Lydia, a Poem from the Appendix Vergiliana: Introduction, Text, Translation, and Commentary (Pseudepigrapha Latina) [Translation ed.] 0192874519, 9780192874511

This volume offers the first comprehensive literary and philological commentary on the Lydia, in any language. At its co

124 79 1MB

English Pages 192 [180] Year 2024

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover
Series Page
Lydia a Poem from the Appendix Vergiliana
Copyright
Acknowledgments
Preface
Contents
Introduction
1. The Lydia and the Dirae: Unity, Chronology, Authorship
2. Poetic Frameworks
3. A Note on the Language and Metre
4. Manuscript Tradition and the Text
Text and Apparatus Criticus
Translation
Commentary
Bibliography
Abbreviations
Early editions cited
Works cited
General index
Index of Greek and Latin words
Recommend Papers

Lydia, a Poem from the Appendix Vergiliana: Introduction, Text, Translation, and Commentary (Pseudepigrapha Latina) [Translation ed.]
 0192874519, 9780192874511

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

P SE U D E P IG R A P HA L AT I NA General Editors

AN TONY AU G OU STAKIS L AU REL FU L K ERSON ROBERT KAST E R C HRIST INA K R AUS

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

P SE U D E P IG R A P HA L AT I NA Pseudepigrapha Latina offers a new series of modern commentaries on frequently neglected, falsely attributed, and anonymous Latin texts. The individual commentaries engage with questions of authorship and dating, traditional philological issues and style, as well as the literary context of these works.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Lydia a Poem from the Appendix Vergiliana Introduction, Text, Translation, and Commentary B O R I S KAYAC H EV

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Boris Kayachev 2023 The moral rights of the author have been asserted All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2023942748 ISBN 978–0–19–287451–1 Printed and bound in the UK by Clays Ltd, Elcograf S.p.A. Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

For Teddy Gaudia semper enim tua me meminisse iuuabit.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Preface This commentary on the Lydia is one of a diptych, the other tackling its more popular sibling, the Dirae. Initially the two were intended to appear together between the same covers, but when the editors of the Pseudepigrapha Latina series offered to publish the Lydia half as an independent volume, this suddenly made sense: the Lydia has never before received a commentary under its own name (nor in fact any substantial piece of scholarship), and it deserves one. I am grateful to the editors and anonymous readers at OUP, and in primis to Antony Augoustakis, for showing an interest in the project—­and giving it this new form. Even if my treatment of the poem fails in every other respect, I hope the Lydia will have secured its own place, however small, on the shelves (and perhaps in the hearts) of classical scholars. But the decision came at a cost: separate publication makes it more difficult to cross-­ reference between the two commentaries, and some of my claims about the Dirae will have to be taken on faith, until its own commentary is published as well. In particular the reader needs to be warned that quota­tions from the Dirae follow the text established in the forthcoming work, without the specific textual arguments being repeated or even referenced here. Both commentaries were produced as part of a project that received funding from the European Union’s Horizon 2020 research and in­nov­ ation programme under the Marie Skłodowska-­Curie grant agreement No 840190. I am also grateful to Oxford’s Faculty of Classics, which hosted the project in 2019–21, and especially to Stephen Heyworth, who supervised it. It is difficult to overestimate Steve’s input: his many acute ideas on the text of the Lydia are explicitly acknowledged throughout the commentary (with an asterisk used to signal that they are previously unpublished), but there is barely a page that has not benefitted from his advice, offered either in person when we regularly met during my work on the commentary, or later in writing when he read a nearly complete

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

viii Preface typescript of the whole thing. I must also give thanks to Tristan Franklinos, who on several occasions joined our discussions, to the colleagues and students who attended a seminar on the Lydia and the Dirae in Hilary 2021, as well as to Lydia Matthews, who lent a hand of support at an initial stage of the project. I am pleased too to thank Gail Trimble, who let me have a glance at, and cite, her forthcoming commentary on Catullus 64. The bulk of this commentary was written during the infamous pandemic, most of which I spent confined within the walls (such as they are) of Wolfson College, Oxford. I owe it a debt of gratitude for awarding me a Non-­Stipendiary Research Fellowship and letting me join its supportive community of scholars, students, and staff. I am also greatly indebted to friends and colleagues, from both within and beyond the college, who through their presence, corporeal or otherwise, made the time of the lockdown seem more of an adventure than an outright disaster: Nikolay Andreev, Arik Avdokhin, Peter Budrin, Aliona Chepel, Ana Dall’Agnol, Richard Davies, Dmitry Dundua, Nikolay Epplée, Cosima Frieden, Mattias Gassman, Arina Guseva, Gregory Hutchinson, Ching-­ haun ‘George’ Lin, Jesse Lundquist, Robert Maltby, Angelo Marra, Valters Negribs, Basil Nelis, Damien Nelis, Dmitry Nikolaev, Ira Pavlova, Elena Racheva, Jenya Sedinkina, Elizaveta Shcherbakova, Mikhail Shumilin, Felix Tennie, Anke Walter, and Maria Yurovitskaya. I am as ever grateful to my parents, Luiza and Alexander, for their unfailing care and support. As I write these lines, my country of birth has for more than eight months been waging an unjust war against her neighbour. My ardent hope is that it will have stopped by the time this book is published. London, All Hallows’ Eve 2022

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Contents Introduction1 1. The Lydia and the Dirae: Unity, Chronology, Authorship 2.  Poetic Frameworks 3.  A Note on the Language and Metre 4.  Manuscript Tradition and the Text

3 14 39 41

Lydia45 Text and Apparatus Criticus 45 Translation49

Commentary51 Bibliography Indexes

151 163

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

IN T RODU C TION

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Introduction 1. The Lydia and the Dirae: Unity, Chronology, Authorship The Lydia came into existence in 1792, when Friedrich Jacobs realized that the last eighty lines (104–83) of the poem transmitted in the manu­ scripts of the Appendix Vergiliana under the title Dirae constitute in fact a separate poem in its own right.1 The title Lydia is in turn attested for a (lost) book by Valerius Cato, with (part of) which Jacobs identified the ‘newly discovered’ text, following Scaliger’s attribution to Cato of the ‘original’ Dirae.2 We shall later return to the question of authorship, but for the moment our focus will stay on that of unity. Jacobs adduced three fundamental arguments against the Dirae proper (1–103) and the Lydia being a single poem: (1) the former exhibits a clear ring com­pos­i­ tion, with the last stanza (98–103)—identified as such in the refrain (97 extremum)—echoing the first stanza (4–8);3 (2) the situations in the two poems are essentially different (in the Dirae the speaker is exiled from his farm, in the Lydia he is dying of lovesickness; in the Dirae the speaker invokes destruction upon his estate, in the Lydia the bucolic landscape continues to thrive); and (3) the style (‘tone’ in Jacobs’s terms) likewise differs notably between the two poems. Subsequent scholars developed and expanded these arguments, with some even coming to believe that the two poems are not by the same author.4 This extreme position had the regrettable consequence that several important studies from the last decades focus exclusively on the Dirae, with the Lydia remaining largely 1  Jacobs 1792. 2  Scaliger 1572, 433–4. 3  To this can be added that the Lydia too has a self-­contained symmetrical arrangement, if less pronounced, see my analysis at the head of the commentary. 4  Note esp. Rothstein 1888, echoed by Fraenkel 1966, 151: ‘Nowadays it is also generally agreed . . . that the two poems, the Dirae and the Lydia, cannot have been written by one and the same man. Metre, style, and mentality . . . are all entirely different.’ Goodyear 1971, 39 is right to be more cautious.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

4 INTRODUCTION overshadowed by its sibling.5 In his monograph, van der Graaf attempted to defend the unity of 1–183, but was duly rebuffed.6 In recent years, however, the unitarian position has gained in popularity, with scholars stressing points of contact between the two poems.7 Yet while these are real and do suggest close intertextual engagement, they can in no way overcome the positive indications that the Dirae and the Lydia are for­ mally separate poems.8 The realization that the Lydia is not part of the Dirae poses the ques­ tion of whether it is in fact a complete poem: as pointed out above, the Dirae exhibits a distinct ring composition, whereas the Lydia’s beginning and end are not explicitly signalled as such. Yet there are more subtle cues that speak against the suspicion of incompleteness. While the opening—­ inuideo uobis, agri—­may indeed take the reader by surprise, the first sen­ tence (104–6) announces the poem’s narrative situation and its dominant theme quite unambiguously: the speaker misses his absent beloved.9 Formal parallels for opening a poem with a first-­person in­di­ca­tive verb can be found, for instance, in Theocritus 12 (1 κωμάσδω ποτὶ τὰν ᾿Αμαρυλλίδα) and Bion’s Lament for Adonis (1 αἰάζω τὸν Ἄδωνιν); though it is true that κωμάσδω and αἰάζω are, unlike inuideo, performative verbs, the repetition of inuideo uobis, agri in refrains at 111 and 123 does give the phrase a metadiscursive dimension, thus supporting its appro­ priateness as an opening (αἰάζω τὸν Ἄδωνιν is similarly repeated in refrains throughout the Lament).10 The Lydia peters out somewhat inconspicuously (and the final lines are marred by textual uncertainty: see 182–3n.), yet although it cannot fully be ruled out that our text is 5  Note esp. Fraenkel 1966, Goodyear 1971, Rupprecht 2007, Breed 2012. 6 Van der Graaf 1945, 127–34; contrast e.g. Schutter 1953, Luiselli 1960, and Van den Abeele 1969, besides Fraenkel 1966 and Goodyear 1971. 7  See e.g. Salvatore 1994, Lorenz 2005, Stachon 2014, 178–200, Holzberg 2020, 16–18, Schniebs 2021. 8  Lorenz (2005, 4) was in fact prudent to admit as much: ‘Ob wir es tatsächlich mit einem oder zwei Gedichten zu tun haben, werde ich im Folgenden zunächst offen lassen. Wichtig ist, dass die Dirae als Einheit überliefert und die genannten Verbindungen zwischen beiden Texten oder Textteilen so offensichtlich sind, dass ihre gemeinsame Interpretation—­als ein Gedicht oder als zusammengehöriges Gedichtpaar—­unumgänglich ist. Dies würde selbst dann gelten, wenn ein weiterer Anonymus von den Dirae (vv. 1–103) zur Abfassung einer Lydia angeregt worden wäre und den vorhandenen Text gewissermaßen komplettiert hätte.’ 9  Cf. in a way the opening of Theocr. 12, ἤλυθες, ὦ φίλε κοῦρε· τρίτῃ σὺν νυκτὶ καὶ ἠοῖ | ἤλυθες· οἱ δὲ ποθεῦντες ἐν ἤματι γηράσκουσιν, which is similarly ‘unprepared’, but likewise specifies the poem’s circumstances—­the speaker’s much-­awaited encounter with his beloved—­clearly enough. 10  On the Lydia’s engagement with the Lament for Adonis, see section 2.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

The Lydia and the Dirae  5 truncated, the suggestion of impending death in 182–3 is a suitably closural move. The poem also appears to have a loosely symmetric arrangement in three sections of three ‘stanzas’ each (see the discussion at the head of the commentary), which likewise seems to confirm that it is essentially complete. Finally, against the possibility that the Lydia may have had (and lost) a narrative frame, it can be noted that the Dirae too is an unframed monologue of a bucolic speaker. Having established that the Lydia is likely to be a separate and complete poem, we are in a pos­i­ tion to assess its relationship to the Dirae. The most salient link between the two poems is the name Lydia, which in both is given to the rustic speaker’s beloved.11 Her identity is not explicitly specified in either poem, but both hint that she may be a kind of nymph, or at least that she has close ties with nature: in the Lydia she has the bucolic landscape all to herself, over which she exercises Orphic powers; in the Dirae Lydia is closely connected with the wood on the speaker’s estate (and with the estate in general) and likewise appears to be a singer (Lydia 110 interdum cantat mihi quae cantabat in aurem, Dirae 41 non iterum dices [sc. carmina] crebro quae, Lydia, dixti: note the formal structural similarity of the two lines).12 On the verbal level, both poems associate her with the landscape by pointedly using the same adjective of Lydia and a topographical feature in a single sentence (Lydia 105–6 hoc formosa [sc. prata] magis, mea quod formosa puella | in uobis nostrum tacite suspirat amorem, Dirae 89 dulcia rura, ualete, et Lydia dulcior illis).13 Both poems have in common the notion that the speaker enjoyed a time of happiness with Lydia, from whom he is now parted (Lydia 123–4 mea gaudia habetis | et uobis nunc est mea quae fuit ante uoluptas, Dirae 103 gaudia semper enim tua me meminisse iuuabit), albeit for different reasons and under different cir­cum­stances.14 Besides these 11  Fraenkel 1966, 151–3 was misguided to excise all references to Lydia from the Dirae, and has found little following, besides Goold in Fairclough and Goold 2000, 386–403; see contra Goodyear 1971, 32–9. 12  Cf., eloquently, Hubaux 1930, 49: ‘les deux poèmes s’inspirent d’une même sensibilité, très caractéristique, faite à la fois d’un vif sentiment de la nature et d’une sorte de galanterie qui n’est point commune chez les poètes latins: dans les deux poèmes, la beauté de la femme aimée est intimement associée à la beauté de la nature’ (pace Luiselli 1960, 104). 13  Note further Dirae 32–3 (cum) formosaeque cadent umbrae, formosior ipsa | silua cadet, of a wood, but in terms more applicable to a person, with the implication that Lydia may be con­ ceived of as a wood nymph. 14  In this respect, they can be contrasted with the typical comastic situation in which the lover addresses a reluctant beloved, as e.g. in Theocr. 3 and 20.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

6 INTRODUCTION thematic and textual resonances centred around the figure of Lydia, two further points of contact are worth registering.15 One involves flowers being referred to as an attribute of Venus (Lydia 116 uarios, Veneris spec­ tacula, flores, Dirae 20 Veneris uario florentia prata decore).16 The other case of coincidence seems more significant: both poems introduce the conceit of rivers either stopping or reversing their flow, and in doing so employ ‘disjunctive’ language to refer to the rivers’ water, and specifically the term lympha, for which such usage is unattested in pre-­Virgilian poetry and appears to allude to Varro’s etymology ab aquae lapsu lubrico (Lydia 121 tardabunt riui labentes murmure lymphas, Dirae 67 flectite currentes lymphas, uaga flumina, retro: see 121n.). On the macro-­level, both poems represent the protagonist’s soliloquies, with no narrative framing: the similarity is the more remarkable as there are no close ana­ logues in Virgil’s Eclogues (the fourth is likewise monologic, but is spoken in an authorial voice, not a bucolic character’s; the seventh is formally spoken by a single herdsman, but is effectively a restaging of a past contest).17 The closest formal comparanda come from Greek bucolic: Theocritean idylls 3, 12, and 20 resemble the Lydia in being unframed monologues of bucolic lovers, featuring series of erotic exempla taken from mythology (3.40–51, 20.34–41).18 In some ways, however, post-­ Theocritean bucolic supplies more relevant models: Bion’s Lament for Adonis and the anonymous Lament for Bion may have inspired the stan­ zaic organization of the Lydia and the Dirae, respectively.19 This shared intertextual heritage forms itself a link between the two Latin poems, as does their joint reception in subsequent poetry (see section 2). If the Lydia and the Dirae are separate but related poems, two ques­ tions arise: whether they are by the same author, and what their chrono­ logic­al relationship is. Fraenkel was probably the most outspoken 15 Enk 1919, 385–95 adduces even more potential similarities, which, however, are less cogent. 16  Though note the textual uncertainty at 116 (see n.). 17  Cf. further Hubaux 1930, 54: ‘Par leur fond, les Dirae et la Lydia appartiennent à ce genre de poèmes que les savants allemands appellent l’élégie subjective, c’est-­a-­dire que l’auteur y parle abondamment de lui, de ses sentiments, de ses amours’; ‘subjective bucolic’ could be a fitting generic description of the Lydia and the Dirae. 18  I owe this observation to Stephen Heyworth. 19  For the Lydia’s engagement with Bion’s Lament for Adonis, see section 2; I discuss the Dirae’s points of contact with the Lament for Bion in my forthcoming commentary on that poem.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

The Lydia and the Dirae  7 proponent of the theory of two authors: ‘Metre, style, and mentality—­if one can speak of mentality in that piece of smooth and shallow versifica­ tion, the Lydia—­are all entirely different.’20 Fraenkel’s perception of the Lydia’s aesthetic inferiority clearly played a crucial role in motivating its ascription to an author other than the Dirae’s.21 If, however, one puts aside subjective assessment of the poem’s artistic quality, the metrical and stylistic differences between the Lydia and the Dirae, though sup­ porting their separation, are insufficient to prove double authorship.22 While these seem real and not insignificant, there is no reason why a single author could not vary his writing mode between two com­pos­i­ tions, different in tone and possibly distant in time.23 Even though it cannot be demonstrated beyond doubt, in view of the poems’ joint transmission and the various links discussed above, single authorship appears to be the most economic hypothesis, which can be accepted as a working assumption until proven false. Whether or not the two poems are by the same author, in principle the Lydia can either pre-­date or post-­date the Dirae; if they are by the same author, they can also be contemporaneous, in the sense of being published simultaneously in a single poetic collection. The last scenario, however, appears to me unlikely, given the marked incompatibility between the Lydia’s and the Dirae’s narrative premises, and even if they were published together, the reader would be justified in wondering 20  Fraenkel 1966, 151. 21  In this he followed Rothstein’s (1888, 508–9) romantic condemnation of the Lydia’s artifi­ ciality: ‘In Dirarum poeta agnoscimus nativum quendam animi irati ardorem, qui multa cum acerbitate et desiderio rerum amissarum de gravi iniuria quam accepit queritur, quam tamen modice fert et fortiter; neque minus clare in Lydia apparet ars longa consuetudine exculta, suco tamen et nervis adeo carens, ut vix feramus hominem nihil maiore studio lectoribus obtru­ dentem quam macie se consumi et ad mortem voluntariam adigi, veri et simplicis doloris ne umbram quidem prodentem.’ In response, Enk 1919, 393–4 aptly noted that others had a higher opinion of the Lydia. 22  Pace Luiselli 1960. 23  Cf. Rodríguez Pantoja 2006, 591: ‘Estas diferencias dan la razón a quienes piensan en dos poemas literariamente diferentes. Ahora bien, los puntos comunes (varios de tipo general) también son apreciables, lo cual permitiría al menos no rechazar de plano la hipótesis (que con los datos aquí expuestos no puede ir ás allá) de un solo autor, que influido por la corriente neotérica, los habría compuesto tratando de acentuar formalmente las diferencias temáticas entre uno y otro, o bien en dos momentos distintos’; for metrical similarities between the two poems, cf. also Duckworth 1969, 85–6. Separate authorship could only be demonstrated if for­ mal characteristics indicated different periods of composition, but in both poems they point to a late Republican dating.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

8 INTRODUCTION about their dramatic chronology. A sequential reading of the two poems, however, is not unproblematic either. Some scholars have assumed that the Lydia describes a situation immediately subsequent to that of the Dirae: having parted from his beloved in the Dirae, the protagonist is dying of lovesickness in the Lydia.24 Yet placing the Lydia, in which both the heroine and the countryside are thriving, after the Dirae, in which both are condemned to destruction, will undermine the latter poem to quite an implausible degree. The reverse chronological arrangement seemingly faces a similar obstacle: the speaker appears to be dying in the Lydia, but is alive and well in the Dirae. However, while the disposses­ sion and exile suffered by the protagonist in the Dirae are presented as objective facts, his proclamation of impending death in the Lydia need not be taken as more than a desperate figure of speech or a reflection of his psychological state, and it is not inconceivable to imagine him re­united with Lydia in the aftermath. Besides, the Dirae contains two potential metaliterary references that might allude to the Lydia. One is 26 multum nostris cantata libellis: although used of Lydia’s wood (cf. 40–1) rather than of Lydia herself (who may, however, be thought of as a wood nymph), the phrase does imply an earlier poem or poems. The other is the Dirae’s very first line, Battare, cycnea repetamus carmina uoce: it is usually taken to imply that the Dirae is a reperformance of an earlier curse, but the notion of ‘singing a swan song again’ could at least as well hint at the Lydia, in which the speaker imagines himself to be dying. If, then, we are to coordinate the Lydia and the Dirae as parts of the same fictional uni­ verse (and the close ties between the two poems suggest that we should), the sequence Lydia–Dirae appears more prob­able, though we must also remain open to the possibility that other, lost, poems may have supplied further pieces of the puzzle that could alter the overall picture. The impression that the Lydia pre-­dates the Dirae is compatible with what can be established about their absolute chronology. The Dirae is a reaction to the veteran settlements of 41 bc, which likewise provided the background for Virgil’s first and ninth eclogues. As I argue in my 24 Cf. e.g. Luiselli 1960, 103: ‘Dovremmo concludere pertanto che il poeta autore delle Dirae, dopo aver cantato le maledizioni per i suoi campi, abbia in un secondo momento com­ posto la Lydia, per esprimere il suo dolore per la lontanza della sua donna, alla quale, del resto, già nelle Dirae si era rivolto con commossi accenti?’

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

The Lydia and the Dirae  9 forthcoming commentary on the Dirae, for that poem a plausible period of composition falls between the sack of Perusia in early 40 bc and the treaty of Misenum in early 39 (though it may conceivably be later by a few months or even years). The Lydia contains no perceptible historical allusions, and the style and metre, while in general compatible with a pre-­Virgilian dating, provide on the whole too shaky a ground to make any firm and specific inferences. There is, however, one formal feature which may help define the Lydia’s date of composition more narrowly, if far from conclusively. As has often been noted, a hallmark of the Lydia’s metrical style, both distinguishing it from the Dirae and creating a link with ‘neoteric’ poetry, are three σπονδειάζοντες; what has not been suffi­ ciently appreciated is that all three of them belong to a specific type, being formed by a quadrisyllabic verb-­form: 136 uestigare, 150 uolue­ bantur, 170 accumbebat.25 It is well known that spondaic verses thrived in Lucretius and ‘neoteric’ poetry, whereas later poets, starting with Virgil, used them much more sparingly. More importantly, not only did these later poets drastically reduce the frequency of spondaic verses in general, but also when they did write such verses, they favoured ending them with exotic-­sounding nouns rather than mundane and cumber­ some verbs. Here is a list of all spondaic verses produced by quadrisyl­ labic finite verbs (and infinitives) in hexameter poetry from Lucretius to Ovid: Lucr. 1.60 usurpare, 64 ostendebat, 764 dissoluuntur, 2.397 per­ manare, 1147 sustentare, 3.249 persentiscunt, 253 perturbentur, 545 obbrutescat, 4.975 usurpare, 978 obuersantur, 5.190 (= 425) pertemptare, Catull. 64.24 compellabo, 67 alludebant, 71 externauit, 80 uexarentur, 83 portarentur, 91 declinauit, 258 incingebant, 274 increbrescunt, 277 discedebant, Varro Arg. fr. 127.2 Hollis conclamarunt, fr. 128 lamentatur, Ciris 495 mutabantur, Verg. Georg. 1.221 abscondantur, Aen. 2.68 cir­ cumspexit, Ov. Met. 6.247 exhalarunt, 7.114 impleuerunt (to add ele­ giacs: Catull. 66.3 obscuretur, 41 adiurarit, 61 fulgeremus, 116.3 conarere, Ov. Her. 12.121 elisissent).26 The attitude to spondaic verses clearly 25  Strictly speaking, 136 uestigare is a conjecture for fastidire, but since the latter likewise produces a spondaic ending, this establishes prima facie evidence for the verse’s being a σπονδειάζων. 26  To match the types found in the Lydia, I have excluded spondaic verses produced by par­ ticiples and gerunds/gerundives (but they would not significantly affect the overall picture).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

10 INTRODUCTION changed between Catullus and Virgil, but our data—­incomplete as they are—­may in fact suggest a more nuanced picture. It does not look like a mere coincidence that out of the twelve surviving verses of Varro’s Argonautae, two are σπονδειάζοντες of our specific type. At the same time, it is remarkable that the Ciris only has one such verse, even though it employs other kinds of σπονδειάζοντες quite liberally. The exact date of composition of Varro’s Argonautae is unknown (though the 40s bc seem a likely option), and my dating of the Ciris within a few years after 44 bc is not uncontroversial.27 There is, however, evidence that the Ciris intertextually engages with the Argonautae, and a date of composition circa 45 bc for the latter appears not implausible.28 If that is the case, we may be able to see the fashion changing in the second half of the 40s bc. Of course, such a change does not happen overnight across the entire writing community; yet for a poet composing a bucolic poem in the wake of Virgil’s Eclogues, employing as many as three such σπονδειάζοντες in the span of some eighty lines would be a decidedly retrograde move. The fact that the Dirae has no spondaic verses need not imply by itself that the Lydia is the earlier poem, since it clearly reflects a difference in tone, but the Lydia’s σπονδειάζοντες do suggest the early-­to-­mid-­40s bc as a plausible chronological setting. It is time to revisit the question of authorship. Scaliger was the first to attribute the Dirae to Valerius Cato, taking it to be a doublet of the Indignatio, in which Cato complained about the loss of his inherited estate licentia Sullani temporis (Suet. Gram. 11.1).29 When Jacobs real­ ized that lines 104–83 form a separate poem, it was only natural that he identified it as part of Cato’s Lydia (taking the latter to be a collection rather than a single poem).30 Since the Dirae can securely be related to the veteran settlements of 41 bc, Scaliger’s attribution of it to Valerius Cato has little substance, and it would appear that as a consequence there is no good reason for ascribing the Lydia to Cato either. Yet the latter attribution has taken on a life of its own, and should be evaluated on its

27  On the Argonautae, cf. Hollis 2007, 178; on the Ciris, see Kayachev 2020, 5–30. 28  For Varro’s possible influence on the Ciris, see Kayachev 2020, 90 ad 32, 93 ad 49–51, 124 ad 184, 174 ad 501, 503, 504. 29  Scaliger 1572, 433–4.    30  Jacobs 1792, 57–9.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

The Lydia and the Dirae  11 own merits. Our only testimony for Cato’s Lydia (besides Suetonius’ attestation of the title at Gram. 11.3) is a fragment by Ticida (likewise preserved by Suetonius): Lydia doctorum maxima cura liber (fr. 103 Hollis). As Hollis comments, Ticida’s fragment ‘suggests that Valerius Cato’s Lydia may have needed (and received) learned exegesis, perhaps a full scholarly commentary’,31 which would seem incompatible with our Lydia, a poem of only some eighty lines and conveying a strong sense of rustic naivety. Courtney conveniently sums up the sceptical case against the ascription to Cato: The inamorata of the poet of the Dirae, who wrote after Vergil’s Bucolics, was called Lydia. This caused someone to attach to it a poem of 80 lines about a girl called Lydia, by a different author writing much more in the style of Catullus . . . but probably later than Catullus, the Bucolics . . ., and Gallus . . . . These factors make it more than difficult to identify the pseudo-­Vergilian poem with that of Valerius Cato; not to mention the triviality of the production, which could hardly have raised such a sen­ sation as Ticida describes. Against this last point it is argued that we may have only one poem from a sequence entitled Lydia, like the Cynthia of  Propertius; but it is hard to conceive such a sequence written in hexameters.32

We have just seen that a pre-­Virgilian dating is not unlikely on internal grounds (the Lydia’s points of contact with the Eclogues will be discussed in section 2). To Courtney’s final objection, it can be replied that our Lydia need not come from a collection of hexameter poetry (and it is indeed unlikely that there existed a whole book of Latin bucolics prior to the Eclogues), but could easily have been the centrepiece of a polyme­ tric collection: we do not know if the Catullan sequence of poems 61–8 reflects authorial arrangement, but from a later period Statius’ Silvae show that there is nothing impossible in combining hexameters with other metres within a single book.33 Courtney’s allegation of triviality, echoing Fraenkel and Rothstein, mistakes the Lydia’s studied naïveté for 31  Hollis 2007, 163.    32  Courtney 1993, 191. 33  Ticida’s term liber does suggest a collection in hundreds of lines, but so might Dirae 26 multum nostris cantata libellis.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

12 INTRODUCTION a quality of its author rather than of its speaker. The text of the Lydia is often uncertain and for the most part understudied, which makes it dif­ ficult to assess the depth and extent of its poetic doctrina. As is argued in the next section and throughout the commentary, however, the Lydia appears to be conversant with a range of Hellenistic poets. A prime example is the citation of Ariadne as a happy case of erotic relationship at 152, which follows Apollonius’ Jason in pointedly suppressing the outcome of her affair with Theseus (see 152–3n.). More than once the Lydia makes subtle allusions to recondite mythological stories, without turning the bucolic protagonist into an implausible poeta doctus. For instance, 149 is a transparent reference to the aetiology of the Pleiades, catasterized to honour their affairs with the Olympian gods; at the same time, it also seems to imply the learned tradition that the seventh Pleiad was made all but invisible, because she had a relationship with a mortal man (see n.). Line 166 similarly alludes to famous stories of Zeus’s assuming various disguises in his pursuit of women, but may also hint at the obscure account of his turning into a cuckoo to seduce Hera (see n.). Lines 172–4 have a touch of burlesque in their explanation of why Vulcan and Mars did not interfere in Venus’ affair with Adonis, but at the same time evoke the less-­known tradition that Adonis was in fact killed by Ares out of jealousy. The subtlety with which the Lydia poet demonstrates his sophistication while maintaining the speaker’s rustic character can also be evidenced in the use of Phoebus to refer first to the sun god at 143 and then to Apollo at 146: the two deities are consistently distinguished in the poetry of the first century bc, whereas the speaker appears to be confusing them; the epithet Phoebus, however, could be applied to either, and by using it the poet delicately signals that he is aware of the difference (see 146n.). Another aspect of the Lydia’s linguis­ tic doctrina is etymological wordplay. A programmatic instance is the figura etymologica 108 Lydia ludit, evoking an etymology attested in Varro (see n.). A more striking example is supplied by 121 labentes mur­ mure lymphas, where the innovative use of lympha for running water appears to be based on Varro’s etymology ab aquae lapsu lubrico (see n.). A further touch of verbal sophistication may be found in 131–2 a te | uaccula non umquam secreta cubilia captans, where the separation of a te from its governing adjective secreta, not only by three words but also by

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

The Lydia and the Dirae  13 a line break, is intended to visualize the separation of the cow from the bull (cf. 131n.). Whether or not these examples put the Lydia as a learned poem in the same league as Cinna’s Smyrna, Ticida’s characterization of Cato’s Lydia as doctorum maxima cura need not be taken as absolutely objective and impartial, but might rather reflect an aspiration implicit in the poem.34 While the Lydia may not be showing off obscure learning, it does make a point of presenting its eponymous character as a docta puella (127–8 non ulla puella | doctior in terris fuit, involving a Sapphic allusion: see n.), and not just in the sense of exacting audience common in elegy (e.g. Prop. 2.13.11–12 me iuuat in gremio doctae legisse puellae, | auribus et puris scripta probasse mea, cf. Verg. Ecl. 10.2 quae legat ipsa Lycoris), but as a poet in her own right (see also section 2).35 One could see Ticida’s epigram acknowledging this claim: he makes doctrina into an attribute of Lydia’s (male) audience, but by implication it also applies to her, and the epigram’s play consists in the ambiguity as to whether Lydia pos­ sesses learning as a person or a book (or indeed both). Another way in which the epigram could be seen to be reflecting our poem is the use of cura in reference to Lydia. The nuance given to this word is not quite the same in the two texts, but both appear to be exploring its conversational usage for ‘(my) love’: the epigram ostensibly uses cura in its neutral sense ‘object of care’ but clearly implies an erotic innuendo, whereas the Lydia pushes the amatory usage by giving cura a referential force (see 122n.).36 These points of contact may not prove that our Lydia is the Lydia read by Ticida, but they make the idea rather more attractive. To recapitulate, the Lydia may have been written by Valerius Cato in the 40s bc, prior to both the Dirae and Virgil’s Eclogues. A nineteenth-­ century scholar would have presented this conclusion as a solid fact—­ until the next scholar ‘proved’ him wrong. My claim is both more cautious 34  Sectarianism may have played a role too: note that Messalla names, and rejects, Ticida and Valerius Cato in one breath: non esse sibi dicens rem cum Furio Bibaculo, ne cum Ticida quidem aut litteratore Catone (Suet. Gram. 4.3, fr. 100 Hollis). 35  Though note that Cynthia too can be presented as a poetess: Prop. 2.3.21–2 et sua cum antiquae committit scripta Corinnae | carminaque Erinnae non putat aequa suis. 36  For the trope by which the relationship between the poem and the commentator is pre­ sented in erotic terms, cf. esp. the anonymous epigram on Crassicius’ commentary on Cinna’s Smyrna (quoted in Suet. Gram. 18.2), with Conrau-­Lewis 2020.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

14 INTRODUCTION and, I hope, more resilient: I believe the available evidence is consistent with some such scenario, but the nature of the problem is of a kind that makes it impossible either to embrace or to reject the solution with absolute confidence, at least in our current state of knowledge. While a conceivable reaction in the face of this uncertainty might be simply to take the Lydia out of the equation in any historical account of Latin ­literature, a more profitable approach will consist in exploring the poem in the context of the poetry of the first century bc as if it were written by Valerius Cato.37 It is a self-­evident fact that the surviving texts of known poets of the first century bc only represent a fraction, however signifi­ cant, of the poetic production of the period, so it is rea­son­able to expect that at least some of the ‘missing’ pieces should turn up among texts transmitted anonymously or under a false attribution. This is not to sug­ gest that all texts from, for example, the Appendix Vergiliana are ‘lost masterpieces’ by leading poets of the first century bc—­quite possibly some are ‘fakes’ (intentional pseudepigrapha), while others simply come from a later period—­but it seems perverse to insist in principle that not a single one could indeed belong to a known con­tem­por­ary of Virgil’s (if not to Virgil himself).38 Whether or not the Lydia actually falls into this category (which we may never know for certain), reading it as such a text can at the very least help us appreciate how limit­ed and lacunose our picture of the golden age of Latin poetry is.

2.  Poetic Frameworks Recent scholarship on the Lydia has highlighted its affinities with Roman elegy and Virgil’s tenth eclogue, as a rule under the assumption that our poem is the dependent text.39 We shall consider (some of) these inter­ texts later, but first I propose to explore the Lydia’s engagement with Greek bucolic. The central contention of the following argument is that 37  Elsewhere I have made similar proposals for a number of other pseudo-­Virgilian texts: the Ciris (Cornelius Gallus?), Catalepton 9 (Valgius Rufus?), and the so-­called Aeneid pre-­ proem (Varius Rufus?); see Kayachev 2016, 2020a, and 2011, respectively. 38  Cf. in general Kayachev 2021a. 39  See esp. Ballarini 1988/9, Lorenz 2005, Rosati 2020.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  15 the figure of the speaker is in important ways modelled on that of the dying Daphnis in Theocritus’ first idyll, and that in general the poem follows in the footsteps of a tradition initiated by Thyrsis’ song in that idyll and continued by Bion’s Lament for Adonis and the anonymous Lament for Bion.40 While the Lydia’s links with Greek poetry do not as such prove that it pre-­dates Virgil, they make it easier to accept that it does not require the Eclogues to account for its existence in genealogical terms.41 I round off the discussion by adducing a number of intertexts from Augustan poetry, which I explore as possible early reactions to the Lydia. The central point of resemblance between Theocritus’ Daphnis and the Lydia speaker is that both are ‘dying’ of lovesickness, in mysterious circumstances (1.130 ἦ γὰρ ἐγὼν ὑπ᾽ ἔρωτος ἐς ῞Αιδαν ἕλκομαι ἤδη, cf. 158 iam mea matura cogor nece soluere fata: note the shared passive; 1.66 Δάϕνις ἐτάκετο, cf. 125 at mihi tabescunt morientia membra dolore). In the standard version of the myth Daphnis is punished for infidelity (or violation of a vow of chastity), whereas in Theocritus he appears to be cursed with a passion for an unattainable woman, perhaps because he boasted that he was immune to love, though any details can only be guessed.42 In the Lydia, the situation appears to be clearer: the speaker suffers because he is separated from his Lydia (127 quod mea non mecum domina est), which in turn he sees as a punishment for defiling her vir­ ginity (156–60). Yet the essential uncertainty persists: it is not explained how the latter leads to the former (and while we know where the speaker is not, we are not told where he is), nor indeed why a sexual relationship between Lydia and the speaker is in any way reprehensible or detrimen­ tal (there is no explicit indication that either of them is bound by a vow of chastity, and the context rather suggests that sex as such, or at any rate

40  For Theocr. 1, Bion’s Lament for Adonis, and the anonymous Lament for Bion as forming a distinct strain of the bucolic tradition, see esp. Hubbard 1993 (reprinted in Hubbard 1998, 19–44) and Kania 2012. 41  For earlier attempts to relate the Lydia to Greek bucolic, cf. e.g. Mackail 1908, 71, Lindsay 1918, 62, Hubaux 1930, 53–6. 42 On Daphnis and his situation in Theocr. 1, see Fantuzzi 1995, Hunter 1999, 63–8, Lightfoot 1999, 526–8, Anagnostou-­ Laoutides and Konstan 2008, Anagnostou-­ Laoutides 2013, 105–97; on his reception in post-­Theocritean bucolic, cf. Gagliardi 2019; see also in ­general Scholl 2014.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

16 INTRODUCTION sex out of wedlock, is considered to be an offence: an outlook quite untypical of erotic poetry). This is, so to speak, a similarity ex silentio, and as such it may not amount to much, but it is silence on a matter the reader will not fail to be curious about, and in both cases must be inten­ tional and pointed. Another motif that links the figures of Daphnis and the Lydia speaker is that of a πρῶτος εὑρετής. Although Theocritus is not quite explicit on the point in the first idyll, in other ancient sources Daphnis often appears as either the inventor of bucolic poetry or the subject of the first bucolic song.43 In the Lydia, this motif takes a different form, but is no less prominent: the speaker at length complains that he cannot claim to be the inventor of love (156–65). This is of course a point of contrast rather than resemblance, but the difference is likely to be meaningful. It is in fact remarkable that in the Lydia there are no explicit indica­ tions that the speaker is a bucolic singer, other than the mere circum­ stance that he speaks in hexameters.44 By contrast, Lydia herself is presented in two contexts as composing and performing poetry (108–10 uobis mea Lydia ludit | et modo submissa meditatur carmina uoce, | inter­ dum cantat mihi quae cantabat in aurem, with 109–10n.; at 119–22, quoted below, Lydia’s words are not explicitly defined as poetry, but formal per­ formance is implied by the presence of an audience), and in a third con­ text her doctrina is praised (127–8). In this the Lydia shows remarkable similarity with a doublet of the Daphnis myth, in which the role of the inventor of bucolic poetry is ascribed not to Daphnis’ counterpart Menalcas, but to his female lover (Clearchus, Erotica fr. 32 Wehrli):45 Ἠριϕανὶς ἡ μελοποιὸς Μενάλκου κυνηγετοῦντος ἐρασθεῖσα ἐθήρευεν μεταθέουσα ταῖς ἐπιθυμίαις. ϕοιτῶσα γὰρ καὶ πλανωμένη πάντας τοὺς ὀρείους ἐπεξῄει δρυμούς . . . ὥστε μὴ μόνον τῶν ἀνθρώπων τοὺς ἀστοργίᾳ διαϕέροντας, ἀλλὰ καὶ τῶν θηρῶν τοὺς ἀνημερωτάτους 43  Theocritus may in fact be conflating both ideas: Daphnis is a bucolic singer of such a c­alibre that he can bequeath his pipe to Pan himself (1.128–9 ἔνθ᾽, ὦναξ, καὶ τάνδε φέρευ πακτοῖο μελίπνουν | ἐκ κηρῶ σύριγγα καλὸν περὶ χεῖλος ἑλικτάν), and at the same time he is the subject of what appears to be a traditional bucolic theme, Δάφνιδος ἄλγεα (1.19, cf. 5.20). 44  Against reading mea (carmina) at 109, see 109–10n. 45 For a discussion, see Gutzwiller 1983, 173, Hunter 1999, 66, Bouchard 2020, 194–5, Massimilla 2021, 133–4.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  17 συνδακρῦσαι τῷ πάθει, λαβόντας αἴσθησιν ἐρωτικῆς ἐλπίδος. ὅθεν ἐποίησέ τε καὶ ποιήσασα περιῄει κατὰ τὴν ἐρημίαν, ὥς ϕασιν, ἀναβοῶσα καὶ ᾄδουσα τὸ καλούμενον νόμιον, ἐν ᾧ ἐστιν “μακραὶ δρύες, ὦ Μέναλκα”. Eriphanis the lyric poetess fell in love with Menalcas as he was hunting and chased after him, driven hither and thither by her desires. As she roamed and wandered, she searched through every mountain thicket . . . so that not only the people who stood out for their heartlessness but also the most savage beasts were brought to tears by her misery, once they observed her passionate hope. That was why she both composed poetry and then went all over the wilderness, as they say, shouting and singing the so-­called pastoral song, which has ‘The oaks are tall, Menalcas’.

Besides the fact that both texts offer, so to speak, a feminist version of the invention of bucolic poetry (note that both involve a girl not just singing but also composing songs), another significant point of conver­ gence between the Lydia and the Menalcas—­or rather Eriphanis—­story can be found in their use of pathetic fallacy (119–22): gaudebunt siluae, gaudebunt mollia prata, tardabunt riui labentes gurgite lymphas et gelidi fontes, auiumque silentia fient, dum mea iucundas expromet cura querelas.

As we shall see shortly, the standard application of the conceit is in con­ texts expressing grief over a dead person, whereas in both these texts nature is presented as sympathizing with a girl in love. This must remain a speculation, but it is tempting to entertain the idea that Lydia is mo­delled on Eriphanis. From the viewpoint of the later European poetic tradition it may appear that pathetic fallacy is ubiquitous, but if one understands it in the narrower sense, excluding other kinds of address to and/or personifica­ tion of natural features and animals, until Virgil’s Eclogues the device only appears sporadically in Graeco-­Roman literature. Its most prom­i­ n­ent pre-­Virgilian occurrences are in the three texts we are closely con­ cerned with here: Theocritus’ first idyll, Bion’s Lament for Adonis, and the anonymous Lament for Bion. In turn, it has been plausibly suggested

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

18 INTRODUCTION that this tradition can be traced back to near-­ eastern laments for Dumuzi (and related figures).46 A notable feature of the conceit, likewise going back to near-­eastern sources, is that it usually involves, at least in its more developed examples, mini-­catalogues listing the various natural features and/or animals participating in the lament. The archetypal pas­ sage, for the classical tradition, is Theocritus 1.71–5: τῆνον μὰν θῶες, τῆνον λύκοι ὠρύσαντο, τῆνον χὡκ δρυμοῖο λέων ἔκλαυσε θανόντα. … πολλαί οἱ πὰρ ποσσὶ βόες, πολλοὶ δέ τε ταῦροι, πολλαὶ δ᾽ αὖ δαμάλαι καὶ πόρτιες ὠδύραντο. For him the jackals, for him the wolves howled, for his death even the lion in the thicket wept. . . . At his feet many cows, many bulls, many heifers and calves wailed.

Although Theocritus refers to animals and the Lydia mainly to inani­ mate nature, and although the emotions expressed are different, the for­ mal stylistic similarity is apparent (if not very striking). We shall later have an occasion to consider the change in mood between the two texts more closely, but at the moment it can perhaps be speculated that the repeated gaudebunt at Lydia 119 may be intended as a ‘mistranslation’ of Daphnis’ farewells in a related passage (1.115–18):47 ὦ λύκοι, ὦ θῶες, ὦ ἀν᾽ ὤρεα ϕωλάδες ἄρκτοι, χαίρεθ᾽· ὁ βουκόλος ὔμμιν ἐγὼ Δάϕνις οὐκέτ᾽ ἀν᾽ ὕλαν, οὐκέτ᾽ ἀνὰ δρυμώς, οὐκ ἄλσεα. χαῖρ᾽, Ἀρέθοισα, καὶ ποταμοὶ τοὶ χεῖτε καλὸν κατὰ Θύβριδος ὕδωρ. Wolves, jackals, and bears that live in mountain caves, farewell: I, Daphnis the oxherd, have forever left your forests, woods, and groves. Farewell, Arethusa, and you rivers that roll your pure waters down the Thybris. 46  See esp. Griffin 1992, 204–11; cf. Hunter 1999, 68, Anagnostou-­Laoutides and Konstan 2008, 500–2, Anagnostou-­Laoutides 2016, 62–3. 47  Or rather, an overly literal translation: Daphnis bids the animals and natural features to ‘rejoice’, and that is precisely what they do in the Lydia.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  19 Above I pointed out that address to nature does not as such constitute pathetic fallacy, but here Daphnis’ apostrophe to the animals is clearly intended as an answer to their weeping (note that λύκοι and θῶες are named in the first lines of both passages; add that 1.120–1 echo 1.74–5). An additional point of contact with the Lydia is here established by the ‘disjunctive’ reference to rivers and their waters (ποταμοὶ . . . ὕδωρ, cf. 121 riui . . . lymphas), as well as, in close proximity, to springs (Ἀρέθοισα, cf. 120 fontes). While pathetic fallacy may be too general a similarity to establish direct relationship between Theocritus 1 and the Lydia, the following point of resemblance is less of a commonplace. In Theocritus Daphnis (a cowherd) is compared with the goatherd, the proverbial example of frustrated love (1.87–8): ᾡπόλος, ὅκκ᾽ ἐσορῇ τὰς μηκάδας οἷα βατεῦνται, τάκεται ὀϕθαλμὼς ὅτι οὐ τράγος αὐτὸς ἔγεντο. The goatherd, when he sees the she-­goats being mounted, weeps his eyes out because he wasn’t born a he-­goat.

How exactly this comparison fits the figure of Daphnis is not entirely clear,48 but the goatherd envying the he-­goat who can easily satisfy his desires with the she-­goats clearly parallels the Lydia speaker who is envi­ ous of the bull and the he-­goat (and other male animals) because they never part from their mates (131–41). This is a striking similarity (even if the reason for envy is not exactly the same in both cases), and it is the more significant as the theme is developed at far greater length in the Lydia. In fact the motif of envy—­for natural features, animals, or gods and heroes—­begins (with 1 inuideo) and pervades nearly two-thirds of the poem (see 104–30n., 131–55n.). One final point of intersection with the Lydia can be found in Daphnis’ contemptuous address to Aphrodite, in which he reminds her of her affairs with mortals (1.105–10):

48  See Hunter 1999, 92–3.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

20 INTRODUCTION οὐ λέγεται τὰν Κύπριν ὁ βουκόλος; ἕρπε ποτ᾽ Ἴδαν, ἕρπε ποτ᾽ Ἀγχίσαν· τηνεὶ δρύες ἠδὲ κύπειρος, αἱ δὲ καλὸν βομβεῦντι ποτὶ σμάνεσσι μέλισσαι. … ὡραῖος χὤδωνις, ἐπεὶ καὶ μῆλα νομεύει καὶ πτῶκας βάλλει καὶ θηρία πάντα διώκει. Don’t they say the cowherd did Cypris? Go to Ida, go to Anchises: there grow oaks and galingale, and the bees hum softly at their hives. . . .  Adonis too is youthful, for he herds sheep, shoots hares, and chases all sorts of animal.

Again, the exact relevance of these allusions is not fully transparent, though it has been argued—­interestingly if inconclusively—­that the girl with whom Daphnis is desperately in love is none other than Aphrodite herself.49 Whether or not this particular interpretation is correct, it is evident that the mythological exempla are intended to serve, in one way or another, as a foil to Daphnis’ own situation.50 The similarity with the Lydia is twofold. On the formal level, the Lydia speaker likewise adduces mythological examples of gods involved in erotic relationships with mortals (169–76, but also 142–5 and 152), one of which is similarly introduced with a negative rhetorical question (175 non Aurora . . . ? cf. οὐ λέγεται τὰν Κύπριν ὁ βουκόλος;). On the level of content, the story of Adonis and Aphrodite features prominently in both texts: in both cases only the positive period of the relationship is explicitly referred to, but in both cases it is clear that the reader is expected to remember how this relationship ended and to relate it to the situations of the poems’ protagonists. While in terms of myth and ritual the figure of Daphnis appears to be shaped, in certain respects, after that of Adonis and his near-­eastern prototypes, in literary terms Thyrsis’ song in Theocritus’ first idyll is a central poetic model behind Bion’s Lament for Adonis.51 This complex relationship can best be exemplified with the case of pathetic fallacy: 49  See Anagnostou-­Laoutides and Konstan 2008. 50  Cf. the mythological exempla in Theocr. 3. 51  For Bion’s engagement with Theocr. 1, see Reed 1997, 22, besides the works cited above in n. 36.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  21 while the thematic link from Bion’s Lament to the near-­eastern dirges is more direct than that to the Theocritean Δάϕνιδος ἄλγεα, on the level of expression Bion’s debt to Theocritus is unmistakable (note e.g. how 18 τῆνον μὲν περὶ παῖδα ϕίλοι κύνες ὠρύονται is modelled on 1.71 τῆνον μὰν θῶες, τῆνον λύκοι ὠρύσαντο). Within this framework, we can see that in some details the Lydia comes closer to Bion than to Theocritus in its application of pathetic fallacy. Above it was suggested that the image of rivers and springs listening to Lydia (121–120) can be linked to Daphnis’ farewell to the spring of Arethusa and the Sicilian rivers (1.117–18); Bion’s depiction of rivers and springs mourning for Adonis is in a num­ ber of ways a closer parallel (33–4): καὶ ποταμοὶ κλαίοντι τὰ πένθεα τᾶς Ἀϕροδίτας, καὶ παγαὶ τὸν Ἄδωνιν ἐν ὤρεσι δακρύοντι. And the rivers cry for Aphrodite’s sorrows, and the mountain springs weep for Adonis.

To begin with, in contrast to Theocritus, here we have a clear case of pathetic fallacy, with natural features actively expressing an emotion, rather than merely being addressed. Another noteworthy point of resemblance consists in the fact that while the springs are said straight­ forwardly to weep for Adonis, the rivers are rather presented as sympa­ thizing with Aphrodite: in the Lydia too, and more consistently, nature’s emotional response is directed not at the dying protagonist, but at his ‘widowed’ girlfriend.52 This shared detail highlights a more general aspect in which the Lament and the Lydia are closer to each other than to Theocritus: whereas in Thyrsis’ song the focus is almost exclusively on Daphnis, in the two later poems the protagonists’ female lovers receive almost equal attention. Aphrodite and Lydia can in fact even be seen, from a certain perspective, as the more central characters: as noted above, it is quite striking that in the Latin poem only Lydia, and not the speaker, is explicitly presented as a bucolic poet, and to an extent simi­ larly Aphrodite’s is, besides the narrator’s, the most prominent voice in 52  Above this feature was linked with the Eriphanis story, but one intertext need not exclude another.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

22 INTRODUCTION the Lament (42–61), while her dirge is accorded the status of a formal performance by the concluding reference to its annual repetition (97–8). If the figures of Lydia and Bion’s Aphrodite are indeed intertextually related, the reader will no doubt be expected to recognize differences as well as similarities. One stanza of the Latin poem draws an idyllic image of Lydia walking around in the fields or lying down on the grass and singing about her love for the speaker (112–18); the contrast with Aphrodite desperately searching for Adonis could not be starker (20–4): λυσαμένα πλοκαμῖδας ἀνὰ δρυμὼς ἀλάληται πενθαλέα νήπλεκτος ἀσάνδαλος, αἱ δὲ βάτοι νιν ἐρχομέναν κείροντι καὶ ἱερὸν αἷμα δρέπονται· ὀξὺ δὲ κωκύοισα δι᾽ ἄγκεα μακρὰ ϕορεῖται Ἀσσύριον βοόωσα πόσιν καὶ παῖδα καλεῦσα. She lets her locks loose and wanders in the woods, heartbroken, dishevelled, unshod. The brambles scratch her as she passes by and draw her holy blood. With shrill wails, she roams through remote ­valleys, shouting for and calling her Assyrian spouse and boy.

The general structural similarity of the passages—­both women are first situated within a natural environment, and then a reference is made to their calling out for their absent lovers—­may not suffice to establish a connection by itself, but it provides a framework in which to consider more specific points of contact. That both are walking barefoot may likewise be no more than a commonplace, but the focus on how Lydia ‘aggressively’ affects the countryside—­she plucks unripe grapes (114 decerpserit uuam) and crushes soft grass (117 illiserit herbam)—looks like a pointed inversion of the passive suffering endured by Aphrodite as she is scratched and bled by brambles (21–2). In fact both grapes and grass can metaphorically suggest the (female) body as a sexual object (see 114–15n., 116–17n.), and in turn Bion’s use of δρέπεσθαι (‘to pluck’, equivalent to decerpere) is an agricultural metaphor, possibly suggesting the idea of sexual violence.53 This connection is further supported by a 53  On the metaphor, cf. Reed 1997, 207, though he does not go as far as to suggest a sexual interpretation.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  23 pointed internal echo: the image of Lydia lying down on soft grass among flowers (116–17) is unmistakably intended to foreshadow the description of Venus and Adonis having sex alfresco in an identical environment later in the poem (169–70). An explicit reference to Venus here (116 uarios, Veneris spectacula, flores) additionally highlights this intratextual link, as well as the connection with Bion. The pointed parallel between the images of Lydia lying in the grass and of Venus and Adonis’ embrace has another important implication: the reader is thus invited to relate the speaker’s situation to that of the dying Adonis, even if it is only the happy period of Venus and Adonis’ affair that is explicitly mentioned in the text. In fact, while lacking exact parallels with the Lament, the Venus and Adonis section of the Lydia does feature suggestive resonances. One point of pregnant contrast is produced by the different conceptions of where Venus and Adonis usu­ ally had sex: the Lament implies that it was on a proper bed with luxuri­ ous sheets (70–4, cf. 3, 79), whereas in the Lydia they are presented as lying on a bed of soft grass and flowers (169–70)—which rather resem­ bles the position of Adonis’ corpse in the Lament (69 οὐκ ἀγαθὰ στιβάς ἐστιν Ἀδώνιδι ϕυλλὰς ἐρήμα).54 It may also be relevant that ‘purple’ is the colour of beddings in the Lament (3 πορϕυρέοις ἐνὶ ϕάρεσι, 79 ἐν εἵμασι πορϕυρέοισιν), and of flowers in the Lydia (170 purpureos flores), though flowers do feature prominently in the Lament as well (esp. 65–6, 75–6). That both texts describe flesh as ‘white’ (171 candida . . . bracchia, cf. 113 pedis niuei; 10 χιονέας κατὰ σαρκός, 26–7 μαζοὶ | χιόνεοι) is of little import, but the shared focus on Adonis’ beauty may be more significant (171 formoso . . . collo, 71 καὶ νέκυς ὢν καλός ἐστι, καλὸς νέκυς, οἷα καθεύδων). There is also a potential parallel between the presentations of the dying Adonis and the Lydia speaker: in the Lament, Adonis is described as λεπτὸν ἀποψύχων, which in the context means ‘faintly breathing his last’, but could also mean ‘slowly growing cold’, and the latter interpretation

54  One might in fact wonder whether the structural similarity between Ep. Adon. 72–3 κάτθεό νιν μαλακοῖς ἐνὶ φάρεσιν οἷς ἐνίαυεν | ὡς μετὰ τεῦς ἀνὰ νύκτα τὸν ἱερὸν ὕπνον ἐμόχθει and Lydia 169–71 illidere in herba | purpureos flores quos insuper accumbebat, | candida formoso supponens bracchia collo is purely coincidental: in both cases the main clause refers to lying on a ‘bed’, the relative clause adds information about the ‘bedding’, and both the ὡς clause and the participial construction specify the circumstances, namely Venus and Adonis having sex.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

24 INTRODUCTION might lie behind the Lydia speaker’s statement that his body is losing the heat of life (126 et calor infuso decedit frigore leti).55 Besides these thematic and lexical points of contact, there are also significant formal similarities between the refrains that separate stanzas in the two poems. One is that the refrains are variable.56 The Lydia only contains three such refrains, which mark off the first two stanzas and then disappear, but all three involve variation: 104 inuideo uobis, agri formosaque prata, 111 inuideo uobis, agri: discetis amare, 123 inuideo uobis, agri: mea gaudia habetis. In the Lament, the refrains continue throughout the poem, and sometimes the same refrain is reused, but in general they display similar variability; note e.g. the first four: 1 αἰάζω τὸν Ἄδωνιν, “ἀπώλετο καλὸς Ἄδωνις”, 6 (= 15) αἰάζω τὸν Ἄδωνιν· ἐπαιάζουσιν Ἔρωτες, 28 “αἰαῖ τὰν Κυθέρειαν”, ἐπαιάζουσιν Ἔρωτες. Another important point of resemblance is that the refrains do not as a rule interrupt the flow of the text, but are thematically, and sometimes syntactically, intertwined with it. This is indeed the case in all of the Lydia examples; from the Lament, compare e.g. 62–3 ὧδ᾽ ὀλοϕύρατο Κύπρις· ἐπαιάζουσιν Ἔρωτες, | “αἰαῖ τὰν Κυθέρειαν, ἀπώλετο καλὸς Ἄδωνις”. The third simi­ larity is that in both poems the refrain (at least, in its most standard form) begins with a verb in the same grammatical form, first-­person singular present indicative active: αἰάζω or inuideo. All these features bind the two poems together in opposition to other bucolic texts with refrains, which are far less variable, far less context-­specific (in the nar­ rower sense), and usually employ imperatives (note e.g. the refrain in Thyrsis’ song: first ἄρχετε βουκολικᾶς, Μοῖσαι ϕίλαι, ἄρχετ᾽ ἀοιδᾶς, then λήγετε βουκολικᾶς, Μοῖσαι, ἴτε λήγετ᾽ ἀοιδᾶς). It seems safe to conclude that in this respect the Lydia took Bion’s Lament for Adonis as a model. The third text in the sequence—­which seals it as a tradition—­is the anonymous Lament for Bion: fittingly it is a poem obsessed with issues of literary succession, while also announcing the end of bucolic poetry (12 ὤλετο Δωρὶς ἀοιδά, which is inherently self-­contradictory), though 55  The two interpretations of ἀποψύχειν, ‘to expire’ and ‘to cool down’, need not in fact be as distinct as may appear from a modern perspective, since the ideas of ‘the breath of life’ and of ‘the heat of life’ were interrelated. 56  On the refrains in the Lament, see Burris 2004, 168–9, noting their uniqueness in Greek bucolic; cf. further Estevez 1981.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  25 this does not prevent the poet from claiming to be Bion’s heir (96 κλαρονόμος μοίσας τᾶς Δωρίδος).57 It is mainly within such a poetologi­ cal framework that points of contact between this second Lament and the Lydia will be discussed, with a focus on the role of female characters. First, however, it is worth briefly pointing out that Bion appears to have succumbed to the same disease as Daphnis and the Lydia speaker: as has convincingly been argued by Manakidou, the ‘poison’ from which Bion is claimed to have died (109 ϕάρμακον ἦλθε, Βίων, ποτὶ σὸν στόμα, ϕάρμακον ἦδες) should be taken as a metaphor for love.58 Above it was suggested that the figure of Lydia is shaped in important ways after that of the Aphrodite of the Lament for Adonis; a similar rela­ tionship can be established between the figures of Lydia and the Galatea of the Lament for Bion. In a fragment of Bion, Galatea makes an appear­ ance as a girl impervious to the speaker’s (Polyphemus’?) pleas (fr. 16), but in the Lament she is presented as an avid listener of Bion’s own sing­ ing (58–63):59 κλαίει καὶ Γαλάτεια τὸ σὸν μέλος, ἅν ποκ᾽ ἔτερπες ἑζομέναν μετὰ σεῖο παρ᾽ ἀιόνεσσι θαλάσσας· οὐ γὰρ ἴσον Κύκλωπι μελίσδεο. τὸν μὲν ἔϕευγεν ἁ καλὰ Γαλάτεια, σὲ δ᾽ ἅδιον ἔβλεπεν ἅλμας, καὶ νῦν λασαμένα τῶ κύματος ἐν ψαμάθοισιν ἕζετ᾽ ἐρημαίαισι, βόας δ᾽ ἔτι σεῖο νομεύει. Galatea too bewails your singing, whom you once entertained as she sat with you at the sea shore, for you didn’t sing like the Cyclops. From him fair Galatea ran, but you she saw as sweeter than sea-­water, and now, forgetting the waves, she sits on a deserted beach and still herds your oxen.

In Lydia 107–10, we find a picture of the eponymous character that is both similar and different. On the one hand, both Galatea and Lydia continue to do the same things they used to do in the company of the male protagonists, only now they do them alone (note 62 νῦν, and cf. 57  Cf. in general Di Nino 2009, and note Spelman 2018. 59  For a general discussion of the passage, see Di Nino 2018.

58  Manakidou 1996, 52–5.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

26 INTRODUCTION repeated nunc in 107–8). On the other, in contrast to Galatea who is presented as Bion’s listener (58 ἅν ποκ᾽ ἔτερπες), Lydia is said to have sung to the speaker herself (110 mihi quae cantabat in aurem), and she continues to sing even after their parting (109–10). There are, however, indications that we should imagine Galatea likewise to be singing—­after Bion’s death, if not before. First, her posture—­sitting in solitude (62–3 ἐν ψαμάθοισιν | ἕζετ᾽ ἐρημαίαισι)—pointedly imitates that of Bion when singing (21 οὐκέτ᾽ ἐρημαίαισιν ὑπὸ δρυσὶν ἥμενος ᾄδει). Second, she now herds Bion’s cattle (βόας δ᾽ ἔτι σεῖο νομεύει), an activity that Bion himself combined with singing (81 καὶ βούτας ἐλίγαινε καὶ ἀείδων ἐνόμευε). Third, the reference to her being entertained by Bion (58 ἅν ποκ᾽ ἔτερπες) echoes an earlier statement that Bion entertained birds and taught them to sing (46–7 ἅς ποκ᾽ ἔτερπεν, | ἃς λαλέειν ἐδίδασκε): one might infer that Galatea could have learned singing from him as well. Finally, the conceit by which a character (Galatea) is presented as inter­ acting with her creator (Bion) is no doubt intended to imply a biographical reading of Bion’s poem in which Galatea appeared. A precedent would have been provided by Phaenias’ allegorical interpretation of Philoxenus’ Cyclops according to which Odysseus was a mask for the author, Polyphemus for Dionysius of Syracuse, and Galatea for the latter’s mis­ tress, seduced by Philoxenus: it may be relevant that Phaenias identified Galatea as an αὐλητρίς.60 Whether we decide that the Lament implicitly presents Galatea as a bucolic singer, or by contrast deliberately sup­ presses this idea, it seems likely that the reader is invited at least to con­ template the possibility. The image of Bion teaching birds to sing (47) appears to be reflected in that of birds listening in silence to Lydia’s singing (120 auiumque silentia fient). While at first glance ‘teaching birds’ may appear to be little more than a hyperbolic figure of speech, the motif of instruction is in fact of central importance for the poetological concerns of the Lament. Besides birds, Bion turns out to have had human pupils whom he instructed in matters of both poetry (95 διδάξαο σεῖο μαθητάς [sc. bucolic poetry]) 60  On Philoxenus’ Cyclops, see esp. Hordern 1999. Note that this biographical reading also underlies Hermesianax fr. 3.73–4 μέγαν πόθον ὃν Γαλατείη | αὐτοῖς μηλείοις θήκαθ᾽ ὑπὲρ [scripsi: ὑπὸ codd.] προγόνοις (Galatea valued Philoxenus’ ardent passion above lambs, offered to her by Polyphemus, one assumes).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  27 and love (83 παίδων ἐδίδασκε ϕιλήματα). Both types of instruction are of course intrinsically linked, since it is through his poems that Bion must be understood to have taught erotics—­as is suggested by an actual fragment of Bion in which the speaker appears as a pupil of Eros, after failing as his teacher (fr. 10.10–13):61 ἀλλά μοι αὐτὸς ἄειδεν ἐρωτύλα, καί με δίδασκε θνατῶν ἀθανάτων τε πόθως καὶ ματέρος ἔργα. κἠγὼν ἐκλαθόμαν μὲν ὅσων τὸν Ἔρωτα δίδασκον, ὅσσα δ᾽ Ἔρως με δίδαξεν ἐρωτύλα πάντα διδάχθην. But he himself sang erotic songs to me and taught me about the love affairs of men and gods and about his mother’s deeds. And I forgot all I taught Eros, but learnt all the erotic songs Eros taught me.

We find the same idea of teaching love through poetry in the Lydia pas­ sage in which the farmlands are said to learn to love by listening to Lydia’s singing (111 inuideo uobis, agri: discetis amare).62 This Lydia line may contain another reflex of the image of Bion teaching birds to sing (in Bion’s actual fragments, nature does not figure as a recipient of instruction, either erotic or poetic), but it may also echo a further pas­ sage from the Lament, in which Bion’s singing is claimed to have been replaced with frogs’ (105–7): καὶ σὺ μὲν ὦν σιγᾷ πεπυκασμένος ἔσσεαι ἐν γᾷ, ταῖς Νύμϕαισι δ᾽ ἔδοξεν ἀεὶ τὸν βάτραχον ᾄδειν. ταῖς δ᾽ ἐγὼ οὐ ϕθονέοιμι, τὸ γὰρ μέλος οὐ καλὸν ᾄδει. And so you will lie in the earth, wrapped in silence, but the frog the Nymphs decreed to sing forever: I welcome them, for he sings an ugly song.

In the last quoted line, the anonymous poet of the Lament states that he does not envy the Nymphs who have to listen to frogs’ ugly singing: this 61  Note that the verb διδάσκειν appears four times in this short passage, with three more occurrences earlier in the fragment. 62  At the same time, Lydia can be seen to be teaching them to love by ‘flirting’ with them: see 107–8n.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

28 INTRODUCTION sentiment is nearly the exact opposite of the Lydia speaker’s words that he envies the farmlands because they learn to love by listening to Lydia’s singing. This underlines the central difference between the poetological outlooks of the two poems: whereas the Lament proclaims the death of poetry (how sincerely or consistently is another matter), the Lydia endorses continuity—­even if the poem’s protagonist is dying, his muse continues to sing. At the same time, and no less prominently, the motif of erotodidaxis is also manifested in the Lydia speaker’s unrealized (and unrealizable) aspiration to be the πρῶτος εὑρετής of ‘clandestine love’ (159–60 istius atque utinam facti mea culpa magistra | prima foret): he may have failed to be the first, but teach love he does, through both his own ex­ample (at the very least, he appears to have initiated Lydia, 156–7) and his words (most clearly, by citing numerous mythological stories with erotic content, cf. Bion fr. 10.10–11 quoted above). This admission of failure has no doubt a metapoetic dimension as well: the Lydia is no more the first-­ever erotic poem than its speaker is the first-­ever lover, and by acknowledging that, it asserts its place in the poetic tradition. More specifically, in his aspiration to be the inventor of love, the Lydia speaker may be seen to emulate Orpheus, who was accredited with the invention of pederasty (see 156–65n.)—and who may thus be accorded the role of archetypal love poet. This brings us to one last feature of the Lament for Bion that needs to be touched upon: the prominent position given in it to the figure of Orpheus (note esp. that Bion is proclaimed as 18 Δώριος Ὀρϕεύς). As has been explored by Acél and especially Kania, the Lament reshapes the bucolic tradition so as to derive it from Orpheus, an innovation subse­ quently embraced by Virgil (and Ovid).63 More recently, it has also been suggested that in particular the portrayal of Bion as a teacher of ­pederasty (83) may be intended as an ‘Orphic’ feature.64 For our present concerns, it is further relevant to note that the Lament conflates two simi­ lar but distinct motifs: the motif of nature mourning for a dead hero (pathetic fallacy), near-­eastern in origin and introduced in Greek

63  Acél 2007, Kania 2012; it is tempting to speculate that Orpheus already appeared in Bion’s poetry, see Knaack 1905, Reed 1997, 27, 152. 64  See Sundt 2021, 40–9.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  29 bucolic by Theocritus, and the motif of nature being enchanted by song, traditionally associated with Orpheus. While from the perspective of Latin poetry the association of pathetic fallacy with Orpheus’ death may appear to be an established topos, in Greek poetry it seems unattested other than in the Lament (where it might be derived from a lost poem by Bion) and in an anonymous epigram.65 In the preceding discussion, I described nature’s response to Lydia’s singing as a variation on pathetic fallacy; now we can see that the Lydia too conflates it with the Orphic topos, albeit in a different way. If the Lament tradition inserts pathetic fallacy into an Orphic context, the Lydia does the reverse: it employs the motif of nature being enchanted by song in a context—­the dying of the protagonist—­in which pathetic fallacy could rather be expected. The above account does not exhaust all the ways in which the Lydia is indebted to Greek bucolic,66 but I hope to have demonstrated that it extensively engages with a specific strain of the tradition: three pro­ grammatic poems (Theocritus 1, Bion’s Lament for Adonis, the an­onym­ ous Lament for Bion) that focus on a bucolic hero dying of love. The argument has been, on the one hand, that the figure of the Lydia speaker should be seen as ‘modelled’ on these earlier instantiations of a dying herdsman; on the other, that our poem departs in significant ways from the Greek treatments of this ‘typical’ situation. Perhaps the most remark­ able difference is the representation of the female protagonist, Lydia, who, unlike her Greek predecessors, is portrayed as a bucolic singer in her own right—­and not just any singer but a master poet whose song possesses the power of instruction (111) and enchantment (119–22). An important corollary of the preceding argument is that the Lydia can thus be assigned an organic—­if still not narrowly defined—­place in the bucolic tradition. While most current research views the poem as derived from Virgil’s Eclogues (with cross-­pollination from Latin elegy), 65  AP 7.10.6–8 ἔρρηξαν Μοῦσαι δάκρυα Πιερίδες | μυρόμεναι τὸν ἀοιδόν [sc. Orpheus]· ἐπωδύραντο δὲ πέτραι | καὶ δρύες, ἃς ἐρατῇ τὸ πρὶν ἔθελγε λύρῃ. Page 1981, 335 suggests that the epigram is Hellenistic and comes from Meleager’s collection; it is, however, remarkable that the similar epitaph by Antipater (AP 7.8) which the present one reworks only has the Muses mourning for Orpheus: the pathetic fallacy is an innovation, possibly an ad hoc invention, here. 66  The most important being the Lydia’s points of contact with comastic monologues featur­ ing mythological exempla of love affairs between gods and mortals, such as Theocr. 3 and 20, as briefly mentioned in section 1.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

30 INTRODUCTION and thus places it in the post-­Virgilian period, the realization that the Lydia can be seen as a development of an earlier Greek tradition reopens the question of relative chronology. Can we really rule out that the Lydia was written prior to, or contemporaneously with, the Eclogues? It seems far from certain that a decided ‘yes’ can be the answer, in which case it will be more fruitful to explore different possibilities. Among Virgil’s poems, the tenth eclogue features the clearest points of contact with the Lydia, on both the thematic and the verbal level. To begin with, it engages with the same set of Greek intertexts: the figure of Gallus is directly modelled on that of Daphnis in Theocritus’ first idyll, but the two laments too have exercised palpable influence.67 At the same time, Gallus’ monologue contains unmistakable textual resonances with the Lydia (10.42–5): hic gelidi fontes, hic mollia prata, Lycori, hic nemus; hic ipso tecum consumerer aeuo. nunc insanus amor duri me Martis in armis tela inter media atque aduersos detinet hostis. Here, Lycoris, are icy springs, here soft meadows, here woods; here with you I would gladly let time itself destroy me. But now a mad passion for cruel Mars keeps me under arms, amid a swarm of spears and charging foes.

The two noun phrases at 10.42 find exact matches in the passage referring to natural features enchanted by Lydia’s singing (119 and 120), whereas the reference to Gallus’ military service at 10.44–5 cannot but evoke the explanation for Mars’s absence given by the Lydia speaker: 172 tum credo fuerat Mauors distentus in armis.68 Deciding which text came first is not easy, though a possible indicator of our poem’s priority might be 67  On the latter, see Gagliardi 2014. 68  If one does read Heumann’s te for me at 10.44, perhaps Martis in armis should be taken to mean ‘in Mars’s embrace’: unlike Venus in the Lydia, Lycoris will have followed her Mars to the front lines. There are further points of contact, though less striking than the ones outlined above: one is the use of cura to refer to the beloved (22 tua cura Lycoris, cf. Lydia 122 dum mea iucundas expromet cura querelas), another is the sentiment expressed by the male protagonists that they would be happy to die if they could become famous (33–4, cf. Lydia 159–60), accom­ panied by a counterfactual wish introduced with atque utinam (35, Lydia 159).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  31 the fact that in Virgil’s eclogue the words and phrases resonating with the Lydia are accumulated in one short passage.69 Whether we date it before or after (or indeed ‘with’) the Eclogues, the intertextual link with the tenth eclogue provides an additional argument for seeing in the Lydia a continuation of the tradition initiated by Thyrsis’ song in Theocritus’ first idyll. Another literary connection suggested by the link with the tenth eclogue is one with Gallus’ poetry. Observing that in the Epistula Sapphus Ovid closely engages with the tenth eclogue, and in particular models the figure of lovesick Sappho on that of Gallus, Thorsen has speculated that the real Gallus may have alluded to Sappho’s poetry in his elegies.70 More recently, Rosati has further argued that, by alluding to both Virgil’s eclogue and Ovid’s epistle, the Lydia confirms Gallus’ use of Sappho’s poetry.71 While Rosati’s argument seems weak (even if one accepts a post-­Augustan dating, the combined allusion to Virgil’s eclogue and Ovid’s epistle could sufficiently be explained by the Lydia poet’s recogniz­ ing their intertextual connection), the idea itself may have something to it. In a remarkable passage, the Lydia speaker proclaims his beloved to be nonpareil, in terms that have often been taken to betray the influence of love elegy (127–8): non ulla puella | doctior in terris fuit aut formosior illa. There is, however, another potential model, Sappho’s fragment 56:72 οὐδ᾽ ἴαν δοκίμωμι προσίδοισαν ϕάος ἀλίω ἔσσεσθαι σοϕίαν πάρθενον εἰς οὐδένα πω χρόνον τεαύταν. No girl who has seen the light of the sun will I believe equal you in skill any time in the future. 69  Courtney 1993, 191 believes that ‘Theocr. 5.33–4 > Buc. 10.42–3 > Lydia 119–20 is a probable sequence of imitation’, but there is no binding reason why Virgil could not lift the phrases gelidi fontes and mollia prata from the Lydia to render Theocritus’ ψυχρὸν ὕδωρ and ποία (similarly there is a good chance that 10.22 cura Lycoris is a Gallan quotation, despite the fact that Virgil uses it to ‘render’ Theocr. 1.82 ἁ δέ τυ κώρα, see Ross 1975, 68–9, cf. Cairns 2006, 115–16); and in any event the case could be reversed by arguing that fontes goes back to Ep. Adon. 34 παγαί, which is an intertext more relevant for the Lydia than for the Eclogues. 70  See Thorsen 2014, 86–91. 71  Rosati 2020. 72  As suggested by Cavallini 1986, 28–30; note that it also appears to lie behind Catull. 61.84–6 nequa femina pulchrior | clarum ab Oceano diem | uiderit uenientem.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

32 INTRODUCTION The Sapphic associations of the conceit in Latin, already explicit in Catullus 35.16–17 Sapphica puella | musa doctior, can further be sup­ ported with two contexts in Martial that likewise name the Greek poetess: 7.69.9–10 carmina fingentem [sc. Theophilam] Sappho laudabat amatrix: | castior haec, et non doctior illa fuit and 10.35.15–16 hac [sc. Sulpicia] con­ discipula uel hac magistra | esses doctior et pudica, Sappho.73 Is Martial echoing specifically the Lydia passage? The situation is complicated by the existence of two further intertexts, an (imperial-­date) inscribed epitaph for an eight-­year-­old slave-­girl, CLE 1166.6 doctior in terris nulla puella foret, and a third context from Martial, 10.32.6 pulchrior in terris nulla tabella foret. The closeness of the epitaph to the Lydia’s wording might indicate that the latter is indeed the immediate model, but the third Martial parallel rather suggests that the common source was a pentameter. The easiest way to account for this constellation of parallels seems to be by positing an influential elegiac text that alluded to Sappho with the pen­ tameter doctior in terris nulla puella foret (fuit): the epitaph will have reused it unchanged, whereas Martial could be seen to reshape it in differ­ ent ways. Who wrote this hypothetic influential elegy, and what is its rela­ tion to the Lydia? It is tempting to speculate that it was Gallus, though even in this case it cannot be excluded that the Lydia was the model. To recapitulate, the thrust of the argument has been so far that the Lydia develops a specific strain of the Greek bucolic tradition, and does so in an entirely original direction by making the female protagonist into the central poet figure. Though insufficient to prove its chrono­logic­al priority over Virgil’s Eclogues, this shows that there is no binding reason to assume a priori that the Lydia is post-­Virgilian either. The examination of intertextually linked passages in the Lydia and the tenth eclogue has proved likewise inconclusive. At the same time, there are stylistic indications aligning the Lydia with pre-­Virgilian poetry, and the ascrip­ tion to Valerius Cato is in some ways not unattractive (see section 1).74 In other words, while the positive evidence placing the Lydia before the 73  See further Luque Moreno 2015. 74  In view of the Gallan connection suggested above, it may be relevant that the Gallus addressed in fr. 85.1 by Furius Bibaculus in connection with Valerius Cato may be the elegist (see Hollis 2007, 140), and that in fr. 145.9 Gallus may himself be referring to Valerius Cato (see Hollis 2007, 248–50).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  33 Eclogues and associating it with Cato does not amount to incontrovertible proof, there is likewise no solid ground for rejecting Catonian ­authorship and insisting on a later composition date.75 Whether or not we can securely attach Cato’s name to the Lydia is ultimately of limited importance, since we only know him by name anyway, but it may prove fruitful to think of the Lydia as (if) written in the 40s bc by a figure like Cato. In what follows I offer, assuming such an early dating, a brief sketch of some possible reactions to the Lydia in subsequent poetry, with no claim to exhaustiveness.76 Horace’s second epode may be one of the earliest distinct echoes of the Lydia. The epode is a praise of country life, and as such naturally has many thematic intersection with our poem, but a number of closer ­text­ual points of contact suggest direct engagement. The μακαρισμός with which the epode opens (2.1–2 beatus ille qui procul negotiis, | ut prisca gens mortalium . . . ) has three formal parallels in the Lydia, where the device, however, has different referents (112, 131, 134); the passage also shares with the Lydia a nostalgic attitude towards the mythic past, to which both refer with the same epic poeticism, mortales (cf. 151 condicio similis fuerat mortalibus illis, with n.). Another lexical link is supplied by Horace’s reference to the plucking of ripe grapes (2.19–20 ut gaudet insitiua decerpens pira | certantem et uuam purpurae), which contrasts with Lydia’s plucking of unripe ones (114 et roseis uiridem digi­ tis decerpserit uuam): note that both use the same almost technical (or at least prosaic) verb decerpere, and that both indicate the grapes’ degree of maturity with colour terms. A rather richer intertext is the description of a locus amoenus in the epode (2.23–8): libet iacere modo sub antiqua ilice,   modo in tenaci gramine: labuntur altis interim riuis aquae,   queruntur in siluis aues fontesque lymphis obstrepunt manantibus,   somnos quod inuitet leues. 75  Though one need not entail the other, as Cato may have been alive into the 20s bc (Hollis 2007, 429). 76  The commentary notes throughout additional parallels with Augustan and later poetry.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

34 INTRODUCTION It’s nice to lie now under an old ilex tree, now in welcoming grass: meanwhile waters run in deep streams, birds chirp in woods, and springs tinkle with flowing water, which induces light sleep.

It has numerous points of contact with the Lydia speaker’s fantasy about his beloved lying in the grass (117) and enchanting nature with her sing­ ing (119–22): gaudebunt siluae, gaudebunt mollia prata, tardabunt riui labentes murmure lymphas     et gelidi fontes, auiumque silentia fient,     dum mea iucundas expromet cura querelas.

121 120

Both passages refer to woods (2.26 siluis, 119 siluae), to rivers (2.25 riuis, 121 riui) and springs (2.27 fontesque, 120 fontes), and to birds (2.26 aues, 120 auiumque).77 There is also a pointed contrast: the water and the birds in Horace are noisy (2.26 queruntur, 27 obstrepunt), inviting sleep; in the Lydia, they fall silent (120 silentia fient), so as to listen to the heroine’s singing.78 References to water contain another significant com­ monality: as already noted, 121 labentes . . . lymphas is a figura etymolo­ gica alluding to Varro’s derivation of the noun ab aquae lapsu lubrico, which allowed the Lydia poet to extend its usage to apply to running water; although Horace splits the figura etymologica (2.25 labuntur, 2.27 lymphis), he uses lympha in the innovative way pioneered by the Lydia—­ and will do so more than once again (see 121n., and cf. below). Finally, the unexpected conclusion revealing the epode to be a fantasy of a ­moneylender (2.67–70) cannot, in the socio-­historical context of the 30s  bc, fail to evoke the land confiscations of 41 bc, thus possibly ­alluding to the situation of the Dirae.79 Tibullus’ elegy 2.3 is a rather tantalizing intertext.80 On the one hand, the situation that Tibullus dramatizes is generally close to that of the 77  For reading riuis rather than ripis at 2.25, and fontesque rather than frondesque at 2.27, see Watson 2003, 101–2. 78  This point of contact adds further support to my conjecture murmure at 120. 79  Note that the fourth epode appears to engage with the Dirae more systematically, as will be discussed in greater detail in my forthcoming commentary on that poem. 80  For a different take, with a focus on the Dirae, cf. Breed 2012, 18–21.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  35 Lydia: he is parted from Nemesis, who, like Lydia, is confined to the countryside. Despite a marked difference in attitude (a starker contrast than between Tibullus’ urbane cynicism and the Lydia speaker’s rustic naivety is difficult to imagine), the two poems share the notion of the remote past as a time of erotic bliss, which they substantiate by alluding to the metallic myth (2.3.35 ferrea non Venerem, sed praedam saecula laudant, contrast 150 aurea quin etiam cum saecula uoluebantur, 177 aurea proles) and to examples of affairs between gods and mortals (Apollo and Admetus in Tibullus, a whole series in the Lydia). Yet close textual points of contact are hard to come by; one possible exception is the link between 2.3.31–2 fabula nunc ille [sc. Apollo] est, sed cui sua cura puella est, | fabula sit mauult quam sine amore deus and 129–30 fabula ni uana est, tauro Ioue digna uel auro | (Iuppiter, auertas aurem) mea sola puella est. Tibullus’ hexameter may be seen as a ‘conflation’ (with variation) of 129 fabula ni uana est and 130 mea sola puella est (note the two instances of prodelided est, both preserved by Tibullus in the same metrical positions), and a general commonality is the hypo­ thetical juxtaposition of a modern human with a god of the myth. Tibullus can also be seen to be providing a playful linguistic commen­ tary on the Lydia: fabula in the neutral sense of fama appears to be an archaism and/or vulgarism, and while in the hexameter Tibullus takes over this usage from the Lydia, in the pentameter he relapses into the urbane idiom of an elegiac poet and employs the term in the modern informal (but not inelegant) sense of ‘gossip, slander’.81 A more explicit case of linguistic commentary may be found in an allusion to the Dirae at the beginning of the elegy (2.3.3–4): ipsa Venus latos iam nunc migrauit in agros,   uerbaque aratoris rustica discit Amor. Venus herself has now relocated to open fields, and Amor is learning the ploughman’s uncultivated speech.

81  Note that Virgil avoids it altogether; in all other contexts in Catullus (69.5), Propertius (2.13.14, 24.1, 32.26, 3.15.45), and Tibullus (1.4.83), fabula has a more or less distinct pejora­ tive connotation (except at Prop. 3.5.45, where it is mock-­epic). Cf. further 129n.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

36 INTRODUCTION The verb migrare is largely avoided in higher-­register poetry and appears to be a vulgarism, as is implicitly acknowledged in the pentameter.82 Here it serves to evoke the curse of Dirae 50 migret Neptunus in arua, which likewise speaks metaphorically of a deity ‘migrating’ to the coun­ tryside. This is, moreover, not the only point of contact with the Lydia’s sibling; as Breed points out, ‘Tibullus 2.3 invites direct comparison with the Dirae with its own “dirae”, a section of curses on the land to punish it for taking away the poet’s beloved (61–6)’.83 While on the surface Tibullus’ curse against his rival’s estate is motivated by erotic jealousy, it also has a socio-­historical dimension, not unlike Horace’s second epode: Tibullus’ rival is a nouveau riche of servile background (2.3.59–60), who may have acquired his estate in the aftermath of the upheaval caused by the land redistribution of 41 bc.84 A third intertext that appears to combine allusions to the Lydia and the Dirae is Ovid’s elegy 3.7, in which he narrates an erotic encounter ruined by his sudden impotence. The suspicion that he may have been hexed leads to a digression on the power of magic (3.7.31–4): carmine laesa Ceres sterilem uanescit in herbam,   deficiunt laesi carmine fontis aquae, ilicibus glandes cantataque uitibus uua   decidit, et nullo poma mouente fluunt. Damaged with song, wheat fades away into barren grass; damaged with song, spring water disappears; bewitched with spells, acorns fall down from the ilex trees and grapes from the vines; and fruits drop shaken down by no-­one.

This catalogue, painstakingly listing different parts of a country estate affected by an incantation, cannot but evoke a similarly detailed curse in the Dirae (15–18):85 82  Cf. Hor. Ars 229 migret in obscuras humili sermone tabernas, where humili sermone simi­ larly comments on migret. 83  Breed 2012, 20, with further pertinent observations. 84  As Horace’s moneylender contemplates doing; as an ex-­slave, Tibullus’ rival is unlikely to have been a legion veteran. 85  This stanza is textually uncertain in many places; I print here the text of my forthcoming edition.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Poetic Frameworks  37 effetas Cereris sulcus producat auenas, pallida flaccescant aestu sitientia prata, immatura cadant ramis pendentia poma, deficiant siluis frondes et fontibus umor. May the wheat furrow raise barren oats, may the meadows wilt and pale as they thirst with drought, may fruits fall unripe from branches on which they grow, may leaves disappear in the woods and water in the springs.

Ovid’s wheat crops turning into barren oats, water disappearing from the springs, and fruits falling down prematurely all find counterparts in the Dirae; oaks losing their acorns can be compared with trees losing their leaves; and grapes can be paralleled from the preceding stanza (12). Finally, sterilem uanescit (in herbam) may serve as a periphrasis for Dirae 9 sterilescant, intended to acknowledge the verb’s subliterary character (it is otherwise unparalleled in poetry). Echoes of the Lydia are less distinct, but Corinna’s active role and position ‘on top’ (3.7.7–8 illa quidem nostro subiecit eburnea collo | bracchia, Sithonia candidiora niue) evokes that of Venus in the Lydia’s account of her affair with Adonis (171 candida formoso supponens bracchia collo: see n.); and the striking comparison of Ovid’s flaccid penis to a wilted rose (3.7.65–6) may allude to the erotically charged image of Venus crushing flowers (170: see n.). My last case of the Lydia’s reception in Augustan poetry comes like­ wise from Ovid, from a passage in the Remedia in which he recom­ mends retiring to the countryside as a strategy against lovesickness (177–84): aspice labentes iucundo murmure riuos;   aspice tondentes fertile gramen oues. ecce, petunt rupes praeruptaque saxa capellae:   iam referent haedis ubera plena suis;     pastor inaequali modulatur harundine carmen,   nec desunt comites, sedula turba, canes; parte sonant alia siluae mugitibus altae,   et queritur uitulum mater abesse suum.

180

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

38 INTRODUCTION Look how the streams flow with pleasant rumble, look how the sheep graze bountiful grass. See, the goats climb cliffs and steep rocks: soon they will bring full udders back to their kids. The shepherd plays a song on his uneven pipes, his trusty dogs—­a restless throng—­are there too. Elsewhere the high woods resound with lowing, and the mother moans for her absent calf.

The expressed goal is that the rural landscape should draw the sufferer’s attention away from his amatory woes (cf. Rem. 169–70), yet on the intertextual level the passage does the exact opposite. A programmatic allusion is effected by 181, which evokes Virgil’s account of Gallus’ seek­ ing in vain respite from love in the countryside (Ecl. 10.51 carmina pas­ toris Siculi modulabor auena). The apparent intention of the Ovidian passage is similarly subverted by allusions to the Lydia. Ovid’s reference to goats’ grazing implies that they do so in order to produce milk for their kids (Rem. 179–80); the Lydia line which the Ovidian hexameter only slightly varies (135 petis montes praeruptaque saxa pererras) belongs in fact to a passage describing the marital bliss enjoyed by goats in contrast to humans (134–7). The couplet referring to a cow mooing in grief for her lost calf (Rem. 183–4) is likewise misleading: although the verbal correspondences are less exact, the Lydia parallel suggests that cows’ mooing may rather have a sexual motivation (133 frustra te pati­ tur siluis mugire dolorem).86 Finally, even the line mentioning rivers (Rem. 177) may make one think, not forget, about love, by evoking the Lydia’s reference to rivers stopping their flow (121 tardabunt riui labentes murmure lymphas) so as to listen to the heroine’s amatory outpourings (122 iucundas . . . querelas).87

86  The literal sense of the couplet alludes to Lucretius’ famous account of a cow having lost her calf (2.352–66), but the term mugitus (not used by Lucretius) rather evokes Virgil’s allu­ sions to the erotically charged myths of Pasiphae and Io (Ecl. 6.48, Georg. 3.150; one suspects that this use of mugitus may go back to Calvus’ Io). 87  A further important Ovidian intertext is the Epistula Sapphus, whose connections with the Lydia have recently been discussed by Rosati 2020 (albeit under the assumption of Ovid’s priority).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

A Note on the Language and Metre  39

3.  A Note on the Language and Metre The most substantial study of the language and metre of the Dirae and the Lydia remains van der Graaf ’s, which takes up just over a half of his monograph.88 A quick comparison of his dossier with the text printed here, however, reveals the fundamental obstacle to any quantitative interpretation of such data taken from a text as uncertain as the Lydia’s (besides all the usual difficulties). For instance, van der Graaf identifies the following colloquialisms in the Lydia: 137 capella, 108 ocelli, 115 uitecula, 132 uaccula, 112 multum beatus, 141 and 175 plorare, 170 insu­ per, 177 numquid;89 out of the nine contexts, I believe four to be corrupt (112, 115, 141, and 175), and a fifth possibly corrupt (132). This does not mean that studies such as van der Graaf ’s are useless, but it does suggest that any systematic description of the linguistic and metrical profile of the Lydia is bound to involve numerous uncertainties. Rather than offering such a systematic account, I limit myself to making two qualitative observations, which I hope together can give some idea of the kind of poetic texture that the Lydia poet aimed to create. One point was already noted in the discussion of the Lydia’s date of composition, when we observed its three distinctive σπονδειάζοντες: 136 uestigare, 150 uoluebantur, 170 accumbebat. As was suggested in section 1, this kind of spondaic verse produced by a verbal form links the Lydia to Lucretius and Catullus 64 (and Varro’s Argonautae), whereas in later poetry (possibly beginning with the Ciris) this specific kind is avoided. Besides being a potential indicator of the composition era, this feature no doubt also has a stylistic aspect, though its exact nuance will prob­ ably depend on the dating. The connection with Catullus (and Varro) suggests ‘neoteric’ mannerism, whereas that with Lucretius may rather imply linguistic austerity. Which is the intended effect? There is a not­ able difference between Catullus’ and Lucretius’ practice: whereas in the former a significant proportion of σπονδειάζοντες are produced by exotic-­ sounding quadrisyllabic Greek words (64.3 Aeetaeos, 11 Amphitriten, 28 Nereine, 36 Larisaea, 74 Piraei, 79 Minotauro, 358 Hellesponto; cf. in the 88  Van der Graaf 1945, 44–122. On the metre, see also Rodríguez Pantoja 2006. 89  Van der Graaf 1945, 116–18.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

40 INTRODUCTION elegiacs 68.87 Argiuorum, 89 Europaeque, 109 Cyllenaeum, 100.1 Aufillenam: two are Latin, but their effect is similar), the latter only has one such verse (4.125 centaurea).90 This brings the Lydia’s practice closer to that of Lucretius, and it seems justified to conclude that its intended effect is studied simplicity, appropriate for a bucolic singer, rather than the hellenizing virtuosity of a poeta doctus (or perhaps the latter under the pretence of the former). My second observation concerns words and phrases that belong to a ‘substandard’ register (from the viewpoint of poetry). Van der Graaf interpreted these as colloquialisms, in the sense of Plautine conversa­ tional language.91 While there are indeed elements that may evoke the expressive informality of unrefined speech (note esp. the discourse marker credo at 172, cf. further 110 in aurem, 148 secum . . . gestat, perhaps 135 pererras with nn.), most lexical departures from the neutrally poetical register seem rather to betray a prosaically mundane and/or indelicate character (note 106 tacite, 114 decerpserit, 117 reclinarit, 122 iucundas, 125 tabescunt, 129 fabula, 136 uestigare, 139 interpellatos, 162 tribuerunt, 174 turpabant and fuligine, 181 libido, which are more often paralleled in Lucretius and/or prose than in comedy).92 This alignment is even more evident in a number of composite expressions and constructions, which for the most part consist of words individually admissible in poetry, but which as idioms have a prosaic and/or tech­nical colouring (105 hoc . . . magis . . . quod, 120 silentia fient, 141 patior . . . dolorem, 151 condicio similis, 155 condicio . . . uitae . . . durior, 173 opus faciebat, 177 numquid minus, 180 sors . . . nascendi). The infusion of such mundane elements seems to be a deliberate strategy intended to create a distinct bucolic ‘dialect’ (a task notoriously more difficult in Latin than in

90  One assumes that Varro’s practice was closer to Catullus’, but with only twelve hexameters surviving of the Argonautae, there is no way to be certain. 91  Van der Graaf 1945, 115–16: ‘It is hardly necessary to say that the sermo familiaris asserts itself especially in a spontaneous and subjective poem like the Dirae. It is certainly curious that several characteristics of the popular language are found for the first time in Plautus, next in the Dirae and only after that in post-­classic Latin.’ Cf. in a way Lipka’s (2001, 144) assessment of colloquialisms in Virgil’s Eclogues. 92  My conjectures 115 uindemia, 116 spectacula, 139 tolerauit, 174 massae may be added to the dossier.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Manuscript Tradition and the Text  41 Greek), in a way rather different to Virgil’s in the Eclogues.93 It is more difficult, however, to pinpoint exactly the styl­is­tic effect being aimed at, but as in the case of the Lydia’s σπονδειάζοντες, affinities with Lucretius may be indicative.

4.  Manuscript Tradition and the Text As already noted, the Lydia is transmitted in the manuscripts as part of the Dirae. The Dirae is in turn listed, along with other poems of the Appendix Vergiliana, in a ninth-­century catalogue description of a lost manuscript from Murbach Abbey (known as ‘Murbacensis’), which with high probability can be identified as the archetype of the entire ­trad­ition.94 Subsequently, however, different poems of the Appendix appear to have more often been copied individually or in smaller groups, so that each has its own stemma as a result. The Dirae and the Lydia survive together in two independent branches. One is represented by two primary manuscripts: m (Munich 305, eleventh–twelfth cent.) and n (Munich 18059, eleventh cent.), which only have very few discrepan­ cies and are usually cited together as M (‘Monacensis’). The other branch is more complex and is represented by two surviving manuscripts, S (‘Stabulensis’, Paris 17177, tenth cent.) and F (‘Fiechtianus’, Melk 717, tenth cent.), and a lost one, L (‘Iuuenalis ludi libellus’).95 L is in turn reconstructed from a family of manuscripts, whose mutual relation­ ships are difficult to establish exactly: W (Trier 1086, ninth–tenth cent.), B (‘Bembinus’, Vatican, Lat. 3255, ninth–tenth cent.), E (Paris 8093, tenth cent.), A (Paris 7927, tenth–eleventh cent.), T (Paris 8069, eleventh cent.), P (‘Petavianus’, Vatican, Reg. Lat. 1719, eleventh cent.), O (Oxford, Bodl. Auct. F I 17, fourteenth cent.).96 S was reused as a cover for another manuscript, so that its text is partially effaced throughout the Lydia; F is missing lines 104–22 of the Lydia. 93  Note Lipka 2001, 131: ‘for generic reasons the number of “pure” prosaisms in the Eclogues is very small’; the bucolic colouring of the Eclogues seems largely to be due to an admixture of colloquialisms along with botanical and agricultural terminology. 94  See esp. Zogg 2018, cf. Kayachev 2020, 31–2. 95  On the contested relationship between SFL, see Clausen 1964, 123–4, Courtney 1968, 138. 96  For the addition of the two last manuscripts, see Reeve 1976, 242.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

42 INTRODUCTION Although disagreements between the manuscripts are not always negligible, in most cases the reading of the archetype can be inferred with sufficient probability.97 The Murbacensis itself, however, was already badly corrupt (even if to a lesser extent than in the Dirae), which makes bridging the gap between the archetype and the original a far more urgent and difficult part of the editor’s job. I have avoided the obelus and printed a conjecture wherever the paradosis appears to be corrupt, even if in some cases the accepted emendation may be highly tentative. The concise critical apparatus is constructed accordingly: it only has an entry when there is significant doubt as to what the reading of the archetype was or that it is correct (this in particular means that if at least one of the two branches preserves what appears to be the genu­ ine reading, it is assumed that it reflects the archetype rather than inno­ vating, and is printed without an entry in the apparatus). The siglum Ω is used to designate the archetype’s likely readings rather than the strict agreement of all the primary witnesses (in practice this usually means the agreement of M with at least one of SFL). The apparatus aims to regis­ter the earliest appearance for every conjectural reading that is either accepted or cited, but it is likely that there are misattributions, especially in the case of later manuscripts and early printed editions, which I have not explored systematically.

97  Existing editions on the whole succeed in establishing the paradosis. Vollmer 1910 was the first to be based on an essentially correct understanding of the relationships between the manuscripts. Giomini 1953 offers a rich apparatus, though sadly not fully reliable. Kenney 1966 had for a long time been the best critical text, despite its liberal use of the obelus, but Goold in Fairclough and Goold 2000, 386–403 has provided strong competition, even if its textual interventions are not always successful. Salvatore 1997 offers some improvements in the reporting of manuscript readings, but is reprehensibly even more conservative than Kenney. Rupprecht 2020 is unremarkable (except for integrating the Lydia back into the Dirae).

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

LY DIA ( APPE N DIX V E RG I LIA NA )

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

Lydia Inuideo uobis, agri formosaque prata, hoc formosa magis, mea quod formosa puella in uobis nostrum tacite suspirat amorem. uos nunc illa uidet, uos nunc arridet ocellis, uos nunc alloquitur, uobis mea Lydia ludit et modo submissa meditatur carmina uoce, interdum cantat mihi quae cantabat in aurem. inuideo uobis, agri: discetis amare. o fortunati nimium nimiumque beati, in quibus illa pedis niuei uestigia ponet et roseis uiridem digitis decerpserit uuam (dulci namque tumet nondum uindemia Baccho) aut inter uarios, Veneris spectacula, flores membra reclinarit teneramque illiserit herbam et secreta meos furtim narrabit amores. gaudebunt siluae, gaudebunt mollia prata, tardabunt riui labentes murmure lymphas et gelidi fontes, auiumque silentia fient, dum mea iucundas expromet cura querelas.

105

110

115

121 120

inuideo uobis, agri: mea gaudia habetis et uobis nunc est mea quae fuit ante uoluptas. a Dirarum carmine seiunxit Lydiaeque Catonianae nomen addidit Jacobs   106 in Heinsius: est Ω   nostrum tacite scripsi: t. n. Ω   107–8  uos – ocellis versu 108, uobis – ludit versu 107 Ω, transmutavi   109 modo scripsi: mea Ω   110  interdum cantat scripsi: cantat et interea Ω, interdum Wagner   112 nimiumque Parrhasius: multumque Ω    114 et scripsi: aut Ω   115 dulci Boccaccius: -is Ω   uindemia scripsi: uitecula Ω    116 Veneris cod. Helmst. 332: -em Ω   spectacula scripsi: spumantia M, stipendia SL    119 gaudebunt2] an ridebunt?   121  ante 120 Ribbeck   murmure scripsi: currite Ω: currere ed. Ald. 1517: gurgite Ellis, limite Heyworth   lymphas scripsi: -ae Ω    122 expromat Heinsius, -et Heyworth: exponat Ω, -et Scaliger   

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

46  APPENDIX VERGILIANA at mihi tabescunt morientia membra dolore et calor infuso decedit frigore leti, quod mea non mecum domina est. non ulla puella doctior in terris fuit aut formosior illa: fabula ni uana est, tauro Ioue digna uel auro (Iuppiter, auertas aurem) mea sola puella est. felix taure, pater magni gregis et decus, a te uaccula non umquam secreta cubilia captans frustra te patitur siluis mugire dolorem. et pater haedorum felix semperque beate, siue petis montes praeruptaque saxa pererras siue tibi siluis noua pabula uestigare siue libet campis, tecum tua laeta capella est. et mas quicumque est, illi sua femina iuncta interpellatos numquam tolerauit amores. cur non et nobis facilis, natura, fuisti? cur ego crudelem patior tam saepe dolorem? sidera per nitidum cedunt cum pallida mundum inque uicem Phoebi surgens redit aureus orbis, Luna, tuus tecum est: cur non est et mea mecum? Luna, dolor nosti quid sit: miserere dolentis. Phoebe, gerens laurus semper celebrabis amorem. et quicumque deus, nisi fictum fama locuta est, omnibus in terris secum sua gaudia gestat aut inspersa uidet mundo (quae dicere longum est).

125

130

135

140

145

125 mihi ed. Ald. 1517: male Ω   tabescunt ed. Rom. 1471: -ant Ω   126 leti scripsi: mortis Ω   127 ulla Boccaccius: illa Ω   128 illa Heyworth: ac si Ω   129 ni Heyworth: non Ω   132 uaccula] bucula Heinsius   135 praeruptaque n: -tos Ω    pererras Heinsius ex ‘codice veteri’: -ans Ω   136 siluis L: silluis F, siluas S, si uis M    uestigare Mähly: fastidire Ω   138 quicumque Badius: quo- Ω: quod- cod. Monac. 21562   139 tolerauit scripsi: plorauit Ω   140 fuisti Salmasius: fuisset Ω    142 nitidum Giardina: uiridem Ω   cedunt Haupt: redeunt Ω   143 Phoebi cod. Vat. lat. 3255: -e FL, -o S, -us M   surgens Mähly: currens Ω   redit Haupt: atque Ω    144 tuus Boccaccius: tui Ω   146  Phoebe (pheope n) M: -o FL   l. semper scripsi: in te l. Ω   celebrabis Scaliger: -auit Ω   147 quicumque Schmidt: quae pompa Ω    deus scripsi: -um Ω   fictum scripsi: siluis Ω: lusus Schmidt, uilia Vollmer   148 omnibus in terris scripsi: omnia uos estis Ω: an omnia uestigans vel sim.?   

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

LYDIA  47 aurea quin etiam cum saecula uoluebantur, condicio similis fuerat mortalibus illis. haec quoque praetereo: notum Minoidos astrum quaeque uirum uirgo sicut captiua secuta est. laedere, caelicolae, potuit uos nostra quid aetas, condicio nobis uitae quo durior esset? ausus egon primus castos uiolare pudores sacratamque meae uittam temerare puellae iam mea matura cogor nece soluere fata? istius atque utinam facti mea culpa magistra prima foret: letum uita mihi dulcius esset. nam mea non ullo moreretur tempore fama dulcia cum Veneris furatus gaudia primus dicerer atque ex me dulcis foret orta uoluptas. at mihi non tantum tribuerunt inuida fata auctor ut occulti noster foret error amoris. Iuppiter ante, sui semper mendacia fingens, cum Iunone, prius coniunx quam dictus uterque est, gaudia libauit dulcem furatus amorem.  et tecum tenera est gauisa illidere in herba purpureos flores quos insuper accumbebat, candida formoso supponens bracchia collo (tum credo fuerat Mauors distentus in armis, nam certe Vulcanus opus faciebat et illi turpabant massae tristi fuligine barbam).

150

155

160

165

168a 170

151  similis fuerat ed. Paris. 1540: similisque foret Ω   152 praetereo ed. Paris. 1540: -a Ω: -am Boccaccius   156 egon ed. Paris. 1540: ego Ω   157 uittam Laetus: uitam Ω    temerare Broukhusius: temptare Ω   158  iam mea matura Heyworth: immatura meae Ω: iam matura mea iam Ellis, mea iam Wernsdorf   161 nam cod. Ambros. O 74 sup.: non Ω    162 primus ed. Paris. 1534: -um Ω   163 dicerer SFB: -eret ML   164 at scripsi: nam Ω: nunc Hermann, sed Cabaret-­Dupaty   inuida fata Heinsius: impia uota Ω    166 fingens scripsi: factus Ω   168a  lacunam Heinsius, supplevi exempli gratia    169 tecum Gronovius: mecum SFL, mea cum M   est add. László   illidere Heyworth: est ludere Ω: est laedere Canter   171 candida ed. Argent. 1527: grandia Ω    bracchia ed. Ald. 1517: gaudia Ω   174 turpabant Giardina: -atque Ω: (-)que del. iam Scaliger   massae scripsi: mala Ω: malam Scaliger: flammae Giardina   tristi huc ­transtuli, versus initio Ω   

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

48  APPENDIX VERGILIANA non Aurora nouos totiens raptauit amores atque rubens oculos croceo celauit amictu? talia caelicolae, numquid minus aurea proles? ergo quod deus atque heros, cur non minor aetas? infelix ego non illo qui tempore natus quo facilis natura fuit. sors o mea laeua nascendi miserumque genus cui sera libido est. tantam fata meae uitae fecere rapinam ut maneam quod uix oculi cognoscere possint.

175

180

175 totiens raptauit scripsi: etiam plorauit Ω   176 croceo Boccaccius: roseo Ω    177 proles Vonck: promo Ω   181 cui Putsche: quo Ω   182  tantam fata meae uitae Canal (t. f. iam Heinsius): tantum uita (uitae WATP) meae cordis MWATP: tanta meae (me S) uitae cordis SFBEO   183  oculi . . . possint scripsi: oculis . . . possis Ω   

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

Translation I envy you, fields and beautiful meadows, the more beautiful because my beautiful girl quietly sighs among you with love for me. It is you she now looks at, it is you she smiles at with her eyes, it is you she speaks to, for you does my Lydia play as she sometimes rehearses songs in a subdued voice and sometimes sings what she sang in my ear. I envy you, fields: you will learn to love. [111] Highly fortunate and highly blessed the fields in which she will put the soles of her snow-­white feet and pluck unripe grapes with her rosy fingers (for the harvest does not yet swell with sweet wine), or reclining among motley flowers, Venus’ spectacle, will crush tender grass and, alone, stealthily tell of her love for me. The woods will rejoice, the soft meadows will rejoice, the streams will restrain their waters flowing with noise, and so will the icy springs, and among the birds there will be silence, while my love utters her melodious complaints. [122] I envy you, fields: you possess my delights and to you now belongs the pleasure that was mine before. But my dying limbs waste away with pain and warmth departs with the suffusion of death’s chill, because my lady is not with me. No other girl on earth was more skilled or more beautiful than her: unless the tale is false, my girl alone is worthy of Jupiter as a bull or gold (Jupiter, don’t listen). [130] You happy bull, father and pride of a big herd, the cow never seeks a secluded lair away from you and lets you bellow with pain in the woods in vain. And you, father of kids, always happy and blessed: whether you climb mountains and roam over steep rocks, whether it takes your fancy to track down new fodder in the woods or on the plain, your nanny-­goat is happily with you. And whatever male there is, his own female, ­in­sep­ar­able from him, has never suffered their love to be broken. Nature, why have you not been benign to me as well? Why do I all the time suffer cruel pain? [141] When the pale stars disappear across the luminous sky and in turn Phoebus’ golden disc rises back, Moon, your love is with you: why is not mine with me too? Moon, you know what pain is: take pity on one in

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

50 Translation pain. Phoebus, by wearing laurel you will always celebrate your love; and any god whatever (unless Rumour has told a lie) carries his delights with him all over the world or sees them scattered in the sky (these things it is too long to narrate). [149] Yes, likewise when the golden ages were passing, the rules were similar for those mortals. But I pass over this: the constellation of Minos’ daughter is familiar, as is the girl who followed a man like a captive. Heaven-­ dwellers, in what way could our age have offended you, so that the rules of life should be harder for us? [155] Is it because I was the first to dare defile pure chastity and profane the holy headband of my girl that I am now forced to cut short my destiny with a speedy death? Would that my offence were the first teacher of this practice: then would death be sweeter to me than life. For my fame would never die, if I were said to have been the first to steal the sweet delights of Venus and if sweet pleasure had originated from me. But envious Fate did not bestow on me so great an honour, that my misdeed should be the origin of secret love. [165] Previously, Jupiter (who always devises disguises of his own self) tasted delights with Juno, before they were made man and wife, by stealing sweet love. and in tender grass with you enjoyed crushing purple flowers on which she was reclining, her white arms placed under your fair neck (I believe Mars was then detained in service, as for Vulcan, he certainly was working, and the hot lumps of metal were defiling his beard with sombre soot). Did not Aurora again and again abduct new lovers and, blushing, hide her eyes with a saffron cloak? [176] That’s what the heaven-­dwellers did, and the golden breed, did they do it any less? So, what the god and the hero did, why not the lesser age? Unlucky me that I was not born at the time when nature was benign. How unpropitious is the lot of my birth, and how wretched the race for whom desire is late. Fate reaved so much of my life that eyes can barely recognize what is left of me. [183]

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

COMMENTARY

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary Structure. The transmitted text of the Lydia is eighty lines long (104–83). It can be divided into three sections, each of which can be isolated on grounds of content even if their boundaries are not formally demarcated; the sections are roughly of the same length: 104–30 (twenty-­seven lines), 131–55 (twenty-­five lines), 156–83 (twenty-­eight lines, plus a lacuna after 168 that need not be longer than one line). The first section expresses the speaker’s envy for the fields which, unlike him, continue to enjoy Lydia’s presence. The second redirects his envy at the animals, the gods, and the heroes, who, the speaker claims, are inseparable from their mates, unlike him. The third section tries and fails to explain why the speaker has been treated unjustly in comparison with the gods and heroes. Each section can in turn be divided into three blocks, which I for convenience’s sake call stanzas, even though they are only demarcated by (quasi-)refrains in the first section (104, 111, 123). Although the stanzas range in length from six (150–5) to over eleven lines (166–76, involving a lacuna), each displays a degree of thematic unity (see introductory nn. to the individual sections for details). On the formal level, in each section the conclusion of the third stanza appears to echo that of the first: 182–3 pick up the topic of fate’s involvement in the speaker’s life from 164–5; 154–5 reiterate the speaker’s question why he has been treated unjustly, first posed at 140–1; in the first section this pattern is less evident, probably because the presence of refrains makes it redundant, but note e.g. the repetition of aurem from 110 in 130 (note also that while the first and the third stanzas address the fields in the second person, in the middle stanza they are referred to in the third person). The third section can also be seen as answering the first. In its first stanza the speaker professes his unattainable aspiration to be the teacher of (clandestine) love (159–60)—a role ascribed to Lydia in the first stanza of the first section (111). In both sections the third stanzas

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

54  Commentary ON 104–30 contain references to the speaker’s physical condition (125–6 and 182–3). Their middle stanzas are linked by the pointed parallelism of language between the depictions of Lydia (116–17) and Venus (169–70) reclining in the grass. While it is possible that this symmetrical arrangement is merely a reflection of the poet’s compositional technique, rather than being intended as a manifest formal feature of the poem, I separate the stanzas in print so as to make the poem’s structure immediately transparent, while also bringing its presentation closer to that of the Dirae with its much more pronounced stanzaic organization. 104–30  The speaker’s envy for the fields which share Lydia’s company in his (unexplained) absence is the dominant leitmotif throughout the first section, emphasized by the repetition of inuideo uobis, agri in the refrains (104, 111, 123). We can perhaps observe how his feeling of benign envy gradually develops into that of bitter jealousy over the ­stanzas: while in the first stanza the speaker imagines Lydia merely present in the fields, in the second she makes physical contact with them, which in the third leads to his resentful comment that the fields ‘possess’ his beloved (123–4). More explicitly the third stanza names Jupiter as a potential rival (129–30), and the allusion to his metamorphoses retrospectively suggests that the fields too may be less passive than they appear. (And if the speaker’s contention that Lydia might attract Jupiter is true, this could explain why inuida fata [164] broke up their relationship.) Another important theme—­which will disappear in the subsequent sections—­is Lydia’s proficiency as a bucolic singer: in the first stanza she is presented as practising and performing (109–10), in the second her singing is endowed with Orphic powers (119–22), in the third she is claimed to be puella doctissima (127–8); all this in contrast to the speaker, who is not explicitly signalled as a poet figure (except for the fact of his speaking in bucolic hexameters). 104–11  The stanza apostrophizes the ‘fields and meadows’ which continue to share Lydia’s company while the speaker is kept away. This could be seen as a development of the (originally) tragic conceit by which a departing hero takes leave of his land (cf. e.g. Hunter 1999, 98), and the most relevant intertext here are the farewells of the dying Daphnis in Theocr. 1, note esp. 116–17 ὁ βουκόλος ὔμμιν ἐγὼ Δάϕνις οὐκέτ᾽ ἀν᾽ ὕλαν, |

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 104–11  55 οὐκέτ᾽ ἀνὰ δρυμώς, οὐκ ἄλσεα: while Daphnis focuses on his impending disappearance from the bucolic landscape, our speaker emphasizes Lydia’s continuing presence. Lydia’s rootedness in the landscape is to an extent adumbrated by the depiction of Aphrodite in Bion’s Lament which likewise places her in a markedly bucolic environment (23 ὀξὺ δὲ κωκύοισα δι᾽ ἄγκεα μακρὰ ϕορεῖται, 36 πάντας ἀνὰ κναμώς, ἀνὰ πᾶν νάπος οἰκτρὸν ἀείδει); yet while in Bion the dying Adonis remains the emotional focus for both Aphrodite and the animated landscape, here it is Lydia who is the centre of attention for both the agri and the speaker. The leitmotif of envy, bordering on jealousy, for the fields enjoying Lydia’s presence is likewise an innovation that seems to lack exact parallels in earlier poetry (cf. DMALL 91–2 s.v. ‘c[elo] de un objeto’). The traditional conceit by which the lover wishes to be transformed into an object that has access to his beloved (see McKeown 1998, 316–17, Hunter 1999, 114–15) may have provided inspiration. Theocr. 3.12–14 αἴθε γενοίμαν | ἁ βομβεῦσα μέλισσα καὶ ἐς τεὸν ἄντρον ἱκοίμαν, | τὸν κισσὸν διαδὺς καὶ τὰν πτέριν ἅ τυ πυκάσδει, being the only prior bucolic instance, could be a relevant intertext, esp. as the motif of envy appears later in the idyll, 49–50 ζαλωτὸς μὲν ἐμὶν ὁ τὸν ἄτροπον ὕπνον ἰαύων | Ἐνδυμίων· ζαλῶ δέ, ϕίλα γύναι, Ἰασίωνα. (Note that Theocr. 3 shares a number of formal features with the Lydia: it is a lover’s monologue lacking a dramatic frame, opens with a first-­person verb, κωμάσδω, and makes extensive use of mythological exempla, 40–51.) The closest parallel comes from Ovid’s elegy addressed to a ring gifted to Corinna, which appears to be humorously engaging with the Lydia: Am. 2.15.7–9 felix, a domina tractaberis, anule, nostra; | inuideo donis iam miser ipse meis. | o utinam fieri subito mea munera possem. While featuring the traditional wish to be transformed into an object accompanying the beloved (which our poem lacks), Ovid parallels the Lydia by making the ring the subject of a μακαρισμός predicated on the beloved’s presence (cf. 112–13) as well as by expressing his envy for it. Finally, one might wonder whether Meliboeus’ non equidem inuideo at Verg. Ecl. 1.11 (the only other pre-­ Ovidian occurrence in poetry of the first-­person form) is a subversive echo: like the Lydia speaker, Meliboeus has left his farm, yet he does not envy Tityrus who, like Lydia, did not leave the countryside and con­ tinues to sing there.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

56  Commentary ON 104 104  The line is a quasi-­refrain, which shares its first half (inuideo uobis, agri) with 111 and 123 (then refrains disappear). The closest formal parallel are the refrains in Bion’s Lament, many of which similarly vary the same beginning, αἰάζω τὸν Ἄδωνιν (note also that the first word is a first-­person singular verb, like inuideo); see intr., section 2. inuideo is a strong word to begin a poem with (the effect being further strengthened by its repetition at 111 and 123): it is more common and conventional to deny that one is envious or jealous (cf. Verg. Ecl. 1.11 quoted 104–11n., Ov. Her. 2.75 nec inuideo, Pont. 1.8.8). Comparable in thrust is the famous speech of Agamemnon early in Eur. IA: 16–19 ζηλῶ σέ, γέρον, | ζηλῶ δ᾽ ἀνδρῶν ὃς ἀκίνδυνον | βίον ἐξεπέρασ᾽ ἀγνὼς ἀκλεής· | τοὺς δ᾽ ἐν τιμαῖς ἧσσον ζηλῶ (where ζηλῶ is glossed with μακαρίζω in scholia and lexica: cf. the μακαρισμοί in 112, 131, 134). An interesting parallel is also supplied by a letter from Caelius to Cic. (Fam. 8.4), which opens with inuideo tibi, motivated by Cic.’s having witnessed events missed by Caelius; even closer to the Lydia speaker’s is the mo­ti­ v­ation found in another letter to Cic. (Pollio in Fam. 10.31): inuideo illi [Cornelius Gallus? see Hollis 2007, 226] tamen quod ambulat et iocatur tecum; by contrast, Ovid opens the Tristia with the claim that he does not envy his book travelling to Rome without him (1.1.1 nec inuideo). formosaque prata For aesthetic appreciation of agricultural land, cf. esp. Var. Rus. 1.4.2 nemo enim eadem utilitati non formosius quod est emere mauult pluris, quam si est fructuosus turpis; cf. further, in praise of natural landscapes, Sen. Ep. 90.43 perlucidi fontes riuique non opere nec fistula nec ullo coacto itinere obsolefacti sed sponte currentes et prata sine arte formosa. The adj. has a remarkable distribution in Latin poetry: Catullus uses it in one epigram (86.1, 3, 5: see below), Virgil favours it in Ecl. (15x) but not in Georg. (1x) or Aen. (0x), Hor. admits it sparingly (3x), Prop., Tib., and Ov. (incl. Met.) use it freely, but then it is completely absent in Val. Fl., Stat. (except Silv. 3.4.44), and Sil. (cf. e.g. Watson 1985, 440–2, Coleman 1999, 52, Konstan 2014, 148–69). Axelson (1945, 60–1) explained this distribution by the adj.’s ‘etwa triviales Gepräge’, yet at least two distinct usages are involved: (1) with an original sense ‘well-­shaped, able-­bodied, thriving’, in particular applied to animals (but also to other natural features and phenomena), which may have had a mundane or prosaic colouring (cf. e.g. Pl. Merc. 229

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 104–6  57 mercari uisus mihi sum formosam capram, Var. Rus. 2.2.5 agnos procreant formosos); and (2) with a newer urbane sense ‘shapely, attractive, sexy’, which may have felt to be too modern for higher-­register poetry (note the substantival use of formosa for ‘a beauty, belle’): it is with this latter usage that Catullus engages in 86 Quintia formosa est multis, mihi candida, longa, | recta est. haec ego sic singula confiteor, | totum illud ‘formosa’ nego . . . —Quintia may be ‘well-­shaped’ but is not ‘attractive’— apparently without fully embracing it. The Lydia plays with the two usages, by first using the adj. of the meadows (104–5) and then of Lydia (105, 128, where see n.) and Adonis (171), most clearly at 105 where the meadows’ beauty is explicitly predicated on Lydia’s (note that in the Dirae the adj. is repeatedly used of Lydia’s wood, at 27 and twice at 32). The same two usages are likewise prominent in Ecl., and are similarly juxtaposed at 5.44 (quoted 105n.); it has been speculated that formosus may have been popularized in poetry by Calvus or Gallus (Lipka 2001, 8–10, Gagliardi 2015), but the Lydia might be able to account for Virgil’s treatment of the adj. in a more straightforward way. 105–6  For the idea that the presence of the beloved can transform a landscape, cf. esp. [Theocr.] 8.41–8, which likewise emphasizes his/her beauty (47 ἔνθ᾽ ὁ καλὸς Μίλων βαίνει ποσίν, 43 ἔνθα καλὰ Ναῒς ἐπινίσσεται). 105 hoc . . . magis . . . quod,  ‘the more . . . because’: the construction is largely prosaic; in poetry, cf. Lucr. 2.125–8 hoc etiam magis haec animum te aduertere par est | corpora quae in solis radiis turbare uidentur, | quod tales turbae motus quoque materiai | significant clandestinos caecosque subesse, Luc. 8.578–9 hoc magis impatiens egresso desse marito | quod metuit clades, Sen. Phoen. 204–5 et hoc magis te, genitor, insontem uoca, | quod innocens es dis quoque inuitis; the parallels show that formosa magis should not be construed as an analytic comparative form, but rather the phrase means something like ‘there is the more reason for your being (perceived as) beautiful because’. formosa . . . formosa For the repetition of the adj., cf. Dirae 32–3 (cum) formosaeque cadent umbrae, formosior ipsa | silua cadet, Verg. Ecl. 5.44 formosi pecoris custos, formosior ipse, Col. 10.299 formoso Nais puero formosior ipsa, Prop. 2.18.29–30 deme mihi: per te poteris

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

58  Commentary ON 105–6 formosa uideri; | mi formosa sat es, si modo saepe uenis, and see Wills 1996, 232–5. puella is the standard term for the beloved in classical poetry, esp. elegy, since Catullus, and here it reoccurs at 127, 130, 157, against a sole occurrence of its alternative, domina, likewise at 127, which is consistent with their relative frequency elsewhere (cf. Thorsen 2021, 219–23; for further alternatives, cf. DMALL 30–1). 106  est uobis (Ω) is not impossible in itself (cf. 124 uobis nunc est), but the asyndeton with suspirat is unattractive (adding et after tacite, with Broukhusius 1708, 130, is not a plausible solution). One line of correction has been to replace est with an interjection: en (Heinsius in Burman 1773, 668), heu (Ribbeck 1895, 97), st (Lindsay 1918, 62). Another has been to assume a lacuna between 105 and 106, while writing et (Boccaccio, cod. Laur. 33.31, f. 26r.) for est (Kenney 1966, 10, after Goodyear). Two objections arise. First, and fundamentally, construing nostrum suspirat amorem with dative uobis (‘she sighs in love for me into your ears’) does not produce a plausible sense: while at 107–10 Lydia is indeed portrayed as conversing with the farmlands, lovesick sighing is not an activity performed for the benefit of an audience (contrast 108 uobis . . . ludit). Second, thus interpreted, 106 provides a somewhat misplaced explanation for the speaker’s envy: ‘I envy you because she lets you hear her sighing for me’ rather than ‘because, even if she sighs for me, she shares your company’ (positing a lacuna is of course immune to this objection). The sense we need is ‘she is with you’, and the easiest way to obtain it is by writing in (likewise Heinsius), for which 113 in quibus provides an inexact but sufficient parallel (Petry 1895, 15 proposed ex, ‘aus eurer Mitte’, as an easier correction, but it produces questionable Latin, and ī → ẽ is easy enough). tacite nostrum (Ω) invites taking the adv. closely with in uobis (‘secretly in your midst’), rather than with its verb from which it is ­sep­ar­ated by nostrum as well as the caesura (contrast 118 meos furtim narrabit amores); I think transposing nostrum tacite produces a more plausible word order, first by bringing together the adv. and its verb, but also by juxtaposing nostrum with in uobis and thus highlighting the reason behind the speaker’s jealousy (‘she sighs for me, but keeps your

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 106–8  59 company’), cf. 108 uobis mea Lydia ludit (and further 123–4); for displacement of adjacent spondaic and/or anapestic words in dactylic verse, see Heyworth 2015, 224, and cf. the variation between the manuscripts at 114 (uiridem digitis M: digitis uiridem SL). nostrum . . . suspirat amorem: ‘expresses her love for me by sighing’ (cf. OLD s.v. suspiro 1c), for the construction cf. 133 mugire dolorem ‘to express his pain by lowing’ (in support of this construal, cf. also 118 meos furtim narrabit amores, which echoes the present line); Tib. 1.6.35 te tenet, absentes alios suspirat amores (imitated in [Tib.] 3.11.11 alios iam nunc suspiret amores) is somewhat different: Maltby 2002, 271 takes it in the sense of suspirare in aliquo (OLD s.v. suspiro 1b), but amores seems rather to be an internal acc., as in Prop. 2.34.25 Lynceus ipse meus seros insanit amores (for this pregnant sense of suspirare ‘to be lovesick’, cf. Hor. Carm. 3.7.10–11 suspirare Chloen et miseram tuis | dicens ignibus uri); cf. further Lucr. 4.1192 nec mulier semper ficto suspirat amore; for the motif of sighing, cf. in general DMALL 411 s.v. ‘suspiros de amor’. tacite: probably ‘secretly’ (OLD s.v. 2, cf. 118 furtim narrabit) rather than ‘quietly’ (for the idea of sighing wordlessly, however, cf. Ov. Her. 21.201 ingemit et tacito suspirat pectore, Stat. Theb. 9.711 tacito ducunt suspiria uoto); the adv. does not otherwise appear in the poetry of the first cent. bc, then first at Ov. Fast. 1.65, Pont. 4.10.54 (on the avoidance in poetry of adverbs in -e, see Axelson 1945, 62–3; on Ovid’s aberrant practice, cf. Kenney 2002, 42). 107–8  As I suggest in Kayachev 2022, 712–13, the second hemistichs of 107 and 108 are transmitted in the wrong places, and swapping them produces a far more logical sequence of actions: Lydia sees (uidet) the fields, smiles (arridet) at them, speaks (alloquitur) to them, and performs (ludit) for them (the last being elaborated in 109–10). This progression is no doubt intended to evoke the conceit of the ‘five stages of love’ (quinque lineae amoris), whose most explicit ancient statement may be that of Don. on Ter. Eun. 640 quinque lineae perfectae sunt ad amorem: prima uisus, secunda alloquii, tertia tactus, quarta osculi, quinta coitus (there follow numerous mediaeval parallels: see e.g. Friedman 1965) but which has plausibly been traced back to Plato and shown to be ubiquitous in Greek and Latin literature (see Krause 2014; cf. further

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

60  Commentary on 107–8 DMALL 158 ‘fases del e[namoramiento]’); note esp. Pl. Phaedr. 255e3 ὁρᾶν, ἅπτεσθαι, ϕιλεῖν, συγκατακεῖσθαι, Pl. Merc. 745 uidere, amplecti, osculari, alloqui, Eun. 373 cibum una capias adsis tangas ludas propter dormias, Ov. Met. 10.343–4 ut praesens spectem Cinyran tangamque loquarque | osculaque admoueam, si nil conceditur ultra, Rufinus AP 5.94.3–4 εὐδαίμων ὁ βλέπων σε, τρισόλβιος ὅστις ἀκούει, | ἡμίθεος δ᾽ ὁ ϕιλῶν, ἀθάνατος δ᾽ ὁ γαμῶν (note the μακαρισμοί); more than once, the ‘stages of love’ appear in erotodidactic contexts (which can add nuance to 111 discetis amare): Ov. Ars 1.660–71, Longus 2.7.7 Ἔρωτος γὰρ οὐδὲν ϕάρμακον, οὐ πινόμενον, οὐκ ἐσθιόμενον, οὐκ ἐν ᾠδαῖς λαλούμενον, ὅτι μὴ ϕίλημα καὶ περιβολὴ καὶ συγκατακλιθῆναι γυμνοῖς σώμασι (spoken by Philetas: note how Daphnis and Chloe misunderstand his euphemistic συγκατακλιθῆναι at 2.8.5 ὅσα εἶπεν ἄρα ϕάρμακα, ταῦτα ζητητέα, ϕίλημα καὶ περιβολὴν καὶ κεῖσθαι γυμνοὺς χαμαί—­similarly Lydia’s lying down on the ground at 116–17 is suggestive of intercourse), Ach. Tat. 2.4.3–4 δεῖ δέ σε καὶ τὴν κόρην μὴ μέχρι τῶν ὀϕθαλμῶν πειρᾶν, ἀλλὰ καὶ ῥῆμα δριμύτερον εἰπεῖν. τότε δὲ πρόσαγε τὴν δευτέραν μηχανήν. θίγε χειρός, θλῖψον δάκτυλον, θλίβων στέναξον. ἢν δὲ ταῦτά σου ποιοῦντος καρτερῇ καὶ προσίηται, σὸν ἔργον ἤδη δέσποινάν τε καλεῖν καὶ ϕιλῆσαι τράχηλον. While the present passage (107–10) deals mainly with the first two stages (with 108 ludit possibly hinting at the third: see n.), the next stanza may be seen to be alluding to the final three: stepping on the meadow with bare feet (113) may evoke ‘touches’, plucking (and eating?) grapes (114) perhaps ‘kisses’ (or likewise ‘touches’), and reclining among flowers (116–17) ‘intercourse’. 107 uidet  For the nuance of the verb, cf. perhaps OLD s.v. uideo 6: ‘To see and converse, etc., with’. arridet ocellis: no exact parallels; for arridere with an abl. referring to the face, cf. Sil. 5.227–8 laeto Victoria uultu | arridens, Calp. Ecl. 4.86 laetus et augusto felix arrideat ore [sc. the emperor], 5.46–7 fronte serena | blandius arrisit [sc. the sky], which all refer to a figure of power showing benevolence; ‘laughing eyes’ seem otherwise to express mockery and/or superiority, cf. Calp. Ecl. 6.15 ridens oculis crinemque simillimus auro, Hymn. Hom. 7.14–15 μειδιάων ἐκάθητο | ὄμμασι κυανέοισι, Theocr. 7.19–20 καί μ᾽ ἀτρέμας εἶπε σεσαρὼς | ὄμματι μειδιόωντι, γέλως δέ οἱ

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 107–10  61 εἴχετο χείλευς (with Hunter 1999, 157–8); possibly the implication is that Lydia is a local deity of some standing rather than an ordinary country girl, though the verb itself occurs in contexts merely expressing friendliness (e.g. Ter. Ad. 864 nulli laedere os, adridere omnibus); for smiling in erotic contexts, cf. e.g. Sappho fr. 31.5 γελαίσας ἰμέροεν (with Catull. 51.5 dulce ridentem, Hor. Carm. 1.22.23), Ov. Ars 3.513 ridenti mollia ride. 108 alloquitur The verb is predominantly epic, largely avoided in erotic elegy (except at Prop. 1.17.2 nunc ego desertas alloquor alcyonas, where it may have an ironic nuance); note, however, also its use in reference to the ‘second’ stage of love at e.g. Pl. Merc. 745 (quoted 107–8n.). uobis: ‘ad delectationem vestram, vix pro abl. accipiendum’ (TLL 7(2).1770.62–3, cf. Kayachev 2022, 712–13). Lydia ludit The etymological wordplay (Lorenz 2005, 16–17) has a strong pedigree: while the juxtaposition at Pl. Bac. 129 non omnis aetas, Lyde, ludo conuenit may be motivated solely by phonetics, the derivation of ludius ‘dancer’ from Lydius—­based on the idea that ludi were an institution taken over from the Etruscans, who in turn were believed to be of Lydian descent—­was an established etymology by the time of Varro (see Waszink 1948, van Rooy 1952, 240–1); the same wordplay is also present (if less conspicuous) in Hor. Carm. 3.11.7–10 dic modos, Lyde quibus obstinatas | applicet aures, | quae uelut latis equa trima campis | ludit exsultim (López-­Cañete Quiles 2013), which appears to be a related context (note that at 13–14 the ode also develops the Orphic topos of the power of music over nature, cf. 119–22). The wordplay may support taking the verb in a more specialized sense, ‘to perform’ (cf. OLD s.v. ludo 6). Given the context, though, the erotic connotations, suggesting flirting or foreplay (cf. OLD s.v. 4), are no doubt present too (cf. 107–8n., in particular Pl. Eun. 373 quoted there); cf. in general DMALL 215–16 s.v. ‘juego de amor’. 109–10  exhibit some remarkable points of contact with, and may have served as an intertext for, Manilius’ account of people born under the constellation of Lyra, who sing even when they are anxious or alone (5.334–6): quin etiam curas inter secreta mouebit | carmina furtiuo ­modulatus murmure uocem, | solus et ipse suas semper cantabit ad aures.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

62  Commentary ON 109–10 Both points apply to Lydia: she is alone in the fields and misses the speaker’s company (note 106, 118), but continues to sing. The passage has two interrelated problems (see Kayachev 2022, 713– 14). At 110, interea (Ω) makes no sense, and Wagner’s (1867, 235) interdum is a compelling correction, also implying that the passage is a case of a ‘now – now’ construction: Lydia sometimes meditatur, sometimes cantat; however, cantat et interdum is still implausible, since in this construction the second clause seems never to be introduced by et; besides, interdum normally stands right at the head of the second clause (though cf. e.g. Ov. Met. 4.200 deficis interdum, 11.501 sternitur interdum, where it is shifted to the second place for metrical reasons), so it is necessary to transpose interdum cantat (while deleting et). At 109, mea (Ω) is problematic because meditari properly means ‘to compose’ or ‘to rehearse’ (see n.), which suggests that the direct object should be Lydia’s own songs, not the speaker’s (it is of course possible to rehearse a song written by someone else, but such a notion would require some elab­or­ation); at the same time a counterpart for 110 interdum is desirable (if not strictly indispensable), which is easily obtained by replacing mea with modo (I am not convinced that Prop. 2.33.38, quoted 109n., gives mea much support, as the similarity may well be coincidental). For the construction, cf. e.g. Prop. 2.15.5–6 nam modo nudatis mecum est luctata papillis, | interdum tunica duxit operta moram, Ov. Met. 2.188–90 animo metitur utrumque | et modo . . . | prospicit occasus, interdum respicit ortus, Moret. 29–31 modo rustica carmina cantat . . . interdum clamat Scybalen. 109 submissa . . . uoce,  ‘in a quiet voice’: cf. e.g. Cic. Flac. 66 sic submissa uoce agam tantum ut iudices audiant, Ciris 355 temptantur patriae submissis uocibus aures (Prop. 2.33.38 mea deducta carmina uoce legis is different, as it rather refers to pitch than volume, cf. Hollis 2007, 152). meditatur carmina The exact nuance of the verb in reference to mu­sic­al composition or performance is elusive (OLD s.v. meditor 7 ‘To work over (a song, etc.) in performance’ is not very helpful; TLL 8.579.75–6 ‘canere, recitare’ is likewise misleading), but it never seems to be used of formal singing in front of an audience and rather implies solitary practising (as in Verg. Ecl. 1.2), just as its standard sense ‘to prepare, practise, rehearse’ lets one expect; cf. esp. Hor. Ep. 2.2.76 i nunc et uersus

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 109–11  63 tecum meditare canoros, and note that at Ecl. 6.82–3 omnia, quae Phoebo quondam meditante beatus | audiit Eurotas (cf. Apul. Fl. 17 olores apud auios fluuios carmen senectae meditantur), the Eurotas overhears Apollo. That Lydia is here presented as composing or rehearsing songs is confirmed by the detail that she is doing so in a quiet voice (contrast e.g. Catull. 61.12–13 nuptialia concinens | uoce carmina tinnula), as well as by the opposition with 110 cantat. 110  The line appears to be echoed in Dirae 41 non iterum dices crebro quae, Lydia, dixti, which has, with bitter irony, an opposite sense: Lydia will no longer sing the songs she sang in the speaker’s company. cantat: the standard term for (esp.) bucolic singing, see Lipka 2001, 21–2. mihi quae cantabat in aurem Singing in someone’s ear is an odd notion, as is illustrated by Mart. 1.89.4 cantas in aurem, iudicas, taces, clamas (things normally done openly), cf. also Petr. 28.5 ad caput eius cum minimis symphoniacus tibiis accessit et, tamquam in aurem aliquid secreto diceret, toto itinere cantauit; the idiom in aurem is probably used here in the looser sense ‘privately’ rather than literally ‘in the ear’ (cf. e.g. Pl. Trin. 207 sciunt id quod in aurem rex reginae dixerit, Cic. Ver. 2.1.120 a qua muliere cum erat ad eum uentum et in aurem eius insusurratum, and note 106 tacite with n.), the point being that the fields now share Lydia’s intimacy which used to be reserved for the speaker; it is notably colloquial and largely avoided in higher-­register poetry (in the first cent. bc, only Hor. Sat. 1.9.9–10, Ov. Her. 3.23); for singing to the beloved, cf. DMALL 329 s.v. ‘poeta de amor’. 111  It might seem natural that a refrain should introduce the following stanza, as 104 and 123 do, but there are two reasons for taking 111 as the conclusion of the preceding stanza: on the one hand, its statement appears to be an inference from the situation described in 107–10 (‘you will learn to love by listening to Lydia’s singing’), though conceivably it could imply the activities of 113–17 (‘you will be taught to love by Lydia’s making love to you’); on the other, and in my view decisively, in 112–22 the farmlands are not referred to in the second person. discetis amare The fields will learn (on the force of the future tense, cf. 112–22n.) to love by listening to Lydia’s songs in which she

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

64  Commentary ON 111–22 expresses her love for the speaker (cf. in a way Verg. Ecl. 1.5 formosam resonare doces Amaryllida siluas: by echoing Tityrus’ song, the woods proclaim Amaryllis to be beautiful, which amounts to a profession of love, cf. 128n.). The idea of learning love through poetry is in a sense an inversion of the more common conceit by which love is the ul­ti­ m­ate source of poetry (cf. Kayachev 2021), but the underlying assumption is essentially the same: Lydia’s poetry can teach love because it is itself inspired by love. The figure of a bucolic poet who is an instructor in matters of love (as well as poetry) appears to go back to Bion (see intr., section 2); on erotodidaxis, cf. in general DMALL 257–9 s.v. ‘­magisterio de amor’. 112–22  The second stanza switches from the present to the future tense (though note already 111 discetis), but as 115 tumet shows, the future does not have here a strictly temporal value; its modality is not assertive but potential/assumptive (see Pinkster 2015, 426–7), intended to indicate that this is a fantasy. In the speaker’s imagination, Lydia’s inter­action with the ‘fields’ (agri) is subconsciously eroticized (cf. Lorenz 2005, 17): in the clearest way this is brought out in 116–17, where Lydia’s reclining among flowers pointedly anticipates the image of Venus and Adonis’ embrace in 169–70, but see also 107–8, 113 and 114–15 with nn. The combination of two motifs—­envy for (106 = 111 = 123 inuideo uobis, agri) and μακαρισμός of (112 o fortunati . . . ) an inanimate object enjoying intimacy with the speaker’s beloved—­creates a link with Ov. Am. 2.15 (see 104–11n.), which develops a similar fantasy of the ring gifted by Ovid having sexual intercourse with Corinna much more explicitly (on which, besides McKeown 1998, 316–27, see Rivero García 2004; cf. further DMALL 184–5 s.v. ‘fantasías eróticas’). 112–18  Μακαρισμοί expressing envy and/or admiration for someone sharing the company of the person spoken to or about are attested since Homer and appear to be a traditional feature of wedding songs, note esp. Od. 6.158–9 κεῖνος δ᾽ αὖ περὶ κῆρι μακάρτατος ἔξοχον ἄλλων, | ὅς κέ σ᾽ ἐέδνοισι βρίσας οἶκόνδ᾽ ἀγάγηται (Odysseus to Nausicaa; see Hague 1983, 134–8, Winkler 1990, 178–80, Nobili 2006; compare Sappho fr.  31.3–5 πλάσιον ἆδυ ϕωνεί|σας ὐπακούει | καὶ γελαίσας ἰμέροεν with Lydia 107–8); cf. further Apul. Met. 2.7 felix et certo certius beatus, cui

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 112–13  65 permiseris illuc digitum intingere (with Harrison 1990, 197–8). Mel. AP 12.52.3–4 τρὶς μάκαρες νᾶες, τρὶς δ᾽ ὄλβια κύματα πόντου, | τετράκι δ᾽ εὐδαίμων παιδοϕορῶν ἄνεμος (Mel.’s ἐρώμενος is sailing away) seems to be the earliest parallel for applying this kind of μακαρισμός to inanimate figures (cf. in general Gutzwiller 2002/3), and the address to natural forces brings the epigram particularly close to the Lydia. For erotically motivated μακαρισμός more widely, cf. e.g. Theocr. 12.34 ὄλβιος, ὅστις παισὶ ϕιλήματα κεῖνα διαιτᾷ. On μακαρισμός in general, cf. Watson 2003, 87, with further references. 112–13 For a μακαρισμός whose subject is expressed by a relative clause, cf. e.g. Verg. Aen. 1.94–6 o terque quaterque beati, | quis ante ora patrum Troiae sub moenibus altis | contigit oppetere, 437 o fortunati, ­quorum iam moenia surgunt, Hor. Carm. 1.13.17–20 quoted 131–41n. 112  multumque beati (Ω) is problematic on two counts: first, multum after nimium is anticlimactic; second, it seems unparalleled with an adjective meaning ‘happy’. Parrhasius’ nimiumque (in Ramires 2005, 143) is clearly right (cf. Kayachev 2022, 704); for the repetition of nimium, cf. Mart. 8.3.17 scribant ista graues nimium nimiumque seueri, [Tib.] 3.6.21 non uenit ingratis [Heyworth*: iratus codd.] nimium nimiumque seueris, Ov. Her. 1.41 ausus es, o nimium nimiumque oblite tuorum. fortunati . . . beati The accumulation of synonyms for ‘happy’ is idiom­at­ic in μακαρισμοί since Hes. Op. 826–7 εὐδαίμων τε καὶ ὄλβιος ὃς τάδε πάντα | εἰδὼς ἐργάζηται (though note already Il. 3.182 ὦ μάκαρ Ἀτρεΐδη μοιρηγενὲς ὀλβιόδαιμον); cf. further e.g. Theogn. 1013 ἆ μάκαρ εὐδαίμων τε καὶ ὄλβιος, Stat. Silv. 2.7.24 felix heu nimis et beata tellus, and esp. Philemon fr. 93.1–2 Kock ὦ τρισμακάρια πάντα καὶ τρισόλβια | τὰ θηρί᾽, οἷς οὐκ ἔστι περὶ τούτων λόγος, which likewise uses μακαρισμός to express envy for non-­humans. On fortunatus, cf. Dickey 2002, 327: ‘Term of praise or envy’; on beatus, 313: ‘Respectful and distant address, or statement of fact with no polite implications’ (the latter is the case here). Verg. Georg. 2.458–9 o fortunatos nimium, sua si bona norint, | agricolas (and up until 474 Iustitia excedens terris uestigia fecit, echoing 113 uestigia ponet?) may be seen as a polemical allusion, saying as it were that the Lydia speaker should be content to inhabit the agricultural world that is the last remnant of the golden age, rather than complaining

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

66  Commentary ON 112–13 that he did not live then. For combination with nimium, cf. further Luc. 8.843 satis o nimiumque beatus. 113  may be an echo of [Theocr.] 8.47 (quoted 105n.). On a cursory reading, the verb phrase can be taken as an inert periphrasis for ‘walking’, with the preposition construed in a spatial sense (‘within which’); yet the number of words used to refer to Lydia’s feet suggests a more physical interpretation, ‘upon which she treads’ or even ‘sets the soles of her feet’ (cf. below). In view of the increasingly manifest sexual undertones in the following description of Lydia’s interaction with the fields, it appears likely that the speaker’s fantasy should be seen as erotically tinted already here; note that women are often presented as barefoot before they attract the attention of a god, cf. e.g. Ov. Fast. 4.426 errabat nudo per sua prata pede (Persephone), and see further Heyworth 2019, 205–6. For (female) feet as an object of erotic attention in Latin poetry, see Glenn 1980/1, 113–14, cf. DMALL 139 s.v. ‘pies’; for touching with a foot as a sexually charged gesture, cf. e.g. Pl. As. 775 neque illaec ulli pede pedem homini premat, Naev. com. 77 alii peruellit pedem, Ov. Am. 1.4.16 clam mihi tange pedem (see further TLL 10(1).1897.62–6, 1898.68–70); cf. further Sappho fr. 16.17 τᾶς κε βολλοίμαν ἔρατόν τε βᾶμα. pedis niuei uestigia ponet The expression uestigia ponere can function as an idiom for ‘to walk’ or ‘to enter’ (cf. OLD s.v. uestigium 4b, and note Hor. Sat. 2.6.101–2 ponit uterque | in locuplete domo uestigia, Prop. 2.32.48 hic posuit nostra nuper in urbe pedem), but it can also be used in a physically more literal sense, ‘to step (set one’s feet) upon’ (cf. Lucr. 3.3–4 inque tuis nunc | ficta pedum pono pressis uestigia signis, 389–90 sentimus nec priua pedum uestigia quaeque, | corpore quae in nostro culices et cetera ponunt, Ov. Met. 14.49 in quibus [sc. undis] ut solida ponit uestigia terra). There are two reasons for construing the expression here in the more physical way. First, uestigia is here expanded by the singular pedis (cf. Verg. Aen. 7.689–90 uestigia nuda sinistri | instituere pedis, where it is a true singular, and Ov. Met. 5.592 primumque pedis uestigia tinxi, where it could be too) rather than the more usual plural pedum (note e.g. the above Lucr. examples): while metrical and euphonic considerations might have played a role, the singular draws attention to the physicality of the foot as a body part. Second, and more crucially, the

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 113–15  67 colour epithet niuei further emphasizes the bodily reality of Lydia’s feet by indicating their attractiveness (cf. esp. Tib. 1.5.66 uinclaque de niueo detrahet ipse pede, of a lover taking the shoes off his girlfriend’s feet; note also Catull. 61.9–10 huc ueni niueo gerens | luteum pede soccum, addressed to Hymenaeus; for white skin as erotically attractive, cf. in general DMALL 137–8 s.v. ‘color de piel’). We should probably surmise that Lydia is walking barefoot, and that uestigia may be intended to mean ‘soles’ rather than ‘steps’ (cf. esp. Catull. 64.162 candida permulcens liquidis uestigia lymphis, with Trimble 2010, 268–70). 114–15  The fantasy of Lydia plucking unripe grapes evokes the metaphor of grapes for woman as a sexual object (cf. e.g. AP 5.304 ὄμϕαξ οὐκ ἐπένευσας· ὅτ᾽ ἦς σταϕυλή, παρεπέμψω· | μὴ ϕθονέσῃς δοῦναι κἂν βραχὺ τῆς σταϕίδος), in particular that of unripe grapes for a prepubescent girl (see individual nn.); on the metaphor, see in general Cairns 2016, 343–4, and note esp. Alc. (?) fr. 119.15–16 [τά]ρ̣β̣ημι μ̣ ὴ δρόπ̣[ω]σιν αὔταις | [ὄμϕ]ακας ὠμοτέραις ἐοίσαις, which may be an influential early model (for the erotic interpretation, see recently Lentini 1999, though it is uncertain, as is the ascription to Alc.; cf. also Macleod 1979, 97–9). Here the presence of an erotic sous-­entendu is supported, on the one hand, by the lack of other motivations for making Lydia pluck specifically unripe grapes and, on the other, by the fact that Lydia herself had been a virgin (156–7): the speaker imagines Lydia doing to the grapes what he himself did to her. For the kind of activity implicitly suggested, cf. perhaps [Luc.] Am. 53 λάθριος ὑγρῶς ἡ δεξιὰ κατὰ κόλπου δῦσα μαστοὺς βραχὺ τὴν ϕύσιν ὑπεροιδῶντας πιέζει, καὶ σϕριγώσης γαστρὸς ἀμϕιλαϕὲς τοῖς δακτύλοις ἐπιδράττεται ὁμαλῶς, μετὰ τοῦτο καὶ πρωτόχνουν ἄνθος ἥβης (with Krause 2014, 72–5). Milton, Lycidas 3–5 I come to pluck your Berries harsh and crude, | And with forc’d fingers rude, | Shatter your leaves before the mellowing year might suggest that the image could be taken to allude to the speaker’s premature death. From a somewhat different perspective, cf. AP 5.84 (anon.) εἴθε ῥόδον γενόμην ὑποπόρϕυρον, ὄϕρα με χερσὶν | ἀρσαμένη χαρίσῃ στήθεσι χιονέοις. 114  aut (Ω) produces a somewhat illogical and unbalanced sequence of alternatives: Lydia either (1) walks around (113), or (2) plucks grapes (114–15), or (3) lies down and crushes the grass and sings (116–18).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

68  Commentary ON 114 Writing et is an easy change and creates two well-­balanced clauses, in which the first verbs specify Lydia’s position in the fields (113 in quibus: she either treads, or lies down, on the ground), while the second ones describe how she physically affects them (she either plucks grapes or crushes the grass). roseis . . . digitis Beautiful flesh is usually white (cf. 113 niuei with n.), ‘rosy’ colour is as a rule restricted to specific body parts such as mouth (Catull. 63.74 roseis . . . labellis, Verg. Aen. 2.593 roseo . . . ore), cheeks (Aen. 2.606 roseas . . . genas) or breasts (Catull. 55.12 roseis . . . papillis); fingers are rarely defined by a colour epithet, but existing evidence suggests that they may ordinarily be thought of as white too (Prop. 2.1.9 digitis . . . eburnis, if correct; Max. 4.11 niueis digitis); for fingers in erotic descriptions, cf. DMALL 138 s.v. ‘dedos’. Eos is ῥοδοδάκτυλος (and the Moon after her, Sappho fr. 96.8), but otherwise rosy fingers only occur in Nonnus, in three voyeuristic contexts (7.257 νηχομένης πάπταινε ῥοδόχροα δάκτυλα κούρης [sc. Zeus], 11.418 νηχομένων σκοπίαζε ῥοδόχροα δάκτυλα χειρῶν [Calamus gazing at his beloved Carpus], 15.236 ἱμερτῆς ἐδόκευε ῥοδόχροα δάκτυλα κούρης [Hymnus stalking Nicaea]), and thence in Musaeus (114 ῥοδοειδέα δάκτυλα κούρης); in Apul., we encounter roseus used in similarly voyeuristic contexts of a hand (Met. 2.17 glabellum feminal rosea palmula . . . obumbrans, Photis cosplaying Praxiteles’ Aphrodite) or feet (plantisque roseis [sc. Veneris]). Here the epithet may have been motivated primarily by the need to provide colour contrast to 113 niuei and 114 uiridem, but it is erotically charged. Since in this fantasy Lydia takes over the speaker’s active sexual role, and in view of the erotic double entendres in Ov. Am. 2.15.5–6 tam bene conuenias quam mecum conuenit illi, | et digitum iusto commodus orbe teras (cf. 112–22n.), her fingers may not be entirely free from phallic associations (cf. in general Henderson 1991, 114–15). uiridem . . . uuam Plucking (eating) of unripe grapes is a common image, esp. in figurative contexts; cf. e.g. AP 9.375.1–3 τίς ποτ᾽ ἀκηδέστως οἰνοτρόϕον ὄμϕακα Βάκχου | ἀνὴρ ἀμπελίνου κλήματος ἐξέταμεν, | χείλεα δὲ στυϕθεὶς ἀπό μιν βάλεν, Job 15:33 τρυγηθείη δὲ ὥσπερ ὄμϕαξ πρὸ ὥρας. An established usage is in reference to pre­ pubes­cent girls, who nevertheless can already be sexually attractive (cf. 114–15n. and 115n., and note esp. Hor. Carm. 2.5.9–10 tolle cupidinem |

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 114–15  69 immitis uuae, with Nisbet and Hubbard 1978, 84–5, Harrison 2017, 88); here the image may evoke Lydia’s own virginity before her affair with the speaker (156–7); for a bucolic parallel, cf. Theocr. 11.21 ϕιαρωτέρα ὄμϕακος ὠμᾶς, addressed to Galatea by Polyphemus (with Hunter 1999, 231). For uiridis ‘unripe’, see OLD s.v. 3 (contrast Col. 12.44.4 scobe uirides uuas custodiunt, of ‘fresh’ grapes, OLD s.v. 4). decerpserit For the verb as a metaphor for sexual intercourse, esp. defloration, cf. Sen. Contr. 1.2.12 inlibatam uirginitatem decerpunt (with Adams 1982, 196), Asclep. AP 7.217.3–4 νέον ἥβης ἄνθος ἀποδρέψαντες ἐρασταὶ | πρωτοβόλου (with Sens 2011, 284), Nonn. 1.351 ὄμϕακα Κυπριδίων ἐδρέψατο καρπὸν Ἐρώτων (of Zeus deflowering Europa), and note δρόπ̣[ω]σιν at Alc. fr. 119.15 (quoted 114–15n.); cf. further DMALL 54 s.v. ‘frutos de a[mor]’. Van der Graaf (1945, 118–19) assumes that the future perfect is used here indistinguishably from the simple future (esp. 113 ponet), but 117 reclinarit and illiserit clearly imply anteriority to 118 narrabit (Lydia will have lain down and crushed the grass, before she begins confessing her feelings; cf. Dirae 79 peruenerit), as well as as­ ­ pect­ ual difference (note that 117 refers to one-­ time occurrences, whereas 118 implies a protracted action, cf. 122 dum); possibly we should likewise recognize a perfective value in decerpserit, as opposed to continuous ponet (or does it suggest that she will have plucked the grapes before they are ripe?); see in general Pinkster 2015, 466–8. 115  The line adds titillating detail; for the literal sense, cf. Mart. 13.68.2 turget adhuc uiridi cum rudis uua mero; for the sexual implication, cf. Philod. AP 5.124.1–2 οὔπω σοι . . . μελαίνει | βότρυς ὁ παρθενίους πρωτοβολῶν χάριτας (with Höschele 2011, 29–30; Harrison 2017, 85), though the image is different: in Philod. βότρυς stands for pubic hair, here grapes rather evoke breasts, note tumet (cf. Cant. 7:8 οἱ μαστοί σου τοῖς βότρυσιν [sc. ὡμοιώθησαν], further Gerber 1978, 204–5, Clay 2001, 107–8). For the calendric implication, cf. Var. Rus. 1.34.2 uuas autem legere et uindemiam facere inter aequinoctium autumnale et uergiliarum occasum. uitecula (Ω) is unparalleled in verse (and generally quite rare), pro­ d­uces with tumet a harsh metonymy and creates a conspicuous jingle with 116 spectacula (which, though, is my conjecture, see n.); in Kayachev 2022, 715–16 I propose uindemia instead.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

70  Commentary ON 115–17 dulci . . . Baccho: cf. Ciris 229 dulcis pocula Bacchi; for Bacchus used proleptically of grape juice, cf. Ov. Ars 2.316 plenaque purpureo subrubet uua mero, Tr. 5.3.36 incluso plena sit uua mero, 4.6.10 (below). tumet: cf. Ov. Tr. 4.6.9–10 tempus ut extensis tumeat facit uua racemis, | uixque merum capiant grana quod intus habent; but tumere can also be used of different body parts, in particular of breasts, cf. Mart. 8.64.10–11 et talis tumor excitet papillas, | qualis cruda uiro puella seruat. nondum: cf. οὔπω at Philod. AP 5.124.1 (quoted above), Hor. Carm. 2.5.1 nondum. uindemia is a technical term for ‘harvest of grapes’, infrequent in poetry (Verg. Georg. 2.6, 89, 522, Ov. Fast. 4.893, then 9x in Mart.); for the image, cf. Aus. Mos. 195 uindemia turget; for the hexameter ending, cf. Mart. 13.119.1 Nomentana meum tibi dat uindemia Bacchum. 116–17  continue the speaker’s erotic fantasy about Lydia, which is now phrased in language anticipating the image of Venus and Adonis’ embrace in 169–71 (see n.): in both cases grass and flowers form the bed, and in both cases the reclining figure is said to ‘crush’ them (Zaboulis 1978, 211 points out numerical symmetry in the position of the two ­passages: 116–18 are preceded by twelve lines, 169–71 are followed by twelve lines). While in general Lydia’s pose is erotically charged, the action of crushing grass (or flowers) may specifically evoke sexual intercourse. The metaphor of the young girl (and her body) as a flower is ubiquitous and can take different forms: one is e.g. the picking of flowers as a symbol of defloration in rape narratives, another is (as implicitly here) the locus amoenus as a figure for the girl’s body and spe­cif­ic­al­ly genitals (cf. in general e.g. Swift 2016, discussing in particular how in Arch.’s Cologne epode fr. 196A.42–4 West παρθένον δ᾽ ἐν ἄνθε[σιν | τηλ]εθάεσσι λαβών | ἔκλινα, an actual account of an intercourse with a girl, is anticipated by 23–4 π̣ο̣η[ϕόρους | κ]ή̣ πους, a metaphoric reference to the girl’s genitals); cf. in general DMALL 190 s.v. ‘f[lores] como símbolo de virginidad de la amada’. At the same time, the suggestion is present that Lydia is not merely an object of desire but plays an active role herself (cf. Deacy 2013, 399–400: ‘there is something limiting in an interpretation of flower-­ picking mythological females that requires them solely to be seen as victims . . . the flower picking in these instances

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 116–17  71 is being done not by male intruders but by the girls themselves . . . mytho­ logic­al girls who pick flowers are performers of sexualised acts’). While illidere is unparalleled as a sexual double entendre, various verbs denoting physical pressure are used figuratively in reference to the active role (Adams 1982, 152–4, 182–3). Ovid’s account of Venus and Adonis (Met. 10.557–9 pressitque et gramen et ipsum | inque sinu iuuenis posita ceruice reclinis | sic ait) may be intended to bring this out: while on a cursory reading the passage merely appears to refer to Venus’ lying down and then telling a story, 557 pressit (see Adams 1982, 182, cf. e.g. [Sen.] Oct. 764 quem [sc. Iouem] modo Ledae pressisse sinum) seems in fact to imply sex (Venus on top), with 558–9 referring to post-­coital chatting. As in the case of Lydia’s plucking grapes (see 114–15n.), the speaker subconsciously projects on to her the actions he himself performed on her. Cf. in general the destruction of Daphnis’ garden by his rival Lampis in Longus 4.7, with Zeitlin 1990, 447: ‘Given the prenuptial context, Lampis’ violence against the flowers might suggest the impending loss of Chloe’s virginity . . . it is more than implied that the tender garden represents not only the maiden Chloe but also (and even more overtly) the beautiful Daphnis.’ 116  stipendia [SL: spumantia M] is usually printed, but it is proble­ matic on a number of counts: as a technical term, it is largely avoided in poetry; it is rarely used in a figurative sense, and it is not clear exactly what metaphor should be involved here (‘tax levied in Venus’ interest’?); it does not seem to be normally construed with a genitive, either sub­ject­ive or objective. In Kayachev 2022, 714–15 I propose spectacula, cf. below; as an alternative, one might consider e.g. munuscula, cf. Verg. Ecl. 4.18–19 nullo munuscula cultu | errantes hederas passim cum baccare tellus [sc. fundet], Stat. Theb. 8.300 innumerosque deae [sc. Telluri], sua munera, flores, AE 1947.31.6 flores ama Veneris, Cereris bona munera carpe. uarios . . . flores is predictably a common expression, cf. e.g. Verg. Aen. 6.707–8 apes aestate serena | floribus insidunt uariis, Tib. 1.7.45 uarii flores, Ov. Am. 3.5.9 uariis immixtas floribus herbas, Met. 10.123 texebas uarios per cornua flores; cf. further ἄνθεα παντοῖα at Theocr. 15.116, ἄνθεα ποικίλα at Ap. Arg. 4.1144–5, Call. Hymn. 2.82–3 (for the

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

72  Commentary on 116 image in general, cf. also adesp. PMG 8a ἔνθα δὴ ποικίλων ἀνθέων ἄμβροτοι λείμακες | βαθύσκιον παρ᾽ ἄλσος ἁβροπαρθένους | εὐιώτας χοροὺς ἀγκάλαις δέχονται). For the word-­order pattern (a noun phrase split by an apposition), cf. esp. Col. 10.96 pangite tunc uarios, terrestria sidera, flores, Stat. Theb. 8.300 quoted above, but also (if less closely parallel) Hes. Th. 576–7 ἀμϕὶ δέ οἱ στεϕάνους νεοθηλέας, ἄνθεα ποίης, | ἱμερτοὺς περίθηκε καρήατι Παλλὰς Ἀθήνη (cf. further Cypria fr. 6.2 quoted below); this pattern is known as the schema Cornelianum (after Skutsch 1956, 198–9; see in general Solodow 1986, cf. Dainotti 2015, 102 n. 330, Gagliardi 2017), reflecting the theory that it was introduced by Gallus, and is frequent in Verg. Ecl. (cf. Nisbet 1991, 8: ‘no extant work favours it so much as the Eclogues, but it must be regarded as a neoteric rather than a bucolic feature’), then in Hor., Prop., and Ov., though note already Enn. Ann. dub. 6 caelicolae, mea membra, dei (for the fragment’s authenticity, see Traina 1995, cf. Farrell 2020, 87–8). Interestingly, inter flores appears in Repos. 103–4 ille [Mars] inter flores furtiuo lumine tectus | spectat hians Venerem, but is unparalleled in the poetry of the classical period. Veneris spectacula For Venus’ ‘flower show’, cf. Dirae 20 haec Veneris uario florentia prata decore; for the association of flowers with Venus, cf. in general DMALL 189 s.v. ‘f[lores] como atributo de Venus’. The underlying image is that of Venus’ pageant in Lucr. 1.17–18 tibi suaues daedala tellus | summittit flores, 5.737–40 it Ver et Venus, et Veneris praenuntius ante | pennatus graditur, Zephyri uestigia propter | Flora quibus mater praespargens ante uiai | cuncta coloribus egregiis et odoribus opplet; for a related image, cf. Cypria fr. 5.2 where Aphrodite appears in a dress ‘dyed’ ἐν ἄνθεσιν εἰαρινοῖσιν (on which cf. Davies 2019, 63: ‘Aphrodite rather resembled Flora in Botticelli’s Primavera’), as well as fr. 6.2 where she is accompanied by nymphs wearing στεϕάνους εὐώδεας ἄνθεα ποίης (note the appositive construction); for the inserted apposition marking association with a deity, cf. Stat. Theb. 10.254 adoratas, Phoebea insignia, frondes. The noun has a down-­to-­earth flavour, and, outside Ovid, is infrequent in the poetry of the first cent. bc; for its use of a natural phenomenon, cf. Hor. Sat. 2.2.26 picta pandat spectacula cauda (a peacock); for its use in apposition, cf. Ov. Ars 3.351 artifices lateris, scaenae spectacula, amantur, Paul. Nol. 10.176 spernentes uarias, rerum spectacula, formas.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 117–18  73 117 membra reclinarit  The verb is generally rare, possibly somewhat lower-­register; for the sense (‘(to cause) to lie, lean back’), cf. esp. Hor. Epod. 17.24 nullum a labore me reclinat otium, Carm. 2.3.6–8 te in remoto gramine per dies | festos reclinatum bearis | interiore nota Falerni; Ov. Met. 10.558 reclinis may be intended to echo it (cf. 116–17n.); here the verb may also evoke the more frequent use of κλίνειν in this kind of context, cf. e.g. Arch. fr. 196A.42–4 West παρθένον . . . ἔκλινα (cf. 116–17n.), Eur. Hip. 210–11 ὑπό τ᾽ αἰγείροις ἔν τε κομήτῃ | λειμῶνι κλιθεῖσ᾽ ἀναπαυσαίμαν, Opp. Cyn. 2.198–9 οὔτ᾽ ἄρα κεκλιμένοι χθαμαλοῖσιν ἐπ᾽ ἄνθεσι ποίης | θηλυτέραις ἐλάϕοισιν ὁμιλαδὸν εὐνάζονται. With verbs of motion, membra can come close in meaning to the reflexive pronoun (TLL 8.640.77–641.17; cf. e.g. Moret. 5 membra leuat uili sensim demissa grabato), though it also highlights the physicality of Lydia’s body (cf. 125 membra with n.). teneramque illiserit herbam: cf. 169–70 tenera . . . illidere in herba | . . . flores. The verb is rare in this sense (‘To injure by crushing’, OLD s.v. illido 1) and is not applied to plants elsewhere, though cf. Verg. Aen. 7.809 nec teneras cursu laesisset aristas; see further 169–71n. First attested at Cato Agr. 151.4 quam tenerrimis herbis, the noun phrase is common in poetry, from Lucr. 1.259–61 noua proles . . . teneras lasciua per herbas | ludit on, possibly after Od. 9.449 τέρεν᾽ ἄνθεα ποίης, Ap. Arg. 1.1143 (= 3.898) τερείνης ἄνθεα ποίης (tener and τέρην are related). 118  echoes 106 nostrum tacite suspirat amorem. secreta, ‘all by herself ’: cf. Verg. Aen. 4.494–5 tu secreta pyram tecto interiore sub auras | erige; for the image, cf. perhaps Catull. 62.39 ut flos in saeptis secretus nascitur hortis; here it also anticipates 132 secreta cubilia (see n.). furtim refers to the intent (‘stealthily’), secreta to the physical circumstance (‘alone’); the adv. is frequent in accounts of clandestine affairs (since Catull. 66.5–6 ut Triuiam furtim sub Latmia saxa relegans | dulcis amor gyro deuocet aerio, cf. Ciris 209 cum furtim tacito descendens Scylla cubili; λάθρῃ has a near-­technical character already in Hom., e.g. Il. 2.515 ὃ δέ οἱ παρελέξατο λάθρῃ [sc. Ares with Astyoche], 16.184–5 παρελέξατο λάθρῃ | Ἑρμείας [with Polymele], Od. 8.269–70 ἐμίγησαν ἐν Ἡϕαίστοιο δόμοισι | λάθρῃ [Ares and Aphrodite], Hymn. 17.3–4 Λήδη | λάθρῃ

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

74  Commentary ON 118–22 ὑποδμηθεῖσα κελαινεϕέϊ Κρονίωνι, cf. further e.g. Parth. 2.2 Πολυμήλη γὰρ τῶν Αἰολίδων τις ἐρασθεῖσα αὐτοῦ [sc. Odysseus] κρύϕα συνῆν, 13.1 κατεργασάμενος τὴν κόρην λαθραίως αὐτῇ συνῆλθεν [sc. Clymenus with his daughter Harpalyce], 35.1 ἠράσθη Λύκαστος . . . Εὐλιμένης . . . ταύτῃ κρύϕα συνὼν ἐλελήθει; for the notion of ‘clandestine love’, cf. 165 occulti . . . amoris with n.), esp. in elegy (e.g. Prop. 2.9.42 et furtim misero ianua aperta mihi, Tib. 1.2.10 neu furtim uerso cardine aperta sones); here it contributes to presenting Lydia’s interaction with the fields—­ conceived of as a substitute for the speaker—­as an erotically charged activity. meos . . . narrabit amores, ‘confesses her love for me’: the sense could be ‘tells about my love for her’ (cf. e.g. Verg. Ecl. 10.34 uestra meos olim si fistula dicat amores), but 106 nostrum suggests that here too meos should be construed objectively; for the verb, cf. Ter. Andr. 312 amorem huic narrabo meum. 119–22  The passage blends together three distinct motifs: that of nature mourning over a fallen hero (pathetic fallacy), the epiphanic motif of nature exulting at the presence of a deity, and the Orphic one of nature being enchanted by music (cf. 119n.). In structuralist terms, pathetic fallacy is the primary framework of reference, which projects this passage against the descriptions of nature mourning over the deaths of Daphnis (Theocr. 1), Adonis (Bion, Ep. Adon.), and Bion (Ep. Bion.); see intr., section 2. The focus on Lydia (rather than on the dying but absent speaker) finds a parallel of a sort in the focus on Aphrodite’s suffering in Bion (33 καὶ ποταμοὶ κλαίοντι τὰ πένθεα τᾶς Ἀϕροδίτας, contrast Ep. Bion. 2 καὶ ποταμοὶ κλαίοιτε τὸν ἱμερόεντα Βίωνα). Orphic elements are introduced in the Lament for Bion, which dubs its late hero Δώριος Ὀρϕεύς (18); by contrast, here it is Lydia, rather than the dying speaker, who is an Orphic figure, and while the Lament proclaimed the death of bucolic poetry with Bion (11–12), Lydia continues to sing and enchant nature. 119 gaudebunt  The joy of nature seems originally to be an epiphanic motif (cognate with pathetic fallacy?), cf. Hymn. Hom. 3.118 μείδησε δὲ γαῖ᾽ ὑπένερθεν (at Apollo’s birth), Theogn. 9–10 ἐγέλασσε δὲ γαῖα πελώρη, | γήθησεν δὲ βαθὺς πόντος (ditto), Lucr. 1.8 tibi [sc. Veneri] rident aequora ponti, Ps. 95:12–13 gaudebunt campi et omnia quae in eis

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 119–20  75 sunt, tunc exsultabunt omnia ligna siluarum a facie Domini quia uenit, Paul. Nol. 14.45–7 omnia gaudent | terrarum et caeli, ridere uidetur apertis | aethra polis (on Christmas). The sky ‘smiles’ at the Muses’ singing in Hes. Th. 40–2 γελᾷ δέ τε δώματα πατρὸς | Ζηνὸς ἐριγδούποιο θεᾶν ὀπὶ λειριοέσσῃ | σκιδναμένῃ, ἠχεῖ δὲ κάρη νιϕόεντος Ὀλύμπου, on which cf. West 1966, 170: ‘The metaphor is also found in the Ṛgveda, and may therefore be Indo-­European heritage’ (though note also the biblical parallel above). Probably unrelated in origin is the idea that animals react with joy to Orpheus’ singing (Aesch. Ag. 1630 ἦγε πάντ᾽ ἀπὸ ϕθογγῆς χαρᾷ); this is not a metaphor, but Virgil conflates the two conceits (Ecl. 6.29–30 nec tantum Phoebo gaudet Parnasia rupes, | nec tantum Rhodope miratur et Ismarus Orphea). A further development of the motif appears in Man. 2.141–2 sed caelo noscenda canam, mirantibus astris | et gau­ dente sui mundo per carmina uatis. In our context the Orphic connotations prevail, but the original use may be worth keeping in mind. A third potential connection is suggested by the repetition of gaudebunt, which evokes repeated salutations (cf. e.g. Theocr. 2.165–6 χαῖρε, Σελαναία λιπαρόθρονε, χαίρετε δ᾽ ἄλλοι | ἀστέρες, 18.49 χαίροις, ὦ νύμϕα· χαίροις, εὐπένθερε γαμβρέ, Ciris 197 gaudete, uagae, gaudete, uolucres) and here may be intended to allude to Daphnis’ farewells to the animals and ­natural features in Theocr. 1.115–18 (see intr., section 2). For the ‘nature rejoices’ topos, however, variation rather than repetition seems to be the rule (cf. the examples above), and one might wonder if the second gau­ debunt is not a persistence error ousting, say, ridebunt (cf. perhaps Mart. 10.51.3 ridet ager, AL 83.79 rident prata rosis). mollia prata With 120 gelidi fontes, the phrase creates a strong link with Verg. Ecl. 10.42 hic gelidi fontes, hic mollia prata, Lycori (see intr., section 2); the exact collocation is otherwise unparalleled, though cf. Georg. 2.384 mollibus in pratis, 3.520–1 non mollia possunt | prata mouere animum, Prop. 3.3.18 mollia sunt paruis prata terenda rotis, Ov. Ars 1.279 mollibus in pratis admugit femina tauro. 121–120  The transmitted order of lines may conceivably be right, but the inverted order (introduced by Ribbeck 1868, 174) has several advantages. First, rivers and springs form a conceptual pair (note, in a similar context, [Orph.] Arg. 1009 πηγάς τ᾽ ἀενάων ὑδάτων ποταμῶν τε ῥέεθρα

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

76  Commentary ON 121–120 [sc. κοιμίσσας]), and it may be preferable to group them together (rather than meadows and springs, as transmitted). Second, the notion of silence is more effective immediately before the reference to Lydia’s singing. Third, the change of word order from ‘verb-­subject’ (standard in this kind of context, cf. the parallels quoted 119n., both Latin and Greek) to (in sense) ‘subject-­verb’ (auiumque silentia fient ~ auesque silebunt) puts emphasis on the subject (~ ‘and even the birds will be silent’ rather than ‘and there will be silence among the birds’), which again recommends a climactic position for the clause. 121  For the Orphic topos of song enchanting rivers, cf. esp. Ap. Arg. 1.27 θέλξαι ἀοιδάων ἐνοπῇ ποταμῶν τε ῥέεθρα, Verg. Ecl. 4.8 mutata suos requierunt flumina cursus (with Clausen 1994, 241–2), Prop. 3.2.3–4 Orphea detinuisse feras et concita dicunt | flumina Threicia sustinuisse lyra, Hor. Carm. 1.12.9–10 rapidos morantem | fluminum lapsus (with Nisbet and Hubbard 1970, 148–9). labentes currite lymphae (Ω) is impossible on two counts: first, it gives a sense that is the opposite of what one might expect (taking labentes in the pregnant sense of ‘flowing slowly’ is unwarranted); second, a parenthetic apostrophe is out of place in this fantasy phrased in the future tense. As I discuss in Kayachev 2022, 716–17, most corrections accept humanist currere (ed. Ald. 1517), and I myself propose lymphas there, to be construed as an acc. and inf. with tardabunt (for the construction, see OLD s.v. tardo 1b, and cf. TrRF adesp. 1.74.6 illum ut maeror tardaret sequi). Still, the construction is heavy, and the word order implausible. While standing by lymphas, I now think it more likely that currite is an error for the ablative of a noun. Ellis’s (1907, 5-­uv.) gurgite is palaeographically the easiest option (for the corruption, cf. currens for surgens [Mähly] at 143), and conceivably correct. It would be taken in a modal sense, ‘(flowing) in a torrent’, almost ‘violently’, for which the closest parallel is Ov. Am. 3.6.8 turpi crassas gurgite uoluis aquas (‘you roll your muddy waters in a foul torrent’, of a flooded river), though the construction is easier because of the epithet; for gurges ‘turbulent flow’ (rather than ‘waters’ more generally), see TLL 6(2).2362.19–23. A conceivable alternative could be limite (Heyworth*), though it too would be easier with an epithet. Hesitantly, I opt for murmure, on the basis of Ov. Rem.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 121  77 177 aspice labentes iucundo murmure riuos (note that Rem. 179 appears to echo Lydia 135, see n.); for murmure with labi without a preposition, cf. further Ibis 76–7 quique per infernas horrendo murmure ualles | imper­iuratae laberis amnis aquae. I think murmure is somewhat easier to construe than gurgite, and it is more pointed in the context (the rivers and springs will suppress their noise, so as to listen to Lydia’s singing). tardabunt In the usual idiom the subject of tardare is some external obstacle, cf. e.g. Var. Ruf. fr. 150.5 Hollis non amnes illam medii, non ardua tardant, Prop. 4.3.47 nec me tardarent Scythiae iuga, Ov. Her. 10.20 alta puellares tardat harena pedes; in reference to a liquid, cf. Cels. 2.19.2 aliae [sc. res] citant urinam, aliae tardant. A rarer usage is in ­reference to agents holding back their own actions, cf. Prop. 1.20.42 errorem blandis tardat imaginibus [sc. Hylas], Sil. 3.500 miles dubio tardat uestigia gressu, Plin. Nat. 2.64 non quia adcelerent tardentue naturales motus [sc. stellae]. Our context can be seen as an extension of the latter usage; the lack of close parallels is unsurprising, given the artificiality of the conceit by which rivers and their waters are presented as distinct (if related) entities (cf. the similarly unusual usage at Verg. Ecl. 4.8 quoted above). The verb is particularly evocative, as the adj. tardus can be applied to stagnant waters, cf. Var. At. fr. 121.1 Hollis pelagi uolucres tardaeque paludis, Verg. Georg. 4.479 tardaque palus inamabilis unda, AE 1953.98.2 foeda palus tardaque lympha fuit. riui . . . lymphas For the ‘disjunctive’ language differentiating between rivers as active agents and their waters as a substance over which they exercise control, cf. esp. Dirae 67 flectite currentes lymphas, uaga flumina, retro; cf. further e.g. Theocr. 1.118 ποταμοὶ τοὶ χεῖτε καλὸν κατὰ Θύβριδος ὕδωρ, Prop. 2.15.33 fluminaque ad caput incipient reuocare liquo­res, Ov. Am. 3.6.1–2 amnis . . . siste parumper aquas (Hor. Carm. 2.3.11–12 obliquo laborat | lympha fugax trepidare riuo is different, riuo ‘irrigation channel’ being locative, cf. Ep. 1.10.21 quae [sc. aqua] per pronum trepidat cum murmure riuum, pace Nisbet and Hubbard 1978, 61, as well as Coleman 1999, 76 n. 74). lymphas is usually taken as simply a poetic term for ‘water’ (OLD s.v. lympha 2), but it is more complex. In Virgil, lympha refers exclusively to water used for ablution, ritual or medicinal (Aen. 1.701, 4.635, 683, 9.23, 10.834, 12.420); in pre-­Augustan poetry, one context refers to ritual

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

78  Commentary ON 121 aspersion (Ciris 147), two refer to water used to wash feet (Pacuv. trag. 244, Catull. 64.162), and three to potable water, in at least two cases added to wine (Catull. 27.5, Varro Men. 442.1, Lucr. 1.496); of the remaining three contexts, one refers to well water (Lucr. 6.1174), p ­ erhaps implying its potability, another to the mineral springs of Thermopylae (Catull. 68.54), perhaps because they were used for bathing (cf. Pisander fr. 7.2 θερμὰ λοετρά), and only the last denotes water as a natural substance tout court (Var. Men. 50.1), but Varro is a special case (see below). Rather than being an artificial poeticism, in poetry lympha appears in fact to reflect some specialized subliterary usages. Cael. Aur. Chron. 4.3.68 uino lymphato (perhaps = lora, a type of lower-­grade wine produced from water mixed with grape pomace), though a late piece of evidence (the usage lives on in mediaeval Latin, cf. e.g. Thom. Aquin. Sum. Theol. 3a.66.4 uinum lymphatum magis est uinum quam aqua) may be relevant in two connections. On the one hand, it may attest to the use of lympha in medical discourse, a usage reflected in Virgil and possibly underlying the term’s frequency in Seren. (19x). On the other, it parallels Catullus’ use of lympha for water added to wine, which is also prominent in subsequent poetry (Hor. Carm. 2.11.20, Tib. 2.1.46, 3.6.58); Hor. Carm. 1.37.14 mentemque lymphatam Mareotico (‘the mind diluted by wine’) no doubt likewise playfully reflects this colloquial/technical idiom (cf. Nisbet and Hubbard 1970, 415). A further relevant piece of evidence may be provided by the dipinto LYMP (vel sim.), common on amphorae, which has recently been argued to stand for (uinum) lymphatum (see García Vargas 2004, Silvino and Poux 2005). These findings conform well with the etymological connection (proposed by Solta 1967, esp. 95–6) of lympha (limpa) with limpor and limpidus, understood as Italic doublets of Latin liquor and liquidus (cf. EDL s.v. limpidus): its etymological sense of ‘clear (or running?) water’ would account for its usage, and its non-­Latin origin might explain its specialized subliterary currency (more recent etymological attempts by Hyllested 2004 and Prósper 2015, 42–8 do not convince). Varro is the first prose writer to use lympha, proposing two alternative etymologies: L. 5.71 ab aquae lapsu lubrico lympha, in a context discussing the lacus Iuturnae, and 7.87 lymphata dicta a lympha, a nympha (Deschamps 1983

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 121–120  79 attractively suggests that the latter etymology was only intended to explain lymphatus as a calque of νυμϕόληπτος). Varro’s former context may give some support to Solta’s etymology (pace de Melo 2019, 709 and 1002, who follows Wackernagel 1908, 218–21 in accepting Varro’s Greek etymology) and could attest to a religious currency of lympha, but most importantly it can account for the innovative use of lympha here, as well as in Augustan poets (other than Virgil), esp. Horace. Varro’s etymology ab aquae lapsu is no doubt implied here in labentes (cf. 108 Lydia ludit, alluding to another etymology attested in Varro). The Dirae picks up this usage with 67 currentes lymphas (and further extends it at 48 by using lymphis of sea water). Five out of ten contexts in Hor. use lympha of running water, esp. that of a spring: Epod. 2.27 fontesque lymphis obstrepunt manantibus, 16.48 leuis crepante lympha desilit pede, Carm. 2.3.11–12 obliquo laborat | lympha fugax trepidare riuo, 11.20 praeter­ eunte lympha, 3.13.16 lymphae desiliunt. 120 gelidi fontes  Like 119 mollia prata, the exact collocation is only paralleled at Verg. Ecl. 10.42 (see 119n.), until Max. 5.5 hic gelidi fontes, hic dulces arboris umbrae; the combination of the noun and adjective is of course more common, cf. e.g. Ov. Met. 4.90, 14.786, 15.550. auiumque For birds enchanted by Orpheus’ singing, cf. Sim. PMG 567.1–2 τοῦ καὶ ἀπειρέσιοι | πωτῶντ᾽ ὄρνιθες ὑπὲρ κεϕαλᾶς, [Orph.] Arg. 438–9 οἰωνοί τ᾽ ἐκυκλοῦντο βοαύλια Κενταύροιο | ταρσοῖς κεκμηῶσιν, ἑῆς δ᾽ ἐλάθοντο καλιῆς, Mart. Sp. 21.6 supra uatem multa pependit auis, Sil. 11.467–8 immemor et dulcis nidi positoque uolatu | non mota uolucris captiua pependit in aethra, Sen. Her. Oet. 1046–7 auditis uaga cantibus | ales deficiens cadit. silentia fient is a (slightly adapted) prosaic idiom, esp. at home in contexts narrating public occasions, cf. Cic. Div. 1.59 silentium fieri iussisse (so as not to wake Cic. up), Liv. 1.13.4 silentium et repentina fit quies (reaction to the conciliatory speech of the Sabine women), 26.16.3 silentium fieri Flaccum iussisse (ordering quiet at a public execution), Sen. Con. 9.2.10 fit a praecone silentium, Plin. Nat. 32.51 inlatis in populum [sc. rubetis] silentium fieri (toads magically make people speechless); it suggests a crowd suddenly falling silent, thus signalling that Lydia’s performance is taking place in a formal setting, as it were.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

80  Commentary ON 122 122  exponat (Ω) has two weak points. First, exponere denotes ex­pos­i­ tion of content rather than verbal performance as such (‘she formulates her complaint’; cf. OLD s.v. expono 6); expromat (Heinsius in Burman 1773, 669; cf. Giardina 2009, 170, noting that forms of expromere are often corrupted to those of exponere) is the verb we need (‘she utters her complaint’; cf. esp. Catull. 64.223 multas expromam mente querelas, Verg. Aen. 2.280 maestas expromere uoces, note also its use of poetic production at Catull. 65.3, Catal. 9.7). Second, subjunctive with dum can denote intended action (OLD s.v. 1d ‘for as long as is needed (for)’): while this produces a possible sense (‘nature will keep quiet, so that she could utter her complaint’; cf. Cic. Att. 5.16.1 subsedi in ipsa uia, dum haec . . . tibi perscriberem), it inverts the expected logical relation of the ideas (‘nature will keep quiet while (and because) . . . ’); we need future indicative, expromet (Heyworth*, exponet already Scaliger 1572, 310). mea . . . cura: ‘Term of affection and/or praise, depending on modi­ fiers, usually with mea or another possessive’ (Dickey 2002, 318); as with other such terms (see Dickey 2002, 152 for a list, and cf. 123 mea gaudia with n.), cura is overwhelmingly used in a predicative role (see TLL 4.1475.42–60); the only comparable case of referential use is Prop. 2.34.9 Lynceu, tune meam potuisti, perfide, curam | tangere? (Ciris 75 is corrupt: Kayachev 2020, 101). One might speculate that such usage could be facilitated through the association of cura with κούρη, first securely attested in Verg. Ecl. 10.22 tua cura Lycoris ‘rendering’ Theocr. 1.82 ἁ δέ τυ κώρα, though the wordplay has been suggested to go back to Gallus, see Ross 1975, 68–9, cf. Cairns 2006, 115–16. For the possibility that Ticida fr. 103 Hollis Lydia doctorum maxima cura liber echoes the present line, see intr., section 1. iucundas . . . querelas The combination may appear oxymoronic, but querela comes to have a generic (rather than a strictly topical) sense (cf. OLD s.v. 2): Lydia does not so much pour out her grievances as perform a ‘complaint’ that is pleasant to listen to (note its quasi-­technical usage in reference to animal sounds, e.g. Cic. Prog. fr. 4.6 assiduas iacit ore querelas [sc. a bird]); cf. Lucr. 4.584–5 dulcesque querelas, | tibia quas fundit, [Tib.] 3.4.75 ne dubita blandas adhibere querelas, Nemes. Ecl. 2.15 ardentes flammati pectoris aestus | carminibus dulcique parant releuare querela, 4.13 dulces cantu duxere querelas. Still, querela properly

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 122–130  81 belongs in a woman’s mouth, cf. esp. Catull. 66.19 id mea me multis docuit regina querelis (of Berenice missing her husband gone to war), Ciris 174 tristes uoluens in nocte querelas (of Scylla longing for Minos); it is used twice in reference to the speech of abandoned Ariadne (Catull. 64.130, 195), and in Verg. ‘complaining’ is denounced as undignifiedly feminine: Aen. 4.360 desine meque tuis incendere teque querelis (Aeneas to Dido), 10.94–5 nunc sera querelis | haud iustis assurgis (Juno to Venus); in Lucr. 4.1182 et meditata diu cadat alte sumpta querela it refers, satirically, to the speech of a rejected male lover—­a usage that becomes regular in elegy, cf. e.g. Prop. 1.18.29–30 et quodcumque meae possunt narrare querelae, | cogor ad argutas dicere solus aues, Tib. 1.2.9 ianua, iam pateas uni mihi, uicta querelis (cf. DMALL 349–51 s.v. ‘quejas de amor’). The adj. iucundus is far more complex and socially charged than it might appear at first glance (see Krostenko 1993, 40–120), which is probably why Verg. only uses it once (Aen. 6.363 caeli iucundum lumen, echoing a funerary formula, cf. Catull. 68.93 fratri iucundum lumen ademptum, CLE 963.1 o iucundum lumen superum), and Hor. avoids it in Carm., as does Ov. in Met.; here it appears to be used in a traditional way in reference to a performance (‘pleasing to the audience’), cf. esp. Pl. Poen. 206 si uis uidere ludos iucundissimos, Nov. Com. 55 iucundum indaudiui melum, further Prop. 1.2.29 unica nec desit iucundis gratia uerbis (of Cynthia as an accomplished poetess), Ov. Rem. 177 aspice labentes iucundo murmure riuos (alluding to the present context, along with 121: here the rivers stop to listen to Lydia’s pleasant singing, in Ov. the reader is invited to enjoy the rivers’ pleasant noise). 123–30  While the first two stanzas focus exclusively on the inter­action between Lydia and the fields, the present stanza widens the perspective in a number of ways, so as to set the stage for developments in the rest of the poem. Here the speaker for the first time becomes the centre of attention, when he claims to be mortally ill with lovesickness (125–6): his sufferings and impending death will be repeatedly evoked in the poem’s second (141, 145) and third (158, 160, 182–3) sections. His ­sep­ar­ation from Lydia, although implied in the preceding stanzas, is first explicitly mentioned at 127: separation of lovers will be the topic of the whole second section (131–55). Here the speaker also makes his first

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

82  Commentary ON 123–30 descriptive (rather than narrative) statement about Lydia, whom he asserts to be nonpareil and worthy of Jupiter himself: the allusion to Jupiter’s affairs with Europa and Danae (128–30) introduces the mytho­ logic­al topic of the gods’ erotic relationships that will be prominent throughout the rest of the poem (and Jupiter will appear again at 166–8). 123–4 Both gaudium (OLD s.v. 3) and uoluptas (OLD s.v. 2) can be applied to persons, esp. as terms of endearment (cf. Dickey 2002, 152 and passim); this may be the primary sense here, but the basic lexical meaning is also operative: not just ‘you have (the girl who was) my delight’ but also ‘you have (experience) the pleasures I once had’ (cf. 124n.). This second meaning is supported by the pointed parallelism with 162–3, where gaudium and uoluptas specifically refer to sex (see nn.); and the sexual innuendo may be present here as well, esp. in view of 114–17, where the speaker appears to be fantasizing about Lydia having intercourse with the fields (see 114–15n., 116–17n.); the idea of ­sexual jealousy is also suggested in 129–30. 123 mea gaudia habetis  For this term of endearment used referentially, cf. Ov. Her. 19.41 exisse domo mea gaudia; cf. also 122 mea . . . cura with n. In the light of 123–4n., note that habere can be used as a sexual ­eu­phem­ism (see Adams 1982, 187), though a more relevant parallel for gaudia habere can be found in [Tib.] 3.13.5–6 mea gaudia narret, | dicetur si quis non habuisse sua, where talking about ‘delights’ is predicated on not having them—­likewise the Lydia speaker reminisces about his beloved when he has already lost her. 124  Note how 124 slightly modifies the sense of 123: at 123 it is easy to take mea gaudia as a term of endearment, esp. given 122 mea . . . cura, but the syntax of 124 restores to uoluptas and, retrospectively, gaudia their basic lexical meanings: ‘the pleasure that once was mine now belongs to you’. 125–6  Lovesickness is often presented as a physical disease, potentially lethal, which can be manifested by a variety of symptoms (see in general e.g. Caston 2006, Pachoumi 2012, Thumiger 2018, DMALL 259–62 s.v. ‘mal de amores’); for feeling to be dying of love, cf. esp. Sappho fr. 31.15–16 τεθνάκην δ᾽ ὀλίγω ᾽πιδεύης | ϕαίνομ᾽ ἔμ᾽ αὔτᾳ, and in general DMALL

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 125–6  83 289 s.v. ‘morir de amor’, citing aptly Pease 1935, 285 on Verg. Aen. 4.308 nec moritura tenet crudeli funere Dido: ‘Whether Dido . . . means more than hyperbole may be queried, but Virgil interestingly shows how the thought, once introduced, develops and comes to possess her mind’— likewise here we are left to wonder as to how literally we should take the speaker’s claim to be dying; for experiencing chill (as opposed to fever, which is more common), cf. Theocr. 2.106 πᾶσα μὲν ἐψύχθην χιόνος πλέον, Call. Aet. fr. 75.19 Κυδίππην ὀλοὸς κρυμὸς ἐσῳκίσατο (for Cydippe’s symptoms as evoking lovesickness, see Rynearson 2009), Longus 2.7.5 τὸ σῶμα ἐψυχόμην, perhaps Lucr. 4.1060 successit frigida cura, Ov. Am. 3.14.38–9 mens abit et morior quotiens peccasse fateris, | perque meos artus frigida gutta fluit. 125  male (Ω) tabescunt is in itself unobjectionable (and even at­tract­ive, in view of the curse tablet AE 1982.448 [persequa]ris eum, ut male ­contabescat usque dum morie[t]ur); but 125–6 lack an explicit reference to the first-­person speaker, and we do need a contrast to 124 uobis; mihi (ed. Ald. 1517) is an easy enough correction. tabescunt . . . dolore Like its cognate τήκεσθαι, tabescere has two largely distinct usages: it can either refer to the body affected by an illness, or to the spirit afflicted by some distress (for the latter, cf. e.g. Ter. Ad. 602–3 dolore ac miseria | tabescit, Cic. Att. 2.21.4 tabescat dolore); examples in which it is used of the body affected by an emotion are rare and predominantly poetic: note perhaps (somewhat ambiguous) Catull. 68.55 assiduo tabescere lumina fletu, based (directly or indirectly) on Od. 19.208 τῆς τήκετο καλὰ παρήια δάκρυ χεούσης; in reference to love, the sense likewise is as a rule psychological (e.g. Prop. 1.15.20 Haemonio tabuit hospitio, of Hypsipyle falling in love with Jason, cf. Ap. Arg. 3.1019–20 ἰαίνετο δὲ ϕρένας εἴσω | τηκομένη, of Medea, Ascl. AP 5.210.2 τήκομαι, with Sens 2011, 31), though note also Ciris 249 senio . . . tabescere tali, of Scylla in love with Minos, where the language is psychological but the wider context refers to bodily symptoms; the use of τήκεσθαι to refer to lovesickness as a physical disease appears to be a Theocritean innovation, note e.g. 11.14–15 αὐτὸς ἐπ᾽ ἀιόνος κατετάκετο ϕυκιοέσσας | ἐξ ἀοῦς, ἔχθιστον ἔχων ὑποκάρδιον ἕλκος (imitated in Lucr. 4.1120 tabescunt uulnere caeco), which is metaphorical, but esp. 1.66 Δάϕνις ἐτάκετο,

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

84  Commentary ON 125 which seems to be intended more literally: this latter context may be the primary model here (cf. intr., section 2); cf. further Ep. Bion. 75 καινῷ δ᾽ ἐπὶ πένθεϊ τάκῃ. The Latin verb is common in Lucr. (7x), but largely avoided in subsequent poetry (Catull. 1x, Hor. Sat. 1x, Prop. 3x, Ov. 7x). For dolor in connection with (unhappy) love, cf. below 133, 141, 145, and see TLL 5(1).1842.40–64; the notion that love can bring pain is ubiquitous (e.g. Lucr. 4.1067 et seruare sibi curam certumque dolorem; cf. in general DMALL 114–15 s.v. ‘cuitas’), but here we have a more specia­ li­zed metonymic use of dolor in reference not just to a symptom but to the essence of love (most clearly at 145 Luna, dolor nosti quid sit), which seems otherwise only paralleled at Catull. 2.7 et solaciolum sui doloris (textually suspect: see Trappes-­Lomax 2007, 37–8) and Ciris 183 quo uocat ire dolor, before elegy (e.g. Prop. 1.17.19 illic si qua meum sepelissent fata dolorem, Tib. 1.2.1 adde merum uinoque nouos compesce dolores); although Greek poetry regularly refers to erotic suffering (e.g. Theocr. 3.12 θυμαλγὲς ἐμὶν ἄχος, Posid. AP 5.211.3–4 λήγω δ᾽ οὔποτ᾽ ἔρωτος, ἀεὶ δέ μοι ἐξ Ἀϕροδίτης | ἄλγος ὁ μὴ κρίνων καινὸν ἄγει τι Πόθος, Mel. AP 12.49.4 ἔκκρουσον στυγερὰν ἐκ κραδίας ὀδύναν: contrast the last with Tib. 1.2.1), it seems to lack such specialized usage—­which, however, is available in prose with πάθος (it may have influenced Ap. Arg. 3.675 τίπτ᾽ ἔπαθες; τί τοι αἰνὸν ὑπὸ ϕρένας ἵκετο πένθος; where πένθος refers to Medea’s passion for Jason, while allusively suggesting that at Il. 1.362 = 18.73 τέκνον τί κλαίεις; τί δέ σε ϕρένας ἵκετο πένθος; Achilles’ suffering is caused by his passion for Briseis and Patroclus, respectively), note its frequency in Parth.’s Ἐρωτικὰ παθήματα (on πάθημα, see Lightfoot 1999, 367–8), e.g. 5.2 ἐπεὶ μέντοι χρόνου διαγενομένου οὐδὲ ἐπ᾽ ὀλίγον ἐλώϕα τὸ πάθος, ἀνακοινοῦται τῇ μητρὶ καὶ πολλὰ καθικέτευε μὴ περιιδεῖν αὐτὸν ἀπολλύμενον, 11.1 τὸν Καῦνον ἐρασθέντα τῆς ἀδελϕῆς, ὡς οὐκ ἔληγε τοῦ πάθους, ἀπολιπεῖν τὴν οἰκίαν, 17.2 προϊόντος δὲ τοῦ χρόνου τὸ πάθος ἐπὶ μεῖζον ηὔξετο καὶ κατέχειν τὴν νόσον οὐκ ἔτι οἵα τε ἦν, 36.3 καὶ τὸ μὲν πρῶτον ἡσυχάζει αἰδοῖ κατεχομένη, ἐπειδὴ δὲ σϕοδρότερον ἐγίνετο τὸ πάθος, ἀπετόλμησεν εἰς λόγους ἐλθεῖν αὐτῷ (it is tempting to speculate that Parth. may have contributed to the development of this usage in Latin poetry). morientia membra, ‘dying body’: for the phrase, cf. Luc. 5.278–9 morientia caespite membra | ponere. The same noun was used of Lydia

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 125–6  85 only a few lines above (117), likewise in a context emphasizing the speaker’s physical separation from his beloved; this coincidence in the use of membra where a pronoun would have sufficed (‘herself ’ at 117, ‘I’ here) brings out the importance to the speaker of their bodily existence, with a hint that their bodies should be otherwise occupied (cf. Lucr. 4.1105–6 denique cum membris collatis flore fruuntur | aetatis, 1114 membra uoluptatis dum ui labefacta liquescunt). 126  The image and language are essentially Lucretian; for heat as a vital principle (specifically a constituent of the soul: Kenney 2014, 109), cf. 3.128–9 est igitur calor ac uentus uitalis in ipso | corpore, qui nobis moribundos deserit artus; for ‘the chill of death’, cf. 3.400–1 sed comes insequitur [sc. anima] facile et discedit in auras | et gelidos artus in leti frigore linquit, 4.924 aeterno corpus perfusum frigore leti, 6.1191 a pedibusque minutatim succedere frigus; for juxtaposing calor and frigus in a single line, though alio sensu, cf. 6.371 et calor extremus primo cum frigore mixtus. Contrast Virgil’s treatment of the same concepts at e.g. Aen. 1.92 soluuntur frigore membra, 4.705 dilapsus calor atque in uentos uita recessit. mortis (Ω) is conspicuous after 125 morientia: in and of itself this may not be a problem, but note that in other such cases—­e.g. Prop. 3.21.33–4 seu moriar, fato, non turpi fractus amore; | atque erit illa mihi mortis ­honesta dies, Verg. Aen. 9.362–3 ille suo moriens dat habere nepoti; | post mortem bello Rutuli pugnaque potiti, Ov. Fast. 3.547–9 tumulique in marmore carmen | hoc breue, quod moriens ipsa reliquit, erat: | ‘praebuit Aeneas et causam mortis et ensem’—mors has a concrete meaning and picks up the reference of the verb. Heinsius (in Burman 1773, 669) conjectured marcentia for morientia, so as to stress mortis here, but tabescere and marcere are too similar in sense and tone to warrant juxtaposition (in support of morientia, cf. also Claud. Carm. min. 29.20–1 tristi morientia torpent | membra fame, Eug. Tolet. 101.14 febre tabescunt moribunda membra). The Lucr. parallels quoted above rather suggest that mortis may be a synonym substitution for leti; cf. also Ov. (?) Met. 2.611 corpus inane animae frigus letale secutum est, Nemes. Ecl. 1.49 iaces letali [v.l. mortali] frigore segnis; by contrast, frigus mortis is unparalleled (Ov. Ibis 316 sit frigus mortis causa famesque tuae is different, as mortis goes

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

86  Commentary ON 126–7 with causa). While it is not impossible that the corruption was triggered by morientia, it is at least as likely that mortis originates as an exegetical gloss, for which cf. Boccaccio’s glosses mors over 160 letum and mortis over Culex 244 leti (cod. Laur. 33.31, ff. 27r. and 21r.); cf. further Black 1998, 118. 127 The aphoristic phrasing, evocative of omnia mea mecum porto, implies a reason why the actual state of affairs is wrong: what is mine (mea) should be with me (mecum); the same conceit is repeated at 137, 144, and 148. mecum . . . est, ‘is with me, keeps my company’: for the usage (OLD s.v. sum 12), cf. e.g. Pl. Truc. 362 nempe tu eris hodie mecum, mea Phronesium? contrast Dirae 94 esse sine illis; the expression is attested as a sexual euphemism (Adams 1982, 177), though for this sense to be operative we should probably expect the male to be the subject (the idiom seems to imply the active role). domina may be used here for the sake of variety, puella being the primary term throughout the poem (see 105n.), but it appears also to be the more intimate word, a token of private amatory language (cf. Dickey 2002, 81–5, esp. 82: ‘domine and domina were addresses with specifically sexual connotations which could be used like terms of endearment to flatter, win over, and arouse a lover’, citing in particular Prop. 2.14.21 pulsabant alii frustra dominamque uocabant, cf. e.g. Ach. Tat. 2.4.4 σὸν ἔργον ἤδη δέσποινάν τε καλεῖν καὶ ϕιλῆσαι τράχηλον, quoted more fully 107–8n., with Dickey 1996, 99): the speaker may be evoking here a situ­ ation when he could use the term properly, i.e. to address Lydia in private. In reference to the beloved, domina is attested since Lucil. 730 cum mei me adeunt seruuli, non dominam ego appellem meam, though in the classical period Catullus notably does not employ this usage (at 68.68 and 158 he comes close to it, but in both cases domina is motivated by domus: see Fulkerson 2013, 182), and it is Gallus who appears to have pioneered it in elegy (fr. 145.7 quae [sc. carmina] possem domina deicere digna mea, with Hollis 2007, 247), where domina is subsequently ubi­ qui­tous (note the early attestations in both Prop. 1.1.21 en agedum dominae mentem conuertite nostrae and Tib 1.1.46 et dominam tenero continuisse sinu; see further DMALL 152–4 s.v. ‘dueña’).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 127–8  87 127–8  The hyperbolic praise of a girl can be traced back, directly or indirectly, to Sappho fr. 56 οὐδ᾽ ἴαν δοκίμωμι προσίδοισαν ϕάος ἀλίω | ἔσσεσθαι σοϕίαν πάρθενον εἰς οὐδένα πω χρόνον | τεαύταν (note σοϕίαν ~ doctior), which also appears to lie behind Catull. 61.84–6 nequa femina pulchrior | clarum ab Oceano diem | uiderit uenientem; for in terris, cf. also Theocr. 18.20 οἵα Ἀχαιιάδων γαῖαν πατεῖ οὐδεμί᾽ ἄλλα (see Cavallini 1986, 28–30). A strikingly close parallel is supplied by an (imperial-­ date) epitaph for an eight-­year-­old slave-­girl, CLE 1166.6 doctior in terris nulla puella foret; the pentameter, in turn, finds an almost exact parallel in Mart. 10.32.6 pulchrior in terris nulla tabella foret, which suggests that both may be independently imitating a passage very much like the present one (Martial’s is clearly a parody), though perhaps from elegiacs rather than hexameters (see intr., section 2). In a bucolic context, cf. Catal. 9.24 altera non fama dixerit esse prior, of Messalla’s girlfriend celebrated in his Greek bucolics (with Kayachev 2020a, 94–5). For similarly phrased hyperbolic statements, cf. further Catal. 4.9 o quis te in terris loquitur iucundior uno? Hor. Ep. 2.2.157 uiueret in terris te si quis auarior uno, Ov. Am. 2.6.23 non fuit in terris uocum simulantior ales, Grat. 104 haud fuit in terris diuum obseruantior alter. 128 doctior  Lydia is a rustic girl or nymph, but has the accomplishments of an urbane docta puella (a Propertian tag: 1.7.11, 2.11.6, 13.11; cf. further Ov. Ars 2.281–2 sunt tamen et doctae, rarissima turba, puellae, | altera non doctae turba, sed esse uolunt), note 109–10. in terris: standard idiom for ‘on earth’, cf. 127–8n. for parallels. formosior is loaded: beauty entails desirability, as in the so-­called καλός inscriptions (‘X καλός’ ~ ‘I love X’, cf. Call. Aet. fr. 73 ἀλλ᾽ ἐνὶ δὴ ϕλοιοῖσι κεκομμένα τόσσα ϕέροιτε | γράμματα, Κυδίππην ὅσσ᾽ ἐρέουσι καλήν); cf. e.g. Watson 1985, 440: ‘formosus . . . often bears the connotation “attractive”; in other words, it refers not only to outward appearance but also to the pleasurable feelings aroused thereby in another person’ (cf. further 104n.). From Catull. 86 on, the adj. appears frequently in erotic contexts in bucolic and elegy; for the specific nuance of the comparative form, bringing out the inherent subjectivity of the adj. (~ ‘attractive, desirable’), cf. Ter. Eun. 730–1 quanto nunc formonsior | uidere mihi quam dudum, Prop. 2.29.25–6 non illa mihi formosior

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

88  Commentary ON 128–9 umquam | uisa and 3.23.13 an tibi nescioquae uisa est formosior? (note its repeated use together with uidere, add e.g. 2.18.29–30 per te poteris formosa uideri; | mi formosa sat es, si modo saepe uenis, Hor. Carm. 4.13.3 uis formosa uideri); cf. further Hutchinson 2019 on the im­port­ ance of comparison in erotic evaluation. 128–9  ac si | fabula non uana est (Ω) is suspect: the hexameter ending created by two monosyllables with tenuous syntactic links to the preceding words and between themselves produces a broken rhythm that is avoided in higher-­register poetry; with line-­final si, parallels are only found in Hor. Sat. 1.1.46 ut, si, 1.2.116 num, si, 2.3.92 ut, si, Ep. 1.16. 33 ut, si (lines like Ov. Ars 3.445 nec toga decipiat filo tenuissima, nec si or Met. 5.519 si reperire uocas amittere certius, aut si are different in that the line-­final si is prepared by the anaphoric structure). The harsh effect is increased by the fact that the subordinate clause is only a par­en­the­ t­ic­al aside, whereas the placement of si gives it unmotivated emphasis. It is an easy improvement to write illa for ac si (for illā paired with [n]ullă in this kind of construction, cf. Ov. Met. 1.322–3 non illo melior quisquam nec amantior aequi | uir fuit aut illa metuentior ulla deorum, 7.730–1 tristis erat, sed nulla tamen formosior illa | esse potest tristi), while replacing non with ni (both Heyworth*). The corruption may have been triggered by the erroneous illa (for ulla, restored by Boccaccio) at 127, prompting the scribe to replace illa here. 129 fabula ni uana est  Reference to the poetic (mythological) trad­ition, esp. critical reference, is a common device of ‘learned’ poetry; cf. e.g. Arat. 30 εἰ ἐτεὸν δή (with Kidd 1997, 184–5), Ap. Arg. 1.154–5 εἰ ἐτεόν γε πέλει κλέος ἀνέρα κεῖνον | ῥηιδίως καὶ νέρθεν ὑπὸ χθονὸς αὐγάζεσθαι, Nic. Ther. 10–12 εἰ ἐτεόν περ . . . Ἡσίοδος κατέλεξε (with Overduin 2015, 183–4), Verg. Georg. 3.391 = Aen. 6.173 si credere dignum est, 3.551 si uera est fama, and see in general Stinton 1976, Horsfall 1990. In the above examples, however, the metadiscursive formula is put in the mouth of an authorial narrator and does genuinely signal non-­mainstream material, whereas here (as at 147) the reference is to some of the best-­ known mythological stories and is made by a bucolic character. (Ov. Met. 10.28–9 famaque si ueteris non est mentita rapinae, | uos quoque iunxit Amor, spoken by Orpheus, is ironic, as it probably alludes to

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 129  89 Orphic texts on the subject.) A closer parallel can be found in εἰ δὴ ϕάτις ἔτυμος put in the mouth of the Lydian and Phrygian women by the ­chorus at Eur. IA 795 to express doubt over the traditional account of Helen’s birth: while disbelief that such a monster could be sired by a god may be part of the point (cf. perhaps Ov. Met. 8.120 non genetrix Europa tibi est, sed inhospita Syrtis), a more significant purpose may be to ­convey the sense of the geographical and civilizational distance between Argos and Troy. A similar parallel, in some ways even closer (like Lydia, Poppaea might attract Jupiter), is provided by [Sen.] Oct. 762–72 (cf. Ferri 2003, 344): si uera loquax fama Tonantis furta et gratos narrat amores (quem modo Ledae pressisse sinum tectum plumis pennisque ferunt, modo per fluctus raptam Europen taurum tergo portasse trucem), quae regit et nunc deseret astra, petet amplexus, Poppaea, tuos, quos et Ledae praeferre potest et tibi, quondam cui miranti fuluo, Danae, fluxit in auro.

Here the si formula may serve to stress the chronological distance, and this no doubt is the point in the Lydia: the speaker is at pains to show that he is completely cut off from the world of erotic bliss enjoyed by the gods. As I discuss in intr., section 2, fabula in the neutral sense of ‘lore, trad­ition’ appears to be treated as archaic and/or vulgar in Augustan poetry, with Tib. 2.3.31–2 fabula nunc ille [sc. Apollo] est, sed cui sua cura puella est, | fabula sit mauult quam sine amore deus alluding to the present context in the hexameter but pointedly giving fabula its modern connotation of ‘gossip, slander’ in the pentameter; elsewhere in elegy fabula is sometimes used in reference to mythological stories, but often to ones of a scandalous nature, thus playing on both usages, cf. esp. [Tib.] 3.4.67–8 me quondam Admeti niueas pauisse iuuencas | non est in

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

90  Commentary ON 129 uanum fabula ficta iocum (cf. ni uana est), Ov. Am. 1.9.39–40 Mars quoque deprensus fabrilia uincula sensit: | notior in caelo fabula nulla fuit (cf. 127–8 non ulla puella | doctior in terris fuit with n.). tauro . . . uel auro The laconic references to Jupiter’s seduction of Europe and Danae evoke epigram (e.g. AP 9.48 Ζεὺς κύκνος, ταῦρος, σάτυρος, χρυσὸς δι᾽ ἔρωτα | Λήδης, Εὐρώπης, Ἀντιόπης, Δανάης); the jingle tauro . . . auro may also illustrate that all such myths are but a variation on a single theme, cf. Ov. Am. 3.12.33–4 Iuppiter aut in aues aut se transformat in aurum | aut secat imposita uirgine taurus aquas (Lorenz 2005, 19), where the wordplay, even if less conspicuous, is extended by the addition of aues (cf. also the playful pairing of fulmine et imbre ‘lightning and rain’ at Catal. 9.34, likewise in reference to Jupiter’s affairs). (Cf. in a way Naeke 1847, 154: ‘Cato autem, quum potuisset cycno Iove digna, vel auro, maluit tauro Iove digna, vel auro, gratum sibi suaeque aetati homoeoteleuton’; Franklinos* suggests that cycno should actually be printed, but I am not convinced that the jingle is ob­jec­tion­ able, and I find the following apostrophe of an actual bull at 131 humorously pointed, cf. in a way the repetition 143 Phoebi–146 Phoebe, in reference to two different deities.) The grecism (pace EDL s.v.) taurus appears initially to have been used to refer to bulls from Greek myth­o­ logy (note that Plautus only uses the term at Ps. 200–1, in reference to Dirce’s bull), as it is here, later being naturalized as a term for ‘bull’ (as opposed to ‘ox’ or ‘cow’), as at 129; at Catull. 64.40 non glebam prono conuellit uomere taurus (and a number of dependent contexts), it has been taken to mean ‘ox’ (cf. Trimble 2010, 74–5), but it probably is intended to evoke Aeetes’ ταύρω χαλκόποδε, στόματι ϕλόγα ϕυσιόωντε (Ap. Arg. 3.410 = 496). Ioue digna The practice of measuring a woman’s desirability by her (potential) appeal to Zeus and in comparison with his paramours goes ultimately back to a passage in the Διὸς ἀπάτη (Il. 14.315–28) in which Zeus himself tells Hera that she is more desirable than any of his mistresses, human or divine, has ever been (cf. Kayachev 2016a, 182). For the phrase, cf. esp. Prop. 1.13.29–30 Ioue digna, et proxima Ledae | et Ledae partu, gratior una tribus (with Booth 2006 and Heyworth 2007, 63), of Gallus’ girlfriend, as well as 2.2.6 Ioue digna soror, of Juno, which may hint at the Homeric origin of the conceit. Whether intentionally or

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 129–55  91 not, the conjunction of digna with auro evokes the topos of Danae as an archetypal prostitute, cf. e.g. Antipater AP 5.31.5–6 Δανάῃ Ζεὺς | οὐ χρυσός, χρυσοῦς δ᾽ ἦλθε ϕέρων ἑκατόν, Parmenion AP 5.3–4 ὁ Ζεὺς τὴν Δανάην χρυσοῦ, κἀγὼ δὲ σὲ χρυσοῦ· | πλείονα γὰρ δοῦναι τοῦ Διὸς οὐ δύναμαι (note that cattle too can serve as currency in Homer, and cf. pecunia ← pecus, with Maltby 1991, 459); for the ‘pecuniary’ sense of digna, cf. Ov. Her. 14.99–100 ut posses etiam Ioue digna uideri, | diues eras, of Io, with Reeson 2001, 292–3. 130 Iuppiter Alcaeus AP 12.64 stands at the head of a series of epigrams (64–70) that develop the idea that Zeus may be attracted by the poet’s beloved and snatch him away, like Ganymedes; some address to Zeus a plea not to do so, e.g. 69.1 Ζεῦ, προτέρω τέρπευ Γανυμήδεϊ, as does the present line, though other scenarios are also possible, e.g. 66.1–2 κρίνατ᾽, Ἔρωτες, ὁ παῖς τίνος ἄξιος [cf. 129 Ioue digna]. εἰ μὲν ἀληθῶς | ἀθανάτων, ἐχέτω· Ζανὶ γὰρ οὐ μάχομαι. Cf. further Ov. Am. 1.10.7–8 talis eras; aquilamque in te taurumque timebam, | et quidquid magno de Ioue fecit amor, with McKeown 1989, 285–6. auertas aurem The expression is infrequent and seems to have a ­solemn and/or deferential tone, suitably for the context, cf. e.g. Prop. 2.13.15 si forte bonas ad pacem uerterit aures (of Cynthia as a judicious docta puella), Mart. 6.64.8–9 quibus aurem aduertere totam | non aspernantur proceres urbisque forique, Sen. Contr. 1.2.5 auertite aures petiturae sacerdotium. mea sola puella est pointedly echoes 127 non ulla puella; for a com­ par­able combination of a negative and a positive formulation of the same idea, cf. [Tib.] 3.8.15 sola puellarum digna est and 24 dignior est uestro [sc. Pieridum] nulla puella choro (of Sulpicia); for the profession of a girl’s uniqueness, cf. further Ov. Fast. 1.650 sola toro magni digna reperta Iouis (of Livia being worthy of Augustus), Prop. 2.3.29–30 gloria Romanis una es tu nata puellis; | Romana accumbes prima puella Ioui. 131–55  In the first section (104–30), the topic was the speaker’s envy, almost jealousy, for the fields with which Lydia shares intimacy while he is far away. The second section develops the theme of envy by directing it at figures who, unlike the speaker, enjoy the company of their own partners (rather than of Lydia, as in the case of the fields). The transition

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

92  Commentary ON 131–55 is effected smoothly: the first stanza (131–41) apostrophizes domestic animals, the bull and the he-­goat, which follows naturally on the apo­ strophes to the fields in the first section; note also that both the animals (131 and 134) and the fields (112) are objects of μακαρισμοί. The second and the third stanzas make the same point about the gods (142–9) and the heroes (150–9), respectively: according to the speaker, both cat­ego­ r­ies could enjoy unhindered relationships. Besides envy, a feeling of indignation develops over the two latter stanzas: against the gods, because they are responsible for putting the speaker in a disadvantaged position (cf. 154), and against the heroes, because they were mortals like the speaker, but were treated more favourably. The third section (156–83) will attempt to explain why the speaker was prevented from enjoying his affair with Lydia indefinitely, but will find no better answer than to blame the malignity of fate. 131–41  The idea of domestic animals as a model of sexual behaviour may have been suggested by Theocritus’ comparison of lovesick Daphnis to the goatherd envying the he-­goat (1.87–8 ᾡπόλος, ὅκκ᾽ ἐσορῇ τὰς μηκάδας οἷα βατεῦνται, | τάκεται ὀϕθαλμὼς ὅτι οὐ τράγος αὐτὸς ἔγεντο), cf. Hubaux 1930, 55; the exact point of the passage is somewhat ambiguous (see Hunter 1999, 92–3), as is Daphnis’ situation in general, but overall it is clear that, in contrast to the he-­goat which can freely satisfy his desires with the she-­goats, both Daphnis and the goatherd find themselves sexually frustrated. A striking parallel is presented by Philetas’ speech in Longus, which likewise combines the bull’s and the he-­goat’s erotic bliss as a foil for the speaker’s own misery, followed by an invocation of a sympathetic deity: 2.7.4–6 ἔγνων δὲ ἐγὼ καὶ ταῦρον ἐρασθέντα, καὶ ὡς οἴστρῳ πληγεὶς ἐμυκᾶτο [~ 133 mugire dolorem]· καὶ τράγον ϕιλήσαντα αἶγα, καὶ ἠκολούθει πανταχοῦ [~ 137 tecum tua laeta capella est]. αὐτὸς μὲν γὰρ ἤμην νέος καὶ ἠράσθην Ἀμαρυλλίδος· καὶ οὔτε τροϕῆς ἐμεμνήμην οὔτε ποτὸν προσεϕερόμην οὔτε ὕπνον ᾑρούμην. ἤλγουν τὴν ψυχήν, τὴν καρδίαν ἐπαλλόμην, τὸ σῶμα ἐψυχόμην [~ 127 calor infuso decedit frigore leti]· ἐβόων ὡς παιόμενος, ἐσιώπων ὡς νεκρούμενος, εἰς ποταμοὺς ἐνέβαινον ὡς καόμενος [~ 141 crudelem patior tam saepe dolorem]. ἐκάλουν τὸν Πᾶνα βοηθόν, ὡς καὶ αὐτὸν τῆς Πίτυος ἐρασθέντα [~ 145 Luna, dolor nosti quid sit: miserere dolentis] (it is tempting to

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 131–41  93 speculate that both Longus and the Lydia are echoing independently the poetry of Philitas of Cos: for his potential presence in Longus, in particular in Philetas’ speech, cf. e.g. Spanoudakis 2002, 64–6; though the possibility that Longus knew the Lydia should not perhaps be ruled out, cf. in general Jolowicz 2021). The theme of domestic animals as erotic exempla then reappears in Longus as he presents Daphnis being inspired by the sight of rams and he-­ goats mounting sheep and she-­ goats (3.13.1–14.5); in particular, Daphnis points out to Chloe that, after mating, the male and the female continue to enjoy each other’s company (3.14.3 μετὰ τοῦτο τὸ ἔργον οὔτε ἐκεῖναι ϕεύγουσιν ἔτι αὐτοὺς οὔτε ἐκεῖνοι κάμνουσι διώκοντες, ἀλλ᾽ ὥσπερ κοινῆς λοιπὸν ἀπολαύσαντες ἡδονῆς συννέμονται), which closely resembles the presentation of the he-­ goat here (134–7) and contrasts the speaker’s own situation (after their first sexual intercourse, he and Lydia were separated). The idealized view of animal sexuality developed in this stanza also anticipates the idea of the golden age as an erotic paradise hinted at below (cf. 150–5n.): the common denominator is provided by the widespread belief that primitive people lived a θηριώδης βίος in more than one respect, including sexual behaviour (cf. in general Campbell 2003, 193–4, and note e.g. Hdt. 1.204 μίξιν δὲ τούτων τῶν ἀνθρώπων εἶναι ἐμϕανέα κατά περ τοῖσι προβάτοισι, of primitive tribes living in the Caucasus). This positive attitude to the sexual mores of animals and primitive humans is of course in keeping with the bucolic character of the Lydia speaker, but more sober outlooks also existed, note e.g. Hor. Sat. 1.3.109–10 quos uenerem incertam rapientes more ferarum | uiribus editior caedebat ut in grege taurus, of prehistoric people, which, real­is­tic­ al­ly, emphasizes violence in the sex life of both animals and primitive humans. It seems in fact likely that the reader is expected to recognize that the speaker’s viewpoint is one-­sided at best (contrast e.g. Longus, who does not suppress the presence of violence: 3.13.2 ἐμάχοντο περὶ τῶν αἰγῶν [sc. τράγοι]), just as his ideas about the divine (142–9) and the heroic (150–5) worlds are strikingly naïve. In any event, domestic animals could more properly serve as a model of unrestricted sexuality than of marital fidelity, cf. Myrrha’s similarly structured argument at Ov. Met. 10.321–35, where she observes that there is no ban on incest among animals, including cows and goats (325–7 nec habetur turpe iuuencae |

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

94  Commentary ON 131–41 ferre patrem tergo, fit equo sua filia coniunx, | quasque creauit init pecudes caper; note the μακαρισμός at 329 felices quibus ista licent), as well as in certain tribes (331–3 gentes tamen esse feruntur | in quibus et nato genetrix et nata parenti | iungitur; note her regretting that she was not born in such a place, 334 me miseram, quod non nasci mihi contigit illic, cf. the Lydia speaker’s complaint that he was not born during the golden age, 179–80 infelix ego non illo qui tempore natus | quo facilis natura fuit). The string of μακαρισμοί serving as a foil for the speaker’s own situ­ ation finds a parallel (inspiration?) in Bion fr. 12 listing ideal examples of a homoerotic relationship from the heroic age (1–2 ὄλβιοι οἱ ϕιλέοντες ἐπὴν ἴσον ἀντεράωνται. | ὄλβιος ἦν Θησεὺς τῶ Πειριθόω παρεόντος . . . ); as Reed (1997, 176) comments, ‘the exempla might have been invoked by one whose beloved was unresponsive’; cf. further Hor. Carm. 1.13.17–20 felices ter et amplius | quos inrupta tenet copula nec malis | diuulsus querimoniis | suprema citius soluet amor die, Tib. 2.3.29–30 felices olim, Veneri cum fertur aperte | seruire aeternos non puduisse deos. 131 felix taure  On address to animals in Latin, see Dickey 2002, 301; on felix, 325: ‘Term of praise or envy, esp. for the dead or those distanced from the speaker’ (contrast 179 infelix, used by the speaker of himself); on taurus, cf. 129n. pater magni gregis et decus In reference to a bull, cf. Stat. Ach. 1.313 pater armenti, though such periphrases are more frequent in reference to a he-­goat, cf. 134 pater haedorum with n., Verg. Ecl. 7.7 uir gregis ipse caper, Petr. 133.3.14 hircus, pecoris pater, also Ov. Ars 1.522 nec laedat nares uirque paterque gregis (for caper in the metaphorical sense OLD s.v. 1b); for decus, cf. Lucr. 3.3 Graiae gentis decus, Ov. Fast. 1.415 hortorum decus et tutela, Priapus, Mart. 8.28.2 esse uelis cuius fama decusque gregis? Claud. Pros. 2.209 stabuli decus armentique iuuencam. a te In view of the positioning, it may be tempting to take a te absolutely in the sense of sine te, but such usage is unparalleled (sum a te would mean ‘I come from you’ rather than ‘I am away from you’); a te must be construed with 132 secreta, ‘withdrawn (hidden?) from you’, cf. Ov. Fast. 6.9–10 secretus ab omni | uoce locus, Cic. Q. fr. 1.1.43 si mihi quicquam esset abs te separatum; the strained word order is probably intended to visualize the separation of the cow from the bull (cf. the

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 131–3  95 effect achieved at Lucr. 1.452 seiungi seque gregari, on which see Hinds 1987). For the line ending, cf. Hor. Sat. 1.2.69–70 numquid ego a te | magno prognatum deposco consule cunnum? Prop. 3.9.59–60 a te est | quod ferar in partes ipse fuisse tuas, Ov. Pont. 4.10.75 conditur a te. 132 uaccula  The diminutive is only paralleled in classical literature at [Verg.] Priap. 2.14, which appears corrupt, and Apul. Met. 7.25 pastores enim mei perditam sibi requirentes uacculam; but it is also attested in epigraphy as the cognomen of two Nigidii living at Pompeii shortly before the eruption (see e.g. Franklin 1997, 435). Still, Heinsius’ bucula (Burman 1773, 670) may well be right (cf. esp. Verg. Ecl. 8.85–6 fessa iuuencum | per nemora atque altos quaerendo bucula lucos), as there appears to have been a tendency for bucula to be corrupted to uaccula (cf. TLL 2.2235.77: ‘confunditur cum vaccula, a qua vix discernas’), no doubt under the influence of uacca which has largely supplanted bos in Romance languages. secreta cubilia The noun is doubly appropriate: it is a proper term for an animal’s lair (OLD s.v. cubile 3), cf. e.g. Col. 8.5.11 quam secretissima cubilia legunt (for hens to lay eggs), Nemes. Cyn. 236 leporum secreta cubilia monstrant, but also for ‘(marriage) bed’ (OLD s.v. 2), cf. e.g. Ov. Fast. 1.427 ut tetigit niueae secreta cubilia nymphae [sc. Priapus]. Contrast Lydia, who is described as secreta at 118. captans, ‘seeking’: cf. OLD s.v. capto 3 (b rather than a); for cubilia as direct object, cf. Ov. Fast. 2.337 uenerat ad strati captata cubilia lecti (Faunus searching for Omphale’s bed). 133 siluis  Woods are the place to which cattle run away, esp. bulls having lost their herds, cf. Theocr. 14.43 ἔβα ποκὰ ταῦρος ἀν᾽ ὕλαν, Verg. Georg. 3.149–50 quo [sc. oestro] tota exterrita siluis | diffugiunt armenta, Ov. Met. 13.871–2 taurus uacca furibundus adempta | stare nequit silua­ que et notis saltibus errat, Luc. 2.601–2 pulsus ut armentis primo certamine taurus | siluarum secreta petit, Sil. 16.4–5 abditus ut silua, stabulis cum cessit ademptis, | amisso taurus regno gregis. mugire dolorem While mugire is the proper (onomatopoeic) term for cows’ mooing, it often has the connotation of suffering, partly because cows moo to express discomfort, but also through association with human moaning (note TLL 8.1559.15–25, and cf. cognate μύζειν ‘to

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

96  Commentary ON 133–5 moan’). Mooing can be provoked by separation from close ones: cf. e.g. Ov. Fast. 4.459 ut uitulo mugit sua mater ab ubere rapto (note how Ovid suggests both a psychological and a physiological cause: the cow misses her calf, but also suffers from her udder being overfilled with milk), Rem. 183–4 parte sonant alia siluae mugitibus altae, | et queritur uitulum mater abesse suum, Theocr. 4.12 ταὶ δαμάλαι δ᾽ αὐτὸν μυκώμεναι αἵδε ποθεῦντι (of cows missing their herdsman, who cannot milk them); but often it expresses sexual desire: cf. Stat. Silv. 4.5.18 nec uacca dulci mugit adultero, Ap. Arg. 1.1269 ἵησιν μύκημα, κακῷ βεβολημένος οἴστρῳ (of a bull, in a simile describing Heracles in search for lost Hylas; οἶστρος is a standard metaphor for frantic desire), Mosch. Eur. 97 μειλίχιον μυκήσατο (of the bull seducing Europa), Longus 2.7.4 quoted 131–41n. The acc. dolorem (mugire is normally intransitive) is used by analogy with 106 nostrum suspirat amorem (see n.); for dolor in the sense of erotic longing, cf. 125n. 134 pater haedorum  Cf. 131 pater magni gregis with n., add further Theocr. 8.49 ὦ τράγε, τᾶν λευκᾶν αἰγῶν ἄνερ, Leonid. AP 9.99.1 αἰγὸς πόσις, Hor. Carm. 1.17.7 olentis uxores mariti (with Nisbet and Hubbard 1970, 220). felix semperque beate On felix, see 131n.; on beate, 112n.; the pairing of two adjectives meaning ‘happy’ for emphasis is idiomatic, cf. 112 with n., and add e.g. Ov. Pont. 2.8.5 argentum felix omnique beatius auro, Plin. Ep. 1.3.2 felix beatusque es, Apul. Met. 2.7 quoted 112–18n. Cic. cites Epicurus’ tenet that the wise man is semper beatus over two dozen times, otherwise the collocation is extremely rare. 135  siue petis montes praeruptos [-aque n] saxa pererrans: Ω’s text is possible, but a number of considerations speak in favour of reading praeruptaque saxa pererras (Burman 1727, 726, claiming that ‘pererras in codice veteri invenit Heinsius’). (1) Postpositive participial constructions have a certain feebleness to them, and it is preferable to have a finite verb at the end of a phrase and/or line (but Burman’s petens for petis is unattractive). (2) Adjective-­ noun word order (praeruptaque saxa) is more effective than noun-­adjective (montes praeruptos). (3) The two phrases—‘to climb mountains’ and ‘to roam over rocks’—express basically the same idea, which suggests that they should be coordinate,

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 135–6  97 rather than one being subordinate to the other. (4) Ov. Rem. 179 ecce petunt rupes [montes cod. Voss.] praeruptaque saxa capellae must be related to our passage (Ovid recommends countryside as a respite from love, whereas for the Lydia speaker it is erotically charged). The fact that n reads praeruptaque tilts the balance in favour of Burman’s proposal: while it could in principle be a reminiscence of the Ovidian passage (or an independent error), it is not impossible that it reflects a variant reading of the archetype. petis montes Besides Ov. Rem. 179 quoted above, cf. e.g. Culex 45–6 propulit e stabulis ad pabula laeta capellas | pastor et excelsi montis iuga summa petiuit, Sen. Tr. 1084 extrema montis ille praerupti petit; cf. further Theocr. 3.1–2 ταὶ δέ μοι αἶγες | βόσκονται κατ᾽ ὄρος. praeruptaque saxa The adj. is a standard epithet for cliffs and such, in poetry as well as prose, cf. e.g. Caes. BC 1.68.2 saxa multis locis praerupta iter impediebant, Verg. Georg. 2.156 praeruptis oppida saxis, Catull. 64.126 praeruptos tristem conscendere montes. pererras The verb is otherwise first attested at Verg. Ecl. 1.61, then several times in Aen., once in Hor. Sat., and becomes common in Ov.; is absent in prose until the first cent. ad; this might suggest colloquial pedigree. In reference to a mountainous locale, cf. e.g. Val. Fl. 3.537 saltusque tuos montesque pererrat. 136  fastidire (Ω) produces a sense (‘to turn away from unfamiliar grass’) that is the opposite of what we expect. One line of approach has been to restore adire, while making it depend on an impersonal verb/phrase analogous to 137 libet: sat sit adire (Scaliger 1572, 442), fas sit adire (Heinsius in Burman 1773, 671), praestat adire (Shackleton Bailey 1978, 309); but pabula adire ‘to approach fodder’ is a rather feeble expression. Mähly’s (1870, 782) uestigare strikes exactly the right note: ‘to search out new pastures’ (cf. below); cf. Watt 2001, 280–1. For the corruption, cf. e.g. fastigia for uestigia at Man. 1.616 (with Housman 1937, 57), though the reverse error seems more common. siluis For forest pastures, cf. Var. Rus. 2.10.3 quibus [sc. to cattle and goats] rupes ac siluae ad pabulandum cordi, Verg. Georg. 3.314 pascuntur uero siluas et summa Lycaei (of a goat breed), Calp. Ecl. 5.44 nec nimis amotae sectabere pabula siluae (for young lambs).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

98  Commentary ON 136–9 noua pabula: ‘fresh fodder, untouched pastures’, cf. Ov. Met. 10.121–2 tu pabula ceruum | ad noua . . . ducebas, Sen. Oed. 845–6 laetus Cithaeron pabulo semper nouo | aestiua nostro prata summittit gregi. uestigare The verb (contrast ubiquitous uestigia) is largely avoided in the poetry of the first cent. bc and including Ovid (only Cic. Arat. fr. 34.142, Verg. Aen. 6.145, 7.132, Ov. Met. 3.52, though note e.g. Pl. Rud. 111, Enn. Ann. 41), except in Aen. 12 where it is used several times of Aeneas ‘tracking down’ Turnus (467, 482, 557, plus in a simile at 588); it must have a prosaic, almost technical, ring to it; cf. esp. Col. 8.17.14 piscis magis naribus escam quam oculis uestigat. On the spondaic ending, see intr., sections 1 and 3. 137 campis  It is more often other kinds of domestic animals that graze on plains (e.g. Lucr. 2.660–2 ex uno tondentes gramina campo | lanigerae pecudes et equorum duellica proles | buceriaeque greges), whereas goats prefer hills and woods (cf. 135 and 136 with nn.), but note Verg. Aen. 3.220–1 laeta boum passim campis armenta uidemus | caprigenumque pecus nullo custode per herbas. tecum . . . laeta . . . est: probably ‘is happily with you’ (cf. 127 mecum . . . est with n.) rather than ‘enjoys your company’; cf. Ov. Ars 2.485–6 laeta salitur ouis, tauro quoque laeta iuuenca est; | sustinet immundum sima capella marem, where laeta serves first as a predicative epithet and then as a nominal predicate (a comment on the potential ambiguity here?). 138–9  The passage has two substantial difficulties: first, in order to provide a fitting parallel for the speaker’s situation, the subject of 193 plorauit (Ω) ought to be the male, not the female; second, plorare is not normally used of animals; both problems can be solved by writing tole­ rauit: ‘the female keeps the male’s company and never suffers their union to be broken’, a sentiment close to that of 131–3. A third, minor, issue is quocumque (Ω), for which Badius’ tentative quicumque (ed. Ascens. 1501, fol. XVIIIr.) is the obvious solution. Heyworth* additionally points out (1) that the construction involving a quicumque clause answered in the main clause by ille in an oblique case is irregular (but see 138n.), (2) that mas and femina could potentially imply humans (but  this is made less likely by the preceding context), and (3) that

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 138–9  99 interpellare is avoided in higher-­register poetry (but the Lydia is generally more tolerant), and suggests that the lines are interpolated; I do not find these objections strong enough, and the parallelism with 147–9, which similarly sum up the preceding stanza on gods, is attractive and supports the lines’ authenticity. 138 mas . . . femina:  a male and a female animal. sua The reflexive possessive pronoun normally refers to the subject, but can have other referents as well, esp. when it has an emphatic force (cf. OLD s.v. suus A.2, Pinkster 2015, 1132–3). Here it may simply be emphatic (OLD s.v. B.2 ‘(in emph. use) His own’), cf. e.g. Pl. Merc. 713 iubet saluere suos uir uxorem suam, St. 133 suos rex reginae placet, but probably has a more specific nuance, ‘Especially dear to him, his beloved’ (OLD s.v. B.7), cf. the substantivized pronouns tuus . . . mea at 144 (with n.). quicumque . . . illi For a quicumque relative clause answered by ille in an oblique case, cf. Man. 2.537–8 quaecumque nitent humana condita forma | astra, manent illis inimica et uicta ferarum. iuncta: probably ‘accompanying’ in the first place (OLD s.v. iungo 6), but with a hint of ‘wed’ (s.v. 7b), cf. Ov. Met. 2.701 iuncta suo pretium dabitur tibi femina tauro; though note that the verb can also be used specifically of intercourse (Adams 1982, 179), cf. in a way Ov. Ars 2.481–2 cum quo sua gaudia iungat, | inuenit in media femina piscis aqua. 139  is a metrical rarity, being formed by only four words, a kind of verse that gained some currency in ‘neoteric’ poetry in imitation of Greek practice (on which see Bassett 1919). For the scheme ‘(X –) adj. – (X –) verb – noun’ found here, cf. e.g. Catull. 64.77 Androgeoneae poenas exsoluere caedis and esp. 319 (mollia lanae) uellera uirgati custodibant calathisci, which likewise is in enjambement from the preceding verse and concludes a period (cf. further e.g. Ap. Arg. 3.764–5 ὁππότ᾽ ἀνίας | ἀκάματοι πραπίδεσσιν ἐνισκίμψωσιν ἔρωτες); on the effect, cf. Hill 2021, 44–5, Dainotti 2015, 79–80 n. 261. interpellatos The verb is avoided in higher-­register poetry: from the first cent. bc on, its only attestations are Hor. Sat. 1.6.127, 1.9.26, 2.4.5, Man. 3.559; for the context, cf. Liv. 3.57.4 stupro interpellato, though

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

100  Commentary ON 139–41 there the sense is ‘to impede’ (OLD s.v. interpello 4) rather than ‘to interrupt’ (s.v. 3). tolerauit In comparison with pati, the verb is sparsely used in the poetry of the first cent. bc and including Ovid (Cic. carm. fr. 23.1, Verg. Aen. 8.409, 515, Hor. Carm. 2.5.4, Sat. 2.5.20, Ov. Ars 2.669, Rem. 231, Fast. 2.300, Pont. 1.2.87), except in Lucr. (3x) and Ov. Met. (5x); its distribution suggests a somewhat archaic, and possibly colloquial, ­ ­colouring; for use with an acc. and inf. construction, cf. Enn. Ann. 130 ferro se caedi quam dictis his toleraret. 140 cur non:  repeated at 144, 178. For the anaphora of cur (here with 141 cur), cf. e.g. Prop. 4.7.31–2 cur uentos non ipse rogis, ingrate, petisti? | cur nardo flammae non oluere meae? Ov. Her. 19.69–70 cur ego tot ­uiduas exegi frigida noctes? | cur totiens a me, lente natator, abes? facilis, natura, fuisti Elsewhere facilis natura describes the character of a substance or individual (OLD s.v. natura 8 and 11a): Lucr. 5.1288 quo facilis magis est natura et copia maior (of copper as opposed to iron), Plin. Nat. 36.160 faciliore multo natura (of mica or selenite), Suet. Gram. 7.1 comi facilique natura (of Dionysius Scytobrachion); here (and at 180) natura refers to the principle that governs the behaviour of animals and humans (cf. perhaps OLD s.v. 4 and 11b), though there still may be an echo of Lucr. 5.1288, a context suggestive of the metallic races myth. The apostrophe may evoke the personification of natura in Lucr. 3.931– 51, where she scolds a person afraid to die (see Kenney 2014, 199–200); it is in turn echoed in Ov. Am. 3.8.5–6 contra te sollers, hominum natura, fuisti | et nimium damnis ingeniosa tuis, which likewise comments on the downsides of progress. 141  pointedly echoes 133 frustra te patitur siluis mugire dolorem, while changing the construction governed by pati. crudelem is glossed as ‘hard to bear, grievous’ by OLD s.v. crudelis 3b; in this kind of context, the adj. seems in fact to have a more physical and more active sense, ‘intense, unabated, relentless’ (here supported by tam saepe), cf. e.g. Catull. 64.76 crudeli peste coactam [sc. Athens] (the pestilence was unremitting, which made the Athenians comply), Verg. Aen. 2.368–9 crudelis ubique | luctus (note ubique, and cf. Horsfall 2008, 302); in this sense, it can in particular be used (as here) of obsessive passion,

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 141–9  101 cf. Ecl. 10.29 nec lacrimis crudelis amor [sc. saturatur], Aen. 6.442 durus amor crudeli tabe peredit, Tib. 1.8.7–8 deus crudelius urit, | quos uidet inuitos succubuisse sibi. patior . . . dolorem: a prosaic expression, in poetry only at Juv. 13.229–30 lateris uigili cum febre dolorem | si coepere pati; cf. further e.g. Larg. 130 oportet autem impositum esse medicamentum, donec dolorem pati non possint, [Quint.] Decl. 14.1 remedii mei patior dolorem, Cic. Ver. 2.5.112 ille morte proposita facile dolorem corporis patiebatur, Hyg. Fab. 103.2 dolorem pati non potuit Laodamia. tam saepe: on ‘the language of habit’ frequently employed in reference to the situation and activities of the elegiac lover, see Heyworth 2013. 142–9 This stanza is one of the most damaged in the poem (see 142–3n. and 147–8n.), but the overall sense is clear: in the previous stanza, the speaker claimed that in the animal world couples in­sep­ar­ ably enjoy each other’s company, and here he makes a similar statement about the divine world. Yet just as the idea that animals live in an erotic idyll is not unproblematic (cf. 131–41n.), so the claim that gods never part from their beloved is not easily substantiated. The first example, Luna’s daily visits to Endymion, sounds plausible, unless it is a distortion of a more common version associating these visits with lunar eclipses or new moons (cf. 142–5n.). But the second example, Phoebus’ affair with Daphne, is outright preposterous: their marital bliss consists in nothing more than Apollo’s always wearing a laurel wreath (146). And this latter example serves as a model for the generalizing statement that follows: gods are inseparable from their beloved, because they either carry them around in a metamorphosed shape or see them catasterized in the sky (147–9). 142–3 sidera per uiridem redeunt cum pallida mundum inque uicem phebus [M: -o S: -e FL] currens atque aureus orbis. The lines should refer to the time of the Moon’s visits to Endymion (144 Luna, tuus tecum est), which are usually imagined to coincide with either lunar eclipses or new moons (though the available evidence is insufficient to draw reliable inferences as to which version of the myth

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

102  Commentary ON 142–3 could be considered standard: cf. Jackson 2006, esp. 11–14, Hunter 2015, 92–3). It seems unlikely, however, that this can be the implication here: on the one hand, these occurrences are too infrequent to illustrate the idea that gods are inseparable from their beloved; on the other, 142–3 can hardly convey such a specific reference; the passage evidently refers to what occurs on a daily basis. The paradosis appears to indicate the time when the stars come back (redeunt) and the sun, in turn (inque uicem), disappears (e.g. cadit for atque by Baehrens 1880, 81). Endymion was in fact interpreted as a night hunter (schol. Ap. Arg. 4.57–8 νύκτωρ πρὸς τῇ σελήνῃ κυνηγεῖν) or an astronomer (νύκτωρ σχολάζοντα): in view of 149 inspersa uidet mundo, it may not be inconceivable that the Moon could be said to be with Endymion when she looks at him down from the sky. Yet this goes blatantly against tradition, which presented the Moon and Endymion as sleeping together. Keeping redeunt also prod­uces an unmotivated ὕστερον πρότερον: it would be more accurate to say that first the sun sets and then the stars reappear (the distortion of the normal order is aggravated by inque uicem). It appears more likely that the reference should be to morning, the Moon’s relationship with Endymion mirroring that of the Dawn with Tithonus. Sonntag (1896) also aptly compares Ov. Met. 4.217–18 dumque ibi quadrupedes caelestia pabula carpunt | noxque uicem peragit, thalamos deus intrat amatos: ‘Während der Sonnengott die Nacht benutzen muß, um die Geliebte aufzusuchen, bleibt der noctivaga Phoebe nur der Tag zu solchem Verkehr übrig.’ Appropriate verbs have been restored by Haupt 1874, 13: cedunt for redeunt and redit for atque; around these, each line has to be fixed up. At 142, a glaring problem is the adjective uiridem defining mundum: out of a number of proposals (e.g. uarium ed. Ven. 1555, teretem Heinsius in Burman 1773, 671, uiduum Goold in Fairclough and Goold 2000, 398), Giardina’s (2009, 171) nitidum seems the easiest (the same corruption apparently occurs at Plin. Nat. 17.74 nitido [Kayachev 2021b, 901: uiridi codd.] caelo). A potential difficulty is that with a verb of motion, per could be taken to mean ‘through’ (cf. Man. 1.395 hoc duce per totum decurrunt sidera mundum) rather than ‘throughout, across’ (cf. Man. 1.848 cum uaga per nitidum [liquidum Goold] scintillant lumina mundum); while this may indicate that cedunt is not the right word (though the objection will apply even more strongly to redeunt), in

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 142–3  103 fact cedere does not seem to be a full-­fledged verb of motion: ‘to withdraw, disappear’ rather than ‘to move away’ (it regularly occurs with an indication of the point of departure, but not of the destination or route, contrast e.g. abire). At 143, it seems clear that aureus orbis (cf. Man. 1.644) must be the subject, which means that we have to accept Phoebi (cod. Vat. lat. 3255; cf. e.g. Ov. Rem. 256 Phoebi pallidus orbis) and restore a verb in place of atque: Haupt’s redit conveys exactly the right sense (cedunt . . . inque uicem . . . redit). The one remaining weakness is currens, which is pointless and for which surgens (Mähly 1870, 782) is an easy improvement (cf. Juvenc. Evang. 1.652 spontanea cetera current, where C adds the variant surgent). For the passage thus restored, cf. Cic. Arat. fr. 34.336–7 sex omni semper cedunt labentia nocte, | tot caelum rursus surgentia [Cochanovius: fugientia codd.] signa reuisunt. 142 sidera . . . pallida  The adj. is to be taken predicatively, ‘dimmed’, cf. Stat. Theb. 2.120–1 iam pallida turbant | sidera lucis equi, 11.589–90 pallida turbet | astra [sc. Charon], 12.406–7 iam sidera pallent | uicino turbata die, Serv. GL 4.461.16 sidera pallida diffugiunt face territa luminis; the noun-­adj. word order may support this construal as it places the adj. close to the verb (note the Serv. quotation), and it is paralleled in a way at 143, where the sun is likewise named first (Phoebi) and then its colour specified (aureus orbis); it is no doubt significant that pallida and aureus occupy the same position, strengthening the contrast (‘the stars set, dimmed; the sun rises, golden’). The more usual adj.-noun word order has been hypothetically considered by van der Graaf 1945, 83; this could be the original order, but in that case we might need to look for an ­alternative to nitidum, as pallida per nitidum would juxtapose two ­morphologically similar adjectives, a feature largely avoided in poetry (Heyworth*). nitidum . . . mundum At the risk of oversimplification, the relevant difference between caelum and mundus seems to be that the former refers to the sky in the meteorological sense (and hence usually at daytime), the latter in the astronomical (and hence more often at night); both terms are etymologically controversial—­for plausible explanations, see EDL s.v. caelum (the word may have obtained ‘its meaning in the augural sphere, as “the whole” in contrast with templum “the part” ’) and

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

104  Commentary ON 142–3 Georgescu 2019 (pp. 219–20: ‘Lat. mundus underwent the same type of semantic change as Gr. χάος . . . from the original “yawning gap” . . . its usage became restricted to the upper vault or the subterranean vault’). With caelum the adj. means ‘bright, clear’, implying good weather (cf. OLD s.v. nitidus 1b); note e.g. Stat. Silv. 3.3.36 missuri nitido pia nubila caelo (of a fire). When defining mundus, the adj. seems to be used as a standing epithet meaning ‘radiant, starry’ (Man. 1.848 cum uaga per nitidum scintillant lumina mundum, Sen. Med. 402 nitidusque certas mundus euoluet uices, Mart. 10.28.1 annorum nitidique sator pulcherrime mundi); this is probably the sense the adj. has here, even if it may seem less apt in reference to a morning sky (cf. the Homeric idiom by which the sky can be ἀστερόεις even in daytime). I have wondered if it could have the pregnant sense of ‘brightening’ (cf. Mart. 8.21.9 tarda tamen nitidae non cedunt sidera luci), but I cannot find a parallel with a term for ‘sky’. cedunt: here ‘disappear, fade’, though the sense of ‘moving away, setting’ can also be present in astronomical contexts; cf. Cic. Arat. fr. 34.336 quoted 142–3n., Ov. Fast. 5.545–6 sidera mundo | cedere festinant. 143 inque uicem:  inuicem in tmesis, otherwise intractable in dactylic verse; cf. e.g. Verg. Georg. 3.188, 4.166, Aen. 12.502. Phoebi Poets of the first cent. bc use Phoebus of both Apollo and the Sun, but otherwise do not confuse the two gods (see Fontenrose 1939, esp. 451, and 1940); likewise here 143 Phoebi and 146 Phoebe are clearly linked, but not necessarily to be identified with one another; the implication is perhaps that a bucolic character might mix them up, though the association may be more sophisticated than appears at first sight (see 146n.). surgens redit: see 142–3n. for the text, and cf. esp. Cic. Arat. fr. 34.337 caelum . . . surgentia . . . reuisunt (quoted there in full). aureus orbis, ‘the golden disc’; cf. Man. 1.643–4 ubi se primis extollit Phoebus ab undis, | illis sexta manet, quos tum premit aureus orbis (when the sun rises for us, it is midday for the people of the east over whom the sun is in zenith). ‘Golden’ is a natural epithet of the sun, but note also contexts like Il. 5.509 Φοίβου Ἀπόλλωνος χρυσαόρου, which could easily lend themselves to allegorizing interpretation.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 144–6  105 144–5  The Moon is often apostrophized in amatory contexts, sometimes with explicit reference to her own affair with Endymion, cf. e.g. Philod. AP 5.123.5–6 ὀλβίζεις καὶ τήνδε καὶ ἡμέας, οἶδα, Σελήνη· | καὶ γὰρ σὴν ψυχὴν ἔϕλεγεν Ἐνδυμίων (with Sider 1997, 115). The theme may go back to Sappho (schol. Ap. Arg. 4.57–8 περὶ δὲ τοῦ τῆς Σελήνης ἔρωτος ἱστοροῦσι Σαπϕὼ καὶ Νίκανδρος), and the Moon’s inner monologue at Ap. Arg. 4.57–8 (οὐκ ἄρ᾽ ἐγὼ μούνη μετὰ Λάτμιον ἄντρον ἀλύσκω, | οὐδ᾽ οἴη καλῷ περὶ δαίομαι Ἐνδυμίωνι) no doubt mirrors this convention. 144 tuus . . . mea The use of substantivized possessive pronouns for ‘beloved’ is idiomatic, see OLD s.v. meus 2c, s.v. tuus 2d; cf. e.g. Hor. Carm. 1.25.7–8 me tuo longas pereunte noctes, | Lydia, dormis? with Nisbet and Hubbard 1970, 295. tecum est . . . non est . . . mecum Cf. 127 mea non mecum domina est with n. 145 Cf. schol. Theocr. 2.10 ὅτι καὶ ἡ Σελήνη περὶ ἐρωτικά τινα ἐνενοσήκει, παρὸ καὶ ταῖς τῷ αὐτῷ πάθει κεχρημέναις συμπράσσει. Hubaux 1930, 57 interestingly compares Mel. AP 12.70.4 οἶδα παθὼν ἐλεεῖν (Zeus speaking). dolor nosti quid sit Cf. Verg. Ecl. 8.43 nunc scio quid sit Amor (based on Theocr. 3.15 νῦν ἔγνων τὸν Ἔρωτα· βαρὺς θεός), Prop. 2.30.24 non nescit, quid sit amare, chorus, Ov. Met. 13.762 quid sit amor, sensit; on dolor, see 125n. miserere dolentis, ‘take pity on me in my suffering’: typically of the language of supplication, the object of misereri does not merely refer to the supplicant but also suggests a ground for granting the request, cf. e.g. Ov. Met. 14.12 diua, dei miserere (as in our case, the supplicant draws a parallel between himself and the deity addressed), further Verg. Aen. 4.435 miserere sororis, Ov. Met. 14.691 miserere ardentis, Sen. Phaedr. 620–3 tu qui iuuentae flore primaeuo uiges . . . miserere uiduae, Val. Fl. 2.256 exime nos sceleri, pater, et miserere piorum. 146  Apollo and Daphne is a rather questionable example of a happy relationship, given that it was never consummated and Daphne was transformed into a laurel so as to escape Apollo (as is explicitly admitted

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

106  Commentary ON 146 e.g. in Ov. Met. 1.557–8 cui deus: at quoniam coniunx mea non potes esse, | arbor eris certe, dixit, mea). gerens in te laurus celebrauit amorem (Ω) is problematic on more than one count, and correction remains uncertain. The clearest pro­ blem is that gerens, which can only apply to Apollo addressed here in the voc., is incompatible with third-­person celebrauit. One line of approach has been to replace gerens with an adj. that can be taken with laurus (interpreted as a nom. sing.), such as Naeke’s (1829, 152) recens; however, gerens laurus is an intrinsically plausible expression (cf. below), and the participle is additionally supported by 148 gestat. The alternative is to restore a second-­person form in place of celebrauit, and Scaliger’s (1572, 443) celebrabis is the obvious option. There remain two issues: first, and fundamentally, in te cannot mean ‘(wearing) on yourself ’ and is superfluous; second, bare future is not suitable for making a general statement. I suggest in te is an interpolated gloss, which has ousted semper (probably after rather than before laurus). As a potential alternative for the line end, one might think of e.g. semper laetaris amore (cf. Serv. on Verg. Ecl. 3.63 qui [sc. Hyacinthus] cum magis Apollinis amore laetaretur). gerens laurus: δαϕνηϕορεῖν is a proper term for a specific ritual practice (cf. in Latin Tib. 1.7.7 te uictrices laurus, Messalla, gerentem, 2.5.117 ipse gerens laurus, lauro deuinctus agresti), and δαϕνηϕόρος is an established title of Apollo, esp. frequent in inscriptions (cf. also e.g. Anacr. fr.  12.6 δαϕνηϕόροιο Φοίβου, Ov. Ars 3.389 laurigero sacrata Palatia Phoebo). Interestingly, there is evidence for the Sun being worshipped as δαϕνηϕόρος at Laurentum (Lydus Mens. 4.155 Ἀγωνάλια δαϕνηϕόρῳ καὶ γενάρχῃ Ἡλίῳ): Galinsky (1967, 626–7 and 633) suggests that the shared cult epithet may have contributed to the association of Sol and Apollo, and here we may have a learned allusion to this piece of antiquarian knowledge (cf. 143n.). semper celebrabis amorem The core meaning of the verb is here probably ‘To practise, exercise, perform’ (OLD s.v. celebro 5), but no doubt with a touch of ‘To honour with ceremonies’ (OLD s.v. 4); I cannot parallel it with amorem, but its usage appears flexible enough; with semper, cf. e.g. Verg. Aen. 8.76 semper celebrabere donis; for semper with the future conveying the idea of perpetuity, cf. Dirae 103 gaudia semper

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 146–8  107 enim tua me meminisse iuuabit, Ov. Am. 1.15.3 Battiades semper toto cantabitur orbe. 147–8 et quae pompa deum nisi siluis fama locuta est omnia uos estis secum sua gaudia gestat. The idea and articulation of the passage seem clear in general outline, but details are less so; in the order of increasing uncertainty: 148b is sound and supplies the predicate; 147a should provide the subject; 147b is a parenthetic remark of the kind of 129 fabula ni uana est; 148a is a crux. The subject of gestat should be ‘every god’, and e.g. Naeke (1847, 176) takes quae pompa deum to mean ‘qui deorum . . . numerus’. I see at least three objections: first, pompa ‘procession’ is not a plausible term to refer to a totality of (male) gods; second, it introduces a concrete image which undermines the universality of the statement (and, besides, it does not work as well with 149); and third, it implies a visual image and thus makes fama pointless. Schmidt’s (1853, 190) quicumque (Heinsius’ quaecumque, in Burman 1773, 673, may be cited as a pointer in the right direction) is necessary and lacks alternatives, and I think we should also write deus, to put slightly more emphasis on the gender; 147a thus comes to resemble 138 et mas quicumque est. The next step is to supply a direct object for locuta, and the main question is whether the clause is confined to 147 or continues into 148. If the latter, something along the lines of Ribbeck’s (1868, 176) nisi siluis fama locuta est | somnia [Heinsius in Burman 1773, 673] pro ueris could be considered. It seems more likely, however, that the clause is confined to 147 and that the missing direct object is concealed by siluis: Schmidt’s (1853, 190) lusus and Vollmer’s (1910, 76) uilia may be mentioned, but do not convince; I have also considered uanum and falsum, but fictum (fıctū → ſıluı?) seems preferable (cf. Sil. 6.554 uera ac ficta simul spargebat Fama per urbem). This leaves us with 148a, which can be treated only in a tentative way. An easy change would be to write nostis (ed. Ven. 1555) or scitis (Sillig 1826, 340) for estis, construing the phrase as a parenthetic apostrophe (cf. 130 Iuppiter, auertas aurem). One option would then be to take it as addressed to the readers, while perhaps also changing uos to iam (Heyworth*) to make the point clearer (‘you already know all these

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

108  Commentary ON 147–8 stories’). Alternatively it could be addressed to some eyewitnesses, but then we probably would also have to introduce a vocative in place of omnia, such as numina (Schmidt 1853, 190) or nymphae (Heyworth*); incidentally this would make testes (Scaliger 1572, 443) a further possibility for estis. More specific objections could be raised against individual variants, but a general weakness of this approach is that the passage thus becomes overburdened with meta-­references (it already has 147 nisi fictum fama locuta est and 149 quae dicere longum est). Perhaps we should look in a different direction; as 135–7 culminating in tecum tua laeta capella est suggest, 148b secum sua gaudia gestat may likewise presuppose a generalizing concessive clause (‘wherever he goes’), which we could try to restore for 148a. One conceivable option would be to replace uos estis with a participle, but the variants that come to mind do not seem plausible (e.g. perlustrans or uestigans). The alternative I cautiously embrace is to change omnia uos to omnibus, while supplying a local ablative for estis; in terris seems the easiest possibility (ī ťıſ → ẽ tıſ  ); for the expression, cf. e.g. Ov. Met. 5.439 omnibus est terris, omni quaesita profundo, Stat. Theb. 11.577–8 omnibus in terris scelus hoc omnique sub aeuo | uiderit una dies. It could be objected that earth is not the only (or even the most likely) place where gods can be found, but 149 likewise implies a terrestrial perspective, and terris supplies an apt counterpart to mundo. Finally, the possibility of even graver corruption cannot be excluded: e.g. Goebbel 1865, 21 posited that each hemistich is a fragment torn out of its context. More interestingly Heyworth* suggests that 147b–148a is an interpolation, the passage originally reading et quae pompa deum secum sua gaudia gestat (sc. semper celebrabit amorem): this is an elegant solution, esp. as it allows us to keep the transmitted quae pompa; I think, however, secum sua gaudia gestat must be the main clause: that the gods are inseparable from their beloved (sua gaudia) is not a premise but the central contention. 147 nisi fictum fama locuta est  Cf. 129 fabula ni uana est with n., add Drac. Laud. dei 3.214 si tamen hunc uerax per saecula fama locuta est; for the metaphor, cf. further e.g. Catull. 78.9–10 nam te omnia saecla | noscent et qui sis fama loquetur anus. 148  Apollo and Daphne (cited at 146) are the best-­known case of a god secum sua gaudia gestans, but there are numerous other aetiological

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 148–9  109 stories of a god’s beloved giving origin to a plant associated with that god, such as Pan and Pitys (see Naeke 1847, 178–81 for more examples). These stories, however, largely follow one of two scenarios, neither of which can be said to exemplify a happy lasting relationship: in one type the beloved is turned into a plant so as to avoid intercourse with the pursuing god (as in the case of Apollo and Daphne, or Pan and Pitys), in the other the beloved is transformed posthumously after some tragic accident cutting short the relationship (as e.g. in the case of Helios and Leucothoe or Clytie). omnibus in terris: ‘all over the world’, see 147–8n. above, and cf. 127 in terris with n. secum . . . gestat The frequentative (contrast 146 gerens) is already pointed, but it is further intensified by the addition of secum, almost suggesting physical contact (‘on him’ rather than merely ‘with him’, as one might carry a photograph or a lock of hair; cf. Mart. 12.17.3 gestatur tecum pariter pariterque lauatur [sc. tua febris]); the idiom is colloquial (in the centuries bc, only at Pl. Epid. 617 in manibus gestant copulas secum, Ps. 10 gestas tabellas tecum, Truc. 274 gestas tecum aenos anulos) and can imply affection (cf. Pl. Ps. 10 above, M. Aur. in Fro. Aur. 1.2.1 tuas litteras mecum gestabo, further Plin. Nat. 35.5 Epicuri uultus per cubicula gestant ac circumferunt secum); cf. also 127n. sua gaudia Cf. 123 mea gaudia with n. 149  The Pleiades are the best-­known and most straightforward case of a catasterism effected specifically to celebrate an affair with a divine lover, and probably are meant here in the first place (contrast e.g. Callisto, who was placed among the stars to avoid being killed in a hunt, or Ariadne, who had only her crown made a constellation). For their story, see Hellan. fr. 19a (schol. Il. 18.486) τὰς μὲν ἓξ θεοῖς συνελθεῖν· Ταυγέτην Διί, ὧν γενέσθαι Λακεδαίμονα· Μαῖαν Διί, ἀϕ᾽ ὧν Ἑρμῆς· Ἠλέκτραν Διί, ὧν Δάρδανος· Ἀλκυόνην Ποσειδῶνι, ὧν Ὑριεύς· Στερόπην Ἄρει, ὧν Οἰνόμαος· Κελαινὼ Ποσειδῶνι καὶ αὐτὴν συγγενέσθαι, ὧν Λύκος· Μερόπην δὲ Σισύϕῳ θνητῷ ὄντι, ὧν Γλαῦκος, διὸ καὶ ἀμαυρὰν εἶναι (with Fowler 2013, 415–18), cf. Eratosth. Cat. 23, Ov. Fast. 4.170–6 quae septem dici, sex tamen esse solent: | seu quod in amplexum sex hinc uenere deorum . . . septima mortali Merope tibi, Sisyphe, nupsit; | paenitet, et facti

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

110  Commentary ON 149–55 sola pudore latet; note that the one sister who had an affair with a mortal man rather than a god (Merope: note the name) was identified as the faintest star in the group, usually invisible to the naked eye (note uidet): this may additionally evidence the disadvantaged position of men as compared to the gods (on the number of the stars, see Kidd 1997, 277, and cf. Noris and Noris 2021 for cross-­cultural parallels, and further Ni’am 2020, 132–3). inspersa . . . mundo For the metaphor, cf. Cic. Arat. fr. 34.397 late caelo dispersa tenetur [sc. Hydra], Germ. Arat. 376–7 sunt etenim toto sparsi sine nomine mundo | inter signa ignes; Ov. Met. 11.309 nox caelum sparserat astris is somewhat different. dicere longum est: a metadiscursive tag, typical of rhetoric, cf. esp. Cic. Ver. 2.1.156 quas iste praedas, quam aperte, quam improbe fecerit, longum est dicere, Clu. 36, 107, Sest. 12, ND 1.30; in poetry, cf. Lucr. 4.1170 cetera de genere hoc longum est si dicere coner, Ov. Met. 5.207–8 nomina longa mora est media de plebe uirorum | dicere, 462–3 quas dea per terras et quas errauerit undas, | dicere longa mora est; in Greek, cf. e.g. Pl. Crit. 119b ἃ μακρὸς ἂν χρόνος εἴη λέγειν, though the more common version is μακρὸν ἂν εἴη λέγειν. 150–5  The preceding two stanzas adduced the animal and the divine worlds as the realms in which lovers can stay inseparably together (in the latter case, the contention is esp. strained, cf. 142–9n.). The present stanza argues that the same happy state of affairs existed for the humans of the ‘golden’ age, even if the examples it uses (Ariadne and Theseus, Medea and Jason) are not overly optimistic. Besides, they rather suggest the more specific idea of the permissibility of transgressive relationships (both couples eloped)—which will indeed be developed in the following stanzas. In fact, looking backwards, the reader realizes that the bull and esp. the he-­goat could more suitably serve as illustrations not of marital fidelity, but of unrestricted sexuality. The Lydia speaker’s view of the golden age as an erotic paradise is in fact not unproblematic. To begin with, in the mythological tradition the golden age is not associated with sexual freedom (rather, if at all, with asexual reproduction); the belief that the primitive man was sexually uninhibited belongs to philosophical (historical) discourse (cf. Campbell

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 150–5  111 2003, 348), and in Latin poetry appears to have been introduced by Lucretius (5.962–5): et Venus in siluis iungebat corpora amantum; conciliabat enim uel mutua quamque cupido uel uiolenta uiri uis atque impensa libido uel pretium, glandes atque arbuta uel pira lecta.

Gale (1994, 171) interestingly observes that the gifts of acorns, arbute berries, and pears with which the primitive man bought sex evoke the bucolic topos of fruits and flowers as erotic offerings (cf. Campbell 2003, 223–4). When Tibullus makes unrestricted sexuality a hallmark of idyllic pre-­agricultural society (2.3.69–74), he is careful not to brand it explicitly as the golden age, even if he refers to the present as the iron age (35 ferrea non Venerem sed praedam saecula laudant). What makes the association of sexual freedom with the golden age even more pro­ blematic here is that the Lydia speaker in fact refers not to the Hesiodic golden race, or to the primitive man in general (as would make sense in a bucolic poem), but to the age of heroes. The only prior attestation of the idea of the golden age as an era of erotic happiness comes from Theocr. 12, in which the speaker fantasizes about his relationship with his ἐρώμενος being remembered as an ex­ample worthy of the golden-­age men (15–16 ἀλλήλους δ᾽ ἐϕίλησαν ἴσῳ ζυγῷ· ἦ ῥα τότ᾽ ἦσαν | χρύσειοι πάλιν ἄνδρες, ὅτ᾽ ἀντεϕίλησ᾽ ὁ ϕιληθείς). As Gow (1952, 225) observes, there may be a connection with Arcesilaus’ anecdotal praise of Polemo and Crates—­his predecessors at the head of the Academy and famous lovers—­as λείψανα τῶν ἐκ τοῦ χρυσοῦ γένους (D.  L.  4.22, cf. Philod. Acad. 15.3–10). The exact point of the compliment is not clear (cf. Lampe 2013, 416–17), but it probably is intended to reflect Polemo’s ideal of living according to nature (as the golden-­age people of course did), and given the importance of erotics in the Academy at the time, it appears unlikely that love was not part of that ideal (one might speculate that homoerotic relationships could be thought to have flourished during the golden age, because there was then no need of sexual reproduction). This is a plausible source for the treatment of the theme in Theocr. 12, and it is not impossible that the

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

112  Commentary ON 150–5 Lydia derived it from Theocr. 12 in turn (note that both speakers express concerns about the posthumous fame of their erotic achievements). The golden age is sometimes also defined as the age when love is bought with gold (Antipater AP 5.31.1–4 χρύσεος ἦν γενεὴ καὶ χάλκεος ἀργυρέη τε | πρόσθεν, παντοίη δ᾽ ἡ Κυθέρεια τὰ νῦν· | καὶ χρυσοῦν τίει καὶ χάλκεον ἄνδρ᾽ ἐϕίλησεν | καὶ τοὺς ἀργυρέους οὔ ποτ᾽ ἀποστρέϕεται, Ov. Ars 2.277–8 aurea sunt uere nunc saecula: plurimus auro | uenit honos, auro conciliatur amor), but this strain of tradition is probably irrelevant here. 150 aurea . . . saecula  The speaker confuses the golden and the heroic ages (the first and the fourth in Hes. Op. 109–201), cf. 177n., or rather refers by that name to ‘the happy past’ in general. quin etiam: ‘(in weakened sense, adding a new point, a new item in an enumeration) yes, and—, and furthermore’ (OLD s.v. quin 3a). uoluebantur The metaphor properly implies the yearly return of the seasons and usually denotes the passage of time, cf. e.g. Verg. Georg. 2.401–2 redit agricolis labor actus in orbem, | atque in se sua per uestigia uoluitur annus (echoing Hes. Op. 386 περιπλομένου ἐνιαυτοῦ), Aen. 1.234 olim uoluentibus annis (echoing Od. 1.16 ἀλλ᾽ ὅτε δὴ ἔτος ἦλθε περιπλομένων ἐνιαυτῶν), 269–70 triginta magnos uoluendis mensibus orbis | imperio explebit; for the duration of time, cf. esp. Lucr. 5.931–2 multaque per caelum solis uoluentia lustra | uulgiuago uitam tractabant more ferarum, where the image is somewhat different but the reference is likewise to primeval times; note also Q. S. 2.506–7 περιπλομένοιο κατ’ ἦμαρ | νωλεμέως αἰῶνος ἑλισσομένων ἐνιαυτῶν. As in Lucr. (see Campbell 2003, 192–3), here the metaphor serves to convey the mea­ sured cyclicity of the course of time during the golden age, and the spondaic ending uoluebantur, in both sound and rhythm, reinforces the impression; the only parallels for a σπονδειάζων produced by a form of uoluere are Lucr. 5.971 circum se foliis ac frondibus inuoluentes [sc. the primitive people] (though 4.978–9 per multos itaque illa dies eadem obuersantur | ante oculos may be more similar in effect), and later Avien. Arat. 366 and 933 conuoluuntur, 1853 inuoluatur; however, in Ap. Arg. and esp. Arat. εἱλίσσοντο (-αι) is a frequent hexameter ending, often referring to the movement of celestial bodies.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 151  113 151  similisque foret mortalibus illis (Ω) has two main problems. First, the conjunction -que is superfluous; second, and related, foret is in the wrong tense. An old correction is similis fuerat (ed. Paris. 1540), which in all probability is right (fuerat was evidently first corrupted to foret, after which -que was inserted to amend the metre). Still, similisne [Herrmann 1949, 123] fuit [Ribbeck 1868, 176]—a rhetorical question—­ may be a viable alternative. Haupt’s (1874, 13) condicio similisque fuit mortalibus usus produces an implausible word order, but calls attention to a potential third problem: while illis is not obviously wrong, its deixis seems somewhat misplaced. The reader can infer from the context that the humans of the ‘golden age’ are meant, but illis implies an opposition between the past and the present generations, whereas the point is rather the similarity of the gods and the mortals, even if confined to a specific period. Herrmann (1949, 123) conjectured ipsis, but its force is inapposite (OLD s.v. ipse 9: ‘(to emphasize something regarded as exception or extreme) The very, himself ’): the speaker’s point is not that humans are as such all but incapable of happiness in love, but rather the reverse—­ the past proves that they could attain happiness, but the present-­ day circumstances prevent them. I tentatively suggest illic, ‘under those circumstances, then’; for the rare temporal use of the adverb, see TLL 7(1).373.14–40 (in reference to the past, cf. e.g. Var. Rus. 1.13.6 illic laudabatur uilla); an anaphoric illic referring to a preceding temporal clause is far from indispensable, but the added emphasis (with a touch of restrictive force?) may not be unwelcome. condicio similis For the noun (‘terms, status, position’), cf. also 155n.; for the phrase, which is notably prosaic and occurs often in various kinds of legal context, cf. e.g. Sen. Ep. 85.32 nisi dissimilis esset gubernatoris condicio et sapientis (‘situation, role’), Plin. Nat. pr.6 neque enim similis est condicio publicantium et nominatim tibi dicantium (‘position’), Ep. 10.56.5 si qui forte in simili condicione inuenirentur (‘legal circumstances’), Gaius Inst. 3.118 sponsoris uero et fidepromissoris similis condicio est, fideiussoris ualde dissimilis (‘legal status, obligations’). fuerat For the use of pluperfect for imperfect or preterite forms, cf. Heyworth 2007, 482–3 on such instances in Propertius: ‘Most of these fall into two categories: (a) the pluperfect of esse . . . ; (b) those with reference to the distant, mythical past’; our case belongs in both.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

114  Commentary ON 151–3 mortalibus: a poeticism with Homeric roots (calquing θνητός and/or βροτός: on their difference see Frère 2014, 138–9; cf. e.g. Il. 24.525–6 ὡς γὰρ ἐπεκλώσαντο θεοὶ δειλοῖσι βροτοῖσι | ζώειν ἀχνυμένοις), though also at home in philosophical discourse (note its frequency in Lucr., on both counts); particularly appropriate here as the context speaks of the humans of the heroic age, while also making a ‘scientific’ statement (note by contrast its infrequency in erotic elegy: only Prop. 2.27.1 [unless mortales is an adj.], Tib. 1.7.41 Bacchus et afflictis requiem mortalibus affert [evoking Il. 14.325 ἣ δὲ Διώνυσον Σεμέλη τέκε χάρμα βροτοῖσιν? note also Orph. Hymn. 51.16 σὺν Βάκχῳ Δηοῖ τε χάριν θνητοῖσι ϕέρουσαι, sc. Νύμϕαι], Ov. Am. 2.17.15 nymphe mortalis amore Calypso, Rem. 489–90 si quid Apollo | utile mortales perdocet ore meo, Her. 18.65 tu, dea, mortalem caelo delapsa petebas). 152–3 This praeteritio is more complex than is apparent at first glance. Clare (1997, 73–4) has discussed how Catull. 64.116–17 sed quid ego a primo digressus carmine plura | commemorem? pointedly echoes Jason’s suppression of details about Ariadne’s fate (Ap. Arg. 3.1096 ἀλλὰ τίη τάδε τοι μεταμώνια πάντ᾽ ἀγορεύω;), letting Medea believe that her affair with Theseus had a happy ending. Our passage primarily alludes to an earlier context in the same conversation between Jason and Medea, in which Ariadne is introduced for the first time (1001–4): σὺν τῷ ἐϕεσπομένη [scripsi: ἐϕεζ- codd.] πάτρην λίπε· τὴν δὲ καὶ αὐτοί ἀθάνατοι ϕίλαντο, μέσῳ δέ οἱ αἰθέρι τέκμωρ ἀστερόεις στέϕανος, τόν τε κλείουσ᾽ Ἀριάδνης, πάννυχος οὐρανίοις ἐνελίσσεται εἰδώλοισιν. Jason misleads Medea into thinking that the constellation of Corona Borealis was created by the gods to celebrate the love affair of Theseus and Ariadne, and the same idea must be implied in 152 notum Minoidos astrum. In the light of 149 inspersa uidet mundo, however, the reader cannot but remember the actual aition, which is also explained in Ap.’s Aratean intertext (71–3): αὐτοῦ κἀκεῖνος Στέϕανος, τὸν ἀγαυὸν ἔθηκεν σῆμ᾽ ἔμεναι Διόνυσος ἀποιχομένης Ἀριάδνης, νώτῳ ὑποστρέϕεται κεκμηότος Εἰδώλοιο.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 152–3  115 This ambiguity, and the intertexts it evokes, problematize Ariadne as an example of a happy erotic relationship between mortals (151) in the heroic age, and in so doing also alert the reader to the fact that Medea’s relationship with Jason proved in the end no more successful. 152 praetereo,  ‘I pass over, omit’: a common metadiscursive tag, with a rhetorical/prosaic ring; in poetry, cf. e.g. Verg. Georg. 4.147–8 uerum haec ipse equidem spatiis exclusus iniquis | praetereo atque aliis post me memoranda relinquo, Hor. Sat. 1.10.87–8 complures alios, doctos ego quos et amicos | prudens praetereo, Ov. Fast. 6.319–20 praeteream referamne tuum, rubicunde Priape, | dedecus? for familiarity of the topic as the reason for omission, cf. e.g. Cic. Ver. 2.3.106 audistis haec, iudices; quae nunc ego omnia praetereo et relinquo. notum: a reference to the literary tradition, cf. e.g. Prop. 2.24.45 (quoted 153n.), Ov. Ars 1.681–2 fabula nota quidem, sed non indigna referri, | Scyrias Haemonio iuncta puella uiro. Minoidos astrum: Corona Borealis, see 152–3n. The patronymic ­echoes Call. Aet. fr. 110.60 νύμϕης Μινωίδος (contrast Catull. 66.60–1 ex Ariadneis aurea temporibus | . . . corona); the noun for the constellation is lost (στέϕανος is a possibility), but note fr. 110.64 ἄστρον in reference to the Coma Berenices (besides Harder 2012, 833–4, see Massimilla 2010, 489–90). 153 quaeque For the relative clause (= et nota est quae), cf. Prop. 2.24.45–6 iam tibi Iasonia nota est quae uecta carina | et modo seruato sola relicta uiro est (with Heyworth 2007, 215–16). uirum uirgo For the double antonomasia calling attention to the gender roles, cf. esp. Ov. Her. 6.133 turpiter illa uirum cognouit adultera uirgo (likewise Jason and Medea), note too Call. Aet. fr. 67.1–2 αὐτὸς Ἔρως ἐδίδαξεν Ἀκόντιον, ὁππότε καλῇ | ᾔθετο Κυδίππῃ παῖς ἐπὶ παρθενικῇ (cf. Ov. Ars 1.682 quoted 152n.); the suppression of their proper names may also convey the idea that they were all but strangers, cf. Met. 4.681–2 primo silet illa nec audet | appellare uirum uirgo (Andromeda meeting Perseus for the first time), Prop. 2.34.8 Colchis et ignotum nonne secuta uirum est? (note ignotum, perhaps playing with 152 notum?). (Heyworth* suggests that uirgo could be an interpolated gloss on e.g. Aeetis or Colchis, but the allusive reference may be intended

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

116  Commentary ON 153–5 to make the reader pause and be reminded of the intertext in Ap. Arg. 3.1001.) sicut captiua Medea is compared to a captive (ληιάς) in Ap. Arg. 4.35–6, but the exact point of comparison is unclear; in Eur. Med. 256 she describes herself as ἐκ γῆς βαρβάρου λελῃσμένη, and one may wonder if the present context is intended to evoke a Latin adaptation. Here the point is that Medea placed herself completely under Jason’s power, without the protection of marriage; cf. Ov. Her. 3.69 uictorem captiua sequar, non nupta maritum: Briseis is willing to follow Achilles even as a slave. At the same time, there may be allusion here to the ritual abduction that was part of the Roman wedding ceremony, in imitation of the rape of the Sabines (see in general Hersch 2010, 69–75). secuta: women commonly follow men, cf. e.g. Eur. Tr. 946–7 ἅμ᾽ ἑσπόμην | ξένῳ, προδοῦσα πατρίδα καὶ δόμους ἐμούς, Ap. Arg. 3.1001 (quoted 152–3n.), 4.433–4 παρθενικῆς Μινωίδος, ἥν ποτε Θησεύς | Κνωσσόθεν ἑσπομένην Δίῃ ἔνι κάλλιπε νήσῳ, Mosch. Eur. 146–7 δῶμα | πατρὸς ἀποπρολιποῦσα καὶ ἑσπομένη βοῒ τῷδε, Catull. 64.180–1 ipsa reliqui [sc. patrem], | fraterna iuuenem respersum caede secuta, Sen. Med. 456 adulterum secuta per Symplegadas; and more generally, Verg. Ecl. 10.23 perque niues alium perque horrida castra secuta est. 154–5  ‘In what way has our age offended the gods?’ Perhaps the most relevant answer to the rhetorical question is given in Catull. 64.397–406, an intertext activated by the allusion to Ariadne at 152 (cf. Lorenz 2005, 22), note esp. 404–403 impia non uerita est diuos scelerare penates | ­ignaro mater substernens ilia [Trimble forthcoming: se impia codd.] nato (transposition after Barth 1664, 596), as well as by the echo of 64.386 in 154 (see n.). The speaker is of course no Oedipus, but he is naïve enough to ignore the established tradition that the modern age displays the greatest degree of moral decline (cf. 156n.). 154 laedere:  ‘to offend’, often in reference to gods, cf. e.g. Catull. 50.21 est uemens dea; laedere hanc caueto, Verg. Aen. 1.8 quo numine laeso, Hor. Epod. 15.3 numen laesura deorum, Prop. 2.15.48 laeserunt nullos pocula nostra deos, Tib. 1.9.5–6 parcite, caelestes: aequum est impune licere | numina formosis laedere uestra semel.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 154–5  117 caelicolae: an Ennian compound (Ann. 445 optima caelicolum, Saturnia), perhaps οὐρανίωνες, with a strong epic colouring (8x Verg. Aen., 2x Ov. Met., but otherwise unattested in the first cent. bc, except 3x in Catull.; remarkably not in Lucr.); here probably intended to evoke Catull. 64.384–6 praesentes namque ante domos inuisere castas | heroum et sese mortali ostendere coetu | caelicolae nondum spreta pietate solebant (cf. 154–5n.), though note also 30.4 nec facta impia fallacum hominum caelicolis placent (apparently based on Od. 14.83 οὐ μὲν σχέτλια ἔργα θεοὶ μάκαρες ϕιλέουσιν, cf. Lyne 2002, 602–3). nostra . . . aetas: the same as 178 minor aetas, ‘the post-­heroic age’; for aetas ‘the people of a particular era or age, a generation’, see OLD s.v. 8c, cf. TLL 1.1137.10–72. quid The postponement of the interrogative (naturally gravitating leftward) is conspicuous, but cf. e.g. Lucr. 1.619 ergo rerum inter summam minimamque quid escit? 155  may be echoed by Max. 1.113 o quam dura premit miseros condicio uitae (referring to old age). condicio . . . uitae, ‘status in, rules of, life’, is a prosaic idiom (cf. also 151 condicio with n.); cf. e.g. Cic. Fam. 5.16.4 a dis immortalibus ereptus ex his miseriis atque ex iniquissima condicione uitae, Phil. 14.33 pro mortali condicione uitae immortalitatem estis consecuti, Balb. 18 non grauior L.  Cornelio quam multis uiris bonis atque fortibus constitui lex uitae et condicio uideretur, Sall. Cat. 20.6 quae condicio uitae futura sit, nisi nosmet ipsi uindicamus in libertatem. durior, ‘disadvantaged’: while e.g. Quint. uses the positive form dura with condicio, all five occurrences in the Cic. corpus feature the comparative form, the implication apparently being that a condicio is dura as such (Clu. 150 quis umquam hoc senator recusauit ne, cum altiorem gradum dignitatis beneficio populi Romani esset consecutus, eo se putaret durioribus legum condicionibus uti oportere? Mur. 47, Rab. Post. 15, Plancus in Fam. 10.24.8, Brutus in ad Brut. 24.10 fateor enim duriorem esse condicionem spectatae uirtutis quam incognitae); the expression is esp. at home in legal contexts, cf. e.g. Gaius Dig. 4.7.3.2 si hominem quem petebamus manumiserit, durior nostra condicio fit, quia praetores faueant libertatibus, Ulp. Dig. 3.3.35.3 nec debebit durior condicio procuratoris

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

118  Commentary ON 155–83 fieri quam est domini, 40.5.24.21 ne contra uoluntatem defuncti durior eius condicio constituatur. In erotic poetry the adj. often means ‘unyielding’ (to the lover’s wishes), and can first be applied to a resistant beloved (e.g. Prop. 1.7.6 aliquid duram quaerimus in dominam, 1.17.16 quamuis dura, tamen rara puella fuit, 2.1.78 huic misero fatum dura puella fuit, Tib. 1.8.50 in ueteres esto dura, puella, senes, Ov. Am. 1.9.19 durae limen amicae), then to any concrete obstacles to the lover’s desires (e.g. Prop. 1.16.18 quid mihi tam duris clausa taces foribus? Tib. 1.2.5–6 nam posita est nostrae custodia saeua puellae, | clauditur et dura ianua firma sera, 2.6.47–8 saepe ego cum dominae dulces a limine duro | agnosco uoces, haec negat esse domi, Ov. Am. 2.1.22 mollierunt duras lenia uerba fores), and finally to more general circumstances restricting his amorous pursuits (Prop. 1.7.6 aetatis tempora dura queri, Tib. 1.6.69 mihi sint durae leges)—a connotation no doubt present here as well. 156–83  In the third section the speaker tries, and fails, to find a satisfactory answer to the question he asked throughout the second section (131–55): why cannot he enjoy his beloved’s company, like the animals (140–1), the gods (144), and the heroes (154–5)? The first stanza (156–64) formulates a hypothetical answer that might satisfy the speaker (had he been the first to engage in extramarital sex, such a punishment would have been justifiable), only to reject it and blame fate for not making it true. The central stanza (166–76) offers three mythological exempla demonstrating that the speaker was indeed not the first: Jupiter (166–8), Venus (169–74), and Aurora (175–6) had all seduced unexperienced partners long before he did. The final stanza (177–83) reiterates the question, but finds no better solution than to blame fate for his impending death. The speaker’s indignation is in fact based on the faulty premise, introduced in the second section, that the gods and the heroes had happy lasting relationships; the actual mythological references cited there can be seen to cast doubt over the speaker’s assumption (cf. esp. 142–9n., 150–5n.), and this is even more true of the exempla here: all three gods were serial adulterers, and many of the goddesses’ mortal lovers met a premature death, not unlike the speaker. 156–65  The preceding stanza closed with the tentative suggestion that in the modern age erotic relationships are censured in retribution for

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 156–65  119 some offence against the gods, which here the speaker develops by considering, and rejecting, the possibility that this offence consisted in the invention of ‘clandestine love’ (165). The idea that love, or more spe­ cif­ic­ al­ ly heterosexual extramarital relationships, had a πρῶτος εὑρετής seems otherwise unattested. The closest parallel comes from Phanocles (followed by Ovid, cf. 165n.), who makes Orpheus the pi­on­eer of pederasty in Thrace (fr. 1.7–10 Powell; see esp. Burges Watson 2014): τὸν μὲν Βιστονίδες κακομήχανοι ἀμϕιχυθεῖσαι   ἔκτανον, εὐήκη ϕάσγανα θηξάμεναι, οὕνεκα πρῶτος ἔδειξεν ἐνὶ Θρῄκεσσιν ἔρωτας   ἄρρενας, οὐδὲ πόθους ᾔνεσε θηλυτέρων. The Lydia speaker would be prepared, as it were, to share Orpheus’ fate if he could equal his achievement. His aspiration to be the first teacher of ‘clandestine love’ may also reflect the figure of Bion, whom the Lament presents as an expert in pederasty (83 παίδων ἐδίδασκε ϕιλήματα) and in whose own poetry the topic of erotic instruction may have featured prominently (note esp. fr. 10 Reed). It is no coincidence that Lydia too has links with both Orpheus and Bion (see intr., section 2), and is in fact a teacher of love (111). While the speaker is clearly right that he was not the first to enjoy sex outside marriage, it is tempting to contextualize his ambition to be the πρῶτος εὑρετής of uoluptas (163) within the framework of ‘the invention of sexuality’, a socio-­cultural process that Habinek 1997 places in the Rome of the Late Republic and Early Principate. This process consists in sex, and esp. sexual pleasure, becoming the centre of a new system of social values and relationships, and for Habinek the crucial witness to this emerging discourse of sexuality as an autonomous domain of human experience is Ovid (see also Ziogas 2021). In the Lydia, which makes gaudia (123, 148, 162, 168, cf. 169) and uoluptas (124, 163) into a leitmotif (note that gaudia is also what Lydia is remembered for at Dirae 103) and which emphatically confines sexual pleasure outside the institution of marriage, we perhaps can see a predecessor. At Ars 1.31–4:

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

120  Commentary ON 156–65 este procul, uittae tenues, insigne pudoris,   quaeque tegis medios, instita longa, pedes. nos uenerem tutam concessaque furta canemus,   inque meo nullum carmine crimen erit.—

Ovid can perhaps be seen taking pains to avoid the pit into which the Lydia speaker had fallen while trying to divest sex of its traditional social constraints. 156–7  Note the language of violence with which the speaker refers to his intercourse with Lydia: this might be seen to cast doubt over her willingness, but the passage rather draws a line between socially imposed constraints (which the speaker does violate) and the woman’s own wishes (which, we can infer, he respects); in fact, ‘violence’ is an established euphemism for ‘sex’, which can be used even when there is no coercion at play, cf. Adams 1982, 198: ‘A sexual act may be emotively spoken of as an act of violence or corruption, even if it is not regarded as such by its perpetrator (or even victim).’ In traditional society, rape is a matter of violence not so much against an individual as against social norms (thus, marital rape is often better tolerated than consensual sex outside marriage), and the Lydia (like much of Roman love poetry in general) strives to prioritize personal feelings and relationships. For the (socially imposed) notion that women should ‘defend’ their virginity even in marriage, cf. e.g. Catull. 66.13–14 dulcia nocturnae portans uestigia rixae, | quam de uirgineis gesserat exuuiis (of Ptolemy and Berenice on their wedding night); for the (socially subversive) notion that women enjoy forced sex, cf. e.g. Ov. Ars 1.673–6 uim licet appelles: grata est uis ista puellis: | quod iuuat, inuitae saepe dedisse uolunt. | quaecumque est ueneris subita uiolata rapina, | gaudet (though the different perspectives also reflect the difference in women’s social status). On sexual violence in Roman society and literature, see Fredrick 1997, Witzke 2016, DMALL 366–9 s.v. ‘riñas’, 453–8 s.v. ‘violación’. 156 ausus egon primus,  ‘did I dare to be the first to?’: the language suggests the idea of a πρῶτος εὑρετής, the inventor of love, which is indeed developed in the following lines; a central model may have been Cic. Arat. fr. 18.1–2 ferrea [an aerea?] tum uero proles exorta repente est |

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 156  121 ausaque funestum prima est fabricarier ensem (based on Arat. 130–1 χαλκείη γενεή, προτέρων ὀλοώτεροι ἄνδρες, | οἳ πρῶτοι κακόεργον ἐχαλκεύσαντο μάχαιραν), which appears also to be echoed at 177 (the Aratean intertext supports the speaker’s argument: the modern age invented war, not love); cf. further Lucr. 1.66–7 primum Graius homo mortales tollere contra | est oculos ausus primusque obsistere contra, Hor. Sat. 2.1.62–3 est Lucilius ausus | primus in hunc operis componere ­carmina morem. castos uiolare pudores: ‘to defile her virginal chastity’. The noun has here the specific euphemistic sense of ‘chastity’ (OLD s.v. pudor 2b; probably after more technical pudicitia, on which cf. Adams 1982, 195–6, note e.g. Liv. 2.7.4 ultor uiolatae pudicitiae, of Brutus as Lucretia’s avenger), apparently unattested earlier, cf. Tib. 1.3.83–4 at tu casta precor maneas, sanctique pudoris | assideat custos sedula semper anus, Verg. Aen. 4.27 ante, pudor, quam te uiolo (Dido fighting against her feelings for Aeneas), Ov. Met. 1.600 rapuitque pudorem (Jupiter raping Io), 6.616–17 quae tibi membra pudorem | abstulerunt (Tereus raping Philomela), Germ. Arat. 649 quoted 157n., Serv. on Verg. Aen. 8.646 uiolatus pudor (of Lucretia), Dion. Hal. 4.82.3 τὴν ἀμίαντον ἀϕαιρεθεῖσ᾽ αἰδῶ μετὰ βίας (ditto; cf. 160n.), cf. further TLL 10(2).2501.41–2502.11; on the concept, represented at 158 by concrete uittam (see n.), cf. further DMALL 345–8 s.v. ‘pudor’, 458–60 s.v. ‘virginidad’. The verb pro­ perly belongs in the religious sphere, and in poetry when it refers to transgressive sexual behaviour, it usually preserves its sense of ritual defilement, note Cic. Arat. fr. 34.420–1 Orion manibus uiolasse Dianam | dicitur (manibus shows that it does not directly refer to the sexual act), Catull. 67.23–4 sed pater ipse sui gnati uiolasse cubile | dicitur, Ciris 73 coniugium castae uiolauerat Amphitrites, as it does here; when it is used as a straightforward sexual euphemism (‘to seduce, corrupt, ruin’ rather than ‘to violate sexually’, pace OLD s.v. uiolo 2c; note esp. Var. L. 6.80 ‘uiolauit uirginem’ pro ‘uitiauit’ dicebant), it seems to have an ironic ring to it, perhaps reflecting its old-­fashioned (or prudish?) character, cf. Tib. 1.6.51 parcite, quam custodit Amor, uiolare puellam (the command is given by a priestess), Ov. Ars 1.375 quaeris an hanc ipsam prosit uiolare ministram? and 675 quaecumque est Veneris subita uiolata rapina, note also CLE 953 si quis forte meam cupiet uiolare puellam, | illum in desertis

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

122  Commentary ON 156–7 montibus urat amor (which has been suggested to be a quotation from Gallus, see Lee 1990, 132, also underlying CLE 954 crescens quisque meam futuet riualis amicam, | illum secretis montibus ursus edat, cf. Goold 1998, 23). 157 (uittam) temptare (Ω), ‘to make an attempt on’, is objectionable on two counts (uittam is Pomponius Laetus’ correction for Ω’s uitam, see Lanzarone 2018, 52). First, the verb merely denotes a (potentially unsuccessful) attempt (which would be anticlimactic after 156 uiolare), as well as implying unwillingness or even resistance on the girl’s part (cf. Tib. 1.3.73 Iunonem temptare Ixionis ausi). Second, although the headband may be a symbol of respectability (note Ov. Ars 1.31 quoted 156–65n., Pont. 3.3.51–2 nec uitta pudicos | contingit crines, and see in general Fantham 2008), it can hardly stand by metonymy for access to the ­woman’s body (as e.g. limina can at Tib. 1.2.18 iuuenis noua limina temptat; contrast Verg. Aen. 2.168 uirgineas ausi diuae contingere uittas, where uittae are an emblem of sacredness rather than chastity). (The physical sense, ‘to touch, feel’, on which cf. Shackleton Bailey 1956, 12, would be even less suitable.) Broukhusius’ (1708, 138) temerare is necessary (for the corruption, cf. temerare for temptare [Braun 1869, 13] at Val. Fl. 1.627; for pairing uiolare with temerare, cf. Ov. Met. 8.741–2 ille etiam Cereale nemus uiolasse securi | dicitur et lucos ferro temerasse uetustos; note also Fast. 6.457 quoted below). sacratam . . . uittam temerare OLD suggests that two distinct kinds of uitta (‘headband, ribbon’) existed in two distinct kinds of context: uittae worn by respectable women, both married and unmarried (s.v. 1), and uittae worn by priests (or supplicants), both male and female, in a ceremony (s.v. 2); sacratam would make one think of the latter rather than the former. The distinction appears not to have been watertight, though, as e.g. the Vestals’ uittae were associated with those worn by brides, being seen as a symbol of chastity (see Gallia 2014); the stripping of the uittae was part of the Vestals’ punishment for sexual transgressions, and Ovid makes defilement of the uittae a metonymy for the crime itself (Fast. 6.457–8 nullaque dicetur uittas temerasse sacerdos | hoc duce, nec uiua defodietur humo, cf. 3.29–30, with Gallia 2014, 234). Here the association appears to work in the opposite direction: even if she is not explicitly

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 157–8  123 presented as e.g. a follower of Diana, sacratam invites the reader to view Lydia as a quasi-­Vestal, her virginity being a question not only of decency but also of ritual; this is in accord, emotionally, with the speaker’s depicting his offence as punishable by death (note 158, though at 125–7 he appears to be dying of lovesickness). The parallel with Germ. 647–9 haec ego non primus, ueteres cecinere poetae, | uirginis intactas quondam contingere uestes | ausum hominem diuae sacrum temerasse pudorem is suggestive, though note that both Aratus’ and Cic.’s references to Orion’s assault on Artemis are less descriptive. Like uiolare, temerare primarily belongs in the ritual sphere; it is otherwise unattested prior to Verg. Aen. 6.840 templa et temerata Mineruae (though Norden 1927, 332 suggests that it is in fact an archaism), occurring next in poetry at Ov. Am. 1.8.19 thalamos temerare pudicos (for which see McKeown 1989, 211), though note also intemerata at Verg. Aen. 2.143, 3.178, and esp. 11.582–4 sola contenta Diana | aeternum telorum et uirginitatis amorem | intemerata colit, of Camilla (with McGill 2020, 206–7). 158  may be intended to evoke Theocr. 1.130 ἦ γὰρ ἐγὼν ὑπ᾽ Ἔρωτος ἐς Ἅιδαν ἕλκομαι ἤδη. immatura me(a)e [me S] cogor [quoque SFL] nece soluere fata [ facta SE]. With mea (Wernsdorf 1782, 20) for meae, the line yields a rea­son­ able sense: ‘Am I forced to dissolve my destiny, still unripe, with violent death?’ Taking immatura predicatively does not seem to be an issue (cf. Dirae 17 immatura cadant), but mea with nece is redundant: ‘Am I to cut my life short with my own death?’ (with whose else?). Ellis (1907, 5-­uiv.) saw that immatura can be split into iam matura, and Heyworth* proposes swapping matura and mea, which results in more plausible phrasing: ‘Am I fated to cut short (“disintegrate”) my destiny with speedy death?’; for matura . . . nece, cf. Ov. Her. 2.143 quoted below. The Ovidian parallel might also support reading furta (Heyworth*) for fata. matura . . . nece While maturus can mean ‘premature’ (OLD s.v. 9), in reference to death it rather means ‘speedy, undelayed, immediate’ (OLD s.v. 8, TLL 8.502.10–50); cf. e.g. Cic. Div. 1.36 maturam oppetere mortem (cf. breui tempore esse moriendum before), Luc. 9.834 uires maturae mortis habere (of a scorpion), CLE 995.18 ut te matura per Styga morte sequar (if genuine), [Tib.] 3.7.205 seu matura dies celerem properat mihi

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

124  Commentary ON 158–60 mortem, and esp. Ov. Her. 2.143–4 stat nece matura tenerum pensare pudorem: | in necis electu parua futura mora est (matura is explained by parua . . . mora), which likewise refers to paying with one’s death for love; for the idea, cf. also Hor. Carm. 3.7.15–16 casto Bellerophontae | maturare necem, which similarly speaks of a man being punished for a sexual transgression (if a falsely alleged one). cogor The passive of cogere often conveys necessity rather than coercion (OLD s.v. cogo 12b), suggesting the agency of fate; for cogor ‘I am fated’, cf. e.g. CLE 1005.9–10 hic ego nunc cogor Stygias transire paludes, | sedibus aeternis me mea fata tenent. For the sense of inevitability ­pro­d­uced in combination with iam, cf. e.g. Var. Ruf. fr. 156 Hollis iam fero infandissima, | iam facere cogor. soluere fata: ‘to undo fate, to change the destined course of life’ (for the metaphor, cf. Tib. 1.7.1–2 hunc cecinere diem Parcae fatalia nentes | stamina, non ulli dissoluenda deo, Serv. on Verg. Aen. 7.772 quae [sc. ars] etiam fata dissolueret), which can in principle mean both ‘to cut life short’ (as here) and ‘to avoid destined death’, though the expression seems otherwise unattested; comparable, and more common, is rumpere fata, which similarly can have both senses: (1) Luc. 2.107 infantis miseri nascentia rumpere fata, ‘to cut short the newly born destiny of a pitiful infant’, Val. Fl. 4.458–9 nec rumpere fata | morte licet, of Phineus, AL 83.93–4 crescentis Iuli | rumpere fata manu, CLE 1156.4 Parcae crudeles, nimium properastis rumpere fata mea; (2) e.g. Verg. Aen. 6.882–3 si qua fata aspera rumpas, | tu Marcellus eris. 159–60  The aspiration to be a teacher in matters erotic infringes on Ovid’s territory: Ov. Rem. 466 atque utinam inuenti gloria nostra foret may be an ironic echo, the more humorous as it refers to a strategy of dealing with unhappy love (which is the speaker’s problem here) rather than to any erotic achievement as such. For the form of the statement (an irreal wish followed by an assertion that death would have been welcome if it were fulfilled), cf. the paradoxical witticism from Neoptolemus’ Περὶ ἀστεϊσμῶν (fr. 6 Powell): ὦ θάνατ᾽, εἴθ᾽ εἴης αὐτάγρετος, ὄϕρ᾽ ἂν ἑλοίμην | πρώτιστος· καί κ᾽ ἐχθρὸς ἐὼν πολὺ ϕίλτατος εἴης; cf. further Hymn. Hom. 5.153–4 βουλοίμην κεν ἔπειτα, γύναι εἰκυῖα θεῇσι, | σῆς εὐνῆς ἐπιβὰς δῦναι δόμον Ἄϊδος εἴσω (Anchises to Aphrodite), and on the topos of ‘extreme wish’, see in general Gregory 2023.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 159–61  125 159 istius . . . facti  may be vaguely euphemistic; cf. Adams 1982, 204, on Juv. 6.271 occulti conscia facti: ‘factum is used suggestively of a misdemeanour of a sexual kind . . . , but in classical Latin the noun was not regularly specialised’. atque utinam For the postponement, cf. Caes. fr. 1.3 Courtney lenibus atque utinam scriptis adiuncta foret uis (with Courtney 1993, 154). culpa magistra The figurative use of magistra (-ter) applied to inanimate entities is common, cf. e.g. Cic. De orat. 2.36 historia . . . magistra uitae; for culpa used of sexual misconduct, cf. OLD s.v. 3b; for the phrase, one must cite Hor. Sat. 2.2.123 ludus erat culpa potare magistra, even if the text and interpretation are uncertain (see Konstan and Muecke 1992/3, 182–4; did Ofellus play ‘Never have I ever’?). 160  The sentiment sounds commonplace, but in fact close parallels are scarce. An archetypal model is Aesch. fr. 90 βίου πονηροῦ θάνατος εὐκλεέστερος, which brings out a certain lack of logic in the speaker’s statement, or rather a confusion of two ideas: ‘my life is so miserable that I should rather die’ and ‘I should gladly die if a certain condition is fulfilled’. The one parallel for ‘death sweeter than life’ in bc texts comes from Brutus’ apostrophe to Lucretia at Dion. Hal. 4.82.3: καὶ σοὶ μέν, ὅτι μίαν ἐτυραννήθης νύκτα τὴν ἀμίαντον ἀϕαιρεθεῖσ᾽ αἰδῶ μετὰ βίας, ἡδίων καὶ μακαριώτερος ἔδοξεν ὁ θάνατος εἶναι τοῦ βίου. In view of the similarity of language between the references to Tarquin’s rape of Lucretia and the speaker’s seduction of Lydia (cf. 156 with n.), it appears likely that Dion. Hal. and the Lydia engage with the same lost account of the rape of Lucretia (cf. perhaps also 1 ὥσπερ αἰχμάλωτος ὑπ᾽ ἀνάγκης κρατηθεῖσα and 153 sicut captiua). Just like the evocation of a Vestal at 157, the allusion to Lucretia at 156 (and here) associates the speaker’s offence with examples of an utmost sexual crime. Note that the line echoes 155. uita . . . dulcius: otherwise mostly used to express affection, cf. e.g. Cic. Fam. 14.7.1 nobis nostra uita dulcior est [sc. Cic.’s daughter], Catull. 68.106–7 uita dulcius atque anima | coniugium. 161  non mea non ullo [nullo SL] (Ω). Although it is difficult to prod­uce decisive objections against it, the first non is superfluous. On the one hand, the emphasis of a repeated non is unmotivated here. On the other, repeated non normally introduces coordinate constituents (e.g. Cons.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

126  Commentary ON 161–3 Liv. 234 non tibi, non ullis uincere fata datur; cf. OLD s.v. 1b); the one exception analogous to the present case is Germ. 647 non ego, non primus, ueteres cecinere poetae, where, however, emphasis would be more justified, but Gain (1976, 41 and 120) may in fact be right to print haec ego non primus. I adopt nam (cod. Ambros. O 74 sup.), which adds a welcome (if not strictly necessary) discourse connective. non ullo moreretur tempore fama The speaker’s logic evokes Achilles’ dilemma (Il. 9.412–16): he would have been content with an early death if he could secure eternal fame (cf. 413 κλέος ἄϕθιτον); but the phrasing rather renders another Homeric formula, κλέος οὔ ποτ᾽ ὀλεῖται (Il. 2.325, 7.91, Od. 24.196, of which the last, glorifying Penelope’s fidelity, may be the most relevant if subversive intertext; see in general Finkelberg 2007). Close Latin parallels seem to be wanting, but cf. Cic. Red. Sen. 3 quod enim tempus erit umquam cum uestrorum in nos beneficiorum memoria ac fama moriatur? Culex 362 fama uetus numquam moritura Camilli, CLE 659.3 nescit fama mori sed semper uiuit ubique (cf. 1376.16). For the aspiration to be immortalized for erotic achievements, cf. Theocr. 12.10–11 εἴθ᾽ ὁμαλοὶ πνεύσειαν ἐπ᾽ ἀμϕοτέροισιν Ἔρωτες | νῶιν, ἐπεσσομένοις δὲ γενοίμεθα πᾶσιν ἀοιδή. 162–3  The emphatic double formulation of the same idea over two lines is conspicuous and evokes 123–4, with dulcia . . . gaudia . . . dulcis . . .  uoluptas here echoing mea gaudia . . . mea . . . uoluptas there. 162  anticipates 168 gaudia libauit dulcem furatus amorem; cf. in general Pind. fr. 217 γλυκύ τι κλεπτόμενον μέλημα Κύπριδος, Anth. App. 71.2 κλεπτομένην χάρισαι Κύπριδος εὐϕροσύνην, Nemes. Ecl. 2.7 tum primum dulci carpserunt gaudia furto (of two young herdsmen raping a girl). dulcia . . . Veneris . . . gaudia The noun ‘commonly ha[s] a sexual implication’ (Adams 1982, 198; cf. TLL 6(2).1712.33–53, DMALL 323–6 s.v. ‘placer’; the usage lives on in Romance languages: Sampson 2009, 37–41), cf. 163 uoluptas with n., and often implies specifically orgasm (e.g. Lucr. 4.1106 iam cum praesagit gaudia corpus, Tib. 1.5.39 iam cum gaudia adirem, cf. Arist. GA 724a1 πρὸς τῷ τέλει ἡ χαρά; cf. further Fr. jouir and Sp. gozar), whereas, in erotic contexts, a purely psychological sense is rare (e.g. Prop. 2.25.30 in tacito cohibe gaudia clausa sinu). For the epithet, cf. Ov. Ars 3.798 dulcia mendaci gaudia finge sono, though of

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 162–3  127 course dulcis is frequent in reference to love; gaudia Veneris is otherwise Ovidian (Am. 2.3.2, Ars 2.459, 3.805; Met. 12.198 nouae Veneris . . . gaudia and [Tib.] 3.9.18 Veneris cupidae gaudia are somewhat different), but note also the Greek parallels quoted above, cf. too Ascl. AP 5.85.3 τὰ τερπνὰ τὰ Κύπριδος, anon. AP 7.221.1 ἡδέα Κύπριδος ἔργα. furatus While furtum is an established (if predominantly poetic) euphemism for illicit sex (OLD s.v. 2b, cf. Adams 1982, 167–8), and furtiuus likewise common (e.g. Catull. 7.8 furtiuos . . . amores), furari is rarer, and more vividly metaphoric: cf. Prop. 3.8.39 furandae copia noctis, Stat. Theb. 3.701–2 non egomet tacitos Veneris furata calores | culpatam­ue facem, cf. also the Greek parallels quoted above, add Philod. AP 5.120.1 τὸν ἐμὸν κλέψασα σύνευνον. 163  says largely the same thing as 162, but with one notable omission: while at 162 the speaker hypothetically aspires to be the first to have engaged in illicit sex, here he fantasizes about inventing sexual pleasure as such; the underlying idea is no doubt that sex is worthy of its name and a source of pleasure only outside marriage (also implicit in the examples of Venus and Aurora below), cf. Pind. fr. 217 quoted 162n., Ov. Am. 3.4.31 iuuat inconcessa uoluptas, Ars 2.687 quae datur officio, non est mihi grata uoluptas, 3.603 (on Ovid’s attitude, see Gibson 2022); cf. Witzke 2016, 265: ‘Officially, Roman marriages were business transactions, and Roman ­ideology dismissed the idea of “romance” in the married state. Roman wives existed for procreation, not pleasure.’ dicerer is pointed and must be taken literally (rather than as an inert periphrasis): ‘if I were spoken about, if I were a common tale’ (it thus brings out the etymological implication of 161 fama, ‘talk’); cf. esp. Hor. Carm. 3.30.10–14 dicar . . . princeps Aeolium carmen ad Italos | deduxisse modos. foret orta: pluperfect subjunctive of oriri. uoluptas, like gaudium, ‘commonly ha[s] a sexual implication’ (Adams 1982, 198; cf. also DMALL 323–6 s.v. ‘placer’); the two nouns are often paired, cf. e.g. Pl. Poen. 1217 gaudio ero uobis—­at edepol nos uoluptati tibi, Petr. 119.1.7–8 non uulgo nota placebant | gaudia, non usu plebeio trita uoluptas. While at 162 gaudia can in principle retain its ­literal meaning ‘(sexual) pleasures’, here uoluptas must be a euphemism

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

128  Commentary ON 163–4 for ‘sexual act’ as such (cf. e.g. Lucr. 4.1263 tractetur blanda uoluptas, ‘sweet sex is being engaged in’). 164  nam (Ω) cannot be right and is possibly a scribal reminiscence from 161. Modern editors tend to accept Hermann’s nunc (in Naeke 1847, 193), in the sense of OLD s.v. 11: ‘(introducing a fact or con­sid­er­a­ tion opposed to a previous speculation, wish, or sim.) (But) as it is; (often w. adversative conjs. or advs.)’; yet as the examples listed in the OLD suggest, nunc is used in the following kind of sentence: ‘if A, then B (counterfactual); (but since not A,) nunc C’; in other words, nunc introduces the actual consequence of a (desirable) condition not being the case; our passage, by contrast, is not concerned with consequences, but merely states that the condition was not fulfilled (‘if A, then B; but not A’). This seems to call for a more straightforward adversative: Cabaret-­Dupaty (1842, 302) proposed sed, but at seems to fit better (on the difference between at and sed, cf. Kroon 1995, 368: ‘in the case of at the author chooses a multiperspective presentation of the argumentation, in the case of sed a monoperspective presentation’; in other words, at would put more emphasis on the agency of fate, which is also consistent with the verb-­subject word order). impia uota (Ω) is no doubt a scribal reminiscence of Dirae 3 and 62. It seems clear that the noun should be fata (Parrhasius in Ramires 2005, 135); for fata with tribuerunt, cf. AE 2004.1266.3 nisi quot fata tribuerunt, 1396a.5 sic fata hoc tribue⟦t⟧run[t]. While impia fata is not an im­pos­ sible expression, it would be appropriate in a context speaking of fate actively depriving of something valuable (e.g. life: cf. CLE 566.1, 652.10) rather than merely denying an advantage (as here). Heinsius’ inuida (in Burman 1773, 675) appears therefore justified (esp. bearing in mind the likelihood of impia uota being interpolated as a phrase); cf. e.g. Phaed. 5.6.5 fato inuido, Luc. 1.70 inuida fatorum series, Stat. Theb. 10.384 inuida fata piis, CLE 465.2 inuida fata, 647.2 inuida fata. tantum tribuerunt: ‘bestowed so great an honour’, followed by an ut explicativum clause at 165; for the construction, cf. e.g. Prop. 2.1.17–18 quod mihi si tantum, Maecenas, fata dedissent, | ut possem heroas ducere in arma manus; with the same verb, Vell. 2.128.3 M.  Tullio tantum tribuere ut paene adsentatione sua quibus uellet principatus conciliaret,

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 164–76  129 Liv. 39.50.10 ab scriptoribus rerum Graecis Latinisque tantum huic uiro [sc. Philopoemeni] tribuitur, ut a quibusdam eorum . . . memoriae mandatum sit tres claros imperatores eo anno decessisse. The verb is avoided by Verg. and Tib., rare in Catull. and Hor., but frequent in Lucr. and Ov. inuida fata (if correct: see above) recalls the leitmotif of the first third of the poem: the speaker’s envy for the fields sharing Lydia’s company (104, 111, 123 inuideo), though now he becomes the object of envy himself. 165  rephrases 159. auctor: ‘the founder or originator (of a branch of knowledge, way of life, etc.)’ (OLD s.v. 14b); cf. Ov. Met. 10.83–4 ille [sc. Orpheus] etiam Thracum populis fuit auctor amorem | in teneros transferre mares; the metaphorical use of auctor predicated of noster . . . error parallels that of magistra predicated of mea culpa at 159 (for auctor used of inanimate objects, cf. TLL 2.1211.76–1212.52); at the same time, in its sense of ‘originator’ it refers back to 163 ex me . . . foret orta. occulti . . . amoris: ‘clandestine love’, with the implication that it is also illicit (on the concept, see in general DMALL 54–5 s.v. ‘[amor] furtivo’, and cf. 118 furtim and 162 furatus with nn.); for the phrase, cf. Prop. 2.33.7–8 tu certe Iouis occultis in amoribus, Io, | sensisti multas quid sit inire uias, Ov. Fast. 2.81 seu fuit occultis felix in amoribus index, Stat. Theb. 1.573–4 felix, si Delia numquam | furta nec occultum Phoebo sociasset amorem, Ach. 1.857 occultus amor. noster . . . error: a synonym for 159 mea culpa, though juxtaposed with amoris, error cannot but evoke ἔρως (cf. Verg. Ecl. 8.41 ut me malus abstulit error ‘translating’ Theocr. 3.42 ὣς ἐς βαθὺν ἅλατ᾽ ἔρωτα, with Kayachev 2021, 106); for error of a sexual transgression, cf. Ciris 240 laedere utrumque uno studeas errore parentem (of incest). 166–76 Strictly speaking, the mythological exempla adduced in this stanza—­ the affairs of Jupiter (166–8), Venus (169–74), and Aurora (175–6)—are intended to provide evidence that the speaker was not the first to engage in an extramarital relationship, but for his contention that he was treated unjustly to be valid it is also necessary to assume (as seems in fact to be implied) that, unlike him, these gods avoided punishment for their sexual transgressions. While this may indeed be true

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

130  Commentary ON 166–76 of Jupiter and Juno (who married in due course and lived happily ever after), Venus’ affair with Adonis had a tragic outcome, and none of Aurora’s could be considered happy in the long run either. The two latter exempla also highlight another respect in which the speaker’s view of his situation may depart from reality: while he imagines himself to be responsible for seducing Lydia, the precedents of Venus and Aurora suggest that she may not have been as innocent as he thinks her to be, which is in a way consonant with her prominence as a poet figure in the first section of the poem. 166–8  The passage merely claims that Jupiter had engaged in (illicit) sex prior to the speaker, but the implication (or at least intimation) may in fact be that he was the first person ever to do so (note the similarity between 162 and 168), the πρῶτος εὑρετής of ‘clandestine love’ (165). The reference to Zeus and Hera’s premarital intercourse, (in)famously alluded to in the episode of Διὸς ἀπάτη (Il. 14.295–6 οἷον ὅτε πρῶτόν περ ἐμισγέσθην ϕιλότητι | εἰς εὐνὴν ϕοιτῶντε, ϕίλους λήθοντε τοκῆας: note λήθοντε), does not just provide a close parallel to the speaker’s relationship with Lydia, but also caps the prior reference to Zeus’s affairs with Europa and Danae (129). In fact, the allusion subtly brings out a potential implication of the Homeric intertext, namely that Zeus’s initial seduction of Hera later served as a model and inspiration for his affairs with other goddesses and women (listed at Il. 14.315–28, cf. 129n.). This paradigmatic role of Zeus and Hera’s first sexual experience makes the passing reference to his subsequent escapades (166 semper mendacia fingens) less surprising and more pointed. 166 The claim that Jupiter ‘always disguised himself ’ is probably intended to include, besides his better-­known metamorphoses, the un-­ Homeric version of the myth in which, in order to seduce Hera, he turned himself into a cuckoo (cf. e.g. Paus. 2.17.4 κόκκυγα δὲ ἐπὶ τῷ σκήπτρῳ καθῆσθαί ϕασι λέγοντες τὸν Δία, ὅτε ἤρα παρθένου τῆς Ἥρας, ἐς τοῦτον τὸν ὄρνιθα ἀλλαγῆναι). mendacia factus (Ω), ‘made into disguises (of his own self)’, is im­plaus­ibly vague and forced; in place of factus, we need a participle that would take mendacia as a direct object (Petry’s [1895, 25] fassus, ‘he who confessed’, does not produce a fitting sense), and fingens seems the

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 166–8  131 easiest solution (for fingere mendacium, cf. e.g. Ciris 362 conficta dolo mendacia turpi). sui . . . mendacia,  ‘disguises of his own self ’: for this concrete sense, cf. impia dum Phoebi Caesar mendacia ludit (from an anonymous epigram referring to Octavian dressed up as Apollo, quoted in Suet. Aug. 70.1), Ov. Her. 8.67 non ego fluminei referam mendacia cycni (of Jupiter transformed into a swan; probably interpolated: see Knox 1995, 8–11), though I cannot parallel the genitive in the sense of ‘(disguise) of ’ rather than ‘(disguise) as’. 167 cum Iunone  For the force of cum, see OLD s.v. 13 (‘(in actions which require two participants) With’), and cf. esp. Pl. Am. 114 dum cum illa quacum uolt uoluptatem capit (of Jupiter with Alcmene), Mil. 320 osculantem atque amplexantem cum altero, Hem. fr. 32 cum eabus stuprum fecerat. coniunx . . . dictus uterque est: possibly ‘the two were pronounced husband and wife’, even if the usage does not seem quite idiomatic (cf. OLD s.v. dico2 10c, and note Ciris 414–15 sacrato foedere coniunx | dicta tibi, where dicta may rather mean ‘promised, betrothed’, Verg. Georg. 3.125 pecori dixere maritum, where dixere means ‘appointed’), or perhaps rather ‘became spouses’, as a calque of Greek καλεῖν ‘to be called, almost = εἰμί, esp. with words expressing kinship or status’ (LSJ s.v. καλέω II b; cf. Aen. 2.678 coniunx quondam tua dicta, based on Il. 3.128 τῷ δέ κε νικήσαντι ϕίλη κεκλήσῃ ἄκοιτις, and Ilias 100–1 quae coniunx dicor tua quaeque sororis | dulce fero nomen, based on Il. 18.365 σὴ παράκοιτις | κέκλημαι, the latter two contexts of Hera). For con­ iunx . . . uterque, cf. Seren. 612 si cubitum noctu coniunx festinat uterque. 168  echoes 162 (see n.). gaudia libauit The verb is a religious term for ‘to offer the first-­fruits of ’, and the idea of taking, so to speak, the first bite out of something lives on in many non-­technical usages (which OLD s.v. libo does not bring quite out); the closest parallels here are Ov. Her. 2.115 cui mea uirginitas auibus libata sinistris (where libata does not mean ‘was offered up’, pace Knox 1995, 133 and Barchiesi 1992, 159, but ‘was first touched, tasted, enjoyed’, cui being a dative of agent) and 16.161 mihi uirginitas esset libata (similarly misread by Michalopoulos 2006, 186, while

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

132  Commentary ON 168–74 Kenney 1996, 105 is more cautious), cf. Sil. 15.271 indelibata . . . uirgine, Copa 15–16 uirgineo libata Achelois ab amne | lilia, Sen. Contr. 1.2.12 quoted 114n., Val. Max. 6.1.4 praecepit ut non solum uirginitatem inlibatam, sed etiam oscula ad uirum sincera perferret, Serv. on Verg. Ecl. 10.57 ibi uirginitas Callistonis delibata sit; cf. further Stat. Theb. 10.63 pueri Iouis oscula libat [sc. Juno]. On gaudia, see 162n.; the lack of a possessive pronoun or genitive may suggest that gaudia has a more abstract meaning: ‘first tasted pleasure’, cf. Pl. Am. 114 quoted 167n., Nemes. Ecl. 2.7 quoted 162n., rather than ‘was the first to enjoy Juno’, cf. Ov. Met. 12.198 nouae Veneris Neptunus gaudia cepit. dulcem furatus amorem Cf. in general 162n.; add perhaps Anth. App. 71.5–6 καὶ γὰρ Ζεὺς θεὸς ἦν· ἀλλ᾽ ἡνίκα καιρὸν ἔκλεπτεν, | αἰετὸς ἢ δαμάλης ἢ κύκνος ἐβλέπετο. The participial construction bears the emphasis: Jupiter gained his first sexual experience through illicit intercourse. 169–74  In Greek bucolic Aphrodite is more than once cited as a goddess who did not consider an affair with a herdsman beneath herself: Theocr. 1.105–10, 3.46–8, 20.34–6. Here the point, at least on the surface level, is different (the relevant detail is that, like the speaker, Venus seduced an unexperienced lover), but the love story of Venus and Adonis plays a paradigmatic role in two further ways. On the one hand, the earlier depiction of Lydia reclining in the grass is phrased in such a way as to evoke the present account of Venus and Adonis having sex alfresco (see 112–22n.). On the other, the speaker’s dying from lovesickness parallels Adonis’ untimely death (see intr., section 2); though the tragic outcome of Venus and Adonis’ affair is suppressed here, note that it is mentioned in two out of the three Greek bucolic contexts (3.48, 20.36 ἐν δρυμοῖσι ϕίλασε καὶ ἐν δρυμοῖσιν ἔκλαυσεν [sc. Κύπρις Ἄδωνιν]: note the implied identity of Adonis’ love- and death-­bed). Finally, given that the reference to Venus’ lover is lost in the lacuna in front of 169 (see 169–71n.), there is a slight chance that the passage alludes to her affair with Anchises—­or in any event it constitutes a relevant mythological parallel, particularly significant in that they only had one rendezvous after which they were separated, just like the speaker and Lydia (at Hymn. Hom. 5.239–48 Aphrodite explains to Anchises why she cannot

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 169–74  133 have a mortal husband, and this might hint at one possible reason for the speaker’s being apart from Lydia; note also that while at Hymn. Hom. 5.155–67 Aphrodite has sex with Anchises on a bed of furs, Theocr. 1.106–7 places their encounter in a locus amoenus). 169–71 et mecum [mea cum M] tenera gauisa est ludere in herba purpureos flores quos insuper accumbebat, grandia formoso supponens gaudia collo. The lines refer to Venus’ affair with someone other than Mars, in all probability Adonis (cf. Ov. Her. 4.97–8 saepe sub ilicibus Venerem Cinyraque creatum | sustinuit positos quaelibet herba duos). The central problem is that neither Venus nor her lover is named, and it is not immediately clear where references to them can be restored. Before we address it, one minor issue can be dealt with: 169 ludere in all likelihood conceals a verb that has 170 purpureos flores as its direct object (cf. 117 teneramque illiserit herbam). Canter’s (1571, 340) laedere is usually adopted, but the unprefixed form seems to have the wrong emphasis: when used of plants (TLL 7(2).867.21–32, 868.2–15), it rather refers to the effect (‘to damage’, cf. Verg. Aen. 7.809 nec teneras cursu laesisset aristas) than to the physical action as such (‘to crush’). (Scaliger 1572, 445 proposed lidere, a by-­form potentially attested at Lucr. 5.1001, but its advantages are not obvious.) Gronovius conjectured elidere (in Arnold 1652, 264), but the verb seems to denote intentional (and intensive) rather than accidental ‘crushing’ (note e.g. Ov. Fast. 4.371 elisae miscetur caseus herbae, Sen. Her. F. 221 tumida tenera guttura elidens manu, Nemes. Ecl. 3.42–3 celerique elidere planta | ­concaua saxa super properant [sc. uuas]). What we need is illidere (Heyworth*), the verb used in a similar context at 117 (it can account for the transmitted est ludere as well as, if not better than, the alternatives: ī → ẽ, lidere → ludere); we probably should add est elsewhere, either after tenera (László 1903, 376), or perhaps at the end of the line. Now the main problem. One way of supplying a reference to Venus has been to write Cypria at 171, in place of either grandia (Putsche 1828, 117) or gaudia (Naeke 1847, 202), with bracchia instead of the other. A reference to her lover would then be restored at the beginning of 169:

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

134  Commentary ON 169–71 e.g. et mare cum (Lachmann 1850, 324), et moechum (Baehrens 1880, 82), et tecum, Anchisa (Purser 1909, 453). There are at least three objections. (1) The corruption of Cypria and bracchia to grandia and gaudia is implausible palaeographically. (2) From a stylistic point of view, the optimal way to restore 171 would be to write bracchia (ed. Ald. 1517) for gaudia, while supplying an epithet for bracchia—­candida (ed. Argent. 1527) is the likeliest—­instead of grandia, cf. Catull. 64.332 quoted 171n. (this would also be easier to explain in terms of palaeography: grandia is a fairly easy error for candida, whereas gaudia is a further corruption of grandia that has ousted, one way or another, bracchia; cf. Max. 5.33 grandia, clamabat, tua nunc me bracchia [grandia Fe] laedunt). (3) None of the proposed ways to supply a reference to Venus’ partner in 169 is convincing. Other scholars have attempted to introduce both lovers in 169, but this inevitably results in extensive rewriting: e.g. et dea cum tenero gauisa est ludere Adoni or et dea cum Cinyrae gauisa est ludere ephebo (both by Wernsdorf 1782, 21), et dea cum puero gauisa est ludere Myrrhae (Saenger 1885, 42). The conclusion seems warranted that there is a lacuna before 169: Heinsius thought ‘deesse forsan nonnulla’ (in Burman 1773, 676), but I see no reason to think that it exceeded one line. In view of the lacuna, any attempts to restore the head of 169 are bound to be more than uncertain, but some guesswork may not be entirely pointless. Assuming et is correctly transmitted, mecum [mea cum M] could well conceal a reference to Adonis: it ought to be made clear that what Venus enjoyed was not crushing ­flowers as such, but doing it in Adonis’ company (cf. the more explicit account at Ov. Her. 16.203–4 Phryx etiam Anchises, uolucrum cui mater Amorum | gaudet in Idaeis concubuisse iugis). If that line of reasoning is valid (though of course such a reference could have been made in the lost line), Gronovius’ tecum (in Arnold 1652, fol. 12v. and p. 267) seems an obvious, perhaps the only, option (cf. in a way 167–8 cum Iunone . . .  gaudia libauit), which will imply that the lacuna contained an apostrophe to Adonis (cf. esp. the addresses to the Moon and to Apollo in 144–6). Yielding to horror vacui, and merely exempli gratia, I offer ipsa [ante?] Venus pariter furtis indulsit, Adoni (could the omission have been triggered by the scribe’s eye jumping down from 168 amorem to adoni? or perhaps adoni was first corrupted to amori?).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 169–71  135 The image of Venus and Adonis having sex on a flowery meadow goes ultimately back to Il. 14.346–9, narrating the successful outcome of Hera’s seduction of Zeus: ἦ ῥα καὶ ἀγκὰς ἔμαρπτε Κρόνου παῖς ἣν παράκοιτιν· τοῖσι δ᾽ ὑπὸ χθὼν δῖα ϕύεν νεοθηλέα ποίην, λωτόν θ᾽ ἑρσήεντα ἰδὲ κρόκον ἠδ᾽ ὑάκινθον πυκνὸν καὶ μαλακόν, ὃς ἀπὸ χθονὸς ὑψόσ᾽ ἔεργε. Ovid too alludes to this Homeric passage (Her. 4.98 sustinuit ~ ἀπὸ χθονὸς ὑψόσ᾽ ἔεργε), as does Virgil, in a somewhat different context (Ecl. 6.53 ille [the bull] latus niueum molli fultus hyacintho: likewise, the striking detail of a flower—­ὑάκινθος is a Homeric hapax—‘supporting’ a reclining body). A more extended, and explicit, allusion appears in Petronius’ account of the erotic rendezvous between Encolpius and Circe (127.9): Idaeo quales fudit de uertice flores terra parens, cum se concesso iunxit amori Iuppiter et toto concepit pectore flammas: emicuere rosae uiolaeque et molle cyperon, albaque de uiridi riserunt lilia prato— talis humus Venerem molles clamauit in herbas, candidiorque dies secreto fauit amori.

It has been suggested that Petronius’ cyperon is an echo of Theocr. 1.105–6 ἕρπε ποτ᾽ Ἴδαν, | ἕρπε ποτ᾽ Ἀγχίσαν· τηνεὶ δρύες ἠδὲ κύπειρος (Connors 1998, 41–2, Setaioli 1999, 252–3). Note also that, like Venus in our passage, Circe takes the leading role in Petronius (127.8 implici­ tumque me bracchiis mollioribus pluma deduxit in terram uario gramine indutam). 169 gauisa illidere The phrase is allusively euphemistic: Venus, one assumes, did not as such enjoy crushing grass and flowers (contrast Hor. Sat. 1.4.78 laedere gaudes), but rather enjoyed the activity that led to the grass and flowers being crushed (cf. Ov. Her. 16.203–4 quoted 169–71n.,

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

136  Commentary ON 169–70 Met. 10.557–9 quoted 116–17n.); for gaudere used of sensual (esp. sexual) pleasure, cf. Catull. 68.103 Paris abducta gauisus . . . moecha, ­ 125–6 nec tantum niueo gauisa est ulla columbo | compar, Ov. Ars 1.295 Pasiphae fieri gaudebat adultera tauri, 2.719 cum loca reppereris, quae tangi femina gaudet, somewhat further Prop. 1.3.23 gaudebam lapsos formare capillos, and note 162 gaudia with n. (though the verb is not in Adams 1982). At the same time, the suggestion is present that crushing flowers is itself an erotically tinted activity, as at 116–17 (see n.). 170 purpureos flores  One might expect ‘purple flowers’ to be ubi­qui­ tous in ancient poetry, but while there are a few parallels in Latin (Verg. Georg. 4.54, Aen. 5.79, 6.884, 9.435, 12.413–14, Hor. Carm. 3.15.15, Ov. Fast. 5.363), in Greek Sappho fr. 105c οἴαν τὰν ὐάκινθον ἐν ὤρεσι ποίμενες ἄνδρες | πόσσι καταστείβοισι, χάμαι δέ τε πόρϕυρον ἄνθος seems to be the only context, until well into the imperial period; the simile probably comes from an epithalamium and alludes to defloration: this would make a fitting intertext here, as in both cases crushing ‘purple flowers’ has erotic significance, and the reference to ποίμενες ἄνδρες might have provided additional incentive to a bucolic writer (the fragment is also believed to lie behind Catull. 11.22–4 and Aen. 9.435 purpureus ueluti cum flos succisus aratro: see Harrison 2020, 35–7, Thomas 2021, 60–1). In our context, though, flowers may rather have phallic symbolism, since it is Adonis who is the virginal figure, while Venus takes the initiative; cf. Ov. Am. 3.7.65–6 nostra tamen iacuere uelut praemortua membra | turpiter hesterna languidiora rosa, Petr. 132.11 lassoue papauera collo with Adams 1982, 26. accumbebat The singular might suggest that Venus was lying on the grass alone rather than with Adonis (note that both TLL 1.341.18–19 and OLD s.v. accumbo 2b take the verb here in its literal sense ‘to lie’ rather than as a sexual euphemism); accumbebant (cod. Monac. 21562) might thus seem attractive. One reason for keeping the singular is that the participial phrase of 171 would make better sense if construed with accumbebat than with 169 gauisa illidere (‘she enjoyed crushing flowers as she put her arms under his neck’ is not quite what the text should be saying). Another is that it is questionable whether the plural accumbebant can have the pregnant sense ‘to have sex’ (no parallels, though there

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 170–1  137 are only a handful of contexts). The verb properly means ‘to lie beside’ rather than ‘to lie down’, and as a sexual euphemism it occurs in two constructions: with a direct object in the sense ‘to lay beside oneself ’, the subject being the male (Pl. Men. 476 scortum accubui, Bac. 1189 scortum accumbas; Apul. Fl. 14 duxit Cynicus in porticum; ibidem . . . accubuit [sc. uirginem] implies the same construction), or with a dative in the sense ‘to lie beside’, the subject being the female or the ἐρώμενος (Tib. 1.9.75 huic tamen accubuit noster puer, Prop. 2.3.30 Romana accumbes prima puella Ioui, 3.15.12 Antiopen accubuisse Lyco); the latter construction is implied in the elliptical usage at Prop. 2.32.35–6 quamuis Ida deam pastorem dicat amasse | atque inter pecudes accubuisse deam (Venus with Anchises), which is the closest parallel for our context, with inter pecudes matching quos insuper in a way (Ov. Fast. 3.526 accumbit cum pare quisque sua refers not to sex but to feasting, pace OLD s.v. accumbo 3a, hence the different construction, the implication of accumbere being ‘at the table’); in other words, it is always the female who lies beside the male, not the other way round (cf. succumbere, which is likewise gender-­ specific, and note in general that Latin tends to differentiate between the male and the female role more consistently than modern English does). 171  Venus is reclining over Adonis lying on his back, her arms around his neck, cf. Catull. 64.332 leuia robusto substernens [Fisch 1875, 25: s. r. codd.] bracchia collo (pointedly contrasted with 403 ignaro mater substernens ilia [Trimble forthcoming: se impia codd.] nato: the underlying idea is probably that the one on top is giving pleasure, the one under receiving) and Ov. Am. 3.7.7–8 illa quidem nostro subiecit eburnea collo | bracchia Sithonia candidiora niue. These need not imply the mulier equitans position (on which cf. Heath 1986, 32–5, Davidson 1997, 196–7, Barchiesi 2006, 109–10), as they may not refer to intercourse as such, but it is important that the woman takes the leading role (this is particularly clear in the case of Ov. Am. 3.7, in which the male protagonist suffers from impotence); for the position, cf. perhaps the right-­hand fresco from the north-­east antechamber wall of room D of the Casa della Farnesina (Museo Nazionale Romano, Palazzo Massimo, inv. no. 1188; reproduction in Levin-­Richardson 2019, plate IX; Clarke 2013, 149–50 interprets the couple as a newly wed husband and wife, which would be

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

138  Commentary ON 171–4 a close match for the Catullus context; cf. also Valladares 2021, 42–51 and plate 4). Contrast Repos. 109–11 stringebat Paphiae Mauors tunc pectore dextram | et collo innexam ne laedant pondera laeuam | lilia cum roseis supponit candida sertis, where Mars is caressing Venus (Lucr. 1.33–9 seems rather to describe a post-­coital position, cf. Edmunds 2002). candida: Ov. Am. 3.7.8 (quoted above) is no doubt intended to cap this. formoso is a suitable adjective to apply to Adonis (contrast Peleus’ heroic robusto . . . collo at Catull. 64.332), cf. Sappho fr. 140.1 ἄβρος Ἄδωνις, Theocr. 1.109 ὡραῖος χὤδωνις, ἐπεὶ καὶ μῆλα νομεύει, with Verg. Ecl. 10.18 formosus ouis ad flumina pauit Adonis, Bion Ep. Adon. 1 ἀπώλετο καλὸς Ἄδωνις, 79 κέκλιται ἁβρὸς Ἄδωνις. supponens: ‘to place beneath so as to support’ (OLD s.v. suppono 1c), cf. esp. Prop. 3.7.69 uos decuit lasso supponere bracchia mento. bracchia collo The juxtaposition of the two words is common in ­references to hugging (e.g. Verg. Aen 2.792 [= 6.700] collo dare bracchia circum, Ov. Her. 16.167 [= Met. 9.605] circumdare bracchia collo; note Virgil’s tmesis: just as dare . . . circum encircles bracchia rather than collo, so Aeneas ends up hugging himself instead of his wife or father), which makes it easier for the reader to infer here that Venus puts her arms under Adonis’ neck by embracing him. 172–4  The explanation of why Vulcan and Mars did not interfere in Venus’ affair may sound merely humorous in its evocation of Demodocus’ song in Od. 8 (where note the key term λάθρῃ in 269–70 ἐμίγησαν ἐν Ἡϕαίστοιο δόμοισι | λάθρῃ, for which cf. 118 furtim with n.), but a more sombre reading is suggested by the version of the myth in which Adonis was killed by Ares out of jealousy (possibly alluded to at Prop. 2.13.53–4: see Heyworth 1992, 57–8). The bT scholia on Il. 5.385 explain Ares’ imprisonment by Otus and Ephialtes as a punishment for killing Adonis whom Aphrodite entrusted to the brothers for protection: Ἀδώνιδος τοῦ Κινύρου ἐρασθεῖσα Ἀϕροδίτη καὶ τὸν ϕθόνον ὑϕορωμένη παρέθετο τοῖς περὶ Ὦτον καὶ Ἐϕιάλτην . . . τοῦτον θηρολετοῦντα ἐν τῷ Λιβάνῳ τῆς Ἀραβίας ἀναιρεῖ Ἄρης. οἱ δὲ ὀργισθέντες ἐπὶ τρισκαίδεκα μῆνας ἀποκλείσαντες τὸν Ἄρεα εἶχον ἐν εἱρκτῇ (Homer may have implied this story: see Cassio 2012, 422–3, Margulies 2020, 497).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 172–4  139 One might speculate that the idea of Mars being ‘detained’ at a campaign could be intended to evoke the story of his imprisonment. A slightly different account is preserved in a Syriac text attributed to Melito of Sardis: ‘before Tamuz [Adonis] she was in love with Ares, and committed adultery with him, and Hephaestus her husband caught her, and [Ares] was jealous over her, and came and slew Tamuz in Mount Lebanon, while he was hunting the wild boar’ (tr. Brown 1995, 245, adapted; the implication is apparently that, after the disgrace of being caught in flagrante, Aphrodite left Ares for Adonis); note that it brings together all three of Aphrodite’s partners in one context, just as the Lydia does. 172  The passage appears to be alluded to in Ov. Am. 1.8.41–2 nunc Mars externis animos exercet in armis, | at Venus Aeneae regnat in urbe sui, on which cf. McKeown 1989, 222: ‘Dipsas implies that, while the armies are on campaign, the soldiers’ wives are living decadently in Rome.’ credo: ‘doubtless, I suppose, presumably’ (OLD s.v. 8c), a conversational tag ubiquitous in comedy but much less frequent in higher-­ register poetry (Catull. 2.8, 84.5, Verg. Ecl. 3.10, Aen. 1.387, 6.368, 7.297, 10.29, 865, Hor. Sat. 2.2.90, 3.88, 5.87, 7.68, Ov. Met. 9.611, Pont. 1.7.56; the avoidance in erotic elegy might suggest an archaic and/or vulgar colouring). fuerat . . . distentus, ‘was kept away’: a pluperfect with a stative force, cf. Pinkster 2015, 458, Wistrand 1958, 40–2; on the use of a pluperfect auxiliary in pluperfect passive forms, see Pinkster 2015, 473–6. The verb is frequent in military contexts, esp. in reference to forces being tied up on different fronts (TLL 5(1).1523.7–19), and some such usage is implied here: Mars is held up by a campaign at the outer frontiers, so that he is unable to keep a check on his own mistress at home. in armis: ‘under arms, on military service, mobilized’ (OLD s.v. arma 5e); cf. esp. Verg. Ecl. 10.44–5 nunc insanus amor duri me Martis in armis | tela inter media atque aduersos detinet hostis. 173  For Vulcan working in the forge while Venus is enjoying herself, cf. Hor. Carm. 1.4.5–8 iam Cytherea choros ducit Venus . . . dum grauis Cyclopum | Vulcanus ardens uisit officinas.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

140  Commentary ON 173–4 nam certe The force of nam cannot be strictly causal, it rather introduces background information more generally (on this function of nam, see Kroon 1995, 147–8): ‘I assume (credo) that Mars was at a war, given (certe) that Vulcan was at work’ (i.e. in Vulcan’s absence Mars would normally be with Venus, but since he was not, he must have been occupied with something else); possibly with a hint that it is adduced by way of an afterthought (see Kroon 1995, 155–8): ‘as for Vulcan, he certainly was at work’. For a parallel of a sort (if a more straightforward one), cf. Apul. Met. 9.22 pudica uxor . . . mensam largiter instruit; denique, ut dei cuiusdam aduentus, sic expectatur adulteri. nam et opportune maritus foris apud naccam proximum cenitabat. (Heyworth* suggests writing tum for nam, but the anaphora of tum does not seem to me to work well with the change from credo to certe.) opus faciebat, ‘was working’: the idiom appears to be a colloquial and/or technical term for manual labour, cf. e.g. Ter. Eun. 220 opus faciam, ut defetiger usque (‘I’ll work at the farm until I’m exhausted’), Cato Agr. 56.1 qui opus facient (‘farm labourers’), Var. Rus. 3.2.11 in ­aluariis opus faciunt (‘the bees toil in the hives’), Catull. 68.50 in deserto Alli nomine opus faciat (of a spider weaving its web), Ov. Met. 14.268 quod hae faciunt opus exigit (of Circe overseeing the nymphs’ work); opus can be construed concretely (Var. Rus. 3.16.5 intus opus faciunt, quod dulcissimum . . . est, of honey, 20 si opus quod faciunt est aequabile ac leue, of the comb), but this is probably not how it should be taken here. 174 (et illi) tristi turpabatque [turba- MS] mala fuligine barbam (Ω). Scaliger’s (1572, 445) turpabat malas fuligine barba successfully solves two out of the three more obvious problems: it eliminates an excessive connective (-que) and an otiose and inapposite adjective (mala), but fails to supply a plausible subject (‘the beard soiled the chin with soot’ is absurd). Alton (1922, 314) proposed carbo for barbam: this is not im­pos­ sible, but we should probably keep the reference to Vulcan’s beard (cf. below), which malas cannot convey (malae can refer to the part of the face where the beard grows but not to the beard itself: cf. TLL 8.160.57–66, and note e.g. Lucr. 5.674 mollem malis demittere barbam). Giardina (2009, 172–3) conjectured flammae for mala (with turpabant), which has the advantage of retaining barbam, but flame could rather be

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 174  141 said to scorch than merely to soot the beard (cf. Verg. Aen. 12.300–1 occupat os flammis: olli ingens barba reluxit | nidoremque ambusta dedit). I believe we should follow Giardina’s approach, but write massae instead (this would be an easier error, esp. given the relative infrequency of the term in verse); for the workpiece as a source of soot (as well as for the image in general), cf. Juv. 10.130–2 quem [sc. Demosthenes] pater ardentis massae fuligine lippus | a carbone et forcipibus gladiosque paranti | incude et luteo Vulcano ad rhetora misit (red-­hot iron emits smoke when it is being forged). The sense thus restored is satisfactory, but there remains a fourth problematic aspect, namely the word order. In its transmitted position tristi seems objectionable on two counts: first, there is danger of correlating it with illi; second, the rhythm – – | – – – tends to be avoided at the beginning of a hexameter in favour of – – – | – –. Transferring tristi to the third rather than the second position seems to produce a more plausible word order (with turpabant massae paralleling 119 gaudebunt siluae and 121 tardabunt riui, among other things). turpabant The verb is infrequent in verse, but seems idiomatic in ­reference to defiling hair (Verg. Aen. 10.832 sanguine turpantem comptos de more capillos, 12.611 canitiem immundo perfusam puluere turpans, Stat. Theb. 3.680 laceris pridem turpata capillis, cf. Ciris 284 incomptos multo deturpat puluere crines, Suet. Cal. 35.2 pulchros et comatos, quotiens sibi occurrerent, occipitio raso deturpabat, further Hor. Sat. 1.5.60–1 at illi foeda cicatrix | saetosam laeui frontem turpauerat oris, Carm. 1.13.9–11 seu tibi candidos | turparunt umeros immodicae mero | rixae [an ode addressed to a Lydia]). The detail of Vulcan’s beard being smeared with soot may be intended to convey Venus’ perspective and to explain why she started cheating on him in the first place (cf. in a way the reference to Tithonus’ grey hair and beard as the reason for Eos’ leaving him at Hymn. Hom. 5.228–9). massae are pieces of (heated) raw metal being worked on by the blacksmith (cf. TLL 8.429.62–75); cf. e.g. Verg. Georg. 4.170–1 lentis Cyclopes fulmina massis | cum properant. tristi fuligine The noun is mundane and ‘unpoetisch’ (only Pl. Poen. 1195 os oblitum est fuligine, Verg. Ecl. 7.50 assidua postes fuligine nigri, before Juv.); on the adj., cf. OLD s.v. tristis 6b: ‘(of shade, dark-­coloured things, etc.) gloomy, sombre’.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

142  Commentary ON 174–6 barbam Vulcan regularly appears with a beard in visual representations, but I am unaware of analogous literary references. 175–6  The passage is corrupt at a crucial point (see 175n.), but clearly must refer to Aurora’s seducing a mortal lover, possibly several (on the Indo-­European background of the Dawn goddess’s rapacious sexuality, cf. Boedeker 1974, 64–84, West 2007, 224–5). Aurora is the last mytho­ logic­al exemplum in the poem, no doubt intended to evoke, by way of ring composition, the first, Luna and Endymion (144–5); for pairing the two, cf. Ov. Ars 3.82–3 (quoted 175n.), Nonn. Dion. 5.516–17 (ibid.). As in the case of other exempla in the poem (cf. 156–83n.), its positive evaluation by the speaker is one-­sided at best; a potentially relevant subversive intertext is supplied by Od. 5.118–24, where Calypso accuses the (male) gods of never letting goddesses enjoy happy relationships with mortal men, citing the example of Eos and Orion, killed by Artemis, as a case in point. 175  etiam plorauit (Ω). At 176 Aurora is described as blushing and cover­ing her eyes, and 175 should supply a reason for her feeling ashamed, which must be her nymphomania, cf. Ov. Ars 3.82–3 Latmius Endymion non est tibi, Luna, rubori, | nec Cephalus roseae praeda pudenda deae; the idea that Aurora ‘wept over her new lover’ can only suggest this in the most indirect way if at all, and the admission that she was unhappy in love undermines the speaker’s overall sentiment. In Kayachev 2022, 717–18, I propose raptauit; since etiam is misplaced (it should mean ‘did not Aurora too’, but in order to do so it should stand next to Aurora), I change it to totiens: Eos had more than one lover (cf. below). non For the negative rhetorical question introducing an erotic mytho­ logic­al exemplum, cf. Theocr. 1.105 οὐ λέγεται τὰν Κύπριν ὁ βουκόλος; 3.47 οὐχ οὕτως Ὥδωνις ἐπὶ πλέον ἄγαγε λύσσας [sc. Κυθέρειαν]; 20.37 Ἐνδυμίων δὲ τίς ἦν, οὐ βουκόλος; 40–1 οὐχὶ δὲ καὶ τύ, | ὦ Κρονίδα, διὰ παῖδα βοηνόμον ὄρνις ἐπλάγχθης; (One might suspect non to be an error for explicitly interrogative nonne, but the latter is overall infrequent in classical poetry and never occurs in elision; cf. in general Murphy 1991.) totiens raptauit Aurora was notorious for abducting mortal lovers: Ov. Her. 15.87–8 hunc ne pro Cephalo raperes, Aurora, timebam, | et

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 175–6  143 faceres, sed te prima rapina tenet, Met. 7.703–4 lutea mane uidet pulsis Aurora tenebris | inuitumque rapit [sc. Cephalum], Od. 15.251 Κλεῖτον χρυσόθρονος ἥρπασεν Ἠώς, Hymn. Hom. 5.218 Τιθωνὸν χρυσόθρονος ἥρπασεν Ἠώς, Eur. Hip. 454–5 ἀνήρπασέν ποτε | ἡ καλλιϕεγγὴς Κέϕαλον ἐς θεοὺς Ἕως, Nonn. Dion. 5.516–17 νυμϕίον ἀργέτις Ἠὼς | ἥρπασεν Ὠρίωνα καὶ Ἐνδυμίωνα Σελήνη. For totiens, cf. Prop. 2.18.17–18 cum sene non puduit talem dormire puellam | et canae totiens oscula ferre comae, where the implication of totiens is probably ‘every night’: likewise here the intimation may be that Aurora abducted a new lover every night, almost; cf. further συνεχῶς at [Apollod.] 1.27 Ὠρίωνος δ᾽ Ἠὼς ἐρασθεῖσα ἥρπασε καὶ ἐκόμισεν εἰς Δῆλον, ἐποίει γὰρ αὐτὴν Ἀϕροδίτη συνεχῶς ἐρᾶν, ὅτι Ἄρει συνευνάσθη. 176  roseo celauit amictu (Ω). Eos is ῥοδοδάκτυλος, but κροκόπεπλος, and so she has a saffron cloak at Ov. Ars 3.179 croceo uelatur amictu (cf. Giardina 2009, 173). While this need not establish an unbreakable rule, the colour contrast introduced by Boccaccio’s croceo (in cod. Laur. 33.31, f. 27r.)—rubens and roseo would refer to more or less the same hue—­is also attractive (cf. Verg. Ecl. 4.43–4 iam suaue rubenti | murice, iam croceo mutabit uellera luto, Aen. 7.26 Aurora in roseis fulgebat lutea bigis, Ov. Fast. 4.714 Memnonis in roseis lutea mater equis; I am not persuaded that Ars 3.82–3, where rubori and roseae appear in proximity, is a valid counter-­argument, since the two colour terms are applied to two different figures). Besides, garments (even gods’) are not otherwise described as ‘rosy’. By contrast, uelauit (likewise Boccaccio) would produce the wrong sense, pace Giardina: the appropriate gesture is not to ‘veil’ one’s eyes (contrast Luc. 2.361 lutea demissos uelarunt flammea uultus, cf. Mart. 12.43.3, describing a specific detail of the wedding attire), but to ‘hide’ them (cf. Stat. Ach. 1.233–4 udaque celat | lumina, Ov. Met. 4.682–3 manibusque modestos | celasset uultus). atque Unelided atque is increasingly avoided in classical poetry (Butterfield 2008), and is particularly disliked in the first foot, though Virgil is exceptionally tolerant of that placement in Ecl. (2.15, 3.11, 5.23, 8.99) and Georg. (1.85, 2.46, 491, 3.265, 427, 4.463). A recognizable pattern is for atque to join coordinate clauses occupying two consecutive lines, as in Lucr. 6.599–600 neu distracta suum late dispandat

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

144  Commentary ON 176–83 hiatum | atque suis confusa uelit complere ruinis, Verg. Ecl. 8.98–9 Moerim saepe animas imis excire sepulchris | atque satas alio uidi traducere messis, Georg. 1.84–5 saepe etiam steriles incendere profuit agros | atque leuem stipulam crepitantibus urere flammis (note saepe in the Virgilian ex­amples: the pattern evidently conveys the impression of a regularly repeated action, cf. perhaps 175 totiens). Here the syntactic parallelism is further underlined by the ‘rhyme’ 175 raptauit amores–176 celauit amictu, and the pattern may be intended to pro­ d­uce a closural effect, rounding off a catalogue of divine lovers (166– 76), for which cf. the conclusion of the proem in the Ciris: 99–100 praecipue nostro nunc aspirate labori | atque nouum aeterno praetexite honore uolumen (the only occurrence of unelided atque in the first foot in that poem). rubens Dawn can naturally be said to ‘redden’, cf. Verg. Aen. 3.521 iamque rubescebat stellis Aurora fugatis, 12.77 puniceis inuecta rotis Aurora rubebit, Ov. Met. 3.600–1 Aurora rubescere primo | coeperat; for the atmospheric phenomenon interpreted as blushing caused by shame, cf. Ov. Ars 3.82 (quoted 175n.). oculos . . . celauit For the eyes as expressing shame, see Cairns 1993, passim (and note e.g. Call. Hec. fr. 49.14–15 Hollis ἀναιδέσιν . . . ὀϕθαλμοῖσι with Hollis 2009, 201), though the more usual gesture is averting or casting down the eyes; for covering the face, cf. above. croceo . . . amictu is a rendition of Eos’ Homeric epithet κροκόπεπλος (cf. above); in Latin poetry the colour term is regularly transferred to the goddess herself or other details of her appearance, cf. e.g. Verg. Georg. 1.447 Tithoni croceum linquens Aurora cubile, Ov. Am. 2.4.43 placuit croceis Aurora capillis, Met. 3.150 croceis inuecta rotis Aurora, 7.703 lutea mane uidet pulsis Aurora tenebris, Fast. 3.421–2 croceis rorare genis Tithonia coniunx | coeperit, Cons. Liv. 281–2 hunc Aurora diem . . . croceis roscida portet equis. 177–83  The last stanza recalls 150–5, both in language (177 caelicolae, cf. 154; 177 aurea, cf. 150; 178 aetas, cf. 154; 182 uitae, cf. 155) and thought: the speaker complains that, whereas both gods and heroes had been able in the past happily to pursue illicit relationships, his own affair with Lydia was cut short. It also echoes the conclusion of the stanza on

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 177–83  145 the sex life of livestock (180 facilis natura, cf. 140), effectively summing up the speaker’s whole argument: he sees himself as having been treated unjustly in comparison with gods, heroes, and animals. 177 talia sc. fecerunt: the omission of a verb, here and three more times over these two lines, suggests euphemism, even though the implied idea (‘committed adultery’) is indelicate at worst, not obscene (on omission as euphemism, see Adams 1981); arguably the omission instils an innuendo that otherwise would have been avoided (though facere can itself be a euphemism). caelicolae pointedly echoes 154 caelicolae: there the gods were presented as being offended by human misdeeds (cf. n.), which have transpired to consist in illicit sexual relations, here they are accused of behaving no differently. numquid minus, ‘can one believe that the golden generation (did these things) any less?’: minus is to be construed adverbially, cf. Alf. Dig. 28.5.45 numquid minus heres esset ob eam rem, quod coheres eius hereditatem non adisset. aurea proles: χρύσεον . . . γένος (Hes. Op. 109), γένος χρύσειον (Arat. 114), though, as 178 heros shows, the heroic age is in fact implied (cf. 150n.). For proles (conjectured by Vonck 1766, 175 for Ω’s promo), cf. Cic. Arat. fr. 18.1 ferrea [aerea?] . . . proles, possibly a model (cf. 156n.). 178 deus atque heros: a conspicuous synecdoche; for pairing gods and heroes, cf. e.g. Hor. Ars 227 quicumque deus, quicumque adhibebitur heros. In Hes., heroes are the fourth race, immediately preceding the current iron one (Op. 159–60 ἀνδρῶν ἡρώων θεῖον γένος, οἳ καλέονται | ἡμίθεοι, προτέρη γενεή). minor aetas: ‘the younger generation’ (OLD s.v. minor 3), contrast Hes. Op. 160 προτέρη γενεή, though possibly also ‘the inferior race’ (OLD s.v. 7), cf. Op. 127 γένος πολὺ χειρότερον (of the silver race), Hor. Carm. 3.6.46–7 aetas parentum peior auis tulit | nos nequiores. 179–81  For the wish to have been born in a different age, cf. Hes. Op. 174–5 μηκέτ᾽ ἔπειτ᾽ ὤϕελλον ἐγὼ πέμπτοισι μετεῖναι | ἀνδράσιν, ἀλλ᾽ ἢ πρόσθε θανεῖν ἢ ἔπειτα γενέσθαι; cf. also Tib. 1.10.11 tunc [sc. in the

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

146  Commentary ON 179–81 golden age] mihi uita foret dulcis (on reading dulcis, see O’Hara 2005), AL 914.83–4 cur mihi non illis nasci, mea uita, diebus | contigit? 179 infelix ego . . . qui:  the construction is similar to the one used in μακαρισμοί (cf. 112 with n.), but is naturally rarer in the first person; in poetry I can only cite (with ego) Hor. Sat. 2.3.236 segnis ego, indignus qui tantum possideam (though note the subjunctive), Ars 301–2 o ego laeuus, | qui purgor bilem sub uerni temporis horam, CLE 1578.5 sed ego infelix, qui te talem carui ecce modo; the construction is surely to be interpreted as an exclamatory one (cf. Phaedr. 1.12.13 o me infelicem, qui nunc demum intelligo), rather than as one involving the omission of sum (cf. e.g. Pl. Trin. 1057 sed ego sum insipientior qui rebus curem puplicis, Aul. 769 sanus tu non es qui furem me uoces, Ter. Eun. 865–6 nam si ego digna hac contumelia | sum maxume, at tu indignus qui faceres tamen, Andr. 750 satin sanus qui me id rogites?). tempore natus: an echo of Catull. 64.22–3 o nimis optato saeclorum tempore nati | heroes, saluete, deum genus, which refers precisely to the time when the speaker would like to have lived (cf. Trimble 2010, 51). 180 facilis natura  echoes 140, see n. sors o mea laeua | nascendi: sors nascendi, ‘the lot of being born (under particular circumstances)’ and hence ‘the circumstances of birth’, is an established idiom, predominantly prosaic, cf. e.g. Sen. Con. 1.6.3 si possent homines facere sibi sortem nascendi, nemo esset humilis, Sen. Dial. 4.22.1 sors nascendi et educatio, Quint. Inst. 12.11.22 nulla sorte nascendi aetas felicior quam nostra, Man. 4.231–2 quibus in bifero Centauri corpore sors est | nascendi concessa; in poetry, the gerund occurs once in Lucr., and then only in Ov. (note esp. Met. 2.650 nascendi lege), until Man.; for the adj. in reference to ‘adverse fate’, cf. Stat. Theb. 11.444 laeuae post praemia sortis, Sil. 3.93–4 quod si promissum uertat Fortuna fauorem | laeuaque sit coeptis, CLE 506.2 si Fortuna quidem fatis non laeua fuisset. 181  genus quo sera libido est (Ω). It may be marginally possible to keep quo if genus is interpreted in a temporal sense (‘generation during which’), but Putsche’s (1828, 119) cui (quoi Naeke 1847, 214) yields an easier text.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 181–3  147 miserumque genus: perhaps an echo of Lucr. 5.1194–5 o genus infelix humanum, talia diuis | cum tribuit facta (cf. 177 talia?), based in turn on Emp. fr. 124.1 ὢ πόποι, ὢ δειλὸν θνητῶν γένος, ὢ δυσάνολβον; cf. also Opp. Hal. 2.669 ἡμερίων δειλὸν γένος (referring to people before the rule of Augustus). cui sera libido est: ‘whose lust is late’, i.e. ‘who experience lust when it is too late (to enjoy it freely)’; cf. esp. Sen. Phaed. 595 serus est nobis pudor, Val. Fl. 4.247 redit Alcidae iam sera cupido, though here the nuance is somewhat different: in the case of Phaedra and Hercules the reference is to an earlier moment in their lives, whereas in the case of the Lydia speaker it is to an earlier age. Libido is an indelicate word, shunned by Verg. and Tib. (for some usages, see DMALL 144–6 s.v. ‘deseo’); here it may evoke Lucr. 5.964 (cf. 150–5n.). 182–3  While 182 is heavily, and obviously, corrupt (see n.), 183 is seemingly sound, but contains an anomaly that might be taken to indicate interpolation: the second-­person address (possis) can only be directed to Lydia (it would be odd for the reader to be appealed to out of the blue), but she is not otherwise addressed by the speaker, and their separateness is very much the point throughout the poem. The suspicion is strengthened by the close parallel at Ov. Pont. 1.4.5–6 nec, si me subito uideas, agnoscere possis, | aetatis facta est tanta ruina meae, which could have provided the interpolator with a model. Heyworth* suggests that 181 may likewise be interpolated, but it does not seem linguistically objectionable (see n.), and the way 180b–181 more or less repeat 179–180a is characteristic of our poet (cf. 123–4, 140–1, on a greater scale 159–65). After some hesitation, I opt for keeping the lines, while emending away the second-­person reference in 183 (see n.). By evoking the agency of fate, the passage supplies a positive counterpart to the negative statement of 165–6 (likewise at the close of a stanza); and by referring to the speaker’s extreme emaciation and thus picking up a theme introduced in the first section (125–6), it provides a fitting if somewhat abrupt conclusion to the whole poem. The suggestion that fate has taken away almost all of the speaker’s life parallels in a way the ending of Thyrsis’ song in Theocr. 1 which also refers to Daphnis’ life running out (139–40 τά γε μὰν λίνα πάντα λελοίπει | ἐκ Μοιρᾶν).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

148  Commentary ON 182 182  tantum uita [uitae WATP] meae [MWATP: tanta meae (me S) uitae SFBEO] cordis fecere rapinam (Ω). For an account of the discrepancies between the manuscripts, cf. Courtney 1968, 138. Heinsius’ tantam fata (in Burman 1773, 678) appears all but certain; his ruinam is not un­attract­ive, both in view of Ov. Pont. 1.4.5–6 (quoted 182–3n.) and because ruina is more readily used in this kind of figurative context and in periphrastic constructions with e.g. facere (add e.g. Lucr. 1.740 principiis tamen in rerum fecere ruinas, Ov. Her. 12.32 illa fuit mentis prima ruina meae, Tr. 4.8.36 curriculo grauis est facta ruina meo, CLE 441.3–4 mors tacita obrepsit subito fecitque ruinam | quae tibi crescenti rapuit iuuenile figuram), but rapinam appears altogether preferable, esp. as it works well with 183 maneam (‘so much has been taken away that the remainder is barely recognizable’). The central problem is meae cordis: besides the wrong gender of meae, cordis falls under suspicion because 183 makes it clear that the speaker is affected physically, not just emotionally. Heinsius’ sortis yields a plausible turn of phrase (‘the ruin of my fortune’), but is unlikely with fata as the subject. Mähly’s (1870, 783) carnis (with rapinam) produces an image of emaciation that works well with 183, but caro is a term avoided in poetry in reference to human flesh (cf. Axelson 1945, 52). I think we should read tantam fata meae uitae with Canal (1840, 1341 and 1393–4): while cordis is an unlikely copying error for uitae, it could be a conjecture aimed at correcting a line with uita and uitae (cf. Putsche 1828, 120). uitae . . . fecere rapinam Normally ‘snatching away of life’ in such a context would imply death (cf. e.g. Ov. Am. 3.9.35 rapiunt mala fata bonos, CLE 379.1 florentem speciem rapuere nouissima fata, 1149.1 inuida florentem rapuerunt fata iuuenta), but the periphrastic construction (fecere rapinam for rapuerunt, cf. Prop. 4.11.62 nec mea de sterili facta rapina domo for rapta sum), with qualifying tantam, makes it possible to view this as a gradual process, for which cf. Sen. Ep. 104.12 omnis dies, omnis hora te mutat; sed in aliis rapina facilius apparet; for the ‘quantitative’ conception of life implied here, cf. esp. Lucr. 3.546–7 magis ac magis undique sensus | deficit et uitae minus et minus undique restat (with Kenney 2014, 146–8); note that at 126 Lucr.’s account of the physiology of dying appears likewise to be a reference point (see n.).

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Commentary on 183  149 183  oculis cognoscere possis (Ω): the second-­person address is out of place (see 182–3n.). Although possis at Ov. Pont. 1.4.5 (quoted ibid.) might seem to protect possis here, writing oculi . . . possint seems an easy and plausible solution (for the corruption, cf. Catull. 64.17 with Trimble 2010, 41); for oculi as subject of (cog)noscere, cf. Lucr. 4.385 nec possunt oculi naturam noscere rerum, Ov. Met. 11.467 oculi nequeunt cognoscere uultus, Pedo fr. 228.20–1 Hollis di reuocant rerumque uetant cognoscere finem | mortales oculos (by contrast, -noscere with instrumental oculis appears unparalleled in the poetry of the classical period). maneam, ‘what is left of me’: a vivid usage that is difficult to parallel exactly; manere is occasionally used with a noun predicate (TLL 8.289.26–36, note e.g. Ov. Met. 1.366 hominumque exempla manemus, 14.288 pecoris pars una manerem), but not in the pregnant sense employed here (‘to remain nothing more than’); the one close parallel is Ov. Am. 3.9.39–40 iacet, ecce, Tibullus: | uix manet e toto parua quod urna capit (cf. TLL 8.288.18–20), if we take Tibullus (rather than the quod clause) to be the subject of manet. cognoscere possint: a frequent hexameter ending in Lucr. (though as a rule with possis), e.g. 2.462, 3.117, 4.642, 749, 5.285, 6.113; Ov. Met. 13.835 inter se similes, uix ut dignoscere possis (sc. two bear cubs, gifted by Polyphemus to Galatea) may allude to our context. ❧ O FENIX ❧

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

Bibliography Abbreviations AE L’Année épigraphique (Paris, 1888–). AL A. Riese, Anthologia latina, 2nd ed., vol. 1, 2 fascs. (Leipzig, 1894–1906). CLE F. Bücheler, Carmina latina epigraphica, 2 vols. (Leipzig, 1895–7). DMALL R. Moreno Soldevila, ed., Diccionario de motivos amatorios en la literatura latina (Huelva, 2011). EDL M.  de Vaan, Etymological Dictionary of Latin and the Other Italic Languages (Leiden, 2008). GL H. Keil, Grammatici latini, vol. 4 (Leipzig, 1864). LSJ H.G.  Liddell, R.  Scott, H.S.  Jones, A Greek-­English Lexicon, rev. ed. (Oxford, 1996). OLD P.G.W.  Glare, ed., Oxford Latin Dictionary, 2nd ed., 2 vols. (Oxford, 2012). PMG D.L. Page, Poetae melici graeci (Oxford, 1962). TLL Thesaurus linguae latinae (Leipzig, 1894–).

Early editions cited Romana 1471 P. Virgilii Maronis Bucolica [. . .] Conradus Suueynheym Arnoldus Pannartzque magistri Rome impresserunt [. . .] [a. 1471]. Ascensiana 1501 Vergiliana opuscula familiariter exposita. Impressa [. . .] solertia Thielmanni Kerver in inclyti Parrhisorum academia ad Idus Martias post Iubileum. Aldina 1517 Diversorum veterum poetarum in Priapum lusus. P.M.V.  Catalecta. Copa. Rosae. Culex. Dirae. Moretum. Ciris. [. . .] Venetiis in aedibus Aldi [ . . . ] mense Decembri M.D.XVII. Argentoratensis 1527 P. V. Maronis opera. Argentorati, Ioannes Cnoblochus excudebat anno M.D.XXVII. Parisina 1534 Publius Vergilius Maro, Philippi Melanchthonis adnotatiunculis [. . .] illustratus. Parisiis, Apud Tielmanum Vivian. 1534. Parisina 1540 P. Virgilii Maronis opera. Parisiis, Ex officina Rob. Stephani [. . .] M.D.XL.

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

152 Bibliography Veneta 1555 Pub. Virgilii Maronis poetae Mantuani opera. Venetiis, Apud Hieronymum Scotum. MDLV.

Works cited Acél, Z. 2007. ‘Der Orpheus des Epitaphios Bionos in den Werken von Vergil und Ovid’, AAntHung 47, 349–68. Adams, J.N. 1981. ‘A type of sexual euphemism in Latin’, Phoenix 35, 120–8. Adams, J.N. 1982. The Latin Sexual Vocabulary (London). Alton, E.H. 1922. ‘Notes on the Dirae and the Lydia’, Hermathena 19, 308–14. Anagnostou-­Laoutides, E. 2013. Eros and Ritual in Ancient Literature: Singing of Atalanta, Daphnis, and Orpheus (Piscataway). Anagnostou-­Laoutides, E. 2016. ‘Daphnis, deus pastoralis: the trail of his advent’, GIF 68, 43–88. Anagnostou-­Laoutides, E., and D.  Konstan. 2008. ‘Daphnis and Aphrodite: a love affair in Theocritus Idyll 1’, AJPh 129, 497–527. Arnold, C. 1652. Val. Catonis Grammatici Dirae (Leiden). Axelson, B. 1945. Unpoetische Wörter: Ein Beitrag zur Kenntnis der lateinischen Dichtersprache (Lund). Baehrens, E. 1880. Poetae latini minores, vol. 2 (Leipzig). Ballarini, S.M. 1988/9. ‘Lydia entre la elegía latina’, REC 20, 33–45. Barchiesi, A. 1992. P. Ovidii Nasonis Epistulae Heroidum 1–3 (Florence). Barchiesi, A. 2006. ‘Women on top: Livia and Andromache’, in R. Gibson et al., eds., The Art of Love: Bimillennial Essays on Ovid’s Ars Amatoria and Remedia Amoris (Oxford), 96–120. Barth, C. 1664. P. Papinii Statii quae exstant, vol. 2 (Zwickau). Bassett, S.E. 1919. ‘Versus tetracolos’, CPh 14, 216–33. Black, R. 1998. ‘Boccaccio, reader of the Appendix Vergiliana: the Miscellanea Laurenziana and fourteenth-­century schoolbooks’, in M. Picone, C. Cazalé-­Bérard, eds., Gli zibaldoni di Boccaccio: memoria, scrittura, riscrittura (Florence), 113–28. Boedeker, D.D. 1974. Aphrodite’s Entry into Greek Epic (Leiden). Booth, J. 2006. ‘Amazing grace: reading between the lines in Propertius 1.13.29–32’, CQ 56, 528–37. Bouchard, E. 2020. ‘The self-­revealing poet: lyric poetry and cultural history in the Peripatetic school’, in B. Currie, I. Rutherford, eds., The Reception of Greek Lyric Poetry in the Ancient World: Transmission, Canonization and Paratext (Leiden), 182–202. Braun, P. 1869. Observationes criticae et exegeticae in C. Valerii Flacci Argonautica (Marburg). Breed, B.W. 2012. ‘The pseudo-­Vergilian Dirae and the earliest responses to Vergilian pastoral’, TC 4, 3–28. Broukhusius, J. 1708. Albii Tibulli equitis Rom. quae exstant (Amsterdam). Broukhusius, J. 1727. Sex. Aurelii Propertii Elegiarum libri quatuor (Amsterdam). Brown, J.P. 1995. Israel and Hellas (Berlin).

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

Bibliography  153 Burges Watson, S. 2014. ‘Orpheus’ erotic mysteries: Plato, pederasty, and the Zagreus myth in Phanocles fragment 1’, BICS 57, 47–71. Burman, P. 1727. Publii Ovidii Nasonis opera omnia, vol. 1 (Amsterdam). Burman, P. 1773. ‘Valerii Catonis Dirae in Battarum’, in P.  Burman, Anthologia veterum Latinorum epigrammatum et poematum, vol. 2 (Amsterdam), 647–78. Burris, S.P. 2004. ‘Refrains in ancient Greek poetry’ (PhD diss., Cornell Univ.). Butterfield, D.J. 2008. ‘The poetic treatment of atque from Catullus to Juvenal’, Mnemosyne 61, 386–413. Cabaret-­Dupaty, J.-R.-T. 1842. Poetae minores (Paris). Cairns, D.L. 1993. Aidōs: The Psychology and Ethics of Honour and Shame in Ancient Greek Literature (Oxford). Cairns, F. 2006. Sextus Propertius: The Augustan Elegist (Cambridge). Cairns, F. 2016. Hellenistic Epigram: Contexts of Exploration (Cambridge). Campbell, G. 2003. Lucretius on Creation and Evolution (Oxford). Canal, P. 1840. ‘Due idilii attribuiti al grammatico Valerio Catone’, in V.  Monti, Le satire di A. Persio Flacco (Venice), 1273–1394. Canter, W. 1571. Novarum lectionum libri octo, 3rd ed. (Antwerp). Cassio, A.C. 2012. ‘Kypris, Kythereia and the fifth book of the Iliad’, in F. Montanari et al., eds., Homeric Contexts: Neoanalysis and the Interpretation of Oral Poetry (Berlin), 413–26. Caston, R.R. 2006. ‘Love as illness: poets and philosophers on romantic love’, CJ 101, 271–98. Cavallini, E. 1986. ‘Due note all’Appendix Vergiliana’, GFF 9, 27–30. Clare, R.J. 1997. ‘Catullus 64 and the Argonautica of Apollonius Rhodius: allusion and exemplarity’, PCPhS 42, 60–88. Clarke, J.R. 2013. ‘Before pornography: sexual representation in ancient Roman ­visual culture’, in H. Maes, ed., Pornographic Art and the Aesthetics of Pornography (Basingstoke), 141–61. Clausen, W. 1964. ‘The textual tradition of the Culex’, HSCPh 68, 119–38. Clausen, W. 1994. A Commentary on Virgil, Eclogues (Oxford). Clay, D. 2001. ‘The scandal of Dionysos on Paros (the Mnesiepes inscription E1 III)’, Prometheus 27, 97–112. Coleman, R. 1999. ‘Poetic diction, poetic discourse and the poetic register’, PBA 93, 21–93. Connors, C. 1998. Petronius the Poet: Verse and Literary Tradition in the Satyricon (Cambridge). Conrau-­Lewis, K. 2020. ‘Cinna and Crassicius: promiscuity, incest and the erotics of commentary’, Mnemosyne 73, 243–60. Courtney, E. 1968. ‘The textual transmission of the Appendix Vergiliana’, BICS 15, 133–41. Courtney, E. 1993. The Fragmentary Latin Poets (Oxford). Dainotti, P. 2015. Word Order and Expressiveness in the Aeneid (Berlin). Davidson, J. 1997. Courtesans and Fishcakes: The Consuming Passions of Classical Athens (London). Davies, M. 2019. The Cypria (Washington, DC).

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

154 Bibliography Deacy, S. 2013. ‘From “flowery tales” to “heroic rapes”: virginal subjectivity in the mythological meadow’, Arethusa 46, 395–413. Deschamps, L. 1983. ‘Varron, les lymphes et les nymphes’, in H. Zehnacker, G. Hentz, eds., Hommages à Robert Schilling (Paris), 67–83. Di Nino, M.M. 2009. ‘Le “verità nascoste”: consapevole appartenenza a un genere, autoinvestitura e bugie metapoetiche in [Mosco] III’, Philologus 153, 86–108. Di Nino, M.M. 2018. ‘Οὐ γὰρ ἴσον Κύκλωπι μελίσδεο: intertextuality, metalepsis, and eulogistic strategies in EB 58–63’, Philologus 162, 25–54. Dickey, E. 1996. Greek Forms of Address: From Herodotus to Lucian (Oxford). Dickey, E. 2002. Latin Forms of Address: From Plautus to Apuleius (Oxford). Duckworth, G.E. 1969. Vergil and Classical Hexameter Poetry: A Study in Metrical Variety (Ann Arbor). Edmunds, L. 2002. ‘Mars as Hellenistic lover: Lucretius, De rerum natura 1.29–40 and its subtexts’, IJCT 8, 343–58. Ellis, R. 1907. Appendix Vergiliana (Oxford). Enk, P.J. 1919. ‘De Lydia et Diris carminibus’, Mnemosyne 47, 382–409. Estevez, V.A. 1981. ‘Ἀπώλετο καλὸς Ἄδωνις: a description of Bion’s refrain’, Maia 33, 35–42. Fairclough, H.R., and G.P. Goold. 2000. Virgil: Aeneid VII–XII, Appendix Vergiliana (Cambridge, MA). Fantham, E. 2008. ‘Covering the head at Rome: ritual and gender’, in J. Edmondson, A. Keith, eds., Roman Dress and the Fabrics of Roman Culture (Toronto), 158–71. Fantuzzi, M. 1995. ‘Mythological paradigms in the bucolic poetry of Theocritus’, PCPhS 41, 16–35. Farrell, J. 2020. ‘The gods in Ennius’, in C. Damon, J. Farrell, eds., Ennius’ Annals: Poetry and History (Cambridge), 63–88. Ferri, R. 2003. Octavia: A Play Attributed to Seneca (Cambridge). Finkelberg, M. 2007. ‘More on κλέος ἄϕθιτον’, CQ 57, 341–50. Fisch, R. 1875. De Catulli in vocabulis collocandis arte quaestiones selectae (Berlin). Fontenrose, J.E. 1939. ‘Apollo and Sol in the Latin poets of the first century B.C.’, TAPhA 70, 439–55. Fontenrose, J.E. 1940. ‘Apollo and the sun-­god in Ovid’, AJPh 61, 429–44. Fowler, R.L. 2013. Early Greek Mythography, vol. 2 (Oxford). Fraenkel, R. 1966. ‘The Dirae’, JRS 56, 142–55. Franklin, J.L. 1997. ‘Cn. Alleius Nigidius Maius and the amphitheatre: munera and a distinguished career at ancient Pompeii’, Historia 46, 434–47. Fredrick, D. 1997. ‘Reading broken skin: violence in Roman elegy’, in J.P.  Hallett, M.B. Skinner, eds., Roman Sexualities (Princeton), 172–93. Frère, J. 2014. ‘Mortals (βροτοί) according to Parmenides’, in N.-L.  Cordero, ed., Parmenides, Venerable and Awesome (Las Vegas), 135–46. Friedman, L.J. 1965. ‘Gradus amoris’, Romance Philology 19, 167–77. Fulkerson, L. 2013. ‘Seruitium amoris: the interplay of dominance, gender and poetry’, in T.S. Thorsen, ed., The Cambridge Companion to Latin Love Elegy (Cambridge), 180–93.

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

Bibliography  155 Gagliardi, P. 2014. ‘L’ecl. 10 di Virgilio e la poesia bucolica post-­teocritea’, REL 92, 123–36. Gagliardi, P. 2015. ‘Formosus in Virgilio e dintorni’, Eirene 51, 45–58. Gagliardi, P. 2017. ‘Tua cura, Lycoris: lessico erotico e schema Cornelianum da Virgilio agli elegiaci nel segno di Gallo’, WS 130, 183–200. Gagliardi, P. 2019. ‘The metamorphosis of Daphnis from Theocritus to Virgil’, Phasis 21–2, 119–39. Gain, D.B. 1976. The Aratus ascribed to Germanicus Caesar (London). Gale, M. 1994. Myth and Poetry in Lucretius (Cambridge). Galinsky, G.K. 1967. ‘Sol and the Carmen Saeculare’, Latomus 26, 619–33. Gallia, A.B. 2014. ‘The vestal habit’, CPh 109, 222–40. García Vargas, E. 2004. ‘El vino de la Bética altoimperial y las ánforas: a propósito de algunas novedades epigráficas’, Gallaecia 24, 117–34. Georgescu, S. 2019. ‘The world as a yawning gap: new insights into the etymology of Lat. mundus “world” ’, in N. Holmes et al., eds., Lemmata linguistica latina, vol. 1 (Berlin), 206–23. Gerber, D.E. 1978. ‘The female breast in Greek erotic literature’, Arethusa 11, 203–12. Giardina, G. 2009. ‘Nuovi emendamenti al testo delle Dirae e della Lydia pseudovirgiliane’, QUCC 92, 167–73. Gibson, R. 2022. ‘Ovid’s Ars Amatoria and the Epicurean hedonic calculus’, in K.  Volk, G.D.  Williams, eds., Philosophy in Ovid, Ovid as Philosopher (New York), 63–83. Giomini, R. 1953. Appendix Vergiliana (Florence). Glenn, J. 1980/1. ‘Ariadne’s daydream (Cat. 64. 158–63)’, CJ 76, 110–16. Goebbel, F.C. 1865. Valeri Catonis quae feruntur carmina (Warendorf). Goodyear, F.R.D. 1971. ‘The “Dirae” ’, PCPhS 17, 30–43. Goold, G.P. 1998. ‘A paraclausithyron from Pompeii’, in P. Knox, C. Foss, eds., Style and Tradition: Studies in Honor of Wendell Clausen (Stuttgart), 16–29. Gow, A.S.F. 1952. Theocritus, vol. 2, 2nd ed. (Cambridge). Graaf, C. van der. 1945. The Dirae (Leiden). Gregory, J. 2023. ‘The extreme wish in Greek literature’, Mnemosyne 76, 357–74. Griffin, J. 1992. ‘Theocritus, the Iliad, and the East’, AJPh 113, 189–211. Gutzwiller, K.J. 1983. ‘Character and legend in Idyll 8’, TAPhA 113, 171–82. Gutzwiller, K.J. 2002/3. ‘Meleager as erotic dolphin: a reading of AP 12.52’, Hermathena 173/174, 91–105. Habinek, T. 1997. ‘The invention of sexuality in the world-­ city of Rome’, in T.  Habinek, A.  Schiesaro, eds., The Roman Cultural Revolution (Cambridge), 23–43. Hague, R.H. 1983. ‘Ancient Greek wedding songs: the tradition of praise’, Journal of Folklore Research 20, 131–43. Harder, A. 2012. Callimachus, Aetia: Introduction, Text, Translation, and Commentary, vol. 2 (Oxford). Harrison, S.J. 1990. ‘Some Odyssean scenes in Apuleius’ Metamorphoses’, MD 25, 193–201.

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

156 Bibliography Harrison, S.J. 2017. Horace: Odes, Book II (Cambridge). Harrison, S.J. 2020. ‘Serial similes in the battle-­narrative of Virgil’s Aeneid’, in C. Burrow et al., eds., Imitative Series and Clusters from Classical to Early Modern Literature (Berlin), 27–38. Haupt, M. 1874. ‘Coniectanea’, Hermes 8, 1–17. Heath, J.R. 1986. ‘The supine hero in Catullus 32’, CJ 82, 28–36. Henderson, J. 1991. The Maculate Muse: Obscene Language in Attic Comedy, 2nd ed. (Oxford). Herrmann, L. 1949. ‘Trois poèmes de P. Valerius Cato’, Latomus 8, 111–44. Hersch, K.K. 2010. The Roman Wedding: Ritual and Meaning in Antiquity (Cambridge). Heyworth, S.J. 1992. ‘Propertius 2.13’, Mnemosyne 45, 45–59. Heyworth, S.J. 2007. Cynthia: A Companion to the Text of Propertius (Oxford). Heyworth, S.J. 2013. ‘Keeping on keeping on: repetition in the Propertian narrative’ (unpublished manuscript, deposited in Oxford Research Archives). Heyworth, S.J. 2015. ‘Notes on the text and interpretation of Vergil’s Eclogues and Georgics’, in H.-C. Günther, ed., Virgilian Studies: A Miscellany Dedicated to the Memory of Mario Geymonat (Nordhausen), 195–249. Heyworth, S.J. 2019. Ovid, Fasti Book 3 (Cambridge). Hill, J.M. 2021. ‘The Latin past and the poetry of Catullus’ (PhD diss., Univ. of Toronto). Hinds, S. 1987. ‘Language at the breaking point: Lucretius 1.452’, CQ 37, 450–3. Hollis, A.S. 2007. Fragments of Roman Poetry c.60 BC–AD 20 (Oxford). Hollis, A.S. 2009. Callimachus, Hecale, 2nd ed. (Oxford). Holzberg, N. 2020. ‘Einführung’, in F.  Zogg, ed., Appendix Vergiliana: lateinisch-­ deutsch (Berlin), 11–42. Hordern, J.H. 1999. ‘The Cyclops of Philoxenus’, CQ 49, 445–55. Horsfall, N. 1990. ‘Virgil and the illusory footnote’, PLLS 6, 49–63. Horsfall, N. 2008. Virgil, Aeneid 2: A Commentary (Leiden). Höschele, R. 2011. ‘Inscribing epigrammatists’ names: Meleager in Propertius and Philodemus in Horace’, in A.  Keith, ed., Latin Elegy and Hellenistic Epigram: A Tale of Two Genres in Rome (Newcastle), 19–31. Housman, A.E. 1937. M. Manilii Astronomicon liber primus, 2nd ed. (Cambridge). Hubaux, J. 1930. Les thèmes bucoliques dans la poésie latine (Brussels). Hubbard, T.K. 1993. ‘Poetic succession and the genesis of Alexandrian bucolic’, SyllClass 4, 27–42. Hubbard, T.K. 1998. The Pipes of Pan: Intertextuality and Literary Filiation in the Pastoral Tradition from Theocritus to Milton (Ann Arbor). Hunter, R. 1999. Theocritus: A Selection (Cambridge). Hunter, R. 2015. Apollonius of Rhodes: Argonautica, Book IV (Cambridge). Hutchinson, G.O. 2019. ‘On not being beautiful’, in S.  Matzner, S.  Harrison, eds., Complex Inferiorities: The Poetics of the Weaker Voice in Latin Literature (Oxford), 169–204. Hyllested, A. 2004. ‘L’esprit des eaux: grec νύμϕη, sanskrit Rámbhā, lithuanien Laumė̃ et quelques autres formes semblant apparentées’, in A. Hyllested et al., eds., Per

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

Bibliography  157 aspera ad asteriscos: studia Indogermanica in honorem Jens Elmegård Rasmussen sexagenarii Idibus Martiis anno MMIV (Innsbruck), 219–33. Jackson, S. 2006. ‘Apollonius of Rhodes: Endymion’, QUCC 83, 11–21. Jacobs, F. 1792. ‘Ueber die Diras des Valerius Cato’, Bibliothek der alten Litteratur und Kunst 9, 56–61. Jolowicz, D. 2021. Latin Poetry in the Ancient Greek Novels (Oxford). Kania, R. 2012. ‘Orpheus and the reinvention of bucolic poetry’, AJPh 133, 657–85. Kayachev, B. 2011. ‘Ille ego qui quondam: genre, date, and authorship’, Vergilius 57, 75–82. Kayachev, B. 2016. Allusion and Allegory: Studies in the Ciris (Berlin). Kayachev, B. 2016a. ‘Catalepton 9 and Hellenistic poetry’, CQ 66, 180–204. Kayachev, B. 2020. Ciris: A Poem from the Appendix Vergiliana (Swansea). Kayachev, B. 2020a. ‘Catalepton 9 and Valgius Rufus’, in T.E. Franklinos, L. Fulkerson, eds., Constructing Authors and Readers in the Appendices Vergiliana, Tibulliana, and Ovidiana (Oxford), 83–95. Kayachev, B. 2021. ‘Love and poetry in Virgil’s sixth eclogue: a Platonic perspective’, in T.S.  Thorsen et al., eds., Greek and Roman Love: The Poetic Connection (Berlin), 105–23. Kayachev, B. 2021a. ‘Introduction’, in B.  Kayachev, ed., Poems Without Poets: Approaches to Anonymous Ancient Poetry (Cambridge), 1–9. Kayachev, B. 2021b. ‘A green sky and a green sun? (Pliny, NH 17.74 and Manilius 2.941)’, CQ 71, 900–2. Kayachev, B. 2022. ‘Emendations in the Dirae and the Lydia’, CQ 72, 703–18. Kenney, E.J. 1966. ‘Dirae (Lydia)’, in W.V. Clausen et al., Appendix Vergiliana (Oxford), 1–14. Kenney, E.J. 1996. Ovid: Heroides XVI–XXI (Cambridge). Kenney, E.J. 2002. ‘Ovid’s language and style’, in B.W. Boyd, ed., Brill’s Companion to Ovid (Leiden), 27–89. Kenney, E.J. 2014. Lucretius: De rerum natura, Book III, 2nd ed. (Cambridge). Kidd, D. 1997. Aratus: Phaenomena (Cambridge). Knaack, G. 1905. ‘Ein verlorenes Epyll des Bion von Smyrna’, Hermes 40, 336–40. Knox, P.E. 1995. Ovid, Heroides: Select Epistles (Cambridge). Konstan, D. 2014. Beauty: The Fortunes of an Ancient Greek Idea (New York). Konstan, D., and F.  Muecke. 1992/3. ‘Richard Bentley as a reader of Horace’, CJ 88, 179–86. Krause, M. 2014. ‘The quinque lineae amoris’, C&M 65, 55–85. Kroon, C. 1995. Discourse Particles in Latin (Amsterdam). Krostenko, B.A. 1993. ‘Pleasure, truth, and beauty in the language of Cicero and Catullus: a cognitive approach’ (PhD diss., Harvard Univ.). Lachmann, K. 1850. In  T.  Lucretii Cari De rerum natura libros commentarius (Berlin). Lampe, K. 2013. ‘Rationality, eros, and daemonic influence in the Platonic Theages and the Academy of Polemo and Crates’, AJPh 134, 383–424. Lanzarone, N. 2018. Il commento di Pomponio Leto all’Appendix Vergiliana (Pisa). László, G. 1903. ‘Szövegkritikai és magyarázó megjegyzések Valerius Cato költeményeihez’, EPhK 27, 361–78.

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

158 Bibliography Lee, G. 1990. Tibullus: Elegies, 3rd ed. (Leeds). Lentini, G. 1999. ‘Un carme allegorico in POxy 1788: il fr. 119 V.  di Alceo’, RFIC 127, 5–31. Levin-­Richardson, S. 2019. The Brothel of Pompeii: Sex, Class, and Gender at the Margins of Roman Society (Cambridge). Lightfoot, J.L. 1999. Parthenius of Nicaea: The Poetical Fragments and the Ἐρωτικὰ παθήματα (Oxford). Lindsay, W.M. 1918. ‘Notes on the Lydia’, CR 32, 62–3. Lipka, M. 2001. Language in Vergil’s Eclogues (Berlin). López-­Cañete Quiles, D. 2013. ‘Lyde ludens (Hor. Carm. 3.11.8–10): avatares y azares de un juego de palabras’, CFC(L) 33, 7–34. Lorenz, S. 2005. ‘Inuideo uobis, agri: mea gaudia habetis: Bukolische Verwünschungen und elegische Eifersucht in den Dirae’, in N.  Holzberg, ed., Die Appendix Vergiliana: Pseudepigraphen im literarischen Kontext (Tübingen), 1–27. Luiselli, B. 1960. ‘Le Dirae e la Lydia’, AFLC 28, 96–124. [=Studi sulla poesia bucolica (Cagliari, 1967), 117–44.] Luque Moreno, J. 2015. ‘Clodia, ¿Sapphica musa?, ¿doctior puella?: nota a Catulo 35,16–17’, Revista de Estudios Latinos 15, 11–28. Lyne, R.O.A.M. 2002. ‘Notes on Catullus’, CQ 52, 600–8. Mackail, J.W. 1908. ‘Virgil and Virgilianism: a study of the minor poems attributed to Virgil’, CR 22, 65–73. Macleod, C.W. 1979. ‘Horatian imitatio and Odes 2.5’, in D. West, T. Woodman, eds., Creative Imitation and Latin Literature (Cambridge), 89–102. Mähly, J. 1870. [Review of Ribbeck 1868,] Heidelberger Jahrbücher der Literatur 63, 769–96 and 801–39. Maltby, R. 1991. A Lexicon of Ancient Latin Etymologies (Leeds). Maltby, R. 2002. Tibullus: Elegies (Cambridge). Manakidou, F.P. 1996. ‘Ἐπιτάϕιος Ἀδώνιδος and Ἐπιτάϕιος Βίωνος: remarks on their generic form and content’, MD 37, 27–58. Margulies, Z. 2020. ‘Like golden Aphrodite: grieving women in the Homeric epics and Aphrodite’s lament for Adonis’, CQ 70, 485–98. Massimilla, G. 2010. Callimaco, Aitia: libro terzo e quarto (Pisa). Massimilla, G. 2021. ‘Theocritus and bucolic poetry’, in P. Kyriakou et al., eds., Brill’s Companion to Theocritus (Leiden), 131–53. McGill, S. 2020. Virgil, Aeneid Book XI (Cambridge). McKeown, J.C. 1989. Ovid: Amores, vol. 2 (Leeds). McKeown, J.C. 1998. Ovid: Amores, vol. 3 (Leeds). Melo, W.D.C. de. 2019. Varro: De Lingua Latina, vol. 2 (Oxford). Michalopoulos, A.N. 2006. Ovid, Heroides 16 and 17 (Cambridge). Murphy, P.R. 1991. ‘On questions introduced by non and nonne’, CJ 86, 226–32. Naeke, A.F. 1829. [Review of Putsche 1828,] RhM 3, 148–52. Naeke, A.F. 1847. Carmina Valerii Catonis (Bonn). Ni'am, M.I. 2020. ‘Tsuroyya’s star as a sign of pandemic’s end’, Elfalaky 4, 124–43. Nisbet, R.G.M. 1991. ‘The style of Virgil’s Eclogues’, PVS 20, 1–14. [= K.  Volk, ed. Vergil’s Eclogues: Oxford Readings in Classical Studies (Oxford, 2008), 48–63.]

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

Bibliography  159 Nisbet, R.G.M., and M.  Hubbard. 1970. A Commentary on Horace: Odes, Book I (Oxford). Nisbet, R.G.M., and M.  Hubbard. 1978. A Commentary on Horace: Odes, Book II (Oxford). Nobili, C. 2006. ‘Motivi della poesia nuziale in Odissea VI 149–185’, RIL 140, 57–72. Norden, E. 1927. P. Vergilius Maro, Aeneis Buch VI, 3rd ed. (Leipzig). Noris, R.P., and B.R.M. Noris. 2021. ‘Why are there Seven Sisters?’, in E. Boutsikas et al., eds., Advancing Cultural Astronomy: Studies in Honour of Clive Ruggles (New York), 223–35. O’Hara, J.J. 2005. ‘War and the sweet life: the Gallus fragment and the text of Tibullus 1.10.11’, CQ 55, 317–19. Overduin, F. 2015. Nicander of Colophon’s Theriaca: A Literary Commentary (Leiden). Pachoumi, E. 2012. ‘Eros as disease, torture and punishment in magic­al literature’, SO 86, 74–93. Page, D.L. 1981. Further Greek Epigrams (Cambridge). Pease, A.S. 1935. Publi Vergili Maronis Aeneidos liber quartus (Cambridge, MA). Petry, J. 1895. Quaestiones criticae ad Diras et Lydiam pertinentes (Munich). Pinkster, H. 2015. The Oxford Latin Syntax, vol. 1 (Oxford). Prósper, B.M. 2015. ‘Celtic and non-­Celtic divinities from ancient Hispania: power, daylight, fertility, water spirits and what they can tell us about Indo-­European morphology’, JIES 43, 1–56. Purser, L.C. 1909. [Review of Ellis 1907,] Hermathena 15, 441–57. Putsche, K. 1828. Valerii Catonis poemata (Jena). Ramires, G. 2005. ‘Parrasio lettore dell’Appendix Vergiliana nell’ incunabolo Neap. V. A. 36’, AION 27, 129–54. Reed, J.D. 1997. Bion of Smyrna: The Fragments and the Adonis (Cambridge). Reeson, J. 2001. Ovid: Heroides 11, 13 and 14 (Leiden). Reeve, M.D. 1976. ‘The textual tradition of the Appendix Vergiliana’, Maia 28, 233–54. Ribbeck, O. 1868. Appendix Vergiliana (Leipzig). Ribbeck, O. 1895. Appendix Vergiliana, 2nd ed. (Leipzig). Rivero García, L. 2004. ‘A reading of Ovid, Amores II 15’, Hermes 132, 186–210. Rodríguez Pantoja, M. 2006. ‘Dirae-Lydia: los argumentos de la métrica’, in C.  Santini et al., eds., Concentus ex dissonis: scritti in onore di Aldo Setaioli (Naples), 577–92. Rooy, C.A.  van. 1952. ‘Livy VII 2 and Valerius Maximus II 4, 4: two notes’, Mnemosyne 5, 236–42. Rosati, G. 2020. ‘Gallo in Virgilio e Saffo in Ovidio: due meta-­poeti nella riflessione della Lydia pseudo-­virgiliana’, Dictynna 17. Ross, D.O. 1975. Backgrounds to Augustan Poetry: Gallus, Elegy and Rome (Cambridge). Rothstein, M. 1888. ‘De Diris et Lydia carminibus’, Hermes 23, 508–24. Rupprecht, K. 2007. Cinis omnia fiat: Zum poetologischen Verhältnis der pseudo-­ vergilischen »Dirae« zu den Bucolica Vergils (Göttingen).

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

160 Bibliography Rupprecht, K. 2020. ‘Dirae’, in F. Zogg, ed., Appendix Vergiliana: lateinisch-­deutsch (Berlin), 44–59. Rynearson, N. 2009. ‘A Callimachean case of lovesickness: magic, disease, and desire in Aetia frr. 67–75 Pf.’, AJPh 130, 341–65. Saenger, G. [Зенгеръ, Г.] 1885. ‘Эпиграфическiя и другiя замѣтки’ [‘Epigraphic and other notes’], Извѣстія Историко-­филологическаго института [Bulletin of the Historic-­Philological Institute] 9, 1–51. Salvatore, A. 1994. ‘Da un «dramma» politico a un dramma esistenziale: le Dirae dell’Appendix Vergiliana e il problema dell’unità’, in Storia, poesia e pensiero nel mondo antico: studi in onore di Marcello Gigante (Naples), 549–64. Salvatore, A. 1997. ‘Dirae’, in A. Salvatore et al., Appendix Vergiliana (Rome), 10–23. Sampson, R. 2009. ‘The joys of Romance’, in S. Harrow, T. Unwin, eds., Joie de vivre in French Literature and Culture: Essays in Honour of Michael Freeman (Amsterdam), 35–50. Scaliger, J. 1572. Publii Virgilii Maronis appendix (Lyon). Schmidt, M. 1853. ‘Zu Valerius Cato’, Philologus 8, 190–2. Schniebs, A. 2021. ‘Lugares de mujer: espacio, genre y gender en las Dirae’, in S.M. Moro, M.E. Martí, eds., Mujeres en la literatura grecolatina: imágenes y discursos. Homenaje al Dr. Andrés Pociña (Rosario), 301–17. Scholl, W. 2014. Der Daphnis-­Mythos und seine Entwicklung: Von den Anfängen bis zu Vergils vierter Ekloge (Hildesheim). Schutter, K.H.E. 1953. ‘De Lydia et Diris carminibus’, Mnemosyne 6, 110–15. Sens, A. 2011. Asclepiades of Samos: Epigrams and Fragments (Oxford). Setaioli, A. 1999. ‘La poesia in Petronio Sat. 127.9’, Prometheus 25, 247–58. Shackleton Bailey, D.R. 1956. Propertiana (Cambridge). Shackleton Bailey, D.R. 1978. ‘Notes on minor Latin poets’, Phoenix 32, 305–25. Sider, D. 1997. The Epigrams of Philodemos: Introduction, Text, and Commentary (New York). Sillig, J. 1826. [Review of Eichstädt 1826,] Jahrbücher für Philologie und Paedagogik 1, 333–43. Silvino, T., and M. Poux. 2005. ‘Où est passé le vin de Bétique? Nouvelles données sur le contenu des amphores dites «à sauces de poisson et à saumures» de types Dressel 7/11, Pompéi VII, Beltrán II (Ier s. av. J.-C.–IIe s. apr. J.-C.)’, in L. Rivet et al., eds., Spécificités et diffusion de la céramique gallo-­ romaine en région Centre (Marseille), 501–14. Skutsch, O. 1956. ‘Zu Vergils Eklogen’, RhM 99, 193–201. Solodow, J.B. 1986. ‘Raucae, tua cura, palumbes: study of a poetic word order’, HSCPh 90, 129–53. Solta, G.R. 1967. ‘Lat. limpidus und seine Verwandten’, in W. Meid, ed., Beiträge zur Indogermanistik und Keltologie: Julius Pokorny zum 80. Geburtstag gewidmet (Innsbruck), 93–7. Sonntag, M. 1896. ‘Pseudo-­Vergil Lydia 39–41’, WKPh 13, 478. Spanoudakis, K. 2002. Philitas of Cos (Leiden). Spelman, H. 2018. ‘Heirs and educators: Pindar in the Epitaph for Bion’, Hermes 146, 107–9.

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

Bibliography  161 Stachon, M. 2014. Tractavi monumentum aere perennius: Untersuchungen zu vergilischen und ovidischen Pseudepigraphen (Trier). Stinton, T.C.W. 1976. ‘Si credere dignum est: some expressions of disbelief in Euripides and others’, PCPhS 22, 60–89. Sundt, P.A. 2021. ‘Orpheus and Sappho as model poets: blurring Greek and Latin love in Lament for Bion, Catullus 51, and Horace Odes 1.24’, in T.S. Thorsen et al., eds., Greek and Roman Love: The Poetic Connection (Berlin), 39–57. Swift, L. 2016. ‘Poetics and precedents in Archilochus’ erotic imagery’, in L. Swift, C. Carey, eds., Iambus and Elegy: New Approaches (Oxford), 253–70. Thomas, R.F. 2021. ‘Catullan intertextuality’, in I. Du Quesnay, T. Woodman, eds., The Cambridge Companion to Catullus (Cambridge), 47–69. Thorsen, T.S. 2014. Ovid’s Early Poetry: From His Single Heroides to His Remedia Amoris (Cambridge). Thorsen, T.S. 2021. ‘The beloved: figures and words’, in T.S. Thorsen et al., eds., Greek and Roman Love: The Poetic Connection (Berlin), 213–30. Thumiger, C. 2018. ‘Liebe als Krankheit: Eine Geschichte von Leib und Seele in griechischer und römischer Literatur und Medizin’, in N.  Reggiani, F.  Bertonazzi, eds., Parlare la medicina: fra lingue e culture, nello spazio e nel tempo (Milan), 253–73. Traina, A. 1995. ‘Un probabile verso di Ennio e l’apposizione parentetica’, MD 34, 187–93. Trappes-­Lomax, J.M. 2007. Catullus: A Textual Reappraisal (Swansea). Trimble, G. 2010. ‘A commentary on Catullus 64, lines 1–201’ (DPhil. diss., Univ. of Oxford). Trimble, G. forthcoming. Catullus: Poem 64 (Cambridge). Valladares, H. 2021. Painting, Poetry, and the Invention of Tenderness in the Early Roman Empire (Cambridge). Van den Abeele, E. 1969. ‘Remarques sur les “Dirae” et la “Lydia” de l’ “Appendix Vergiliana” ’, RhM 112, 145–54. Vollmer, F. 1910. Poetae latini minores, vol. 1 (Leipzig). Wackernagel, J. 1908. ‘Qua – qua. Lympha. Eruptum = ereptum’, ALLG 15, 213–21. Wagner, W. 1867. ‘On Ribbeck’s Virgil’, TPhS 12, 198–239. Waszink, J.H. 1948. ‘Varro, Livy and Tertullian on the history of Roman dramatic art’, VChr 2, 224–42. Watson, L.C. 2003. A Commentary on Horace’s Epodes (Oxford). Watson, P. 1985. ‘Axelson revisited: the selection of vocabulary in Latin poetry’, CQ 35, 430–48. Watt, W.S. 2001. ‘Notes on the Appendix Vergiliana’, Eikasmos 12, 279–92. Wernsdorf, C. 1782. Poetae latini minores, vol. 3 (Altenburg). West, M.L. 1966. Hesiod: Theogony (Oxford). West, M.L. 2007. Indo-­European Poetry and Myth (Oxford). Wills, J. 1996. Repetition in Latin Poetry: Figures of Allusion (Oxford). Winkler, J.J. 1990. The Constraints of Desire: The Anthropology of Sex and Gender in Ancient Greece (New York).

OUP CORRECTED PROOF – FINAL, 22/09/23, SPi

162 Bibliography Wistrand, E. 1958. Horace’s Ninth Epode and Its Historical Background (Gothenburg). Witzke, S.S. 2016. ‘Violence against women in ancient Rome: ideology versus reality’, in W.  Riess, G.G.  Fagan, eds., Topographies of Ancient Greek and Roman Violence (Ann Arbor), 248–74. Zaboulis, H. 1978. ‘Appendix Vergiliana: Dirae’, Philologus 122, 207–23. Zeitlin, F.I. 1990. ‘The poetics of erōs: nature, art, and imitation in Longus’ Daphnis and Chloe’, in D.M.  Halperin et al., eds., Before Sexuality: The Construction of Erotic Experience in the Ancient Greek World (Princeton), 417–64. Ziogas, I. 2021. Law and Love in Ovid: Courting Justice in the Age of Augustus (Oxford). Zogg, F. 2018. ‘Carmina Virgilii mitte minora, precor: Die Überlieferung der ‹Appendix Vergiliana› im Mittelalter’, MLatJb 53, 27–45.

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

General index abduction  116, 142–3 see also rape, violence Academy 111 Achilles 126 Adonis  12, 20–3, 55, 57, 64, 70–1, 74, 130, 132–9 see also Dumuzi, Tamuz age golden  65–6, 93–4, 110–13, 145–6 heroic  94, 111–12, 114–15, 117, 145–6 iron  111, 145 allegory, biographical  26 Anchises 132–3 animals  17–20, 53, 56, 92–4, 98, 101, 110, 118, 145 see also bull, goat Aphrodite see Venus Apollo  12, 35, 101, 104, 105–6, 108–9, 131 Appendix Vergiliana  3, 14, 41 Ares see Mars Ariadne  12, 81, 110, 114–15, 116 Aurora  68, 102, 118, 127, 129–30, 141, 142–4 beard 140–2 beauty  23, 57, 64, 68, 87–8 Bion  24–9, 64, 119 fragments  25, 27–8, 94 Lament for Adonis  4, 6, 15, 17, 20–5, 29, 55, 56, 74 see also Lament for Bion birds  26–7, 34, 76, 79 body  22, 24, 66, 68, 70, 73, 83, 85, 122 Botticelli 72 breasts  68, 69, 70

bucolic  5, 6, 10–12, 14–17, 21, 24, 26, 28–9, 32, 40–1, 54–5, 64, 74, 87, 111, 132, 136 bull  19, 92, 94–5, 110 Calvus  38 n. 86, 57 Calypso 142 catasterism  12, 101, 109 Cato  see Valerius Catullus  9–10, 11, 39–40, 56–7, 78, 86, 116–17, 137–8 Cephalus 142–3 chill  83, 85 Ciris  9–10, 39 Clitus 143 colloquialism  39–41, 63, 78, 109, 140 Corona Borealis  114–15 countryside  3, 5, 8, 22–3, 33, 35, 36, 37–8, 55–7, 97 Crates of Athens  111 Danae  82, 89–91, 130 Daphne  105–6, 108–9, 110 Daphnis (Longus)  60, 71, 93 Daphnis (Theocritus)  15–16, 18–21, 25, 30, 54–5, 74–5, 92, 147 Dawn  see Aurora death of a bucolic hero  29 of Adonis  23, 55, 132 of Bion  26, 74 of Daphnis  15, 18, 54, 74 of Orpheus  29, 74 of the Lydia speaker  3, 5, 8, 21, 28–9, 67, 74, 81, 83, 118, 123–4, 126, 132, 148 see also lovesickness defloration  15, 69, 70, 121–2, 136 see also virginity

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

164  General index Dionysius I of Syracuse  26 Διὸς ἀπάτη  90, 130 Dirae  3–11, 13, 34–7, 39, 41–2, 54, 57, 63, 79, 86, 119 docta puella  see learning doctrina  see learning double entendre  68, 71 see also euphemism, wordplay Dumuzi 18 see also Adonis elegy  13, 14, 29, 31, 35, 58, 61, 81, 86, 87, 89, 114, 139 enchantment  29–30, 34, 74, 76, 79 Endymion  101–2, 105, 142 envy  19, 27–8, 53–6, 58, 64–5, 91–2, 129 see also jealousy Eos  see Aurora epic  33, 61, 117 epigram  13, 90 epiphany 74 Eriphanis 16–17 Eros 27 erotodidaxis  see instruction etymology  6, 12, 34, 61, 78–9, 103–4, 127 euphemism  60, 82, 86, 120, 121, 125, 127, 135, 136–7, 145 Europa  82, 130 exempla  6, 20, 55, 93–4, 118, 129–30, 142 exile  3, 8 fallacy, pathetic  17–21, 28–9, 74 fame  112, 126 fantasy  34, 64, 66–8, 70, 82, 111, 127 see also wish farm  3, 5, 10, 27–8, 36, 55 see also countryside fate  53–4, 92, 118, 123–4, 128–9, 146–8 feet  60, 66–7, 68 fever 83 fingers 68 flowers  6, 23, 37, 60, 64, 70–1, 72, 111, 134–6 see also defloration

Fraenkel, Eduard  6–7 frame, narrative  5–6, 55 Galatea 25–6 Gallus  11, 14 n. 37, 30–2, 56, 57, 72, 80, 86, 122 see also schema Cornelianum genitals 70 see also penis goat  19, 38, 92–4, 97–8, 110 gods  12, 20, 35, 53, 66, 82, 89, 92, 99, 101, 102, 107–10, 113, 116, 118–19, 129, 142, 144–5 Graaf, Cornelis van der  4, 39–40 grapes  22, 33, 60, 67–71 grass  22–3, 34, 54, 67–70, 132, 135–6 habit, language of  101 heat  24, 85 Hephaestus  see Vulcan Hera  see Juno herdsman  6, 18–20, 25–6, 29, 92, 132 heroes  19, 53, 92, 118, 144–6 see also age, heroic homoeroticism  94, 111 see also pederasty Horace  33–4, 36, 79 hyperbole  26, 83, 87 see also uniqueness iconicity, linguistic  12–13, 94–5 impotence  36, 137 injustice  53, 118, 129, 145 instruction erotic  26–9, 53, 60, 64, 119, 124 poetic  26–7, 64 see also learning invention of love  16–17, 28, 119–21, 127, 130 of poetry  16 of sexuality  119 Jacobs, Friedrich  3, 10 Jason  12, 110, 114–16 jealousy  12, 36, 54–6, 58, 82, 91, 138–9 see also envy

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

General index  165 Juno  12, 90, 130–2, 135 Jupiter  12, 54, 82, 89–91, 118, 129–32 Lament for Bion  6, 15, 17, 24–30, 74, 119 land confiscation  8, 10, 34, 36 learning, poetic  11–13, 16, 31–2, 40, 54, 87 see also instruction love clandestine  28, 53, 74, 119, 129–30 stages of  60–1 see also homoeroticism, instruction, invention, sex lovesickness  3, 8, 15, 25, 29, 31, 37–8, 58–9, 81–4, 92, 97, 123–4, 132 Lucretia  121, 125 Lucretius  9, 39–41, 111 Luna  101–2, 105, 142 Lydia as lover  8, 13, 15, 22–3, 28, 53–4, 59–60, 63–4, 66–71, 74, 81–2, 86, 91–2, 119, 120, 125, 130, 132–3 as nymph  8, 13, 61, 87, 123 as poet/singer  16–17, 21–2, 25–31, 61–4, 74, 76, 77, 79–81, 87 magic 36 μακαρισμός  33, 55–6, 64–6 marriage  116, 119–20, 122, 127, 130 Mars  12, 30, 133, 138–40 Martial  32, 87 Medea  110, 114–16 Meliboeus 55 Melito of Sardis  139 metamorphosis  54, 55, 101, 105, 109, 130 Milton 67 monologue  5–6, 55 Moon  see Luna mortals  12, 19–20, 33, 35, 38, 92, 93, 100, 110, 113–15, 118, 133, 142 see also animals, heroes nature  5, 17–19, 21, 27, 28–9, 34, 61, 74–5, 80, 111 joy of  74–5 see also countryside, fallacy

Orion  123, 142 Orpheus  28–9, 74–6, 79, 119 Ovid  9, 28, 59, 119–20, 124 Amores  36–7, 55, 64 Ars 119–20 Epistula Sapphus 31 Remedia  37–8, 97, 124 pastoral  see bucolic pederasty  28, 119 see also homoeroticism penis  37, 136 performance  16, 22, 54, 61, 62, 80–1 Phaenias 26 Philetas 92–3 Philitas 92–3 Philoxenus 26 Phoebus  see Apollo Plautus  40, 90 pleasure  87, 119, 127, 132, 136, 137 Pleiades  12, 109 poeta doctus  see learning poetry Augustan  15, 33–8 Greek  6, 12, 14–30, 32 ‘neoteric’  9, 39–40, 72, 99 see also bucolic, elegy, epic, epigram Polemo of Athens  111 Polyphemus 25–6 prosaism  33, 40–1, 56–7, 79, 84, 98, 101, 113, 115, 117, 146 πρῶτος εὑρετής  see invention rape  70, 116, 120, 125 see also abduction, seduction, violence refrains  3, 4, 24, 53–4, 56, 63 rivers  6, 18–19, 21, 34, 38, 75–7, 81 Sappho  13, 31–2, 87, 105, 136 Scaliger, Joseph Justus  3, 10 schema Cornelianum 72 seduction  12, 26, 90, 118, 121, 125, 130, 132, 135, 142 see also abduction, defloration, rape Selena  see Luna

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

166  General index separation (of lovers)  12–13, 15, 53, 81, 85, 93, 94, 96, 101–2, 108, 110, 132, 147 settlement of veterans  see land confiscation sex  15–16, 22–3, 60, 64, 66–71, 82, 86, 92–3, 99, 109, 110–11, 119–121, 122, 124–5, 126–8, 129, 130, 132–3, 136–7, 142, 145 alfresco  23, 135 extramarital  15–16, 118–20, 127, 129 illicit  121–2, 124, 127, 129–30, 132, 144–5 orgasm 126 see also defloration, love, pleasure, seduction, violence singer  5, 16, 26, 29, 40, 54 see also learning Sol  12, 102, 104, 106 σπονδειάζοντες  9–10, 39–40 springs  19, 21, 34, 36–7, 75–9 stanza  3, 5, 24, 53–4, 60, 63–4, 81, 92, 93, 99, 101, 110, 118, 129, 144, 147 Sun  see Sol Tamuz 139 teacher  see instruction Theocritus  4, 6, 15–21, 29–31, 54–5, 83, 92, 111–12, 132 Theseus  12, 110, 114 Tibullus  34–6, 111 Ticida  11, 13, 80 Tithonus  102, 141, 143

Tityrus  55, 64 tradition, literary  15, 17–18, 24, 28–32, 88–9, 115 transgression  110, 121–2, 124, 129 uniqueness, of the beloved  31, 82, 91 see also hyperbole Valerius Cato  3, 10–11, 13–14, 32–3 Varro, M. Terentius  6, 12, 34, 61, 78–9 Varro Atacinus  10, 39–40 Venus  12, 19–23, 25, 37, 54, 55, 64, 70–2, 74, 118, 129–30, 132–41 violence  22, 71, 93, 111, 120 see also rape Virgil  9–10, 14, 28 Aeneid  77–9, 138 Eclogues  6, 8, 10–11, 13, 14–15, 17, 29–33, 38, 41, 56–7, 72, 75, 135, 143 Georgics 143 virginity  67, 69, 70, 71, 120–1, 123, 136 Vulcan  12, 138–42 wish, irreal  124, 145–6 see also fantasy woods  5, 8, 34, 57, 95, 98 word order  58, 72, 76, 94, 96, 102, 103, 113, 128, 141 wordplay  12, 61, 80, 90 see also double entendre, etymology Zeus  see Jupiter

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Index of Greek and Latin words δαϕνηϕορεῖν, δαϕνηϕόρος 106 δρέπεσθαι 22 λάθρῃ  73–4, 138 accumbere 136–7 alloqui 61 atque 143–4 caelicola 117 caelum 103–4 cedere 102–3 cogere 124 credere  40, 139 crudelis 100–1 cura  13, 30 n. 68, 31 n. 68, 80 decerpere  33, 40, 69 dicere 131 dolor 84 domina 86 durus 117–18 error 129 fabula  35, 40, 89–90 formosus  56–8, 87–8 furtim 73–4 gaudium  82, 126–7 interdum 62 inuidere 56 iucundus  40, 81 libare 131–2 Lydia 61

lympha  6, 12, 34, 77–9 manere 149 maturus 123–4 meditari 62–3 migrare 35–6 misereri 105 mortalis  33, 114 mugire 95–6 mugitus  38 n. 86 mundus 103–4 nitidus 104 nunc 128 Phoebus  12, 104 pudor 121 puella 58 querela 80–1 sterilescere 37 stipendium 71 suspirare 59 tabescere  40, 83–4 taurus 90 temerare 122–3 turpare  40, 141 uaccula  39, 95 uestigare  40, 98 uinum lymphatum 78 uiolare 121–2 uoluptas  82, 127–8

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Index of passages discussed AP 7.10.6–8:  29 n. 65 Apollonius Rhodius Arg. 3.675:  84 Arg. 3.1001–4:  114 Bion Ep. Adon. 22:  22 Ep. Adon. 58–63:  25–6 Catullus 64.332: 137 64.403–4: 116 86: 57 Cicero Arat. fr. 18.1:  120–1 Arat. fr. 34.337:  103 Ciris 99–100: 144 414–15: 131 CLE 954: 122 1166.6:  32, 87 Dionysius of Halicarnassus 4.82.3: 125 Dirae 15–18: 36–7 Germanicus 647: 126 Hermesianax fr. 3.73–4:  26 n. 60 Horace Carm. 2.3.11–12:  77

Epod. 2.1–2, 19–20:  33 Epod. 2.23–8:  33–4 Longus 2.7.4–6: 92–3 Martial 10.32.6:  32, 87 Ovid Am. 3.7.7–8:  37 Am. 3.7.31–4:  36–7 Am. 3.7.65–6:  37 Ars 1.31–4:  119–20 Fast. 4.459:  96 Her. 2.115:  131 Her. 4.98:  135 Her. 16.161:  131–2 Met. 10.557–9:  71 Rem. 177–84:  37–8 Rem. 466:  124 Pliny Nat. 17.74:  102 Propertius 1.17.2: 61 Theocritus 4.12: 96 Tibullus 2.3.3–4: 35–6 2.3.31–2, 35:  35 [Tibullus] 3.6.21: 65 Ticida fr. 103 Hollis:  11, 13, 80

OUP CORRECTED PROOF – FINAL, 25/09/23, SPi

Index of passages discussed  169 Valerius Flaccus 1.627: 122 Varro L. 5.71, 7.87:  78–9

Virgil Aen 2.792 [= 6.700]:  138 Ecl. 1.11:  55 Ecl. 6.53:  135 Ecl. 10.42–5:  30–1