεῦξεν) but suppresses the name of the shipbuilder Argus, ascribing the building to the goddess alone. This rules out Apollonius’ implicit etymology (also Arg. 1.18–19), strengthens the emphasis on the gods’ presence among men during the heroic age, and suggests that the voyage was divinely approved, not hubristic (6 ausi sunt n.). The ship is now seen as an artefact, made not born: contrast 1 prognatae. leui … flamine: The ease and speed of the miraculous ship: cf. 84n. and e.g. Lucr. 4.901–2, a heavy ship moved by uentus … tenuis. Catullus may draw inspiration from the fair winds at Ap. Arg. 1.423–4 ἀήτης | μείλιχος, 566 λιγὺς … οὖρος (Westwood (2022)). The wording, with ipsa at the beginning of the line, is echoed at Manil. 5.565–6 ipsa leui flatu … | aura, of the wind refreshing the chained Andromeda; for the influence of this poem on Manilius’ Andromeda episode see the Introduction, 3a(ii), and cf. esp. 14n. uolitantem: A new image, the Argo as a bird, its sails (rather than oars, given flamine) standing for wings; cf. 12 rostro, literally ‘beak’. A ship ‘flies’ (uolat) at Enn. Ann. 376 Skutsch/FRL (7 uerrentes n.), and Lucretius has 4.390 quos agimus praeter nauem uelisque uolamus, while uolitare is attested of ships in fr. 4 of Furius Antias, a late second- or early first-century poet criticised by the grammarian Caesellius Vindex specifically for his neologistic verb forms (Courtney (20032) 97–8). currum: Of a ship only here in extant Latin; ὄχος, ὄχημα, and ἀπήνη sometimes have this meaning in Greek tragedy (e.g. Aesch. Suppl. 32, [Aesch.] Prom. 468, Soph. Trach. 656, Eur. Med. 1123, IT 410), but usually with a clarifying epithet, e.g. νάϊον. Cf. however auriga of a

150

COM M ENTAR Y: 10–11

­ elmsman at Var. At. fr. 4 Courtney = 124 Hollis (cf. 6 cita … puppi n.), h Ov. Tr. 1.4.16, with Courtney on the Varro fragment for the wider metaphorical association. This is the culmination of the metaphors avoiding nautical terms (1 pinus n.): more standard references began with with 6 puppi, and follow after this line with 10–13 carinae, rostro, remigio. Disconcertingly, the narrator is simultaneously a learned poet imitating a Greek mannerism and a naïve observer lacking the word for ‘ship’. 10  The line expands on Minerva’s carpentry (no details in Ap. Arg.) in a miniature version of Homeric practice (e.g. Odysseus building his raft, Od. 5.244–61). pinea … texta: Cf. 292 contexta of similarly miraculous craftsmanship with trees. Here, a more realistic understanding is possible: the ribs of the carina (warp) are crossed perpendicularly by the curved slats (weft) fitted on to form the ship’s sides. V.Fl. 1.125 clarifies and humanises the picture, showing planks being softened by steaming and bent into shape. texta completes a set of references here to Minerva as patron of shipbuilding, chariot-making, and weaving (9 currum; cf. e.g. Il. 15.410–12, Hymn. Hom. 5.12–15). The Indo-European root of texere meant ‘build, construct (of wood)’ as well as ‘weave’ (see Skutsch on Enn. Ann. 504): Ennius has textrinum there of a dockyard, and texere continued to be used of shipbuilding, especially in poetry (see Fordyce and cf. esp. Enn. trag. 43–4 Jocelyn = fr. 151a.12–13 FRL classis cita | texitur with 6n.). However, pinea … texta still sounds oxymoronic here, implying that the pines are for Minerva as flexible as woven threads, and foreshadowing the woven uestis, which is an object of wonder (267–8) just as the Argo is a monstrum (15); with 15 admirantes cf. 51 mira … arte n. Both artefacts have metapoetic significance as analogues for the poem itself, a resonance cued here by texta: see further 12–18n. coniungens: The verb recurs in the poem only with reference to marriage (331, 335, 372; coniunx, 123, 182, 298, 329, 373): perhaps another programmatic hint (cf. 8n.). inflexae … carinae: Here ‘curved hull’ rather than ‘keel’, as shown by the details of its construction. See also 249n. 11  A statement that the Argo was the first ever ship (cf. 1–11n.). This is in any case implied by the wonder of the Nereids in 12–18, and so alternative readings that make the line refer to this particular ship’s having never sailed before do not remove the temporal paradox caused by the appearance of Theseus’ ship on the uestis (53 cum classe n.); they are also anticlimactic given the line’s position as a hinge between the

151

COM M ENTAR Y: 11

initial description of the Argo, which it summarises, and the nymphs’ reaction, which it motivates. I would therefore rule out illa rudem cursu proram imbuit Amphitrite (a text based on the marginal variant proram and correction aphitrite both offered by O1), whether understood with Amphitrite ablative and illa = Minerva (Baehrens), or with Amphitrite nominative with illa, ‘the Amphitrite of that time’ (Ellis); for discussion of other still less plausible possibilities see Zicàri (1961), Nuzzo, and Pinotti (2019) 923–8. Many modern editors print illa rudem cursu prima imbuit Amphitriten, in which illa refers to 10 carinae, now understood synecdochically as ‘ship’, and the primacy of the Argo is emphasised with the manuscript conjecture prima. The addition of primus to imbuo is paralleled at e.g. Prop. 4.10.5, Sen. HF 455, Stat. Theb. 3.219, Sil. 3.64–5. Specifically, prima may be supported by Prop. 3.15.5–6 illa rudes animos per noctes conscia primas | imbuit (so Zicàri (1961), although the syntactic role played by primas is ­different), or by the appearance of prima in subsequent Argonautic passages (Ov. Am. 2.11.1, Luc. 3.195, 6.400, V.Fl. 1.1; cf. 1–11, 1nn. and Degl’Innocenti Pierini (1999) 193 n. 39). It is also plausible that primam, found in GR, was the reading of A, since it could easily lie behind O post eam via a misreading of the abbreviation for pri- (see Thomson, and cf. 145n. on the confusion of post- and prae‑). However, this text creates one serious stylistic problem, the combination with 12 quae, since, as Heyworth (in Harrison–Heyworth (1998) 105) points out, ‘it is not normal Latin style to pick up the subject of the preceding sentence with a nominative relative’. (This objection would also apply to prora (mentioned as long ago as Voss), if understood as nominative with illa, a text promoted by Postgate (1886) 436, (1888) 244–6 and recently printed by Lee and Nuzzo; moreover, carinae picked up by prora would be an odd identification of one part of a ship with another.) Heyworth therefore proposes reintroducing prora, but as an instrumental ablative (a possibility discussed by Voss), and understanding Minerva as the subject, referred to in the first word of the line not with the transmitted illa, inept after 9 ipsa, but with a second ipsa in anaphora (tentatively suggested by Baehrens; see also Kroll). Heyworth notes that this strengthens the allusion to Apollonius, since Arg. 1.109 and 111 (cf. 9n.) both begin with αὐτή of Athena; however, with this text Catullus has varied his Argonautic predecessor again by reprising the second of Apollonius’ lines in 9 before adding the anti-Apollonian detail in 11 (since in Apollonius the Argo is not the first ship). Meanwhile prora is not only a clear antecedent for 12 quae but appropriately anticipates the rest of that line, since the beak is attached to the prow, and

152

COM M ENTAR Y: 11

proscidit might suggest an etymology for prora (though see also 12n.). imbuit can make its point without prima, as Heyworth notes, comparing the allusion to this line (and to the metaphor of 12) at V.Fl. 1.69–70 ignaras Cereris qui uomere terras | imbuit. Finally, against the passages of Propertius and Ovid cited in support of prima (above) may be set further allusions to Catullus’ famous proem at Prop. 3.22.13–14 qua rudis immissa natat inter saxa columba | in faciem prorae pinus adacta nouae, Stat. Theb. 5.335–7 aerata dispellens aequora prora | Pelias intacti late subit hospita ponti | pinus. If the correct text is prora, then perhaps either A read proram and X and O made separate corruptions but only O1 added A’s reading as a variant, or A did indeed read primam (perhaps as a result of misreading another p- abbreviation, pro‑) and O1’s proram was either a marginal variant in A (subsequently omitted either by X copying A, or by R or R2 copying X) or a fortunate conjecture. For prorae see below. rudem cursus: ‘inexperienced in travel’; rudis has sexual overtones, as frequently in elegy (see Pichon s.v.). It also appears of the sea before the first voyage at Stat. Silu. 3.2.61–2 quis rude et abscisum miseris animantibus aequor | fecit iter? The combination with ipsa creates a strange effect as the female goddess seems to take the virginity (see below and 12n.) of the equally divine and female (Amphitriten) sea. Heyworth’s cursus for transmitted cursu is needed next to the ablative prora, and should also be considered by editors who print e.g. illa … prima, since genitive after rudis is common (also in + ablative, ad + accusative), whereas rudis + bare ablative seems to be attested only in the expression arte or artibus rudis (Ov. Tr. 2.424, Pont. 3.3.38, Stat. Theb. 6.437) and at Vell. 2.73.1 hic adulescens erat studiis rudis, Mart. 6.42.4 fontes Aponi rudes puellis. An alternative solution would be to take rudem as absolute and cursu with imbuit, as Mitscherlich suggests alongside genitive prorae, ‘initiated in the movement of the prow’. imbuit: The sense ‘initiate (to a new activity)’ is activated by the personal object Amphitriten, but imbuere reached this meaning through ‘dip or wet for the first time’ (cf. 4.17) from its basic sense ‘drench’ (397n.): paradoxically, the ship ‘wets’ the sea. Amphitriten: Wife of Poseidon/Neptune (e.g. Hes. Theog. 930–3); see further Roscher i.1.318–21 (Stoll). The appearance of both members of this divine couple, in usages that are just on the edge of metonymy (2 Neptuni n.), helps to create an appropriate background for the meeting of Peleus and Thetis, although Michalopoulos (2009) is not convincing in his argument that Poseidon and Amphitrite, and a dithyramb of Bacchylides in which they appear (Bacchyl. 17 Snell–Maehler), are significant presences throughout the poem. The use of Amphitrite

153

COM M ENTA R Y: 12–18

for the sea is first extant here in Latin, possibly taken directly from a Hellenistic source, or created on the model of similar metonymies in (certain kinds of) Hellenistic poetry: see Hunter (2006) 77–9, who notes that it appears in the same Isis hymn from Andros (Totti text 2, at line 145) as that cited at 15n. Virgil turns Thetis herself into a metonym when alluding to Catullus at Ecl. 4.32 temptare Thetin ratibus: see further Polt (2016). The name appears in the spondaic line-ending as in all its appearances in Homer (Od. 3.91, 5.422, 12.60, 97), with prosodic lengthening of an internal vowel before mute + liquid (cf. 291n.). The Greek ending is a safe restoration (see 3 Phasidos n. and Housman (1972) 817–39); cf. the transmission of Acmen in poem 45, where the fact that it looked less like a name and more like other Latin words perhaps preserved ‑n (45.1 ac men OGR, 45.21 agmen OGR).

12–18: THE NEREIDS The Nereids initially seem to be presented as a decorative accompaniment to the apparent protagonist, the Argo, rather than as a group that could include one of the poem’s central characters. Although a few have their own stories (Thetis, Amphitrite, Galatea), the daughters of Nereus usually appear in literature in large groups, and are not highly individualised, even when named, since this is usually in catalogues (e.g. Il. 18.37–49, Hes. Theog. 240–64). In art, and in some Hellenistic and later literary representations (e.g. Mosch. Eur. 118–19, Lucian, Dial. mar. 15.3), they are even more obviously pieces of decoration, appearing alongside other typical sea creatures such as dolphins and Tritons. See LIMC s.v. ‘Nereides’ (Icard-Gianolio–Szabados), Barringer (1995). As the first obvious visual tableau in the poem, yet occurring in the supposedly narrative outer story rather than the ecphrasis, this passage often plays a key role in critical approaches emphasising the poem’s ‘pictorialism’ and artistic imagery (e.g. Granarolo (1972), O’Connell (1977)). The gaze is an important unifying factor in the poem (see e.g. Fitzgerald (1995) 140–68): recurring aspects introduced here are the voyeuristic viewing of the female body (cf. the Fates (305–22), Polyxena (362–70), and especially Ariadne), love at first sight (cf. Ariadne at 86–93), and the gaze towards a ship (Ariadne at 52–70, 126–7, 249–50; Aegeus at 241–5). The mutual divine–human gaze leading to marriage contrasts with mortals no longer being able to see the gods in the epilogue (esp. 407–8). But the passage also indicates how the process of

154

COM M ENTAR Y: 12–18

reading might almost be able to bridge this gap: like the uestis (10n.), the Argo offers itself as a potential metaphor for the text, a work of art to be gazed and wondered at, while, as narrative time slows down to linger on one moment in the story, the Nereids are presented pictorially. Reminiscent of a common artistic subject (LIMC s.v. ‘Nereides’, Griffin (1985) 93 n. 24), and presumably seeking to appeal to the reader by its erotic visual details, the description programmatically invites such a reader to read this and the poem’s other visual scenes simply by gazing at them. The implication (for now) is that the poem can provide access to the heroic age via the pictures it creates, even though more narrative, less descriptive passages distance both narrator and reader from the past (1 quondam, 2 dicuntur nn.). See further esp. 50–75n. and the Epilogue to the commentary. However, the scene’s engagement with earlier epic treatments of nymphs rising from the water gives it a narrative aspect of its own, which suggests a less happy tone than the pictorial reading. The two main passages evoked are Ap. Arg. 1.540–58 (gods wonder at the newly launched Argo) and 4.930–63 (Thetis and the Nereids help it through the Planctae). On one level these allusions to basically joyous occasions of divine support for the Argonauts underline the optimistic mood; but they also alert the reader to Apollonius’ specific treatment of Peleus and Thetis (Hunter (1991)), especially since both emphasise the unhappy end of the marriage and the separation of Achilles from his parents (at Ap. Arg. 1.553–8, the baby Achilles appears with Chiron; in book 4 the Nereids’ appearance is part of an episode, 4.753–967, in which the whole story of the marriage and estrangement is told). Moreover, in Homer Thetis and the Nereids mainly rise from the sea in contexts involving Achilles’ grief or death (Il. 1.357–60, 18.65–70 (cf. 323–81n.), Od. 24.47–8); and Clare (1996) 63–4 plausibly identifies another Apollonian model for this scene, the rape of Hylas (Ap. Arg. 1.1228–39; cf. also Theocritus’ version, 13.39–54, closer to Catullus in having plural nymphs). Allusion to this story is not propitious either for a meeting of lovers, since love at first sight ends tragically for Hylas, or in an Argonautic context, since it causes the loss of Heracles. Catullus’ scene influenced many later renderings of the appearance of water-nymphs. At Virg. Aen. 9.117–22 Aeneas’ transformed ships ‘turn into a kind of composite of the nymphs and the ship at which they marvel in Catullus 64’ (Hardie (2012) 227; cf. 15 monstrum n.). At Manil. 5.563–4 extulit et liquido Nereis ab aequore uultus | et, casus miserata tuos, rorauit et undas the narrator addresses an Andromeda reminiscent of Catullus’ Ariadne, while rorauit et undas reprises the paradox of 11

155

COM M ENTAR Y: 12

imbuit ­Amphitriten. Statius’ Achilleid opens in Catullan style with a fateful voyage, that of Paris and Helen from Sparta to Troy, which causes Thetis, described a little later as pectore nudo (Ach. 1.77), to rise anxiously from the sea, accompanied by her sisters (Ach. 1.25–9); cf. 183n. for a Catullan phrase at Ach. 1.26 and see further esp. 1–11n. and the Introduction, 3a(ii). Finally, an exceptionally close and sustained adaptation of the imagery and diction of this passage occurs at Sil. 7.411–14 sulcabat rostris portusque intrarat apertos | ac totus multo spumabat remige pontus, | cum trepidae fremitu uitreis e sedibus antri | aequoreae pelago simul emersere sorores. The evocation of the Argo suggests the audacia of the Carthaginians as they sail into the harbour of Caieta, a place that had not previously known enemy ships, while Catullus’ Nereids become very Italian nymphs concerned for the fate of Rome: on seeing the ships they flee to ask Proteus what is happening, motivating a prophecy about Rome’s future (Sil. 7.435–93). See further Littlewood on Sil. 7.409–93 and 414. 12  quae simul ac: Monosyllable + simul ac at line-beginning also at 86, 147, 233, 366; once in Cic. Arat. (350 Soubiran), eight times in Lucretius, only twice in Virgil (Aen. 4.90, 12.222). Cf. 31 quae simul (where see n. on the relative), again a line-­beginning opening a new section of narrative, where simul alone is used in the same way as a conjunction, ‘as soon as’, as at 22.15, 51.6, 99.7. rostro: The technical term for the beak on a ship’s prow; it also means ‘ploughshare’ (Plin. NH 18.171, 34.144), contributing to the metaphor in proscidit aequor (Konstan (1977) 19 n. 25). Cf. Cic. Arat. 129 Soubiran rostris Neptunia prata secantes; as this describes how other ships, unlike the celestial Argo, travel forwards, not stern-foremost (6 cita … puppi n.), Catullus may be alluding to Cicero’s word and imagery (the phrase has no original in Aratus) to make the point that while still an earthly ship the Argo did not travel backwards. uentosum … aequor loosely translates Ap. Arg. 1.541 λάβρον ὕδωρ, the ‘boisterous’ water hit by the Argonauts’ oars (the beak is not mentioned). aequor in either singular or plural is used by Catullus with more (7, 206, 4.17, 11.11, 68.3) or less (179, 11.8, 101.1) sense of the sea’s ‘flat surface’ (see Ernout–Meillet, ­Walde–Hofmann, de Vaan s.v. for the derivation from aequus). For the plural cf. Enn. praetextae fr.  2 FRL = scen. 367 Vahlen2, Ann. 505 Skutsch/FRL; singular, Cic. Arat. 67 Soubiran, frequently in Lucretius. As aequor can also mean the level surface of a field, it suggests the new fluidity of the boundary between land and sea; cf. Virg. Georg. 1.50 ac prius ignotum ferro quam scindimus aequor, where ploughing untilled soil seems comparable to the first ploughing

156

COM M ENTAR Y: 13

or seafaring in history, and thus evokes the idea of the end of the golden age (Hardie (2012) 219; cf. 2, 6nn.). On poetic ‑osus adjectives see Ross (1969) 53–60, Knox (1986b). Virg. Georg. 1.206 conflates uentosum … aequor with Cat. 101.1 to produce uentosa per aequora uectis, echoed at Aen. 6.335, Ov. Her. 17.5. proscidit: ‘Ploughing’ or ‘cutting through’ the sea is a regular poetic image (e.g. Ap. Arg. 4.225, Ov. Met. 4.706–7, Homeric τέμνειν, e.g. Od. 3.174–5); proscindere is especially appropriate for the first seafaring as it strictly means the first ploughing of the year (e.g. Lucr. 5.209, Virg. Georg. 1.97 and DServ. ad loc., Var. RR 1.29.2), while rostro … proscidit might make an etymological point (but see also 11n.). The image also develops the sexual metaphor of 11, here underlined by the phallic shape of the rostrum. 13  Whitening of seawater by a ship is described at e.g. Od. 12.171–2, Eur. Cycl. 16–17 (caused by oars); Eur. fr. 757.844–5 TrGF (caused by the Argo’s prow). The implied colour contrast with the blue sea (7) is more explicit at Enn. Ann. 377–8 Skutsch/FRL and at Ap. Arg. 1.542, mentioning both foam and dark sea (ἀφρῷ … κελαινὴ … ἅλμη) and 545–6, comparing the white wake to a path on a green plain (χλοεροῖο … πεδίοιο; cf. 12 aequor). Catullus emphasises this visual aspect of Apollonius’ launching scene (Ap. Arg. 1.519–58), excluding references to the sound of the roaring sea and Orpheus’ lyre-playing (Syndikus (1990) 119–21). tortaque: OGR totaque would leave remigio unsupported. For tortaque (a conjecture made in manuscripts and by Avancius) against Baehrens’ motaque cf. Virg. Aen. 3.207–8 nautae | adnixi torquent spumas et caerula uerrunt, V.Fl. 1.362 tortas … undas: it demonstrates acute ‘visual precision’ (  Jenkyns (1982) 104), as each dipped oar creates a small vortex. The shape of this line, with ‑que introducing an end-stopped, tacked-on detail after another end-stopped line that could have completed a sentence or clause, recurs frequently in the poem (30, 36, 96, 191, 235, 264, 300, 310, 316, 331, 345, 351, 398); cf. 5 optantes n. and see the Introduction, 2d(ii). incanuit: The conjecture is owed to the first Aldine edition. OGR incanduit is retained by e.g. Baehrens, Friedrich, Kroll, and Eisenhut, but is unlikely so close to 14 candenti. This could be explained as ‘participial resumption’ (Wills (1996) 311–25), but this figure (otherwise absent from Catullus, except possibly for 67.12–13; cf. Wills (1996) 315) would not be particularly pointed here; more importantly, incandescere meaning ‘shine white’ (clearly the sense of 14 candenti; cf. 45 candet ebur n.) would be unparalleled, although candescere = ‘shine white’ is ­attested

157

COM M ENTAR Y: 14

(TLL i i i .236.20–6). incanduit could mean ‘became shining’ only in the sense ‘became glowing red-hot’, which would be nonsensical (the conceit of the river growing hot with sexual excitement at Ov. Am. 3.6.25–6, cited by Konstan (1977) 13, is much more explicit). canus, canere, etc., referring to grey or matt white, are common of sea-foam (e.g. Virg. Aen. 8.672, Ov. Her. 3.65, 5.54, Sen. Ag. 441, V.Fl. 3.32), following Homeric ἁλὸς πολιῆς (e.g. Il. 1.350). While canus and other words for ‘white’ can be used together of the same object with little distinction of sense (André (1949) 67–8), here a contrast of textures is implied between incanuit and candenti (  Jenkyns (1982) 104): as at Lucr. 2.766–7 mare … | uertitur in canos candenti marmore fluctus, the image is probably matt foam with a surface gleam from the sunlight. This also explains why 14 candenti e gurgite can be followed by 18 e gurgite cano. 14  OGR’s text is printed by e.g. Ellis, Lenchantin, Quinn, and Thomson and defended by Puelma (1977) 156–72. Despite the lack of close parallels feri … uultus might just be acceptable as nominative plural in apposition with 15 aequoreae … Nereides (more likely than genitive singular qualifying it), but the sense of ferus creates a more serious problem. Its primary meaning is ‘undomesticated’; the only possible parallel for an extension of this to ‘timid’ (Quinn) is Virg. Aen. 4.152 ferae … caprae, which is more likely to mean simply ‘wild goats’ rather than domesticated ones. Of the face, where emotions are shown, ferus would much more naturally have its second main meaning, ‘savage’. But the Nereids are friendly, reacting with wonder (15), not anger or fear; the sea they personify is stirred up by oars (13) but otherwise calm, whereas feri understood as transferred to the sea would mean ‘stormy’, ‘hostile’ (63.40 mare ferum). This also counts against other conjectures which attribute ferus to a word for ‘sea’ in this line: either Schrader’s fero … gurgite (adopted by Baehrens alongside Eldick’s candentes, which he spells candentis and understands as accusative plural) or Koch’s feri … ponti (see below). Schrader’s freti on the other hand gives an example of the generalised usage (since Accius’ Medea, trag. 477 Dangel; cf. 4.18) of a word literally meaning a certain part of the sea, a ‘strait’ (cf. the treatment of e.g. unda, uada, aequor, gurges). The corruption of freti to feri would be relatively easy: see Friedrich for the palaeography, and note the same corruption in one branch of the manuscript tradition at Sen. Epist. 79.1 (see ­Reynolds’ apparatus on line 20). There is some evidence that later authors read freti here: Manil. 1.165 emersere fretis montes ­(Biondi (1981); cf. 3, 9, 251nn. for other echoes in Manilius of the openings of lines from this poem), [Sen.] Oct. 706–7 talis emersam freto | spumante Peleus

158

COM M ENTAR Y: 15

coniugem accepit Thetin (Ferri ad loc.). freti is particularly appropriate for this description of foaming water because of the connection with ferueo (Var. LL 7.22; see further Maltby s.v. fretum and cf. puns at e.g. Lucr. 6.427–8, Virg. Georg. 1.327). Cf. similar periphrases for ‘sea’ at 127, 178–9; although freti … gurgite is unparalleled, comparable genitives with gurgite include Virg. Aen. 3.421 imo barathri … gurgite, Ov. Fast. 4.48 Tuscae gurgite … aquae as well as the convenient line-ending gurgite ponti (Cic. Progn. i i i .7 Soubiran, Lucr. 5.387, Luc. 7.813, Sil. 12.117, Stat. Theb. 7.143). Trappes-Lomax, repeating a conjecture originally made by Koch, would print feri … gurgite ponti here. But uultus is an appropriate detail: it goes well with 15 admirantes (cf. 34 declarant gaudia uultu) and also creates the sense that the nymphs emerge gradually, heads first and then upper bodies (17–18). If freti is printed, however, then unless the apposition of uultus and Nereides is to be accepted after all (as by Eldick, with candentes nominative), emersere must be taken as transitive, subject Nereides, object uultus. This is difficult, but less problematic than keeping feri. For full discussions of possible parallels see Puelma (1977) 158–60 and Thomson: in all except Dirae 56–7 (below) the text has been challenged. However, reflexive se emergere is established by e.g. Nep. Att. 11.1 quibus ex malis ut se emersit, Sall. Hist. 4.28 ubi se laniata naufragia fundo emergunt, Manil. 5.198 ex undis … sese emergit in astra, and the past participle usually occurs in a middle sense, ‘having raised oneself out’ (cf. OLD s.v. 1b), e.g. Cic. Diu. 2.140 subito sum uisus emersus e flumine, [Ov.] Hal. 36 demum emersus in auras. Although this does not prove that emergere could be used fully transitively, the use here with a body part as object feels similar (quasi-reflexive). This would also c­ over Dirae 56–7 monstra repentinis terrentia saepe figuris, | cum subito emersere furenti corpora ponto (where, however, it is ‘just possible’ that emersere might be intransitive, corpora a rather flat nominative: Richmond (1962) 44). I would also understand Sil. 7.414 emersere sorores (12–18n.) as a regularisation of Catullus’ verb to its usual intransitive sense. For gurgite cf. 18, 178, 183. gurgite always occurs in this place in the line in Catullus (also 65.5) and Lucretius (three times), and Luck (1976) 233 suggests that this results from the influence of Cicero’s poetry (Arat. 422, Progn. i i i . 7 Soubiran); but there are limited places for dactylic words in the hexameter, and perhaps this is only a slight stereotyping within the tradition. 15  The first of four four-word hexameters in the poem (also 77, 115, 319, not elsewhere in Catullus): see Bassett (1919) on the history and

159

COM M ENTAR Y: 15

distribution of such lines in Greek, Kidd (1997) 35–6 on Aratus’ fondness for the version that is also a spondeiazon, as here, and Dainotti (2015) 79–80 n. 261 for Virgil’s four-word lines (often with names: cf. in Catullus the three-word pentameter at 68.112 audit falsiparens Amphitryoniades). The rarity of four-word lines encourages the inference that this one translates Ap. Arg. 1.550 Πηλιάδες κορυφῇσιν ἐθάμβεον εἰσορόωσαι (Thomas (1982) 158 n. 54). aequoreae: The adjective appears first here, possibly coined (TLL s.v. ‘adi. a neotericis formatum’); see further Ross (1969) 60–3 on Catullus’ use of poetic adjectives in ‑eus. aequoreus became fairly common in poetry, describing sea-gods at e.g. Prop. 3.7.67, Ov. Met. 12.197, Fast. 1.372, Manil. 5.433, Sil. 7.414. monstrum: Originally a religious ‘portent’ (cf. moneo), but often more widely ‘strange phenomenon’, ‘wonder’, though here as for the Trojan Horse (Virg. Aen. 2.245) the sense ‘object presaging things to come’ is still appropriate – the apparition of the Argo marks the beginning of seafaring, Peleus’ and Thetis’ relationship, and the poem. A monstrum would normally be a sign from the divine world seen by humans: here this is reversed, since, although the Argo was divinely made, it is now in human hands, appearing to the hitherto solely divine society of the sea. Quite possibly the sense of ‘monster’ or strange animal (cf. 101) is also active (cf. 2 nasse, 7 palmis, 9 uolitantem n., 12 rostro). According to Cic. ND 2.89 the shepherd in Accius’ Medea thought the Argo was a diuinum et nouum uehiculum (see below), while his predecessors at Ap. Arg. 4.315–22 see the Argonauts’ ships as θῆρας. At Virg. Aen. 9.120 the ship-nymphs are mirabile monstrum, a phrase common in Virgil but also echoing this line (Hardie (1987) 163 n. 4); cf. also Drac. Med. 38 nam monstra putat. Nereides: Late in the sentence as the Argonauts and the reader finally discover the creatures’ identity. The Greek form of the Greek name (cf. 3 Phasidos n.) is usual in Latin for third-­declension nominative plural (Housman (1972) 836–9); cf. 23 heroes, 193 Eumenides. Nereides appears first in Latin at Cic. Arat. 446 Soubiran (also with Greek short ‑es). Catullus, like Cicero, transliterates Νηρηΐδες rather than the alternative Νηρεΐδες: the former occurs at Ap. Arg. 4.930 but is also the consistent Homeric form (Il. 18.38, 49, 52) whereas Νηρεΐδες does not appear in epic until Mosch. Eur. 118 (Campbell ad loc.). admirantes translates ἐθάμβεον (Ap. Arg. 1.550), perhaps also drawing on the description in Accius’ Medea, trag. 467–78 Dangel, of amazement at the Argo felt by a shepherd, admirans et perterritus in Cicero’s summary (ND 2.89). Further examples of wonder at the Argo include

160

COM M ENTAR Y: 16–18

V.Fl. 1.149, 4.711–12, Claud. Get. 8–9, Orph. Arg. 237. Cf. also the Nereids’ reaction to the first ship in the Isis hymn from Andros cited at 11 Amphitrite n. (Totti text 2, at lines 152–7 = 1028.62–7 Kaibel): the description ends with παπταινοίσαις, similarly a participle describing the gaze and creating an evocative spondeiazon (for which cf. e.g. Virg. Aen. 2.68 circumspexit). The similarity may be owed to an earlier Hellenistic model (Kroll, Kaibel ad loc., Wilamowitz (1924) ii.300), but Hunter (2006) 57 argues that first-century Roman poets may have known Greek texts like this one (itself probably from the first century). This line is evoked at Ciris 391 complures illam nymphae mirantur in undis (see also 19–21n.). At Virg. Aen. 8.91–2 mirantur et undae, | miratur nemus and Ov. Am. 2.11.1 mirantibus aequoris undis, the pathetic fallacy ascribes the emotion to inanimate nature. 16–18  The narrative shifts from reactions to the Argo to the Argonauts’ own point of view; the importance of this turn away from the model of Ap. Arg. 1.540–58, where the divine gaze simply emphasises the miraculous nature of the Argo, becomes clear from line 19 onwards. The auspicious vision of the nymphs in these lines accords with references in Pindar to ­Peleus as ­particularly blessed for having seen the gods (Pyth. 3.94, Nem. 4.66–8), but contrasts with the tradition that seeing naked nymphs is a danger to be avoided, causing blindness or madness (Balland (1976) 3–6); Tiresias (e.g. Call. Hymn. 5) and Actaeon (e.g. Ov. Met. 3.173–255) are both punished for seeing naked goddesses with their attendant nymphs. This tradition may undermine, or simply be a foil for, Catullus’ description of a supposedly happy outcome from mutual gaze between heroic man and naked goddess. 16  illa, atque alia: The paradosis is incomplete. Supplements giving the meaning ‘on that day (luce), and on no other’ began to be proposed in the Renaissance: illa haudque Sabellicus (but the hiatus is implausible and the form haudque unparalleled), illaque Puccius. illa atque (Bergk, if not a­ nticipated by recentiores: see app. crit.) is the clearest of such supplements in expression and the closest to the transmitted text, since the readings in GR and in O would simply show different extents of omission. On haud see 339n. Other conjectures avoid saying that the event was unique, since there are numerous instances of Nereids accompanying ships on ‘later’ occasions (e.g. Soph. OC 716–9, Eur. El. 432–51, and indeed Ap. Arg. 4.930–63). But e.g. ‘on that day, if on any’ (illa si qua alia Lachmann), ‘on that day and on other succeeding days’ (illa, atque alia Vahlen; cf. 68.152 haec atque

161

COM M ENTAR Y: 16

illa dies atque alia atque alia) seem too vague for a context in which the importance of a single moment is emphasised (12 simul ac, 19–21 tum), while ‘on that day and never before’ (illa, ante alia Friedrich, illa, ante alia Goold) might, as Godwin points out, suggest the odd inference that while this was the first time that sailors saw nymphs, such an occurrence is now commonplace. Some conjectures introduce a spatial element: Schulze, for instance, argued for reading the same words as Vahlen but understanding them as ‘on that day (illa … luce) on one side and another (alia atque alia)’, comparing Ap. Arg. 4.930–1 ἔνθα σφιν κοῦραι Νηρηίδες ἄλλοθεν ἄλλαι | ἤντεον, but as Vahlen (1911) i i .721–5 and Friedrich note, it would be very difficult to read alia as agreeing with an understood feminine noun (uia?) when it comes between illa and luce. Alternatively, if we suppose that some of the words transmitted by OGR may be intrusive glosses or corrections rather than remnants of the original text (for instance, Baehrens argues that O illa alia and GR illa atque alia could represent different results of copying an archetype which had atque alia in the text and illa written above atque as an attempted correction of alia), it is possible to assume wider corruption and perhaps to introduce an adjective for luce (e.g. Baehrens atque illa uidere luce marinas). For further suggestions see Catullus Online and the textual discussions of Thomson, Nuzzo, Balland (1976) 7 n. 25, Biondi (1980) 143 n. 51. A statement of complete uniqueness, however, still seems the best fit for this emphatic position. The specificity of the rest of the sentence is also important: we are told that this was the only day on which mortals saw naked nymphs, in the water, exposed as far as the breasts (perhaps this was the only time when the innocent Nereids did not know that a ship would contain mortals who might see them). This softens the line’s inconsistency with the references to Nereid sightings in other texts (above) and, more importantly, does not contradict the epilogue’s description of gods and men intermingling in the heroic age in various (other) contexts (384–96, 407–8). luce ‘day’ (illa … uiderunt luce translates Ap. Arg. 1.547 λεῦσσον … ἤματι κείνῳ), but with connotations of sunlight (cf. 14 candenti) and clear vision (cf. 17 oculi and 408 nec se contingi patiuntur lumine claro). Cf. 31, 237, 325 and see further Lyne on Ciris 417. It is perhaps a piece of quasi-Hellenistic philological cleverness to put mortales and oculi so close to luce when the use of lux = ‘day’ is, at least in poetry, inspired by that of Greek φῶς, also primarily ‘light’: φῶς also = ‘eye’, and φώς (from a different root) = ‘mortal man’. marinas: First here of sea deities; frequently so in subsequent Latin poetry (TLL viii.397.82–398.10), sometimes, as here, with specifying force (e.g. Prop. 4.6.61–2; Ov. Met. 11.228, 13.964, 14.566). R  E xvii.1

162

COM M ENTAR Y: 17–18

s.v. ‘Nereiden’ 3–4 (­Herzog-­Hauser) collects Greek and Latin examples of similar phrases, e.g. Soph. Phil. 1470 νύμφαις ἁλίαισιν. marinas varies 15 aequoreae, the repetition of sense helping to convey the reciprocity of the gaze. This is the first strongly enjambed line in the poem, but the wide separation of marinas from nymphas is mitigated by the position of both at line-end; cf. qualibus … | fluctibus opening lines 97–8. 17 mortales: Usually taken as substantive, subject of uiderunt, but adjectival elsewhere in Catullus (385, perhaps 168 (see n.), no other appearances): see below. oculi: A manuscript conjecture for OGR oculis; the editor ­faces the same textual choice at Virg. Aen. 8.222–3 nostri … | … o­ culis (oculi ‘alii’ according to Servius). An ablative strengthening the verb of seeing would follow Homeric ὀφθαλμοῖσιν ἰδεῖν (e.g. Il. 1.587, Od. 4.269); in Latin this usage is much more common in comedy and prose than in epic (OLD s.v. 2b; in Virg. Aen. only certain at 2.68, 589, 12.151). Here it sounds awkward next to the unrelated ablative nudato corpore, while nominative mortales oculi forms a balanced contrast with aequoreae … Nereides and marinas | … nymphas. For mortalis with a part of the body, and a similar contrast with the divine, cf. e.g. Virg. Aen. 1.327–8 haud tibi uultus | mortalis, Prop. 2.32.50 altaque mortali deligere astra manu; with oculi at Albinovanus Pedo 1.20–1 Courtney = 228.20–1 Hollis di reuocant rerumque uetant cognoscere finem | mortales oculos, Lucr. 1.66–7 (of Epicurus, another hero viewing, near the start of the poem, what was not previously open to mortal sight), Apul. De mund. 31.6, and cf. Nonn. Dion. 5.402 ὀφθαλμοὺς βροτέους (also at the start of the line, as Nuzzo notes; for the possible relationship between Catullus and Nonnus see the Introduction, 3a(ii)). Cf. also 385–6 sese mortali ostendere coetu | caelicolae. nudato corpore: A detail from artistic rather than literary representations of Nereids: in art after the classical period they are usually naked or semi-naked (LIMC s.v. ‘Nereides’ (Icard-­Gianolio–Szabados); cf. Cairns (1984) 96 n.  3), while no particular state of dress is mentioned in Ap. Arg. 1, Il. 18, Od. 24. See further 18 tenus n. nudato for nudo suggests they have been actively ‘stripped’, perhaps by the poet. nymphas: Greek, but already adopted into Latin at Liv. Andron. Od. 14 FRL = 13 Blänsdorf. The generalising word allows a more direct allusion to the νύμφαι | Πηλιάδες of Ap. Arg. 1.549–50 (Syndikus (1990) 120 n. 80). 18  Although 17 nudato corpore is literally true, the Nereids turn out to have only their upper bodies above the water: the construction in which a participle (exstantes) adds a further detail a­ fter the apparent end of

163

COM M ENTAR Y: 18

a sentence (5 optantes n.) is teasingly effective. Cairns (1984), working from Artemidorus’ Oneirokritikon 2.37–8, where dreaming of Aphrodite emerging from the sea as far as the ζώνη is a good omen for seafarers, ingeniously suggests that both Catullus and Lucian, Dial. mar. 15.3 refer to a tradition of marriage-omens involving dream-apparitions of Aphrodite, the Nereids, and Nereus: however, the connection between Aphrodite and the Nereids together and marriage is unconvincing, resting only on Aphrodite’s associations with marriage and with the Nereids and on the appearance of Nereids at the weddings of two of their own number (Amphitrite and Thetis). Moreover, ἡμίγυμνοι of Lucian’s Nereids could mean ‘negligently dressed’ rather than specifically ‘naked to the waist’. nutricum: Literally ‘nurses’, here ‘breasts’. Apparently a display of Greek etymological learning, since, according to Hesych. s.v., τιτθή, feminine of the more common τιτθός, ‘breast’, could also mean ‘nurse’; nutricum, moreover, would translate τιτθῶν, genitive plural of either noun. No literary examples of τιτθή = ‘nurse’ survive: whether or not Catullus imitates a lost Hellenistic use, on this explanation the point is only comprehensible to a reader aware of a rare meaning of a Greek word. This could be part of a strategy to construct, or attract, a learned reader at the beginning of the poem. Alternatively, Adkin (2002) explains nutricum = ‘breasts’ as an example of transumptio (Quint. Inst. 8.6.37–9; cf. Lausberg §571), whereby since mamma can mean ‘nurse’ (e.g. Pers. 3.18) as well as ‘breast’, the meaning of nutrix, synonymous with mamma in the sense of ‘nurse’, can be extended to ‘breast’. This keeps within Latin the reader’s process of understanding nutricum, but still gives a clear impression of the author’s shifting of ordinary meanings. Some sense connected with ‘nursing’ would also still be felt, strengthening the links with the other appearances of breasts in the poem (65 lactentes … papillas, 351 pectora of mothers: see nn.). In other ways too the detail allusively draws attention to the unhappiness of Thetis’ future motherhood: when the Homeric Thetis mourns with Achilles the Nereids beat their breasts (Il. 18.50–1), and one etymology for Achilles’ name was ἀ-χεῖλος, he who did not put his lips to the breast (Apollod. 3.13.6; cf. Ap. Arg. 4.813 and see Hunter (1991) 255, Heslin (2005) 175–81). tenus: Almost always placed after its noun (N–W ii.948–9); the genitive with tenus is a rare, usually poetic alternative to the ablative (see Harrison on Virg. Aen. 10.209–11). The apparent p ­ recision may titillate the reader, since it is not clear whether tenus = ‘as far as and including’ or ‘as far as and not including’. The detail also neatly reverses Ap. Arg. 4.940 and (in a simile) 949, where the Nereids have rolled up

164

COM M ENTAR Y: 19–30

their clothing to expose their legs (Cairns (1984) 100; cf. 129n.); it may even offer a ‘humorous misinterpretation’ of the κόλπος that has been rolled away in 949, since κόλπος would normally refer to the top half of a garment (Hunter (1991) 254). e gurgite cano closes a ring from 13–14, uniting incanuit with candenti e gurgite. This phrase is the closest to H ­ omeric ἁλὸς πολιῆς (13 incanuit n.); it is reprised at Ciris 514, Stat. Theb. 11.43, Avien. Or. mar. 67. Its appearance next to nutricum (cf. 14 candenti and uultus) suggests both a similar whiteness to the nymphs’ bodies and a textural contrast, this time between smooth skin and ‘hoary’ foam. The reader is reminded that the only thing covering the Nereids’ lower halves is water; cano may suggest at least semi-opaque foam, but Friedrich thinks it would be transparent enough to allow the Argonauts ‘to guess’ (‘erraten’).

19–30: BETROTHAL OF PELEUS AND THETIS This introduction to the wedding scene centres on the hymnic invocation of the heroes at 22–4, and hymnic markers continue into the address to Peleus at 25–30. Such elements are appropriate for the opening of an otherwise non-hymnic text: cf. hymnic proems in didactic poetry (e.g. Hes. Theog. 1–115, Arat. Phaen. 1–18, Lucr. 1.1–49, Virg. Georg. 1.1–42) and the proem to Achilles in Simonides’ Plataea elegy (fr. 11 West; see Magnelli (2018)). The apparent rhapsodic farewell echoing the Homeric Hymns (24n.) might appropriately mark the transition from ­introductory material to the main subject, since the Homeric Hymns were apparently proemia to epic recitations (see e.g. Richardson (1974) 3–4, Aloni (2001) 92–3), and the αὐτὰρ ἐγώ of the farewell can also be used in the middle of a poem to indicate a similar transition to narrative (Obbink (2001) 67–73). However, such formulae close the Homeric Hymns as transmitted, and invocations frequently characterise the endings of other hymns (a tradition represented by e.g. 63.91–3 (on 63 and hymn see Harrison (2005) 18–23), Call. Hymn. 1.93–6, 5.140–2, 6.134–8), while address is a closural marker in other Catullan poems (Peden (1987) 97–8). Lines 22–4 allude to the end of the Argonautica, Ap. Arg. 4.1773–5 (see further Zetzel (1983) 261); yet that is an end whose own use of hymnic formulae suggests that it might be a beginning, ‘as if the complete Argonautica has been a (hymnic) prelude’ (Goldhill (1991) 287). These conflicting signals about location in the text mirror the disorientation over location in mythical chronology caused by the innovation of Peleus’ and Thetis’ wedding coming after the Argo’s v­ oyage

165

COM M ENTA R Y: 19–21

(19–21n.): both draw attention to how much the reader is in the power of the narrator and are part of one of the poem’s many explorations of the workings of narrative time. It is initially unclear whether the repeated tum in 19–21 is freezing the moment described in 12–18 or moving on from it (see n.), while the change at 22 to second-person address means that by 28–30 the reader is unsure whether the text connects with any identifiable moment of story time at all: cf. Culler (1981) 148–54, (2015) 225–9 on the ways in which address or apostrophe can create a lyric ‘now’, and see Trimble (2020) 126–7 for a suggestion of how this might also work for apostrophe in a narrative text. It does not seem reasonable for Friedrich to claim that 28 tenuit cannot mean ‘embraced’ (see n.) on the grounds that Thetis is still in the water, Peleus on board the Argo; yet the text has not told us where else they might be. There is also generic ambiguity, comparable to that between tragedy and epic in the opening (1–11n.), over whether the addresses in 22–30 are really hymnic after all. Though praise of ancestry is typical of hymns (23 deum genus n.), the heroes are otherwise praised only for their enviable blessedness in being born when they were: this arguably makes 22–4 as well as the address to Peleus in 25–8 simply a macarismos, an exclamation at others’ good fortune which occurs in various genres, not always with hymnic force, and which lacks the element of prayer. At the point where in a hymn a prayer would be expected (Furley–Bremer (2001) i.60–1), or where in a hymnic proem the deity addressed would be asked to inspire the following song (e.g. Hes. Theog. 104–15, Arat. Phaen. 16–18), the heroes are not asked to do anything: the only imperative is saluete (contrast even ἵλατε at Ap. Arg. 4.1773, which suggests that the Argonauts are being asked to fulfil the immediately following wish that ‘these songs’ may be ‘sweeter’ in the future; 24n.), and the questions to Peleus do not request any new information (28–30n.; cf. also 94–8n.). The narrator emerges in the first person for the first time at 22, apparently full of enthusiastic emotion in response to the connection between the human and the divine that has been established in 19–21 (on our first impressions of his subjectivity here see also Fernandelli (2012) 418–38); but despite his apparent piety, he does not relinquish his own authority over his poem. This complicates even further than is generally recognised the question of how seriously we should take the enthusiasm for the heroic age that he expresses (see e.g. Kinsey (1965), Curran (1969), Harmon (1973), Courtney (1990), and cf. esp. 382–408n.). 19–21 The acoustically and visually striking tum … tum … tum and Thetidis … Thetis … Thetidi emphasise the long-delayed announcement

166

COM M ENTAR Y: 19

of the real subject of the outer story and ­Catullus’ innovative treatment of it. Though other examples of triple anaphora occur at 117–18, 280–1 (cf. Évrard-Gillis (1976) 197–205), this form, with the repeated word opening three consecutive lines, is particularly distinctive: see Wills (1996) 400–5. It occurs in early Greek hexameter poetry (e.g. Il. 2.382–4, Od. 3.109–11; Hes. Theog. 833–5, Op. 5–7, 317–19) but proportionally more often in Callimachus (e.g. Hymn. 6.4–6, 108–10; extended versions, Hymn. 3.262–6, 4.70–5, 6.18–22), and in 64 more often than in any other Latin poem (39–41, 63–5, 257–9; cf. 387–94). Triple line-initial anaphora of tum is paralleled only at Ciris 387–9 (Lyne ad loc.), near other echoes of this poem (Ciris 391–2; cf. 15n.). It conveys the speed with which events followed one another (tum = ‘next’), or might even suggest simultaneity (tum = ‘at that moment’, each referring to the same point in time); each tum also gains the force of ‘it was then that’ as Catullus’ unexpected treatment of the myth, making the wedding follow the launching of the Argo (contrast e.g. Ap. Arg. 1.533–8, 4.757–960, V.Fl. 1.130–3, Apollod. 3.13.7), is emphasised; cf. 334–6 on the similarly innovative mutual happiness of the marriage, linked to these lines by similar repetitions. Meanwhile, the polyptoton of a name is paralleled at Call. Hymn. 2.44–7, 55–8, Stat. Ach. 1.473–4, but more closely in Archil. fr.  115 West, Anacr. 359 PMG ; as Anacreon repeats the name of a beloved, the narrator mimics Peleus’ love for Thetis. Wills (1996) 401–2 demonstrates the frequency of other types of repetition, especially polyptoton, alongside triple line-initial anaphora in Homer and Hesiod (e.g. Il. 23.315–18, Hes. Op. 182–4), and identifies the closest Homeric model for this passage as Od. 16.118–20; but neither these nor the closest Latin parallels for such a combination (Ov. Met. 5.341–3, Manil. 4.813–16) combine paired polyptoton and triple anaphora with a comparable triple polyptoton. The effect here may in general be closer to lyric (Syndikus (1990) 124–5; Wills (1996) 402 on 51.13–16 otium … otio … otium) or to ‘incantatory’ repetition in Hellenistic hexameters influenced by lyric hymn or encomium (e.g. Call. Hymn. 1.93–6, 2.43–5, 4.83–5; Theoc. 16.3–4, 17.26–7, 22.23–6) or by magical and other sub-­literary song (especially Theocritus, e.g. 1.97–8, 100–1, 2.38–41, 5.68–75, 11.22–3; cf. Dover (1971) xlv–l); the lines therefore form a stylistic bridge to the hymnic passage at 22–30. 19  Peleus falls in love at first sight in the space of one line: cf. 86n. incensus … amore: Repeated of Bacchus at 253, inviting a parallel between the two divine–human couples, Peleus/Thetis and Bacchus/ Ariadne; cf. also 97 incensam n. Since incensus can describe the effect of

167

COM M ENTAR Y: 20

any strong emotion (e.g. Liv. 2.6.1 incensus Tarquinius non dolore solum … sed etiam odio iraque), only at amore do the surprising chronology and relative importance in the poem of the wedding and the voyage emerge: at Ap. Arg. 4.866 Thetis’ appearance causes Peleus regretful ἄχος αἰνόν, and here, even after the erotic presentation of the Nereids, tum Thetidis P ­ eleus incensus could begin a description of a bitter detail incidentally relevant to the Argo’s setting out (e.g. ‘Then Peleus, noticing Thetis’ face, was inflamed with grief, remembering …’). fertur: Probably incensus fertur = incensus esse fertur, ‘is said to have been inflamed’. The appeal to tradition (2 dicuntur n.) is here ironic, emphasising Catullus’ originality in stating that Peleus’ relationship with Thetis began with him seeing and falling in love with her: even in the happiest accounts of the wedding (25n.), his feelings towards Thetis are normally no more relevant than hers towards him (20n.; see also the Introduction, 2b(i)). Quinn thinks this construction ‘out of harmony with the positive statements of 20 and 21’ and suggests ‘swept along on fire’ (cf. Virg. Aen. 4.376); but, as Quinn notes, the present tense would clash with despexit, sensit (20–1). 20 humanos: Like homo itself, thought to be derived from ­humus (e.g. Priscian, GL i i .79.8–9 ab humo humanus ; Maltby s.v. ­humanus, humus, homo) and therefore adding an elemental aspect to the contrast between the sea creature Thetis and her earthly husband (Feeney (2007) 123). non despexit: Although her opinion is often not mentioned at all (cf. 19n.), there is no other version in which Thetis is said to be willing to marry Peleus (unless one counts Culex 300 hunc rapuit serua, ast illum Nereis amauit, where Seelentag ad loc. overstates the romantic implications of amauit in the context). Contrast rather the strand of the myth in which Peleus wrestles with her while she struggles to escape by metamorphosis (e.g. Pind. Nem. 3.35–6, 4.62–5, Soph. frr. 150, 618 TrGF, Ov. Met. 11.221–65, Paus. 5.18.5, Apollod. 3.13.5; see the Introduction, 2b(i)). Here non juxtaposed with the verb evokes Homer’s brief reference to her feelings, Il. 18.434 οὐκ ἐθέλουσα, to state the opposite meaning. Though surprising, this first indication of her consent is not strong compared to the Fates’ later references (328–36, 372–80): it is difficult to judge the level of emotion implied by the litotes (‘did not disdain’, ‘did not refuse’), which contrasts markedly with the description of Peleus’ emotion in 19. Literally ‘looked down on’, despexit also continues the motif of the gaze (12–18n.), contrasting with the Nereids coming up from the sea to look at the Argonauts.

168

COM M ENTAR Y: 21

Catullus combines the Greek tetrasyllable at line-end with another Greek metrical ‘licence’, the lengthening in arsi of a short closed syllable. This is consistent before hymenaeos in Catullus (62.4, 66.11) and in Virgil (Aen. 7.398, 10.720), where it seems to be a marker of allusion to Catullus (Wills (1996) 20 n. 18; there are no extant examples before ὑμέναιος in Greek hexameters). Cf. also V.Fl. 8.259 impediit hymenaeos and see Norden (19574) 450–2, Fordyce on Virg. Aen. 7.398, Harrison on Virg. Aen. 10.720. hymenaeos: ὑμήν and lengthened form ὑμέναι’ ὦ were ritual cries at wedding ceremonies (reproduced in poems 61 and 62); a version of the latter, ὑμέναιος/hymenaeus, can mean either ­‘wedding-song’ (62.4; cf. e.g. Il. 18.493, Pac. trag. fr. 88.1 Schierl, Lucr. 1.97, and see also 323–81n.), or, as here, often in the plural, synecdochically ‘wedding’ or ‘marriage’ (141, 66.11; attested from the fifth century, e.g. Soph. Ant. 813, Eur. Ion 1475; cf. e.g. Lucr. 4.1251). Catullus always has it at line-end and is followed by Virgil (15 out of 16 occurrences); this is also its position in the majority of appearances in Nonnus and Musaeus, the only two Greek hexameter poets to use it frequently, so that Kroll suspects a shared Hellenistic source. See further 141n. 21 Thetidi: Metre confirms Latin dative with -ıˉ, perhaps supporting the manuscript reading of Latin genitive Thetidis at 19 (3 Phasidos n.). Cf. Thetidıˉ (336) but Greek Minoidı˘ (247), Tethyı˘ (66.70). pater ipse: Usually interpreted as Jupiter, the line making the same point as 26–7. This in itself seems suspect: should the text say the same thing twice? It is not clear that ‘the father (himself)’ without defining genitive (contrast 27 diuum genitor, 298 pater diuum) would always be understood as a reference to Jupiter; apparent examples involve him in the context (Tib. 1.4.23, Virg. Georg. 1.121, Aen. 2.617) or feature defining activities, e.g. hurling thunderbolts (Lucr. 6.398, Virg. Georg. 1.328, 353, Hor. Carm. 3.29.44). Mayer (1980) makes a good case for a return to Renaissance commentators’ identification of pater ipse as Nereus (accepted by Thomson, Gaisser (1995) 584). The natural meaning of the phrase is the father of someone currently being mentioned, and this would be particularly expected in the context of a betrothal given the crucial role played by the bride’s father in arranging any marriage; cf. esp. 62.60–1 pater cui tradidit ipse, | ipse pater. Unlike Jupiter, Nereus has already been briefly alluded to (in 15 Nereides), and it would be odd to mention Thetis’ grandparents (29–30) without the most obviously concerned family member. The contrast with Ariadne’s relationship with Theseus being unsanctioned by her father (esp. 117) is also notable. On the other hand,

169

COM M ENTAR Y: 22

accounts of this particular marriage often seem to omit the father of the bride: see e.g. Gantz 230 (noting that Nereus is surprisingly far back in the procession on the François vase and absent altogether from the Erskine: 265–302n.), Hunter on Ap. Arg. 4.808–9, and cf. esp. Eur. IA 703 Ζεὺς ἠγγύησε καὶ δίδωσ᾿ ὁ κύριος with Fernandelli (2012) 266–7 n. 411. Yet at Alc. fr. 42.7 Voigt Peleus takes Thetis ἐκ Νή[ρ]ος … [μελάθρων]. I am not sure that the point can be decided. Cf. the ambiguity of Virg. Aen. 9.449 pater Romanus with Conington in Conington–Nettleship, and more generally Fowler (1996). ipse here, instead of emphasising a definite identity, paradoxically draws attention to these doubts (cf. Hardie (1993) 33, 39, 109, 113 n. 47 on ironic uses of epic ipse), and perhaps therefore to the tendentiousness of the entire presentation of the betrothal here. iugandum … sensit: iugare meaning ‘marry’ (poetic: TLL vi i .2.633.3–29) is first attested here and at Calvus fr. 6.1 Courtney = 31.1 Hollis, probably following Greek ζεύγνυμι (e.g. Soph. Trach. 536). sensit = ‘realised’, with the gerundive conveying the ‘must’ of fate, shows the pater, aware of the prophecy about Thetis (27n.), realising that Peleus’ love reveals him as her destined husband, and of the two possible identities fits Nereus particularly well since he is a prophet in his own right (e.g. Hes. Theog. 233–6, Pind. Pyth. 3.92, Ap. Arg. 1.1311, Hor. Carm. 1.15.4–5). Alternatively, if pater is taken to refer to Jupiter, sensit could be understood as ‘judged’ (e.g. Cic. Orat. 195, Leg. 3.39, Hor. CS 73–4 haec Iouem sentire deosque cunctos | spem bonam certamque d­ omum reporto). Pontanus’ sanxit ‘decreed’ is therefore unnecessary for Jupiter, and less appropriate for Nereus. 22 o marks the turn from narration to address; repeated o ­occurs in hymnic series of invocations (Norden (1913) 144). This o is also reminiscent of Catullus’ use of emotional, exclamatory o (e.g. 26.5, 43.8, 107.6): see further Ross (1969) 49–53. nimis: Baehrens and Fordyce point to 56.4 nimis iocosa (cf. 43.4, 93.1) to argue that this means simply ‘very much’, but in this elevated context it is highly unlikely to be read in this colloquial sense (cf. affected English ‘just too funny’); though more widespread with a negative (e.g. Cic. Fam. 12.30.7, Caes. BG 7.36.6), the use is not found in positive sentences after C ­ atullus. The choice of an explanation for the note of excess in ‘too much longed for’ is important for the reader’s perception of the narrator’s initial attitude to the heroic age (19–30n.): perhaps simply ‘too much to say how much’ (Ellis ‘happy beyond measure’); or, regretfully, ‘all too much’ for the peace of mind of men of today (cf. Virg. Georg. 2.458 o fortunatos nimium, Aen. 4.657 heu ­nimium

170

COM M ENTAR Y: 23–23b

­f­elix); or, bitterly, ‘too much’ because the heroic age was not really worth such desire (  Janan (1994) 108). optato: Stronger than ‘gratus iucundusque’ (Baehrens), ‘­felix’ (Kroll): the adjectival form in particular is associated in Catullus (and later in elegy) with erotic longing (5 optantes n.; cf. 62.30, 66.79, and Pichon s.v. optare). saeclorum tempore: For praise of heroes via the time in which they were born cf. Virg. Aen. 6.649 magnanimi heroes, nati melioribus annis. The periphrasis probably alludes to the myth of ages (first at Hes. Op. 106–201; cf. esp. Arat. Phaen. 100–36 and see further 31–49, 397–408nn.): the saecla would then be the five successive golden, silver, bronze, heroic, and iron races or ages of humans, from which the tempus of the heroes is singled out. The implication that the narrator knows about the whole series is complicated by the ambiguous status of the sailing of the Argo (1–11n.), by the appearance of golden-age motifs in Peleus’ and Thetis’ time (31–49n.), and especially by the focus in the epilogue on a sharp division between the age of theoxeny and the wicked present (382–408n.). Disyllabic saeclum, more tractable in dactylic verse than saeculum and restored here by R2, is the older form, used consistently by Catullus except at 95.6. Cf. 326, 367nn. 23–23b  The first half of 23b was restored by Orioli from an incomplete citation of Catullus in the note to Virg. Aen. 5.80 given by the Verona scholia (on which see Zetzel (2018) 265). The restoration produces a tricolon crescens in heroes, deum genus, bona matrum progenies : cf. esp. Arat. Phaen. 15 χαῖρε, πάτερ, μέγα θαῦμα, μέγ’ ἀνθρώποισιν ὄνειαρ. OGR o bona mater was obviously intended to make the line grammatical after 23b had been omitted, and editors once struggled to make sense of it as a reference either to Thetis or to the ‘mother’ Argo (see Ellis) in the middle of an address to the heroes (clearly also 24 uos); a reminder that apparent ineptitudes in Catullus’ text may be the result of unsuspected lacunae. heroes: Emphasised by enjambment from the previous line with following pause; this method of isolating a single word ‘in rejet’, found throughout Greek and Latin dactylic verse, is recurrent in the poem (e.g. 30, 32, 33, 44, 87, 108, where see n.), standing out against the typical end-stopping (see also the Introduction, 2d(ii)). Here the repetition of the technique in consecutive lines (23b progenies) creates by far the most violent use of enjambment so far, strongly conveying a heightened emotional tone.

171

COM M ENTAR Y: 23–23b

In Latin, heros is not usually particularly technical (TLL gives many more examples s.v. i ‘uir fortis’ than s.v. ii, iii, deified men or lares, ‘daemones’, etc.), but the rest of 23–23b almost defines heroes according to the widespread post-Homeric Greek understanding as the sons of one divine, one human parent (e.g. Hes. Op. 159–60, Pind. Pyth. 4.12–13, Plat. Rep. 3, 391d). This sense of a defining, all-encompassing address encourages the identification of these heroes with the heroes of the heroic age in general rather than just the Argonauts, though the specific echoes of Ap. Arg. 4.1773–5 rather suggest a hymnic farewell to them in particular (see below and 24n.). Cf. also 51n. saluete … | … saluete iter