Dialect, Diction, and Style in Greek Literary and Inscribed Epigram 9783110498790, 9783110496499

Citation preview

Dialect, Diction, and Style in Greek Literary and Inscribed Epigram

Trends in Classics – Supplementary Volumes

Edited by Franco Montanari and Antonios Rengakos Scientific Committee Alberto Bernabé · Margarethe Billerbeck Claude Calame · Philip R. Hardie · Stephen J. Harrison Stephen Hinds · Richard Hunter · Christina Kraus Giuseppe Mastromarco · Gregory Nagy Theodore D. Papanghelis · Giusto Picone Kurt Raaflaub · Bernhard Zimmermann

Volume 43

Dialect, Diction, and Style in Greek Literary and Inscribed Epigram Edited by Evina Sistakou and Antonios Rengakos

ISBN 978-3-11-049649-9 e-ISBN (PDF) 978-3-11-049879-0 e-ISBN (EPUB) 978-3-11-049702-1 ISSN 1868-4785 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dn.de. © 2016 Walter de Gruyter GmbH, Berlin/Boston Logo: Christopher Schneider, Laufen Printing and binding: CPI books GmbH, Leck ∞ Printed on acid-free paper Printed in Germany www.degruyter.com

Preface Epigram is a broad generic category of ancient Greek poetics, as it encompasses both the literary form written by distinguished poets and the anonymous verses inscribed on stone and various objects, and spans almost a dozen centuries from the archaic and classical period until the rise of the Byzantine empire and beyond. Versatility and variety—in content, tone and diction—are the cornerstones of this genre, whose popularity is evidenced by the numerous epigrammatic corpora that have come down to us. Yet it was the artistic innovation introduced by the Hellenistic learned poets to the epigram that marks the turning point in the history of the genre. The main ideas explored in the present volume were probably conceived for the first time in Ptolemaic Alexandria by the very same poets who saw in the epigram, a poetic form whose brevity dramatically increases the power of the word, the λεπτός genre par excellence. The aim of the volume at hand is to offer new insights into a key feature of the epigrammatic genre, namely its rhetoric in all its possible manifestations in both literary and inscriptional epigram. The spotlight is turned on Hellenistic and post-Hellenistic epigrams, precisely because it was then that the poets became acutely aware that in epigram ‘every word counts’. Twenty-one contributors bring to the fore the various aspects of this premise, and highlight language, style, and structure of words as a heuristic tool for the interpretation of epigram. The first part is dedicated to dialect and diction. Choice of dialect and dialect mixing are seen as part of the poet’s strategy in the early Hellenistic era, whereas imitation and variation in terms of diction become a rule in later epigram. Epigram as a formal construction and a genre dependent on sophisticated word design is explored in the second part; discussions feature letter- and word-play, the use of pentameters and syntactical patterns, the rhetoric of riddles and technopaegnia, and the introduction of the epigrammatic diction into epic. The remaining chapters of the volume come under the heading of ‘style’. In literary epigram style seems to be defined by theme. Three different thematic areas are viewed from the perspective of their idiosyncratic style, and hence three styles emerge as paradigmatic, namely the sepulchral, the philosophical, and the pastoral. In the last part of the volume a fresh approach to epigrammatic discourse is pursued: the concluding chapters demonstrate that inscriptions feature elaborate linguistic and rhetorical patterns and that style in these anonymous verses is equally essential as in their artistic counterparts. The papers assembled in this volume were presented during the 9th Trends in Classics conference held in Thessaloniki (29 – 31 May, 2015), which was organ-

VI

Preface

ized by the Department of Classics of the Aristotle University and the Centre for the Greek Language. This international conference was part of the Action “Ancient Greek Dialects of Vital Importance for the Continuity of the Greek Language and the Cultural Tradition”, co-financed and managed by the Greek Ministry of Education and Religious Affairs and the European Union (NSRF 2007– 2013). Our thanks go to the Welfare Foundation for Social and Cultural Affairs (K.I.K.P.E.) for sponsoring the conference; we owe a special debt of gratitude to the VicePresident of the foundation, Manos Dimitrakopoulos, who believed in the Trends in Classics project in the first place. Evina Sistakou Antonios Rengakos

Thessaloniki, August 2016

Table of Contents Preface

V

Dialect and Diction Ewen Bowie Doing Doric

3

Dee L. Clayman Callimachus’ Doric Graces: 15 GP = 51 Pf.

23

Taylor Coughlan Dialect and Imitation in Late Hellenistic Epigram

37

Lucia Floridi The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

Form and Design Regina Höschele “Unplumbed Depths of Fatuity?” Philip of Thessaloniki’s Art of 105 Variation Gregory O. Hutchinson Pentameters 119 Demetra Koukouzika Epigrams in Epic? The Case of Apollonius Rhodius Jan Kwapisz When Is a Riddle an Epigram?

139

151

Giulio Massimilla The ἀπὸ κοινοῦ Construction of Prepositions as a Feature of the Epigrammatic Style 173

71

VIII

Table of Contents

Style in Literary Epigram a) Sepulchral Style Egbert J. Bakker Archaic Epigram and the Seal of Theognis Michael A. Tueller Words for Dying in Sepulchral Epigram

195

215

b) Philosophical Style Benjamin Acosta-Hughes A Little-Studied Dialogue: Responses to Plato in Callimachean 237 Epigram Kathryn Gutzwiller Style and Dialect in Meleager’s Heraclitus Epigram Richard Hunter A Philosophical Death?

253

269

c) Pastoral Style Marco Fantuzzi Novice Pastoral Eros and Its Epigrammatic Critics

281

Nita Krevans Pastoral Markers in Hellenistic Epigram: The Fan-Fiction Approach

Style in Inscribed Epigram Francesca Angiò A Sundial for a Deceased Woman: Two Epigrams from Pamphylia (I – II A. D.) 311

297

Table of Contents

Valentina Garulli Playing with Language in Everyday Poetry: hapax legomena in Inscribed 323 Funerary Epigrams David Petrain Hearing Heracles on the Tabula Albani

335

Andrej Petrovic Casualty Lists in Performance. Name Catalogues and Greek Verse-Inscriptions 361 Ivana Petrovic The Style and Language of Epigrammatic Programmata List of Contributors Index

421

415

391

IX

Dialect and Diction

Ewen Bowie

Doing Doric This paper has two parts. In the longer, first part I offer some observations on the use of Doric by poets whose work has been transmitted initially via the Garland of Philip and then via the Anthology. In the second and shorter part I consider the case of five poems from a sepulchral monument in Nicaea, poems not read by modern epigraphists but preserved by Book 15 of the Anthology.

The Garland of Philip Some of Philip’s poets, especially those of the middle and latter part of the first century BC, have no truck with Doric. This is no surprise in Philodemus, whom neither his Gadarene origin nor his Italian residence, whether in Rome or on the bay of Naples, would have given an incentive to compose any of his 29 surviving epigrams in Doric. Nor is it a surprise in the 15 poems of Parmenion and Diodorus of Sardis respectively, despite the Spartans at Thermopylae being the subject of Parmenion x Gow-Page (= AP 9.304) and a Tarentine setting being that of Diodorus ix Gow-Page (= AP 7.700). But if it is right to think that Honestus was domiciled in Corinth it is somewhat surprising that none of his 22 poems shows any sign of Doric,¹ and also perhaps surprising that Crinagoras, despite his Mytilenean origin, adhered throughout his 51 surviving poems to Attic-Ionic.² Less can be argued on the basis of poets of whose epigrams only very few survive, but for what it is worth I note that there is no Doric in the single surviving epigrams of Apollonius, Automedon, Boethus, Diotimus, Etruscus, Quintus, Scaevola or Serapion; in the two epigrams of Sabinus; in the three of Macedonius, Maeandrius, or Polemo; or in the four of Diocles or Secundus. In certain poets who did choose Doric for some of their surviving poems it might be argued that the dialect is chosen either when there is a thematic or contextual trigger in the poem, or when the poem follows or refashions the work of an earlier poet strongly associated with Doric.

 For Honestus see Jones ,  – ; Bowie  [], .  But for the absence of local dialect features from the works of both Theophanes and Crinagoras see Bowie , esp. .

4

Ewen Bowie

Thus several poems using Doric evoke either a Leonidean or a Theocritean pastoral world and their poets may be choosing to use Doric so as to stress that aspect of a poem. Of Myrinus’ four transmitted epigrams it is only iii Gow-Page (=AP 7.703), on Thyrsis asleep, that uses Doric: Θύρσις ὁ κωμήτης, ὁ τὰ νυμφικὰ μῆλα νομεύων, Θύρσις ὁ συρίζων Πανὸς ἴσον δόνακι, ἔνδιος οἰνοπότης σκιερὰν ὑπὸ τὰν πίτυν εὕδει φρουρεῖ δ’ αὐτὸς ἑλὼν ποίμνια βάκτρον Ἔρως. ἆ Νύμφαι, Νύμφαι, διεγείρατε τὸν λυκοθαρσῆ βοσκόν, μὴ θηρῶν κύρμα γένηται Ἔρως. Thyrsis the rustic, who pastures the nymphs’ sheep, Thyrsis whose piping equals Pan’s reed, the noon wine-bibber, is under the shady pine, asleep, and Eros himself has taken a crook and guards his flocks. – oh Nymphs! Nymphs! Wake up the wolf-bold herdsman, lest the beasts’ prey should be Eros.

Of Thallus’ poems it is just one, v Gow-Page (= AP 9.220), on the pleasures of sex under a plane-tree entwined with a vine, his only surviving treatment of a rural (though not properly pastoral) scene, that is also his only epigram to exploit Doric: Ἁ χλοερὰ πλατάνιστος ἴδ’ ὡς ἔκρυψε φιλεύντων ὄργια τὰν ἱερὰν φυλλάδα τεινομένα· ἀμφὶ δ’ ἄρ’ ἀκρεμόνεσσιν ἑοῖς κεχαρισμένος ὥραις ἡμερίδος λαρῆς βότρυς ἀποκρέμαται. οὕτως, ὦ πλατάνιστε, φύοις· χλοερὰ δ’ ἀπὸ σεῖο φυλλὰς ἀεὶ κεύθοι τοὺς Παφίης ἑτάρους. The green plane-tree – see how it has hid the lovers’ rites, stretching out its sacred foliage: and around its own branches, a delight to the Seasons, a sweet tended vine’s bunch dangles down. So may you grow, plane-tree, and may your growth of green foliage ever conceal the Paphian’s companions.

Another example is Adaeus ii Gow-Page (= AP 6.258), a dedication by Crethon to Demeter ‘characteristic of the imitators of Leonidas’ (Gow-Page ad loc.), ‘in the Leonidean manner’:³

 Cf. Magnelli , .

Doing Doric

5

Τὰν ὄιν, ὦ Δάματερ Ἐπόγμιε, τάν τ’ ἀκέρωτον μόσχον καὶ τροχιὰν ἐν κανέῳ φθοΐδα σοὶ ταύτας ἐφ’ ἅλωος, ἐφ’ ᾇ πολὺν ἔβρασεν ἄντλον Κρήθων καὶ λιπαρὰν εἶδε γεωμορίαν, ἱρεύει, πολύσωρε· σὺ δὲ Κρήθωνος ἄρουραν πᾶν ἔτος εὔκριθον καὶ πολύπυρον ἄνοις. The ewe, Demeter of the Furrow, and the hornless calf, and the wheel-cake in its basket, to you, on this threshing-floor, where a rich pile he has winnowed, and seen a brilliant harvest, Crethon consecrates, o many-heaped one! May you make Crethon’s plough-land fine in barley and plentiful in wheat every year.

The name Crethon is already in Adaeus’ model, Leonidas lxxv Gow-Page (= AP 7.740): αὕτα ἐπὶ Κρήθωνος ἐγὼ λίθος, οὔνομα κείνου δηλοῦσα· Κρήθων δ’ ἐγχθόνιος σποδιά. ὁ πρὶν καὶ Γύγῃ παρισεύμενος ὄλβον, ὁ τὸ πρὶν βουπάμων, ὁ πρὶν πλούσιος αἰπολίοις, ὁ πρίν – τί πλείω μυθεῦμ’ ἔτι; πᾶσι μακαρτός, φεῦ, γαίης ὅσσης ὅσσον ἔχει μόριον. Here upon Crethon stand I, a stone, and his name I display; but Crethon’s in the earth, he’s ash. He who before equated his prosperity even with Gyges, who before, was cattle-rich, who before was wealthy in flocks, who before – why do I say more? Blessed in every way alas! of so much land how small a piece he has!

But as Gow-Page noted Adaeus is also influenced by Theocritus 7.154– 7.⁴ Both such a pastoral setting and Doric subjects seem to have been relevant to Erucius’ decisions to use Doric. Of his fourteen transmitted poems Erucius chooses Doric for no less than eight, in four of these apparently for its pastoral associations; and of those for which he did not choose to use any Doric two were virtually excluded by their subject matter – an epigram on an Athenian woman who died at Cyzicus which begins programmatically with the ethnic ᾿Aτθίς, Gow-Page vi (= AP 7.368),⁵ and another on the Athenian poet Sophocles,

 οἷον δὴ τόκα πῶμα διεκρανάσατε, Νύμφαι, | βωμῷ πὰρ Δάματρος ἁλωίδος; ἇς ἐπὶ σωρῷ | αὖτις ἐγὼ πάξαιμι μέγα πτύον, ἃ δὲ γελάσσαι | δράγματα καὶ μάκωνας ἐν ἀμφοτέραισιν ἔχοισα.  Merkelbach-Stauber , // = GVI .

6

Ewen Bowie

Gow-Page xi (= AP 7.36). The poems with some Doric have almost all a pastoral setting: Erucius i Gow-Page (= AP 6.96), where two ox-herds sacrifice to Pan:⁶ Γλαύκων καὶ Κορύδων, οἱ ἐν οὔρεσι βουκολέοντες, ᾿Aρκάδες ἀμφότεροι, τὸν κεραὸν δαμάλην Πανὶ φιλωρείτᾳ Κυλληνίῳ αὐερύσαντες ἔρρεξαν καί οἱ δωδεκάδωρα κέρα ἅλῳ μακροτένοντι ποτὶ πλατάνιστον ἔπαξαν εὐρεῖαν, νομίῳ καλὸν ἄγαλμα θεῷ. Glaucon and Corydon, who herd cattle in the mountains, Arcadians both, took the horned calf for Pan, the mountain-loving Cyllenian, drew back its head, and slaughtered it; and its twelve-palmed horns they pinned with a long-tapered nail to a plane tree that stood broad, a fair ornament for the god of pastures.

Erucius iii Gow-Page (= AP 9.558), in which dogs frighten off wolves: Ὁ τράγος ὁ Κλήσωνος ὅλαν διὰ πάννυχον ὄρφναν αἶγας ἀκοιμάτους θῆκε φριμασσόμενος— ὀδμὰ γάρ μιν ἔτυψε λύκου χιμαροσφακτῆρος τηλόθε πετραίαν αὖλιν ἀνερχομένου— μέσφα κύνες κοίτας ἀνεγέρμονες ἐπτοίασαν θῆρα μέγαν· τραγίνους δ’ ὕπνος ἔμυσε κόρας. Clyson’s billy-goat throughout all the dark night kept the nanny-goats awake with his snorting – for the scent of a she-goat-slaying wolf had struck him as it made its way up from afar to the rock-built fold – until the dogs, aroused from their rest, scared off the great beast, and sleep closed the goats’ eyes.

Erucius iv Gow-Page (= AP 9.824), where Pan promises success to hunters. The dialect is very Doric, the subject reworks Leonidas xxix Gow-Page (= AP 9.337): Εὔστοχα θηροβολεῖτε, κυναγέται, οἱ ποτὶ ταύταν Πανὸς ὀρειώτα νισσόμενοι σκοπιάν, αἴτε λίνοις βαίνοιτε πεποιθότες, αἴτε σιδάρῳ, αἴτε καὶ ἰξευταὶ λαθροβόλῳ δόνακι·

 The Doric forms are φιλωρείτᾳ () and ἅλῳ, ποτὶ, and ἔπαξαν (). For a succinct and judicious statement of the debated relationship between this poem and the Arcades ambo of Vergil, Eclogue . –  see Clausen .

Doing Doric

7

κἀμέ τις ὑμείων ἐπιβωσάτω· οἶδά ποτ’ ἄγραν κοσμεῖν καὶ λόγχαν καὶ λίνα καὶ καλάμους May you hit your target, huntsmen, who to this peak of mountain-dwelling Pan come, whether you advance trusting in nets, or in iron, or indeed are fowlers with a stealthy-striking rod; and let each of you call upon me: I know how to set the foot-trap, and spear, and nets, and reeds.

The Leonidas poem is clearly Erucius’ model, but Erucius has eight Doricisms in his six lines (κυναγέται, ποτὶ, ταύταν, ὀρειώτα, σιδάρῳ, ἐπιβωσάτω, ποτ’(ὶ), λόγχαν), Leonidas only one in his four (βόασον, line 3): Εὐάγρει, λαγόθηρα, καὶ εἰ πετεεινὰ διώκων ἰξευτὴς ἥκεις τοῦθ’ ὑπὸ δισσὸν ὄρος, κἀμὲ τὸν ὑληωρὸν ἀπὸ κρημνοῖο βόασον Πᾶνα· συναγρεύω καὶ κυσὶ καὶ καλάμοις. Good hunting, hare-chaser, or if pursuing winged things you come with lime to beneath this twin peak, and from a crag call upon me, the forest-watcher, Pan: I shall join your hunt with dogs and reeds.

Erucius vii Gow-Page (= AP 7.174), an epitaph for a cowherd, is marked by light Doric: Οὐκέτι συρίγγων νόμιον μέλος ἀγχόθι ταύτας ἁρμόζῃ βλωθρᾶς, Θηρίμαχε, πλατάνου· οὐδέ σευ ἐκ καλάμων κερααὶ βόες ἁδὺ μέλισμα δέξονται σκιερᾷ πὰρ δρυῒ κεκλιμένου. ὤλεσε γὰρ πρηστήρ σε κεραύνιος· αἱ δ’ ἐπὶ μάνδραν ὀψὲ βόες νιφετῷ σπερχόμεναι κατέβαν. No more do you fashion your pastoral tune near this tall plane tree, Beast-fighter, nor will the horned cattle listen to the sweet tuning from your reeds as you recline beside a shady oak. For a lightning bolt destroyed you: and to their byre your cows came down late, hurried by snow.

Two other poems seem to use Doric not only because of a pastoral setting but also because their speaker is Dorian. One, Erucius ii Gow-Page (= AP 9.237), on a statue of Heracles about which an unnamed interlocutor asks a cowherd, is noted by Gow-Page as exhibiting a ‘severer Doric dialect’:

8

Ewen Bowie

‘Βουκόλε, πρὸς τῶ Πανός, ὁ φήγινος, εἰπέ, κολοσσὸς οὗτος, ὅτῳ σπένδεις τὸ γλάγος, ἔστι τίνος;’ ‘Τῶ λειοντοπάλα Τιρυνθίω· οὐδὲ τὰ τόξα, νήπιε, καὶ σκυτάλην ἀγριέλαιον ὁρῇς’. ‘Χαίροις, ᾿Aλκείδα δαμαληφάγε, καὶ τάδε φρούρει αὔλια κἠξ ὀλίγων μυριόβοια τίθει’. ‘Cow-herd, in Pan’s name tell me, the oak statue here, for whom you pour your libation of milk – whose is it?’ ‘The Tirythian lion-wrestler’s: do you not even see his bow, you idiot, and his club of wild-olive?’ ‘Hail, calf-eating Alcides, and guard these byres, and from few make their cattle countless’.

The second is Erucius v Gow-Page (= AP 6.255), both pastoral and at the same time a vehicle for the thoughts of an Ambraciot dedicator: it uses several Doric forms despite touches of epic colour:⁷ Τοῦτο Σάων τὸ δίπαχυ κόλον κέρας Ὡμβρακιώτας βουμολγὸς ταύρου κλάσσεν ἀτιμαγέλου, ὁππότε μιν κνημούς τε κατὰ λασίους τε χαράδρας ἐξερέων ποταμοῦ φράσσατ’ ἐπ’ ἀιόνι ψυχόμενον χηλάς τε καὶ ἰξύας· αὐτὰρ ὃ βούτεω ἀντίος ἐκ παγέων ἵεθ’· ὁ δὲ ῥοπάλῳ γυρὸν ἀπεκράνιξε βοὸς κέρας, ἐκ δέ μιν αἰπᾶς ἀχράδος εὐμύκῳ πᾶξε παρὰ κλισίᾳ This docked horn of two cubits Saon the Ambraciot, the cow-milker, broke off from a bull that scorned his herd, after he had sought it out along the ridges and shaggy torrent-beds and spotted it on a river’s bank cooling its hooves and flanks. But right for the cowherd it lunged from the stream – and he with his club de-headed the curved horn from the ox’s head, and on a tall pear tree fixed it, beside his fine-lowing byre.

The remaining poem of Erucius to use Doric seems to do so exclusively because of its Dorian subject: in Erucius xii Gow-Page (= AP 7.230) a Spartan mother castigates and kills her son who had fled from the battlefield (i. e. a τρέσας). The Doric appears both in the description and in the mother’s words that are quoted:

 Epic colour noted by Gow-Page , . – . For a defence of the Palatinus’ αὐτᾶς and a rejection of Gow-Page’s emendation αἰπᾶς see Giangrande ,  – .

Doing Doric

9

Ἁνίκ’ ἀπὸ πτολέμου τρέσσαντά σε δέξατο μάτηρ, πάντα τὸν ὁπλιστὰν κόσμον ὀλωλεκότα, αὐτά τοι φονίαν, Δαμάτριε, αὐτίκα λόγχαν εἶπε διὰ πλατέων ὠσαμένα λαγόνων· ‘Κάτθανε, μηδ’ ἐχέτω Σπάρτα ψόγον· οὐ γὰρ ἐκείνα ἤμπλακεν, εἰ δειλοὺς τοὐμὸν ἔθρεψε γάλα’. When you fled battle and your mother received you after you had lost all your hoplite gear, it was she herself, Damatrios, who forthwith a deadly spear thrust through your broad flanks, with the words ‘Die, and may Sparta bear no blame: for it is not she who erred, if it was cowards my milk nourished’.

By contrast I can see no obvious explanation for the choice of Doric in Erucius x Gow-Page (= AP 6.234), in which a Gallus dedicates his kit to Cybele: Γάλλος ὁ χαιτάεις, ὁ νεήτομος, ὡπὸ Τυμώλου Λύδιος ὀρχηστὰς μάκρ’ ὀλολυζόμενος, τᾷ παρὰ Σαγγαρίῳ τάδε Ματέρι τύμπαν’ ἀγαυᾷ θήκατο καὶ μάστιν τὰν πολυαστράγαλον ταῦτά τ’ ὀρειχάλκου λάλα κύμβαλα καὶ μυρόεντα βόστρυχον, ἐκ λύσσας ἄρτ’ ἀναπαυσάμενος. The long-haired Gallus, gelded young, the one from Tmolus, the Lydian dancer whose shrieks carry far, has dedicated to the noble Mother by the Sangarius these timbrels and scourge with many knuckle-bones and these chattering copper cymbals and a perfumed lock, having just now entered rest after frenzy.

Since Erucius’ origin is Cyzicus it might seem probable that his choice of dialect related to the city’s claims to have been originally an ἀποικία of Corinth, though the city was later refounded by Miletus. Epigraphically preserved epigrams from Cyzicus do indeed use some Doric, though neither extensively nor consistently, and its proportion diminishes slightly between the Hellenistic and imperial periods. But even what seems to be our earliest case, from the third or second or first century BC, an iambic trimeter epitaph for Menecrates, has only three or four Doric forms in eight lines other than the personal name of his father, Matrodorus.⁸ From the second or first centuries BC an elegiac poem for Maean-

 Merkelbach-Stauber , // = GVI  = IK .: [τᾶς] ᾿Aφροδίτας ναός (), ἐτερπόμαν (), Ματροδώρου (): ναός, regular in Attic tragedy, might well not be perceived as Doric. For the special case of personal names see Sens ,  with n..

10

Ewen Bowie

dria surprisingly has twelve Doric forms in ten lines,⁹ whereas one for Menander has only two in six lines.¹⁰ A first century AD epitaph for Alexander from Alexandria has only two Doric forms in eight lines,¹¹ while the elaborate and ambitious acrostich poem for Apollonides, of the first or second century AD, two in its ten lines – the first (ἁ …) required by the acrostich, and the second (ποτὶ) perhaps following suit: the remaining nine and a half lines are consistently AtticIonic.¹² Likewise two imperial epigrams from Cyzicene territory for statues of poets eschew Doric – though of course that both statues are of epic poets may have been sufficient reason for their epigrammatists to maintain Attic-Ionic. That for Homer’s statue is short and simple, IMT Kyz Kapu Dağ 1425:¹³ ἡρώων κλέα πολλὰ καὶ Ἰλιακοῦ πολέμοιο κοσμήσας ὁ θεοῖς ἶσος Ὅμηρος ὅδε. The man who the heroes’ many deeds of renown and the war for Ilium arrayed, the man equal to the gods, Homer, is before you.

The other, much later, is for the poet Nestor of Laranda, IMT Kyz Kapu Dağ 1451:¹⁴ ἡ βουλὴ{ι} τείμεσσεν ἀγασσαμένη τὸν ἀοιδόν Νέστορα καὶ μολπῆς εἵνεκα καὶ βιοτοῦ· [ε]ἰκόνα ἐξετέλεσσεν καὶ εἵσατο πατρίδος ἄρχων Κορνοῦτος θαλερῆς ἐν τμένεσσ Κόρης· ὄφρα καὶ οψίγονί περ ἐν ἄστεϊ παῖδες ἔχοιεν σῆμα φιλοξενίας καὶ δέλεαρ σοφίης. The Council honoured in admiration the bard Nestor, both for his singing and his lifestyle. The image was made and set up by the fine city’s archon Cornutus in the precincts of Korē, so that children in later generations too might have a marker of friendly action and a lure to wisdom.

 Merkelbach-Stauber , // (assigning to the late Hellenistic period) = GVI  (assigning to the second or first century BC) = IK .: Φερσεφόνα (), θνατοὺς ἁλικίαν θεμένα (), τᾷ (), ἀλλοδαπὰν (), δεικνυμένα (), γενόμαν … λειπομένα (), Φερσεφόνας (), Λάθας λουσαμένα ().  Merkelbach-Stauber , // = GVI  = IK . τὰν (), ἀνδρολέταν ().  Merkelbach-Stauber , // = GVI  = IK .: δμαθεὶς (), Λάθας ()  Merkelbach-Stauber , // = GVI  = IK .: ἁ … ποτὶ ().  Merkelbach-Stauber , //.  CIG  = Merkelbach-Stauber , //. On Nestor of Laranda see Ma .

Doing Doric

11

If there is a pattern to be discerned at all, it is one which reveals a decline in Doric by the late first century BC.¹⁵ Erucius’ choice, then, was not determined by his Cyzicene origin. But similar choices to those of Erucius of Cyzicus may also be seen in the epigrams of Antiphilus of Byzantium,¹⁶ who seems to eschew Doric for paradoxography, but uses it in his Leonidean xv Gow-Page (= AP 6.95), a dedication by the rustic Parmis, a ploughman (γατόμος, line 6), whereas in his other Leonidean dedications by tradesmen he does not:¹⁷ Βουστρόφον, ἀκροσίδαρον, ἀπειλητῆρα μύωπα καὶ πήραν μέτρου σιτοδόκον σπορίμου γαμψόν τε δρέπανον, σταχυητόμον ὅπλον ἀρούρης, καὶ παλιουροφόρον, χεῖρα θέρευς, τρίνακα καὶ τρητοὺς ποδεῶνας ὁ γατόμος ἄνθετο Δηοῖ, Πάρμις, ἀνιηρῶν παυσάμενος καμάτων. His ox-turning, iron-tipped, threatening goad, and wallet that held corn in the measure to be sown, and curved sickle, ear-cutting tool of the field, and fork that casts back to the wind, the harvest’s hand, and pierced funnels – the ploughman dedicated to Dēo, Parmis, when he retired from his painful toils.

This reworks a Leonidas poem which is also a dedication by a rustic Parmis, Leonidas lxvi Gow-Page (= AP 7.504), where the Palatinus has Doric ὀλόαν and παλλομένα but Gow-Page print Ionic. Antiphilus also uses Doric for the several poems with arguably Doric themes. One of these is Antiphilus vi Gow-Page (= AP 9.178), immortalising the moment when Nero give freedom to the Dorian island of Rhodes: Ὡς πάρος ᾿Aελίου, νῦν Καίσαρος ἁ Ῥόδος εἰμὶ νᾶσος, ἴσον δ’ αὐχῶ φέγγος ἀπ’ ἀμφοτέρων·

 I set aside the sporadically Doric poems describing to mythological scenes represented on columns of the early second century BC temple of Apollonis, mother of Attalus and Eumenes of Pergamum: it seems clear that the  poems, constituting Book Three of the Palatine Anthology, are late, perhaps as late as Nonnus, cf. Demoen , Cameron , .  Gow-Page , . rightly see his geographical references as confirming the ethnic ‘Byzantine’ attached to  poems, though they miss the mise-en-scène of Elaeous at xxiii (= AP .). See also Robert .  On these other dedications see Ypsilanti . For an interesting proposal that Antiphilus xi GP (= AP .) was a programmatic first poem in an Antiphilan collection and xvi (= AP .) an epilogue see Höschele ,  – .

12

Ewen Bowie

ἤδη σβεννυμέναν με νέα κατεφώτισεν ἀκτίς, Ἅλιε, καὶ παρὰ σὸν φέγγος ἔλαμψε Νέρων. πῶς εἴπω, τίνι μᾶλλον ὀφείλομαι; ὃς μὲν ἔδειξεν ἐξ ἁλός, ὃς δ’ ἤδη ῥύσατο δυομέναν; Just as before I was the island of the Sun, so now I, Rhodes, am Caesar’s, and I boast an equal light from both: when I was already being extinguished a new beam lit me up, Sun, and alongside your light shone Nero. How can I put it? To which of you do I owe more? To the one who displayed me from the sea, or the other who saved me when I was already sinking?

Another is Antiphilus xiii Gow-Page (= AP 5.307), on a painting of Leda and the swan, where the choice of dialect may be a response to the very explicitly Spartan setting: Χεῦμα μὲν Εὐρώταο Λακωνικόν, ἁ δ’ ἀκάλυπτος Λήδα χὠ κύκνῳ κρυπτόμενος Κρονίδας. οἱ δέ με τὸν δυσέρωτα καταίθετε. καὶ τί γένωμαι; ὄρνεον. εἰ γὰρ Ζεὺς κύκνος, ἐγὼ κόρυδος. The stream is the Eurotas, in Laconia, the unclothed woman Leda, the male concealed as a swan, Zeus. You inflame my fragile passion – just what am I to become? A bird. If Zeus is a swan, then I am a lark.¹⁸

The Trojan horse of Antiphilus xxxv Gow-Page (= AP 9.156) is a more precarious case. The Palatinus reads ὅλα in line 4 and μάταν in line 5, both of which GowPage print; the Planudean reads ὅλη and μάτην. Homer had Menelaus describe the Trojan horse filled with the best of the Argives (Od. 4.271– 3), which may be enough to explain the Doricisms.¹⁹ Δέρκεο τὸν Τροίας δεκέτη λόχον, εἴσιδε πῶλον εὐόπλου Δαναῶν ἔγκυον ἡσυχίης. τεκταίνει μὲν Ἐπειός, ᾿Aθηναίη δὲ κελεύει ἔργον, ὑπὲκ νώτου δ’ Ἑλλὰς ὅλα δύεται.

 Gow-Page see this as an adaptation of the ‘proverbial phrase’ comparing the singing of larks and swans, AP .. and Dioscorides xxxvi  –  Gow-Page (= AP .. – ). Konstan ,  –  suggests that Antiphilus’ point is that like Zeus he will be a singing bird – not indeed a swan, but ‘at all events’ a lark. This still lacks force, and I suggest there may also be a pun – with itacism κόρυδος can sound like κόρη δός, ‘Give, girl’.  Or perhaps Antiphilus has been reading the Iliou Persis of Stesichorus, a well-known poem, or that of the Argive poet Sacadas, still known to Athenaeus .c some  years after Antiphilus, cf. Bowie a.

Doing Doric

13

ἦ ῥα μάταν ἀπόλοντο τόσος στρατός, εἰ πρὸς Ἄρηα ἦν δόλος ᾿Aτρείδαις ἐσθλότερος πολέμου. Behold the tenth-year ambush of Troy, behold the filly pregnant with Greeks’ well-armed silence. The carpenter is Epeius, and Athena commands the work, and up into its back all Greece enters. In vain did so large an army perish, if for Ares the Atreidae found guile better than battle.

The Doric of Antiphilus xxxviii Gow-Page (= AP 9.294) on the corpse of Leonidas is much more pervasive, though at line 6 the Palatine has Ionic ἐλευθερίης where the Planudean has ἐλευθερίας: ‘Πορφυρέαν τοι τάνδε, Λεωνίδα, ὤπασε χλαῖναν Ξέρξης ταρβήσας ἔργα τεᾶς ἀρετᾶς’. ‘οὐ δέχομαι· προδόταις αὕτα χάρις· ἀσπὶς ἔχοι με καὶ νέκυν· ὁ πλοῦτος δ’ οὐκ ἐμὸν ἐντάφιον.’ ‘ἀλλ’ ἔθανες. τί τοσόνδε καὶ ἐν νεκύεσσιν ἀπεχθὴς Πέρσαις;’‘οὐ θνᾴσκει ζᾶλος ἐλευθερίας.’ ‘This purple cloak, Leonidas, is given you by Xerxes, in fear of your valour’s deeds’. ‘I do not take it. This is a favour to traitors. May a shield cover me even as a corpse. Wealth is not my shroud.’ ‘But you have died. Why, when dead, so hostile to Persians?’ ‘Love of freedom never dies.’

Antiphilus xlviii Gow-Page (= APl (A) 136) has six instances of Doric in eight lines in describing Timomachus’ Medea, perhaps gesturing to the myth’s location in Corinth:²⁰ Τὰν ὀλοὰν Μήδειαν ὅτ’ ἔγραφε Τιμομάχου χεὶρ ζάλῳ καὶ τέκνοις ἀντιμεθελκομέναν, μυρίον ἄρατο μόχθον, ἵν’ ἤθεα δισσὰ χαράξῃ, ὧν τὸ μὲν εἰς ὀργὰν νεῦε, τὸ δ’ εἰς ἔλεον. ἄμφω δ’ ἐπλήρωσεν· ὅρα τύπον· ἐν γὰρ ἀπειλᾷ δάκρυον, ἐν δ’ ἐλέῳ θυμὸς ἀναστρέφεται. ἀρκεῖ δ’ ἁ μέλλησις, ἔφα σοφός· αἷμα δὲ τέκνων ἔπρεπε Μηδείῃ, κοὐ χερὶ Τιμομάχου. When Timomachus’ hand painted murderous Medea by her jealousy and her children pulled to and fro,

 For a discussion of this and the three other epigrams on Timomachus’ Medea see Gurd .

14

Ewen Bowie

he took on a huge task, to sculpt the two sides of her character, of which one inclined to anger, the other to pity, and achieved both: see the image – for in her threat a tear lurks, and in her pity, anger. ‘Intent suffices’, said the sage: the children’s blood fitted Medea, not Timomachus’ hand.

Similarly Antiphilus xlix Gow-Page (= APl (A) 147) weaves some Doric into its description of a painting of Perseus and Andromeda, presumably because Perseus is a hero of Argive descent: Αἰθιόπων ἁ βῶλος· ὁ δὲ πτερόεις τὰ πέδιλα Περσεύς· ἁ δὲ λίθῳ πρόσδετος ᾿Aνδρομέδα· ἁ προτομὰ Γοργοῦς λιθοδερκέος· ἆθλον ἔρωτος κῆτος· Κασσιόπας ἁ λάλος εὐτεκνία. χἀ μὲν ἀπὸ σκοπέλοιο χαλᾷ πόδας ἠθάδι νάρκᾳ νωθρόν· ὁ δὲ μναστὴρ νυμφοκομεῖ τὸ γέρας. The soil is Ethiopian; the man with winged sandals Perseus; the woman tied to the rock Andromeda; the head, a petrifying Gorgon’s; love’s ordeal a sea-monster; Cassiopeia’s, the big-mouthed child-pride. She loosens her legs from the crag, numb with now-familiar torpor; her suitor scoops his bridal prize.

But this explanation does not work for Antiphilus xxxvi Gow-Page (= AP 9.192) on the Iliad and Odyssey, unless the prominence of ‘Argives’ in the Iliad is relevant: Αἱ βίβλοι, τίνες ἐστέ; τί κεύθετε; ‘Θυγατέρες μὲν Μαιονίδου, μύθων δ’ ἵστορες Ἰλιακῶν· ἁ μία μὲν μηνιθμὸν ᾿Aχιλλέος ἔργα τε χειρὸς Ἑκτορέας δεκέτους τ’ ἆθλα λέγει πολέμου· ἁ δ’ ἑτέρα μόχθον τὸν Ὀδυσσέος ἀμφί τε λέκτροις χηρείοις ἀγαθᾶς δάκρυα Πηνελόπας’. Ἵλατε σὺν Μούσαισι· μεθ’ ὑμετέρας γὰρ ἀοιδὰς εἶπεν ἔχειν αἰὼν ἕνδεκα Πιερίδας. ‘Books, who are you? What have you within?’ ‘Homer’s daughters, tellers of the tales of Troy. One tells of Achilles’ wrath and the deeds of Hector’s hand, and ordeals of a ten-year war, the other Odysseus’ toil and the tears of noble Penelope over her widowed bed’. ‘Be gracious, with the Muses: for after your songs Time has said it has eleven Pierians’.

Doing Doric

15

I find Antiphilus’ choice of Doric here as puzzling as that of Argentarius for his only Doric epigram out of his 37 surviving poems, i. e. xxxv Gow-Page (= AP 9.221), a carving of Eros on a sphragis: Αὐγάζω τὸν ἄφυκτον ἐπὶ σφραγῖδος Ἔρωτα χερσὶ λεοντείαν ἁνιοχεῦντα βίαν, ὡς τᾷ μὲν μάστιγα κατ’ αὐχένος, ᾇ δὲ χαλινοὺς εὐθύνει· πολλὰ δ’ ἀμφιτέθηλε χάρις. φρίσσω τὸν βροτολοιγόν· ὁ γὰρ καὶ θῆρα δαμάζων ἄγριον οὐδ’ ὀλίγον φείσεται ἁμερίων. I behold on the seal inescapable Eros reining in a mighty lion with his hands, with one guiding the whip to its neck, with the other the bit: much grace blooms all around. I shudder at the man-destroyer: for he who tames even a wild beast will not spare mortals at all.

As suggested by Kathryn Gutzwiller in the discussion at Thessaloniki, the best explanation for this unusual move by Argentarius is that the recipient of the poem, perhaps also recipient or existing owner of the seal, had a Dorian connection. In the case of Antiphilus we might perhaps expect Doric to be his dialect of choice, since Byzantium was, and sometimes remembered that it was, a Megarian colony. But in fact Byzantium’s inscribed epigrams are resolutely Attic-Ionic, and by the imperial period we find Doric only in occasional personal names like Asclapiadas, or in the names of offices like that of the hieromnamon.²¹ A nice example of an inscribed epigram of the early first century AD (‘or shortly after’) is the following poem with some literary aspirations but no scent of Doric, presumably composed for the sculpture of a dolphin on a nymphaeum:²² [τόν με κυ]βιστητῆρα τὸν ἐξ ἁλὸς οὐκέτι Νηρεὺς [ποιμαίνει]· χέρσονδ’ ὧδε μετῳκισάμην· [ἤλλαγμ]αι Νύμφας Νηρηΐσι καὶ πεπέδημαι [νάματος ἀ]λλοτρίου τερπνοτέραις σταγόσιν· [χαίροιτ’ εἰ]ς κόλποιο μυχοὺς εὐίχθυες ἄγραι· [………]ς χέρσῳ πόντον ἀναινόμεθα. A diver from the sea, I no longer have Nereus [as my shepherd]: I have moved house to here on land. [I have exchang]ed the Nereids for Nymphs, and am tied to the pleasanter droplets of another’s [stream].

 IK .  IK ..

16

Ewen Bowie

[Farewell] hunts rich in fish into the gulf’s depths: [content with?] dry land we reject the sea.

Sacerdos of Nicaea Against this background of poets composing during the century or so before the reign of Nero I turn to five splendid poems from Nicaea. We might approach these epigrams of Nicaea with no special expectation that they would be Doric. Indeed there is only one other poem which uses some Doric in the currently known epigraphy of Nicaea, an epitaph for a weaver called Myrmex:²³ […] σου χθὼν ἥδε […………]κεύθει [Μ]ύ̣ρμηξ, ἱστεῶνος πρόστατα καλλιπέπλου· [σᾶι τ]έχναι γὰρ ἔφυς [……] ἔξοχος ἄλλων [κ]αὶ δόκιμος βιοτᾷ καὶ φρενὶ πιστότατος· [οὔ] σε μάταν μύρμηκος ἐπώνυμον ἔφρασαν [ἄ]ν[δρες]. […]this earth[……………]holds Ant, guardian of the fair-clothed loom For in your skill you were [ ] above others and approved in your lifestyle and most trusted for your character; not for nothing did men call you by the name ‘Ant’.

More telling, perhaps, is the fact that of the thirty one surviving epigrams of Apollonides, who seems to be from Nicaea, not one uses Doric, even though xxi Gow-Page (= AP 9.280) has a Spartan setting and xxviii Gow-Page (= APl (B) 49) is set in Rhodes. One might add that Doric is not used by Diodorus x Gow-Page (= AP 7.701) in his poem for Diomedes of Nicaea; but since Diodorus seems never to have used Doric anyway not much weight can be placed on this detail. Overall, however, Nicaea seems very rarely to exhibit Doric in its ‘literary’ or in its other inscribed epigrams. It is therefore surprising that Doric is on show in no less than three of the five poems on the grandiose funeral monument of a Nicaean who is there referred to only by his ὄνομα, in Roman terms his cognomen, Sacerdos. Its five elegiac epigrams are transmitted by Book 15 of the Palatine Anthology, where the poems are numbers 4 to 8.²⁴ A note by J tells us that AP 15.4 was a sepulchral epigram copied ‘at Nicaea near the lake on the obelisk’ (ἐπιτύμβιον ἐν Νικαίᾳ πλησίον

 IK ...  Merkelbach-Stauber , //, , ,  and  (on pages  – ).

Doing Doric

17

τῆς λίμνης ἐν τῷ ὀβελίσκῳ). The writer of this note, suggested by Cameron to be Constantine of Rhodes and perhaps to be getting his information from Alexander, metropolitan of Nicaea in the tenth century,²⁵ evidently uses the term ‘the obelisk’ because this was the architectural form of Sacerdos’ tomb. It must have been similar to the obelisk of Cassius Philiscus, some five kilometres out of town away from the lake, and erected c. AD 120.²⁶ I have discussed these poems recently elsewhere,²⁷ but for convenience I also print both texts and translations here: Αὔχησον, Νίκαια, τὸν οὐρανομάκεα τύμβον καὶ τὰν ἀελίῳ γείτονα πυραμίδα, ἃ τὸν ἐνὶ ζῳοῖς βεβοαμένον ἱεροφάνταν κρύπτει ἀμετρήτῳ σάματι θαπτόμενον. ἔστι Σακέρδωτος τόσον ἠρίον, ἔστι Σεουήρας μνᾶμα τόδ’, ᾧ γείτων οὐρανός, οὐκ ἀίδας. Boast, Nicaea, of the tomb as tall as heaven and the pyramid that is neighbour to the sun, which hides the hierophant renowned among mortals buried in its measureless monument. This is the great sepulchre of Sacerdos, of Severa is this memorial, to which heaven, not Hades, is neighbour. AP 15.4 Oὐράνιον τὸ μνᾶμα καὶ ἁ χρυσήλατος ἀκτὶς ἀνδρὸς ἴσον βιότῳ καὶ τάφον εὑραμένου ἄστροις γειτονέοντα· φέρει δ’ ὅσον οὔτινα τύμβος ἀνέρα, τὸν τελετᾶς οὐρανίδος ζάκορον, τὸν πάτραν ἐριποῦσαν ἀπὸ χθονὸς ὑψώσαντα, τὸν φρενὸς, ᾗ γλώσσας, ἄκρα λαχόντα γέρα· ᾧ πέρι δηρίσαντο καὶ ἁ νέκυν ἐν πυρὶ θεῖσα ᾿Aτθὶς χἀ κόλποις ὀστέα δεξαμένα. The heavenly memorial and the ray of beaten gold matches the life of a man who found even his burial neighbouring the stars: the tomb harbours a peerless man, sacristan of the heavenly mystery, who raised up his fallen country from the ground, who attained the highest honours for mind as for tongue, over whom the land fought that set his body on a pyre, Attica, and that which took his bones to its bosom. AP 15.5

 Cameron , .  IK .: Γ. Κάσσιος Φιλίσ|κος, Γ. Κασ|σίου ᾿Aσκληπιοδότου υἱός,| ζήσας ἔτη πγ. Cf. BekkerNielsen , ,  Bowie b,  – .

18

Ewen Bowie

Τοῦτο Σακέρδωτος μεγάλου μέγα σῆμα τέτυκται παμφαές, ᾿Aσκανίης ἄστρον ἐπιχθόνιον, ἀκτίνων ἀντωπόν· ὁ δ’ ἥσυχος ἔνδοθι δαίμων κεῖται ὁ καὶ πάτρῃ δεξιτερὴν τανύσας κεκλιμένῃ καὶ στέμμα περὶ κροτάφοισιν ἀνάψας ἱερὸν ἐκ πατρὸς παιδὶ νεαζόμενον, ὃν πάτρη μὲν ἔδεκτο φίλον νέκυν, ἥγνισε δ’ ᾿Aτθὶς πυρκαϊῇ, σέβεται δ’ Ἑλλὰς ἅπασα πόλις. This is the great monument of great Sacerdos, shining everywhere, the earthly star of Ascania, its face to the sun: but inside in peace the spirit rests, who stretched out his hand to his country in collapse, and tied a sacred fillet on the head of a son when it was rejuvenated by a father, whose dear body his country received, Attica sanctified on a pyre, and every Greek city reveres. AP 15.6 Ἁ πάτρα Νίκαια, πατὴρ δέ μοι ὀργιοφάντας οὐρανοῦ, αὐτὰρ ἐγὼ κλαρονόμος τελετᾶς· οὗτος ὁ καὶ σεισθεῖσαν ἐμὰν πόλιν ἐξ ἀίδαο ῥυσάμενος δώροις Αὐσονίοιο Διός· θνᾴσκω δ’ ᾿Aσκανίας μὲν ἀπόπροθεν ἠδ’ ἐπὶ γαίας ᾿Aτθίδος ἀρχεγόνου πυρκαϊᾶς ἐπέβαν. μνᾶμα δέ μοι περίσαμον ὁμώνυμος εὕρατο πάππῳ παῖς ἐμός· ἁ δ’ ᾿Aρετὰ λεύσσει ἐς ἀμφοτέρους. My country is Nicaea, my father displayer of the holy objects of heaven, and I the heir of his office. I am the man who also saved from Hades his city shaken by earthquakes, by the gifts of Ausonian Zeus. I died far from Ascania, and in the Attic land, whence my family came, I mounted the pyre. My famed memorial was devised by one who has his grandfather’s name, my son: excellence looks upon both of them. AP 15.7 Εἷς γάμος ἀμφοτέρων, ξυνὸς βίος, οὐδὲ θανόντων μνήμονες ἀλλήλων ἔσχον ἀποικεσίην· καὶ σεῦ μὲν τελεταί τε καὶ ἄρρενος ἔργα, Σακέρδως, κηρύξει βίοτον πάντας ἐς ἠελίους· αὐτὰρ ἐμὲ Σευουήραν ἀνήρ, τέκος, ἤθεα, κάλλος τῆς πρὶν Πηνελόπης θήσει ἀοιδοτέρην. One marriage united both, a shared life, and not even dead did they part, thinking of each other. And your offices and manly deeds, Sacerdos, will proclaim your life each day the sun rises;

Doing Doric

19

me, Severa, my husband, son, character, beauty, will make more hymned than Penelope of old. AP 15.8

One surprising features of these five poems commemorating Sacerdos is that neither his father nor the homonymous son who erected the obelisk is named. But the riddling phrase that closes AP 15.7 covertly reveals their names: μνᾶμα δέ μοι περίσαμον ὁμώνυμος εὕρατο πάππῳ παῖς ἐμός· ἁ δ’ ᾿Aρετὰ λεύσσει ἐς ἀμφοτέρους. My famed memorial was devised by one who has his grandfather’s name, my son: excellence looks upon both of them.

Not many Greek names mean ‘virtuous’ (ἀγαθός, for example, is used only as part of a compound personal name in Bithynia),²⁸ but the name Chrestus does (of course with the shift of accent from χρηστός to Χρῆστος that is regularly found when nouns or adjectives are used as personal names). The father of Sacerdos is therefore none other than a man well attested in Nicaean epigraphy, C. Cassius Chrestus, who died aged 58.²⁹ The Sacerdos of the tomb-obelisk thus had as his full Roman name C. Cassius Sacerdos: he will have taken on the position of ἀρχιερεὺς καὶ σεβαστοφάντης,̣ archiereus and sebastophant, on his father’s death, perhaps around AD 90, and seems himself to have died around AD 130. He joins Cassius Asclepiodotus, Cassius Philiscus and (a century later) the historian Cassius Dio Cocceianus as a distinguished member of one of the leading families in imperial Nicaea, the Cassii. We do not know what Sacerdos was doing in Attica when he died. Merkelbach suggested he was there for the ceremonies attending establishment of the Panhellenion in AD 131/2,³⁰ but there is no hint of that in the poetic texts. His presence in Attica may equally have related to his intellectual and rhetorical eminence, picked out in the second poem in the attributive phrase τὸν φρενὸς ᾗ γλώσσας ἄκρα λαχόντα γέρα, ‘who attained the highest honours for mind as for tongue’(AP 15.5.6)?³¹ Nothing helps answer these questions as yet. Unfortunately

 But it does appear as a name in other provinces, e. g. in Egypt, Baillet  – , no. .  C. Cassius Chrestus is described on his late first-century sarcophagus, found outside Nicaea’s East (Lefke) gate (for whose reconstruction between AD  and  he was responsible, IK .), as πρέσβυ[ς καὶ] (or πρεσβύ̣[τερος]) | ἀρχιερεὺς καὶ σεβαστοφάντης ̣ | ἐτῶν νηʹ, IK ..  Merkelbach , repeated Merkelbach and Stauber , .  In this line the Palatinus reads ἢ, ‘or’, giving the feeble sense ‘who attained the highest honours for mind or for tongue’. Merkelbach-Stauber ,  translate ἢ ‘und’. I print and trans-

20

Ewen Bowie

the name of Sacerdos’ wife, Severa, is too common to be diagnostic either for her place of origin or her status in society, but certainly admits a relationship with some of the top families in western Asia Minor. One mark of Sacerdos’ paideia may be the Doric dialect that was chosen for three of the five poems. Like Byzantium, Astacus, on the gulf of Nicomedia and a little west of Nicaea, claimed foundation by Megara,³² and the poet may be gesturing towards the Dorian origins of parts of his and Sacerdos’ region. The identity of that poet cannot be resolved – indeed it cannot be assumed that the same epigrammatist was responsible for all five poems. Sacerdos was grand enough for more than one poet to have been approached, and the practice of inscribing poems by more than one poet on the same monument is a very old one.³³ That said, it is possible that the poet responsible for the poems that use Doric might be the sophist Memmius Marcus, a member of the governing class of Byzantium and a man whose floruit is Hadrianic.³⁴ Among the anecdotes reported by Philostratus is one concerning his visit to Megara and his successful intervention to quell Megarian hostility to Athenians, allegedly stemming from Pericles’ Megarian decree of the fifth century BC.³⁵ Memmius Marcus was clearly conscious of Byzantium’s foundation by Megara – indeed he traced his family’s origin to the founder Byzas.³⁶ As one might expect, and as the poets collected by Philip of Thessalonice in his Garland exemplify, the intellectual links between the cities of the northern Aegean, the Propontis and north-western Asia Minor (i. e. Bithynia and Mysia) were sometimes closer than their links with Attica or with the great cities of provincia Asia.³⁷ Marcus might well have had connection with the elite of Nicaea. The other sophist from Byzantium who makes it into Philolate ᾗ, ‘just as’, which will of course have been indistinguishable from ἢ in the epigraphic text copied at Nicaea.  Bekker-Nielsen ,  n.. Note also the Doric form γυνά in IK . Ὀλυμπιὰς Διονυσίου γυνὰ ἐτῶν σεʹ χαῖρε of which Robert Bull.  no.  wrote that the woman must have come from the southern shore of the gulf of Nicomedia, around Pylae, which was Byzantine territory.  Cf. the story about a competition between Aeschylus and Simonides for a ἐλεγεῖον to commemorate the battle of Marathon, with Bowie ,  – ; more generally on epigram competitions Petrovic ,  – .  ἠγάσθη αὐτὸν καὶ ᾿Aδριανὸς ὁ αὐτοκράτωρ πρεσβεύοντα ὑπὲρ Βυζαντίων, ἐπιτηδειότατος τῶν πάλαι βασιλέων γενόμενος ἀρετὰς αὐξῆσαι, Philostr. VS .. – . For a good discussion distinguishing the (Memmius) Marcus noted as magistrate on coins representing Pius and then later as ἥρως (i. e. after his death) on coins representing Annia Lucilla (i.e. pre-) from the Memmius Marcus, his son, on coins of Marcus and Commodus, see Puech ,  – .  Philostr. VS ...  Philostr. VS ...  On the economic basis for this network see Gren .

Doing Doric

21

stratus’ selective series of biographies is called Chrestus.³⁸ This Chrestus lived at least a generation after our grandee commemorated at Nicaea, Sacerdos, son and father of a Chrestus. But his name raises the possibility that there was some connection between the families located in Byzantium and Nicaea respectively. It will not only be in the fictional world of Achilles Tatius that a one of the Byzantine elite married somebody of their own class from another city.

Abbreviations AP Anthologia Palatina. CIG Boeckh, A. (ed.) (1828 – 1877), Corpus Inscriptionum Graecarum, Berlin. GP Gow, A./Page, D. (eds.) (1968), The Greek Anthology: The Garland of Philip and Some Contemporary Epigrams, 2 vols., Cambridge. GVI Peek, W. (ed.), (1955), Griechische Vers‒Inschriften, Berlin. IK Inschriften griechischer Städte aus Kleinasien. 1972 ff., Bonn.

Bibliography Baillet, J. (1920 – 1926), Inscriptions grecques et latines des tombeaux des rois ou Syringes à Thèbes, Cairo. Bekker-Nielsen, T. (2008), Urban life and local politics in Roman Bithynia, Aarhus / Oakville, CT. Bowie, E. L. (2010), ‘Marathon in fifth-century epigram’, in: K. Buraselis / K. Meidani (eds.), Marathon: The ancient deme and the Battle, Athens, 203 – 219. Bowie, E. L. (2011), ‘Men from Mytilene’, in: T. Schmitz / N. Wiater (eds.), The struggle for identity. Greeks and their past in the First Century BCE, Stuttgart, 49 – 63. Bowie, E.L. (2008 ([2012]), ‘Luxury cruisers? Philip’s epigrammatists between Greece and Rome’, in: Aevum Antiquum 8, 223 – 258. Bowie, E.L. (2014a), ‘Rediscovering Sacadas’, in: A. Moreno / R. Thomas (eds.), Patterns of the Past: Epitedeumata in the Greek Tradition, Oxford, 39 – 55. Bowie, E.L. (2014b), ‘Greek culture in Arrian’s Bithynia’, in: S. Lalanne / A. Hostein (eds.), Le monde d’Arrien de Nicomédie, in: Ktèma 39, 37 – 49. Clausen, W.V. (1994), Virgil Eclogues. Oxford. Cameron, A. (1993), The Greek Anthology from Meleager to Planudes, Oxford. Demoen, K. (1988), ‘The date of the Cyzicene epigrams. An analysis of the vocabulary and metrical technique of AP, III’, in: L’Antiquité classique 57, 231 – 248. Giangrande, G. (1975), ‘Fifteen Hellenistic epigrams’, in: JHS 95, 31 – 44

 Philostr. VS .. – . That Chrestus had been a pupil of Herodes, and that the Athenians wanted the emperor to appoint him as successor to Hadrianus of Tyre in the imperial chair at Athens, suggests a floruit in the s and early s.

22

Ewen Bowie

Gren, E. (1941), Kleinasien und Ostbalkan in der wirtschaftlichen Entwicklung der römischen Kaiserzeit, Uppsala. Gurd, S.A. (2007), ‘Meaning and material presence: four epigrams on Timomachus’s unfinished Medea’, in: TAPhA 137, 305 – 331 Höschele, R. (2007), ‘The travelling reader: journeys through ancient epigram books’, TAPhA 137, 333 – 369. Jones, C.P. (2004), ‘Epigraphica VIII-IX’, in: ZPE 146, 93 – 98. Konstan, D. (2008), ‘Antiphilus’ Erotic Epigrams: Two Notes’, in: Mnemosyne 61, 290 – 7. Ma, J. (2007), ‘The worlds of Nestor the poet’, in: S. Swain / S. J. Harrison / J. Elsner (eds.), Severan Culture, Cambridge, 83 – 113. Magnelli, E. (2007), ‘Meter and diction: from refinement to mannerism’, in: P. Bing / J.S. Bruss (eds.), Brill’s Companion to Hellenistic Epigram, Leiden / Boston, 165 – 183. Merkelbach, R. (1987), Nikaia in der römischen Kaiserzeit, Opladen. Merkelbach, R. / Stauber, R. (2001), Steinepigramme aus dem griechischen Osten, vol. 2, Bonn. Petrovic, A. (2009), ‘Epigrammatic contests, poeti vaganti and local history’, in: R.L. Hunter / I.C. Rutherford (eds.), Wandering Poets, Cambridge, 195 – 216. Puech, B. (2002), Orateurs et Sophistes Grecs dans les Inscriptions d’Époque Impériale. Avec préface de L. Pernot. Paris. Robert, L. (1979), ‘Un voyage d’Antiphilos de Byzance, Anthologie palatine XI,17, Géographie antique et byzantine’, in: JS, 257 – 294. Sens, A. (2004), ‘Doricisms in the New and Old Posidippus’, in: B. Acosta-Hughes / E. Kosmetatou/M. Baumbach (eds.), Labored in Papyrus Leaves. Perspectives on an epigram collection attributed to Posidippus (P. Mil. Vogl. VIII 309), Center for Hellenic Studies, Washington DC, 65 – 83. Ypsilanti, M. (2006), ‘An aspect of Leonidas’ reception in later epigrammatists and the art of variation’, in: CPh 101, 67 – 73.

Dee L. Clayman

Callimachus’ Doric Graces: 15 GP = 51 Pf.* τέσσαρες αἱ Χάριτες, ποτὶ γὰρ μία ταῖς τρισὶ κείναις ἄρτι ποτεπλάσθη κἤτι μύροισι νοτεῖ. εὐαίων ἐν πᾶσιν ἀρίζηλος Βερενίκα, ἇς ἄτερ οὐδ’ αὐταὶ ταὶ Χάριτες Χάριτες. 1 κείναις AP : τήναις Wilamowitz 3 ἀρίζηλος AP : ἀρίζαλος Brunck, Wilamowitz Four are the Graces; in addition to those three, one has just been molded and is still damp with myrrh. Fortunate Berenike, conspicuous among all, without whom the Graces themselves are not Graces. 15 GP = 51 Pf. = AP 5.146

Epigram 15 GP is one of a handful of Callimachus’ epigrams that contain more than a few Doric word forms.¹ This is literary Doric, not a language that anyone actually spoke, but a selection of Doric forms inherited from earlier literature mixed with others including Homeric, poetic, and koine.² Below I argue that Callimachus’ selection of Doric words and their precise deployment in the epigram constitute a rhetorical strategy designed to compliment his dedicatee, Berenike II, and to implicate her in his stylistic agenda. The texts of Callimachus’ Doric poems are slippery objects of study because ancient editors and copyists had a tendency to replace unfamiliar spellings with standard ones, and modern editors tend to do the opposite; they correct standard spellings with Doric ones.³ Two examples of this are below, which I will argue are both ill-conceived. But editing is not the whole story. Dialect instability is also found in inscribed epigrams in the third century BC, as koine began to assert itself against dialects generally, and Alexander Sens has shown that it characterizes the text of the “new” Posidippus, which was produced within a generation

* Grateful thanks are due to the organizers of “th Trends in Classics: International Conference on Greek Epigram,” held in Thessaloniki, May  – , , where this paper was first presented, and to my fellow participants who offered such useful comments, especially Franco Montanari, Ivana Petrovic, Benjamin Acosta-Hughes and Peter Bing.  Callimachus’ other epigrams with Doric features include  GP = AP .;  GP = AP .;  GP = AP .;  GP = AP .;  GP = AP . and  GP = AP .. Others are possibly Doric.  On the Doricisms of Callimachus’ Hymn to Demeter where the poet uses a similar melange of dialect forms see Hopkinson ,  – ; on the literary Doric of bucolic poetry, Molinos Tejada ; and on Theocritus’ literary Doric, Hunter ,  – .  On the treatment of dialect forms by ancient editors, Colvin ,  – .

24

Dee L. Clayman

of the poet’s death in the 3rd c. BC.⁴ This, he argues, implies that Posidippus himself was the source of the dialectic mix, and that dialect itself could be an element in his poetic toolbox. Callimachus’ famous pronouncement on dialect mixing in Iamb 13 seems to corroborate this. τοῦτ’ ἐμπ[έ]πλεκται καὶ λαλευσ[…].[ Ἰαστὶ καὶ Δωριστὶ καὶ τὸ σύμμεικ[τον. This is interwoven and chattering/ in Ionic and Dorian, and the mixed type. Call. fr. 203.17– 18 Pf.

Here Callimachus reserves the right to use any dialect forms that he likes consistently or inconsistently. Why Callimachus includes so many Doric forms in 15 GP is not a mystery. The subject of his poem, named in line 3, is Berenike. This is generally agreed to be Berenike II, the wife of Ptolemy III and Callimachus’ great patron.⁵ It is significant that the poet uses the Doric form of her name here, Βερενίκα. Both Callimachus and Berenike were from Cyrene and native Doric speakers. He is not speaking Cyrenean here,⁶ and we do not know whether he ever had opportunities in real life to speak with her in the accents of their native city, but it is selfevident that he exploited their common homeland to claim a special relationship with her in the endless game of one-upsmanship that characterized life among the royal courtiers in Alexandria.⁷ In all likelihood she was among the intended readers of this poem, so this is a subtle or not so subtle way to remind her of their shared linguistic heritage. Beyond her Cyrenean birth and childhood, Berenike was Doric by virtue of her family’s Macedonian background. Her father Magas was a son of the first Berenike by her first husband, the otherwise unknown Philip.⁸ And we know from Posidippus (AB 88.4) that Berenike I came from Macedonian Eordea like her second husband, Ptolemy I, so the social and political values that became attached to Macedonian Doric speech by Callimachus’ time are operative here.⁹ A key component of the Ptolemies’ claim to be Dorians is the myth of Macedonian ties to Doric speaking Argos in the Peloponnese, which Callimachus highlights and associates with Berenike II in the begin Sens . Horrocks ,  notes the tenacity of Dorian spelling and Dorian identity in the face of the spread of Attic-based koine.  Wilamowitz , vol. , , nt. ; Pfeiffer , ; Gow-Page , vol. , . On Berenike II as a patron of Callimachus, Clayman ,  –  and passim.  On Doric dialect as Cyrenean, Ruijgh , and for a critique of this claim, Abbenes .  On Alexandrian court life, Weber  and .  On Berenike’s parents, Clayman ,  – .  On the politics of Macedonian, Hunter .

Callimachus’ Doric Graces: 15 GP = 51 Pf.

25

ning of the ‘Victoria Berenikes’ (54 – 60j Harder = 143 – 156 Massimilla) which also features Heracles, that quintessentially Doric hero.¹⁰ In short, Berenike II was three times Doric. Though Callimachus had personal and political motives for spelling her name in the Doric manner he was not in the habit of doing so. In the three other places where he names his queen in his extant poetry, she is Berenike, not Berenika.¹¹ This form, as well as its accompanying adjective, ἀρίζηλος is a literary reference, a quote from Theocritus Id. 17.57, the ‘Encomium to Ptolemy’,¹² ᾿Aργεία κυάνοφρυ, σὺ λαοφόνον Διομήδεα μισγομένα Τυδῆι τέκες, Καλυδωνίῳ ἀνδρί, ἀλλὰ Θέτις βαθύκολπος ἀκοντιστὰν ᾿Aχιλῆα Αἰακίδᾳ Πηλῆι· σὲ δ’, αἰχμητὰ Πτολεμαῖε, αἰχμητᾷ Πτολεμαίῳ ἀρίζηλος Βερενίκα. Argive woman with dark brows, you bore man-slaying Diomedes in union with Tydeus, the Calydonian hero; and deep-bosomed Thetis [bore] the warrior Achilleus to Peleus son of Aeacus; and you, spearman Ptolemy, conspicuous Berenika [bore] to spearman Ptolemy. Theoc. Id. 17.53 – 57

Theocritus’ poem describes the birth of Ptolemy II, whose mother was Berenike I, and celebrates their closeness to the gods. The concept is reinforced by ἀρίζηλος, ‘very conspicuous’, which may be appropriately applied to both Berenikes. ἀρίζηλος has struck some editors as an undoric form, and since the Doric ζᾶλος is attested, Brunk and others corrected it to ἀρίζαλος, but this would be a mistake as Pfeiffer knew.¹³ Not only do the best manuscripts of Theocritus have ἀρίζηλος, but so does Pindar (Ol. 2.55) and the spelling is supported by evidence from inscriptions (IG 9.1.270).¹⁴ We can be confident it is correct and that the whole phrase is a literary reference intended to confer some of the weight of Theocritus’ argument for the Ptolemies’ divine, heroic, and Doric qualities on Callimachus’ own slight epigram. Just as Theocritus tells us that Ptolemy II resembled  On Heracles as a Ptolemaic ancestor, Griffiths ,  – . Berenike’s husband, Ptolemy III claims descent from Heracles in the opening lines of the Adoulis decree (OGIS ).  Other passages where Callimachus names Berenike include frr. .; .; and . Pf.  Gow-Page , vol. ,  explain, following Kaibel, that Βερενι ́κα at the end of the verse violates a metrical rule observed by Callimachus in his elegies that a -syllable word at the end of a verse in the shape ˘ ˘ ¯ ¯ must be preceded by a monosyllable. There are a number of examples of verses that ignore this rule in Theocritus, but this is the only one in Callimachus, so it seems all but certain that Callimachus borrowed the words and placement from Theocritus. See also McLennan .  Pfeiffer  – , .  There is more on ἀρίζηλος below.

26

Dee L. Clayman

his father, Ptolemy I (in the manner of Achilles and Diomedes) so, Callimachus subtly suggests, Berenike II resembles the first Berenike. There is a point, then, to Berenike’s Doricness, but the epigram does not begin with it. It opens with the startling fact that there are now more Graces than there once were: Τέσσαρες αἱ Χάριτες, ‘Four are the Graces.’ Τέσσαρες is epic in form. The Doric would be τέτορες¹⁵ which would not scan, but that is not the only reason Callimachus chose Τέσσαρες. In this brief narrative the Graces start with their epic/Ionic identity, but by the end of line four, αἱ Χάριτες have become ταὶ Χάριτες.¹⁶ They become Doric Graces right before our eyes. Though ται ́ is also epic, it appears in Homer very rarely as an article followed by a noun, as it does here.¹⁷ To be precise, though, we could say that the epic/Ionic Graces become epic/Doric Graces. Not only have the Graces become Doricized, but they have multiplied. At the conclusion αἱ Χάριτες are ταὶ Χάριτες Χάριτες. The increase in number is the first message of the epigram and it is reinforced by Callimachus elegant word order. But how and when was the transition made? To observe the process let us look again at line 1. In addition to those three (ταῖς τρισὶ κει ́ναις), a new one has just now been created. Those three are the traditional three, the Graces from the Ionic island of Paros, who appear near the beginning of bk 1 of the Aitia (fr. 3 – 7b Harder).¹⁸ The manuscripts of the AP make their origen clear with κει ́ναις, an Ionic form equivalent to the Attic ἐκει ́ναις. This reading is not found in every text. Wilamowitz and Pfeiffer after him corrected κει ́ναις to the Doric τήναις. The emendation was made on the grounds that Callimachus uses τήνος consistently in his Doric hymns 5 and 6,¹⁹ but consistency is not a characteristic of Hellenistic poets using dialect forms. Gow-Page retains the manuscript reading and it is likely correct. Before the miraculous creation of a new Grace, the original three were firmly epic/Ionic. In line 1 the new Grace has not yet been identified. She is simply ‘one’, μι ́α. This form is dialectically ambiguous. It could be Attic, but also West Greek or epic. The first unambiguously Doric word is ποτι ́ which joins the One to the other Graces, while it retains the distinction between them.²⁰ This epic/Doric one has been added to ‘those Ionians.’

 Tribulato , .  On the articles ται ́ and τοι ́ as characteristic of West Greek dialects, Colvin ,  –  and Tribulato , .  The overwhelming majority of Homer’s ται ́’s and τοι ́’s are pronouns.  As noted in I. Petrovic & A. Petrovic , .  Wilamowitz ; Pfeiffer  – .  On ποτι ́ for Attic/Ionic πρός, Tribulato , .

Callimachus’ Doric Graces: 15 GP = 51 Pf.

27

The two linguistic groups come together in a process that has just been completed as line 2 begins, ἄρτι ποτεπλάσθη. Ποτεπλάσθη with its Doric prefix, reinforces ποτι ́ in line 1. It was ‘molded in addition to’, or ‘molded on’ not simply ‘added’ as LSJ suggests. This is an act of creation and extension, and the process that it describes is doricizing. The verb is not simply πλάσσω, or προσπλάσσω, but ποτιπλάσσω. It is a Doric fashioning that has just been completed as Callimachus begins his poem. At this point the moment of inception has already passed, but the new cré is in krasis Doric-style with και ́,²¹ and its poation is still damp. The adverb ἐτι sition following the diaeresis is emphatic. Though the process is complete, the dampness persists beyond expectation. νοτεῖ, ‘it is damp’, is not marked as Doric, but νοτέω is an interesting word. Its first appearance in Greek literature is right here, and the only other citations in LSJ come from Eratosthenes, Callimachus’ younger contemporary, and the still later Nicander. νοτεῖ, then, is apparently a Hellenistic formation and seems to have a scientific ring. In the Alexipharmaca it is used with ἱδρῶς, sweat (Al. 24), but also oil precisely as it is pressed out of olives (Al. 494). In Eratosthenes, it is dampness from melting ice (fr. 16.10 Powell). The older, more usual form of the verb, νοτι ́ζω, also refers to perspiration and other bodily fluids. In our epigram the liquid in question is myrrh, a perfume, and a commonly repeated idea is that perfume was poured or rubbed on a statue of the queen during its dedication ceremony. This interpretation was questioned by Gow-Page, who do not doubt that the poet is describing a statue, but note that there is no evidence that perfume was used in dedication ceremonies.²² The notion of pouring or annointing is absent from the poem itself which offers only νοτεῖ. The new creation is wet with a liquid like sweat, freshly pressed olive oil, or water from melting ice, i. e. liquid that originates within its source and then appears on the surface. No agent applies it. This suggests, in turn, that Callimachus is in the presence of that uniquely Hellenistic marvel: a sweating statue.²³ Sweating statues are omens, usually bad ones, and Posidippus includes one among his ‘Oinoskopika’: ́ ξέσματος ἱδρώσαντος ὁσος πόνος ἀνδρὶ πολίτηι καὶ δοράτων ὅσσος προσφέρεται νιφετός· ἀλλὰ τὸν ἱδρ[ώσα]ντα κάλει θεόν, ὅστις ἀπώσε[ι πῦρ ἐπὶ δυ[σμε]νέων αὔλια καὶ καλάμα[ς.

 The poet wrote κἤτι, not Ionic κἄτι.  Gow-Page , vol. , .  For references to sweating statues in antiquity, Bastianini / Gallazzi ,.

28

Dee L. Clayman

When a statue has perspired what trouble it is for the citizens, and how great a snow storm of spears approaches. But call on the sweating god, who will push the fire onto the buildings and crops of the enemy. Posidippus AB 30

Posidippus’ straightforward description of the phenomenon raises an alarm at the omen, but offers a solution to the threat: call upon the god who has perspired and he will drive away the enemy. There is a seamless transition from the perspiring statue to a perspiring god, who will avert the danger for a worshipper who is reverent. No prayer is evident in Callimachus’ epigram, but the suggestion that the statue is the same as its subject and that both have qualities of divinity is certainly present. Sweating statues were usually understood as negative omens like Posidippus’, but there was a notorious exception. Arrian (1.11.2) reports that shortly before Alexander set out on his expedition to Asia a statue of the poet Orpheus in Pieria sweated continuously. Though various explanations were offered, the seer Aristander of Telmissus soothed the king by explaining that the omen meant that epic and choral poets, as well as those who wrote odes, would perform much hard labor singing of Alexander and his works.²⁴ A perspiring Berenike would have the same effect on the poets of Alexandria. That Callimachus might present a sweating Berenike is not in the least inappropriate. Posidippus AB 36 marks the dedication of a headband to Arsinoe Philadelphus by a young girl who dreamed that the queen had used it to wipe the sweet sweat from her brow as she rested after heavy toils (3 – 4). The queen was holding a spear and a shield, so she had been toiling on the battlefield. In Posidippus’ text Arsinoe’s sweat is sweet (γλυκὺν), but Callimachus goes one step further, Berenike sweats myrrh. If this is an omen, it is certainly a good one. Though Berenike, too, was represented as a warrior,²⁵ her perfumed fragrance does not suggest toil so much as divinity. She is like Artemis whom Hippolytus perceives not by sight, but by the ‘divine breath of her fragrance’ (Eur. Hipp. 1391). In the same way Prometheus knows that the Oceanids are near by the sound of their winged chariot and their scent (Aes. Prom. 115 – 16). He wonders whether the fragrence is ‘divine, mortal or a mixture’, and Callimachus’ reader could ask the same question of Berenike’s. In this too, Alexander set

 Versions of the story are also in Plut. Alex. . and Ps.-Callisth. . where another seer connects the sweat to Alexander’s own future toils on the battlefield.  Berenike also was portrayed by Callimachus as active on the field of battle (Hy. Astr. .. – ).

Callimachus’ Doric Graces: 15 GP = 51 Pf.

29

the precedent. As Plutarch says on the authority of Aristoxenus, ‘From his skin a sweetness emenated and there was a fragrance around his mouth and all his flesh so that his clothing was full of it’ (Plut. Alex. 4.2). Berenike’s sweat is not simply perfume, but precisely myrrh. This was generally harvested by tapping the myrrh tree i. e. slashing its bark and letting its juices run. These would soon concretize, and the residue would be packaged and sold. But unlike ordinary myrrh, Pliny (NH 12.35) reports that at the height of summer, the tree spontaneously sweats (sudant) a liquid called stacte, which is the most precious of all myrrh.²⁶ Berenike’s own sweat is surely stacte, and this, in turn, associates her with Myrrha (or Smyrna), who was impregnated by her own father and gave birth to her son Adonis after changing into a myrrh tree. The story of Myrrha’s incestuous passion with its dramatic consequences was known by Panyasis in the 5th c. BC and was current in the Hellenistic age.²⁷ Ovid’s retelling of the tale highlights its eroticism, a quality inherent in the story itself that may explain why Callimachus’ epigram was classified by its earliest compilers with other erotic epigrams.²⁸ Myrrha’s son, Adonis, the beloved of Aphrodite, is one of those liminal figures who hovers between mortal and immortal. He plays a leading role in the constellation of myths used by the Ptolemies and their apologists to situate their dynasty culturally between the east and west and themselves, between mortal and immortal.²⁹ An important document in this program is Theocritus’ Id. 15 which features an elaborate description of an Adonia produced by Arsinoe II to celebrate her mother’s deification. At the center of the festivities is a statue of Adonis surrounded by flowers, fruits and costly offerings including perfumes from Syria. In the words of ‘the Argive woman’s daughter’, a singer who performs at the festival, ‘Cypris, daughter of Dione, you made the mortal Berenike immor-

 Theophr. (de Odor. .), in contrast, says that stacte is the oil that flows from the myrrh tree after it has been bruised and notes that others describe it as the product produced when myrrh is processed with oil of balanos. Discussion in Lucas .  In his rendition of Myrrha’s tale Apollodorus (..) cites Panyasis (th cen. BC) as a source, and his version of Myrrha’s story was current in the Hellenistic age: Schol. on Lycophron ; Ovid Met. . – .  The AP locates the poem in book , a collection of erotic epigrams probably reflecting the original arrangement of Meleager (I. Petrovic / A. Petrovic ,  – ). The Petrovics further identify three motifs in the epigram which make it similar to other erotic epigrams: Charis and the Charites, perfume, and statues ( – ). Given the Ptolemies’ penchant for incestuous marriages, the nature of Myrrha’s relationship with her father was not a narrative detail that needed to be suppressed or explained away (Griffith ,  n. ). Berenike, too, was a woman, like Myrrha, who chose her own husband (Clayman :  – ).  Griffith ,  – ; Reed .

30

Dee L. Clayman

tal, as men say, when you dripped (ἀποστάξασα) ambrosia into the woman’s breast. You who are much-hymned, you who have many temples, to gratify you Berenike’s daughter Arsinoe, as beautiful as Helen, cherishes Adonis with many lovely things’ (Id. 15.106 – 11). Here at a festival for Adonis Arsinoe II celebrates the deification of her mother, Berenike I, as the gift of Aphrodite. The metamorphosis is accomplished by the goddess’s dripping ambrosia into Berenike’s breast.³⁰ The unique benefits of divine fluids were well known to Callimachus. In his Hymn to Apollo (38 – 41) the god himself produces a magical oil that drips from his hair bringing healing to mortals, and the poet asks his Parian Graces in Aitia 7.11– 14 to wipe their annointed hands on his elegies in order to give them long life. Their hands are slippery with oil that flows, like Apollo’s, from their locks. With a unique reconfiguration of these same imaginative elements, our epigram implies that Berenike II herself generates a divine fluid in the guise of the myrrh tree that brought forth Adonis. Unlike her predecessor, she is self-annointing, like Apollo and the Graces, and therefore already on a path to immortality. That Callimachus presents this image in Dorian accents is also a reflection of Id. 15. There the description of the Adonia is embedded in a mime featuring two Alexandrian women visiting the festival who vigorously defend their right to speak Dorian when a man in the crowd complains about their speech (87– 95). Callimachus claims the same right in his epigram which puts another Berenike at its center. The third line of our epigram finally reveals the name of the newest Grace, which is accompanied by some of her attributes as if the poem were a very short hymn. εὐαίων, ‘fortunate’ Berenika, ‘happy in life’ and ἀρίζηλος ‘very conspicuous.’ Both also appear elsewhere describing other Ptolemies, e. g. εὐαίων is attached to Ptolemy IV in SH 979.2, an anonymous epigram in a school text, and Eratosthenes (35.13 Powell) applies it to Ptolemy III, Berenike’s husband. Likewise ἀρίζηλος ‘very conspicuous’ appears in an anonymous fragment (fr. 734 Pf.) attributed to Callimachus himself, attached to Ptolemy son of Lagus. It seems, then, that these adjectives were authorized epithets for Ptolemaic rulers, but they are not banal or irrelevant to this poem. In Iliad 22.28 – 29 and Pindar Ol. 2.55 ἀρίζηλος describes a star that recalls Berenike’s ‘Lock’, which became a constellation (fr. 110 – 110 f Harder); the ‘star of Berenike’ (fr. 387 Pf.); and generally, the Graces’ association with light (Pind. P. 9.89 – 90). Likewise εὐαίων  Ambrosia, conventionally called the “food” of the gods, was often represented as a liquid rather than a solid (Wright ). There are early precedents for rendering a mortal immortal by the application of ambrosia in the Homeric Hymn to Demeter ( – ) and Pind. P. .  – .

Callimachus’ Doric Graces: 15 GP = 51 Pf.

31

often accompanies Apollo in tragedy (Eurip. Ion 126, 142; Soph. Phil. 829 – 30) whose closeness to the Graces is well attested. But there is an edge to Callimachus’ praise, as if, like Pindar, he felt the need to remind his laudanda that even charis can have a dark side. In his fifth hymn ‘The Bath of Pallas’, εὐαίων describes the nymph Chariclo whose son Teiresias has just been blinded because he inadvertantly saw the goddess naked in her bath. Chariclo is ‘blessed’, but only in comparison with the mother of Actaeon, who suffered a much worse punishment for a similar crime. In comparison with being torn apart and eaten by your own dogs blindness is a blessing. ἀρίζηλος also is double-edged. In Iliad 22.27– 29 it describes the armor of Achilles which shines like the light of the Dogstar, the brightest of all, but a sign of evil.³¹ And in Iliad 2.318 it is linked to another portent: a serpent, which appears while the Greeks are sacrificing to the gods at Aulis. The serpent jumps to a nearby tree where it eats eight baby sparrows together with their mother. Then, the poet says of the portent, τὸν μὲν ἀρίζηλον θῆκεν θεός, ὁς́ περ ἔφηνε· λᾶαν γάρ μιν ἔθηκε Κρόνου πάϊς ἀγκυλομήτεω. 318 ἀρίδηλον Zenodotus: ἀΐζηλον Aristarchus 319 athetized The god who revealed it, made it ἀρίζηλον / the son of crooked-counseling Kronos, turned it to stone. Il. 2.318 – 19

As Schmid notes, these verses are especially relevant to Callimachus’ epigram because here ἀρίζηλος describes a living creature that is turned to stone, i. e., it becomes a statue.³² The passage was of interest to Homer’s Alexandrian editors and to Callimachus himself because Aristotle had commented on it.³³ In using ἀρίζηλος to describe Berenike as a statue, the poet may have been thinking not only of a divine portent turned to stone, but of its ultimate form, the constel The rising of the Dogstar is also precisely the moment when Myrrh trees are ready to be tapped (Theophr. Hist. Plant. ..).  Schmid , .  The question asked by Aristotle was why does Chalcas’ speech (Il. . – ) interpret the snakes’ eating of the birds as a portent, while ignoring the metamorphosis of the snake into stone (Porph.  Rose =  Gignon). Aristotle concludes that the petrification of the serpent does not refer to the return of the Greeks, but to the slowness and long duration of the war, which was not relevant in the dramatic circumstances. Thus, Il. . presents no difficulty. Both Zenodotus and Callimachus seem to accept this. Later, Aristarchus tried to improve the cö́ for ἀρι ́ζηλον and athetizing .. A detailed study herence of the passage by reading ἀιζηλον with bibliography is in Montanari .

32

Dee L. Clayman

lation Draco, which is in the same family of constellations as the Coma Berenices. The fourth verse is the poem’s elegant conclusion, ‘Berenika, without whom the Graces themselves are not Graces.’ The relative pronoun, ἇς, is Doric like its referent, and now the article accompanying the Graces is also Doric. So great is her power, that though she is only one, and they, three, the combined ensemble takes on her identity. They and she are all Doric now and they have become indistinguishable – a group of four Doric Graces. To say that a queen is like a goddess or in this case, like a Grace, or that the queen is a Grace, is not the height of poetic creativity. The conceit that a human could be like a deity begins with Homer, and was taken to a new level of literalness by Alexander and his successors. But this is not exactly Callimachus’ point. By manipulating dialect forms, Callimachus has found a subtle way to say not only that Berenike has become a Grace, but that the Graces have become Berenikes. This puts a new spin on an old cliche, and is a much greater compliment. The newly doricized Graces are also uniquely Callimachus’. Now that Berenke has been grafted onto them, he no longer has to make do with the usual three Ionians. Now he has a unique set of four, specifically Dorian Graces to bring their many gifts to his work. These would include not only their inspiration and beauty, but their newly acquired financial and political clout. No one would doubt the power of that combination or Callimachus’ success in forging it. It is clear, then, that Callimachus uses dialect here, with all of its literary, social and political connotations, to claim a unique and powerful position for himself and his art. Though attention to dialect reveals something new about Callimachus’ artistry with words, it does not addresss all of the interpretative issues of the poem. One that has generated ongoing discussion is the question of what relation the poem may have to Callimachus’ literary agendas. Ivana and Andrej Petrovic, for example, have argued that the four Graces represent the four books of the Aitia, which begins with the Parian Graces and ends with Berenike’s ‘Lock’, and that it may have announced the publication of books 3 and 4.³⁴ Benjamin AcostaHughes and Susan Stephens come to the same conclusion using a slightly different approach.³⁵ They understand the epigram’s referent not as a perfumed statue, but a perfumed lock of hair, representing Berenike’s ‘Lock.’ Neither argument takes into account the corporality of the statue which was molded (ποτεπλάσθη) and is wet (νοτεῖ), but the statue’s apparent solidity does not invalidate them.

 I. Petrovic & A. Petrovic ,  – .  Acosta-Hughes & Stephens ,  – .

Callimachus’ Doric Graces: 15 GP = 51 Pf.

33

The poet does not make clear whether he is describing a real statue or an imagined one, and real or imagined, it could be understood metaphorically. Though attention to dialect will not solve this particular problem, it is possible to associate the epigram with Callimachus’ poetic program in another way. At the conclusion of his Hymn to Apollo Envy speaks to Apollo and declares, ‘οὐκ ἄγαμαι τὸν ἀοιδὸν ὃς οὐδ’ ὅσα πόντος ἀείδει.’ τὸν Φθόνον ὡπόλλων ποδί τ’ ἤλασεν ὧδέ τ’ ἔειπεν· ‘᾿Aσσυρίου ποταμοῖο μέγας ῥόος, ἀλλὰ τὰ πολλά λύματα γῆς καὶ πολλὸν ἐφ’ ὕδατι συρφετὸν ἕλκει. Δηοῖ δ’ οὐκ ἀπὸ παντὸς ὕδωρ φορέουσι μέλισσαι, ἀλλ’ ἥτις καθαρή τε καὶ ἀχράαντος ἀνέρπει πίδακος ἐξ ἱερῆς ὀλίγη λιβὰς ἄκρον ἄωτον.’ ‘I do not like the singer who does not sing as much as the sea.’ Then Apollo struck Envy with his foot and said, ‘Great is the flood of the Assyrian river, but it hauls much refuse and garbage in its water. The bees carry water to Deo not from every source but from that which flows, pure and immaculate, from a holy spring, a little trickle, the topmost flower.’ Call. Hymn to Apollo, 106 – 112

This passage has long been understood as a statement of Callimachus’ literary values that contrasts two kinds of style.³⁶ One is like a river that is large, copious, and filled with garbage. The other resembles tiny liquid droplets that are pure, sacred, and harvested from perfection itself, the topmost flower. Though a variety of interpretations have been advanced for the identity of the great Assyrian stream,³⁷ it is clear that the pure droplets carried by the bees³⁸ represent Callimachus’ own poetry. The droplets are not exactly water but nectar, the clear, thin liquid that bees transform into honey. Berenike’s statue, damp with stacte, the pure liquid which produces the finest myrrh is equivalent to the ‘topmost flower’,

 A bibliography of the modern discussion is in Williams ,  which is supplemented with more recent bibliography in Cheshire (,  nt. ). Typically this passage is read as Callimachus’ defense against his critics, especially Apollonius of Rhodes. Cheshire argues that Apollo is portrayed here as defending the chorus which performs in the hymn.  The great stream has been variously identified with the Euphrates (Schol. ad loc.); the Pontic Assyrian rivers such as the Thermodon, Iris and Halys (Huxley ); or Homer’s Ocean (Williams ,  – ).  The bees have been understood as poets (Bacchylides . –  S.-M.; Aristoph. Birds  – ), priestesses of Demeter (Schol. ad Theoc. ./a; Schol. ad Pind. Pyth. .a; Apollodorus of Athens, FGrH  F  = P.Oxy.  col. .), or literally bees.

34

Dee L. Clayman

the perfect source of nectar from which Callimachus distills his honeyed verse.³⁹ In this way, his fourth Dorian Grace is not only a patron and inspiration for the poet, but a template for precisely the kind of poetry that he wishes to write.

Abbreviations AB AP GP

Austin, C./Bastianini, G. (eds.) (2002), Posidippi Pellaei quae supersunt omnia, Milan. Anthologia Palatina. Gow, A./Page, D. (eds.) (1968), The Greek Anthology: The Garland of Philip and Some Contemporary Epigrams, 2 vols., Cambridge. LSJ Liddell, H.G. / Scott, R. A Greek-English Lexicon, rev. and augm. throughout by Stuart Jones, H. with the assist. of McKenzie, R. and with the cooperation of many scholars, Oxford 1940 + A Supplement, ed. by Barber, E.A. with the assist. of Maas, P. / Scheller, M. / West, M.L., Oxford 1968 + Revised Supplement, ed. by Glare, P.G.W. with the assist. of Thompson, A.A., Oxford 1996. OGIS W. Dittenberger (ed.) (1903 – 1905), Orientis graeci inscriptiones selectae, Leipzig.

Bibliography Abbenes, J.G.J. (1996), ‘The Doric of Theocritus: A Literary Language’, in: M.A. Harder / R.F. Regtuit / G.C. Wakker (eds.), Theocritus, Groningen, 1 – 19. Acosta-Hughes, B. / Stephens, S.A. (2012), Callimachus in Context, Cambridge. Bastianini, G./Gallazzi, C. (eds.), con la collab. di C. Austin (2002), Posidippo di Pella Epigrammi (P.Mil.Vogl. 309), Milan. Brunck, R.F.P. (1772 – 76), Analecta Veterum Poetarum Graecorum, 3 vols., Strassburg. Cheshire, K. (2008), ‘Kicking Phthonos: Apollo and his Chorus in Callimachus’ Hymn 2’, in: CPh 103, 354 – 73. Clayman, D.L. (2014), Berenice II and the Golden Age of Ptolemaic Egypt, Oxford. Colvin, St. (1999), Dialect in Aristophanes and the Politics of Language in Ancient Greek Literature, Oxford. Crane, G. (1987), ‘Bees without Honey and Callimachean Taste’, in: AJPh 108, 399 – 403. Deichgräber, K. (1971), Charis und Chariten. Grazie und Grazien, Munich. Gow, A.S.F. / Page, D. (eds.) (1965), The Greek Anthology: Hellenistic Epigrams, 2 vols., Cambridge. Griffiths, F.T. (1979), Theocritus at Court, Leiden. Harder, A. (ed.) (2012), Callimachus: Aetia, 2 vols., Oxford. Hopkinson, N. (ed.) (1984), Callimachus: Hymn to Demeter, Cambridge.

 Honey is a conventional marker for the sweetness of poetry (Hes. Th.  – ; Vita Pindari . –  Drachmann.; Theoc. Id. . – ). It is noteworthy that the bees do not bring Callimachus honey, but its less viscous precursor, nectar (not literally water, pace Crane ) which the poet himself will transform into verse.

Callimachus’ Doric Graces: 15 GP = 51 Pf.

35

Horrocks, G. (2010), Greek: A History of the Language and its Speakers, 2nd ed., Malden, MA. Hunter, R. (1996), Theocritus and the Archaeology of Greek Poetry, Cambridge. — (2005), ‘Speaking Glossai: Dialect Choice and Cultural Politics in Hellenistic Poetry’, in: W.M. Bloomer (ed.), The Contest of Language: Before and Beyond Nationalism, Notre Dame IN, 187 – 206. Huxley, G. (1971), ‘Kallimachos, the Assyrian River and the Bees of Demeter’, in: GRBS 12, 211 – 15. Kaibel, G. (1877), ‘Observationes criticae in Anthologiam Graecam’, in: Commentationes philologae in honorem Theodori Mommseni scripserunt amici, Berlin, 326 – 336. Lloyd-Jones, H. / Parsons, P. (1983), Supplementum Hellenisticum, Berlin. Lucas, A. (1937), ‘Myrrh and Stacte’, in: Journal of Egyptian Archaeology 23, 27 – 33. Massimilla, G. (ed.) (1996), Callimaco: Aitia Libri Primo e Secondo, Pisa. — (2010), Callimaco: Aitia Libro Terzo e Quarto, Pisa. McLennan, G.R. (1971), ‘A Note on Callimachus, Hymn 5.83’, in: CQ 21, 425. Molinos Tejada, T. (1990), Los dorismos del Corpus Bucolicorum, Amsterdam. Montanari, F. (2008), ‘Aristotele, Zenodoto, Aristarco e il serpente pietrificato di Iliade II 319, in: P. Arduini / S. Audano et al. (eds.), Studi offerti ad Alessandro Perutelli, vol. II, Rome, 237 – 44. Petrovic, I. / Petrovic, A. (2003), ‘Stop and Smell the Statues: Callimachus’ Epigram 51 Pf. Reconsidered (Four Times)’, in: MD 51, 179 – 208. Pfeiffer, R. (1928), ‘Ein neues Altersgedicht des Kallimachos’, in: Hermes 63, 302 – 41. — (ed.) (1949 – 53), Callimachus, 2 vols., Oxford. Powell, J.U. (ed.) (1925), Collectanea Alexandrina, Oxford. Reed, J.D. (2000), ‘Arsinoe’s Adonis and the Poetics of Ptolemaic Imperialism’, in: TAPhA, 130, 319 – 51. Ruijgh, C. (1984), ‘Le dorien de Théocrite: Dialecte cyrénien d’Alexandrie et d’Égypte’, in: Mnemosyne 37, 56 – 88. Schmid, W. (1923), ‘Ἀρίζηλος Βερενίκα (Callimach. epigr. 51 Wilamowitz)’, in: Philologus 32, 176 – 79. Sens, A. (2004), ‘Doricisms in the New and Old Posidippus’, in: B. Acosta-Hughes / E. Kosmetatou / M. Baumbach (eds.), Labored in Papyrus Leaves: Perspectives on an Epigram Collection Attributed to Posidippus (P.Mil.Vogl. VIII 309), Washington DC, 65 – 83. Tribulato, O. (2012), Language and Linguistic Contact in Ancient Sicily, Cambridge. Weber, G. (1993), Dichtung und höfische Gesellschaft: Die Rezeption von Zeitgeschichte am Hof der ersten drei Ptolemäer, Stuttgart. — (1997), ‘Interaktion, Repräsentation und Herrschaft: Der Königshof im Hellenismus’, in: A. Winterling (ed.), Zwischen Haus und Staat: Antike Höfe im Vergleich, Munich, 28 – 71. — (2011), ‘Poet and Court’, in L. Lehnus / B. Acosta-Hughes / S.A. Stephens (eds.), Brill’s Companion to Callimachus, Leiden, 225 – 44. Wilamowitz-Moellendorff, U. von (1924), Hellenistische Dichtung in der Zeit des Kallimachos, 2 vols., Berlin. — (ed.) (1925), Callimachi Hymni et Epigrammata, 4th ed., Berlin. Williams, F. (ed.) (1978), Callimachus: Hymn to Apollo, Oxford. Wright, C.R. (1917), ‘Food of the Gods’, in: CR 31, 4 – 6.

Taylor Coughlan

Dialect and Imitation in Late Hellenistic Epigram The selection, distribution, and mixture of dialect in Hellenistic book epigram are complex and meaningful components of an epigrammatist’s presentation of his text.¹ As scholars have recently begun to recognize, dialect has the potential to mark the ethnic identity of an epigrammatic subject, suggest a context in which to imagine the display of a dedication or epitaph, construct a poetic selfidentity, or recall a literary model.² This paper addresses this last aspect of dialect usage, focusing on the use of dialect in the imitations of third-century BC book epigrams by later Hellenistic epigrammatists, and paying particular attention to Antipater of Sidon, Archias, and Meleager. While imitation was long considered a hallmark of poetic inferiority, recent work has shown the meaningful contribution that imitation makes to our understanding of the self-conscious poetics of book epigram.³ Imitation is an art in its own right.⁴ The aspects of the model epigram that the imitation adopts, varies, or discards contextualizes the later poet’s reading, allowing the imitator to construct new meaning out of an established generic tradition. Imitation, in other words, is a form of reader-response, providing insight into the imitator’s reception of the model as much as it spurs our own reading.⁵ My analysis will show that later Hellenistic imitators were sensitive readers who were keenly aware of the dialect choices in their model texts. Reading lit-

 The dialectal composition of Hellenistic book epigram is undoubtedly compromised by the process of textual transmission (see Gow and Page , .xlv-xlv; Cameron ,  – ; and Sens , lxv-lxvi). While the authenticity of an individual dialectal form is ultimately uncertain, my own survey of the transmission of common dialect forms finds that certain manuscripts are not so unreliable as to prevent detecting aggregate patterns of dialect usage across the corpora of individual authors.  See Palumbo Stracca  and  – ; Sens ; Guichard ; Bettarini ; Sens ; and Gutzwiller .  Fundamental discussions of artistic imitation in Hellenistic book epigram are Ludwig , Tarán , Gutzwiller ,  – ; see also Sens .  On literary concept and practice of μίμησις (or imitatio in Latin), see Reiff ; the essays in West and Woodman , particularly Russell ; Conte ; and Halliwell .  The scholarship on reader-response theory is substantial; fundamental critical discussions are Iser  and  and Fish . For the application of reader-response theory to ancient texts see the introductory essay by Pedrick and Rabinowitz to their  guest edited volume of Arethusa along with the collected articles.

38

Taylor Coughlan

erary imitations through the lens of dialect choice offers a significant opportunity to explore how ancient poets understood and responded to the dialect choices of their literary models. In the following analysis, I explore the development of dialect as a literary device in epigram imitations. In the first section, I explore the close imitations of Antipater of Sidon. I demonstrate that Antipater reuses the dialect of his model as a literary device, employing the matching or opposing dialectal color in key words that draw attention to his reading and variation of the model. In the second section, I examine a sequence of imitations. Archias follows Antipater in his imitations of Leonidas, inserting himself into the chain of compositions by borrowing language and structure from both poems. I argue that this creative contamination extends to the use of dialect. In the final section, I turn to Meleager, paying attention to his role both as poet and editor. In addition to continuing the use of dialect in individual imitations seen in Antipater and Archias, Meleager, I contend, closely reads dialect from an editorial perspective, deploying it as an organizational device in the Garland.

Reading third-century epigrams with Antipater of Sidon Antipater of Sidon, a second-century epigrammatist of Phoenician extraction, is our most prolific surviving imitator and whose securely ascribed epigrams number over sixty.⁶ The Sidonian was famous in antiquity for his improvisational versification.⁷ And indeed his close imitations of earlier epigrams (perhaps recited immediately before his impromptu response), especially the surviving series of self-variations on a topic such as Myron’s Cow (HE 36 – 40) or the poet Ana-

 Antipater was a common name in antiquity and the Sidonian shares it with another, later epigrammatist Antipater of Thessalonica who was collected by Philip in his Garland. In a number of instances the lemma to an epigram lacks an ethnic, often making the identification between the Sidonian and Thessalonican uncertain. The positioning of an epigram in the Greek Anthology (e. g. Meleagrean or Philipan sequences) and various stylistic criteria have been used by scholars to assign these ambiguous poems to one of the two authors; Argentieri ,  –  collects these epigrams and offers a detailed analysis of their ascriptions ( – ). Here I treat only epigrams whose ascriptions to Antipater of Sidon are generally undisputed.  See famously the comments of Lucius Licinius Crassus in Cicero’s De oratore (.), who, while extoling the benefits of varying the structure and rhythm of the final words of a sentence, recalls Antipater of Sidon and his great facility at fashioning verses ex tempore.

Dialect and Imitation in Late Hellenistic Epigram

39

creon (HE 13 – 16), do have an air of ex tempore composition in performance. Whatever the extent of Antipater’s improvisational versification, these oral compositions were eventually written down, likely by the poet himself (rather than some sort of transcription in performance), at which point they underwent some form of revision and refinement for presentation in a pamphlet or poetry book. Accordingly, I approach Antipater’s imitations of earlier epigrams and his use of dialect from the perspective of a reader of these epigrams, rather than from the perspective of a listener to an improvisational performance. The appeal of Antipater’s imitative technique resides in his ability to vary the diction, structure, or point of his model so that the imitation becomes an intellectually satisfying act of reception. That dialect is a component of this imitative technique is evident from Antipater’s imitation (7 HE=AP 7.146) of Asclepiades 29 HE=AP 7.145 on the mourning figure of Virtue likely to be imagined as carved in relief on Ajax’s tomb.⁸ Ἅδ’ ἐγὼ ἁ τλάμων ᾿Aρετὰ παρὰ τῷδε κάθημαι Αἴαντος τύμβῳ κειραμένα πλοκάμους, θυμὸν ἄχει μεγάλῳ βεβολημένα, εἰ παρ’ ᾿Aχαιοῖς ἁ δολόφρων ᾿Aπάτα κρέσσον ἐμεῦ δύναται. 2 κειραμένα Plpc: κειρονένα PPlac ἐμοῦ Pl, κρέσσων ἐμοῦ Eust.

4 κρέσσον ἐμεῦ CPepl. Tzetzes: κρέσσονα μεῦ P, κρεῖσσον

I, bold Virtue, sit here by this Tomb of Ajax, having cropped my hair, stricken with a great grief in my heart, since among the Achaeans cunning Deceit is stronger than me.⁹ Asclepiades 29 HE=AP 7.145 σῆμα παρ᾽ Αἰάντειον ἐπὶ Ῥοιτηίσιν ἀκταῖς θυμοβαρὴς ᾿Aρετὰ μύρομαι ἑζομένα, ἀπλόκαμος, πινόεσσα, διὰ κρίσιν ὅττι Πελασγῶν οὐκ ἀρετὰ νικᾶν ἔλλαχεν ἀλλὰ δόλος.

 A note on my conventions for marking dialect: Doric forms appear in bold; Ionic forms are underlined; koine forms, in those instances where they are juxtaposed with another dialect, are italicized. In those cases where a form is dialectally ambiguous, such as α following ε, ι, or ρ (Doric or Attic) or the contraction of the diphthong εo to ευ (Doric or Ionic), I mark the form in accordance with the surrounding dialectal environment. I print the text of epigrams from the editions of Gow and Page; the critical apparatuses are composite created by the author.  All translations are my own.

40

Taylor Coughlan

τεύχεα δ᾽ ἄν λέξειεν ᾿Aχιλλέος, “ἄρσενος ἀλκᾶς¹⁰ οὐ σκολιῶν μύθων, ἄμμες ἐφιέμεθα.” 5 ἀλκᾶς Page: ἀκμᾶς PPl Sitting alongside the tomb of Ajax on the Rhoiteian promontory I, Virtue, heavy at heart, weep, having shorn my hair, unwashed, because of the judgment of the Pelasgians that not virtue but deceit has obtained victory. Achilles’ armor would have said, “We long for manly courage not crooked tales.” Antipater of Sidon 7 HE=AP 7.146

With one exception, the manuscript witnesses to Asclepiades 29 HE transmit a consistent Doric color principally consisting of the preservation of inherited *ā as α. Planudes (Pl) transmits ἐμοῦ for ἐμεῦ in the Palatine Codex (P), which contains a resolution of the εο diphthong prevalent in contemporary Doric,¹¹ and should be understood as one of the manuscript’s many regularizations of Doric forms to their Attic-Ionic equivalents.¹² Asclepiades’ dialect usage likely takes its inspiration from inscriptional and literary antecedents. As regards the former, it was common practice for inscribed epitaphs, whether erected at home or abroad, to be composed in the native dialect of the deceased. Ajax’s homeland of Salamis, before being turned over to Athenian control following his death, was Megarian; Asclepiades’ use of Doric, as Alexander Sens has argued, can be read as an acknowledgement of the Telamonian’s imagined native dialect.¹³ Sens’ reading is highly suggestive and has parallels in other third-century BC epigrammatists. Most notable are Posidippus’ two Doric epigrams on Cretan subjects from the Milan papyrus, AB 64 on Cresilas’ bronze statue of  I print the simple and plainly correct emendation of Page for the ἀκμᾶς of PPl; despite the error the dialectal color found in the manuscript is almost certainly secure.  The contraction is widely documented in Aegean Doric inscriptions by the middle of the fourth century BC (e. g. Thera, Astypalea, Cnidos, Rhodes, and Cos) except on Crete and Heraclea; see Buck , §.; Bartoněk ,  –  and ; and Bubeník ,  –  and . The diphthong is also a feature of Hellenistic literary Doric; see Bulloch ,  –  on Callimachus and Molinos Tejada, ,  –  on Theocritus. This contraction is not, however, unambiguously Doric, since it is also a feature of Ionic from the fourth century BC, particularly in the East (e. g. Chios, Miletus, and Teos); see Stüber ,  – . In Hellenistic book epigram, the attribution of a Doric or Ionic color to the contraction should be based on the surrounding dialectal composition. The predominant Doric color of Asclepiades’ epigram makes it very likely that we should interpret ἐμεῦ as Doric.  On this phenomenon in Pl, see Waltz , .li; Gow and Page , .xxxvi-xxxvii; and Cameron , .  Sens , .

Dialect and Imitation in Late Hellenistic Epigram

41

the Cretan hero Idomeneus and AB 102 an epitaph for a certain Menoetius of Crete who died abroad.¹⁴ As regards literary influences, scholars have long recognized that Aristotle’s Hymn to Virtue (PMG 842) likely inspired Asclepiades to write an epitaph for the Homeric hero that included the figure of Virtue.¹⁵ In the second-half of the song Aristotle invokes Achilles and Ajax as heroic exempla for the greatness that Virtue can grant to those who follow her. Now ranking among her sacrificial dedicatees is Hermias, Aristotle’s father-in-law and the philosopher tyrant of Atarnaeus, whose murder the song memorializes. For our interests, the Hymn to Virtue opens with Doric ἀρετά, a choice that initiates the light Doric coloring present throughout the poem. Asclepiades also features Doric ἀρετά in the first line of his epitaph and it is quite probable that the similarity in dialectal coloring is designed to recall the Aristotelian model and its mournful register. Antipater’s dialect usage is more varied than his model, incorporating koine and Ionic features alongside the Doric α vocalism. This variance is clear from the opening couplet. Where Asclepiades places Virtue in the hexameter and Ajax’s tomb in the pentameter, Antipater inverts this organization. The hexameter is completely lacking in Doric forms, where they could easily be incorporated; in the pentameter, however, Antipater includes the Doric color that appears consistently throughout Asclepiades. Here Antipater reserves the dialect for two words that signals his reception of Asclepiades. The first doricized word is ἀρετά and its significance is directly related to Antipater’s variation of the structure of his model’s opening couplet. By first stating the tomb’s occupant and location, such information would direct the reader, if one knew the model poem, to expect the introduction of Virtue in the pentameter. And indeed, when Virtue does appear in the pentameter as the subject of the sentence, her naming includes the first Doric form, a dialectal notice of Antipater’s imitative position in relation to Asclepiades. This self-conscious use of Doric color in ᾿Aρετά is reinforced in line four. Personified Virtue admits the victory of deceit over virtue (οὐκ ἀρετὰ νικᾶν ἔλλαχεν ἀλλὰ δόλος), recalling the final line in Asclepiades (ἁ δολόφρων ᾿Aπάτα κρέσσον ἐμεῦ δύναται), where the use of Doric in the abstract conceptualization of virtue (ἀρετά) varies the Doric pronoun ἐμεῦ for the personified Virtue. The second Doric form in the first pentameter is ἑζομένα. As is common in Antipater’s imitations, the Sidonian varies the syntax of the model. Antipater re-

 For further discussion of the dialect as a marker of Cretan identity in these epigrams, see Palumbo Stracca  – ,  –  (AB ) and Sens ,  –  (AB ) and  –  (AB ).  See Guichard ,  –  and Sens , . The points of contact with Aristotle’s Hymn to Virtue partially explains why this epigram was incorporated into the Aristotelian peplos; see Cameron , .

42

Taylor Coughlan

places Asclepiades’ main verb κάθημαι (in an unmarked dialect) with ἑζομένα a participial form of a synonym. Whereas the use of Doric in ᾿Aρετά communicates an initial point of contact with the model, Antipater uses the doricized ἑζομένα to draw attention to its variation.¹⁶ Finally, Antipater adds a further couplet in which Virtue imagines that Achilles’ armor prefers the manly courage of Ajax over the wily machinations of Odysseus. Speeches that envision what people might say are a feature of Homer and became a game played in epigram, mostly in Book 11 of the Greek Anthology. Inspired by this Homeric practice, Antipater employs some epicizing language in the uncontracted nouns τεύχεα and ᾿Aχιλλέος within the frame of the speech. In the armor’s imagined speech we find another epicism ἄμμες, but alongside it the doricized ἀλκᾶς, Page’s convincing emendation of the manuscripts’ ἀκμᾶς. As a marker of Virtue’s voice and the identity of Ajax, the Doric color of ἀλκᾶς calls attention to the connection Virtue draws between Ajax and “manly virtue” (ἄρσενος ἀλκᾶς)—in contrast to Odysseus’ deceitful cunning (σκολιῶν μύθων)—which she seeks to inject into the imagined judgment of Achilles’ armor. Altogether, then, in the imitation of Asclepiades we have the hallmarks of Antipater’s use of dialect as a literary device that we shall see repeated in various ways throughout his other imitations. Leonidas of Tarentum is the most commonly imitated poet among Antipater’s extant epigrams. A centerpiece of Leonidean epigram is the integration of elevated language into the simple and straightforward structure and subject matter of dedicatory and sepulchral epigram.¹⁷ Epigrams of this sort were also ideally suited to close imitation and likely explain, in part, the many imitators of Leonidas.¹⁸ Constructed out of relatively simple and interchangeable syntactic building blocks, these verses easily accommodated the substitution of adjectives and phrases that are foundational, as we shall see, to Antipater’s art of imitation. Leonidas 41 HE=AP 6.288 and HE 42=AP 6.289, two dedicatory epigrams for groups of sisters whose occupation was weaving, exemplify the Tarentine’s trademark combination of an elevated stylistic register with banausic subject matter.

 While not dialectally marked, Antipater’s use of the neologism θυμοβαρής for τλάμων in the earlier epigram acts as an explanatory gloss to remove the ambiguity of the Homeric term, which can mean both stout-hearted and wretched, and so further reinforces the Sidonian’s reading and interpretation of Asclepiades’ language.  On Leonidas’ application of elevated language to low or banausic situations and figures, see Gigante  and Criscuolo .  For a survey of the various techniques of imitation and variation of Leonidas by later epigrammatists, see Ypsilanti .

Dialect and Imitation in Late Hellenistic Epigram

43

Αἱ Λυκομήδευς παῖδες ᾿Aθηνὼ καὶ Μελίτεια καὶ Φιντὼ Γληνίς θ’, αἱ φιλοεργόταται, ἔργων ἐκ δεκάτας ποτιθύμια τόν τε πρόσεργον ἄτρακτον καὶ τὰν ἄτρια κριναμέναν κερκίδα, τὰν ἱστῶν μολπάτιδα, καὶ τὰ τροχαῖα πανία †κερταστὰς τούσδε ποτιρρόγεας†, καὶ σπάθας εὐβριθεῖς †πολυάργυρα τὼς δὲ† πενιχραί ἐξ ὀλίγων ὀλίγην μοῖραν ἀπαρχόμεθα. τῶν χέρας αἰέν, ᾿Aθάνα, ἐπιπλήσαις μὲν ὀπίσσω,¹⁹ θείης δ’ εὐσιπύους ἐξ ὀλιγησιπύων. 1 Λυκομήδευς Meineke: λυκομήδης C, λακαμήδης P We, the industrious daughters of Lycomedes, Atheno and Meliteia and Phinto and Glenis, destitute, offer as a tithe from our work these welcome gifts, the laborious spindle and the shuttle that passes between the thread of the warp, the songstress of the loom, and the spinning spools……and wool-laden blades… a little portion from our meager holdings. Ever after, Athena, fill our hands, and render us rich in bread out of our current poverty. Leonidas 41 HE=AP 6.288 Αὐτονόμα, Μελίτεια, Βοΐσκιον, αἱ Φιλολᾴδεω καὶ Νικοῦς Κρῆσσαι τρεῖς, ξένε, θυγατέρες, ἁ μὲν τὸν μιτοεργὸν ἀειδίνητον ἄτρακτον, ἁ δὲ τὸν ὀρφνίταν εἰροκόμον τάλαρον, ἁ δ’ ἅμα τὰν λεπτῶν εὐάτριον ἐργάτιν ἱστῶν κερκίδα, τὰν λεχέων Πανελόπας φύλακα, δῶρον ᾿Aθαναίᾳ Πανίτιδι τῷδ’ ἐνὶ ναῷ θῆκαν, ᾿Aθαναίας παυσάμεναι καμάτων. 4 ὀρφνίταν Grüter: ὀρφνείταν P

7 ᾿Aθαναίᾳ C: ᾿Aθαναίῳ P

Autonoma, Meliteia, Boiskion, the three Cretan daughters of Philolaides and Nico, stranger,—the first the thread making, ever whirling spindle, the second the dark wool basket, the third the shuttle of the loom, the well-weaving worker of the delicate warp, the guard of Penelope’s marriage bed— dedicated these things as gifts to Weaver Athena in her temple, having retired from the labors of Athena. Leonidas 42 HE=AP 6.289

 I print Stadtmüller’s metrically required emendation of C’s ἴσως.

44

Taylor Coughlan

The dialect color of both epigrams is predominantly Doric, which consists exclusively of the preservation of inherited *ā as α. Leonidas’ usage of Doric in 42 HE, in a fashion similar to Asclepiades’ Ajax epitaph, evokes the native language of the Cretan dedicators. While Leonidas does not supply an ethnic identity to the sisters in 41 HE, the association between Doric and Cretan identity in 42 HE likely colored the reading of the same dialect in this variation of the theme, especially since these epigrams may well have sat next to each another in a collection of Leonidas’ epigrams as they did in the Garland where Meleager had possibly excerpted them in a similar arrangement. In his imitation, Antipater borrows features of diction and phrasing from both models, creating a single new dedication out of the two variations: Παλλάδι ταὶ τρισσαὶ θέσαν ἅλικες ἶσον ἀράχνᾳ τεῦξαι λεπταλέον στάμον᾽ ἐπιστάμεναι, Δημὼ μὲν ταλαρίσκον ἐύπλοκον, ᾿Aρσινόα δέ ἐργάτιν εὐκλώστου νήματος ἠλακάταν, κερκίδα δ᾽ εὐποίητον, ἀηδόνα τὰν ἐν ἐρίθοις, Βακχυλίς, εὐκρέκτους ᾇ διέκρινε μίτους. ζώειν γὰρ δίχα παντὸς ὀνείδεος εἵλεθ᾽ ἑκάστα, ζεῖνε, τὸν ἐκ χειρῶν ἀρνυμένα βίοτον. 2 στάμον᾽ P: στήμον᾽ Pl 6 ᾇ Pl: ἐῦ P, εὖ Sud. 7 ἑκάστα P: ἑκάστη Pl, ἔκάστα Sud. Pallas Athena received as dedications from three age-mates, who knew how to fashion a delicate web as well as a spider, Demo a well-plaited basket, Arsinoe a distaff, the workwoman of well-spun thread, and Bachulis a well-made shuttle, the nightingale among weavers, with which she parted the finely woven threads. For each of them wished to live free from every reproach, stranger, gaining their livelihood from their hands. Antipater 5 HE=AP 6.174

As he varies the internal construction and the motivating purpose for the dedication, Antipater retains the general Doric color of his model (again limited to the preservation of inherited *ā), bringing with it the imagined voices of the original dedicators. Antipater’s first line is a summary of the components of the two Leonidean epigrams that allows him to quickly declare his dependence on the models. Line one is also the most doricized of the epigram, containing three forms: ταὶ, ἅλικες, and ἀράχνᾳ, which are distributed evenly throughout the hexameter in feet two, four, and six. Antipater has encapsulated the basic content of the epigram through the use of Doric dialect in these three words, all of which recall their original use in his models—three women (ταὶ τρισσαὶ), of some relation (ἅλικες), making dedications related to weaving (ἀράχνᾳ).

Dialect and Imitation in Late Hellenistic Epigram

45

At the same time, Antipater introduces variety into his imitation with the substitution of ἅλικες for παῖδες or θυγατέρες. By making the three women age-mates, Antipater essentially rules out that they are sisters like in the model poems. The Sidonian accentuates his rewriting of the relationship between the female dedicators through his choice of dialect. The Doric color that Antipater gives to ἅλικες integrates the variation within the framework of the imitation, which partially relies on a replication of the model dialect. An instructive parallel can be found in Antipater’s imitation (1 HE=AP 6.14) of Leonidas 46 HE=AP 6.13, a dedication of hunting nets by three brothers.²⁰ Due to its simple structure and patterning of three brothers and three types of nets, it inspired numerous imitations.²¹ We will return to this epigram below, examining it in light of Archias’ series of imitations. οἱ τρισσοί τοι ταῦτα τὰ δίκτυα θῆκαν ὅμαιμοι, ἀγρότα Πάν, ἄλλης ἄλλος ἀπ᾽ ἀγρεσίης, ὧν ἀπὸ μὲν πτανῶν Πίγρης τάδε, ταῦτα δὲ Δᾶμις τετραπόδων, Κλείτωρ δ᾽ ὁ τρίτος εἰναλίων. ἀνθ᾽ ὧν τῷ μὲν πέμπε δι᾽ ἠέρος εὔστοχον ἄγρην, τῷ δὲ διὰ δρυμῶν, τῷ δὲ δι᾽ ἠιόνων. 3 πτανῶν P: πτηνῶν Pl The three brothers dedicated these nets to you, Pan the Hunter, each from a different chase; Pigres these from the birds, Damis these from the beasts, Kleitor, the third one, from the fish. In return for these send a successful hunt through the air, through the woods, and through the waters. Leonidas 46 HE=AP 6.13 Πανὶ τάδ᾽ αὔθαιμοι τρισσοὶ θέσαν ἄρμενα τέχνας· Δᾶμις μὲν θηρῶν ἄρκυν ὀρειονόμων, Κλείτωρ δὲ πλωτῶν τάδε δίκτυα, τῶν δὲ πετανῶν ἄρρηκτον Πίγρης τάνδε δεραιοπέδην.

 Leonidas’ epigram was painted below an image depicting the three brothers making the dedication as part of the frescoes in the “House of the Epigrams” at Pompeii; see Dilthey  for the publication of the dipinti of Leonidas’ epigram and three others. The epigram is now poorly preserved, but what does survive presents no differences in dialect from the text transmitted in the manuscripts. The incipit to the epigram also survives on an ostrakon that contains a list of thirteen incipits (including Leonidas  HE=AP .) from the second century BC (SH ).  See Gutzwiller ,  –  on the appeal of Leonidas’ poem for later imitators. For a treatment of the entire sequence of imitations, including epigrams that fall outside of the chronological boundaries of the present discussion, see Laurens ,  – .

46

Taylor Coughlan

τὸν μὲν γὰρ ξυλόχων, τὸν δ᾽ ἠέρος, ὅν δ᾽ ἀπὸ λίμνης οὔ ποτε σὺν κενεοῖς οἶκος ἔδεκτο λίνοις. 1 τέχνας PSud. s.v. ὀρειονόμων et ἄρμενα: τέχνης PlSud. s.v. αὔθαιμοι 3 τῶν δὲ Pl: τὰν δὲ P, τάνδε Sud. πετανῶν C: πετηνῶν PSud., πετεινῶν Pl 4 τάνδε PSud.: τᾶνδε C, τήνδε Pl 5 λίμνης PPl: λίμναις C, fort. λίμνας Three brothers dedicated these tools of the trade to Pan: Damis a net for mountain-ranging beasts, Kleitor lattice nets for fish, and Pigres this unbreakable net, a collar for birds. Not once did their household receive them with empty nets, the one from the thickets, the second from the air, and the third from the sea. Antipater 1 HE=AP 6.14

In the Leonidean model the nets are referred to collectively in the first line as τὰ δίκτυα. The first line of Antipater 1 HE echoes the language of this first line but substitutes τὰ δίκτυα for the circumlocution ἄρμενα τέχνας. The vagueness of ἄρμενα τέχνας (produced partially in a Doric color whose significance will become apparent shortly), combined with the faithful reproduction of the rest of the first line, points to the imitative position of the epigram—those who recognize the line as an imitation already know the dedicated objects are going to be nets. The dialectal content of the model poem is mixed, with Leonidas inserting two Doric forms at the halfway point of the epigram (l.3: πτανῶν [P: πτηνῶν Pl]…Δᾶμις²²) in an epigram that is predominantly Ionic. Given the propensity of epigram to retain native dialectal forms in names, it could be argued that the reader is to understand Damis and his brothers are from a Dorian area. On the whole, however, the Ionic color of the epigram lends an elevated or heroicizing tone to the dedication of various hunting implements. Like Leonidas, Antipater appears to mix koine and Ionic forms with Doric, although textual uncertainties complicate our understanding of the Sidonian’s reception of the model poem’s dialect in several places.²³  EM ,  (Kallierges) understands the name as formed on analogy from δᾶμος, the Doric form of Attic-Ionic δῆμος.  This is particularly the case with the dialect form of the adjective πετεινός in line three of Antipater  HE. P, the generally more reliable witness for dialectal features, transmits most of the Doric forms in comparison to the Atticized forms in Pl (the Suda is more variable). The one exception is the Corrector’s (C) correction of P’s πετηνῶν to πετανῶν. C did have access to another exemplar, the anthology of Michael Chartophylax, which could have contained the Doric form; however, Planudes had similar access to this exemplar, as attested by a significant agreement between C’s corrections and the readings in Pl, but here Pl transmits πετεινῶν. Given the tendency for Pl to regularize Doric forms (especially when they appeared in a dialectally mixed context), πετεινῶν could mask an original πετανῶν, but we cannot also discount C’s

Dialect and Imitation in Late Hellenistic Epigram

47

Antipater’s reception of his model poem’s dialect is most clearly registered in the use of the phrase ἄρμενα τέχνας.²⁴ As I argued above, ἄρμενα τέχνας is a vague periphrasis for Leonidas’ τὰ δίκτυα that signals Antipater’s imitative position to the model. Within this periphrasis, however, Antipater also embeds information that communicates not only his knowledge of Leonidas’ epigrams for the hunting brothers, but also the Tarentine’s corpus more broadly. The Doric color in which he renders τέχνας is the linguistic mode through which he communicates this engagement with the corpus of Leonidas. There exists an interesting pattern of dialect usage, to my knowledge previously uncommented upon, in relation to τέχνη in Leonidas. Every use of the noun in the extant epigrams of Leonidas appears in the genitive and in Doric.²⁵ Including 46 HE, all of the epigrams in which τέχνας appears are dedications of implements by craftspeople (7 HE=AP 6.204 and 8 HE=AP 6.205 both on carpenters) or hunters (52 HE = AP 6.4 on a fisherman), often on the occasion of retirement from their laborious livelihood. More significantly, the use of Doric τέχνας appears in dialectally mixed contexts, suggesting that its dialect color is not simply influenced by the predominant dialectal color of the epigram in which it is used.²⁶ It is quite plausible, then, that the Doric form of τέχνη functioned as some type of signature of Leonidean aesthetics, especially when we consider his noted fascination with members of the working poor, such as weavers, carpenters, and other craftspeople. Why exactly Leonidas chose Doric is unclear—there is nothing in

own intervention into the text, especially given the position of Antipater’s imitation in the Palatine Anthology immediately following the Leonidean model. Despite the textual uncertainties, given the presence of a similar Doric form in Leonidas . HE (πτανῶν P: πτηνῶν Pl), it is tempting to print the reading of the Corrector, since πετανῶν would recall the one Doric form in the model and thus provide tantalizing evidence of Antipater’s reading of Leonidas’ dialect choices. I thus provisionally print the reading of C with the understanding that the koine πετηνῶν of P (or πετεινῶν of Pl), which would impart a degree of dialectal variation to Antipater’s imitation, is also a possible original reading.  The reading of τέχνης in Pl and Suda s.v. αὔθαιμοι is a patent example of the normalization of Doric vocalisms.  . HE=AP ..; . and  HE=AP .. and ; . HE=AP ... See also possibly Perses . HE=AP .. (μανυταὶ τέχνας P: μηνυταὶ τέχνης Pl), an epitaph for two woodsmen, which is the only transmitted Doric forms in the quatrain. We know very little about Perses, but it is possible that he pre-dates Leonidas. If that is the case the association of τέχνη with Doric color may well be a wider literary phenomenon, but we cannot completely dismiss the possible influence of Leonidas on Perses either.  In  HE=AP . τέχνας (l.) appears alongside Ionic χὠρμιήν (l.) and τρηχύν (l.); Doric also appears again in relation to craft in the final line: ἀρχαίας (C:-αίης Pl, -αῖα P)…τεχνοσύνας (P: -νης Pl).  HE=AP . (τᾷ Παλλάδι; ῥυκάναν) and  HE=AP . (᾿Aθάνᾳ) both display a limited Doric color.

48

Taylor Coughlan

particular to connect the various dedicatory contexts of its usage with this dialect choice—but what is rather more clear is that Antipater likely recognized this pattern of dialect usage in Leonidas’ dedications of craftspeople and hunters and used it to playfully communicate his interpretation and reception of the model poet’s corpus to his own readers. It was not only Leonidas’ dedicatory epigrams that Antipater used as models for his imitations. Antipater 45 HE=APl 178 on Apelles’ Aphrodite Anadyomene takes its inspiration from Leonidas 23 HE=APl 182. Again we find Antipater strategically reusing the dialect of his model to guide the reader through his own reading of the epigram. τὰν ἐκφυγοῦσαν ματρὸς ἐκ κόλπων, ἔτι ἀφρῷ τε μορμύρουσαν, εὐλεχῆ Κύπριν ἴδ᾽ ὡς ᾿Aπελλῆς κάλλος ἱμερώτατον οὐ γραπτὸν ἀλλ᾽ ἔμψυχον ἐξεμάξατο. εὖ μὲν γὰρ ἄκραις χερσὶν ἐκθλίβει κόμαν, εὖ δ᾽ ὀμμάτων γαληνὸς ἐκλάμπει πόθος, καὶ μαζός, ἀκμῆς ²⁷ ἄγγελος, κυδωνιᾷ. αὐτὰ δ᾽ ᾿Aθάνα καὶ Διὸς συνευνέτις φάσουσιν “ὦ Ζεῦ, λειπόμεσθα τῇ κρίσει.” 8 ᾿Aθάνα PlSQ: ᾿Aθηνά Σπ Having fled from her mother’s bosom, still frothing with foam, Cypris who brings wedded bliss— look how Apelles figured her most lovely beauty not painted but alive. Beautifully does she squeeze out her hair with her hands, beautifully does gentle desire flash forth from her eyes, and her breast, the prime of her youth, swells. Athena herself and the wife of Zeus will say: “O Zeus, we are left wanting in this judgment.” Leonidas 23 HE=APl 182

 While epic-Ionic μαζός is most likely present for metrical reasons (Attic-Ionic μαστός and Doric μασδός would not allow an opening dactyl), the presence of ἀκμῆς rather than Doric ἀκμᾶς (a modern emendation in HE) does, I believe, have a literary purpose. ἀκμή, as an art-historical term, was a core aspect of Apelles’ aesthetic mode. A mosaic from Byblos, which Moreno ,  –  identifies as deriving from a painting of Apelles, contains the personifications of καιρός, ἀκμή, and χάρις, the essential qualities of Apelles’ art. The ἀκμή in line seven, then, refers not only to the developmental stage of Aphrodite’s body but to a “flowering” intrinsic to Apelles’ aesthetic. Consideration of the word in art-historical terms relies, in part, on the retention of the Attic-Ionic form; the Doric, while perhaps more fitting to the dialect composition of the epigram, would obscure.

Dialect and Imitation in Late Hellenistic Epigram

49

τὰν ἀναδυομέναν ἀπὸ ματέρος ἄρτι θαλάσσας Κύπριν, ᾿Aπελλείου μόχθον ὅρα γραφίδος, ὡς χερὶ συμμάρψασα διάβροχον ὕδατι χαίταν ἐκθλίβει νοτερῶν ἀφρὸν ἀπὸ πλοκάμων. αὐταὶ νῦν ἐρέουσιν ᾿Aθηναίη τε καὶ Ἥρη “οὐκέτι σοι μορφᾶς εἰς ἔριν ἐρχόμεθα.” Cypris rising just now from her mother, the sea, Behold the labor of Apelles pencil, how having grasped her mane dripping with water in her hand she squeezes out the foam from her wet locks. Athena and Hera themselves will now say, “No longer will we enter into a contest of beauty with you.” Antipater of Sidon HE 45=APl 178

Apelles’ painting was originally displayed at the temple of Aphrodite on the island of Cos, which preserved its local Doric dialect well into the Hellenistic period.²⁸ On this account, I suggest that we read the dialect of the epigram as reproducing the painting’s original display context. Leonidas has cleverly used dialect to color the description of the work of art with an aural realism unable to be captured by paint. In a condensed elegiac version of the original, Antipater retains the general Doric color of the model. Before discussing its relationship to the imagined focalization of the painting, I want to make a few smaller observations on Antipater’s dialect usage. First, Antipater reproduces the opening Doric feminine accusative article and participle combination present in Leonidas. Here, however, in usual Antipaterian fashion, his variation of ἐκφυγοῦσαν with ἀναδυομέναν likely echoes the title of the painting, thus rendering the subject matter of the epigram identifiable from the start.²⁹ Second, where Leonidas does not identify Aphrodite’s mother as the sea by name, likely assuming his readers will find some intellectual pleasure in distinguishing the μάτηρ as they decode the scene, Antipater, as he does with τὰν ἀναδυομέναν, makes the recognition of the scene and its characters explicit with θαλάσσας. Finally, in line three Antipater inserts the synonymn χάιταν in Doric for Leonidas’ similarly colored κόμαν, making the dialectal variatio a piece of his deliberate dialogue with Leonidas. In each in-

 Bubeník ,  – .  The painting is so called by Pliny at NH . (Anadyomene vocatur, versibus Graecis tantopere dum laudatur, aevis victa sed inlustrata), and we can plausibly assume it was known by the title in the late Hellenistic period.

50

Taylor Coughlan

stance, then, Antipater highlights his reading and explication of the model epigram by inserting a Doric feature drawn from its primary dialectal coloring.³⁰ In the final couplet, the “judgment” of Athena and Hera, Antipater inverts the distribution of dialectal features. Leonidas wrote the name of Athena in Doric, used a dialectally unmarked periphrasis for Hera, and described the deities’ act of speech in a Doric future form of φημί. In their short declaration the dialect coloring of the manuscripts is koine, perhaps because as gods Athena and Hera would have been assumed not to speak Doric, the one feature notably absent from epic. Conversely, Antipater employed the ionicized forms of Athena and Hera, reapplying an epicizing luster to their names, and replaced Leonidas’ Doric future φάσουσιν with Ionic ἐρέουσιν. In the speech of Athena and Hera Antipater incorporates Doric μορφᾶς, which stands in contrast to the dialect used in the brief narrative that frames it. Like the imagined viewer of the painting, Athena and Hera are stunned by Apelles’ masterpiece and its life-like beauty. In the beauty contest they concede not to Aphrodite herself but to Apelles’ version— they admit, in other words, that they are no match for the μορφά on display in Cos. Antipater’s subtle application of Doric also focalizes the deities’ view of the painting, inviting the reader to imagine Athena and Hera standing in awe before the masterpiece. Finally, we might read Antipater’s choice to make Athena and Hera use μορφᾶς within the context of his imitative position. The epigram is itself a reproduction of Leonidas’ reproduction of Apelles’ painting. The epigram’s meaning, in other words, is intrinsically linked to its writtenness on the landscape of the bookroll, since a reader’s interpretation of the epigram is keyed to Antipater’s handling of Leonidas. While Leonidas primarily uses dialect to capture the aural realism of the painting’s location of display, Antipater uses dialect primarily to reproduce the mental act of reading Leonidas’ epigram on the page of the bookroll. On this model, the comment of Athena and Hera could be taken not only to address their appreciation of Apelles’ beautiful painting of Aphrodite, but also Antipater’s imitation of Leonidas’ epigram on Apelles’ beautiful painting of Aphrodite. The use of Doric μορφᾶς, then, is doubly significant. The Doric color recalls both the “beauty” of Apelles’ Coan Aphrodite and the “form” of Antipater’s imitation, an imitation which both engages with the Doric color found in its original and, more noticeably, transforms the model epigram’s metrical  The association of Doric with these epigrams on Apelles’ Aphrodite Anadyomene continues into the imperial period, where Lucillius (AP .), on the art thief Dion, copies the first line of Antipater of Sidon  HE=APl  along with its Doric color. Lucillius uses dialect to mark out this quotation whose expected ecphrastic context he then undermines in the rest of the epigram, which he composes in Attic-Ionic; see Floridi , .

Dialect and Imitation in Late Hellenistic Epigram

51

shape from trimeters to elegiacs.³¹ In this context, Athena’s and Hera’s declaration that they will no longer enter into a competition regarding μορφή can also be read as a boastful declaration that Antipater’s imitation cannot be matched.

Coming after: Archias reads Antipater of Sidon and Leonidas of Tarentum Antipater’s imitations inspired several younger contemporaries in the late second and early part of the first century BC. I treat here a certain Archias, possibly to be identified with the poet and improvisational versifier Archias of Antioch, to whom several poems that imitate Leonidas and Antipater of Sidon have been ascribed.³² Previous treatments of these imitations have noted Archias’ practice of borrowing diction and structural features from both models and combining them in the creation of his epigram.³³ In his contamination of earlier models, Archias also reveals a sensitivity to the use of dialect as a mode of reading that rivals that found in Antipater, who very likely inspired his younger contemporary from the East.³⁴ In Archias’ hands, as we shall see, dialect becomes a means of

 For this final point, I owe a debt of gratitude to the comments of Peter Bing.  The Greek Anthology transmits epigrams ascribed to five different poets named Archias: Archias of Mytilene, Archias of Byzantium, Archias of Macedonia, Archias the Younger, and Archias the Grammarian. It is almost certain, however, that some of these poets did not exist (such as Archias of Macedonia; see Stadtmüller and Law , ) or are alternative names for another of the Archiai (Law ,  proposed Archias the Younger and Archias the Grammarian). In addition to these poems, there are another twenty-one simply ascribed to an Archias. Law  and Gow and Page , . –  have both attempted to distribute the epigrams transmitted under the name of Archias among the various poets. Law ,  –  argues that these imitations of Leonidas and Antipater of Sidon be ascribed to Archias of Byzantium, since the Corrector has ascribed one of these poems to the Byzantine (the lemma to Planudes has only ᾿Aρχίου). Gow and Page , . and Gutzwiller ,  and n.  claim these epigrams are the work of Archias of Antioch, although Gutzwiller departs from the Gow and Page’s assumption that these epigrams were included in Meleager’s Garland, since not one imitation can be securely placed in a Meleagrean sequence.  For discussion of Archias’ method of imitation, see Law ,  –  and Gutzwiller ,  – .  If we should identify this Archias with Archias of Antioch, Cicero speaks of the latter in similar terms to Antipater of Sidon (Pro Archia ): Quotiens ego hunc vidi, cum litteram scripsisset nullam, magnum numerum optimorum versuum de eis ipsis rebus, quae tum agerentur, dicere ex tempore! Quotiens revocatum eandem rem commutatis verbis atque sententiis! Likely influenced by Cicero, Quintilian, in a discussion of improvisational versifiers, mentions Antipater and Archias in the same breath (..).

52

Taylor Coughlan

signaling from which model he adopted a certain word or phrase and, more broadly, how he understood his own position in the imitative series.³⁵ The four variations of the hunting-net dedications by Leonidas and Antipater of Sidon, which were treated separately above, are exemplative of Archias’ creative combination of dialect in the close-reading of multiple models. I print the poems of Leonidas and Antipater and then the four imitations of Archias. οἱ τρισσοί τοι ταῦτα τὰ δίκτυα θῆκαν ὅμαιμοι, ἀγρότα Πάν, ἄλλης ἄλλος ἀπ᾽ ἀγρεσίης, ὧν ἀπὸ μὲν πτανῶν Πίγρης τάδε, ταῦτα δὲ Δᾶμις τετραπόδων, Κλείτωρ δ᾽ ὁ τρίτος εἰναλίων. ἀνθ᾽ ὧν τῷ μὲν πέμπε δι᾽ ἠέρος εὔστοχον ἄγρην, τῷ δὲ διὰ δρυμῶν, τῷ δὲ δι᾽ ἠιόνων. 3 πτανῶν P: πτηνῶν Pl The three brothers dedicated these nets to you, Pan the Hunter, each from a different chase; Pigres these from the birds, Damis these from the beasts, Kleitor, the third one, from the fish. In return for these send a successful hunt through the air, Through the woods, and through the waters. Leonidas 46 HE=AP 6.13 Πανὶ τάδ᾽ αὔθαιμοι τρισσοὶ θέσαν ἄρμενα τέχνας· Δᾶμις μὲν θηρῶν ἄρκυν ὀρειονόμων, Κλείτωρ δὲ πλωτῶν τάδε δίκτυα, τῶν δὲ πετανῶν ἄρρηκτον Πίγρης τάνδε δεραιοπέδην. τὸν μὲν γὰρ ξυλόχων, τὸν δ᾽ ἠέρος, ὅν δ᾽ ἀπὸ λίμνης οὔ ποτε σὺν κενεοῖς οἶκος ἔδεκτο λίνοις. 1 τέχνας PSud. s.v. ὀρειονόμων et ἄρμενα: τέχνης PlSud. s.v. αὔθαιμοι 3 τῶν δε Pl: τὰν δὲ P, τάνδε Sud. πετανῶν C: πετηνῶν PSud., πετεινῶν Pl 4 τάνδε PSud.: τᾶνδε C, τήνδε Pl λίμνης PPl: λίμναις C, fort. λίμνας Three brothers dedicated these tools of the trade to Pan: Damis a net for mountain-ranging beasts, Kleitor lattice nets for fish, and Pigres this unbreakable net, a collar for birds.

 Law ,  –  and Gutzwiller ,  –  do not comment on dialect in their discussion of Archian imitations. The former () does, however, note the dialectal mixture in a number of poems as another roadblock in establishing criteria to distinguish authorship between the various Archiai. Gow and Page , . briefly note the mixture of Doric and Ionic forms transmitted by P in  GP=AP ., another epitaph for Ajax.

Dialect and Imitation in Late Hellenistic Epigram

53

Not once did their household received them with empty nets, the one from the thickets, the second from the air, and the third from the sea. Antipater 1 HE=AP 6.14 Σοὶ τάδε, Πὰν σκοπιῆτα, παναίολα δῶρα σύναιμοι τρίζυγες ἐκ τρισσῆς θέντο λινοστασίης, δίκτυα μὲν Δᾶμις θηρῶν, Πίγρης δὲ πετηνῶν λαιμοπέδας, Κλείτωρ δ’ εἰναλίφοιτα λίνα. ὧν τὸν μὲν καὶ ἐσαῦθις ἐν ἠέρι, τὸν δ’ ἔτι θείης εὔστοχον ἐν πόντῳ, τὸν δὲ κατὰ δρυόχους. 2 τρισσῆς PlSud.: τρισσᾶς P

3 πετηνῶν P: πετεινῶν Pl

To you, uplands Pan, these manifold gifts did the kinsmen dedicate, three from three types of net hunt, Damis nets for catching beasts, Pigres nooses for fowl, Kleitor cast-nets that roam over the sea. Grant again that the one be successful in the air, another on the sea, and the third among the underbrush. Archias 4 GP=AP 6.16 ᾿Aγραύλῳ τάδε Πανὶ βιαρκέος ἄλλος ἀπ’ ἄλλης αὔθαιμοι τρισσοὶ δῶρα λινοστασίης, Πίγρης μὲν δειραχθὲς ἐύβροχον ἅμμα πετανῶν, Δᾶμις δ’ ὑλονόμων δίκτυα τετραπόδων, ἄρκυν δ’ εἰναλίων Κλείτωρ πόρεν· οἷς σὺ δι’ αἴθρας καὶ πελάγευς καὶ γᾶς εὔστοχα πέμπε λίνα. 3 πετανῶν P: πετηνῶν PlacSudVMac, πετεινῶν PlpcSudrel 5 αἴθρας P: αἴθρης Pl To rustic Pan, these gifts three kinsmen dedicated, each one from a different life-giving net-hunting, Pigres the heavy, neatly knotted noose for birds, Damis the nets for wood-dwelling beasts, Kleitor the reticulated net for the fish; to them deliver successful hunts through the air, sea, and land. Archias 5 GP =AP 6.179 Ταῦτά σοι ἔκ τ’ ὀρέων ἔκ τ’ αἰθέρος ἔκ τε θαλάσσας τρεῖς γνωτοὶ τέχνας σύμβολα, Πάν, ἔθεσαν· ταῦτα μὲν εἰναλίων Κλείτωρ λίνα, κεῖνα δὲ Πίγρης οἰωνῶν, Δᾶμις τὰ τρίτα τετραπόδων· οἷς ἅμα χερσαίᾳσιν, ἅμ’ ἠερίῃσιν ἐν ἄγραις, ᾿Aγρεῦ, ἅμ’ ἐν πλωταῖς, ὡς πρίν, ἀρωγὸς ἴθι. 1 θαλάσσας CPl: θαλάσσης P

2 τέχνας P: τέχνης Pl 5 χερσαίᾳσιν P: χερσαίῃσιν Pl

From the mountains, sky, and sea, these tokens of their skill three brothers, Pan, dedicated to you; Kleitor these nets for fish, Pigres those for birds Damis a third type for beasts;

54

Taylor Coughlan

Be, as you were before, their guide, Hunter, in hunts on land, air, and sea. Archias 6 GP =AP 6.180 Τρίζυγες, οὐρεσίοικε, κασίγνητοι τάδε τέχνας ἄλλος ἀπ’ ἀλλοίας σοὶ †τάδε†, Πάν, ἔθεσαν· καὶ τὰ μὲν ὀρνίθων Πίγρης, τὰ δὲ δίκτυα θηρῶν Δᾶμις, ὁ δὲ Κλείτωρ εἰναλίων ἔπορεν. τῶν ὁ μὲν ἐν ξυλόχοισιν, ὁ δ’ ἠερίῃσιν ἐν ἄγραις αἰέν, ὁ δ’ ἐν πελάγει εὔστοχον ἄρκυν ἔχοι. 5: ἠερίηισιν Pl: ἠερίοισιν P Three brothers dedicated to you, mountain-dwelling Pan, these gifts, each one from a different skill Pigres nets for birds, Damis nets for beasts, Kleitor nets for fish; Of the three, may the one gain success at the hunt in the brambles, the other in the air, and the third on the sea. Archias 7 GP =AP 6.181

In this series of epigrams, Archias displays a studied ability to balance internal variation of his own versions of the dedication with the reuse of diction and phrasing from his two models. It has been observed that Antipater achieves variety in his sequence by using four different words for brother (αὔθαιμοι, σύναιμοι, γνωτοί, and κασίγνητοι) and shuffling the order of the dedications.³⁶ In 7 GP Archias retains the order in Leonidas, the original model, and elsewhere he appropriates the language of Leonidas and Antipater, e. g. ἄλλος ἀπ’ ἄλλης (5.1 GP)≈ ἄλλης ἄλλος ἀπ᾽ (Leonidas 46.2 HE) or εὔστοχα…λίνα (5.6 GP) from εὔστοχον (Leonidas 46.5 HE) and λίνοις (Antipater 1.6 HE).³⁷ These patterns of variation also reveal themselves in Archias’ dialect choice. In terms of internal repetition and variation, note the repetition of Ionic λινοστασίης in the same pentameter-end position in 4.2 GP and 5.2 GP. λινοστασία is a rare word and appears in previous literature only at AP 7.448=Leonidas 12 HE and AP 9.76=Antipater of Thessalonica 30 GP, a poem that should likely be assigned to the Sidonian.³⁸ In addition to possibly evoking

 Gutzwiller , : 4 GP: Damis—Pigres—Kleitor 5 GP: Pigres—Damis—Kleitor 6 GP: Kleitor—Pigres—Damis 7 GP: Pigres—Damis—Kleitor  Gutzwiller ,  –  with further examples and discussion.  On the ascription of this epigram to Antipater of Sidon, see Argentieri ,  – .

Dialect and Imitation in Late Hellenistic Epigram

55

its use in other epigrams of Archias’ two models, λινοστασίης adds further precision, almost akin to a gloss, to Ionic ἀγρεσίης in Leonidas 46.2 HE in whose metrical place it stands in the imitations. Similarly, 4 GP (ἠέρι), 6 GP (ἠερίῃσιν) and 7 GP (ἠερίῃσιν) each contain a form of the epic-Ionic ἠέρ present in the poems of Leonidas (ἠέρος) and Antipater (ἠέρος), and in each instance the form appears in roughly the same metrical position in the final hexameter of the epigram.³⁹ In 6.2 GP (τέχνας σύμβολα) Archias guides the reader to the poem of Antipater, in particular, and in doing so thematizes the epigonal position of his epigrams. The introduction of Doric τέχνας (1.1 HE: ἄρμενα τέχνας) was an innovation of Antipater and one that, I argued, was meant to recall a quirk of Leonidean diction. With τέχνας σύμβολα Archias almost certainly takes up Antipater’s periphrasis for Leonidas’ τὰ δίκτυα, here achieving further variety through the substitution of σύμβολα for ἄρμενα and the reversal of the structure of the phrase.⁴⁰ The choice of σύμβολα is significant in its own right, since it demonstrates an understanding on the part of Archias of his position in the process of replication. As I showed above, Antipater’s use of the less precise ἄρμενα for δίκτυα pointed to the imitative position of the epigram, since those who recognize the line as an imitation already know the dedicated objects are going to be nets. Archias takes this acknowledgement of the distance between model and imitation a step further in his substitution of σύμβολα (“token”)—in other words a (linguistic) place-

 There are other significant dialectal variations possibly present in Archias, but disagreements between the manuscripts hinder our certainty about the exact distribution of the forms. In one instance a pattern of dialectal variance can be glimpsed, namely the forms of πτηνός in Leonidas  HE, Antipater  HE, and Archias  GP and  GP. Above I discussed the Doric πτανῶν (πτηνῶν Pl) in Leonidas  HE, which being the only Doric form transmitted by either witness in the epigram is the lectio difficilior and thus slightly more likely to be original to the text, in relation to the mixed tradition of Antipater  HE (πτανῶν C: πτηνῶν PSud, πετεινῶν Pl). I provisionally printed the reading of C. In Archias  GP (πτηνῶν P: πτεινῶν Pl) and  GP (πετανῶν P: πτηνῶν PlSud) we find a similar variance in the readings of P to the poems of Leonidas and Antipater, except in this instance the Attic-Ionic and Doric forms appear in dialectal contexts that better accommodate these forms. Although an exact picture of the dialectal color of the word in each imitation is not available, it is apparent that the variance is part of the imitative project of at least some of the epigrammatists.  I print, along with Gow and Page, P’s reading of τέχνας (τέχνης Pl). The repetition of the dialect form lends further support to retaining τέχνας in Antipater  HE. While the reading of Pl is almost certainly a regularization, such a reversal of the dialect of Antipater coupled with Archias’ substitution of σύμβολα for ἄρμενα would not significantly detract from my overall point in this reading.

56

Taylor Coughlan

holder for another item, in this case Antipater’s ἄρμενα (“implements”), which is more contextually appropriate.

Poetic and editorial receptions of dialect in Meleager Meleager in both his role as poet and editor offers a final example that allows us to examine the way in which dialect is used as a device to express one’s reading and reception of a model poem. Meleager is not known for his close imitation of model texts; rather he often situates an individual poem within a tradition of epigrammatic composition through the contamination of diction, phrases, and motifs drawn from a string of earlier models, most commonly Asclepiades, Posidippus, and Callimachus, and conventional poetic themes.⁴¹ Although Meleager does not often follow the type of close imitation present in his older contemporary Antipater of Sidon, he occasionally practices the Sidonian’s use of dialect to recall a key word or phrase in a model epigram. Indeed, Meleager recalls Antipater’s reading of model texts through the dialect he uses in his epitaph for his fellow epigrammatist. In HE 122=AP 7.428, Meleager commemorates Antipater in a masterful variation of the Sidonian’s own sequence of riddle epitaphs, incorporating epitaphic imagery (such as the rooster, eagle, and scepter) from his model while also expanding the riddles by using three rather than a single rejected interpretation of said imagery as a guide to the deceased’s identity.⁴² Meleager composed his homage to Antipater’s imitative practice in a predominant Doric color, a choice which makes little sense in the context of Antipater’s ethnic identity, since as a Hellenized Syrian Antipater’s native dialect would almost certainly have been the koine. The use of Doric becomes readily apparent, however, when read in the context of Meleager’s imitation and variation of Antipater’s riddling epitaphs. It has been recognized that the opening words of Meleager’s epitaph (ἁ στάλα) copy those in Antipater of Sidon HE 32=AP 7.427 (ἁ στάλα), one of Antipater’s riddling epitaphs that forms the source material for Meleager’s variation, and which also directly pre-

 On Meleager’s reuse and variation of his epigrammatic predecessors, see commentary on individual epigrams of Meleager in Gow and Page  and epigrams of Asclepaides in Sens ; Tarán , especially  –  and  – .  For a detailed treatment of Meleager various reworkings of Antipater’s riddle epitaphs, see Gutzwiller ,  –  as well as Goldhill ,  – .

Dialect and Imitation in Late Hellenistic Epigram

57

cedes Meleager’s epigram in the Palatine Anthology.⁴³ No one has commented, however, on the fact that Meleager also adopts the dialect of his model, which he then carries through the rest of his inventive variation. Meleager’s dialect choice is, then, its own marker of identity, just of poetic rather than ethnic identity. Like Antipater, Meleager contextualizes his close-reading of his model through the lens of dialect choice. Also like Antipater, Meleager extended his dialect readings to third-century epigrammatists. Meleager 113 HE=AP 12.49 is an exhortation to drink as a remedy for the ills and pains of love: ζωροπότει, δύσερως, καί σου φλόγα τὰν φιλόπαιδα κοιμάσει λάθας δωροδότας Βρόμιος. ζωροπότει, καὶ πλῆρες ἀφυσσάμενος σκύφος οἴνας ἔκκρουσον στυγερὰν ἐκ κραδίας ὀδύναν. Drink pure wine, love-wracked man, and Bromios, the provider of forgetfulness, will calm your boy-loving flame. Drink pure wine, and helping yourself to a cup filled to the brim with wine sweep away the hateful pain from your heart.

Linguistically the epigram is marked by its Doric color and the incorporation of rare or unique compounds. One of these compounds, in particular, holds the interpretive key to understanding the literary purpose behind Meleager’s use of Doric. The adjective φιλόπαιδα in line one is uncommon and is almost certainly borrowed here from Callimachus 3 HE=AP 12.150. In this epigram Callimachus claims to a certain Philip, his addressee, that there are two remedies for love: to the curative powers of song, as discovered by the Cyclops Polyphemus (Πολύφαμος ἀνεύρατο τὰν ἐπαοιδάν | τὠραμένῳ), should be added hunger (χἀ λιμὸς ἔχει μόνον ἐς τὰ πονηρά | τὠγαθόν). Hunger has the potential to bring an end to the “boy-loving sickness” (ἐκκόπτει τὰν φιλόπαιδα νόσον). As we can see from the quoted portions of AP 12.150, Callimachus’ epigram is also composed in a marked and consistent Doric color. Although chronological uncertainties remain, it is quite likely that the Doric has its basis in the verses’ engagement with Theocritus Idyll 11, a bucolic idyll whose dialect is clearly influenced by generic and regional (Polyphemus is a native of Sicily) associations. In addition to the dialect, Callimachus has also borrowed Theocritus’ image of Polyphemus, the use of an abundance of medical language, and the inclusion of an addressee (Theocritus frames his Idyll with an address to the doctor Nicias,

 Gutzwiller ,  – .

58

Taylor Coughlan

whom he may have met on Cos).⁴⁴ Through his use of Doric and τὰν φιλόπαιδα in particular, Meleager guides the reader to the context in which they should read his own epigram on the remedies of love, whose recommendation of drink over song or hunger is a riposte to Callimachus and, through his epigram, ultimately Theocritus. Ever fond of self-variation, Meleager reuses φλόγα τὰν φιλόπαιδα in another of his epigrams (86 HE=AP 12.81): ψυχαπάται δυσέρωτες, ὅσοι φλόγα τὰν φιλόπαιδα οἴδατε τοῦ πικροῦ γευσάμενοι μέλιτος, ψυχρὸν ὕδωρ †νίψαι† ψυχρὸν τάχος ἄρτι τακείσης ἐκ χιόνος τῇ ᾽μῇ χεῖτε περὶ κραδίῃ. ἦ γὰρ ἰδεῖν ἔτλην Διονύσιον· ἀλλ᾽, ὁμόδουλοι, πρὶν ψαῦσαι σπλάγχνων πῦρ ἀπ᾽ ἐμεῦ σβέσατε. 1 τὰν Sylbury: τᾶν P, τὴν Brunck Love-wracked soul-deceivers, you who know the boy-loving flame and have tasted its bitter-sweet honey, ice-cold water…ice-cold from melted snow swiftly now pour about my heart. For I have dared to look upon Dionysios; but, fellow-slaves, quench the fire in me before it reaches my innards.

That 113 HE and not 86 HE is the model example of this phrase is plain from the context. Here φλόγα τὰν φιλόπαιδα is the only Doric color present in the manuscript witnesses of the epigram. It is surrounded by koine with the admixture of Ionic forms (cf. κραδίῃ and likely ἐμεῦ, which could also be categorized as Doric). Such a dialectal context points to a borrowing of the phrase from its original context, which as we saw was entirely Doric. Highlighted by its dialectal difference from the rest of the epigram φλόγα τὰν φιλόπαιδα is present in order to direct the reader to 113 HE, on the curative powers of drink, to which the conceit of this epigram is in direct opposition: Dionysius only exacerbates love rather than soothes it.⁴⁵ Again, dialect paired with diction takes up a model to which the epigram responds. The editorial techniques Meleager used in his Garland, their literary purposes, and their reception and reuse in later poems and poetry collections have re-

 See Hunter ,  and Hordern , .  Indeed the entire first line of the epigram recalls . HE (ζωροπότει, δύσερως, καί σου φλόγα τὰν φιλόπαιδα) with δυσέρωτες for δύσερως and ψυχαπάται approximating the sound patterns of ζωροπότει in the same metrical sedes.

Dialect and Imitation in Late Hellenistic Epigram

59

ceived a good deal of scholarly attention;⁴⁶ however the organizational structure of these sequences have not been subjected to an analysis that takes continuities or shifts in dialect into account. In what follows, I offer two examples that demonstrate that future work on Garland sequences should consider dialect as contributing factor to their organization and interpretation. I begin with a sequence of four epigrams in the twelfth book of the Greek Anthology whose topic is homoerotic love. As has long been recognized, the division of erotic epigrams between heterosexual and homosexual subjects—books five and twelve respectively—is the work of Cephalas.⁴⁷ In the Garland, all the erotic epigrams were found mixed together in one book into which Meleager incorporated a number of his own epigrams. AP 12.75 – 78 are variations on a theme, in which the poet compares the beauty of his male subject to that of the god Eros; they almost certainly derive intact from the Garland. Meleager alternates a third-century model epigram with one of his own variations: Εἰ πτερά σοι προσέκειτο καὶ ἐν χερὶ τόξα καὶ ἰοί, οὐκ ἂν Ἔρως ἐγράφη Κύπριδος ἀλλὰ σύ παῖς. If wings were applied to you and there was a bow and arrows in your hand, Eros would not have been written down as the child of Cypris, but you. Asclepiades 21 HE=AP 12.75 Εἰ μὴ τόξον Ἔρως μηδὲ πτερὰ μηδὲ φαρέτραν μηδὲ πυριβλήτους εἶχε πόθων ἀκίδας, οὐκ, αὐτὸν τὸν πτανὸν ἐπόμνυμαι, οὔποτ’ ἂν ἔγνως ἐκ μορφᾶς τίς ἔφυ Ζωίλος ἢ τίς Ἔρως. 1 φαρέτραν P: φαρέτρην ABV If Eros did not have a bow and wings and a quiver and fevered arrows of desire, I would not in anyway have known, and I swear by the winged boy himself, from shape, who was Zoilos or who was Eros. Meleager 89 HE=AP 12.76

 Important discussions include: Wifstrand ; Cameron ,  – ; Gutzwiller ; Gutzwiller ,  – ; and Höschele ,  – .  For discussion, see Wifstrand , ; Gow , ; Lenzinger ,  – ; and Cameron ,  – . Cephalas’ division is hardly systematic with homosexual poems appearing in Book  and, to a greater degree, heterosexual poems in Book , from which Gutzwiller ,  surmises that “Cephalas evidently did not understand that names, or nicknames, ending in –ιον were given to women.”

60

Taylor Coughlan

Εἰ καθύπερθε λάβοις χρύσεα πτερὰ, καί σευ ἀπ’ ὤμων τείνοιτ’ ἀργυρέων ἰοδόκος φαρέτρη καὶ σταίης παρ’ Ἔρωτα, φιλάγλαον, οὐ μὰ τὸν Ἑρμῆν⁴⁸, οὐδ’ αὐτὴ Κύπρις γνώσεται ὃν τέτοκε. 3 Ἑρμῆν PABV: Ἑρμᾶν BKT 5.1.75 If you should have golden wings above, and from your silver shoulders an arrow-bearing quiver should hang and you should stand alongside resplendent Eros, by Hermes, not even Cypris herself will know whom she bore. Asclepiades 38 HE=AP 12.77⁴⁹ Εἰ χλαμύδ’ εἶχεν Ἔρως καὶ μὴ πτερὰ μηδ’ ἐπὶ νώτων τόξα τε καὶ φαρέτραν, ἀλλ’ ἐφόρει πέτασον, ναὶ ⟨μὰ⟩ τὸν ἁβρὸν ἔφηβον ἐπόμνυμαι, ᾿Aντίοχος⁵⁰ μέν ἦν ἂν Ἔρως, ὁ δ’ Ἔρως τἄμπαλιν ᾿Aντίοχος. If Eros had a cloak and did not have wings or on his back a bow and quiver, but was wearing a broad-brimed hat, indeed I swear by the graceful youth, Antiochos would be Eros, and Eros in turn would be Antiochos. Meleager 83 HE=AP 12.78

 There is great difficulty in knowing the best text between the Attic-Ionic Ἑρμῆν in the manuscripts (P and ABV) and the Doric in the papyrus. Following my general principles of giving more weight both to papyrus readings over manuscript readings (in the small number of instances where such a comparison fruitfully exists) and to transmission of Doric forms non-Doric or dialectally mixed contexts, Doric Ἑρμᾶν would be the preferred form. That said, it is difficult to understand why Asclepiades would insert a Doric form of the god’s name here, although such a practice is not without parallel: AP .=Theodorus HE  and AP .=Leonidas of Tarentum HE .  The lemmata to P and the ABV both ascribed authorship alternatively to Asclepiades and Posidippus and the lemma to BKT .. is not preserved. There is nothing on stylisic or metrical grounds to favor the authorship of one poet over another. Decisions of ascription come down to the belief one holds whether Asclepiades would have composed a self-variation (cf.  ΗΕ=AP .) or whether such a variation is likely to be taken as a sign of Posidippan authorship, who is known to imitate and vary other poems by Asclepiades. Guichard ,  – , who collects the previous bibliography on the question, and Sens ,  each conclude that Posidippus is more likely to be the author of this epigram.  Here and again in the following line P and P.Berol.  disagree on the name of the epigram’s erotic subject with the former transmitting ᾿Aντίοχος and the latter transmitting ᾿Aντιγένης. On the one hand, P’s ᾿Aντίοχος has strong Syrian associations through its connection to the name of Seleucid dynasts, which would make sense for a poem by the Syrian Meleager. On the other hand, ᾿Aντιγένης in P.Berol.  is the name of a Coan aristocrat in Theocritus Id. , and if ᾿Aντιγένης is original to the poem its presence may help the reader to understand the φαρέτραν in line two as Doric.

Dialect and Imitation in Late Hellenistic Epigram

61

Meleager’s imitations hew closely to the structure and diction of his models (cf. πτερὰ in AP 12.75.1, 12.76.1, 12.77.1 and 12.78.1), but they differ in their dialectal color (cf. Doric φαρέτραν in AP 12.76.1 and likely the same in AP 12.78.2, but the absence of any other Doric features here does not allow for a certain identification, whereas Ionic φαρέτρη at AP 12.77.2). So while this sequence is a cohesive whole in terms of subject matter, Meleager demarcates for his readers the boundaries between model and imitation through the alternation of dialects. If Meleager identified the authorship of AP 12.77 with Asclepiades, and this is uncertain given various arguments that the double ascriptions of Asclepiadean and Posidippan authorship ultimately derive from the Garland,⁵¹ the alternation would also bring the voices of the two poets into contrast. The two examples of Asclepiades capture the dialectal color of the majority of his erotic epigrams as well as approximating the native speech of the Samian poet, while Meleager adopted a Doric tinged language which appears prominently elsewhere in his epigrams in realtion to his adopted city of Cos.⁵² On the same model, Posidippus also wrote the majority of his surviving erotic epigrams in koine with Ionic admixture (indeed he was likely influenced by the Samian in his dialect choice as he was in the subject matter and style of his erotic epigrams). In this case, then, Meleager, chose to imitate a model poem and its own earlier imitation. Collectively, Meleager’s decision to write his imitations in a dialect that approximates that spoken on Cos grounds the reading of the sequence in the location where he likely arranged and edited the Garland. In another sequence of erotic epigrams from Book Five of the Greek Anthology, Meleager intertwines two conventional motifs—garlands and the Graces (AP 5.142– 49).⁵³ AP 5.142– 45 are all on garlands and share, with several notable exceptions, a similar koine color. Τίς—ῥόδον ὁ στέφανος Διονυσίου, ἢ ῥόδον αὐτός τοῦ στεφάνου; δοκέω, λείπεται ὁ στέφανος. Who—is the rose the garland of Dionysios or is he the rose of the garland? I believe that the garland is lacking. Anon. 23 HE=AP 5.142

 See Gow ,  – ; and Gow and Page , .xxx.  On Meleager’s use of Doric in relation to Coan identity, see Gutzwiller .  Gutzwiller ,  considers this sequence as the third part of the opening of Meleager’s book of erotic epigrams. Gutzwiller ,  –  treats this sequence in detail, paying particular attention to its organization and programmatic importance in Meleager’s erotic book; for further discussion see Höschele ,  – . For one possible reconstruction of Meleager’s book of erotic epigrams see Gutzwiller ,  –  and table II.

62

Taylor Coughlan

Ὁ στέφανος περὶ κρατὶ μαραίνεται Ἡλιοδώρας, αὐτὴ δ’ ἐκλάμπει τοῦ στεφάνου στέφανος. The garland withers on Heliodora’s head; but she shines forth as the garland’s garland. Meleager 45 HE=AP 5.143 Ἤδη λευκόιον θάλλει, θάλλει δὲ φίλομβρος νάρκισσος, θάλλει δ’ οὐρεσίφοιτα κρίνα. ἤδη δ’ ἡ φιλέραστος, ἐν ἄνθεσιν ὥριμον ἄνθος, Ζηνοφίλα Πειθοῦς ἡδὺ τέθηλε ῥόδον. λειμῶνες, τί μάταια κόμαις ἔπι φαιδρὰ γελᾶτε; ἁ γὰρ παῖς κρέσσων ἁδυπνόων στεφάνων. 3 ἡ φιλέραστος PPl: fort. ἁ φιλέραστος 6 ἁ P: ἡ Pl κρέσσων P: κρείσσων Pl ἁδυπνόων P: ἡδυπνόων Pl The snowdrop is already in bloom, so is the rain-loving narcissus, and the mountain-roaming lilies. Already dear to lovers, the flower timely among flowers, Zenophila, the sweet rose of Persuasion, has bloomed. Meadows, why do you foolishly smile at the horse-tail in your hair? For the girl is better than sweetly-smelling garlands. Meleager 31 HE=AP 5.144 Αὐτοῦ μοι στέφανοι παρὰ δικλίσι ταῖσδε κρεμαστοί μίμνετε μὴ προπετῶς φύλλα τινασσόμενοι, οὓς δακρύοις κατέβρεξα· κάτομβρα γὰρ ὄμματ’ ἐρώντων· ἀλλ’ ὅταν οἰγομένης αὐτὸν ἴδητε θύρης, στάξαθ’ ὑπὲρ κεφαλῆς ἐμὸν ὑετόν, ὡς ἂν ἄμεινον ἡ ξανθή γε κόμη τἀμὰ πίῃ δάκρυα. For my sake, garlands, stay here hung up by these doors, don’t hastily shake off your petals, you that I drench with my tears; for the eyes of lovers are rainclouds. But whenever the doors open and you should see him, drip my rain over his head, so that his yellow hair may better drink my tears. Asclepiades 12 HE=AP 5.145

The sequence begins with two distichs, one by an unknown poet and the other by Meleager, whose joint conceit is that the garlanded beloved (Dionysios and Heliodora, respectively) surpasses that garland in beauty. The topic is expanded upon in AP 5.144, another epigram by Meleager; this time it is the enchanting bloom of Zenophila’s beauty, another of the poet’s female lovers. 5.145, Asclepiades’ epigram on the tear-drenched wreaths hung by a lover over the closed door of the object of attention, shares with the preceding poem by Meleager an identical length and the personification of natural phenomena (smiling

Dialect and Imitation in Late Hellenistic Epigram

63

meadows in Meleager and a weeping garland in Asclepiades),⁵⁴ but it also functions as a bridge epigram to the next half of the sequence and we shall return to it shortly. Doric color is limited to the two epigrams by Meleager, whose subjects —Heliodora and Zenophila—are Coan women featured in erotic cycles from the Gadaran’s poetry.⁵⁵ Besides fittingly using Doric color in the name of Zenophila (Ζηνοφίλα and the nominative feminine articles ἁ at lines three and six) and quite likely also in the name of Heliodora (Ἡλιοδώρας), although the form is ambiguously Doric or Attic, Meleager reserves the use of Doric to the final line of AP 5.144 in ἁδυπνόων. The phrase ἁδυπνόων στεφάνων echoes the final line of Meleager’s proem to the Garland (ἡδυεπὴς στέφανος) and appears in the same metrical position at the end of the pentameter.⁵⁶ Intratextual reminiscences of the proem to the Garland are not limited to this phrase. This listing of flowers echoes Meleager’s catalogue of plants representing various poets that comprises the majority of the proem to his anthology. In particular, it has been observed that by placing the snowdrop first in AP 5.144, Meleager reverses its position from the proem (AP 4.1.56), where it is used to represent the Gadaran’s own poetry.⁵⁷ Through this positioning of λευκόιον, Meleager further draws attention to his allusion to the very end of his proem at the conclusion of AP 5.144. By interweaving this epigram on Zenophila with echoes of the proem to the Garland, Meleager constructs, as Gutzwiller has noted, an “association between the ‘sweet-smelling garlands’ surpassed by Zenophila and the ‘sweet-speaking garland’ that is Meleager’s epigram anthology.”⁵⁸ The surpassing allure of Zenophila finds expression in the doricization of ἁδυπνόων στεφάνων: in this eroticized miniaturization of the Garland proem, the Coan’s beauty is so overwhelming that it can outshine the Garland in which it is interwoven. At AP 5.146=Callimachus 15 HE on the beauty of Berenike, the motif shifts from garlands to Graces and with it so does the predominant dialectal color of the first epigram on the topic.⁵⁹

 See Gutzwiller ,  on the “confounding of nature’s phenomena with human emotion” as the connection between . and .; she does not note how the sequence also divides into two pairs of epigrams of identical length.  On the interpretation of Meleager’s epigrams on Heliodora as “ein fragmentierter Liebesroman”, see Höschele ,  – .  As observed by Gutzwiller , , who does not comment on the contrast in dialect between the two appearances.  Gutzwiller , .  Gutzwiller , .  The identity of this Berenike is uncertain. Based on Callimachus’ dates the queen could be Berenike I, Berenike Syria (the daughter of Ptolemy II), or Berenike II. In this volume and elsewhere (Clayman , ) Clayman has argued persuasively on historical and dialectal grounds

64

Taylor Coughlan

τέσσαρες αἱ Χάριτες, ποτὶ γὰρ μία ταῖς τρισὶ κείναις ἄρτι ποτεπλάσθη, κἤτι μύροισι νοτεῖ, εὐαίων ἐν πᾶσιν ἀρίζηλος Βερενίκα, ἇς ἄτερ οὐδ᾽ αὐταὶ ταὶ Χάριτες Χάριτες. Four are the Graces, for in addition to those three one has just now been modeled, and is still moist with perfume. Berenike is fortunate and envied among everyone without whom the Graces are not themselves Graces.

Since the sequences of Meleager’s Garland were excerpted by Cephalas, a process that could lead to the corruption of individual epigrams and the rearrangement or significant dismantling of sequences, it is possible that one or more epigrams may have fallen out at this point in our sequence. As we shall see, however, there are thematic points of contact (garlands and Graces) between AP 5.145 and AP 5.146, where one might plausibly posit missing epigrams or rearrangement, and the epigrams that follow (AP 5.147– 49) as well as a studied contrast of dialect related to the interlocking of the two themes, which together supports the proposition that the epigrams, particularly AP 5.146, are not completely out of place. As transmitted, Callimachus’ Doric epigram on Berenike II and the Graces stands separated from the other epigrams on the same topic that follow, since Meleager returns to the previous motif of garlands in 46 ΗΕ=AP 5.147 with another of his own epigrams on Heliodora. With the return of garlands the reader once again encounters an epigram with a koine/Ionic color: Πλέξω λευκόϊον, πλέξω δ’ ἁπαλὴν ἅμα μύρτοις νάρκισσον, πλέξω καὶ τὰ γελῶντα κρίνα, πλέξω καὶ κρόκον ἡδύν· ἐπιπλέξω δ’ ὑάκινθον πορφυρέην, πλέξω καὶ φιλέραστα ῥόδα, ὡς ἂν ἐπὶ κροτάφοις μυροβοστρύχου Ἡλιοδώρας εὐπλόκαμον χαίτην ἀνθοβολῇ στέφανος. I will pluck the snow-drop, I will pluck the narcissus tender for the myrtle, I will pluck the smiling lilies, and I will pluck the sweet crocus; I will weave purple hyacinth, and I will pluck roses that are dear to lovers, so that the garland may wreath the well-tressed hair on perfumed Heliodora’s temples.

that we should identify the subject of Callimachus’ epigram with Berenike II and I adopt this identification in my discussion.

Dialect and Imitation in Late Hellenistic Epigram

65

This epigram has many points of contact with the language of Meleager 31 HE=AP 5.144 (cf. AP 5.144.1: ἤδη λευκόϊον θάλλει and 5.147.1: πλέξω λευκόϊον)⁶⁰ and so instantly recalls the preceding sequence of epigrams. On a more thematic level, the falling petals of the garland in AP 5.145=Asclepiades HE 12 anticipates the image of the plucked and falling petals in AP 5.147. The return of Heliodora also acts as bridge to the next epigram in which Meleager makes the bold claim that she surpasses the Graces in their natural quality (47 HE=AP 5.148): Φαμί ποτ’ ἐν μύθοις τὰν εὔλαλον Ἡλιοδώραν νικάσειν αὐτὰς τὰς Χάριτας χάρισιν. 1 τὰν CPl: τὴν P I say that Heliodora sweet-chatting in conversations will conquer the Graces themselves with grace.

While the likely doricized Ἡλιοδώραν connects Heliodora’s two garland poems to this epigram on the graces, the consistent Doric color of the epigram evokes the dialect of Callimachus 15 HE=AP 5.146, the first of the epigrams on the topic, and whose concluding αὐταὶ ταὶ Χάριτες Χάριτες plainly echoes in Meleager’s αὐτὰς τὰς Χάριτας χάρισιν. Altogether AP 5.145 – 48 present an interlocking sequence of epigrams, whose organization is reinforced by dialectal variance: Table: Dialect Sequence in AP 5.145 – 148 AP Number

Topic

.=Asclepiades HE  .=Callimachus HE  .=Meleager HE 

Tear-sodden garland Ionic Berenike II likened to a Grace Doric with one Attic-Ionic form A garland for Heliodora Attic-Ionic with light Ionic and Doric color Heliodora likened to a Grace Doric

.=Meleager HE 

Dialect

Thematically and linguistically, then, AP 5.145 and AP 5.146 together provide models for Meleager’s own epigrams that follow. The importance of AP 5.146 on Berenike II to the sequence as a whole is reinforced in the following and final epigram, which is again by Meleager and also has Doric coloring (32 HE=AP 5.149):

 See Gutzwiller ,  –  and Höschele ,  – .

66

Taylor Coughlan

Τίς μοι Ζηνοφίλαν λαλιὰν παρέδειξεν ἑταίρην;⁶¹ τίς μίαν ἐκ τρισσῶν ἤγαγέ μοι Χάριτα; ἦ ῥ’ ἐτύμως ἁνὴρ κεχαρισμένον ἄνυσεν ἔργον δῶρα διδοὺς καὐτὰν τὰν Χάριν ἐν χάριτι. 1 λαλιὰν P: λαλιὴν Pl 2 ἑταίρην PPl: ἑταίραν Brunck 4 καὐτὰν τὰν P: καὐτὴν τὴν Pl Who has represented in painting my chatty mistress Zenopila? Who has led one Grace to me out of the three? Yes, truly that man has accomplished a graceful work, having granted a gift of the Grace herself with grace.

As we saw in the sequence of garland epigrams (AP 5.142– 45), Meleager alternates a poem on Heliodora (AP 5.148) with one on Zenophila (AP 5.149) and here as well the variation includes an expansion of the model. Zenophila (λαλιάν) is equally as chatty as Heliodora (εὔλαλον). And in the final pentameter he again reworks the final line of his two models (καὐτὰν τὰν Χάριν ἐν χάριτι). Meleager also caps Callimachus with whom he began the sequence. In his presentation of Zenophila, Meleager echoes the language of epigrams on works of art, making his beloved into a statue.⁶² But whereas Callimachus imagines a portrait of Berenike added to a statue group of the Graces in order to complete gracefulness of the Graces (Callimachus 15.4 HE=AP 5.146.4: ἇς ἄτερ οὐδ᾽ αὐταὶ ταὶ Χάριτες Χάριτες), Meleager’s artistic likeness of Zenophilia has enough grace to replace all three of the Graces imagined in Callimachus’ epigram (Meleager 32.2 HE=AP 5.149.2: τίς μίαν ἐκ τρισσῶν ἤγαγέ μοι Χάριτα). All of these variations are couched in the same dialect color as the model, linguistically uniting the sequence. The unification of this sequence is made all the more apparent when read within the wider context of its integration with the epigrams on garlands. The two sequences make a cohesive whole, sharing in similar language and subjects, but their cohesiveness is partially a product of their studied variation. And Meleager, in his general division of garland and Graces epigrams by a juxtaposition of Doric and koine/Ionic dialect color, highlights his own reading of the model epigrams and the structure and movement of his editorial decisions that brings together the old and new. This is especially apposite considering that this sequence of epigrams has been recognized to consistently engage with the image of the garland, and thus the process of epigram se-

 I print the reading of the manuscript witnesses, but Brunk’s emendation ἑταίραν (printed in HE) is a strong possibility given the other Doricisms transmitted by P.  Gow and Page , . on παρέδειξεν “‘represent, of a painter’ LSJ; unique but easy extension of the common meaning exhibit” and ἄνυσεν ἔργον “the phrase strongly suggests the work of art.” See translation and discussion of the epigram in Gutzwiller , .

Dialect and Imitation in Late Hellenistic Epigram

67

lection and anthologization,⁶³ and the motif of grace, both of which are intertwined in the opening quatrain of the proem of the Garland (1 HE=AP 4.1).⁶⁴ This evocation of the proem also finds expression in the interlaced dialect present in the sequence, for we find a similar mixture of Doric and Ionic features there.⁶⁵ This sequence of erotic epigrams, which possibly held an introductory and programmatic position in Meleager’s book of erotic epigrams, replicates the dialect practice originally put on display by Meleager in his general introduction to the anthology.⁶⁶

Conclusion This chapter has ventured to establish dialect choice as a literary device used by epigrammatists when they sought to imitate an earlier epigram. Third-century book epigrammatists such as Asclepiades and Leonidas, whose dedications and epitaphs often provided the raw material for the development of a poetics of imitation, employed dialect in order to enrich the poetic meaning of their works. Having recognized the literary quality of these authors’ dialect practice from distant millennia, the question is how did the readers and fellow epigrammatists, who studied and interpreted these epigrams, understand and receive the dialect of their models. Imitations, in other words, can offer an early history of reader-response to dialect. I have shown patterns of dialectal engagement in the imitations of Antipater of Sidon, Archias, and Meleager, which demonstrate that dialect served two main purposes in the process of imitation. First, dialect choice allows these poets to guide a reader through their own reading of a model text,

 See Höschele ,  on Melager  HE=AP .: “Das Flechten des Kranzes (man beachte die sechsfache Anapher [ἐπι‐]πλέξω!) repliziert auf mikrotextueller Ebene die im Prolog (.) beschriebene Tätigkeit des poeta-editor.”  Μοῦσα φίλα, τίνι τάνδε φέρεις πάγκαρπον ἀοιδάν ἤ τίς ὁ καὶ τεύξας ὑμνοθετᾶν στέφανον; ἄνυσε μὲν Μελέαγρος, ἀριζάλῳ δὲ Διοκλεῖ μναμόσυνον ταύταν ἐξεπόνησε χάριν. Additionally note the presence of Doric ἀριζάλῳ which echoes ἀρίζηλος in Callimachus 15 HE=AP 5.146; should the Doric color here give us pause about the transmission of the AtticIonic form by P in Callimachus?  On the dialect of the proem to the Garland, see Gutzwiller ,  – .  See the comments of Gutzwiller ,  that this reading, in part, bears out: “The proem surely sets as template for the use of dialect in Meleager’s own epigrams as in the epigrams of other anthologized poets, where a dialectal mixture occurs within epigrams and sequentially from one epigram to the next.”

68

Taylor Coughlan

both by highlighting points of similarity and points of departure. Second, it also acts as an organizational device for Meleager as he creates sequences of epigram variations and/or sequences of motifs, contrasting model with variation or one motif with another. The imitation of earlier epigrams and their collection into artfully organized anthologies has rightly been taken as evidence for the canonization of the genre of book epigram at the end of the Hellenistic period. That authors such as Antipater of Sidon and Meleager, who were so integral to this process of canonization, time and again turned to the dialect of the epigrams that they imitated or interweaved with their own to signal their relationship to and reading of the model epigrams only underscores the dynamic role of dialect choice in Hellenistic book epigram.

Sigla/Abbreviations AB

Austin, C./Bastianini, G. (eds.) (2002), Posidippi Pellaei quae supersunt omnia, Milan. ABV Appendix Barberino-Vaticana (Paris suppl. gr. 1199; Vat. gr. 240; Vat. Barb. gr. 123). C Corrector codicis P. GP Gow, A./Page, D. (eds.) (1968), The Greek Anthology: The Garland of Philip and Some Contemporary Epigrams, 2 vols., Cambridge. HE Gow, A./Page, D. (eds.) (1965), The Greek Anthology: Hellenistic Epigrams, 2 vol., Cambridge. PMG Page, D. (ed.) (1962), Poetae melici Graeci, Oxford. P Palatinus gr. 23 + Par. suppl. gr. 384. Pl Marc. gr. 481. Q BM add. 16409. S Sylloge Parisina (Par. suppl. gr. 352 et Par. gr. 1630). Σπ Sylloge Σπ (Palatinus gr. 23, foll. AvBrvCrDrv). SH Lloyd-Jones, H./Parsons, P. (eds.) (1983), Supplementum Hellenisticum, Berlin. Stadtmüller Stadtmüller, H. (ed.) (1894– 1906), Anthologia Graeca epigrammatum Palatina cum Planudes, Leipzig. Sud. Suidae Lexicon. SudA=Par. gr. 2625 et 2626 vetus manus. SudF=Laur. 51.1. SudG=Par. gr. 2623. SudM=Marc. gr. 482. SudV=Leid. Voss. gr. F. 2. ac =ante correctum pc =post correctum rel. =reliqui

Dialect and Imitation in Late Hellenistic Epigram

69

Bibliography Argentieri, L. (2003), Gli epigrammi degli Antipatri, Bari. Bartoněk, A. (1972), Classification of the West Greek Dialects at the Time about 350 B.C., Amsterdam. Bettarini, L. (2004), ‘Posidippo e l’epigramma epinicio: aspetti linguistici’, in: M. Di Marco / B. Palumbo Stracca / E. Lelli (eds.), Posidippo e gli altri. Il poeta, il genere, il contesto culturale e letterario, Rome, 9 – 22. Bubeník, V. (1989), Hellenistic and Roman Greece as a Sociolinguistic Area, Amsterdam. Buck, C. (1955), The Greek Dialects, Chicago. Bulloch, A. (ed.) (1985), The Fifth Hymn. Callimachus, Cambridge. Cameron. A. (1993), The Greek Anthology from Meleager to Planudes, Oxford. Clayman, D. (2014), Berenice II and the Golden Age of Ptolemaic Egypt, Oxford. Conte, G.-B. (1986), The Rhetoric of Imitation: Genre and Poetic Memory in Virgil and Other Latin Poets, Translated by C. Segal, Ithaca. Criscuolo, U. (2004), ‘Mimesi e tecnica espressiva in Leonida di Taranto’, in: A. López Eire / A. Ramos Guerreira (eds.), Registros lingüísticos en las lenguas clásicas, Salamanca, 33 – 49. Dilthey, C. (1976), ‘Dipinti Pompeiani accompagnati d’epigrammi Greci’, Annali dell’Instituto di corrispondenza archeologica 48, 294 – 314. Fish, S. (1980), Is There a Text in this Class? The Authority of Interpretive Communities, Cambridge, MA. Floridi, L. (2014), Lucillio, Epigrammi: introduzione, testo critico, traduzione e commento, Berlin. Gigante, M. (1971), L’edera di Leonida, Naples. Goldhill, S. (1994), ‘The Naïve and Knowing Eye: Ecphrasis and the Culture of Viewing in the Hellenistic World’, in: S. Goldhill / R. Osborne (eds.), Art and Text in Ancient Greece, Cambridge, 197 – 223. Gow, A. (1958), The Greek Anthology: Sources and Ascriptions, London. Gow, A. / Page, D. (eds.) (1965), The Greek Anthology: Hellenistic Epigrams, 2 vols., Cambridge. — (eds.) (1968), The Greek Anthology: The Garland of Philip and Some Contemporary Epigrams, 2 vols., Cambridge. Guichard, L. (2004), Asclepíades de Samos: Epigramas y fragmentos, Berne. — (2005), ‘Dialecto y género literario en los epigramas de Posidipo (PMil. Vogl. VIII 309’, in: A.A. Ezquerra (ed.), Actas del XI Congreso Español de Estudios Clásicos, Madrid, 311 – 320. Gutzwiller, K. (1997), ‘The Poetics of Editing in Meleager’s Garland’, in: TAPhA 127, 169 – 200. — (1998), Poetic Garlands: Hellenistic Epigrams in Context, Berkeley. — (2014), ‘Poetic Meaning, Place, and Dialect in the Epigrams of Meleager’, in: R. Hunter / A. Rengakos / E. Sistakou (eds.), Hellenistic Studies at a Crossroads: Exploring Texts, Contexts, and Metatexts, Berlin, 75 – 95. Halliwell, S. (2002), The Aesthetics of Mimesis: Ancient Texts and Modern Problems, Princeton. Hordern, J. (2006), ‘Cyclopea: Philoxenus, Theocritus, Callimachus, Bion’, in: CQ 54, 285 – 92.

70

Taylor Coughlan

Höschele, R. (2010), Die blütenlesende Muse: Poetik und Textualität antiker Epigrammsammlungen, Munich. Hunter, R. (1999), Theocritus: A Selection, Cambridge. Iser, W. (1976), Der Akt des Lesens: Theorie ästhetischer Wirkung, Munich. — (2000), The Range of Interpretation, New York. Laurens, P. (1989), L’abeille dans l’ambre: célébration de l’épigramme, Paris. Law, H. (1936),‘The Poems of Archias in the Greek Anthology’, in: CPh 31, 225 – 43. Lenzinger, F. (1965), Zur griechischen Anthologie, Zurich. Ludwig, W. (1968), ‘Die Kunst der Variation im hellenistiche Liebesepigramm’, in: A. Dihle (ed.), L’épigramme grecque (Entretiens sur l’antiquité classique), Vandœuvres, 297 – 348. Molinos Tejada, T. (1990), Los Dorismos del Corpus Bucolicorum, Amsterdam. Moreno, P. (2001), Apelles: The Alexander Mosaic, Skira. Palumbo Stracca, B. (1987), ‘Differenze dialettali e stilistiche nella storia dell’epigramma greco’, in: G. Bolognesi / V. Pisani (eds.), Linguistica e Filologia (Atti del VII convengo internazionale di linguisti, tenuta a Milano nei giorni 12 – 14 settembre 1984), Brescia, 429 – 34. — (1993 – 94), ‘Note dialettologiche al nuovo Posidippo’, in: Helikon 33 – 34, 405 – 12. Pedrick, V. / N. Rabinowitz. (1986), ‘Introduction’, in: Arethusa 19, 105 – 114. Reiff, A. (1959), ‘Interpretatio, imitation, aemulatio: Begriff und Vorstellung literarischer Abhängigkeit bei den Römern’, Ph.D. Dissertation, University of Cologne. Russell, D.A. (1979), ‘De Imitatione’, in: D. West / T. Woodman (eds.), Creative Imitation and Latin Literature, Cambridge, 1 – 16. Sens, A. (2004), ‘Doricisms in the New and Old Posidippus’, in: B. Acosta-Hughes / E. Kosmetatou / M. Baumbach (eds.), Labored in Papyrus Leaves: Perspectives on an Epigram Collection Attributed to Posidippus (P.Mil.Vogl. VIII 309), Washington, D.C., 65 – 83. Sens, A. (2007), ‘One Thing Leads (Back) to Another: Allusion and the Invention of Tradition in Hellenistic Epigrams’, in: P. Bing / J. Bruss (eds.), Brill’s Companion to Hellenistic Epigram, Leiden, 373 – 390. — (ed.) (2011), Asclepiades of Samos: Epigrams and Fragments, Oxford. Stüber, K. (1996), Zur dialektalen Einheit des Ostionischen, Innsbruck. Tarán, S.L. (1979), The Art of Variation in the Hellenistic Epigram, Leiden. Waltz, P., et al., (eds.) (1929‐), Anthologie grecque, Paris. Wifstrand, A. (1926), ‘Studien zur griechischen Anthologie’, Ph. Dissertation, University of Lund. Ypsilanti, M. (2006), ‘An Aspect of Leonidas’ Reception in Later Epigrammatists and the Art of Variation: The Case of Fishermen’s Epitaphs’, in: CPh 101, 67 – 73.

Lucia Floridi

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD* In what follows, I propose to describe the style and language of the most important representatives of Greek skoptic epigram. I will focus on the poets of the I ‒ II centuries AD – the heyday of the subgenre, which, as far as we can tell, was only occasionally practised by the poets of Meleager’s and Philip’s Garlands,¹ and became prominent in the Neronian Era. Specifically, I will concentrate on Lucillius, Nicarchus, and Ammianos. Not only can these authors, unlike others preserved in book XI of the Greek Anthology, be dated with a certain degree of confidence,² but their epigrammatic corpora are big enough to allow a comprehensive analysis of each poet’s style and diction.³ The epigrams ascribed to Lucian are excluded from this survey, for the problems of ascription and authorship would require a study beyond the purposes of this paper. I hope, however, to pursue such a study in the near future, in my planned commented edition of these poems.⁴ The aim of the present survey is twofold: on the one hand, I intend to demonstrate that a better understanding of the style of these authors can help in the exegesis of their poems, and sometimes lead to textual improvements in modern critical editions.⁵ Further, I wish to show that, in spite of the stylistic differences

* This paper profited from the advice of Emilia Barbiero (who also helped me to improve the English), Regina Höschele, Francesco Pelliccio, and Francesco Valerio, and from discussion with other participants at the conference.  It is entirely possible that the scarcity of surviving skoptic epigrams before the Neronian Era is due to Meleager’s and Philip’s editorial choices: see e. g. Floridi , ‒; this assumption seems now confirmed by the publication of the ‘Vienna Epigrams Papyrus’ (Parsons-MaehlerMaltomini ): see Floridi-Maltomini ; on the development of Greek skoptic epigram, see especially Blomqvist .  For Lucillius’s chronology, see Floridi , ‒; for Nicarchus’s, Schatzmann , ‒ (with my observations in Floridi , in press); for Ammianos’s, Schulte , .  There are  epigrams by Lucillius labelled as authentic in my edition;  (plus some minor fragments) are ascribed to Nicarchus in Schatzmann’s edition (where the poems of POxy  and –, edited by Parsons in  and  respectively, are included);  are ascribed to Ammianos by Schulte .  On Lucian’s epigrams, see especially Baldwin ; Floridi , ‒, with further bibliography.  It is interesting to note that from the first editions of these authors onward there has been a progressive ‘decrease’ in the numbers of emendations proposed by scholars: while the first mod-

72

Lucia Floridi

among individual poets, we can detect several common traits of skoptic epigram. These tendencies reveal a common interpretation of the subgenre, and attest to a stylistic change in the history of the epigram.

1. Lucillius Lucillius is the ‘father-figure’ of Greek satirical epigram. In an important article published in 1968, Rudolf Keydell builds on the analysis of Linnenkugel 1926, 52 ff. by showing that the language of this Neronian epigrammatist is informal and non-classical. In particular, Lucillius’s lexical, morphological, and syntactical choices are influenced by everyday language. The most distinctive feature of his style, in fact, is its colloquial character and its openness to the linguistic novelties of his time, as testified by papyri and post-classical texts such as the New Testament. The analysis that I undertook while working on a new critical edition and commentary of Lucillius’s epigrams⁶ allowed me to expand on Keydell’s contribution and to reinforce and clarify his conclusions. The most important features of Lucillius’s ‘colloquial’ style can thus be summarised as follows: Lexicon – Words can take a meaning they did not have in Classical Greek: ἐκολύμβα, in AP 6.166.3 = 1.3 Floridi, for instance, does not mean ‘to dive, plunge headlong’,⁷ but ‘to swim’,⁸ as, e. g., in [Hierocl.] Philog. 2a and 2b Dawe;⁹ ἄνωθεν, in AP 11.78.5 = 10.5 F., is perhaps to be taken in the temporal sense of ‘over again’, ‘de nuevo’,¹⁰ as often in post-classical Greek;¹¹ neutral participles, such as τυχόν (AP 11.160.3 = 57.3 F., AP 11.246.5 = 96.5 F.), are used adverbially, according to a prosaic and colloquial use.¹²

ern readers of the Greek Anthology mostly tended to ‘normalise’ syntax and lexicon, a more perceptive understanding of the alleged anomalies of these authors’ style is encouraging scholars to be more conservative. Some examples will be offered in the course of this essay.  Floridi  (in particular, for Lucillius’s style, ‒).  LSJ s.v., .  Keydell , .  Further examples in LSJ s.v., .  See LSJ s.v., II. and DGE s.v., B. respectively.  Examples include Artem. .; NT, Ep. Gal. .. For this interpretation of the adverb in the epigram, see my comm. ad loc.  Linnenkugel , . Τυχόν is sometimes used in the sense of ἴσως, ‘maybe’, already in classical Greek prose: see e. g. Xen. An. ..; Blass‒Debrunner §  n. .

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

–

–

73

Lucillius employs ‘unpoetic’ and post-classical words: e. g. post-classical adverbs, such as ἑκοντί (AP 11.276.1 = 108.1 F.), or πάλι (AP 11.277.2 = 109.2 F.);¹³ words like χοραύλην (AP 11.11.1 = 4.1 F.) which is very rare in Greek Literature, only attested in prose, and never before the Imperial age;¹⁴ κλινικός (AP 11.113.1 = 36.1 F. and AP 11.116.3 = 39.3 F.), a technical term derived from κλίνη, attested, as a substantive, only in Imperial medical prose;¹⁵ χορτασίη (AP 11.313.4 = 117.4 F.), another rare and prosaic word, only found in the LXX, in inscriptions, and then in later prose;¹⁶ the diminutive form ὠτάρια (AP 11.75.2 = 7.2 F.), used as a simple equivalent of οὖς (cf. e. g. Ev. Jo. 18.10);¹⁷ τρώγεις (AP 11.207.1 = 82.1 F.), regularly used in popular language from the I century AD onwards as the present form of ἔφαγον instead of ἐσθίω;¹⁸ the adverb πάντοτε for ἀεί (AP 11.87.2 = 17.2 F., AP 11.210.6 = 84.6 F.)¹⁹ and the verbal form ἵπτασο (AP 11.392.3 = 125.3 F.), both condemned by Atticists;²⁰ κυνόμυια for κυνάμυια (AP 11.265.1 = 106.1 F.), which is never attested before the LXX. Also noteworthy is Lucillius’s use of words such as καννάβινον in AP 11.107.4 = 33.4 F., ‘hemp-like’, borrowed from the language of comedy.²¹ Particularly interesting is the verbal form ἀπεσκάρισεν, transmitted by the manuscripts in AP 11.114.6 = 37.6 F. ᾿Aπασκαρίζω is a rare verb, attested in Aristoph. PCG 510 ἀπασκαρίζειν ὡσπερεὶ πέρκην χαμαί, where it means ‘struggle, be convulsed’, like a dying fish,²² and in Men. PCG 881 ἀπασκαριῶ δ᾿ ἐγὼ γέλωτι τήμερον, while the simple form ἀσκαρίζω is in Hipponax (frr. 19.2,

 Keydell , .  See e. g. Plut. Ant. ..  LSJ s.v.  E. g. LXX Pr. .; Sammelb. . (Axum, IV AD); OGIS . (Taphis, IV/V AD); later examples in Lampe s.v. Interestingly enough, grammarians and lexicographers use χορτασία to gloss other words indicating fullness, feeding and repletion, such as πλησμονή (schol. ad Ar. Pl. b. Chantry), or κόρος (see e. g. Hesych. κ  Latte); cf. Modern Greek χόρταση.  Blass‒Debrunner §  n. ; in κοινή Greek ‒ιον is often used to form non-diminutive new words.  Blass‒Debrunner §  n. ;  n. ;  n. . Cf. Latin edere vs. manducare; the word is still used in Modern Greek.  Blass‒Debrunner § ; this use survives in Modern Greek.  For the first see Ecl.  Fischer; Moer. π  Hansen, for the second Phryn. Ecl.  Fischer (see also Rutherford , ‒).  Hesych. κ  L. ὅθεν καὶ οἱ λεπτοὶ καὶ ἄσαρκοι κάναβοι λέγονται; Poll. . ὅθεν καὶ Στράττις ἐν τῷ Κινησίᾳ τὸν Σαννυρίωνα διὰ τὴν ἰσχνότητα κάναβον καλεῖ (= Stratt. PCG ). For the variant spellings καν(ν)άβινον, see my commentary ad loc.  Cf. Phot. α  Theodoridis = Suda α  Adler; see also Hesych. α  Latte ἀπασκαρίζειν· σπαίρειν.

74

Lucia Floridi

104.12 West2 = 33.2, 107.12 Degani) and Cratinus (PCG 27). Thus in AP 11.114.6 = 37.6 F. Liddell‒Scott, s.v. ἀπασκαρίζω, correct ἀπεσκάρισεν of the manuscripts in ἀπησκάρισεν.²³ However, the form used by Lucillius is paralleled by LXX Jd. 4.21²⁴ and [Zonar.] 282.7 Tittmann ἀποσκαρίζειν· διακεχυμένως γελᾷν, and can be interpreted as a compound from σκαρίζω, an alternative, late form of ἀσκαρίζω, without epenthetic α (it is attested in Gp. 20.7.4).²⁵ For this reason, we can likely keep the transmitted text: ἀποσκαρίζω might be a post-classical and colloquial form, in keeping with Lucillius’s style. The language of ancient comedy is ‘updated’ with an eye to contemporary linguistic fashion. Morphology – There are morphological ‘oddities’, such as the superlative form μελαινοτάτας in AP 11.68.2 = 5.2 F., paralleled by GVI 1993.4 = SGO 04/05/ 07, v. 4 (Lydia, Thyateira, unknown date) νυκτὶ μελαινοτάτῃ,²⁶ or κράξω for κεκράξομαι in AP 11.141.7 = 50.7 F. (cf. e. g. Ev. Luc. 19.40, Apoc. 17.3).²⁷ Syntax – Normally intransitive verbs can become transitive: e. g. συνεστακέναι in AP 11.139.2 = 48.2 F. τὸν δ᾿ υἱὸν τούτῳ … συνεστακέναι, where it means, ‘to entrust somebody (acc.) to somebody (dative)’. – The perfect is used, when an aorist form would be expected in Classical Greek:²⁸ cf. AP 11.75.6 = 7.6 F. κέκριτ’, AP 11.184.3 = 71.3 F. γέγονεν. The identification of these and other similar features can lead us to a better understanding of the text. A single example will suffice to clarify my point. In AP 11.247.1 = 97.1 F., scholars have detected a syntactic difficulty that has sparked various emendations:

 The emendation is accepted by Beckby ‒, III and Aubreton .  Redaction A; in redaction B the phrase is shorter and the verb does not occur: see Swete  ad loc.  DELG s.v. σκαίρω.  Cf. Kühner‒Blass I., ; DELG s.v. μέλας, . Although the form was already printed by Lascaris  in the editio princeps of the Planudean Anthology, some scholars tried to correct it: see Herwerden , , ‘Barbarismum tam immanem non commisit Lucillius. Dedit κελαινοτάτας’.  See Moulton I, .  Linnenkugel , ; for this phenomenon, see Blass‒Debrunner § ; Moulton III, ‒ .

75

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

Εἰς πέλαγος πλέομεν, Διονύσιε, καὶ γεγέμισται τὸ πλοῖον παντὸς πανταχόθεν πελάγους. ἀντλεῖται δ᾿ ᾿Aδρίας, Τυρρηνικός, Ἰστρικός, Αἴγων· οὐ πλοῖον, πηγὴ δ᾿ Ὠκεανοῦ ξυλίνη. ὁπλίζου, Καῖσαρ· Διονύσιος ἄρχεται ἤδη οὐκέτι ναυκληρεῖν, ἀλλὰ θαλασσοκρατεῖν.

5

We sail the sea, Dionysius, and the ship is full of every sea from all parts. The Adriatic, the Tyrrhenian Sea, the Istricus, the Aegean are running dry. This is not a ship, but a wooden fountain of Ocean! To arms, Caesar! Dionysius begins already not to command a ship, but to command the seas.²⁹

At v. 1, all the most recent editors print Dübner’s correction ἦ, instead of the readings of the manuscripts, εἰς (Pl = Marc. gr. 481) or εἰ (P = Pal. Heid. gr. 23). Εἰ of the Palatine manuscript is clearly corrupt, and Planudes’s text has been accused of the same,³⁰ since πλέω, in Classical Greek, is usually followed by the simple accusative, or by ἐν + dat.³¹ According to Dübner, ἦ – an easy change for εἰς or εἰ – would serve the purpose of emphasising the word πέλαγος. This should be understood as a predicative of the subject, aimed at anticipating the ‘hydric’ metamorphosis described in the epigram: ‘ἡμεῖς ὄντες πέλαγος πλέομεν: vere nos mare navigamus, nam mare sumus, non iam nautae’. Such an interpretation, however, is far-fetched, to say the least, as the embarrassed translations of the editors clearly show: e.g. Beckby 19652‒19672, III: ‘Hoi, wir fahren zur See’; Aubreton 1972: ‘tandis que nous naviguons en pleine mer’; Paton 1916‒1918, IV: ‘of a truth, Dionysius, we the seas sail’.³² In my edition of the epigrams, I thus kept Planudes’ text, for not only is εἰς + acc. in a local sense (a construction which was to become ‘regular’ in later Greek) attested in κοινή Greek from c. 150 BC onwards,³³ when the distinction between motion and rest starts to become obscure, but it is also paralleled by Lucillius himself (AP 11.77.5 = 9.5 F. τὸ πρόσωπον ἰδεῖν ἐς ἔσοπτρον), and the very syntagma εἰς πέλαγος πλεῖν is found in Suda π 923 Adler πελαγιζούσαις· εἰς πέλαγος πλεούσαις. In a similar vein, the poet uses πρός + acc. instead of παρά + dat. in AP 11.207.3 = 82.3 F. τοῦ κοφίνου τοῦ πρὸς

 Translations from the Greek Anthology are by Paton ‒ (occasionally modified).  Dübner , .  See LSJ s.v., I.  The scholar adds in a footnote: “πέλαγος may be taken either as accusative or nominative. In the former case the meaning is ‘we sail the seas’, in the latter ‘we, the seas, are sailing’”. Nystrom  also accepts this interpretation.  Cf. Modern Greek στό with accusative (= εἰς τόν).

76

Lucia Floridi

πόδας and AP 11.249.4 = 98.4 F. ἐτάφη μισθοῦ πρός τινα τῶν ὁμόρων (cf. e. g. Ev. Jov. 1.1 καὶ ὁ λόγος ἦν πρὸς τὸν θεόν).³⁴

Colloquial liveliness Other devices reinforce the impression of Lucillius’s ordinary, ‘unpoetic’ speech. Apart from the prominence of parataxis vs hypotaxis, this author’s epigrams often take on a dialogic form, thanks to the insertion of direct discourse,³⁵ an abundant use of interrogative clauses,³⁶ and of abrupt passages. These are often marked by the adverb πλήν, used in prose, from the Hellenistic era onwards, ‘to break off and pass to another subject’³⁷ (e. g. AP 11.78.5 = 10.5 F.). Lucillius also makes use of proverbs as well as expressions reminiscent of proverbs. Just a couple of examples: in AP 11.91.2 = 21.2 F. τριχὸς ἐκδήσας, used in an absurd, literal sense – a very thin person hangs himself with a hair – toys with the proverb ἐκ τριχὸς κρέμαται, which usually indicates a dangerous situation (Zen. Vulg. 3.47 = CPG I.69; Greg. Cypr. 2.50 = CPG I.362; Diogen. 4.41 = CPG I.238); in AP 11.185.3 = 72.3 F. the phrase Ναύπλιος … ἀεὶ κακόν reminds the audience of Diogen. 1.68 = CPG I.192 ἀεὶ φέρει τι Λιβύη κακόν ~ Zenob. 2.51 = CPG I.45.³⁸

A ‘mimetic’ style Although Lucillius’s predilection for informal linguistic features is certainly the most remarkable trait of his poetry, this register does not prevent him from also using more ‘refined’ poetic language, and his style is by no means monotonous and flat. In targeting professionals his satire often appropriates different jargons; indeed one of the most typical skoptic techniques is the reproduction of a given character’s linguistic mannerisms. Medical idiom, for instance, when satire is di-

 Keydell ,  notes this peculiarity, but he does not discuss AP ... For εἰς = ἐν in local sense, Blass‒Debrunner § ; for πρός τινα = παρά τινι Blass‒Debrunner §  n. ; on this phenomenon, see also Moulton I, ‒, ‒; III, ‒.  See e. g. AP . =  F.  A recurring interrogative phrase that makes the diction more lively is τί γάρ;: AP .. = . F., AP .. = . F., AP .. = . F., AP .. = °. F.  LSJ s.v., III..  See Leutsch‒Schneidewin  ad loc. for further parallels.

77

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

rected against doctors;³⁹ agonistic, when against unsuccessful athletes;⁴⁰ astrological, when against seers and astronomers,⁴¹ etc. This technique is particularly effective when the targets of satire are poets, grammarians, and rhetoricians. Often, in these as in other cases, the author not only reproduces the language of his targets with their mannerisms and actual mistakes,⁴² but he also creates new bombastic words that mimic their pretentious language. In AP 11.140 = 49 F., an epigram with a clear programmatic intent, the poet criticises pedantic philologists who are only able to discuss Homeric questions, thus spoiling the joyful atmosphere of the banquet. The poem best exemplifies this aspect of Lucillius’s style: Τούτοις τοῖς παρὰ δεῖπνον ἀοιδομάχοις λογολέσχαις, τοῖς ἀπ᾿ ᾿Aριστάρχου γραμματολικριφίσιν, οἷς οὐ σκῶμμα λέγειν, οὐ πεῖν φίλον, ἀλλ᾿ ἀνάκεινται νηπυτιευόμενοι Νέστορι καὶ Πριάμῳ, μή με βάλῃς κατὰ λέξιν ‘ἕλωρ καὶ κύρμα γενέσθαι’. σήμερον οὐ δειπνῶ ‘μῆνιν ἄειδε θεά’.

5

To these praters, these verse-fighters of the supper table, these slippery dominies of Aristarchus’s school who care not for making a joke or drinking, but lie there playing infantile games with Nestor and Priam, cast me not literally ‘to be their prey and spoil’. To-day I don’t sup on ‘Sing, O Goddess, the wrath’.

At v. 3 the author employs a vulgarism attested in papyri and New Testament Greek, πεῖν, a contract form for πιεῖν,⁴³ which can be considered as the symbol  See, for instance, the use of verbs such as ἅπτομαι, ‘to touch’, a technical term to indicate the ‘touch’ of the physician’s hand (Boehm ) in AP .. = . F. or AP .. = . F.  See e. g. the hapax πεντετριαζόμενος, ‘to be conquered five times’, in AP .. = . F., built on the basis of τριάζειν, ‘to conquer, vanquish, properly of a wrestler, who did not win until he had thrice thrown his adversary, or conquered him in three bouts’ (LSJ s.v., ). In general, athletes are often mocked through the appropriation and perversion of inscriptional forms traditionally used to praise their agonistic deeds. Inscriptional conventions are turned on their head and praise of victories is substituted with mockery of defeats. Eulogy is replaced by abuse (Robert ).  See e. g. μακρόγηρων in AP .. = . F., on charlatan astrologers, most probably to be interpreted as a terminus technicus of their professional language: the adjective, quite rare, is paralleled by Cat. Cod. Astr. []..  See e. g. AP . =  F., where the solecisms of the satirised target – Heliodorus the grammarian – make the poet actually commit a solecism (see my comm. ad loc.).  Some examples of papyri bearing this reading in LSJ s.v.; Gignac I, . Among these, at least PFlor . might be contemporary with Lucillius (the date is not certain, although it seems to be of the end of the I century AD); see further Heraeus ; Keydell , ; for the phenomenon see also Schwyzer, GG I, ; Blass‒Debrunner § .;  n. .

78

Lucia Floridi

of his heterodoxical expressive choices. In this very epigram, however, Lucillius also creates several neologisms.⁴⁴ These are built on clever reworkings of rare Homeric words beset by debate amongst grammarians both ancient and modern: see especially γραμματολικριφίσιν, ‘puzzled-headed’ (v. 2), from λικριφίς, a Homeric δὶς λεγόμενον,⁴⁵ whose accent and etymology were debated;⁴⁶ νηπυτιευόμενοι, ‘to play childish games’ (v. 4), from νηπύτιος, epic diminutive of νήπιος,⁴⁷ much discussed by Homeric scholars.⁴⁸ Elsewhere, the poet adopts Doric colouring in contexts that reproduce the atmosphere of athletic celebration.⁴⁹ In a similar vein, poetic and Ionic-epic forms are used when a ‘higher’ style is requested;⁵⁰ expressions and phrases reminiscent of epic language can occur,⁵¹ nor are actual poetic quotations, either literal or modified for humoristic purposes, missing.⁵² Far from being a gratuitous parade of erudition, all these stylistic features play a specific role in Lucillius’s poetry, where they serve to generate satire, often with intentional contrastive effects.⁵³ Thus although Lucillius privileges a

 Given the sympotic setting of the poem, there might be here also an allusion to the practice of creating funny compounds during the banquet, as it is testified by Plut. Quaest. Conv. . fa (I owe this reference to Prof. G. Hutchinson, whom I would like to thank).  Il. .; Od. ..  Cf., respectively, schol. A ad Il. .b, III,  Erbse ᾿Aρίσταρχος ὀξύνει ὁμοίως τῷ ‘ἀλλήλοισιν ἔφυν ἐπαμοιβαδίς’ (Od. .) and, for the etymology, e. g., schol. A ad Il. .a, III,  Erbse; schol. b (BCEE), T ad Il. .c, III,  Erbse.  Il. . = ., . = ., ., . etc.  See e. g. schol. b (BCE), Til ad Il. ., III,  Erbse.  See especially AP ..‒ = .‒ F. Πᾶσαν ὅσαν Ἕλληνες ἀγωνοθετοῦσιν ἅμιλλαν / πυγμῆς, ᾿Aνδρόλεως πᾶσαν ἀγωνισάμαν. For the connection between the Doric dialect and athletic celebration, see e. g. Tribulato , , ‒; Silk . For Doric in epigrams, see especially Sens .  Cf. e. g. αὔρης and δι᾿ αἴθρης in AP .. = . F., or the genitive ending ‒οιο in AP .. = . F. and AP .. = . F.: all these epigrams present mock mythological examples that require a somewhat more elevated diction.  Cf. e. g. AP .. = . F. τοῖον ἔλεξεν ἔπος, modelled on epic expressions with τοῖον ἔπος + verbum dicendi; AP .. = . F. φέρτατον ἡρώων, for which compare, e. g., Q. S. . and .‒; AP .. = . F. Ζωγρεῖτε … Τρῶες ἀρηΐφιλοι, reminiscent of Homeric language (for ζώγρει/ζωγρεῖτ’, see e. g. Il. . = ., .; for ἀρηΐφιλος cf. e. g. Il. ., ., ., ., ., ., . = ., .).  See, again, AP . =  F.: at l.  the Homeric phrase ἕλωρ καὶ κύρμα γενέσθαι (Il. .; Od. .) is quoted, and the epigram is closed by a quotation of the very beginning of the Iliad, often employed by Lucillius and other skoptic poets in their epigrams (see my comm. ad loc.).  A good example of such a contrast is AP . =  F., Lucillius’s introduction to his second book of epigrams, which offers dedicatory thanks to Nero. The poem combines quotations from

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

79

plain and colloquial language, his style is, on the whole, rich and variegated, and capable of adapting itself to the different needs of satire and of its targets.

2. Nicarchus Informality and stylistic flexibility Nicarchus shares several linguistic features with Lucillius. His epigrams manifest a similar syntactic informality such as preference for parataxis vs hypotaxis, interrogatives clauses,⁵⁴ proverbs (e. g. POxy. 4502.1.7), dialogues (e. g. AP 11.82), the use of the perfect where an aorist form would appear in Classical Greek (e. g. AP 11.120.4) and a free use of verbal diatheses. Particularly interesting is AP 11.162, closely modelled on Lucillius AP 11.163‒164 = 59‒60 F., where the active and middle forms of πλέω recur in the same couplet without any difference in meaning: Εἰς Ῥόδον εἰ πλεύσει, τις Ὀλυμπικὸν ἦλθεν ἐρωτῶν / τὸν μάντιν, καὶ πῶς πλεύσεται ἀσφαλέως (vv. 1‒2). One should also note the generalisation of the indicative form in the indirect interrogative clause,⁵⁵ which led Planudes to the normalising correction Εἰς Ῥόδον εἰ πλεύσοι, τις Ὀλυμπικὸν ἦλθεν ἐρωτῶν / τὸν μάντιν.⁵⁶ The poet also shares Lucillius’s ‘mimetic’ style, although he directs it towards fewer targets. We thus have epigrams on doctors, where technical terms, such as ἅπτομαι, ‘touch’⁵⁷ (AP 11.118) and κλύζω, ‘purge’ (AP 11.118, 11.119), are used.⁵⁸ In a variation on the theme of the murderous doctor, Nicarchus also

Hesiod and Homer with allusions to the prosaic reality of the economic support requested by the poet from the emperor. See also AP . =  F., on the unsatisfying dinner offered by a host more interested in form than substance. Here the combination of poetic traits (e. g. epic εἰλαπίνην, v. ; ὀχθήσας δ᾿ ἄρ᾿ ἔειπον, v. , modelled on Homeric ὀχθήσας δ᾿ ἄρα εἶπε; the precious hapax ἀργυροφεγγέϊ, v. ) and more colloquial words (e. g. χορτασίη, v. ) stylistically replicates the contrast between the pomp of the silverware shown by the host and the emptiness of the plates he actually offers his guests.  Included τί γάρ, so frequent in Lucillius: cf. AP .. (for the opportunity of maintaining the traditional punctuation, against Schatzmann’s , ‒ proposal to change it, see Galán Vioque , ; Floridi , in press).  Εἰ is normally followed by the indicative (either present, future or aorist) in New Testament Greek: cf. Blass‒Debrunner § . and n. .  The reading πλεύσηται at l. , attributed to Planudes by Aubreton  and Schatzmann , is not in the manuscript: Planudes reads πλεύσεται, as P.  See supra, n. .  Also noteworthy is ὑδρωπική (AP ..), a medical term used metaphorically for a boat.

80

Lucia Floridi

adopts the typical Lucillian form of the mock-inscription,⁵⁹ writing a mock-epitaph (AP 11.124). Here the parody of funerary inscriptions implies the adoption of a ‘higher’, more pretentious style (note Doric μάν at l. 1, or the poetic expression οὓς γλυκεροῦ φέγγους … ἐστέρισεν at l. 2, reminiscent of the Homeric iunctura γλυκερὸν φάος),⁶⁰ thus revealing the poet’s ability to adapt his diction to different contexts. Another good example of such stylistic flexibility in Nicarchus is AP 11.407, a variation on the typical Lucillian theme of the λεπτός, the abnormally thin man:⁶¹ Τὸν λεπτὸν θακεῦντα Μενέστρατον εἴαρος ὥρῃ μύρμηξ ἐξελθὼν εἵλκυσεν εἰς ῥαγάδα· μυῖα δ’ ἐπιπτᾶσ’ αὐτὸν ἀνήρπασεν, ὡς Γανυμήδην αἰετὸς εἰς θαλάμους οὐρανίους Κρονίδεω· πίπτεν δ’ ἐκ χειρῶν μυίης, κοὐδ’ ὣς θίγε γαίης, ἐκ δ’ ἀράχνης ἱστοῦ τῶν βλεφάρων κρέμαται.

5

As lean Menestratus was sitting in springtime, an ant came out and pulled him into a crevice; but a fly flew up and carried him off, just as the eagle carried Ganymede to the heavenly chamber of Zeus. He fell from the fly’s hands, but not even so did he light on the earth, but is hanging by his eyelids from a spider’s web.

The poem, which conflates several Lucillian models,⁶² develops the story through a linear narrative, mostly based on common words and unproblematic syntactical structures. Nevertheless, several elements call for comment: θακεῦντα (v. 1), epic-Ionic contraction of the verb θακέω which is mostly used in tragedy,⁶³ creates an impression of elevated diction; the Ionic iunctura εἴαρος ὥρῃ (v. 1)⁶⁴ and the forms of indicative without augment πίπτεν and θίγε (v. 5) create, in their turn, an impression of poeticism. The epigram thus combines prosaic and more poetic traits, with a certain taste for stylistic contrast. AP 11.328,

 See supra, n. .  On this epigram, see Floridi , ‒; Schatzmann , ‒. Another parody of inscriptions in Nicarchus is to be found in AP ., whose specific model is the epigram on Ladas, AP .: see Schatzmann , ‒, with further bibliography.  On which see Floridi  and ,  ff.  The exemplum of the mythical abduction is taken from Lucillius AP . =  F.; the very little man pulled into a crevice by a small animal at a specific time of the year is reminiscent of Lucillius AP . =  F., the spider’s web of Lucillius AP . =  F. and AP . =  F., the ant of Lucillius AP . =  F. (from which also the name of the thin man, Menestratos, is taken).  Cf. e. g. A. Pr. , ; Soph. Aj. , , OT ; Eur. Hec. , Heracl. .  Paralleled, for instance, by [Mel.] AP .. and adesp. AP .., both presumably of late date.

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

81

one of Nicarchus’ best poems, plays entirely upon stylistic contrast by cleverly investing Homeric expressions with new, obscene meanings in the description of a ménage à quatre. ⁶⁵ While most of the stylistic features described so far show how Nicarchus can write very much like Lucillius, the Homeric parody of AP 11.328 is, for its very bawdiness, typically Nicarchean.⁶⁶ It serves, then, as a good introduction to what is perhaps the most distinctive aspect of Nicarchus’s poetry: his characteristic scurrilous vein, which distinguishes his epigrams from those of other skoptic authors and particularly Lucillius.⁶⁷

Scurrility, primary obscenities and double entendres Prostitutes (e. g. AP 11.71, 11.73), pathici (POxy 4502.4), cunnilingus (AP 11.329) and bizarre sexual intercourse (AP 11.328) are at home in Nicarchus’s poetry. As a consequence, his language is often sexually loaded, either through the use of primary obscenities, or through the use of metaphors or lexical ambiguity. These terms might derive from the imagery and language of previous poetry (especially comedy), while others seems to be unique to Nicharchus, and are presumably inspired by contemporary linguistic uses. Apart from several common sexual euphemisms, widespread in Greek literature from comedy, to epigram, to prose,⁶⁸ Nicarchus’s poems show a particular predilection for crude words and sexual metaphors. Among the most remarkable are the following: – The mild obscenity βδέω, ‘break wind’ (AP 11.242.2), an onomatopoeic word of comic derivation, frequent in Aristophanes (see e. g. Pl. 693, Pax 151)⁶⁹ and later found, for instance, in [Hierocl.] Philog. 233 D. (in the very aorist form used by Nicarchus).⁷⁰ – The coarser πρωκτός (AP 11.241.1 and 3), terminus technicus to indicate the anus at least from Hipponax fr. 104.32 West2 = 107.32 Degani onwards, and  It is no coincidence that the epigram has attracted much scholarly attention: see especially Magnelli ; Vergados . They both stress that the text implies an audience familiar not only with Homer, but also with scholarly discussions of Homeric interpretation.  For the use of Homer in epigrams in general, see now Guichard, forthcoming.  Sexual themes, although not completely absent from Lucillius’s corpus of epigrams, are only occasionally touched upon by him: see Floridi , .  E. g. λαμβάνειν and διδόναι for ‘receiving/giving sexual favours’ (AP ..‒), or θέλειν to indicate erotic will (AP ..).  On words designing crepitation in comedy, see Henderson , ‒ (in particular ‒ for βδέω).  Nicarchus also writes an epigram on πορδή, AP ..

82

–

–

Lucia Floridi

then regularly in Aristophanes and other playwrights.⁷¹ In the Greek Anthology, it only occurs in Nicarchus and in an epigram by the linguistically licentious Strato of Sardis, AP 12.6.1 = 6.1 Floridi.⁷² Derived from comedy is also κόλλοψ (POxy 4502.1.4). The word, already used by Dioscorides AP 12.42.6 = HE 1528, does not mean, as suggested by Hesychius and other lexicographers,⁷³ a ‘beloved past the age of youth’,⁷⁴ but simply ‘prostitute’, open to any experience, as recently remarked by Morelli 2015 on the basis of Pöhlmann‒Tichy 1982.⁷⁵ Morelli also rightly stresses that the substantive is to be interpreted as the learned recuperation, on the part of the two epigrammatists, of a linguistic fossil, no longer used in everyday language during the Hellenistic and Roman periods, whose meaning was actually discussed by philologists. Χοῖρος (AP 11.329.2), lit. ‘young pig, porker’, a common slang term for the pudenda muliebra in comedy.⁷⁶ Noteworthy is also, in the same poem, the use of ἄκανθα (v. 2) as a metaphor to indicate the ‘hairy’ dangers met by the cunnilingus: Δημῶναξ, μὴ πάντα κάτω βλέπε, μηδὲ χαρίζου τῇ γλώσσῃ· δεινὴν χοῖρος ἄκανθαν ἔχει. καὶ συζῇς ἡμῖν, ἐν Φοινίκῃ δὲ καθεύδεις κοὐκ ὢν ἐκ Σεμέλης μηροτραφὴς γέγονας. Demonax, do not always turn down your eyes, nor indulge your tongue; the pig has a formidable thorn. And you live with us, but you sleep in Phoenicia, and though not Semele’s son, art nourished by a tight.

While Schulte 1999 tries to keep the metaphor of v. 2 by translating ‘ein Schwein hat eine gefährliche Borste’ (‘a pig has a dangerous bristle’), Schatzmann 2012

 As remarked by Henderson , , the word was a primary obscenity, capable of provoking laughter even ‘in the absence of a joke’.  The poem is a playful isopsephy, which establishes the equivalence between πρωκτός and χρυσός to signify the trivial venality of love for boys.  E. g. Hesych. κ  Latte κόλλοπες […] καὶ τοὺς σκληροὺς δὲ καὶ παρηβηκότας παῖδας […] κόλλοπάς φασιν; Suet. Blasph.  Taillardat Κόλλοψ· ὁ σκληρός, ὑπὲρ τὴν ἀκμὴν πάσχων.  Pace Schatzmann , .  See also Slater , ‒; Morelli ,  and nn.  e ; Galán Vioque , ‒. Doubts on the meaning suggested by the lexicographers were already expressed by Gow‒Page ad loc.  Ar. Ach. ; ‘χοῖρος indicates the pink, hairless cunt of young girls as opposed to that of mature women’ (Henderson , ). In AP, see the pun on the proper name Χοίριλος in Crates AP ..‒ = HE ‒.

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

83

prefers ‘ein Schwein hat einen fürchterlichen Stachel’ (‘a pig has a terrible sting’), and argues that the line contains a reference to the porcupine, ἀκανθόχοιρος. Be that as it may, the metaphorical meaning of the expression is clear: ἄκανθα hints at (pubic) hair, as, for instance, in adesp. AP 12.40.3‒4 = HE 3700‒3701 γυμνὴν ᾿Aντιφίλου ζητῶν χάριν, ὡς ἐπ’ἀκάνθαις / εὑρήσεις ῥοδέαν φυομένην κάλυκα, where the puer’s anus is described as a rose (ῥοδέαν … κάλυκα) among thorns (ἐπ’ἀκάνθαις), according to an image usually exemplified by the opposition ῥόδον/βάτον, rose buttocks vs dark hair, frequent in pederastic contexts.⁷⁷ Obviously, the implied reference in Nicarchus to the rose hints at the pudenda muliebra, as often in comedy.⁷⁸ At v. 3, the description of the drawbacks of Demonax’s activity proceeds through another sexual metaphor: ἐν Φοινίκῃ δὲ καθεύδεις clearly alludes to the act of φοινικίζειν, ‘to imitate the Phoenicians, of unnatural vices’,⁷⁹ that is to act as a cunnilingus, to mouth the vaginal orifice.⁸⁰ One should note that Nicarchus does not employ the actual verb, but alludes to the sexual habit attributed to the Phoenicians through a geographical metaphor (ἐν Φοινίκῃ δὲ καθεύδεις), where the act of sleeping is also obviously significant in the erotic context. A sexual metaphor is finally implied in the last line: here, in a mock-mythological context, the cunnilingus is described as μηροτραφής, an epithet for Dionysus,⁸¹ which playfully alludes to the passion of Demonax for μηροί, or better, for what is between them, μέσος μηρῶν, i. e. the vagina.⁸² AP 11.329 is thus a good example of Nicarchus’s ability to personalise the sexual language and metaphors inherited from the literary tradition with original twists and inventiveness.

 Floridi , . This is the common pederastic motif of the εἰσὶ τρίχες: Tarán .  Henderson , .  LSJ s.v.  Probably upon menstruating women, given the association between the Phoenicians and the colour of blood, i. e. red: see Schulte  and Schatzmann  ad loc. For φοινικίζω in this sexual sense see Gal. simpl. med. . = XII. Kühn; Luc. Pseudol. ; Jocelyn ,  and , n. .  The expression κοὐκ ὢν ἐκ Σεμέλης alludes to Dionysus’s birth from Semele, and μηροτραφής is used for the god, for instance, in Str. ...  Μηροί are usually associated with pederastic love (see e. g. Anacr. fr.  Gentili; Sol. fr. . Gentili‒Prato; Eub. PCG ; Plut. Mor. c); thus the use of διαμηρίζω for intercrural intercourse (Dover , ), or of ὁμηρίζειν, ὁμηρικός in an obscene sense in Ach. Tat. .. and Crates AP .. = HE , respectively, or the lusus in nomine on Μηριόνης in Antip. Sid. AP ..‒ = HE ‒ and Strato of Sardis AP .. = . F. A heterosexual interpretation of the joke on Μηριόνης, which somewhat parallels Nicarchus’s use, is in Rufinus AP .. = . Page (see Höschele , ‒).

84

Lucia Floridi

A similar longing for linguistic novelty is demonstrated by AP 11.7, an epigram on the inevitability, given human nature, of conjugal infidelity: Οὐδεὶς τὴν ἰδίην συνεχῶς, Χαρίδημε, γυναῖκα κινεῖν ἐκ ψυχῆς τερπόμενος δύναται· οὕτως ἡ φύσις ἐστὶ φιλόκνισος, ἀλλοτριόχρως καὶ ζητεῖ διόλου τὴν ξενοκυσθαπάτην. No one, Charidemus, can constantly sleep with his own wife and take heartfelt pleasure in it. Our nature is so fond of titillation, such a luster after foreign flesh, that it persists in seeking the illusion of a stranger κύσθος.

The poem combines the use of a common euphemism for sexual intercourse such as κινεῖν (v. 2) with original neologisms: φιλόκνισος (v. 3) is created on the basis of a common erotic euphemism, κνίζω, widely used metaphorically of love,⁸³ and employed by Nicarchus himself in AP 11.73.7; ἀλλοτριόχρως (v. 3) – if the text is correct⁸⁴ – entails a novel combination of two common words, ἀλλότριος and χρώς, as in Mel. AP 7.207.3 = HE 4322 γλυκερόχρως (itself a hapax, at end of the hexameter, like here); ξενοκυσθαπάτην (v. 4), at the very end of the poem, combines, with comic contrastive effects, the vulgar κύσθος, a comic slang for pudenda muliebra,⁸⁵ with the adjective ξεναπάτης, attested in lyric poetry and in tragedy.⁸⁶ Given the context, one might also wonder whether φύσις (v. 3) is to be taken as a euphemism for πέος (φύσις, as natura in Latin, is commonly used in Greek in this sense).⁸⁷ The poet would thus be wittily hinting at the fact that it is the man’s membrum to always long for new erotic adventures, in keeping with the search for κύσθος declared in the epigram’s closure.⁸⁸ Other words used by Nicarchus in an erotic context and unattested elsewhere are εὐεπίτακτον, ‘submissive’, and παθικεύεται, ‘to be sexually passive’ (both in AP 11.73, vv. 4 and 7 respectively). The hapax παθικεύομαι is particularly interesting in the light of POxy 4502.4.2, where the only occurrence of the word

 LSJ s.v., II.; see also κνίσμα, in a similar sense.  ᾿Aλλοτριόχρως is Toup’s (, I,  and II, ) plausible correction for ἀλλότριος χρώς of the manuscripts (unconvincingly defended by Schneider , ).  Cf. Ar. Ach.  and , Lys. , Ran. ; Henderson , .  E. g. Alc. fr. . Voigt; Pind. O. .; Eur. Med. , Tr. . For similar hapax, cf. Asclep. AP .. = HE  = . Sens φιλεξαπάτης, with Sens  ad loc.; Agath. AP .. βρωματομιξαπάτη.  Cf. LSJ s.v., VII.; Henderson , ; Adams , ‒.  The personification of the membrum virile that we would have in this case is common in Greek and Roman Literature: cf. Citroni  ad Mart. .; Adams ,  ff. For such a reading, see already Schneider , .

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

85

παθικός in Greek Literature is registered. Before the publication of the papyrus, the word, quite common in Latin to indicate the passive partner in the homosexual couple,⁸⁹ and usually held to be a popular borrowing of a Greek word into Latin, was attested in Greek only in a non-literary context, i. e. in a graffito from Aphrodisias.⁹⁰ If, as seems certain, the word is to be interpreted as a mark of sermo cotidianus, Nicarchus’s epigrams show an interesting mixture of actual colloquial vulgarisms with obscene words and metaphors of literary derivation. Along with ‘learned’ obscenities, such as κόλλοψ, or obscene words already used as such by iambic/comic poets, and presumably still in use, such as πρωκτός, or comic obscenities renewed through neologisms, such as κύσθος in ξενοκυσθαπάτην, Nicarchus also uses current obscenities derived from contemporary linguistic use. Nicarchus’s obscene language, then, is not just a collection of coarse words and images, but a complex, playful and clever blending of different lexical registers. In this light, AP 5.38 merits special attention: Εὐμεγέθης πείθει με καλὴ γυνή, ἄν τε καὶ ἀκμῆς ἅπτητ’, ἄν τε καὶ ᾖ, Σιμύλε, πρεσβυτέρη. ἡ μὲν γάρ με νέα περιλήψεται· ἢν δὲ παλαιή, γραῖά με καὶ ῥυσή, Σιμύλε, † δικάσεται †. A fine and largely built woman attracts me, Similus, whether she be in her prime, or elderly. If she be young she will clasp me, if she be old and wrinkled † … †

The epigram is closed by a textual corruption: P’s δικάσεται, both unmetrical and nonsensical, most probably hides a reference to fellatio. ⁹¹ Among the proposed restorations, Brunck’s 1773, 349 λιχμάσεται⁹² can be dismissed, since the verb would be unparalleled in this sense.⁹³ Nor does συκάσεται, proposed by Hecker 1843, 35 on the basis of Hesych. σ 2220 Cunningham συκάζει […] τὸ κνίζειν ἐν ταῖς ἐρωτικαῖς ὁμιλίαις, supply the meaning required by the context.⁹⁴ Al-

 E. g. Cat. ., ..  Bain .  Scholars usually compare the closure of Hor. Ep. . ore adlaborandum est tibi. The poem, in general, has been considered difficult and inconsistent: see Schatzmann , in particular ‒. A new interpretation of these lines is offered by Di Marco .  Approved by Jacobs ‒, I, ; Stadtmüller ; Waltz .  As remarked by Bain ,  n. , ‘the parallels for this manner of speaking of fellatio are as yet confined to Latin’: see Adams , ‒.  Συκάζειν is most probably euphemistic for βινεῖν: see Stratt. PCG . (with Kassel‒Austin ad loc.); Pl. Com. PCG ; Men. PCG ; Henderson , .

86

Lucia Floridi

ternatives are Heraeus’s 1915, 38 n. 1 λαικάσεται⁹⁵ and Toup’s 1790, 169 λειχάσεται (accepted by most of the editors).⁹⁶ Λαικάσεται would both satisfy sense and palaeography, but it is usually rejected for two reasons: (a) it would be the only occurrence of the verb with an object expressed; and (b) the word is considered too crude even for Nicarchus, as first remarked by Jocelyn 1980, 27.⁹⁷ The first difficulty might be removed by accepting Waltz’s τε for transmitted με: the change would be easy and justified by the occurrence of με in the preceding line, in the second foot as here and preceded by γάρ, a word bearing similarities to γραῖα. As an alternative, we might also leave open the possibility that λαικάζω could have actually been followed by an object, due, perhaps, to the influence of Latin fello, which allows a transitive construction.⁹⁸ Nicarchus lived in a Roman milieu, and his language is sometimes reminiscent of Latin, as we will see presently. As far as the second difficulty is concerned, Jocelyn’s view that the verb is too crude to appear in an epigram has, as it seems to me, excessively influenced later scholars. Given Nicarchus’ general tendency towards the vulgar and coarse, now further proved by some of the poems preserved in Oxyrhynchus papyri,⁹⁹ it is not impossible that he might have used such an obscenity of comic derivation. Jocelyn’s study has proved that the word λαικάζω was particularly coarse and uneducated in tone. Nevertheless, it appears not only in Attic comedy,¹⁰⁰ notoriously prone to obscenity, but also in New Comedy, where, as it is well known, indecencies are usually more limited.¹⁰¹ Although in many comic passages the verb is just a generically obscene imprecation¹⁰² while in the epigram it is employed in its proper sexual sense, there is no reason to think that its use in Nicarchus would have been felt as more ‘shocking’ than that of πρωκτός, or than the reference to κύσθος in ξενοκυσθαπάτην.¹⁰³ As regards λειχάσεται, printed by most of

 Already proposed by Jacobs ‒, I, , and then withdrawn at III, , for a misunderstanding of its true meaning: see Jocelyn ,  and  n. .  Dübner ; Paton ‒, I; Beckby ‒, I; Schulte ; Schatzmann .  See then Bain , ; Schatzmann , ‒.  Cat. . Rufa Rufulum fellat; CIL IV. Salvia felat Antiocu(m) luscu(m); on the verb, see Adams , ‒.  As Nisbet ,  rightly remarks, ‘Nikarkhos is noticeably ruder on papyrus’.  Cf. Ar. Eq. , Thesm. ; Cephisodorus PCG ..  Strato CGFPR . and especially Men. Dysc. , with Jocelyn , ‒.  Degani , ; Jocelyn’s  discussions of these passages.  A clear allusion to λαικάζειν is in Strat. AP .. = . F. τοῖς φθονεροῖς λάμβδα καὶ ἄλφα λέγε. That scurrilous Strato does not actually spell out the verb is hardly significant. This epigram on a music teacher who sexually abuses his pupils employs double entendres from the language of school teaching (in particular, it seems parodically reminiscent of Arat. AP . =

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

87

editors, it is considered ‘unacceptable’ by Bain 1991, 75, n. 198, ‘since the word does not enter the Greek vocabulary until later’. In fact, λειχάζω is only attested in Greco-Latin Glossaries, such as the one edited by Bonaventura Vulcanius in 1536,¹⁰⁴ where it might have been drawn, as observed by Heraeus 1915, 38, n. 1, from Martial 11.58.12. Here, part of the tradition has leicazin or leicatin, rightly restored to λαικάζειν already by Schneidewin 1842, 493, and then commonly printed by the editors,¹⁰⁵ instead of Calderini’s 1480 λειχάζειν.¹⁰⁶ Nevertheless, Schatzmann, on the grounds of Martial’s passage, considers the verb a Byzantine form,¹⁰⁷ and tentatively suggests that Nicarchus, in order to avoid the overly coarse λαικάζω, might have been the first to use a new formation from λείχω¹⁰⁸ that would have reappeared only much later in Greek. However, Martial’s leicazin is simply most likely just a medieval misspelling for λαικάζειν, due to a confusion between λαικάζω and λείχω, which are equivalent in meaning, and thus cannot be taken as evidence for the existence of λειχάζω. At least another misspelling of this kind, caused by what Jocelyn defined as a ‘semantic contamination’,¹⁰⁹ is attested, and can further prove the opportunity of rejecting λειχάζειν as non-Greek: λαιχάζειν (or λαιχκάζειν) for λαικάζειν in schol. ad Aristoph. Ecc. 920.¹¹⁰ In conclusion, unless more reliable evidence for the existence of λειχάζω appears, it is certainly safer to print Heraeus’s palmar emendation in Nicarchus’s

HE ‒ αἰάζω Διότιμον, ὃς ἐν πέτραισι κάθηται, / Γαργαρέων παισὶν βῆτα καὶ ἄλφα λέγων; see my discussion ad loc.).  Onomasticon vocum latino-graecarum, col. : fello, las, λειχάζω. See also ThGL, vol. /I, col.  λειχάζω = fello (Gl.). The form, however, is not recorded by modern dictionaries, such as LSJ, Sophocles, LBG, or Lampe.  See also Kay  ad loc.  Calderini, however, apparently intended λεσβίζειν, and λειχάζειν was supplied by whoever looked after the printing in Venice: see Jocelyn , . Paoli , ‒ made an attempt at defending λειχάζειν, but see Degani .  ‘… ist λειχάζω eine vor der byzantinischen Epoche nur schlecht bezeugte Variante (nämlich in einem Teil der Überlieferung bei Martial ,,; dazu s. Heraeus :  m. Anm.) zur eigentlichen vox propria für fellatio, λαικάζω’ (Schatzmann , ).  For its obscene use see e. g. Ammian. AP ..  Jocelyn , .  See Regtuit’s apparatus ad loc. X‒XI century’s codex Marc. gr.  (Luc. Lex. ) does not read, pace Jocelyn , , λεικαλέος for λαικαλέος, but λεκαλέος, as several other manuscripts of Lucian do (see MacLeod’s apparatus ad loc. Francesco Valerio checked the Marc. gr.  for me).

88

Lucia Floridi

epigram,¹¹¹ and to consider λαικάσεται as just another item of comic obscene vocabulary in a notoriously obscene poet.

Neologisms, colloquialisms and Latinisms A certain fondness for linguistic novelty on the part of Nicarchus – both in the creation of neologisms and the adoption of colloquialisms – is also testified in non-erotic contexts. Noteworthy among these neologisms are the compounds λεπτεπιλεπτότερος, ‘thin upon thin’ (AP 11.110.2), a superlative form which is reminiscent of comic creations such as παππεπίπαππος, ‘grandfather’s grandfather’ (Nicopho PCG 23 = Philonid. [dubium] PCG 17), and especially φαυλεπιφαυλότατοι, ‘bad upon bad’ (AP 11.238.4, at the end of the climax φαῦλοι-φαυλότεροι-φαυλότατοι);¹¹² Ἱπποκρατιππιάδης (AP 11.17.4) and Διονυσιοπηγανόδωρος (AP 11.17.5), two funny personal names aimed at satirising the social climbing of a gardener,¹¹³ who changes his name in order to reflect his alleged acquisition of power and prestige.¹¹⁴ Another hapax is the metrically convenient λεπτοσύνης (AP 11.110.1), a later variant of λεπτότης formed by the suffix –συνη, which continued to be used for word coinages from the Hellenistic period onwards. Colloquialisms include the diminutive ὀσταρίοις (AP 11.96.2), in the sense of ‘little bones’;¹¹⁵ ἐδυσώνει (AP 11.169.3), ‘to cheapen, beat down the price’, only attested, in this sense, in Pl. Com. PCG 246; the adjective ἐπιτίτθιον qualifying παῖδα, ‘at the breast’ (AP 11.243.3), elsewhere attested in poetry only as a varia lectio for ὑποτίτθιον in Theocr. 24.54, where it means ‘a suckling’ (substantive),

 It might also be worth noting that the (easy) corruption λαικάσεται > δικάσεται is paralleled in Martial (Gryphianae: see Degani , , n. ).  Schulte  ad Ammian. AP .. and Schatzmann  ad loc.  If Jacobs’ κηπεύς θ’, instead of καὶ παῖς trasmitted by P and Pl at l. , is correct (see Schatzmann  ad loc.).  Ἱπποκρατιππιάδης hints at ‘great wealth and position owing to its very horsey sound’ (Paton ‒, IV). Διονυσιοπηγανόδωρος is a comic exaggeration of the aristocratic name Διονυσιόδωρος with the addition of πηγανό‒, ‘rue’, a common herb that hints at the character’s previous profession, while probably implying, at the same time, a pun on πυγή (Aubreton  ad loc.). For these kinds of comic personal names, see especially Plaut. Capt. .. Thesaurochrysonicochrysides, or the monstrous coinages in Alciphron’s Letters of Parasites, such as Τρεχέδειπνος, or Λοπαδεκθάμβος.  As in, e. g., PTeb. . (II/I BC); Damocr. ap. Gal. .; Tz. H. .. Elsewhere the word means ‘stone, kernel’ (of nuts): see LSJ s.v.

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

89

and then in later prose and lexica;¹¹⁶ βαύκαλιν, ‘vessel’ (AP 11.244.4), another rare word, only found in Imperial prose (e. g. Athen. 11.784b); βουνόν for ὄρος, ‘hill’ (AP 11.406.5), a barbarism,¹¹⁷ according to Hdt. 4.199, frequently used in Imperial prose (cf. e. g. Plb. 3.83.1) and in New Testament Greek (LXX Ex. 17.9, Lc. 3.5, 23.30); later verbal forms such as ἀποκτέννω for ἀποκτείνω (AP 11.395. 1 and 3), common in Imperial prose (cf. e. g. Plb. 1.69.11) and in New Testament Greek (LXX To. 6.13, Mt. 10.28, Lc. 12.4, Apoc. 6.11).¹¹⁸ Noteworthy is the technical term λόγευμα (POxy. 4502.5.3), ‘taxes collected’,¹¹⁹ only attested, before the publication of POxy. 4502, in documentary papyri from the III century BC to the II century AD.¹²⁰ The word is related to the parent substantive participle τὸ λογευθέν, ‘the money raised’ (by the sale of wheat), at Plb. 31.31.1; forms of λογεύω, ‘to collect contributions, taxes, etc.’, are elsewhere attested only in papyri.¹²¹ Nicarchus’ tendency to incorporate linguistic materials from everyday language is also testified by a number of Latinisms:¹²² the Roman measure ξέστης, from Latin sextarius (AP 11.73.6);¹²³ μιλιάριον, from Latin miliarium, a copper ves-

 E. g. Suda ε  Adler Ἐπιτίτθιον· ὑπομάζιον, ὑπὸ τὴν τροφόν.  I.e. one of the few ‘Aeolisms’ that can be found in κοινή: Blass‒Debrunner § , n.  and §  n. .  Blass‒Debrunner § ., n. .  LSJ s.v.; see Preisigke s.v. Parsons , ‒, translates as ‘the entrance fees’ (the word appears in the phrase πιστεύεις … θεατρώνῃ λόγευμα, ‘You are entrusting … the entrance fees to a theatre-manager’) and then comments: ‘Since this is something appetising to the theatre-lessor, does the word here refer to tickets money? Or should we visualise three tiers? – the city rents its theatre to the lessor, who then collects fees from companies using it?’.  P.Cair.Zen. . (III century BC); PCol . (III century BC); PCol . (III century BC); PMich. . (III century BC); PRevLaws . (III century BC); Sb. . (III century BC); PTebt. .. (II century BC); PTebt. .. (II century BC); Sb. . (I century AD); PPrinc. . (II century AD). It is, apparently, a technical term most probably developed in the context of Ptolemaic administration.  See LSJ s.v.  Lexical Latinisms in Lucillius are fewer, in spite of the fact the he certainly had contacts with the Neronian court. Apart from personal names, such as Γάϊος or Μάρκος, Latinisms include a possible bilingual lusus in nomine in AP .. = . F. (τὸν μικρὸν Μάκρωνα, probably playing both on μικρόν/Μάκρωνα/μακρός and on μικρόν/Μάκρωνα/macer: see my comm. ad loc.), the geographical denomination Αἴγων in AP .. = . F., paralleled by Latin Aegon, and the word ἰσίκιον, from Latin i(n)sicium, conjectured by Jacobs ‒, II,  at AP .. = . F. (but not accepted by many editors: see my comm. ad loc.). For a more detailed account of the possible influences of Latin on Lucillius’s language, see Floridi ,  n. .  The word is a back-formation from *ξεστάριον = sextarius, with ξέστης for *σεξτης, since ξτ would have been impossible in Greek; it is used in Imperial prose (see e. g. Gal. XIII.; J. A.J.

90

Lucia Floridi

sel to boil water in (AP 11.244.1);¹²⁴ λάσανον for ‘night-stool’ (AP 11.74.8), singular, as Latin lasanum (cf. e. g. Petr. 41.9; the Greek form is always plural¹²⁵). All these words, significantly, denote common object and belong to everyday speech.¹²⁶ On the whole, Nicarchus’s style can be defined as informal and colloquial, open to linguistic novelties and everyday language and prone to coarseness and obscenity. It is, however, at the same time rich, inventive and variegated, capable of combining different features and registers, learned and literary in its own way and by no means flat and banal. The main differences with respect to Lucillius are its scurrility, its frequency of sexual words and metaphors, and a greater lexical receptiveness towards Latin.

3. Ammianos A taste for puns Ammianos is a poet with a taste for puns:¹²⁷ AP 11.15 toys with the principle of alphabetical order (as clarified by the ‘technical’ expression of l. 3, κατὰ στοιχεῖον); AP 11.98 is based on the pun μητρόπολις/πόλις; AP 11.156 on the assonance φθειρῶν/φρενῶν, ‘lice/brains’. Several epigrams entail word puns with personal names. The companion pieces AP 11.230 and 11.231 serve as good examples.¹²⁸ In AP 11.230 the pun is based on the city name Μάσταυρα: Μασταύρων ἀφελὼν δύο γράμματα, Μάρκε, τὰ πρῶτα ἄξιος εἶ πολλῶν τῶν ὑπολειπομένων.

..) and in New Testament Greek (as a v.l. in Mc. ..; see Blass‒Debrunner § , n. ). In poetry it is also found in adesp. AP .., presumably of a late date. See Schwyzer, GG I, .  Cf. e. g. Athen. .c.  The singular is only paralleled by Hp. Superf. .  A possible influence of Latin is also to be found in AP ., where the name Dinarchus, for a miser, might play on δηνάριον (= Latin denarius; see Blass‒Debrunner § , n.  and § ) + ἄρχω, ‘the one who has power over money’, as suggested by Schulte ,  (contra, Di Marco , , n. ). At l. , the expression φιλαργυρίας δεινῆς might suggest an additional pun, linking together Δείναρχος/δηνάριον/δεινῆς. A bilingual pun involving proper names is probably already in Lucillius (see supra, n. ); lusus in nomine of this kind will then be typical of Ammianos (see infra). In general, for speaking names in skoptic epigram, see especially Conca ‒; Floridi , ‒ (for misers, see in particular Nicarchus AP .. Φείδων ὁ φιλάργυρος, from φείδομαι, ‘save money’).  Nisbet , .  For a detailed analysis of the two epigrams as a pair, see Kirstein , ‒.

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

91

Take away, Marcus, the two first letters from Mastauron, and you deserve many of what is left.

If one takes away the first two letters from Μασταύρων, the addressee, Μάρκος, deserves many of what is left, i. e. σταύρων, ‘crosses’. AP 11.231 exploits the technique of σκώμματα παρὰ γράμμα and διὰ γράμμα.¹²⁹ The addressee is a ‘beast’ but for one letter (παρὰ γράμμα), and a man because of a letter (διὰ γράμμα): Θηρίον εἶ παρὰ γράμμα καὶ ἄνθρωπος διὰ γράμμα· ἄξιος εἶ πολλῶν, ὧν παρὰ γράμμα γράφῃ. You are a wild beast all but a letter and a man by a letter, and you deserve many of the beasts that you are all but a letter.

The missing information to solve the riddle is provided by the preceding epigram, where the name of the man is specified: Μάρκος without a letter is a beast, i. e. ἄρκος. This is a colloquial word, a later form for ἄρκτος, only attested in Imperial prose, New Testament Greek, and non-literary works.¹³⁰ In a similar vein, in AP 11.181, Polemon’s name – it is recognised – is ᾿Aντώνιος,¹³¹ but three letters are suddenly missing (i. e., he is ὤνιος, ‘venal’): Ἤιδειμεν, Πολέμων, ᾿Aντώνιον ὄντα σε πάντες. ἐξαπίνης τρία σοι γράμματα πῶς ἔλιπεν; We all knew, Polemon, that your name was Antonius. How is it that three letters are suddenly missing?¹³²

AP 11.180 is another attack on Polemon: Εἰδοῖς οὐ κρίνει Πολέμων, νώναις κατακρίνει· κἂν δῷς, κἂν μὴ δῷς, ἔστιν ἀεὶ Πολέμων. On the Ides Polemon does not decide the suit, on the Nones he condemns you. Whether you give or don’t give, he is always Polemon.

 Theorised by Arist. Rhet. .a‒. See also Tiberius Rhet., De figuris Demosthenicis  Ballaira.  E. g. LXX.Ki. ., Apoc. .; Ael. NA .; Tab. Def.  Audollent (Carthage, I century AD); IG XIV. (II century AD): see LSJ s.v., I; Blass‒Debrunner § . and n. ; see also Nisbet ,  and n. ; Schulte , .  For the identification with Antonius Polemon, the famous II century rhetor from Smyrna, see Nisbet ,  ff., with further bibliography.  Ammianos might mean that someone ‘bought’ the three missing letters, as Emilia Barbiero cleverly suggests to me.

92

Lucia Floridi

Some kind of play on the two temporal expressions εἰδοῖς¹³³ and νώναις is implied in the poem, as has been long recognised: while the first almost certainly aims at suggesting something like εἰ δῷς,¹³⁴ or εἰ δοίης,¹³⁵ ‘if you give him something’, i. e. ‘if you bribe him’ (see v. 2, κἂν δῷς), νώναις must hint at the opposite (κἂν μὴ δῷς). Οn the basis of AP 11.181, where Polemon is accused of venality, an equivalence νώναις = ἂν ὠνῇς has been suggested.¹³⁶ Other scholars prefer to think that νώναις might play on Latin non ais (‘if you say no’),¹³⁷ or non aes (‘no cash’).¹³⁸ A bilingual pun on the Latin negative non is in fact a likely possibility, especially given the fact that εἰδοῖς and νώναις are Greek transpositions of Latin temporal expressions. These are rarely found in Greek prose, and almost exclusively in authors treating Roman matters, such as Plutarch or Dionysius of Halicarnassus, or in technical contexts, such as epigraphic documents.¹³⁹ One should also note that the puns on personal names analysed so far suggest as well, at least to a certain extent, a bilingual context, in which the audience was presumed to be able to connect Roman names, such as Marcus or Antonius, with Greek words.

Etymologies, myths, and Homeric language Other puns in Ammianos’s epigrams reveal the author’s rhetorical/grammatical training. AP 11.188 plays on the widely discussed etymology of Apollo as ‘destroyer’, from ἀπόλλυμι:¹⁴⁰ Νικήτης ᾄδων τῶν ᾠδῶν ἐστιν ᾿Aπόλλων· ἂν δ’ ἰατρεύῃ, τῶν θεραπευομένων.

 Εἰδοῖς is Huschke’s ,  correction for the manuscript’s εἰδούς (maintained by Beckby ‒, III and Schulte ).  Jacobs ‒, III.,  and ‒, III, ; Aubreton , , n. .  Dübner , .  Aubreton , , n. .  Paton ‒, IV; Beckby ‒, III. A play on Latin non was already suspected by Boissonade ap. Dübner , .  Aubreton , , n. ; Bowie, ap. Nisbet , . According to Nisbet , , the very name of Polemon could entail a bilingual pun: πολλὰ emo.  For (ε)ἰδοί = Idus, see e. g. Plut. Rom. . εἰδοῖς ᾿Aπριλίαις, . εἰδοῖς Ὀκτωβρίαις; LSJ s.v.; for Νῶναι = Nonae, e. g., D.H. .; IGRom. .., Acmonia, I AD; BCH suppl. , n. , Thessalonica,  AD (both quoted by LSJ Supplement s.v.); see also the alternative spelling νόν(ν)αι in Plut. Mor. d, and then in later writers (Lampe s.v. νῶναι).  Such an etymology, widespread in Greek literature (see, e. g., A. Ag. ; Eur. Phaeth. fr. ‒ Diggle), is explicitly rejected by Pl. Crat. c-e.

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

93

Nicetas when he sings is the Apollo of the songs, and when he doctors, of the patients.

AP 11.16 entails a false etymology for the Homeric adjective ἐγχεσίμωροι.¹⁴¹ Ancient grammarians were puzzled by the second element of this word, usually explaining it as οἱ περὶ τὰ ἔγχη μεμορημένοι, ὅπερ ἐστὶ πεπονημένοι, vel similia. ¹⁴² Ammianos connects the word to μωρός, ‘dull, stupid’, as the closure of the epigram makes clear (μωρότερος is the superlative of μωρός): Κύλλος καὶ Λεῦρος δύο Θεσσαλοὶ ἐγχεσίμωροι· Κύλλος δ’ ἐκ τούτων ἐγχεσιμωρότερος. Cyllus and Leurus, two Thessalian bounders with the spear, and Cyllus the bigger bounder of the two.

The poet might here exploit a popular etymology,¹⁴³ so as to humorously engage in the erudite dispute going on about the meaning and origins of the Homeric compound.¹⁴⁴ Other poems use myth to exemplify everyday situations. In AP 11.14 the Gorgon and Niobe with their ‘stony’ associations serve the purpose of describing a very hard pillow; in AP 11.147 Thebes is recalled as a land of prodigy; AP 11.209 mentions Irus, the Homeric beggar; in AP 11.228 Oedipus serves as a prototype for a man who has killed his mother, father and brother. It is noteworthy that such mythical exempla do not imply the adoption of a more elevated diction; on the contrary, the language of these poems is plain. AP 11.14, in particular, exhibits one of the most colloquial terms found in Ammianos’s epigrams, τύλη for ‘cushion’. The word, perhaps taken from Lucillius in AP 11.315 = 119 F., is attested in comedy¹⁴⁵ and condemned by Phryn. Ecl. 145 Fischer as a non-Attic form: τύλην, εἰ καὶ εὕροις που, σὺ κνέφαλον λέγε.

 Il. ., ., .; Od. ..  E. g. Apollonius Soph., Lexicon Homericum, . Bekker; see also Hesych. ε  Latte. Frisk, GEW cites Celtic, German, and Slavic parallels for the second element, and explains the compound as ‘speerberühmt’; see also DELG s.v. ἔγχος.  W. Beck in LfgrE has actually suggested that the second element was connected with μωρός by popular etymology, and that the compound was understood as something like ‘berserk’.  An intense, learned and serious engagement with the Homeric poems and their interpretation was especially typical of Hellenistic poets, epigrammatists included (see e. g. Giangrande ); here, Ammianos seems to offer a humorous counterpart to this philological practice (this is nicely paralleled by c.  in the Corpus Priapeorum, where Priapus interprets several Homeric words in an obscene manner). On AP . and its historical background see Sekunda , –.  Cf. Eupol. PCG ; LSJ s.v., .

94

Lucia Floridi

One of the rare poeticisms in Ammianos is the Homeric epithet for death πορφύρεος in AP 11.13.2,¹⁴⁶ but – once again – the overall style of the poem is quite plain. Epic echoes might also be detected in AP 11.226, a mock-epitaph in Lucillian manner,¹⁴⁷ which reverses the topic funerary motif of sit tibi terra levis:¹⁴⁸ Εἴη σοι κατὰ γῆς κούφη κόνις, οἰκτρὲ Νέαρχε, ὄφρα σε ῥηιδίως ἐξερύσωσι κύνες. May the dust lie light on you when under earth, wretched Nearchus, so that the dogs may easily drag you out.

The wish that dogs may easily remove the light dust on the tomb is reminiscent of epic in formulation: ὄφρα + subjective typically introduces a final clause in Homer.¹⁴⁹ Both the adverb ῥηιδίως and the verb ἐξερύσωσι are frequent in Homer: see, respectively, e. g., Il. 4.390, 5.808, 9.184, 11.114 etc.; Il. 5.112, 666, 10.505, 13.532, and especially Od. 18.87 μήδεά τ’ ἐξερύσας δώῃ κυσὶν ὠμὰ δάσασθαι, which might be the ultimate model of the line. The comic contrast between the ‘high’ diction of the couplet and its abusive content is striking.

Satire of rhetoricians and philosophers Given such a clearly rhetorical personality, it comes as no surprise that among Ammianos’s favourite victims are men of letters and science, such as bad rhetoricians and philosophers. When satirising these targets the poet sometimes adopts a ‘mimetic’ style reminiscent of Lucillius. AP 11.146 is a variation on the motif of the ignorant rhetorician, who commits many solecisms when speaking. As in several Lucillius’s epigrams,¹⁵⁰ the adoption of the rhetorician’s language entails the reproduction of his very errors, with an exchange in meaning between ἀριθμῷ and μέτρῳ.¹⁵¹ AP 11.157 – clearly modelled on Lucillius AP 11.142

 The lexical echo does not entail, however, the recuperation of this word’s original ‘bloody’ associations (death is defined as πορφύρεος by Homer with reference to the warriors’ blood).  Another parody of inscriptions is AP ., a mock-dedication against a bad rhetorician.  Cf. Crinagor. AP ..‒ = GPh ‒; Agath. AP ..‒ and AP ..‒; in Latin poetry Mart. ..‒.  Chantraine, GH, ‒.  This is a typical Lucillian theme: see AP . =  F., AP . =  F.  One should also note that the rather colloquial adverbial expression τοῦ λοιποῦ, only found in comedy and prose, is put in the mouth of the rhetorician, perhaps in order to further suggest his unpretentious expressive manners.

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

95

= 51 F. – ridicules the affectations of the philosophers of Ammianos’ days, prone to using platonic forms of speech (vv. 1‒2) and diminutives (v. 3).¹⁵²

A stylistic ‘middle-way’ The most distinguished features of Ammianos’ epigrams, then, are taste for puns and a flair for bilingualism. Although we do not know much about the author’s biography,¹⁵³ his rhetorical training is evident in the ways he plays with language. Further, Ammianos’ range of satirical targets is more restricted than those of Lucillius and Nicarchus.¹⁵⁴ Nevertheless, as his predecessors, the author occasionally adopts a mimetic style. Elevation of diction is rare. At the same time, Ammianos’ style is never too informal: from a lexical point of view, the number of actual colloquialisms is very limited (in his verses, only τύλη can be considered as such, and ἄρκος, which is, nevertheless, only implied in the pun on Μάρκος). There are no sure obscenities.¹⁵⁵ Ammianos’s style is thus characterised by a certain medietas: it is never too ‘high’ (indeed, actual literary quotations are absent from his verses, a feature which distinguishes him from other skoptic poets) nor is it ever too colloquial or coarse.

Conclusions Three different poetic personalities have emerged from this survey. Lucillius’s style is casual, strongly influenced by common speech both on a lexical and on a syntactical level. But it is also capable of elevation in diction through the creation of ‘precious’ neologisms, the adoption of poetic forms and actual poetic quotations. Nicarchus is especially characterised by his propensity towards obscenity combined with a taste for mixing the low and the high via a marked lin-

 See especially Aubreton  ad loc.  Schulte , .  Other targets include a bad man, defined as ‘a scorpion’, and then satirised through astrological imagery (AP .), a person affected by gout (AP .), a big nose (AP .), a bad host (AP .; the poem presents an enumeration of food items reminiscent of comic style).  AP ., a joke where the obscene λείχω occurs (Henderson , , ), is of doubtful ascription and many scholars consider it to be by Nicarchus. However, even if it were by Ammianos, it would be an isolated case of obscenity (and λείχω is not as coarse as its synonym λαικάζω: see supra).

96

Lucia Floridi

guistic creativity. His lexicon often reflects everyday speech, and testifies to a world characterised by a certain degree of bilingualism. Some of the words found in his epigrams are only paralleled by papyri and graffiti. Ammianos, finally, has a special taste for puns, which also entails a certain degree of bilingualism. He is, in general, less colloquial than both Lucillius and Nicarchus, and at the same time less prone to stylistic elevation and lexical creativity. His epigrams appeal to the literate by virtue of their sophisticated play with language, an indication of the author’s own rhetorical and literary training. Therefore, it would be wrong to treat these skoptic poets as an indistinct stylistic whole. Nevertheless, they certainly share the tendency to incorporate contemporary linguistic material into their poems and to ‘mimic’ linguistic characteristics of their targets while simultaneously engaging with the literary tradition, each in their own way. The stylistic choices of the three poets thus testify to a common interpretation of skoptic epigram, in a time when this subgenre was becoming prominent. The expansion of the skoptic form in the Neronian age opens new avenues for the practitioners of the genre. Epigram admits new themes and becomes the genre par excellence of satire and abuse. During this period epigrammatists tend to portray and satirise their everyday reality, with all of its excesses and its ridiculousness. This implies an aesthetic change with respect to the stylistic trends shown by the authors of the Garland of Philip, vis-à-vis the tendencies of earlier authors included in Meleager’s Garland. Although these two sets of poets vary significantly in diction and style, the following generalisation should, nonetheless, give us a general picture of the most relevant stylistic trends recognisable in the two generations of epigrammatists. The authors of the first Garland are, for the most part, learned poets, prone to experimentalism especially in voice and narrative technique, at a time when epigram was starting to free itself from the object it was inscribed upon and to acquire a purely literary status. These authors are highly conscious of their role in the development of the genre, and pay close attention to their relationship with previous poetry. Although their language and style tend to appear simple, it is the product of a learned engagement with traditional literary genres such as epic, lyric, comedy, and tragedy. The language of authors such as Callimachus, Asclepiades and Posidippus varies greatly with individual words and expressions deriving from many different literary genres and registers. And although these poets often disguise such complexity to convey an impression of stylistic clarity, their language is, on the whole, elaborate and complex. Words uncharacteristic of high poetry may occur, and some of these may reflect the spoken language, especially in poems indebted to genres closest to ordinary speech such as comedy and mime. However, the number of actual colloquialisms, as regards both

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

97

syntax and lexicon, is relatively small,¹⁵⁶ and works within the general Hellenistic tendency to combine the high and the low in a sophisticated manner. The poets of the second Garland, for their part, are often skilful versifiers, who write variations on commonplace topics in an affected and recherché style. They are prone to lexical complexity, and exert their creativity mostly through unexpected neologisms and the use of poetic imagery and tropes. Redundant epithets, rare words and long sequences of almost equivalent ideas are the hallmark of their style, often rightly labelled as ‘baroque’. Their model is not the disguised complexity of Asclepiades or Callimachus but the inventive and sometimes abstruse style of Leonidas of Tarentum.¹⁵⁷ With Lucillius, Nicarchus, and Ammianos yet another change occurs: new subjects, strongly connected to a critical observation of everyday reality, enter the realm of epigram. To describe such an ‘ordinary’ reality, the poets renounce the mannerisms that were in fashion in the age of the second Garland, and opt for a more straightforward style. Both syntax and lexicon echo the ordinary language of the human categories described, and a mimesis of their expressive habits is often felt as the best way to capture their peculiarities and faults. In addition, these skoptic authors do not draw only from ‘formalised’ literature when writing their poems, but also take into account popular forms of narratives such as jokes, punch lines, and funny stories, which exercise an influence over style and structure.¹⁵⁸ On the other hand, satirical exaggeration often leads to the search for stylistic effectiveness: hence, the creation of ‘precious’ neologisms, the expressive contrasts, the recuperation of literary words and images and the often abrupt elevation of diction. Skoptic epigram ‘mimics’ the ordinary (or the extraordinary hidden in everyday reality) while commenting on it: to do so, it often has recourse to the tools of learning and literature. This leads to effects of linguistic diversity so far unparalleled, in the history of the genre: Greek skoptic poets write in a colloquial and informal style, borrowing words and syntactical structures from ordinary language, and creatively combining them with lexical inventiveness, poetic reminiscences and occasional engagement with grand style, always with marked effects of contrasts. After the disguised learned complexity of Meleager’s poets and the strong formalisation of Philip’s Garland, such linguistic freedom and mimesis of ordinary speech mixed with lexical novelty and traditional poetic

 Examples from Asclepiades in Sens , lxxviii‒lxxix.  Marcus Argentarius’s mostly plain and straightforward style can be considered a partial exception. On the stylistic evolution of epigram from the Garland of Meleager to the Garland of Philip, see Magnelli .  See Floridi .

98

Lucia Floridi

language, must have been felt as an important change in the aesthetics of epigram, a long-lasting and continually shifting literary genre.

Sigla/Abbreviations Blass-Debrunner Blass, F. / Debrunner, A. (1976), Grammatik des neutestamentlichen Griechisch, Göttingen. Chantraine, GH Chantraine, P. (1948‒1953), Grammaire homérique, I ‒ II, Paris. CPG von Leutsch, E.L. / Schneidewin, F.G. (eds.) (1839‒1851), Corpus Paroemiographorum Graecorum, I ‒ II, Göttingen. DELG Chantraine, P. (1983‒19842), Dictionnaire étymologique de la langue grecque. Histoire des mots, I ‒ II, Paris. DGE Adrados, F.R. (redactado bajo la dir. de) (1980–), Diccionario Griego-Español, Madrid. Gignac Gignac, F.T. (1976), A Grammar of the Greek Papyri of the Roman and Byzantine Periods, I (Phonology), Milano. HE Gow, A.S.F. / Page, D.L. (1965), The Greek Anthology. Hellenistic Epigrams, I ‒ II, Cambridge. GPh Gow, A.S.F. / Page, D.L. (1968), The Garland of Philip and Some Contemporary Epigrams, I ‒ II, Cambridge. Lampe Lampe, G.W.H. (ed.) (1961), A Patristic Greek Lexicon, Oxford. LfgrE Snell, B. / Erbse, H. et al. (eds.) (1955‒2010), Lexikon des frühgriechischen Epos, I ‒ IV, Göttingen. GVI Peek, W. (ed.), (1955), Griechische Vers‒Inschriften, Berlin. Kühner‒Blass Kühner, R. / Blass, F. (18903‒18923), Ausführliche Grammatik der griechischen Sprache, I ‒ II, Hannover / Leipzig. LSJ Liddell, H.G. / Scott, R. A Greek-English Lexicon, rev. and augm. throughout by Stuart Jones, H. with the assist. of McKenzie, R. and with the cooperation of many scholars, Oxford 1940 + A Supplement, ed. by Barber, E.A. with the assist. of Maas, P. / Scheller, M. / West, M.L., Oxford 1968 + Revised Supplement, ed. by Glare, P.G.W. with the assist. of Thompson, A.A., Oxford 1996. Moulton Moulton, J.H. (19083, 1929 and 1963), A Grammar of New Testament Greek, I (Prolegomena), Edinburgh 19083; II (Accidence and Word-Formation), 1929; III (Syntax), 1963. PCG Kassel, R. / Austin, C. (1983–), Poetae Comici Graeci, Berolini / Novi Eboraci. Schwyzer, GG Schwyzer, E. (19593), Griechische Grammatik, I, Munich. SGO Merkelbach, R. / Stauber, J. (ed.), Steinepigramme aus dem griechischen Osten, Stuttgart / Leipzig 1998 (I); Munich / Leipzig 2001– 2004 (II ‒ V).

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

99

Bibliography Adams, J.N. (1982), The Latin Sexual Vocabulary, Baltimore. Aubreton, R. (1972), Anthologie Grecque. Première partie. Anthologie Palatine, X (livre XI), Paris. Bain, D. (1991), ‘Six Greek Verbs of Sexual Congress (βινῶ, κινῶ, πυγίζω, ληκῶ, οἴφω, λαικάζω)’, in: CQ n.s. 41, 51‒77. — (1997), ‘Two Submerged Items of Greek Sexual Vocabulary from Aphrodisias’, in: ZPE 117, 81‒84. Bakker, E.J. (ed.) (2010), A Companion to the Ancient Greek Language, Oxford. Baldwin, B. (1975), ‘The Epigrams of Lucian’, in: Phoenix 29, 311‒335. Beckby, H. (19652‒19672), Anthologia Graeca, I ‒ IV, München. Blomqvist, J. (1998), ‘The Development of the Satirical Epigram in the Hellenistic Period’, in: Harder, M.A. / Regtuit, R.F. / Wakker, G.C. (eds.) (1998), Genre in Hellenistic Poetry, Groningen, 45‒60. Boehm, I. (2003), ‘Toucher du doigt. Le vocabulaire du toucher dans les textes médicaux grecs et latins’, in: Gaide, F. / Biville, F. (eds.) (2003), Manus medica. Actions et gestes de l’officiant dans les textes médicaux latins. Questions de thérapeutique et de lexique. Actes du Colloque tenu à l’Université Lumière Lyon II, les 18 et 19 septembre 2001, Aix en Provence, 229‒240. Brunck, R.F.P. (1773), Analecta veterum poetarum, II, Argentorati. Calderini, D. (1480), Martialis Epigrammata, Venetiae. Citroni, M. (1975), M. Valerii Martialis Epigrammaton liber I, Firenze. Conca, F. (2004‒2005), ‘Su alcuni epigrammi scoptici’, in: RAAN 73, 323‒334. Degani, E. (1962), ‘Laecasin = ΛΑΙΚΑΖΕΙΝ’, in: RCCM 4, 362‒365. Di Marco, M. (2012), ‘Giocare con i nomi: su alcuni epigrammi di Nicarco (AP V 38; XI 96; XI 169)’, in: RCCM 54, 83‒94. Dover, K.J. (1978), Greek Homosexuality, London. Dübner, F. (1864 and 1872), Epigrammatum Anthologia Palatina cum Planudeis et appendice nova epigrammatum veterum ex libris et marmoribus ductorum, annotatione inedita Boissonadii, Chardonis de la Rochette, Bothii, partim inedita Jacobsii, metrica versione Hugonis Grotii, et apparatu critico, I ‒ II, Parisiis 1864 (I), 1872 (II). Floridi, L. (2007), Stratone di Sardi. Epigrammi, Alessandria. — (2010), ‘Rivisitazione delle convenzioni epigrammatiche nel sottogenere scoptico’, in: MD 65, 9‒42. — (2012), ‘Greek Skoptic Epigram and ‘Popular’ Literature. Anth. Gr. XI and the Philogelos’, in: GRBS 52, 632‒660. — (2013), ‘Considerazioni sul rapporto tra gli epigrammi scoptici sui ‘piccoli’ e le arti figurative’, in: MD 70, 179‒198. — (2014), Lucillio. Epigrammi, Berlin / Boston. — (2016), rev. Schatzmann 2012, in: Prometheus 42, 2016 (in press). Floridi L. / Maltomini, F. (2014), ‘Sui contenuti e l’organizzazione interna di P.Vindob. G 40611 (CPR XXXIII)’, in: Aegyptus 94, 2014 [printed 2016], 19‒62. Galán Vioque, G. (2001), Dioscórides. Epigramas, Huelva. — (2015), rev. Schatzmann 2012, in: Mnemosyne 68, 515‒519. Giangrande, G. (1970), ‘Hellenistic Poetry and Homer’, in: AC 39, 46‒77.

100

Lucia Floridi

Gow, A.S.F. / Page, D.L. (1965), The Greek Anthology. Hellenistic Epigrams, I ‒ II, Cambridge. Guichard, L.A. (forthcoming), ‘Homer in the Greek Epigram of the 1st to 4th Centuries’, in: Manolea, Chr.-P. (ed.), Brill’s Companion to the Reception of Homer from Hellenistic Age to Late Antiquity, Leiden / Boston. Hecker, A. (1843), Commentatio critica de Anthologia Graeca, Lugduni Batavorum. Henderson, J. (19912), The Maculate Muse. Obscene Language in Attic Comedy, New York / Oxford. Heraeus, W. (1915), ‘Προπεῖν’, in: RhM 70, 1‒41 (= Kleine Schriften von Wilheim Heraeus zum 75. Geburtstag am 4. Dezember 1937, ausgewählt und herausgegeben von Johann Baptist Hofmann, Heidelberg 1937, 190‒226). Höschele, R. (2006), Verrückt nach Frauen. Der Epigrammatiker Rufinus, Tübingen. Huschke, I.G. (1800), Analecta critica in Anthologiam Graecam cum supplemento epigrammatum maxima parte ineditorum, Jenae / Lipsiae. Jacobs, F. (1794‒1814), Animadversiones in epigrammata Anthologiae Graecae secundum ordinem Analectorum Brunckii, I ‒ III, Lipsiae. — (1813‒1817), Anthologia Graeca ad fidem codicis olim Palatini, nunc Parisini ex apographo Gothano edita, I ‒ III, Lipsiae. Jocelyn, H.D. (1980), ‘A Greek Indecency and its Students. LAIKAZEIN’, in: PCPhS 26, 12‒66. Keydell, R. (1968), ‘Zur Sprache des Epigrammatikers Lukillios’, in: Philologus 112, 141‒145 (= Kleine Schriften zur hellenistischen und spätgriechischen Dichtung, Leipzig 1982, 315‒319). Kirstein, R. (2002), ‘Companion Pieces in the Hellenistic Epigram’, in: Harder, M.A. / Regtuit, R.F. / Wakker, G.C. (eds.), Hellenistic Epigrams, Leuven / Paris / Dudley (Mass.), 113‒ 135. Lascaris, J. (1494), ᾿Aνθολογία διαφόρων ἐπιγραμμάτων (…). Impressum Florentiae per Laurentium Francisci de Alopa. Linnenkugel, A. (1926), De Lucillo Tarrhaeo, epigrammatum poeta, grammatico, rhetore, diss. Paderbornae. Magnelli, E. (2005), ‘Nicarco, A. P. 11.328: allusioni oscene e allusioni erudite (con osservazioni sulla trasmissione degli epigrammi scoptici)’, in: SemRom 8, 153‒166. — (2007), ‘Meter and Diction. From Refinement to Mannerism’, in: Bing, P. / Bruss, J.S. (eds.), Brill’s Companion to Hellenistic Epigram, Leiden / Boston, 165‒183. Morelli, A.M. (2000), L’epigramma latino prima di Catullo, Cassino. — (2015), ‘Il papiro di Nicarco (P.Oxy. LXVI 4502) e l’epigramma latino’, in: Del Corso, L. / De Vivo, F. / Stramaglia, A. (eds.), Nel segno del testo. Edizioni, materiali e studi per Oronzo Pecere, Florence, 41‒60. Nisbet, G. (2003), Greek Epigram in the Roman Empire. Martial’s Forgotten Rivals, Oxford. Nystrom, B. (2004), An English Translation of the Poetry of Lucillius, a First-Century Greek Epigrammatist, Lewiston / Queenston / Lampeter. Paoli, U.E. (1938), ‘Note petroniane: 1. Laecasin dico; 2. La carne d’orso’, in: SIFC 16, 43‒53. Parsons, P.J. (1987), in: Coles, R.A. / Maehler, H. (eds.), The Oxyrhynchus Papyri, LIV, London, nr. 3725, 82‒84. — (1999), in: N. Gonis et al. (eds.), The Oxyrhynchus Papyri, LXVI, London, nrs. 4501‒4502, 38‒57. Parsons, P.J. / Maehler, H. / Maltomini, F. (2015), The Vienna Epigrams Papyrus (G 40611) (CPR XXXIII), Berlin / München / Boston, 2015.

The Language of Greek Skoptic Epigram of the I ‒ II centuries AD

101

Paton, W.R. (1916‒1918), The Greek Anthology, I ‒ V, London / Cambridge (Mass.). Pöhlmann-Tichy, E. (1982), ‘Zur Herkunft und Bedeutung von κόλλοψ’, in: Tischler, J. (ed.) (1982), Serta Indogermanica. Festschrift für G. Neumann, Innsbruck, 287‒311. Robert, L. (1968), ‘Les épigrammes satiriques de Lucillius sur les athlètes. Parodie et réalités’, in: L’Épigramme grecque, Entretiens Hardt XIV, Vandoeuvres / Geneva, 179‒295 (= Opera Minora Selecta, VI, Amsterdam 1989, 317‒431). Rozema, B.J. (1971), Lucillius the Epigrammatist: Text and Commentary, diss. Madison. Rutherford, W.G. (1881), The New Phrynichus, London. Schatzmann, A. (2012), Nikarchos II. Epigrammata, Göttingen. Schneider, O. (1856), Nicandrea. Theriaca et Alexipharmaca, Lipsiae. Schneidewin, D.F.G. (1842), M. Val. Martialis. Epigrammaton libri, II, Grimae. Schulte, H. (1999), Die Epigramme des Nikarchos, Trier. — (2004), Die Epigramme des Ammianos, Trier. Sekunda, N.V. (1997), ‘The Kylloi and Eubiotoi of Hypata during the Imperial Period’, in: ZPE 118, 207–226. Sens, A. (2004), ‘Doricisms in the New and Old Posidippus’, in: Acosta–Hughes, B. / Kosmetatou, E. / Baumbach, M. (eds.), Labored in Papyrus Leaves: Perspectives on an Epigram Collection Attributed to Posidippus (P. Mil. Vogl. VIII 309), Cambridge (Mass.), 65‒83. Sens, A. (2011), Asclepiades of Samos. Epigrams and Fragments, Oxford. Silk, M. (2010), The Language of Greek Lyric Poetry, in: Bakker 2010, 424‒440. Slater, W.J. (1999), ‘Hooking in Harbours: Dioscurides XIII Gow-Page’, in: CQ n.s. 49, 503‒ 514. Stadtmüller, H. (1894), Anthologia Graeca epigrammatum Palatina cum Planudea, I, Lipsiae. Swete, H.B. (19254), The Old Testament in Greek according to the Septuagint, I, Cambridge. Tarán, S.L. (1985), ‘Εἰσὶ τρίχες. An Erotic Motif in the Greek Anthology’, in: JHS 105, 90‒107. Toup, J. (1790), Emendationes in Suidam et Hesychium et alios lexicographos graecos, II, Oxonii. Tribulato, O. (2010), Literary Dialects, in: Bakker 2010, 388‒400. van Herwerden, H. (1874), ‘Ad Anthologiam Palatinam’, in: Mnemosyne 2, 302‒346. Vergados, A. (2010), ‘Nicarchus AP 11, 328 and Homeric Interpretation’, in: Mnemosyne ser. IV 63, 406‒423. Vulcanius, B. (1536), Onomasticon vocum Latino-Graecarum, prioribus illis glossariis praestantia selectissimorum vocabulorum nihilo inferius, Lugduni Batavorum. Waltz, P. (1928), Anthologie Grecque. Anthologie Palatine, II (livre V), Paris.

Form and Design

Regina Höschele

“Unplumbed Depths of Fatuity?” Philip of Thessaloniki’s Art of Variation Philip of Thessaloniki is one of those authors whose work has elicited much scorn and little interest. According to the verdict of Gow and Page, the editor of the second Garland is a “second-rate dealer in second-hand materials”¹ and himself a “dull writer”, who composed “two or three epigrams of good quality”, while the remainder of his poems include “some … which sound hitherto unplumbed depths of fatuity”.² Over the past few decades, scholars have reappraised many a writer similarly disparaged by previous critics; imitation – a crucial element of post-Classical Greek literature – is no longer considered a fourletter word. Rather than being viewed as a sign of deplorable epigonality, it is seen as a literary phenomenon interesting in its own right.³ Philip’s oeuvre, however, remains largely unexplored, both with regard to his editorial undertaking and to his own poetry – a gap I am currently trying to fill with a monograph on Philip’s Garland. In this paper, I would like to present some observations on his art of variation, an essential feature of the genre,⁴ which is of particular prominence in Philip’s epigrammatic cosmos. As we will see, Philippan variation occurs both on a micro- and a macrotextual level, for he not only rewrites numerous individual epigrams but also systematically varies Meleager’s way of interweaving poems. Significantly, both forms of variation – the textual and the editorial – inform each other and are intrinsically linked through a game of letters, which underlies the Stephanos’ overall design. Let us start by considering one specific example of Philippan variatio, his epigram on a boy meeting a gruesome death in the Thracian river Hebrus (37 GP = AP 9.56), which is closely modeled on Statilius Flaccus 4 GP (AP 7.542): Ἕβρου χειμερίοις ἀταλὸς κρυμοῖσι δεθέντος κοῦρος ὀλισθηροῖς ποσσὶν ἔθραυσε πάγον· τοῦ παρασυρομένοιο περιρραγὲς αὐχέν’ ἔκοψεν θηγαλέον ποταμοῦ Βιστονίοιο τρύφος.

 Gow/Page  II, .  Gow/Page  II, .  For a good overview of common scholarly prejudices against Greek imperial literature and an insightful discussion of mimesis in texts of this period, cf. Whitmarsh ,  – .  On epigrammatic variatio, cf. Ludwig , Tarán  and Laurens ,  – ; see also Gutzwiller ,  –  on Antipater of Sidon as a main practitioner of variatio.

106

Regina Höschele

καὶ τὸ μὲν ἡρπάσθη δίναις μέρος, ἡ δὲ τεκοῦσα λειφθὲν ὕπερθε τάφῳ μοῦνον ἔθηκε κάρα, μυρομένη δὲ τάλαινα ‘τέκος, τέκος’ εἶπε ‘τὸ μέν σου πυρκαϊή, τὸ δέ σου πικρὸν ἔθαψεν ὕδωρ’. When the Hebrus was bound by winter’s frosts, a tender boy, with slipping feet, broke through its ice. As he was swept away, a sharp fragment that had broken out from the Bistonian river pierced through his neck. One part of him was carried away by the flood; but his mother laid the only thing left above the surface – his head – into a grave. And bursting out in tears, she wretchedly cried: “My child, my child, one part of you the funeral pyre, the other one the cruel water has buried”. (Statilius Flaccus AP 7.542) Ἕβρου Θρηϊκίου κρυμῷ πεπεδημένον ὕδωρ νήπιος εἰσβαίνων οὐκ ἔφυγεν θάνατον· ἐς ποταμὸν δ’ ἤδη λαγαρούμενον ἴχνος ὀλισθὼν κρυμῷ τοὺς ἁπαλοὺς αὐχένας ἀμφεκάρη· καὶ τὸ μὲν ἐξεσύρη λοιπὸν δέμας, ἡ δὲ μένουσα ὄψις ἀναγκαίην εἶχε τάφου πρόφασιν. δύσμορος, ἧς ὠδῖνα διείλατο πῦρ τε καὶ ὕδωρ· ἀμφοτέρων δὲ δοκῶν οὐδενός ἐστιν ὅλως. A child, stepping out onto the frost-shackled water of Thracian Hebrus, did not escape death. When he slipped with his foot into the river, which was already thawing, his soft neck was cut off all around by the ice. The rest of his body was swept away, but his head remained and gave necessary reason for a burial. Miserable is she whose child was divided between fire and water; seeming to belong to both, he belongs to neither fully. (Philip AP 9.56)

In the first half of his 8-line epigram, Flaccus narrates how a young boy fell into the frozen Hebrus and was decapitated by a sharp piece of ice; the second half focuses on the paradoxical separation of his head and body: while the latter is carried away by the river and thus buried in water, the former alone is left to be cremated and interred by the child’s mother, who bitterly bemoans this two-part “burial”. Maintaining the length and bi-partite structure of Flaccus’ poem, Philip markedly starts both halfs with the same words as his model: Ἕβρου (v. 1) – καὶ τὸ μὲν (v. 5). Whereas Flaccus builds up tension by describing how the child slips, breaks through the ice (vv. 1– 2), is swept away and then beheaded (v. 3), Philip’s narrative is somewhat less dramatic, as he matter-of-factly gives away the child’s death at the end of the first couplet (οὐκ ἔφυγεν θάνατον). A reader familiar with the original would, of course, already anticipate the outcome of the boy’s venture upon the ice; the statement of his death in line 2 therefore does not so much “spoil” the dramatic effect as reflect the reader’s knowledge of the boy’s fatal end. By noting that the river was “already thawing” (ποταμὸν δ’ ἤδη λαγαρούμενον, v. 3) Philip, moreover, makes explicit the reason for the breaking of the ice, which is only implicit in Flaccus. Last but not least,

“Unplumbed Depths of Fatuity?” Philip of Thessaloniki’s Art of Variation

107

he replaces another dramatic element, the mother’s direct speech, at the conclusion of his model with a reflection of his narrative persona on that mother’s loss, which likewise centers upon the corpse’s division between water and fire. Flaccus replicates this physical separation by placing the words πυρκαϊή and ὕδωρ at the beginning and end of his last verse respectively; Philip, on the other hand, places both nouns together, πῦρ τε καὶ ὕδωρ, at the end of v. 7, which underlines his final point that the corpse seems to belong jointly to these two elements, if not fully to either one.⁵ The division of one body into two halves may, I suggest, be mirrored by the theme’s treatment in two epigrams. This correspondence between the subject matter and its poetic articulation is further underlined by the distinctly bipartite structure of each poem (with the verbatim repetition of each half’s incipit in Philip’s variation). Even as the boy’s body is split into “the one part” (τὸ μέν, Flaccus vv. 5/7; Philip v. 5) and “the other” (τὸ δέ, Flaccus v. 8), so do Flaccus’ and Philip’s versions constitute τὸ μέν and τὸ δέ respectively. One might say, then, that these two poems reflect the doubling effect of variation in general through their specific structure and subject matter. But whereas the head and body of the boy are physically separated, the two texts were, as we will see, in all likelihood joined as a pair within the context of Philip’s anthology. On a verbal level, Philip echoes many of Flaccus’ expressions, while replacing others with chosen synonyms: Flaccus Ἕβρου (v. ) κρυμοῖσι (v. ) δεθέντος (v. ) κοῦρος (v. ) ὀλισθηροῖς ποσσὶν (v. ) παρασυρομένοιο (v. ) αὐχέν’ ἔκοψεν (v. ) ποταμοῦ (v. ) καὶ τὸ μέν (v. ) τάφῳ (v. )

Philip Ἕβρου (v. ), same sedes κρυμῷ (v. ) πεπεδημένον (v. ) νήπιος (v. ), same sedes ἴχνος ὀλισθών (v. ) ἐξεσύρη (v. ) αὐχένας ἀμφεκάρη (v. ) ποταμόν (v. ) καὶ τὸ μὲν (v. ), same sedes τάφου (v. )

 It is indicative of Gow-Page’s bias against Philip that they characterize this last couplet as “among the silliest in the present collection” ( II, ). The corpse’s division between water and funerary pyre varies an earlier epigrammatic motif, the division of a dead body between land and sea. Cf. Leonidas AP . on Tharsys, who was half-eaten by a fish, and Antipater of Thessaloniki (AP .) on a man likewise eaten by fish, who declares in his epitaph that he does not belong wholly to either land or sea (οὐδετέρης ὅλος εἰμὶ θανὼν νέκυς, v.  ~ Philip, v. : οὐδενός ἐστιν ὅλως). On this theme in sepulchral epigram, cf. Tueller (forthcoming), who also discusses Philip’s poem.

108

Regina Höschele

κάρα (v. ) τάλαινα (v. ) πυρκαϊή … ὕδωρ (v. )

ὄψις (v. ); κάρα is echoed acoustically in ἀμφεκάρη (v. ) δύσμορος (v. ) πῦρ τε καὶ ὕδωρ (v. )

He, moreover, introduces a special twist after the bucolic dihaeresis of line 5 by exchanging the participial phrase ἡ δὲ τεκοῦσα with ἡ δὲ μένουσα, which, it turns out, does not refer to the mother but to the head (ὄψις) as sole remainder of the boy (it thus picks up the aorist passive participle λειφθέν, which stands at the beginning of v. 6 in Flaccus, i. e. in the same sedes as Philip’s ὄψις). This verbal substitution goes hand in hand with the elimination of the mother as an active agent in Philip’s poem. Needless to say, the motif of a severed head swimming in the Hebrus recalls the legend of Orpheus, whose head floats down the Thracian river into the sea and from there on to Lesbos⁶ – in the case of our two epigrams, by contrast, it is the head that remains behind, while the body is carried away. Another variation of the motif is to be found in an apocryphal Christian story about the death of Salome, first attested in a pseudepigraphical letter from King Herod to Pontius Pilate⁷: like the boy in our epigram pair, Salome was playing on a frozen river when she broke through the ice. Her mother tried to hold on to her by her head, which, however, was severed by the river’s icy surface. Since her daughter’s body was swept away, the mourning mother was left with nothing but her head. The manner of Salome’s death is evidently conceived as an appropriate punishment for the beheading of John the Baptist, for which she was responsible, with the river as the former locus of John’s baptisms becoming a tool of divine vengeance⁸. According to Wolfgang Speyer, this Christian legend might in fact have been inspired by a pagan anecdote, such as we find reflected in the Palatine Anthology, if not by our two epigrams themselves.⁹ If he is right – and  Cf. Phanocles fr. . on Orpheus’ decapitation and the journey of his head and lyre to Lesbos; their burial on the island marks the beginning of its association with music and song, serving as a memorable aetion for the celebrated tradition of Lesbian lyric.  The Greek text, transmitted in a th-century manuscript, was first edited by James ,  – ; the two letters are, however, also preserved in a Syrian version of the th/th century and probably go back to Late Antiquity; cf. Speyer ,  –  Anm . For other versions of the story in later Greek texts and the Legenda Aurea, cf. James , xlv-xlvii.  On the principle of talio identica (the correspondence of punishment and crime) in this story, cf. Speyer , . As Nicephorus Callistus (th cent, hist. eccl. .) remarks, the ice acts in lieu of a sword (οὐ ξίφει, ἀλλὰ κρυστάλλῳ). Note, too, how the reference to the girl’s childlike play (παίζουσα) may carry a sinister note, evoking her seductive dance that led to John’s death (the verb used in the NT is ὀρχέομαι [Matthew :, Mark :], but the letter’s author seems to play with the fact that παίζω can also refer to the act of dancing).  Cf. Speyer ,  – .

“Unplumbed Depths of Fatuity?” Philip of Thessaloniki’s Art of Variation

109

there is good reason to think so –, this would constitute a fascinating variation of an epigrammatic motif in a non-epigrammatic context, brilliantly inserted into a narrative framework likewise concerned with the theme of decapitation.¹⁰ But let us return to the motif’s variation within the realm of epigram! I have started my considerations with the supposition that Philip was the one imitating Flaccus, not the other way round. Reitzenstein, Beckby and Gow-Page likewise view Philip as the imitator,¹¹ whereas Geffcken, in his RE-article on Statilius, adduces Flaccus’ otherwise “dependent character” as proof for his role as copyist¹² (hardly a valid argument, not least since the same could and has been said about Philip). I should note that we have no external information about the author Statilius Flaccus, under whose name 17 epigrams have come down to us.¹³ Flaccus is neither named among the authors listed in the preface to Philip’s Garland, nor are any of his poems transmitted within securely Philippan sequences of the AP.¹⁴ However, we have, I submit, good cause to believe that Philip composed his epigram as a response to Flaccus’ and juxtaposed both within his Garland. Determining the relative chronology between two texts of unknown date is, of course, always a tricky business, but there are several arguments speaking in

 Another narrative version of the motif is to be found in Achilles Tatius’ novel Leucippe and Clitophon (.): led to believe that his beloved has been decapitated, the hero fishes “her” body out of the water and bemoans Leucippe’s double death (θάνατον διπλοῦν), the unequal division of her corpse between land and sea. Though he seemingly possesses the larger part of her, he would prefer to have the smaller one (i. e. her head), as it contains “all” of his beloved. For a discussion of this passage vis-à-vis the epigrammatic tradition, cf. McGill .  Cf. Reitzenstein ; Beckby , ; Gow/Page  II, .  Cf. Geffcken , : “So schwierig es nun, so unmöglich es oft bleibt, bei gleichzeitigen oder chronologisch nicht bestimmbaren Autoren Original und Nachahmung festzustellen, so nahe legt uns doch der sonstige unselbständige Charakter des Epigrammatikers, in den genannten Fällen nur in ihm den Kopisten zu sehen”. He concludes his evaluation of Statilius Flaccus with the observation: “So ist S. zwar nicht besser als seine Genossen, doch auch nicht schlechter als diese ganze Kumpanei von Literaten”.  A Latin version of his poem on the decapitated boy is, inter alia, attributed to Germanicus ( Riese), in which case Flaccus “must have flourished not much if at all later than the first decade AD” (Gow/Page  II, ). However, this terminus ante quem is problematic, as the Latin epigram in all likelihood is not a product of Antiquity, but was composed by the Carolingian writer Paulus Diaconus (cf. Rubensohn ).  In putting together his anthology, Cephalas alternated between creating thematic groupings of his own and appropriating entire blocks from Meleager, Philip and Agathias without rearranging the epigrams; cf. Cameron , . The sequences left intact by Cephalas are analyzed in Lenzinger . There are  sections, comprising  epigrams, that go back directly to Philip’s Garland, while another  –  epigrams from his collection are transmitted in other parts of the AP; cf. Gow/Page  I, xii-xiii. From the  authors securely attributed to Philip’s Garland, only  are named in its preface; cf. Gow/Page  I, xxi.

110

Regina Höschele

favor of Philippan imitatio. As mentioned above, the less dramatic character of Philip’s narrative may be explained by his (and the reader’s) prior familiarity with the story – we know the outcome, hence no need for suspense. In itself this would hardly stand as evidence for Flaccus’ priority, but the hypothesis is further supported by the substitution of ἡ δὲ τεκοῦσα with ἡ δὲ μένουσα: it is, to my mind, more likely that Philip surprised the reader by replacing a reference to the mother with a reference to the head (with ὄψις serving as an aprosdoketon) than that Flaccus transformed the head into a mother.¹⁵ Most importantly, though, the fact that the two epigrams start with the same word, Ἕβρου, strongly points to Philip as the imitator, since it is an essential element of his variation technique to begin rewritings with the same letters as their models. Unless we were to suppose that Flaccus paid homage to Philip by replicating this practice (a possible, but rather unlikely scenario), the identity of the two incipits confirms the latter as the imitating author. Why does Philip commonly begin his variations with the same letter as their archetypes? Identical incipits or initial letters are not unprecedented among Hellenistic Parallelgedichte,¹⁶ but they are hardly a regular feature. In the case of Philip, this striking characteristic is intrinsically linked with the organization of his Garland, which was arranged κατὰ στοιχεῖον, i. e. alphabetical according to the initial letter of each poem. While critics had long dismissed this arrangement as sterile pedantry,¹⁷ two scholars in the 1960s, Eduard Hirsch and Alan Cameron, recognized – independently from one another – that Philip’s design is anything but artless.¹⁸ As Cameron remarks, “the grouping of poems by initial letter was merely a preliminary, the framework of his system. It was Philip’s variation on Meleager’s preliminary division of material into four basic categories before addressing himself to the more subtle task of arranging the individual poems inside those categories. Philip too had an internal and an external system.”¹⁹ As he himself declares in the proem to his collection, Philip wove his garland in rivalry (ἀντανέπλεξα, AP 4.2.3) with Meleager (τοῖς  As Peter Bing has pointed out to me, the fact that Philip seems to avoid the more common poetic word for head, κάρα, used by Flaccus, even while he plays on it with ἀμφεκάρη, also suggests Philip as the imitating author.  Cameron ,  lists several examples of thematically related Hellenistic epigrams starting with the same word or letter: AP . (ἀκρίδα),  (ἀκρίς),  (ἀχήεις) on crickets; AP . –  (οὐκέτι) on dead animals; AP . –  (ναυηγοῦ),  (ναυτίλοι),  (ναυηγόν) on shipwrecks; AP . –  (ἔγχει) toast for Heliodora; AP . –  (ὄρθρε) on the morning star. Cf. also Hirsch ,  –  with further examples.  On negative views regarding his arrangement, cf. Höschele (forthcoming a).  Cf. Hirsch  and Cameron .  Cameron , .

“Unplumbed Depths of Fatuity?” Philip of Thessaloniki’s Art of Variation

111

Μελεαγρείοις ὡς ἴκελον στεφάνοις, AP 4.2.4) – similar to his, not identical, as scholars would have it who saw in Philip nothing but a slavish imitator of his predecessor. However, just as individual epigrammatists, including Philip, frequently vary poems by earlier writers, so does Philip’s anthology as a whole constitute a macro-variation of sorts on the first Garland.²⁰ While an alphabetical ordering in and of itself may strike us as mechanical and potentially tedious, it does not preclude additional and subtler modes of arrangement: the alphabet just provides an external framework within which Philip was free to connect poems through verbal or thematic links, just as Meleager had done. In opting for this organization, Philip in fact faced a greater challenge than his predecessor, as he could only group together poems with the same (or, possibly, subsequent) initial letters, which considerably limited his choices for the creation of thematic sub-groups. This self-imposed formal constraint adds to the artistry of Philip’s enterprise and makes his achievement all the more remarkable.²¹ The creation of meaningful collocations was, at any rate, facilitated by Philip’s combination of excerpted texts with poems of his own. It was Wifstrand who first noted, back in 1926, that Philip routinely begins variations with the same letter as their models so as to be able to juxtapose the two.²² The examples adduced by him have been further supplemented by Cameron and Hirsch.²³ Building upon their observations, we may count 12 close variations on various themes starting with the same letter, if not the same word²⁴: Philip’s poem . Αἴλιος

Varied epigram . Αἴλιος

by Apollonides

*. . .

. . .

Antiphanes Flaccus Antiphilus

γραμματικοί Ἕβρου ἔζευξ’

γραμματικῶν Ἕβρου εἰπέ

Theme Suicide of Roman soldier²⁵ Against critics Boy decapitated by ice Mole in Puteoli

 As Cameron ,  –  and ,  –  plausibly argued, Meleager divided his Stephanos into four books containing, respectively, erotica, epitumbia, anathematica and epideictica. Within these books, epigrams are ordered according to thematic sub-groups and copies are typically juxtaposed with their respective models. For a detailed structural analysis, cf. Gutzwiller ,  – .  On Philip’s poetics of editing cf. Höschele (forthcoming a) and, with a special focus on the novelty of his undertaking vis-à-vis Meleager, Höschele (forthcoming b).  Cf. Wifstrand ,  n. .  Cf. Hirsch ,  –  and Cameron ,  – .  The cases in which model and copy are juxtaposed are highlighted through bold print; an asterisk marks the Philippan poems which are not placed after their model, but precede it.  For a discussion of this pair, cf. Höschele (forthcoming a).

112

Regina Höschele

Continued Philip’s poem . ἡ . Ἥρη

Varied epigram . ἤμην . Ἥρακλες

by Antiphilus Geminus

. . *. . . *.

. . . . . .

Antipater Antipater Antiphilus Secundus Antiphilus Argentarius

Κύπρι οὔρεα²⁶ πουλύ συλήσαντες τίς ὡραίας

καί οὐχί πορφυρέαν σκυλοχαρεῖς τάν ὥριμος

Theme Hetaira-ship Herakles statue without weapon Spartan Kypris Tomb of Themistokles Leonidas’ corpse Erotes with weapons Timomachus’ Medea Priapus guarding figs²⁷

There are, as Hirsch recognized, two further close variations, which begin with subsequent letters and might thus have been placed together in the transition from one letter section to the next, between eta and theta as well as chi and phi ²⁸: *. .

ἥρως χαίροιθ’

. .

Θεσσαλέ φεύγεθ’

Antiphilus Antipater

Tomb of Protesilaos Against critics

In addition, we encounter seven instances in which Philip begins epigrams on the same general theme with identical incipits or the same letter: .

ἄρτι

.

Αὐσονίη

Antipater

. . . . . *.

εἰ ἐν²⁹ ἠπείρῳ Ἰκαρίην πηρός³⁰ ταῦροι

. . . . . .

εἰ ἔστην ἠιόνιον Ἰκάρου πηρός ταύρῳ

Antiphanes Honestus Philip adespoton Antipater Adaeus

Death instead of wedding Money-lending Thebes Shipwrecked corpse Ikarian sea Blindness Bulls

As can be seen from this survey, most of the epigram pairs – mainly those transmitted within one of the sections going back directly to Philip’s Garland – are still juxtaposed in the Palatine Anthology (in the majority of cases, though not all, Philippan variations are placed after their respective model). There are

 The poem is also ascribed to Alpheius of Mitylene.  In Höschele (forthcoming b) I discuss how this pair, together with a series of further apotropaic poems starting with omega, serves to mark the end of the collection.  Cf. Hirsch , ; Cameron ,  only mentions the transition from eta to theta.  This epigram is, in fact, a close variation of Honestus AP ., which starts with Ἁρμονίης.  Philip’s epigram is a variation of Antiphilus AP .b starting with ἄμφω, see below.

“Unplumbed Depths of Fatuity?” Philip of Thessaloniki’s Art of Variation

113

some instances in which original and copy are transmitted separately (as are Flaccus’ and Philip’s epigrams on the decapitated boy), since the poems are preserved outside of Philippan sequences. It is, however, fair to assume that they, too, were originally placed next to or in close proximity to one another. The underlying principle of arrangement is easily evident. We should, however, note that not all of Philip’s variations replicate the initial letter of their models. I have counted six cases, in which original and copy came to stand in separate sections of the Garland due to their different initial letters. Philip’s poem . σπερμοφόρον . ῥοιήν . πηρός . ἐν

Varied epigram . βουστρόφον . ἀρτιχανῆ .b ἄμφω³¹ . Ἁρμονίης

by Antiphilus Zonas Antiphilus Honestus

.

ἠρίθμει

.

καί

. . .

Antipater Antipater Bianor

εἶχεν πλοῦτος ἠνίδε

Theme Farmer’s dedication Gardener’s dedication Lame and blind Good and evil in Theban myth Suicide of Aristeides Cow plowing/suckling

Hirsch concluded that these variations must have been written before Philip conceived of an alphabetically organized anthology, which presupposes that Philip would have followed only a single mode of variation once he had developed the design for his Garland. But should we not rather view them as variations on his general technique of variation and a means of creating thematic links across various letter sections³²? Of particular interest here is Philip’s epigram on good and evil in Theban myth (AP 9.253, starting with epsilon), which rewrites Honestus, AP 9.216 (starting with alpha). While varying a poem in the alpha-section, Philip’s epigram is, as we have seen above, placed in proximity to another thematically related epigram by Honestus in the epsilon-section, AP 9.250, which deals with the fall of Thebes: built with the help of a lyre, the city was destroyed under

 Note that this epigram varies a poem by Plato the Younger (AP .a) starting with ἀνέρα.  Hirsch ,  observes: “Diese Nachahmungsgedichte werden vor dem Gedanken an einen alphabetischen Kranz entstanden sein: Denn was hätte einen Philipp sonst später gehindert, die gleiche Initiale zu wählen […]?” This conclusion strikes me as erroneous, since Philip could just as well have modified his imitations to begin with the same initial as their models, if he had wanted to juxtapose the two within the context of his anthology. It does, at any rate, not lead anywhere to speculate about the relative chronology between Philip’s composition of individual poems and the genesis of his Garland; what matters – as in the case of single-authored libelli (cf. Höschele , ) – is the resultant collection as a whole.

114

Regina Höschele

the accompaniment of a flute.³³ Significantly, this poem is preceded by four epigrams on wine and Dionysus (AP 9.246 – 249), whose myth is closely connected with Thebes, and immediately followed by an epigram against a book-eating insect, which picks up the theme of destruction. Poem incipit author . ἐθραύσθης Argentarius . εὐθηλῆ

Philip

. εἰ Boethus . εὐπέταλον Q. Maecius . ἔστην Honestus . ἐχθίστη . ἐς . ἐν

content broken wine bottle pouring forth Bromios like Semele Torn-out plane revived with vine: Lyaios, who makes others fall, helped tree stand straight Pylades playing Dionysus Pan guarding a vineyard Destruction of Thebes

Euenus Destruction of books anon./Bianor Man chased by wolves Philip Good and evil in Theban myth

key motif Dionysus Dionysus Dionysus Wine Thebes/ Destruction Destruction Death Thebes

Philip’s copy may be separated from its model, but it is well placed in a sequence of tightly interwoven epigrams. Even if epigrams originally belonging to this sequence should have dropped out or were moved by Cephalas to a different book,³⁴ the overall thematic coherence of this series is still clearly manifest. Consider, too, Philip’s poem on the suicide of Aristeides (AP 9.255), who kills himself after losing his only cow and sheep. Starting with ἠρίθμει, it offers a variation on a pair of epigrams by Antipater of Thessaloniki, whose incipits begin with epsilon (AP 9.149: εἶχεν) and pi (AP 9.150: πλοῦτος) respectively (cf. table above). Philip’s version is meaningfully placed between another Philippan epigram (AP 9.254) on a woman who loses all her children, even an adopted one, and an epigram by Antiphanes on an old appletree which loses its only remaining apple to a caterpillar (AP 9.256).

 Cf. AP .. –  ἔστην ἐν φόρμιγγι, κατηρείφθην δὲ σὺν αὐλῷ / Θήβη· φεῦ Μούσης ἔμπαλιν ἁρμονίης. This motif, treated here in  verses, is also evoked briefly in Honestus’ poem on good and evil in Theban myth (AP .. – : τειχομελὴς κιθάρη· / ἀλλ’ αὐλὸς δύσμουσος) as well as Philip’s variation thereof (AP .. – : τείχεα χορδαῖς / ἔστη καὶ λωτοῖς ἔστενε λυόμενα).  Importantly, Philip’s collection did not encompass several thematically different books, each organized by alphabet, but epigrams of all types were intermingled in one continuous alphabetical series, cf. Hirsch ,  –  and Cameron ,  – . Even though Cephalas excerpted epigrams in a linear manner, he had to distribute them among his books according to subject matter.

115

“Unplumbed Depths of Fatuity?” Philip of Thessaloniki’s Art of Variation

Poem incipit . ἡ . ἠρίθμει

author Philip Philip

key motif Loss Loss

. ἥμισυ

content Woman loses all her children, even an adopted one Having lost his only cow and sheep, Aristeides kills himself Antiphanes Caterpillar deprives old tree of its only apple

. . . . . .

Apollonidas Antiphanes Bianor Secundus Epigonus Philip

Loss, Age Murder Μurder No loss Age Age Loss

ἡ ἡ ἤριπεν ἡ ἡ ἠρίθμουν

. ἡ

Antiphilus

Spring polluted by murder variation of .³⁵ House collapses, child survives Old Lais Old vine, wrinkled grapes Aristodike loses  of her  children to illness,  to the sea Old Euboule’s stone oracle

Age

The theme of old age, which first resounds in Antiphanes’ poem on the appletree, is echoed in three of the following epigrams (AP 9.260, 261 and 263), while the theme of a mother’s loss is picked up by Bianor AP 9.259, where a child miraculously survives the collapse of a house, and Philip, AP 9.262 on a woman named Aristodike, who loses three of her six children to illness and three to the sea. Strikingly, this Philippan poem starts with the same verb as his variation on Aristeides’ suicide (ἠρίθμει – ἠρίθμουν; note, too, that the names of both protagonists begin with Arist‐). Thematically, it offers, moreover, a variation of his epigram on the decapitated boy: just as the boy’s corpse was divided between water and fire, so one half of Aristodike’s children were given to the earth, the other to the sea (as v. 3 states, ὕδωρ and χθών fought together against her). As these examples demonstrate, Philip’s editorial technique involves not just the collocation of copy and original, but also the interweaving of poems more broadly related by theme. Importantly, those of his variations that do not start with the same letter as their archetype are all placed in a later part of the collection – never in an earlier part – than their respective model (beta → sigma, alpha → rho, alpha → pi, alpha → epsilon, epsilon → eta, eta → kappa), which means that a reader perusing the Garland in a linear manner, heading from alpha to omega, first encounters the originals, then their copies.³⁶ He is thus invited to

 On the placement of this epigram pair within the context of Philip’s anthology, cf. Höschele (forthcoming a).  That Philip conceived his collection for a uni-directional reading from alpha to omega (and not the other way round) is not least shown by the fact that he starts his own proem (AP .) with an alpha. In addition, the poems contained in the omega-section very much evoke a sense of an ending (cf. n. ).

116

Regina Höschele

perceive thematic links not just within a given sequence but also across the boundaries of letter sections. As I hope to have shown with this brief survey, it is still possible to detect a sophisticated design behind the seemingly mechanical alphabetization of Philip’s Garland. It is the artfulness of his arrangement – not its fatuity – that remains largely unplumbed. While the initial letters of the epigrams included in his poetic cosmos hardly played a role in their composition, they became significant in the process of Philip’s anthologization and constitute a key factor in his personal technique of variation.³⁷

Abbreviations AP Anthologia Palatina. GP Gow, A./Page, D. (eds.) (1968), The Greek Anthology: The Garland of Philip and Some Contemporary Epigrams, 2 vols., Cambridge.

Bibliography Beckby, H. (1957), Anthologia Graeca, Buch VII-VIII. Griechisch-Deutsch, Munich. Cameron, A. (1968), ‛The Garlands of Meleager and Philip’, in: GRBS 9, 323 – 349. — (1993), The Greek Anthology from Meleager to Planudes, Oxford. Geffcken, F. (1931), ‛Statilius, 19a’, in: RE Suppl. 5, 993 – 994. Gow, A.S.F. / Page, D.L. (eds.) (1968), The Greek Anthology: The Garland of Philip and Some Contemporary Epigrams, 2 vols., Cambridge. Gutzwiller, K. (1998), Poetic Garlands: Hellenistic Epigrams in Context, Berkeley / Los Angeles / London. Hirsch, E. (1966), ‛Zum Kranz des Philippos’, in: Wissenschaftliche Zeitschrift der Universität Halle 15, 401 – 417. Höschele, R. (2010), Die blütenlesende Muse: Poetik und Textualität antiker Epigrammsammlungen, Tübingen. — (forthcoming a): ‛A Garland of Freshly Grown Flowers: The Poetics of Editing in Philip’s Stephanos’, in: C. Carey / M. Kanellou / I. Petrovic (eds.), Reading Greek Epigram from the Hellenistic to the Early Byzantine Era, Oxford. — (forthcoming b), ‛‘Harvesting from a New Page’: Philip of Thessalonike’s Editorial Undertaking’, in: Aitia 6. James, M.R. (1897), Apocrypha Anecdota. Second Series, Cambridge. Laurens, P. (22012), L’abeille dans l’ambre. Célébration de l’épigramme de l’époque alexandrine à la fin de la Renaissance. 2e éd. revue et augmentée, Paris.

 I would like to thank Peter Bing and Niklas Holzberg for their very helpful comments on this paper.

“Unplumbed Depths of Fatuity?” Philip of Thessaloniki’s Art of Variation

117

Lenzinger, F. (1965), Zur griechischen Anthologie, Bern. Ludwig, W. (1968), ‛Die Kunst der Variation im hellenistischen Liebesepigramm’, in: L’épigramme grecque, Entretiens sur l’antiquité classique 14, Geneva, 299 – 348. McGill, S. (2000), ‛The Literary Lives of a Scheintod: Clitophon and Leucippe 5.7 and Greek Epigram’, in: CQ 50, 323 – 326. Reitzenstein, R. (1909), ‛Flaccus, 5’, in RE 6.2, 2433 – 2434. Rubensohn, M. (1893), ‛Eine Übersetzung des Paulus Diaconus aus der Griechischen Anthologie’, in: Neue Jahrbücher für Philologie und Pädagogik 147, 764 – 765. Speyer, W. (1989), ‛Der Tod der Salome’, in: W. Speyer, Frühes Christentum im antiken Strahlungsfeld: Ausgewählte Aufsätze 1, Tübingen, 59 – 63 (= JbAC 10, 1967, 176 – 180). Tarán, S.L. (1979), The Art of Variation in the Hellenistic Epigram, Leiden. Tueller, M. (forthcoming): ‛Sea and Land: Dividing Sepulchral Epigram’, in: C. Carey / M. Kanellou / I. Petrovic (eds.), Reading Greek Epigram from the Hellenistic to the Early Byzantine Era, Oxford. Whitmarsh, T. (2001), Greek Literature and the Roman Empire: The Politics of Imitation, Oxford. Wifstrand, A. (1926), Studien zur griechischen Anthologie, Lund.

Gregory O. Hutchinson

Pentameters*

Epigrams should not be thought of too much as a self-contained genre. For the most part, Greek epigrams are a subset of the genre—or super-genre—of elegiacs. At all events, non-epigrammatic elegiacs form an important point of comparison, particularly when we are looking at metre; and metre appears to be the most important aspect of genre in ancient conceptions of non-dramatic poetry.¹ Callimachus is unusual among Greek poets in presenting us with substantial remains in both epigrammatic and non-epigrammatic elegiacs; in any case, conversation about Hellenistic epigrams tends to drift back to Callimachus. This paper will be chiefly concerned with a feature investigated in an important article by the late S.R. Slings: the arrangement whereby in a pentameter the word at the caesura and the word at the end of the line agree with each other (or are in apposition), as in Call. Ep. 5.12 ῥέζειν, καὶ Ϲμύρνηϲ ἐϲτὶν ἀπ᾿ Αἰολίδοϲ or 18.2 ναῦν ἅμα καὶ ψυχὴν εἶδεν ἀπολλυμένην (or 20.2 δυομένου Βαϲιλὼ κάτθανε παρθενική).² Slings’s immediate purpose was to show that the Tattoo Elegy (SSH 970) was not by Hermesianax, since its handling of this feature was so different. Huys’s cautious attribution of the poem to Hermesianax had not actually seemed to possess a strong basis.³ However, this argument of Slings’s against the attribution is not in itself inescapable, since one poet could have adopted two different approaches to the feature (as I shall call it) in the two poems the Leontion and the Tattoo Elegy. We will see such differences within the work of Callimachus and of Catullus. But beyond tattoos, this distinguished scholar performed the valuable service of considering the feature more closely than before, of collecting some statistics, and indeed of observing the difference between Callimachus’ epigrams and his other elegiac poetry. This paper will attempt to take the investigation a little further; and in particular it will take issue with the claim that Callimachus ignored the feature in his epigrams.⁴

* Many thanks to Professor A. Rengakos, Professor E. Sistakou and Ms. K. Zianna for organizing the conference and this volume.  Cf. Hutchinson a, and b,  – ,  – , etc.  Slings ,  – , . Earlier note Heyken ,  – .  Huys ,  – ,  – . Scepticism on grounds of style: Hutchinson , .  Slings , . Like Slings, I include participles and apposition in instances of the feature; it would seem artificial to exclude them. For discussion of the status of participles, see Bentein ; such debates go back to antiquity, as the name of course indicates: participles have a share in different categories (cf. e. g. D. T.  p.  Uhlig; Plut. Quaest. Plat. c-d).

120

Gregory O. Hutchinson

Callimachus’ Aetia reaches a climactic moment—though in amorous not heroic narrative—when Acontius is able to marry Cydippe. (Item 1). κ̣ἦν αὖ ϲῶϲ ̣ [̣ ]̣ λοιπόν, ᾿Aκόντ̣ι ̣ε, ϲεῖο μετελθεῖν ̣ ̣ ̣ ̣ ̣ ̣ η̣ν ἰδίην ἐ̣ϲ ̣ Διονυϲιάδα. χἠ θεὸϲ εὐορκεῖτο, καὶ ἥλικεϲ α̣ὐ̣τίχ᾿ ἑταίρηϲ ἦιδον ὑμηναίουϲ οὐκ ἀναβαλλομένουϲ. οὔ ϲε δοκέω τημοῦτοϲ, ᾿Aκόντιε, νυκτὸϲ ἐκείνηϲ ἀντί κε, τῆι μίτρηϲ ἥψαο παρθενίηϲ, οὐ ϲφυρὸν Ἰφίκλειον ἐπιτρέχον ἀϲταχύεϲϲιν οὐδ᾿ ἃ Κελαινίτηϲ ἐκτεάτιϲτο Μίδηϲ δέξαϲθαι, ψήφου δ̣᾿ ἂν ἐμῆϲ ἐπιμάρτυρε̣ϲ ̣ ε̣ἶ ̣ε̣ν̣ οἵτινεϲ οὐ χαλεποῦ νήϊδέϲ εἰϲι θεοῦ. Call. fr. 75.40 – 9 Harder (and Pfeiffer), 174 Massimilla⁵

It would seem plausible to suppose some connection between the climax and the remarkable profusion of the feature, which is used for every couplet in this passage. At the same time, the profusion highlights the elegist’s artistry: the feature underlines the elegant symmetry of the pentameter form, and in the last three couplets produces word-order which diverges with particular ostentation from what one might expect in speech. Notions of heightening, of stylistic elevation, are confirmed by the difference between Callimachus’ epigrams, where the feature occurs in 11.85 % of pentameters, and Hymn 5 (38.57 %) and the Aetia (29.32 %).⁶ They are confirmed also by the only two epigrams which have more than two examples of the feature: the unusually long and fanciful poem on Arsinoe’s nautilus (Ep. 5, 16 below), and especially the six-line poem which unfolds an Attic tragedy in Cyrene (Ep. 20, 19 below); the maiden even has the kingly name of Basilo. Some typical examples from the Aetia follow; it should be noted how the word at the caesura is most often an adjective (11 instances out of 13), and in many cases a decorative epithet which could be subtracted and still leave intact the basic point of the clause (see especially 2, 3, 4, 5, 8, and 9). 2 fr. 1.35 … τριγ⸥λ̣ώ̣ ⸤χι⸥ν ὀλ⸤οῶι⸥ νῆϲοϲ̣ ἐ̣π᾿ Ἐγκελά⸤δωι. 3 fr. 7.26 (9 M) αὖτιϲ ἐϲ ἀρχαίην⸥ ἔπλεον Αἱμονίην. 4 fr. 23.1 (25 M) … ἀϲτέρα, ναὶ κεραῶν ῥῆξιν ἄριϲτε βοῶν.’

 Dots etc. for Aetia come from my reading of the papyri.  See Appendix (). The numbers in the Appendix will give the right general impression; but in such matters different scholars usually give slightly different figures: subjectivity and error cannot be excluded.

Pentameters

5 fr. 43.65 (50 M) ϲτείνεα καὶ λ̣ευρὰϲ ὄφρα τά̣μ[ωϲιν ὁ]δ̣ούϲ. ̣ 6 fr. 63.10 (162 M) … παρθενι⸥καῖϲ Δηοῦϲ ὄργια Θεϲμοφόρου 7 fr. 64.14 (163 M) ὤ⸥λ̣ιϲ⸤θ⸥ε̣⸤ν μεγ⸥άλο⸤υϲ⸥ οἶκοϲ ἐπὶ ⸤Ϲ⸥κ⸤ο⸥πάδ⸤α⸥ϲ. 8 fr. 66.9 (165 M) … οἰκία καὶ λιπαραὶ ῥεῖτε Πελαϲγιάδεϲ. 9 fr. 67.10 (166 M) ἑδνῆϲτιν κεραῶν ἤιτεον ἀντὶ βοῶν. 10 fr. 75.77 (174 M) μ̣ῦ̣θο ̣ ̣ϲ ̣ ἐ̣ϲ ̣ ἡ̣μ̣ετέρην ἔδραμε Καλλιόπην. 11 fr. 110.62 (213 M) ̣ ̣ ̣ καὶ Βερ]ενίκειοϲ καλὸϲ ἐγὼ πλόκαμ[οϲ, 12 fr. 110.78 … λιτ̣ά, γυν̣α̣ι ̣κείων δ’ οὐκ ἀπέλαυ̣ϲα μύρων. 13 fr. 112.7 (215 M) χαῖρε, ϲὺν εὐεϲτοῖ δ᾿ ἔρχεο ̣ λωϊτέρηι. 14 fr. 178.4 (89 M) … ᾿Aτθίϲιν οἰκτίϲτη, ϲὸν φάοϲ, Ἠριγόνη,

Now we may look at all the examples from the epigrams. 15 Ep. 1.10 Pf. … ἔϲτρεφον εὐρείηι παῖδεϲ ἐνὶ τριόδωι. 16 5 κόγχοϲ ἐγώ, Ζεφυρῖτι, παλαίτερον, ἀλλὰ ϲὺ νῦν με, Κύπρι, Ϲεληναίηϲ ἄνθεμα πρῶτον ἔχειϲ, ναυτίλοϲ ὃϲ πελάγεϲϲιν ἐπέπλεον, εἰ μὲν ἀῆται, τείναϲ οἰκείων λαῖφοϲ ἀπὸ προτόνων, εἰ δὲ Γαληναίη, λιπαρὴ θεόϲ, οὖλοϲ ἐρέϲϲων ποϲϲίν (†ινωϲπ† ἔργωι τοὔνομα ϲυμφέρεται), ἔϲτ’ ἔπεϲον παρὰ θῖναϲ Ἰουλίδαϲ, ὄφρα γένωμαι ϲοὶ τὸ περίϲκεπτον παίγνιον, ᾿Aρϲινόη, μηδέ μοι ἐν θαλάμηιϲιν ἔθ᾿ ὡϲ πάροϲ (εἰμὶ γὰρ ἄπνουϲ) τίκτηται νοτερῆϲ ὤεον ἀλκυόνοϲ. Κλεινίου ἀλλὰ θυγατρὶ δίδου χάριν· οἶδε γὰρ ἐϲθλά ῥέζειν, καὶ Ϲμύρνηϲ ἐϲτὶν ἀπ᾿ Αἰολίδοϲ. (Cf. Mimn. fr. 9.6.) 17 14.1– 2 … καὶ ϲέ, | Χάρμι, τὸν ὀφθαλμοῖϲ χθιζὸν ἐν ἁμετέροιϲ, … 18 18 Νάξιοϲ οὐκ ἐπὶ γῆϲ ἔθανεν Λύκοϲ, ἀλλ’ ἐνὶ πόντωι ναῦν ἅμα καὶ ψυχὴν εἶδεν ἀπολλυμένην, ἔμποροϲ Αἰγίνηθεν ὅτ᾿ ἔπλεε. χὠ μὲν ἐν ὑγρῆι νεκρόϲ, ἐγὼ δ᾿ ἄλλωϲ οὔνομα τύμβοϲ ἔχων κηρύϲϲω πανάληθεϲ ἔποϲ τόδε· ‘φεῦγε θαλάϲϲηι ϲυμμίϲγειν Ἐρίφων, ναυτίλε, δυομένων.᾿ 19 20 ἠῶιοι Μελάνιππον ἐθάπτομεν, ἠελίου δέ δυομένου Βαϲιλὼ κάτθανε παρθενική αὐτοχερί· ζώειν γὰρ ἀδελφεὸν ἐν πυρὶ θεῖϲα οὐκ ἔτλη. δίδυμον δ᾿ οἶκοϲ ἐϲεῖδε κακόν πατρὸϲ ᾿Aριϲτίπποιο, κατήφηϲεν δὲ Κυρήνη πᾶϲα τὸν εὔτεκνον χῆρον ἰδοῦϲα δόμον. 20 24.2 ἵδρυμαι μικρῶι μικρὸϲ ἐπὶ προθύρωι, 21 30.3 – 4 ἦ ῥά ϲε δαίμων | οὑμὸϲ ἔχει, χαλεπῆι δ᾿ ἤντεο θευμορίηι; 22 31.1– 2 πάντα λαγωόν | διφᾶι καὶ πάϲηϲ ἴχνια δορκαλίδοϲ 23 56

121

122

Gregory O. Hutchinson

φηϲὶν ὅ με ϲτήϲαϲ Εὐαίνετοϲ (οὐ γὰρ ἔγωγε γιγνώϲκω) νίκηϲ ἀντί με τῆϲ ἰδίηϲ ἀγκεῖϲθαι χάλκειον ἀλέκτορα Τυνδαρίδηιϲι· πιϲτεύω Φαίδρου παιδὶ Φιλοξενιδέω. [24 not 59.2, where read λίαν (Maas)] 25 59.6 κἠγὼ τὼϲ πολλὼϲ οὐκέτ᾿ ἔχω Πυλάδαϲ.

The difference in frequency from the Aetia can be connected with the different style of the epigrams: the lack of extended sentences and structures (the nautilus apart), the lack of decoration. Connected with this last point is the difference in the instances themselves. In the epigrams only 9 of the 16 words at the caesura are adjectives; the only ornamental epithets seem to be 15 Ep. 1.10 ἔϲτρεφον εὐρείηι παῖδεϲ ἐνὶ τριόδωι, part of a more extended narrative, and 16 Ep. 5.10 τίκτηται νοτερῆϲ ὤεον ἀλκυόνοϲ (though the nautilus’ distaste is significant). Clearly not in this category is 25 59.6 κἠγὼ τὼϲ πολλὼϲ οὐκέτ᾿ ἔχω Πυλάδαϲ, where the multitude, accented by the article, is a vital part of the irony: it is a salient aspect of the real Pylades that he was unusual. Cf. e. g. Cic. Lael. 24, Man. 2.583 – 4 unus erat Pylades, unus qui mallet Orestes | ipse mori; lis una fuit per saecula mortis, Dio Prus. 74.28.⁷ This brings us to the pointed role of the feature when it does occur in the epigrams. More generally in the epigrams, the position before the caesura is a weighty one; words are often put there to throw them into relief. Cf. e.g. 1.14 παίδων || κλήδονα ϲυνθέμενοϲ, 6.4 Κρεωφύλωι, || Ζεῦ φίλε, τοῦτο μέγα, 7.4 (ἄλλων μέν …) κείνου δ᾿ || Ἑλλὰϲ ἀεὶ ϲοφίην, 9.2 … κοιμᾶται· θνήιϲκειν || μὴ λέγε τοὺϲ ἀγαθούϲ, 13.6 (εἰ δὲ τὸν ἡδύν |) βούλει, Πελλαίου || βοῦϲ μέγαϲ εἰν ᾿Aΐδηι, 16.4 τὸν πάϲαιϲ || ὕπνον ὀφείλομενον, 17.4 κενεὸν || ϲῆμα παρέρχομεθα, 22.4 (᾿Aϲτακίδην … ᾿Aϲτακίδηϲ. … οὐκέτι Δάφνιν |) ποιμένεϲ, ᾿Aϲτακίδην δ᾿ || αἰὲν ἀειϲόμεθα, 27.4 ᾿Aρήτου || ϲύμβολον ἀγρυπνίηϲ, 31.6 τὰ δ᾿ ἐν μέϲϲωι || κείμενα παρπέταται, 35.2 (εὖ μὲν ἀοιδήν |) εἰδότοϲ, εὖ δ᾿ οἴνωι || καίρια ϲυγγελάϲαι, 41.6 δύϲερωϲ ||, 43.6 φωρὸϲ δ᾿ || ἴχνια φὼρ ἔμαθον, 58.4 αἰθυίηι δ᾿ || ἶϲα θαλαϲϲοπορεῖ (cf. and contrast fr. 178.3 – 4), 62.4 (νῦν δὲ πέπαυται, |) αἶγεϲ, ἐπεὶ ϲπονδὰϲ || ἡ θεὸϲ εἰργάϲατο. Such weight is apparent with most instances of the feature in the epigrams, and not least the adjectives; the syntactical separation now adds to the emphasis. Cf. 16 5.4 οἰκείων (paradox of nautilus), 19 20.4 δίδυμον (the tragic development, cf. e.g. Eur. Med. 1315 διπλοῦν κακόν, Hec. 897 διϲϲὴ μέριμνα μητρί), 20.6 εὔτεκνον (set next to χῆρον; the change is strengthened by the article), 20 24.2 μικρῶι (set next to μικρόϲ), 21 30.4 χαλεπῆι (essential to the meaning), 22 31.2 πάϲηϲ (taking up πάντα in emphatic anaphora from after the bucolic diaeresis).

 Note also (apart from Eur. IT  – ,  – ) Eur. El.  – , Or.  – .

Pentameters

123

Thus when the feature appears, it is very far from being the accidental upshot of an indifferent placing of words. It is deployed forcefully, and with different tendencies here from its use in the Aetia (more frequent interest in emphasis here, much less frequent interest in ornament). We must see as significant both the use in the epigrams and the differentiation from the Aetia in overall occurrence. This double significance is confirmed when we consider two zealous followers of Callimachus, Catullus and Meleager. Catullus, in the two halves of his elegiac book (or, if one prefers, in his two types of elegiac poem) clearly responds to the first aspect, the difference between Callimachus’ short poems and his big one. The response has the more importance if it is an authorial book, which demonstrates Catullus’ ability to match Callimachus in both veins; hence he would start and close with Callimachus (65 – 6, 116). Catullus obviously notices the feature in Callimachus, as one can see from 66.58 Graia Canopeis incola litoribus cf. Call. fr. 110.58 (213 M) ̣ ̣ ̣ ̣ Κ]ανωπίτου ναιέτιϲ α[ἰγιαλοῦ. This version of Callimachus (poem 66) probably ends with an instance, ‘67.1– 2’ o dulci iucunda uiro, iucunda parenti, | salue, teque bona Iuppiter auctet ope; it begins with an instance transferred into the hexameter: 66.1 Omnia qui magni dispexit lumina mundi (cf. also 66.4 and 6). The frequency of the feature in the pentameter is markedly higher in the first half of the book, poems 65 – 8b: 37.42 %, compared with 20.75 % in 69 – 116.⁸ This difference is not accidental, as one can see by looking at Catullus’ new development, the thorough-going extension of the feature in 65 – 8b to the hexameter, at third-foot caesura and end: we see it in 20.50 % of hexameters in 65 – 8b, in 6.33 % of hexameters in 69 – 116.⁹ Callimachus shows more interest in Hymn 5 (5.56 %) than in the epigrams (0.75 %); but the proportion is still very small compared to the pentameters. Ennius and Lucretius show little interest: Ennius, Annales 2.07 %, Lucretius Book 1 2.07 %; Cicero shows more: Aratea 8.52 %. In 65 – 8b Catullus makes the hexameters with the feature not only three times more numerous in relation to the hexameters of 69 – 116, but of a much higher proportion in relation to the pentameters of 65 – 8b: in 65 – 8b the hexameters with the

 See Appendix (), and cf. Ross ,  – . For ‘. – ’ see Agnesini , and Du Quesnay ,  – ; cf. also Hunter in Fantuzzi and Hunter (), . A point to be added is that a husband, presented as a husband, can be seen as dulcis by a wife, but hardly by a door; this excludes Kroll’s idea that uiro and parenti are unspecific. It could possibly be maintained that the door is conceived of as a wife (cf. Cat. . postquam es … facta marita; but see Kroll ad loc., and add S. Tr.  –  δόμοϲ … ὁ μελλόνυφοϲ, with Stinton ,  – ). In that case, however, parenti would make no sense. At  in dominum ueterem deseruisse fidem I suggest dominos. I am grateful to Professor S. J. Heyworth for discussion of these matters.  Cf. Ross ,  – , and for Latin hexameter poetry Conrad ,  – .

124

Gregory O. Hutchinson

feature are 54.78 % of the pentameters with the feature, in 69 – 116 they are 30.50 %. The Hellenistic pentameter looks a more plausible inspiration than the Hellenistic hexameter for the feature to enter the Latin hexameter in abundance. Catullus 64, which could be later than the elegiac book of c.55 BC (48 BC or later?), shows 18.18 %, or 20.39 %, according to whether the refrain is counted once or every time; in 62, even if the refrain is counted every time, the proportion is only 7.58 %. At all events, this growth in the elegiac hexameters compared to Callimachus enables Catullus to invest his couplets with a hyper-Callimachean intensity as the first half of the elegiac book reaches its amorous climax. 28 Cat. 68b.119 – 40 nam nec tam carum confecto aetate parenti una caput seri nata nepotis alit, qui, cum diuitiis uix tandem inuentus auitis nomen testatas intulit in tabulas, impia derisi gentilis gaudia tollens suscitat a cano uulturium capiti: nec tantum niueo gauisa est ulla columbo compar, quae multo dicitur improbius oscula mordenti semper decerpere rostro quam quae praecipue multiuola est mulier: sed tu horum magnos uicisti sola furores, ut semel es flauo conciliata uiro. aut nihil aut paulo cui tum concedere digna lux mea se nostrum contulit in gremium, quam circumcursans hinc illinc saepe Cupido fulgebat crocina candidus in tunica. quae tamen etsi uno non est contenta Catullo, rara uerecundae furta feremus erae, ne nimium simus stultorum more molesti: saepe etiam Iuno, maxima caelicolum, coniugis in culpa flagrantem contudit iram, noscens omniuoli plurima furta Iouis.

There is a brief, though significant, turn in the last part of the passage: the feature continues for a little as the climax is undercut. Catullus, then, in the two halves of his elegiac book, captures the difference between Callimachus’ bodies of elegiac poetry; but he does not appear to regard Callimachus as ignoring the feature in his Epigrams. Thus there is an instance in his imitation of Callimachus, Epigram 25, the second of the short poems (70.4

Pentameters

125

rapida … aqua), and there are probably two in the final poem, which names Callimachus (116.4 … caput, 6 nostras … preces).¹⁰ The most notable short poems in this regard well establish that the feature carries point, even though Catullus here pursues ideas of his own. As in Callimachus’ extant epigrams, two of the short poems present the feature in more than two pentameters. The more conspicuous poem is 99, where 6 of the 8 pentameters show the feature (and one hexameter). surripui tibi dum ludis, mellite Iuuenti, suauiolum dulci dulcius ambrosia. uerum id non impune tuli: namque amplius horam suffixum in summa me memini esse cruce, dum tibi me purgo, nec possum fletibus ullis tantillum uestrae demere saeuitiae. nam simul id factum est, multis diluta labella guttis abstersti mollibus articulis, ne quicquam nostro contractum ex ore maneret, tamquam commictae spurca saliua lupae. praeterea infesto miserum me tradere amori non cessasti omnique excruciare modo, ut mi ex ambrosia mutatum iam foret illud suauiolum tristi tristius elleboro. quam quoniam poenam misero proponis amori, numquam iam posthac basia surripiam.

The poem starts with an extreme of amorous delight, and a Callimachean structure: suauiolum dulci dulcius ambrosia (cf. 20 Call. Ep. 24.2). But everything goes wrong, and after amorous torments, the delight is neatly inverted: … suauiolum tristi tristius elleboro. The abundance of the feature is clearly pointed; it seems to present an undoing of the lover’s joy, especially when set against the climax of the book’s first half in 68b (28). The other instance is harder to interpret. 30 Catullus 80 quid dicam, Gelli, quare rosea ista labella hiberna fiant candidiora niue, mane domo cum exis et cum te octaua quiete e molli longo suscitat hora die? nescioquid certe est: an uere fama susurrat grandia te medii tenta uorare uiri?

 With Catullus  cf. also Meleager LXIX.

126

Gregory O. Hutchinson

sic certe est: clamant Victoris rupta miselli ilia, et emulso barba notata sero.

The poem seems to move gradually down from the quasi-amorous beauty of rosy and white, and the more extreme separation of hiberna fiant candidiora niue, into the most drastic obscenity. The point could be an incongruous combination of the elegant and the disgusting, or simply emphasis; in any case, the presence of some point is suggested by the next Gellius poem, 88, which has two instances, 2 prurit et abiectis peruigilat tunicis and 8 non si demisso se ipse uoret capite, and presents an extreme of scelus. Thanks to the feature in the hexameter, there is one other particularly salient short poem: the first of them. We now move downward from Catullus’ union with the mistress in the preceding poem: Rufus is unable to unite with any woman. 29 Catullus 69 noli admirari quare tibi femina nulla, Rufe, uelit tenerum supposuisse femur, non si illam rarae labefactes munere uestis aut perluciduli deliciis lapidis. laedit te quaedam mala fabula, qua tibi fertur ualle sub alarum trux habitare caper. hunc metuunt omnes; neque mirum: nam mala ualde est bestia, nec quicum bella puella cubet. quare aut crudelem nasorum interfice pestem, aut admirari desine cur fugiant.

tenerum supposuisse femur has a graphic physicality absent from 68b; perluciduli deliciis lapidis gestures to epigram (cf. Posidippus’ section on stones). We also have a parody of aetiology, complete with fabula. Catullus seems to be exploiting the feature in a variety of downward departures. The finer detail of Catullus’ use in the short poems does not correspond to Callimachus’ in the Epigrams: so adjectives at the caesura, including ones that merely supplement, are much commoner in these epigrams than in Callimachus’. Catullus is going in his own direction. But his exploitation is based on a model of Callimachus’ use in the Epigrams both as less frequent than in the longer poems and as purposeful. The same is true of Meleager; and here we must take a look at the wider history of the feature, in the pentameter only. Detailed statistics are given in the Appendix (2 – 4). It will be seen that before Hellenistic poetry, the feature, in the scanty remains of individual poets, rises to 12.5 % in Mimnermus, cf. Theognis 1– 270 11.11 %, Xenophanes 11.76 %; the most extensively preserved poets or bodies of poetry give 6.79 % (Theognidea 1– 1006) and 7.34 % (Solon). In Hellenistic

Pentameters

127

poetry outside epigram, if we exclude poets or poems with under 20 pentameters, proportions go from 27.27 % to 59.57 %; it is actually notable how often we get 50 % when only 4 pentameters or so are preserved. In epigram, poets up to and including Callimachus give varied totals: some offer figures like 23.84 % (Posidippus) or 27.6 % (Anyte); some have much lower figures, like 14.58 % (Theocritus), 14.29 % (Asclepiades), 11.85 % (Callimachus). In epigram thereafter, roughly datable poets up to Meleager with 20 pentameters or more do not sink below 25 %; nor do Hellenistic poets of uncertain date with 20 pentameters or more (and note Nicarchus is a doubtful entity). In the datable group, figures above 30 % are common. Of poets with over 100 pentameters, we may note Leonidas (28.80 %), Dioscorides (30.47 %), Antipater of Sidon (31.93 %). By contrast, Meleager yields 12.14 %. After him, Philodemus, say, returns us to 27.96 %. We can now see that there is a context in earlier Hellenistic poetry for Callimachus’ choices in both epigrammatic and non-epigrammatic elegiacs. The higher figures for his non-epigrammatic poems suit the preferences of earlier Hellenistic elegiacs outside epigram; the lower figures for epigrams follow one of two strands in earlier Hellenistic epigram, the strand visible in Asclepiades, his particular model.¹¹ After Callimachus’ generation the other strand takes epigram over entirely, and epigrammatic and non-epigrammatic elegiacs are harder to separate; but then we reach Meleager, and a striking divergence from what had become the norm. Meleager is very much bound up with Callimachus intertextually; it looks highly likely that he reverts to a much lower use of the feature because he is imitating Callimachus’ practice. At the same time, it is very clear that he does not see Callimachus as ignoring the feature, but rather as using it tellingly. A plain example is one of his two poems with three instances (excluding the dubious CXXXII); this links itself with one of Callimachus’ two extant poems with three instances (16 5 on the nautilus, above). 36 LX οὐκέτι Τιμάριον, τὸ πρὶν γλαφυροῖο κέλητοϲ πῆγμα, φέρει πλωτὸν Κύπριδοϲ εἰρεϲίην· ἀλλ’ ἐπὶ μὲν νώτοιϲι μετάφρενον ὡϲ κέραϲ ἱϲτῶι κυρτοῦται, πολιὸϲ δ’ ἐκλέλυται πρότονοϲ, ἱϲτία δ᾿ αἰωρητὰ χαλᾶι ϲπαδονίϲματα μαϲτῶν, ἐκ δὲ ϲάλου ϲτρεπτὰϲ γαϲτρὸϲ ἔχει ῥυτίδαϲ, νέρθε δὲ πάνθ᾿ ὑπέραντλα νεώϲ, κοίληι δὲ θάλαϲϲα πλημύρει, γόναϲιν δ᾿ ἔντρομόϲ ἐϲτι ϲάλοϲ.

 On the feature in Asclepiades see Sens , lxxxix; on Callimachus’ epigrams and Asclepiades, ibid. lvi.

128

Gregory O. Hutchinson

δύϲτανοϲ, τίϲ ζωὸϲ ἔτ᾿ ὢν ᾿Aχερουϲίδα λίμνην πλεύϲετ᾿ ἄνωθ᾿ ἐπιβὰϲ γραὸϲ ἐπ᾿ εἰκοϲόρου;

An instance of the feature in this poem of Meleager’s deploys the very same word as an instance in that poem of Callimachus’: cf. Mel. LX.4 κυρτοῦται, πολιὸϲ δ’ ἐκλέλυται πρότονοϲ with Call. Ep. 5.4 τείναϲ οἰκείων λαῖφοϲ ἀπὸ προτόνων. One line is the opposite of the other: ἐκλέλυται is set against τείναϲ. Both poems use sailing metaphorically: the nautilus was a creature which made itself a ship, but now has risen to become Arsinoe’s plaything, in a sacred and fantastical elevation; the old hetaera remains a ship, but is now no longer fit for use, and the poem ends with death. Cypris comes in the second line of both poems; but in Callimachus the word marks final attainment and Arsinoe’s divine favour, in Meleager the lost past, itself improper. The three uses of the feature in Meleager’s poem mark a decline; the last in Callimachus’ achieves a climax. Meleager’s poem shows downward movement and distortion in the use of the feature; we have seen the same in short poems of Catullus. The only time the feature appears in a pentameter of Meleager’s prologue is precisely when Callimachus is introduced: Meleager I 21– 2 ἡδύ τε μύρτον | Καλλιμάχου ϲτυφελοῦ μεϲτὸν ἀεὶ μέλιτοϲ. The feature is used with sharp paradox, and a refined attempt to catch the essence of Callimachus with brevity. This same phrase is taken up in another instance of the feature: 35 LXXXVI ψυχαπάται δυϲέρωτεϲ, ὅϲοι φλόγα τὴν φιλόπαιδα οἴδατε τοῦ πικροῦ γευϲάμενοι μέλιτοϲ, ψυχρὸν ὕδωρ †νίψαι ψυχρὸν† τάχοϲ ἄρτι τακείϲηϲ ἐκ χιόνοϲ τῆι ᾿μῆι χεῖτε περὶ κραδίηι. ἦ γὰρ ἰδεῖν ἔτλην Διονύϲιον· ἀλλ᾿, ὁμόδουλοι, πρὶν ψαῦϲαι ϲπλάγχνων πῦρ ἀπ᾿ ἐμεῦ ϲβέϲατε.

The lines imitate 1 Call. fr. 75.49 οἵτινεϲ οὐ χαλεποῦ νήϊδέϲ εἰϲι θεοῦ, which itself employs the feature, and at a striking moment. The end of the poem also echoes Callimachus, Ep. 30.6 Pf. τὸν καλόν, ὦ μόχθηρ᾿, ἔβλεπεϲ ἀμφοτέροιϲ. The employment of the feature is clearly to be noted: the poem uses the feature more than once, like only six other poems of Meleager (excluding CXXXII).¹² Many other dimensions come into Meleager’s use of the feature (see further below); the Callimachean element is important, but not unique. There are many other aspects to Meleager’s use of the feature, some Callimachean, some not; but

 With ὁμόδουλοι in line  cf. Sen. Epist. . ‘serui sunt.’ immo conserui …

Pentameters

129

he himself obviously exemplifies the combination found in Callimachus’ epigrams: the feature is used sparingly, but with great effect. It will by now be apparent that Slings’s contention on relatively low usage must be mistaken: he appears to argue (33 – 4) that figures up to 15 % show a poet ignores the feature. We may consider this point more widely. Slings’s suppositions are not spelt out, but for the argument to work, the pre-Hellenistic poets would have to offer a random distribution or the like against which we could measure the Hellenistic poets. The pre-Hellenistic poets, that is, would be unaware of the feature. Such an approach would be legitimate (I very much hope) with pre-Hellenistic prose and Hegesianic rhythm; thus Antiphon the Orator, say, could be supposed to end a sentence with three long syllables and a cretic in complete innocence, quite unaware that this would become in Hegesias’ later system a desirable rhythm.¹³ But there could never be a similar unconsciousness in the separation of noun from adjective and the fitting of them to end each one half of the precisely balanced form. Nor, most of the time, are the absolute numbers, not the percentages, sufficient for statistical tests to be applied with any comfort. And it is by no means apparent that all these consecutive couplets are independent of each other, that choice in one does not influence choice in the next; but that is what statistical tests would require. In any case, let us suppose we could show it was improbable that one group of data displayed the same distribution as a group which we knew to display random distribution.¹⁴ It would not follow that if the figures in the first group had been similar to those in the second, they would both be examples of random distribution; for we could not exclude the situation which we have concluded to obtain with Callimachus: that the feature is deployed much less frequently in epigram than in other elegiac genres, but even in epigram is deployed deliberately and with point. The combination of restraint and forceful usage can be illustrated, and further made plausible, from modern poetry; this is worth doing because we have a more direct access to its metrical traditions. The following two poems from Derek Walcott’s book The Gulf and Other Poems, would not initially strike the reader as rhymed poems; but they make significant use of rhyme, imperfect rhyme, pararhyme, quasi-rhyme between identical words, rhyme between the end of a line and the inside of a line, rhyme between stressed and unstressed syllables.¹⁵

 Cf. Hutchinson .  Cf. Press/Teukolsky/Vetterling/Flannery ,  – .  Derek Walcott, The Gulf and Other Poems, Jonathan Cape, London ; © Derek Walcott, .

130

Gregory O. Hutchinson

‘Mass Man’ (Walcott 1969, 19) Through a great lion’s head clouded by mange a black clerk growls. Next, a gold-wired peacock withholds a man, a fan, flaunting its oval, jewelled eyes; What metaphors! What coruscating, mincing fantasies! Hector Mannix, waterworks clerk, San Juan, has entered a lion, Boysie, two golden mangoes bobbing for breastplates, barges like Cleopatra down her river, making style. ‘Join us’, they shout. ‘Oh God, child, you can’t dance?’ But somewhere in that whirlwind’s radiance a child, rigged like a bat, collapses, sobbing. But I am dancing, look, from an old gibbet my bull-whipped body swings, a metronome! Like a fruit bat dropped in the silk-cotton’s shade, my mania, my mania is a terrible calm. Upon your penitential morning, some skull must rub its memory with ashes, some mind must squat down howling in your dust, some hand must crawl and recollect your rubbish, someone must write your poems.

5

10

15

20

‘Landfall, Grenada’ (Walcott 1969, 49) for Robert Head, mariner Where you are rigidly anchored, the groundswell of blue foothills, the blown canes surging to cumuli cannot be heard; like the slow, seamless ocean, one motion folds the grass where you were lowered, and the tiered sea whose grandeurs you detested climbs out of sound. Its moods held no mythology for you, it was a working place of tonnage and ruled stars; you chose your landfall with a mariner’s casual certainty, calm as that race into whose heart you harboured; your death was a log’s entry, your suffering held the strenuous reticence of those

5

10

15

Pentameters

whose rites are never public, hating to impose, to offend. Deep friend, teach me to learn such ease, such landfall going, such mocking tolerance of those neat gravestone elegies that rhyme our end.

131

20

25

The first poem is particularly sparing in its use of obvious rhyme. The quasi-literary carnival, complete with metaphors and fantasies, dissolves unhappily, with no rhyme after a prominent imperfect rhyme (12 after 10 – 11; the descent is beginning by line 11). In the third stanza, the speaker at first produces a sad mimicry of the carnival in his own historical and poetic thoughts, with pararhymes (lines 13 – 16; note ‘metronome’ as a pararhymed word). But the last stanza moves to a yet barer aftermath of carnival, the actual production of poems; here, ironically, there are no rhymes, emphatically not in the last word ‘poems’. The second poem, filled with oblique variation of rhyme, deals in part with the denial of both literature and a literary reading of nature. The first part ends with no rhyming and the dead man’s detestation of grandeur. The second ends with a firm rhyme, though of a line-end some way back (‘offend’, 20); the dead man’s attitude now is not detestation but amused tolerance. The poem distances itself from ‘those | neat gravestone elegies’; but in some measure it shares, however ironically, their art and their turning of death into form. With words like ‘rhyme’ and ‘end’ there can be no doubt that the play around rhyme is pointed. Callimachus, Meleager, and Catullus are at least as intricate in their use of the feature. The second example from Walcott has affinities with the end of Meleager CXXII, which is the other of his poems with three instances. It is an epitaph for Antipater of Sidon, and particularly engages with Antipater of Sidon XXXII; that poem ends with three instances of the feature, this with two. Meleager is mocking and honouring Antipater and the style he espoused. 36 CXXII.1– 2, 17– 20 ἁ ϲτάλα, ϲύνθημα τί ϲοι γοργωπὸϲ ἀλέκτωρ ἔϲτα καλλαΐναι ϲκαπτροφόροϲ πτέρυγι …; ϲκᾶπτρα δ᾿ ἔχει ϲύνθημα λόγου, θνάιϲκειν δὲ πεϲόντα οἰνοβρεχῆ προπετὴϲ ἐννέπει ἀϲτράγαλοϲ. καὶ δὴ ϲύμβολα ταῦτα· τὸ δ᾿ οὔνομα πέτροϲ ἀείδει, ᾿Aντίπατρον προγόνων φύντ᾿ ἀπ᾿ ἐριϲθενέων.

132

Gregory O. Hutchinson

The last two instances in Antipater’s poem (XXXII) present the pathos of young death and the final word ἀϲτραγάλων (both poems solve riddling symbols on tombstones). The last two here begin with an inglorious drunken death, and some suggestion of Callimachus Epigram 61, but end with the famed lineage of Antipater.¹⁶ Callimachus’ usage, unlike Meleager’s, presents the complexity from which we started: that of operating within an opposition between epigrammatic and non-epigrammatic elegiacs. More broadly, then, his non-epigrammatic elegiacs play with the expansiveness and the catalogues of elegiac poetry, especially Hellenistic elegiac; the epigrams play with their own brevity. In both the play is thematized and complicated by Callimachus’ theoretical professions on length. We have seen that the handling of the feature in the two types of his elegiacs diverges, but also displays points of connection. In Catullus we have a similar opposition, which receives a further layer of complication from the relationship to Callimachus. We have incidentally seen the wider history of the feature, up to the end of the Republic; it continues to be important in Imperial Greek epigram, and in Latin elegy from Tibullus 1 and Propertius 1 onwards.¹⁷ But our greatest concern has been to establish the purposeful handling of the feature in Callimachus’ epigrams, against a generic background, and to see how that purposeful handling is confirmed by developments of it in Catullus and Meleager. Metre is obviously a fundamental part of epigram, and we are only beginning to interpret it.

Appendix *: Slings  gives his figures for these. Bodies of material with fewer than  pentameters or hexameters are enclosed in brackets, with fewer than  in double brackets. : the feature in the elegiac poems of Callimachus, Meleager and Catullus, pentameters and hexameters (third-foot caesura); some parallel material for hexameters in Catullus and other Republican poets. Callimachus, Epigrams*: Callimachus, Epigrams:

 out of  pentameters = . % (not incl. Ep.  Pf., but incl. frr.)  out of  hexameters = . %

 I am grateful to Professor R. Höschele for comments on this poem.  For Latin elegiacs from Tibullus on see Platnauer ,  (pentameter), and also Heyken ,  –  (hexameter and pentameter). It is disputed whether Tibullus I or Propertius I comes first; cf. Knox , Heslin .

Pentameters

(Callimachus, Hymn *: (Callimachus, Hymn : Callimachus, Aetia (*Slings Bk. ):

Callimachus, Aetia

(Callimachus, Victoria Sosibii (Callimachus, Victoria Sosibii Meleager (not incl. CXXXII, where  of  pent.) Meleager Catullus  – b: Catullus  – b: (Catullus : (Catullus : (Catullus : (Catullus : (Catullus : (Catullus : (Catullus a: (Catullus b: (Catullus b: Catullus  – : Catullus  – : Catullus  (counting refrain only once) Catullus  (with refrain) (Catullus  (with refrain) Ennius, Annales: Cicero, Aratea: Lucretius Book :

 out of  pentameters = . %)  out of  hexameters = . %)  out of  pentameters = . % (only those pentameters where enough survives to give a result; note, though, that with incomplete pentameters, remains just of line-end can indicate a negative result, but not a positive)  out of  hexameters = . % (figures still more uncertain; fr. . Θρηϊκίην μὲν || … ἄμυϲτιν included, on analogy with pentameters that show δ’ ||)  out of  pentameters = . %)  out of  hexameters = . %) (the instance is fr. . Ϲιδόνιόϲ με || … γαῦλοϲ, cf. on fr. . above)  out of  pentameters = . %

 out of  hexameters = . % (but  in I)  out of  pentameters = . %  out of  hexameters = . %  out of  pentameters = . %  out of  hexameters = . %)  out of  pentameters = . %, or with . –  as end of   out of  = . %)  out of  hexameters = . % [incl. st line]; or with . –  as end of   out of  = . %)  out of  pentameters = . %, or with . –  as end of   out of  = . %)  out of  hexameters = . %, or with . –  as end of   out of  = . %)  out of  pentameters = . %)  out of  pentameters = . %)  out of  hexameters = . %)  out of  pentameters = . %  out of  hexameters = . %  out of  hexameters = . %  out of  hexameters = . %  out of  hexameters = . %)  out of c. sufficiently complete hexameters = . %  out of  sufficiently complete hexameters = . %  out of  complete hexameters = . %

: the feature in archaic and classical elegiacs; pentameters (Archilochus

133

 out of + completish pentameters =  %)

134

Gregory O. Hutchinson

Theognidea  –  (only thus far to save most problems of roughly repeated lines; such lines are only included once) Theognis  –  (to give  pents., cf. Call. Ep.) (Callinus (Critias* (Ion* (Mimnermus (Simonides* Solon (Tyrtaeus (Xenophanes

 out of  pentameters = . %

 out of  pentameters = . %

       

out out out out out out out out

of of of of of of of of

 pentameters = . %)  pentameters = . %)  pentameters = . %)  pentameters = . %)  pentameters = . %)  pentameters (excluding fr.. West) = . %  pentameters = . %)  pentameters = . %)

: the feature in Hellenistic elegiacs outside epigram; pentameters (Philetas*, not epigr. (Hermesianax* (Phanocles* (Alexander Aetolus*, not epigr. (Archimedes* SH  ((Dorieus SH  ((Eudemus SH  A ((Hedyle SH  (Parthenius frr.  – ,  –  Lightfoot (Philo of Tarsus, SH  ((Philostephanus ((Ptolemy ((Timon SH  ((SH  ((SH  ((SH  –  (Tattoo poem [SSH ]* ((??P. Oxy. 

 out of  pentameters = . %))  out of  pentameters = . %)  out of  pentameters =  %)  out of  pentameters = . %)     

out out out out out

of of of of of

 pentameters = . %)  pentameters =  %))  pentameters =  %))  pentameters =  %))  pentameters = . %)

        

out out out out out out out out out

of of of of of of of of of

 pentameters = . %)  pentameters =  %))  pentameters =  %))  pentameters =  %))  pentameters =  %))  pentameters =  %))  pentameters =  %))  pentameters =  %)  pentameters =  %))

: the feature in Hellenistic epigram; pentameters ((Antagoras (Anyte* ((Duris (Nossis* (Perses

 out of  pentameters = . %))  out of  pentameters = . %)  out of  pentameters = . %))  out of  pentameters = . %)  out of  pentameters = . %)

Pentameters

((Phalaecus (Simias ((Alexander Aetolus, epigrams (Asclepiades* (Hedylus, epigrams ((Hedylus, SH  (Nicias Posidippus (not incl. AB ) (Theaetetus (Theocritus (Hegesippus Leonidas (elegiac poems in I-XCI; *Slings  pent.) (Mnasalces* (Damagetus (including XII) Dioscorides (Nicaenetus ((Phaedimus (Rhianus, epigrams ((Theocritus of Chios SH  (Theodoridas (Alcaeus ((Herodicus ((Samius Antipater of Sidon (following Page, Ep. Gr.) (Dionysius ((Moschus (Phanias ((Polystratus ((Zenodotus (Philodemus ((Agis ((Amyntas ((Archelaus ((Aristodicus (Ariston ((Artemon ((Athenaeus SH  – 

135

 out of  pentameters = . %))  out of  [or ] pentameters = . % [or . %])  out of  pentameters = . %))  out of  pentameters = . %)  out of  pentameters = . %)  out of  pentameters =  %))  out of  pentameters = . %)  out of  pentameters = . %  out of  pentameters = . %)  out of  pentameters = . %)  out of  pentameters = . %)  out of  pentameters = . %

 out of  pentameters = . %)  out of  pentameters = . % [without XII:  out of  pentameters = . %])  out of  pentameters = . %  out of  pentameters = . %)  out of  pentameters = . %))  out of  pentameters = . %)  out of  pentameters =  %))  out of  pentameters = . %)  out of  pentameters¹⁸ = . %)  out of  pentameters = . %))  out of  pentameters = . %))  out of  pentameters = . %  out of  pentameters = . %)  out of  pentameters = . %))  out of  pentameters = . %)  out of  pentameters = . %))  out of  pentameters = . %))  out of  pentameters = . %)  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters = . %)  out of  pentameters = . %))  out of  pentameters = . %))

 Note in Alcaeus:  out of  hexameters =. %;  in X.

136

Gregory O. Hutchinson

((Carphyllides ((Chaeremon ((‘Damostratus’ (Diotimus ((Hegemon ((Hermocreon ((Hermodorus (Lobo, fragmenta certa Garulli¹⁹ ((Menecrates ((Nicander (Nicarchus ((Nicomachus ((Pamphilus ((Pancrates ((Phaennus ((Philetas of Samos ((Thymocles (Tymnes ((SH  ((SH  ((SH  ((SH  ((SH  ((SH  ((SH  ((SH  ((SH  ((SH 

 out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters = .)  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters = . %) (and  in the  pentameters of the fragmenta dubia)  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters = . %)  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters = . %))  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters =  %)  out of  pentameters =  %))  out of  pentameter =  %))  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters = . %))  out of  pentameters =  %))  out of  pentameters =  %))  out of  pentameters = . %))  out of  pentameters =  %))  out of  pentameters =  %))

Abbreviations AB Austin, C./Bastianini, G. (eds.) (2002), Posidippi Pellaei quae supersunt omnia, Milan. SH Lloyd-Jones, H. / Parsons, P. (eds.) (1983), Supplementum Hellenisticum, Berlin. SSH Lloyd-Jones, H. (ed.) (2005), Supplementum Supplementi Hellenistici, Berlin.

 See Garulli .

Pentameters

137

Bibliography Agnesini, A. (2011), ‘Catull. 67,1 s.: incipit della Ianua o explicit della Coma?’, in: Paideia 66, 521 – 40. Bentein, K. (2013), ‘Adjectival periphrasis in Ancient Greek: the categorial status of the participle’, in: Acta Classica 56, 1 – 28. Conrad, C. (1965), ‘Traditional patterns of word-order in Latin epic from Ennius to Vergil’, in: HSCPh 69, 195 – 258. Du Quesnay, I. (2012), ‘Three problems in Poem 66’, in: I. Du Quesnay / A. Woodman (eds.), Catullus: Poems, Books, Readers, Cambridge, 153 – 83. Fantuzzi, M. / Hunter, R. (2004), Tradition and Innovation in Hellenistic Poetry, Cambridge. Garulli, V. (2004), Il Περὶ ποιητῶν di Lobone di Argo, Eikasmos Studi 10, Bologna. Heslin, P. (2010), ‘Virgil’s Georgics and the dating of Propertius’ first book’, in: JRS 100, 54 – 68. Heyken, J. (1916), Über die Stellung der Epitheta bei den römischen Elegikern, Diss. Kiel. Hutchinson, G.O. (1992), review of Huys (1991), in: CR n.s. 42, 483 – 4. — (2013a), ‘Genre and super-genre’, in: T.D. Papanghelis / S.J. Harrison / S. Frangoulidis (eds.), Generic Interfaces in Latin Literature: Encounters, Interactions and Transformations, Trends in Classics, Supplementary vol. 20, Berlin, 19 – 34. — (2013b), Greek to Latin: Frameworks and Contexts for Intertextuality, Oxford. — (2015), ‘Appian the artist: rhythmic prose and its literary implications’, in: CQ n.s. 65, 788 – 806. Huys, M. (1991), Le Poème élégiaque hellénistique P. Brux. inv. E. 8934 et P. Sorb. inv. 2254. Édition, commentaire et analyse stylistique, Papyri Bruxellenses Graecae II 22, Brussels. Knox, P. E. (2005), ‘Milestones in the career of Tibullus’, in: CQ n.s. 55, 204 – 16. Platnauer, M. (1951), Latin Elegiac Verse: A Study of the Metrical Usages of Tibullus, Propertius & Ovid, Cambridge. Press, W.H. / Teukolsky, S.A. / Vetterling, W.T. / Flannery, B.P. (1992), Numerical Recipes in C: The Art of Scientific Computing2, Cambridge. Ross, D.O., Jr. (1969), Style and Tradition in Catullus, Cambridge, Mass. Sens, A. (2011), Asclepiades of Samos, Epigrams and Fragments: Edited with Translation and Commentary, Oxford. Slings, S.R. (1993), ‘Hermesianax and the Tattoo Elegy (P.Brux. inv. E 8934 and P.Sorb. inv. 2254)’, in: ZPE 98, 29 – 37. Stinton, T.C.W. (1990), ‘Heracles’ homecoming and related topics: the second stasimon of Sophocles’ Trachiniae’, in: Collected Papers on Greek Tragedy, Oxford, 402 – 29. Walcott, D. (1969), The Gulf and Other Poems, London.

Demetra Koukouzika

Epigrams in Epic? The Case of Apollonius Rhodius

The last decades have seen a great deal of work on how literary genres interact with each other. The epigram has not remained immune from this scholarly movement–important work has indeed looked at the interactions of, for example, epigram and elegy and epigram and pastoral. The well-known generic inclusivity of epic has, naturally enough, been brought into play, though to speak of the ‘generic inclusivity’ of Homer might be to create a potentially very misleading form of literary history (or it might not). Does Homeric epic offer a home to assimilated forms of non-epic genres, or is it in fact the origin of those other genres? Epitaphic poetry is perhaps the most famous and the most studied case. Epic is full of the dead, and many of the dead are as unfamiliar to us, and probably were to Homer’s audience, as names in a row of gravestones. Most famously of all, Hector’s foreshadowing of post-mortem fame for one of the Greeks in Iliad 7 has drawn more than its fair share of attention: καί ποτέ τις εἴπῃσι καὶ ὀψιγόνων ἀνθρώπων νηῒ πολυκλήϊδι πλέων ἐπὶ οἴνοπα πόντον· ἀνδρὸς μὲν τόδε σῆμα πάλαι κατατεθνηῶτος, ὅν ποτ᾽ ἀριστεύοντα κατέκτανε φαίδιμος Ἕκτωρ. ὥς ποτέ τις ἐρέει· τὸ δ᾽ ἐμὸν κλέος οὔ ποτ᾽ ὀλεῖται. And people will say, even men of generations not yet born, as they sail over the sparkling sea in their many-benched ships: “This is the mound of a man who died long ago. Glorious Hector killed him as he fought gallantly”. That is what they will say: and my glory will never die. Il. 7.87– 91, transl. M. Hammond, adapted

Scholars such as Richard Thomas, Ruth Scodel¹ and Don Lavigne in this volume have fruitfully explored how, on the one hand, consciousness of epigrammatic modes adds meaning to Hector’s words (the dead remains nameless, whereas Hector is named) and, on the other, what these verses may be able to tell us about the history of the funerary epigram itself. More straightforward in some ways seem to be the brief notices for slain warriors with which the battle-scenes of the Iliad are littered, like corpses strewn around a battle-field. Name, parents’

 Thomas , Scodel .

140

Demetra Koukouzika

names and homeland give the necessary detail, just as they were to do in countless later funerary epigrams. Iphition can stand for very many: ἐν δ᾽ ᾿Aχιλεὺς Τρώεσσι θόρε φρεσὶν εἱμένος ἀλκὴν, σμερδαλέα ἰάχων, πρῶτον δ᾽ ἕλεν Ἰφιτίωνα ἐσθλὸν Ὀτρυντεΐδην, πολέων ἡγήτορα λαῶν, ὃν νύμφη τέκε νηῒς Ὀτρυντῆϊ πτολιπόρθῳ Τμώλῳ ὕπο νιφόεντι, Ὕδης ἐν πίονι δήμῳ. But Achilles leapt on the Trojans, his heart clothed in courage, and shouting fearfully. The first he killed was Iphition, brave son of Otrynteus and leader of many people, borne by a water-nymph to Otrynteus, sacker of cities, below snowy Tmolos, in the rich land of Hyde. Il. 20.381– 5, transl. M. Hammond

Later epic picked up the Homeric hint². There are some famously epigrammatic utterances in the Aeneid (the opening verses of Book 7 on Caieta, for example) and Martin Dinter³ has helpfully pursued some of them in an important article. The Argonautica has to date not become a central part of this discussion, although from time to time commentaries cite epigrammatic parallels for particular turns of phrase in Apollonius’ epic. This is somewhat surprising, and not just because so many of the techniques which Virgil elaborated from their Homeric beginnings have been found to be mediated through the Hellenistic epic. Whatever date we want to give to the Argonautica in the form in which we now have it, it will have been written at a time of the flowering of Hellenistic literary epigram. We know – in as much as we can ever claim to know such things – that Apollonius knew Callimachus’ Aitia very well, as well as some at least of the hymns and (almost certainly) the Iambi and the Hecale; why not the epigrams also? Why not, moreover, Theocritus’ epigrams, as well as the Idylls? Put like this there is no argument to be had: we will of course assume that Apollonius was very much at home with all Alexandrian literary genres. Finding epigrammatic form within the Argonautica will not cause anyone any surprise. Moreover, the interplay of the large and the small, the macro- and the microscopic, in terms of commemoration of the dead, of κλέα ἀνδρῶν, suggests that ‘epigram in epic’ would have appealed to Hellenistic sensibility, in something of the same way that Posidippus’ epigrams on precious stones have shown us how the entire

 A number of Iliadic passages have been identified as ‘epigrams’ or ‘epigrammatic’ in the Homeric scholia: () AT . – b; () T .; AbT .d (refers to . as τὰ ἐπιγράμματα); () AbT . –  (τὰ ἐπιγράμματα); () bT .b and in Ps-Plut. De Homero .. Cf. Vox  for a concise commentary and Elmer  for a closer analysis of . –  and . –  (the ‘epigrammatic’ lines spoken by Helen in the τειχοσκοπία).  Dinter .

Epigrams in Epic? The Case of Apollonius Rhodius

141

world can be captured and glimpsed within a tiny space. A full treatment of this topic would also have to look at the whole topic of epic miniaturisation, recently so fruitfully explored by Michael Squire and others⁴, with regard to the tabulae Iliacae, but space does not permit that in this paper. One obvious place to look for ‘epigram in epic’ is a catalogue of heroes, where each hero is given a few lines, usually specifying name, place of origin, family and something special about them. Reading a catalogue of heroes is another activity which is almost inevitably like walking through a graveyard⁵ or reading a collection of funerary epigrams. The work of Kathryn Gutzwiller, Peter Bing and others has taught us so much in recent years about the importance of the coming of epigram collections in the third century that it seems inevitable that we should want to connect the forms of epic catalogue and epigram collection. I will not be doing that at any length in this short paper, but let me use just one example to introduce our quest. Tiphys, the steersman of the Argo, is introduced as follows: Τῖφυς δ᾽ Ἁγνιάδης Σιφαέα κάλλιπε δῆμον Θεσπιέων, ἐσθλὸς μὲν ὀρινόμενον προδαῆναι κῦμ᾽ ἁλὸς εὐρείης, ἐσθλὸς δ᾽ ἀνέμοιο θυέλλας καὶ πλόον ἠελίῳ τε καὶ ἀστέρι τεκμήρασθαι. Tiphys, son of Hagnias, left the region of Siphai, peopled by the Thespians. He was skilled at foretelling when the waves of the broad sea would rise up, skilled too at foretelling windstorms, and at guiding a voyage by the sun and the stars. Arg . 1.105 – 8, transl. R. Hunter⁶

A life summed up in a few verses, made memorable by verbal anaphora and chiasmus. Are we justified in calling this ‘epigrammatic’? Collections of inscribed epigrams, such as Merkelbach-Stauber Steinepigramme aus dem griechischen Osten, would provide very many parallels for epitaphic poems celebrating the skills of the dead (often schoolteachers, grammarians etc), and clearly the verbal elaboration here is meant to catch our eye and stick in our mind. Is the ‘epigram-

 Cf. Squire , Petrain .  Or reading the inscription on a polyandrion. For catalogues in early Greek epigram cf. Petrovic, A. in this volume. An ‘epitaph’ commemorating a different kind of androktasia could be Arg. . – : Ἔνθ᾽ ἄμυδις πᾶς δῆμος ὑπερβασίῃσι γυναικῶν νηλειῶς δέδμητο παροιχομένῳ λυκάβαντι. I have discussed this at greater length in my 2008 dissertation: Apollonius Rhodius, Argonautica 1.605 – 1362: A Commentary.  All translations of the Argonautica come from Hunter .

142

Demetra Koukouzika

matic’, then, merely in the eye of the beholder? Is it relevant here that Tiphys’ disappearance from the poem might also be thought to be ‘epigrammatic’ without being an epigram: Ἁγνιάδην Τῖφυν θανέειν φάτις· οὐδέ οἱ ἦεν μοῖρ᾽ ἔτι ναυτίλλεσθαι ἑκαστέρω· ἀλλά νυ καὶ τὸν αὖθι μινυνθαδίη πάτρης ἑκὰς εὔνασε νοῦσος, εἰσότ᾽ ᾿Aβαντιάδαο νέκυν κτερέιξεν ὅμιλος. The story is that Tiphys, son of Hagnias, died; he was not fated to sail further. He too was put to sleep by a brief illness far from his home, after the crew had buried the corpse of the son of Abas. Arg . 2.854– 7

‘A brief illness put him to sleep far from his home’: multiple epigrammatic parallels come to mind, by poets great and small, though the fullest modern commentary on Book 2 (Matteo) mentions only a contrast with Odysseus. Epigrammatic form, which puts a small seal on great lives, carries its own pathetic force, and pathos is clearly one of the effects at which Apollonius is aiming here: all of Tiphys’ skill was as nothing against ‘a brief illness’. Perhaps Virgil saw the point. Aeneas’ famous ‘epitaph’ for his steersman Palinurus at the end of Aeneid Book 5, o nimium caelo et pelago confise sereno, nudus in ignota, Palinure, iacebis harena Ah, too trustful in the calm of sky and sea, naked you will lie, Palinurus, on an unknown strand! Aeneid 5.870 – 1, transl. H.R. Fairclough

may, as well as rhythmically evoking epitaphic epigram with the break before the vocative, owe something to the epigrammatic framing, as I have argued, of the Apollonian Tiphys (though Damien Nelis does not mention the possibility in his book on the Argonautica and the Aeneid). Is then, to repeat, the ‘epigrammatic’ in the eye of the beholder? Am I alone in imagining that the opening verses of Book 2 might well have stood as an epigram memorialising the mythological past for those on the Argonautic tourist route: Ἔνθα δ᾽ ἔσαν σταθμοί τε βοῶν αὖλίς τ᾽ ᾿Aμύκοιο, Βεβρύκων βασιλῆος ἀγήνορος, ὅν ποτε Νύμφη τίκτε Ποσειδάωνι Γενεθλίῳ εὐνηθεῖσα Βιθυνὶς Μελίη ὑπεροπληέστατον ἀνδρῶν·

Epigrams in Epic? The Case of Apollonius Rhodius

143

ὅς τ᾽ ἐπὶ καὶ ξείνοισιν ἀεικέα θεσμὸν ἔθηκε, μή τιν᾽ ἀποστείχειν πρὶν πειρήσασθαι ἑοῖο πυγμαχίης, πολέας δὲ περικτιόνων ἐδάιξε. There were the cattle-stalls and pens of Amykos, proud king of the Bebrykians, whom a nymph, Bithynian Melie, had borne to Poseidon Genethlios, with whom she had lain. He was the most outrageous of men, and even upon strangers he had imposed a shameless ordinance: no one might depart before trying his luck against the king in boxing, and many men from neighbouring territories had thus met their end. Arg . 2.1– 7

If that seems overly speculative, what of, for example, the famous verses with which Apollonius introduces the narrative of the killing of Medea’s brother, Apsyrtus: Σχέτλι᾽ Ἔρως, μέγα πῆμα, μέγα στύγος ἀνθρώποισιν, ἐκ σέθεν οὐλόμεναί τ᾽ ἔριδες στοναχαί τε πόνοι τε, ἄλγεά τ᾽ ἄλλ᾽ ἐπὶ τοῖσιν ἀπείρονα τετρήχασιν· δυσμενέων ἐπὶ παισὶ κορύσσεο, δαῖμον, ἀερθείς, οἷος Μηδείῃ στυγερὴν φρεσὶν ἔμβαλες ἄτην. Reckless Eros, great curse, greatly loathed by men, from you come deadly strifes and grieving and troubles, and countless other pains on top of these swirl up. Rear up, divine spirit, against my enemies’ children as you were when you threw hateful folly into Medea’s heart. Arg . 4.445 – 9

The narrator intrudes into his narrative with what we might be tempted to call an ‘epigraph’ for the coming narration⁷; at some time someone should write a history of the use of epigrams as epigraphs. The apopemptic function of these verses is familiar, but might it also be fair to describe them as an epigram? No one will need convincing that eros, let alone the dangers of eros, is a suitable subject for epigrams, and the rejection of pain, suffering and strife also conjures a sympotic context; we naturally think of the close of Xenophanes’ famous poem on the ideal symposium: ἀνδρῶν δ᾽ αἰνεῖν τοῦτον ὃς ἐσθλὰ πιὼν ἀναφαίνει, ὡς ᾖ μνημοσύνη καὶ τόνος ἀμφ᾽ ἀρετῆς, οὔ τι μάχας διέπειν Τιτήνων οὐδὲ Γιγάντων οὐδὲ ‹ › Κενταύρων, πλάσμα‹τα› τῶν προτέρων, ἢ στάσιας σφεδανάς· τοῖς οὐδὲν χρηστὸν ἔνεστιν· θεῶν ‹δὲ› προμηθείην αἰὲν ἔχειν ἀγαθήν.

 Cf. Hunter , ad loc.

144

Demetra Koukouzika

But the man whom one must praise is he who, after drinking, reveals noble thoughts, so that there is a recollection of and striving for excellence; it is not right to make an array of the wars of the Titans or Giants or Centaurs, creations of our predecessors, or violent factions – there is nothing useful in them; and it is right always to have a good regard for the gods. Xenophanes 1.19 – 24, transl. D. Gerber, adapted

What eros should bring is pleasure and delight, not real pain or even, as in the Argonautica, narratives of pain. Commentators have pointed out that the nearest parallel for Apollonius’ imprecation against Eros is in fact verses from the Theognidea, indeed the verses which were placed at the head of the so-called Book 2, the collection of paederastic verse: Σχέτλι’ Ἔρως, μανίαι σε τιθηνήσαντο λαβοῦσαι· ἐκ σέθεν ὤλετο μὲν Ἰλίου ἀκρόπολις, ὤλετο δ’ Αἰγείδης Θησεὺς μέγας, ὤλετο δ’ Αἴας ἐσθλὸς Ὀιλιάδης σῇσιν ἀτασθαλίαις. Cruel Love, the spirits of Madness took you up and nursed you. Because of you Troy’s acropolis was destroyed, and great Theseus, Aegeus’ son, and noble Ajax, son of Oileus, through your acts of recklessness. Theognis 1231– 4, transl. D. Gerber, adapted

It is hard to imagine a context for the Theognidean verses other than a symposium, and the appeal to famous ‘epic’ disasters caused by love brings these verses quite close to Apollonius’ imprecation. At a moment of the highest tension, then, a sense of ‘the epigrammatic’ is used by the epic poet, not just to mark his empathetic engagement with his narrative, but also to consign the horrors he is about to narrate to collective memory, as we have seen in the verses from the Theognidea. This is a collective memory which most naturally surfaces in those short poems performed (or imagined as performed) by guests taking their ease, epigrams in other words. The general point should, I hope, by now be clear. We are obviously not going to be able to establish rules for finding ‘the epigrammatic’, but there are clear signs that what we are looking for actually exists and some suggestive criteria can at least be outlined. When Medea leaves behind a lock of her hair as she prepares to flee with the Argonauts, several different poetic forms are evoked: Τόνδε τοι ἀντ᾽ ἐμέθεν ταναὸν πλόκον εἶμι λιποῦσα, μῆτερ ἐμή· χαίροις δὲ καὶ ἄνδιχα πολλὸν ἰούσῃ· χαίροις, Χαλκιόπη καὶ πᾶς δόμος. Αἴθε σε πόντος, ξεῖνε, διέρραισεν πρὶν Κολχίδα γαῖαν ἱκέσθαι.

Epigrams in Epic? The Case of Apollonius Rhodius

145

As I go I leave you this flowing lock, mother, to take my place. Farewell- this is my wish as I depart on a very distant journey; Farewell, Chalkiope and all my home! Stranger, would that the sea had torn you in pieces before you reached the Colchian land! Arg. 4.30 – 3

Is one of those forms the epigrammatic? Dedicatory epigrams are, of course, one of the earliest forms of the genre (and epigrams recording hair-dedications are familiar in our Hellenistic corpus), but also one which we know, from poets such as Callimachus and Posidippus, was developed in ‘literary’ directions in the third century. On the verge of a kind of marriage which is also a departure to a distant and unknown home, Medea dedicates a lock and wishes her family well. So far, so epigrammatic. We know that the dedication of a lock of hair was indeed a familiar pre-nuptial rite for virgins⁸. It is, however, the last verse and a bit which shatters the dedicatory and/or nuptial atmosphere – as in so many epigrams, the sting really is in the tail. This departure is also a departure to a kind of death – ‘partir c’est mourir un peu’ – and the other common setting for the cutting of locks of hair is indeed funerary rites. It is indeed funerary epigram to which epic seems most receptive, and for reasons which do not need spelling out further. In Apollonius, death is often of course also an aetiological moment, and funerary epigram and aetiology are two modes of comemmoration which may exist, as it were, alongside each other. Both project the dead into the future. Consider the sad case of the death of Kleite: Οὐδὲ μὲν οὐδ᾽ ἄλοχος Κλείτη φθιμένοιο λέλειπτο οὗ πόσιος μετόπισθε· κακῷ δ᾽ ἐπὶ κύντερον ἄλλο ἤνυσεν, ἁψαμένη βρόχον αὐχένι. Τὴν δὲ καὶ αὐταὶ Νύμφαι ἀποφθιμένην ἀλσηίδες ὠδύραντο· καί οἱ ἀπὸ βλεφάρων ὅσα δάκρυα χεῦαν ἔραζε, πάντα τά γε κρήνην τεῦξαν θεαί, ἣν καλέουσι Κλείτην, δυστήνοιο περικλεὲς οὔνομα νύμφης. Nor did Kleite survive after her husband’s death, but she added even worse misery to what had gone before by fitting a noose around her neck. The very nymphs of the groves mourned her death, and from all the tears which dropped to the earth from their eyes the goddesses fashioned the spring which men call Kleite, the ever renowned name of the unhappy bride. Arg. 1.1063 – 9

 Cf. Hunter , ad vv.  – .

146

Demetra Koukouzika

The suicide of a young wife (almost certainly) in grief at the death of her husband, and before she had given birth to any children: the situation is made for a metamorphosis, even an Ovidian one, but the pattern of name, narrative, lamentation might also be thought to evoke funeral epigram. The etymological play of 1069 sets an epigrammatic seal upon Kleite’s fame and certainly one, or perhaps two, aetiologies are in play. We learn the origin of the fountain named after the sad bride, but the equivocation of νύμφη between ‘nymphs’ (1066) and ‘bride’ (1069) perhaps suggests that we are to understand that we have here also an aetiology for another pre-nuptial rite: Cyzicene brides are to shed tears for Kleite before their marriage. This must remain only a guess, but I hope that the epigrammatic flavor of the passage is more than that⁹. Compare, for example, Callimachus’ famous poem on the death of Basilo, in grief at her brother’s death: Ἠῷοι Μελάνιππον ἐθάπτομεν, ἠελίου δέ δυομένου Βασιλὼ κάτθανε παρθενική αὐτοχερί· ζώειν γὰρ ἀδελφεὸν ἐν πυρὶ θεῖσα οὐκ ἔτλη. δίδυμον δ᾽ οἶκος ἐσεῖδε κακόν πατρὸς ᾿Aριστίπποιο, κατήφησεν δὲ Κυρήνη πᾶσα τὸν εὔτεκνον χῆρον ἰδοῦσα δόμον. It was morning when we buried Melanippus, and at sunset the maiden Basilo died by her own hand; for after laying her brother on the pyre she could not abide to live. The house of their father Aristippus witnessed a double woe, and all Cyrene stood with downcast eyes, seeing the home bereft of its lovely children. Epigr. 20 Pf., transl. A.W.Mair, adapted

This typically Callimachean revision of the ‘straight’ funeral epigram is obviously very different from Apollonius’ passage on Kleite, despite the shared motifs, and yet the two also share the sense of the confinement of an unspeakable grief within the small space of an epigram. Callimachean understatement moves from the dead brother to Basilo’s suicide to the bereavement of the oikos and finally to that of the whole city; a similar movement is observable through the whole Apollonian narrative, which ends with an aetiology of Cyzicene ritual practice; the community will never forget her. As with Tiphys (see above), it may in fact be that Kleite is framed epigrammatically. Consider the verses which introduce her to us:

 The repetition of νύμφαι and νύμφης forms a type of ring composition which contributes to the epigrammatic tone of the passage (I owe this point to Benjamin Acosta-Hughes).

Epigrams in Epic? The Case of Apollonius Rhodius

147

Ἁρμοῑ που κἀκείνῳ ὑποσταχύεσκον ἴουλοι, οὐδέ νύ πω παίδεσσιν ἀγαλλόμενος μεμόρητο, ἀλλ᾽ ἔτι οἱ κατὰ δώματ᾽ ἀκήρατος ἦεν ἄκοιτις ὠδίνων, Μέροπος Περκωσίου ἐκγεγαυῖα Κλείτη ἐυπλόκαμος. Τὴν μὲν νέον ἐξέτι πατρὸς θεσπεσίοις ἕδνοισιν ἀνήγαγεν ἀντιπέρηθεν. He too was just showing the first beard of manhood, and had not yet been blessed with children. His wife, the lovely-tressed Kleite, daughter of Merops of Perkote, remained in the palace untouched by the pains of child-bearing. He had recently brought her from her father’s house on the coast facing the city, after paying a wondrous bride-price for her. Arg. 1.972– 7

Her name, her father’s name and place of origin, the fact that she was κατὰ δώματ᾽ ἀκήρατος … ἄκοιτις ὠδίνων all suggest epigrammatic motifs; the introduction of the young bride, not yet mother, almost foreshadows the fact that we are to witness her death. The ‘marvellous wedding-gifts’ with which Cyzicus won her are to prove a sad waste; such grim ironies abound in Greek funerary epigram. The purpose of this paper has been just to put out some questions about Apollonius and epigrammatic form. I could go on piling up examples, but let me just make a small soros of further potential Argonautic epigrams. The death of the seers Idmon (nomen et omen) and Mopsus both convey horribly familiar epigrammatic lessons: Ἔνθα δ᾽ ᾿Aβαντιάδην πεπρωμένη ἤλασε μοῖρα Ἴδμονα, μαντοσύνῃσι κεκασμένον· ἀλλά μιν οὔ τι μαντοσύναι ἐσάωσαν, ἐπεὶ χρεὼ ἦγε δαμῆναι. Then it was that fated destiny overtook Idmon, son of Abas and excellent prophet; his skill in prophecy did not save him, since necessity led him on to death. Arg. 2.815 – 7 Ἔνθα καὶ ᾿Aμπυκίδην αὐτῷ ἐνὶ ἤματι Μόψον νηλειὴς ἕλε πότμος· ἀδευκέα δ᾽ οὐ φύγεν αἶσαν μαντοσύναις· οὐ γάρ τις ἀποτροπίη θανάτοιο. There on that same day pitiless fate overtook Mopsos, son of Ampykos. His prophetic skill did not help him escape the bitterness of his allotted end, for death cannot be bought off. Arg. 4.1502– 4

The two deaths are paired in various ways in the text¹⁰, but I wonder if they are also paired in evoking the epigrammatic. No skill can save you when Death  Cf. Hunter , ad vv.  – .

148

Demetra Koukouzika

comes calling, ‘for there is no turning away of death’. I do not think that we need to consult Lattimore’s Themes in Greek and Latin Epitaphs to recognise perhaps the most common funerary commonplace of them all – death is the one thing that awaits us all. As we know, Hellenistic epigram seeks out the unusual and paradoxical: the prophet who did not ‘know when the hour was at hand’ is certainly both of those. Finally, consider the tomb of Sthenelos: Ἔνθεν δὲ Σθενέλου τάφον ἔδρακον ᾿Aκτορίδαο, ὅς ῥά τ᾽ ᾿Aμαζονίδων πολυθαρσέος ἐκ πολέμοιο ἂψ ἀνιών - δὴ γὰρ συνανήλυθεν Ἡρακλῆι βλήμενος ἰῷ κεῖθεν, ἐπ᾽ ἀγχιάλου θάνεν ἀκτῆς. Next they saw the tomb of Sthenelos, son of Aktor, who died there on the sea-shore from an arrow wound; he was returning from the bold expedition against the Amazons on which he had accompanied Herakles. Arg. 2.911– 14, transl. R. Hunter

Death on the seashore is always a sad end, and one we associate with the epigrammatically shipwrecked, and all those who lie behind Horace’s Archytas ode. Did the Argonauts not just see the tomb but also – with the kind of unrealism always available to the epic poet – also read the inscription upon it¹¹? Did the inscription read ‘This is the tomb of Sthenelos son of Aktor, who died here on the seashore from an arrow wound…’. When Sthenelos’ ghost then appears above the tomb, ‘just as he was when he went off to war’, we think naturally of those representations on tombstones of warriors saying farewell as they go off to fight, never – as it turned out – to return. Here then epitaphic and epigrammatic practice is central to the meaning of the passage; that part of this ‘generic inclusivity’ depends upon an interest in the practice of reading¹² provides a suitably Hellenistic conclusion.

 See also Il. . – , above p. . For more Homeric tombs visible from the sea cf. Pearce ,  – .  According to Bing , it was not common practice to read inscriptions in the archaic and classical period before they acquire book form in the Hellenistic era; lack or sparsity of evidence however could simply be attributed to the fact that it was so common a practice as to be rendered insignificant.

Epigrams in Epic? The Case of Apollonius Rhodius

149

Bibliography Bing, P. (2002), ‛The Un-Read Muse?’ Inscribed Epigram and its Readers in Antiquity’, in: M.A.Harder / R.F.Regtuit / G.C.Wakker (eds), Hellenistic Epigrams, Leuven, 39 – 66. Dinter, M.T. (2005), ‛Epic and Epigram – Minor Heroes in Virgil’s Aeneid’, in: CQ 55, 153 – 169. Dinter, M.T. (2009), ‛Epic from Epigram: The Poetics of Valerius’ Flaccus Argonautica’, in: AJPh 130, 533 – 566. Elmer, D.F. (2005), ‛Helen Epigrammatopoios’, in: CQ 24, 1 – 39. Fantuzzi, M. / Hunter, R. (2012), Tradition and Innovation in Hellenistic Poetry, Cambridge. Gutzwiller, K.J. (1998), Poetic Garlands: Hellenistic Epigrams in Context, Berkeley / Los Angeles. Hunter, R. (1993), Apollonius of Rhodes, Jason and the Golden Fleece (Argonautica), Oxford. — (2015), Apollonius of Rhodes. Argonautica Book IV, Cambridge. Lattimore, R. (1942), Themes in Greek and Latin Epitaphs, Urbana. Matteo, R. (2007), Apollonio Rodio, Argonautiche, Libro II. Introduzione e commento. Lecce. Nelis, D. (2001), Vergil’s Aeneid and the Argonautica of Apollonius Rhodius, Leeds. Pearce, T.E.V. (1983), ‛The Tomb by the Sea: The History of a Motif’, in: Latomus 42, 110 – 115. Petrain, D. (2014), Homer in Stone: The Tabulae Iliacae in their Roman Context, Cambridge. Scodel, R. (1992), ‛Inscription, Absence, and Memory: Epic and Early Epitaph’, in: SIFC 10, 57 – 76 Squire, M. (2011), The Iliad in a Nutshell: Visualizing Epic on the Tabulae Iliacae, Oxford / New York. Svenbro, J. (1993), Phrasikleia. An Anthropology of Reading in Ancient Greece, Ithaca / London. Thomas, R.F. (1998), ‛Melodious Tears, Sepulchral Epigram and Generic Mobility’, in: G.C.Wakker (ed.), Genre in Hellenistic Poetry, Groningen, 205 – 22. Vox, O. (1975), ‛Epigrammi in Omero’, in: Belfagor 30, 67 – 70.

Jan Kwapisz

When Is a Riddle an Epigram?* Introduction (1) Εἷς ὁ πατήρ, παῖδες δυοκαίδεκα· τῶν δὲ ἑκάστῳ παῖδες τριάκοντα διάνδιχα εἶδος ἔχουσαι· αἱ μὲν λευκαὶ ἔασιν ἰδεῖν, αἱ δ’ αὖτε μέλαιναι· ἀθάνατοι δέ τ᾿ ἐοῦσαι ἀποφθινύθουσιν ἅπασαι. There is one father and twelve children. Each of these has twice thirty children of different aspect; some of them we see to be white and the others black, and though immortal, they all perish.¹ AP 14.101 (= Stob. 1.8.37; D.L. 1.90; Suda s.v. Κλεοβουλίνη [κ 1718 Adler])² What are they? The year, months, days and nights. (2) Εἶδον ἐγώ ποτε θῆρα δι᾿ ὕλης τμητοσιδήρου ὕπτιον ὀρθὰ τρέχοντα, ποσὶν δ’ οὐχ ἥπτετο γαίης. I once saw a beast running forward backwards through a wood cut by the steel, and its feet touched not the earth. AP 14.19 What did I see? A louse. (3) Εἷς ἄνεμος, δύο νῆες, ἐρέττουσιν δέκα ναῦται· εἷς δὲ κυβερνήτης ἀμφοτέρας ἐλάει. One wind, two ships, ten sailors rowing, and a steersman directs both. AP 14.14 What is it? A double-reed aulos. (4) Πικρή μοι ζωή, θάνατος γλυκύς, ὕδατα δ᾿ ἄμφω, θνῄσκω ἀναιμάκτοις ἔγχεσι νυσσόμενος·

* I intended this to merely be a working title, as it too presumptuously appropriates, without real relevance, the title of West , but I have decided to leave it as it is as a tribute to the author of, inter alia, one of the most illuminating recent discussions of Indo-European, and specifically Greek, riddles (West ,  – ). The present discussion was prepared with the financial support of the Polish National Centre of Science under grant No. DEC-//B/HS/.  The translations of the riddles of the Anthology are from Paton , slightly adapted.  With the exception of the Suda, our sources ascribe this riddle to Cleobulus.

152

Jan Kwapisz

ἢν δέ τις ἐν ζώοντι νέκυν τύμβῳ με καλύψῃ, αἵματι συγγενέων πρῶτον ἀποβρέχομαι. Bitter is my life, my death is sweet, and both are water [salt water and sweet sauce]. I die pierced by bloodless spears [fire and heat]; but if anyone will cover me when dead in a living tomb [stomach], I am first moistened by the blood of my relations [sauce]. AP 14.36 What am I? A fish. (5) Κτανθεὶς τὸν κτείναντα κατέκτανον· ἀλλ᾿ ὁ μὲν οὐδ᾿ ὣς ἤλυθεν εἰς ᾿Aίδην· αὐτὰρ ἔγωγ᾿ ἔθανον. Slain, I slew the slayer, but even so he went not to Hades; but I died. AP 14.32 Who am I? The Centaur Nessus. (6) Εἰμὶ χαμαίζηλον ζῴων μέλος· ἢν δ᾿ ἀφέλῃς μου γράμμα μόνον, κεφαλῆς γίνομαι ἄλλο μέρος· ἢν δ᾿ ἕτερον, ζῷον πάλιν ἔσσομαι· ἢν δὲ καὶ ἄλλο, οὐ μόνον εὑρήσεις, ἀλλὰ διηκόσια. I am the part of an animal which affects the ground, and if you take a single letter away from me I become a part of the head. If you take away another I shall again be an animal, and if you take yet another away you will not find me one, but two hundred. AP 14.105 What am I? A foot (πούς), and an ear (οὖς), and a pig (ὗς), and ς (the number 200).

These six riddles, which I have chosen to exemplify different classes of riddles, are a part of a collection of about fifty riddles which we find in Book 14 of the Palatine Anthology. This is one of the two largest extant corpora of Greek riddles, alongside the excursus that is mostly based on Clearchus of Soli’s treatise Περὶ γρίφων which is preserved in Athenaeus’ Book 10.³ These are riddles proper, such as are attested in many Indo-European cultures, i. e. short poems formulated by the Riddler in such a way so as to propound a problem to which the Guesser is expected to find the solution (not a solution; the traditional riddle is not a postmodernist game).⁴ The Byzantine compiler of Book 14 of the Anthology ⁵ clearly felt that there was a generic affinity between riddles and epigrams, as he appended this collection to a generically uniform collection of epigrams. To

 On these two corpora, see esp. Berra ; on the riddles of the Anthology, see Forster .  Cf. Potamiti , .  On the origin of Book , see Cameron ,  – .

When Is a Riddle an Epigram?

153

adduce two further examples – there is a riddle among Ausonius’ epigrams (Epigr. 85 Green), and Symphosius’ riddles are in many ways linked with the epigrams collected in Martial’s books of Xenia and Apophoreta. ⁶ These are all illustrations of when a riddle clearly is an epigram; in later eras this process of a generic transformation of riddles into a subclass of epigram was complete.⁷ On the whole, the notion that proper riddles form a subcategory of the epigram is a common sentiment.⁸ K. Ohlert, who more than a century ago authored what still remains one of the most comprehensive discussions of ancient Greek riddles,⁹ expressed the same opinion: ‘Zahlreiche Epigramme könnten als Rätsel gelten, wenn man ihnen die Überschrift nimmt, zahlreiche Rätsel als Epigramme, wenn man die Aufforderung zum Raten entfernt’.¹⁰ This intuition, sound enough as it is, brings us no closer to explaining this phenomenon. Is it possible to define with more precision the nature of the relationship between riddles and epigrams? An attempt to answer this question will take us far back in time.

 See Leary ,  – . There is a tenuous possibility that yet another example, which would be remarkably early and therefore highly relevant for the present discussion, lurks among the recently published ‘Vienna epigrams’ of P. Vindob. G , as the incipit at col. ii  is, as was noticed by Cornelia Römer, identical with the incipit of AP .; see Parsons et al. , .  Also in post-mediaeval Europe the category of poesis epigrammatica encompasses poesis artificiosa, which includes riddles; see Kwapisz , .  In what follows, my concern is only with such ‘proper riddles’, strictly defined, and the possible points of contact between them and the epigram; as a consequence, I will not be interested in various modes of enigmatic or playful discourse which can be detected in epigrams of different eras but whose connection to the proper riddles is indirect, such as, e. g. the so-called trigônon epigramma, on which see sch. Il. . – , with Elmer ,  (I am grateful to Ivana Petrovic for turning my attention to this phenomenon), or the manifestations of ‘irony’ in epigrams as discussed by Pittore . Kirstein  is an interesting proposal of approaching ancient riddles (note the term Rätselepigramm as used by the author) from the viewpoint of G.E. Lessing’s theory of the epigram, and as such also offers a reflection on the interconnectedness of the two genres.  Yet see also Schultz  and, among the recent publications, Luz ,  – , Monda , Beta , Kwapisz et al. ,  –  and Potamiti .  Ohlert , ; see also Ohlert ,  – .

154

Jan Kwapisz

Epigrammatisation through Hellenisation? Greek riddles are notoriously difficult to date;¹¹ most of those that have reached us might be located at any point in the history of Greek culture. Yet they are not completely ahistorical. What allows us to distinguish between various types of Greek riddles is the fact that the genre is older than Greek culture itself. This becomes clear when we compare the extant Greek riddles, as the late M. West did in a recent discussion,¹² with the Indo-European material from outside the Mediterranean world. What makes the ‘year riddle’ (No. 1 as cited above) more ancient than the other examples I supplied above is not merely the fact that it was ascribed to Cleobulus, hence even the Greeks must have recognised its antiquity. What is more significant is that similar riddles are attested in many cultures outside the Greek world,¹³ and further parallels show that prehistoric riddles often mystified the phenomena of Nature, in particular those whose essence continues to be elusive and which are, so to speak, enigmatic in themselves: e. g., Day and Night, Smoke, Sleep, Echo, Silence.¹⁴ Lice may be less mysterious, yet the riddle of the louse (No. 2) belongs to the same category as it also has parallels in non-Hellenic cultures.¹⁵ Moreover, the narrative structure that this riddle exhibits is of a sort that is familiar to us from a number of Indo-European riddles which, in view of this structure, can be labelled ‘I saw riddles’.¹⁶ Of course, this riddle is only a variation on the ancient theme and I do not intend to state that it was actually composed by a prehistoric barber (for one thing it does mention iron). Yet we can safely assume that some of the extant Greek riddles exhibit a more ancient ancestry than others because of how they are rooted in the Indo-European tradition. One may suppose that the earliest Greek riddles were much unlike epigrams. They exhibited fewer variations and a limited number of themes, and they were bound by strict narrative conventions, such as the ‘I saw’ pattern.¹⁷ They were probably in hexameters, such as the year riddle and the other riddles that

 Cf., e. g. Luz ,  –  n. .  West ,  – .  West ,  – .  Cf. the material discussed by West ,  – .  Symphosius  (=  Leary) may be an instance of how ‘S[ymphosius] sometimes recycles traditional riddles’ (Leary , ), but similar louse riddles are widespread even outside Indo-European cultures; cf., e. g. Sila Basak ,  and Miruka , .  On this class, see West ,  – .  See further West ,  – .

When Is a Riddle an Epigram?

155

have a tinge of archaicness.¹⁸ They shared metre with the earliest inscriptional epigrams,¹⁹ and as such must have sounded somewhat archaic to later ears which were accustomed to elegiac epigrams. In a number of later riddles we see clear enough traces of the locus in which the process of Hellenisation of the riddle took place. We have plenty of evidence to show that the most common venue for engaging in riddle games in the Greek world was the symposium.²⁰ This is also evidenced by the riddles themselves, which often describe the sympotic apparatus. This includes musical instruments (as in No. 3) and snacks to serve with wine (as in No. 4; an elaborate gastronomical riddle).²¹ The hexameter may have been recognised as the traditional metre of riddle, but most of the riddles preserved in the Palatine Anthology are elegiac. This shift in metrical preferences was probably due to the fact that riddles shared their venue with the eagerly cultivated genre of sympotic elegy.²² In view of this evidence about the importance of the symposium for the evolution of the genre, I should like to suggest that the riddle may owe to the symposium the development of one of its most emblematic narrative conventions. The Ich-Rede is the usual narrative mode of Greek riddles (as in Nos. 4, 5 and 6). It is a widespread feature that is familiar to us from many riddles from outside Greece. However, M. West warns us that the ‘European distribution’ of this feature is ‘compatible with a Classical origin’.²³ In other words, there is no evidence for prehistoric first-person riddles. Perhaps this should not surprise us. The IchRede, with the game of changing identities which it entails when the first-person text is read aloud or heard, is a striking and even curious feature. Analogously, its employment by the Greeks in epigraphic texts is one of those ‘odd notions’ that are at once so emblematic of Greek culture and that make it so ‘exotic’ within the universe of comparable world cultures.²⁴ As R. Wachter puts it with regard to oggetti parlanti:

 Cf. West , . It is telling that Athenian playwrights used hexameters for the riddles embedded in their dramas; see West , .  On the metre of early epigrams, see Bowie ,  – .  Cf. Potamiti  and already Ohlert ,  – .  Cf. Kwapisz b, .  My view is influenced by the discussion of the metrical preferences of the authors of epigrams in Bowie ,  –  (: ‘A factor in both the adoption and then the dominance of elegiacs seems to me very likely to have been the rise of the elegiac couplet as a performance metre in symposia and festivals’).  West , .  For another ‘odd notion’ – note, again, the sympotic context – and for the source of this expression see Rotroff ,  on Hellenistic pottery: ‘The Greeks abandoned the odd notion of

156

Jan Kwapisz

It is a fact that we do not have any pre-Greek attestations of the speaking-object type of inscription; there are none from Egypt or the Near East or Mesopotamia, as many colleagues have assured me. It seems to be another clever invention by the Greeks in the earliest day of their alphabet, illustrating their fast-developing sense of individuality.²⁵

In view of the fact that the Greeks betray a distinctive predilection for giving voice to inanimate objects, I feel tempted to follow West’s faint suggestion and to posit that the peculiar type of riddle-solving game in which the Riddler plays the role of the object to be guessed was a Greek innovation. It is not unlikely, I believe, that this innovation originated in the playful and performative atmosphere of the Greek banquet. As a parallel, one can think of modern drinking games, such as the ‘Who Am I’ game as portrayed in Quentin Tarantino’s move Inglourious Basterds: the spontaneous sympotic role-playing would have provided a natural context for the invention of a variation of the riddling contest which became, effectively, a sort of role-playing game. Can the Ich-Rede of riddles have anything to do with the common epigraphic and epigrammatic convention? The debate on the origins of this narrative feature and of the tradition of the so-called ‘speaking objects’ has been long and fervent.²⁶ Without entering it, I should nevertheless like to point out that recent archaeological evidence suggests the relevance of the sympotic context for the early history of the epigraphic Ich-Rede. Strikingly, the ground-breaking recent finds from Pierian Methone have resulted in bringing to light a late eighth-century first-person iambic inscription – a playful ownership formula – on what is clearly a piece of sympotic equipment dating from the late eighth or early seventh century BC (an Euboean skyphos; Μθ 2248; Besios et al. 2012, 341):²⁷ ̓ [ὶ ……….c.22……….]ει ̣τετο[..c.6..]μεκ[..c.6..]ατον στερε̄-́ ← hΑκεσάνδρō ε̄μ σ[ετ]α̣ι. ←

This Hakesandros inscription is parallelled by finds from later periods (in particular by SEG 47.1475, the aryballos from Kyme with Tataie’s inscription, c. 675 –

decorating their pottery with human figures – for it is an odd notion, shared by only a handful of other ceramic traditions in the long history of fired pottery’.  Wachter , .  Important recent studies include Wachter , Day ,  – , Tueller ,  – , Meyer .  Cf. Wecowski ,  – , who convincingly explains the meaning of such ownership formulas in the sympotic context by arguing that ‘asserting the ownership of one’s cup during a symposion was crucial since cups were supposed to circulate’; see also Wecowski forthcoming.

When Is a Riddle an Epigram?

157

650?); on the basis of these analogies the following reconstruction of the inscription’s text was suggested:²⁸ I belong to Hakesandros; [whoever steals me from him (?)] will lose his eyes (or will be deprived of his money?).

Here is a remarkably early document of the presence of a playful first-person poetic text at the symposium.²⁹ Of course, this is not a riddle, and this inscription can by no means serve to definitely prove that first-person riddles were also present at the symposium at such an early date. What this ownership formula does show, however, is that even relatively early in the history of Greek culture was the use of the epigraphic Ich-Rede extended beyond dedicatory or sepulchral inscriptions, as it clearly felt at home at symposia. Moreover, my impression is that there is some natural kinship between the ‘odd notion’ of making objects speak and the distorted reality that was temporarily created within the peculiar atmosphere of sympotic entertainment. This kinship extends, I would argue, to yet another ‘odd notion’, namely that of the Riddler virtually becoming the object to be guessed. The conclusion which these remarks lead me to is that certain formal similarities between riddles and epigrams, such as the penchant for the first-person narrative and elegiac metre, may not be due to a mutual dependence of one genre upon the other, but rather due to the fact that both genres evolved in the same peculiar Hellenic environment. In particular, they both remained within the sphere of the far-ranging influence of the most productive institution of Greek culture – the symposium. To be sure, Hellenisation of the riddle was deeper than the process I am trying to sketch here. Numerous typically Hellenic riddles did not necessarily test the erudition directly connected with the symposium but the sort of erudition that was expected from anyone endeavouring to participate in Greek paideia. Such were the mythological riddles (such as No. 5 above) or riddles whose sole concern was to test linguistic competence by a play of words (No. 6).³⁰ We may imagine that both types were common at symposia as well as, for instance, at school.³¹ Yet the general pattern which now emerges  Besios et al. , ; cf. Wecowski ,  n. .  A similar example of an even more elaborate contemporary inscription of this sort is probably provided by the famous ‘Cup of Nestor’, yet the presence of the Ich-Rede in this tercet is uncertain due to its fragmentariness, and therefore I have excluded it from the present discussion. For a relevant and illuminating discussion of the ‘Cup of Nestor’ inscription, including remarks on the Hakesandros inscription, see Wecowski  and forthcoming.  These and other categories of Greek riddles are discussed by Luz .  On riddles as school texts, see the remarks in Bartol , .

158

Jan Kwapisz

appears to be this: the more distant the riddle was from its Indo-European origin and the more Hellenised it became, the less clear its generic identity was and the more it resembled epigrams. This process was to reach its climax in an entirely new era of Greek culture.

Literary riddles and their offspring What changed the rules of the game was the transition of the riddle from the oral domain (especially the domain of the symposium) to written culture. Contrary to the epigram, the riddle is an oral genre par excellence;³² it entails direct interaction between the Riddler and the Guesser. However, at the dawn of Hellenistic book culture, when poets were more prone than ever to experiment with writing, a new type of elaborate literary riddle was conceived. In such bookish riddles the lack of the play of oral interaction was compensated by an increased richness of sophisticated allusiveness. As a result, what characterises such riddles is that either the solution can be found only after repeated rereadings or at least a number of rereadings are necessary to fully appreciate the intricate way the object to be guessed has been concealed.³³ A famous example of such a literary riddle is the Oyster Riddle, preserved on papyrus on which it is accompanied by an extensive commentary (P. Louvre inv. 7733v + D’Alessio 1990 = SH 983 (poem) + 984 (commentary) = anon. 153 FGE).³⁴ The general consensus is that the poem was composed at some point in the third century BC.³⁵ On one level, this is a gastronomical riddle, similar to the one cited above (No. 4).³⁶ As such it might be seen to fit the context of a sympotic riddling contest;³⁷ predictably, the narrative is first-person and the metre is elegiac. In yet other respects this is a traditional riddle; perhaps because oysters are enigmatic per se (but the sexual connotations may have something to

 For a discussion of Greek riddles as an oral genre, see Potamiti . Bevilacqua and Ricci (, ) comment on the scarcity of evidence for ancient epigraphic riddles and explain it by the oral nature of riddles.  This procedure is, of course, applicable to other Hellenistic book poetry; its description is provided by Bing ,  – .  The text is highly fragmentary and too problematic to be supplied here; an extensive discussion, with a detailed account of the status quaestionis, has recently been provided by Martis . For the English translation, see Parsons ; note, however, a reappraisal of the text as offered in D’Alessio  (yet see H. Lloyd-Jones’ comment on line  in SSH p. ).  Martis ,  and  n. .  This parallelism is noted by Sbardella , ; cf. Martis , .  Cf. Kwapisz b,  – .

When Is a Riddle an Epigram?

159

do with this popularity), they appear in many riddles around the world,³⁸ and the ‘I am X, not Y’ pattern which most likely appeared in lines 2– 3 (‘I was not nursed by the Nile, but by the Ocean’) is familiar to us from other Greek riddles and may be rooted in their Indo-European past.³⁹ Yet it would obviously be a mistake to regard this curious short poem as no more than a traditional riddle. The commentary, much longer than the poem itself, is testimony to the poem’s stunning exegetical richness. The comments of an unknown scholiast are, to us, a display of enviable learnedness; the fragments by Sophocles (fr. 966a Radt), Diphilus (fr. 59 Kassel-Austin) and even by the third-century epigrammatist Theodoridas (SH 743) are adduced as comparanda. Since the sources of these quotations are lost to us, we may miss some allusions, but we are still capable of appreciating the refined diction of this poem. It contained at least two hapaxes⁴⁰ and riddling periphrases such as Δωσοῦς [νυμφί]ο̣ς,̣ ‘Aphrodite’s lover (?)’, for a knife (lines 5 – 6).⁴¹ Proper appreciation is possible only when the poem is carefully read; the lengthy commentary which accompanies it is a necessary one. According to P. Parsons, the Oyster Riddle ‘has only one congener, among the Παίγνια of Philetas’.⁴² This takes us to the no less famous Alder Riddle, which our source attributes to the book of Παίγνια by Philitas of Cos:⁴³ οὐ μέ τις ἐξ ὀρέων ἀποφώλιος ἀγροιώτης αἱρήσει κλήθρην, αἰρόμενος μακέλην· ἀλλ᾿ ἐπέων εἰδὼς κόσμον καὶ πολλὰ μογήσας, μύθων παντοίων οἶμον ἐπιστάμενος. No lumbering rustic from the mountains shall bear me, snatching up a hoe – me, an alder tree; but one who knows the marshalling of words, who toils, who knows the pathways of all forms of speech. Philit. fr. 8 Lightfoot = 25 Spanoudakis = 12 Sbardella ap. Stob. 2.4.5

We note, again, the elegiacs and the Ich-Rede. The modern scholarship that surrounds this quatrain is a clear illustration of the exegetical torments to which a literary riddle may subject the Guesser. What is this alder which addresses us? Is

 E. g., the Exeter Book collection of Old English riddles includes an oyster poem ( Williamson), and there is one more to dishcover in Lewis Carroll’s Through the Looking Glass.  See Kwapisz a,  –  and cf. West ,  – .  See Martis , ; cf. Sbardella , .  Martis (,  – ) rightly points to the affinity with the cryptic language of Lycophron’s Alexandra, Pseudo-Theocritus’ Syrinx and Dosiadas’ Altar.  Parsons , ; this follows Lasserre ,  – , who argued for the Philitan authorship of the Oyster Riddle (cf. Martis , ).  The translation is from Lightfoot , slightly adapted.

160

Jan Kwapisz

it a writing tablet? A poetic staff? The title of a poetry book by Philitas? Or something altogether different?⁴⁴ My firm conviction is that the correct solution is the one that was propounded a decade ago by G. Cerri.⁴⁵ In order to find it he went much along the same path that the Oyster Riddle takes the Guesser. The first step was to read through the poem’s intertextuality. Cerri saw the Alder Riddle to be heavily allusive to the Odyssey;⁴⁶ it echoes the descriptions of Odysseus in which he is referred to as οὐκ ἀποφώλια εἰδώς, ‘knowing clever things’, and as the one who speaks ἐπισταμένως, ‘wisely’, and in whom there is μορφὴ ἐπέων, a ‘grace of words’.⁴⁷ Once we take account of these allusions, we begin to understand that this riddle is more about Odysseus than about the alder. Guided by Homer, we find an alder in the Odyssey, in Book 5 – it is one of the trees Odysseus used to build a raft on Calypso’s island. The word κλήθρη may seem to be neither particularly rare nor specially refined, but it occurs only twice in Homer (Od. 5.64 and 239); both occurrences are in a description of Calypso’s island and, furthermore, this is the sole appearance of this word in poetry before Philitas. There can be no doubt now that this alder, mentioned in passing by Homer, is the alder which Philitas wants us to find. The solution is surprising; once we have it we realise that we have asked the wrong question. Instead of asking ‘What is the alder?’ we should have asked ‘Which alder is this?’. For the answer is ‘the Homeric alder’. Indeed, it takes one who knows the pathways of poetry to solve this riddle! In a recent discussion of the Oyster Riddle, C. Martis devoted a good deal of attention to the problem of the poem’s generic identity.⁴⁸ Is it an elegy or rather an epigram? We certainly feel that it is more than just a riddle, and it is probably evident by now that the main difference between this poem and, for instance, the gastronomical riddle from the Palatine Anthology (No. 4), and other ‘proper riddles’ as well, is that the former is imbued with enhanced, or bookish, litera-

 For a list of the proposed solutions, see Sbardella ,  –  and Spanoudakis ,  – .  Cerri ; cf. Kwapisz b,  – .  Cf. also Bing ,  n. , Garriga  and Sier .  Cf. esp. Od. . –  and . – . Note that the words chosen by Philitas to trigger these intertextual links turn out, once the epic intertext of the Alder Riddle is identified, to have a special ring to them, as they give a very apt characterisation precisely of the Odyssey; it does pay off to know some ἔπεα by heart, especially those that tell the story about οἶμος – and especially when it is οἶμος μύθων! (I am grateful to Benjamin Acosta-Hughes for a comment that took me on this path; in addition, he pointed out to me that μάκελλα is a hapax at Il. . – surely a good enough reason for Philitas to have used a variant of this word.)  Martis ,  – ; there is no conflict between her conclusions and the somewhat terser conclusion presented here.

When Is a Riddle an Epigram?

161

riness. This effect of literariness is an aspect which both the Oyster Riddle and the Alder Riddle share with the Hellenistic epigram. What traditional riddles are about, even those that are more complex, is, in the end, always about asking a simple question, whereas the bookish riddles such as the two we have seen do much more, i. e. they are not done with the reader even when he or she manages to find the solution; they want to be repeatedly reread and to dazzle us with their ingenuity. What they offer is, in effect, indistinguishable from the many literary epigrams, or at least the more ambitious ones, an attempt to impress the reader, the only difference being the clearly marked presence of the question, which makes every riddle, even a literary one, a riddle. We have already seen that Stobaeus, who preserves Philitas’ Alder Riddle, refers to it with the term paignion, and although its precise meaning and this category’s relationship to the epigram are unclear,⁴⁹ one feels that ‘the little plaything’ is an appropriate label for such playful poems which are neither exactly riddles nor epigrams, but a little of both. Although no more specimens of the class of literary riddles are extant, this pair, whose primary concern is ex definitione with language and play, sufficiently shows the literary potential of such elaborate compositions to Hellenistic scholar-poets, who were at once the authors and the audience of such poetry.⁵⁰ It was seen that the Alder Riddle carries strong programmatic overtones; in accordance with Philitas’ poetic programme, the quatrain offers a leçon par l’exemple, which at once tells of and illustrates what poetry should be like.⁵¹ Yet also the Oyster Riddle as we find it on a papyrus scrap – a short poem with a long commentary – provides a vivid illustration of what pleasures a poet could bring to a well-read reader. These are path-showing poems, and we will now see how this path was pursued by other poets. The working definition of the paignion which we now have may encompass other poems besides this pair, and it is not surprising that all of these poems which bear some structural resemblance to riddles are, strictly speaking, epigrams (in one case an epigrammatic technopaegnion) – a fact which supplies further testimony to the interconnectedness between the two genres. On the spectrum between the riddle and the epigram, these epigrammatic paignia on which the remaining part of the present discussion will focus are located further away from ‘proper riddles’ than the Oyster Riddle and even

 For different, though not necessarily conflicting, views, see Reitzenstein ,  –  n. , Gutzwiller ,  – , Sbardella ,  – , Spanoudakis ,  – , Kwapisz b,  – , Martis ,  – .  That such poems are composed by grammatikoi for an audience of grammatikoi is implied by SH , a fragment of Philicus’ Hymn to Demeter. Bing (, ) remarks that the riddle is ‘a form particularly suited to current [i. e. Hellenistic] tastes’.  E. g. Bing , Sbardella ,  – , Cerri , .

162

Jan Kwapisz

the Alder Riddle, yet, as I hope to show, not entirely outside this spectrum. Their ties with the tradition of the riddle to a certain extent define, and complicate, their generic identity. It is customary to link the class of paignia with the famous six Greek pattern poems which have been dubbed by modern criticism as technopaegnia, although the latter denomination, if surprisingly apt, is definitely of late origin.⁵² I have extensively discussed this relationship elsewhere;⁵³ for the purpose of the present discussion, however, it will be useful to have a closer look at one of the technopaegnia, namely Simias of Rhodes’ Wings of Eros. That this poem was composed in the same milieu in which Philitas’ Alder Riddle was conceived is practically guaranteed by the fact that Simias not only belonged to the same generation as Philitas but also was, like the famous Coan, both a poet and a scholar, and his poetry was abundant in rare glosses and learned allusions.⁵⁴ The Wings of Eros purports to be an inscription on a statuette of Eros. A mini-cosmology jostles the epigrammatic frame; with strong Platonic overtones, the poem describes the paradoxes of being the God of Love:⁵⁵ Λεῦσσέ με τὸν Γᾶς τε βαθυστέρνου ἄνακτ’ ᾿Aκμονίδαν τ’ ἄλλυδις ἑδράσαντα, μηδὲ τρέσῃς, εἰ τόσος ὤν δάσκια βέβριθα λάχναις γένεια· τᾶμος ἐγὼ γὰρ γενόμαν, ἁνίκ’ ἔκραιν’ ᾿Aνάγκα, πάντα δὲ γᾶς εἶκε φραδαῖσι λυγραῖς ἑρπετά, πάνθ’, ὅσ’ ἕρπει 5 δι’ αἴθρας, Χάους τε, οὔτι γε Κύπριδος παῖς, ὠκυπέτας δ’ ἁβρὸς Ἔρος καλεῦμαι· οὔτι γὰρ ἔκρανα βίᾳ, πραϋνόῳ δὲ πειθοῖ· 10 εἶκε δέ μοι γαῖα θαλάσσας τε μυχοὶ χάλκεος οὐρανός τε· τῶν δ’ ἐγὼ ἐκνοσφισάμαν ὠγύγιον σκᾶπτρον, ἔκρινον δὲ θεοῖς θέμιστας. Look on me, the lord of broad-bosomed Earth, who established the Heaven elsewhere, and tremble not that someone so mighty should have a shade of down on his cheeks. For I was born when Necessity was ruler, and all creatures of earth and those that move through the sky yielded to her dire decrees. A child of Chaos I was born, not of Aphrodite; I am called

 For the story of the term, see Kwapisz a,  – , with a supplementary remark in Kwapisz ,  n. .  Kwapisz a,  – . See already Wilamowitz ,  –  and cf. Martis ,  –  (already quoted) and Pittore ,  – .  See Kwapisz a,  – , Perale ,  n. . I am currently preparing an extensive discussion of Simias’ poetic and grammatical fragments.  The text and translation are as printed in Kwapisz a, yet the latter is adapted from Paton . See Kwapisz a,  –  for my discussion and interpretation of the poem.

When Is a Riddle an Epigram?

163

swift-flying pretty Love, for in no wise did I rule by force, but by gentle-minded persuasion. Earth, the depths of the sea and the brazen heaven yielded to me; I robbed an ancient sceptre to rule them and gave laws to the gods.

The Ich-Rede of this poem may be more than a nod towards the epigrammatic convention.⁵⁶ The crucial information about the identity of the speaker is withheld from the reader until line 9. In fact, even this late self-introduction of Eros is uncertain; the transmitted text is corrupt and among the proposed emendations there are conjectures which altogether remove the name of Eros from this poem, as the following excerpt from my apparatus to line 9 demonstrates:⁵⁷ δ᾿ ἁβρὸς Ἔρως Bergk (cf. Σ οὐκ εἰμὶ δὲ ὁ ᾿Aφροδίτης υἱός, καλοῦμαι δὲ Ἔρως); Ἔρος scripsi; δ᾿ ἀέρος G K; δ᾿ ἀέριος Anth.; δ᾿ ἔρως Iunt.; δ᾿ Ἄρεος Call.; οὐδ᾿ Ἄρεος Wilamowitz; ἠδ᾿ Ἄρεος Powell; ᾿Aρέιος Edmonds; δ᾿ ἁδὺς Ἔρως Buffière

Furthermore, the persona loquens is, quite emphatically, characterised through the same pattern ‘I am X, not Y’, which we have already seen as typical of riddles: ‘I am not a child of Aphrodite, but of Chaos’ (lines 7– 8); ‘I do not rule by force, but by gentle persuasion’ (line 10), ‘I may look young, but I am very old’ (lines 2– 3 and ff.).⁵⁸ Perhaps even the opening Λεῦσσέ με, ‘behold me’, in which modern critics saw (and rightly so) a generic marker which, on the one hand, immediately places the poem in the tradition of the ecphrastic epigram and, on the other, emphasises its ontological status as a visual textphenomenon,⁵⁹ may be seen, in addition, to faintly suggest the poetics of a riddle. It reverses εἶδον, ‘I saw’, which is already familiar to us as a conventional riddle incipit (as in No. 2 above). Although the genre towards which the technopaegnia tend to gravitate is, above all, the epigram,⁶⁰ the Wings is arguably influenced also by the tradition of the riddle – quite aptly, as the riddling diction corresponds with how Eros presents himself as a paradox and an enigma and, moreover, enhances the effect of disorientation which the poem’s visual form creates by making the reader ask him or herself the crucial question of ‘What do I see?’. If the Wings is not exactly a paignion in the same sense that the Alder Riddle is, then at least its ties with this poetic class are notable.

 Cf. Kwapisz a,  –  and  – .  Kwapisz a, .  See further Kwapisz a,  and  –  on the curious presence of this pattern in all six extant technopaegnia.  Cf. Kwapisz a,  –  and Pappas , .  Cf. Kwapisz a,  – .

164

Jan Kwapisz

Simias’ liking for such quasi-riddles is attested by yet another composition. Among his few epigrams that have reached us we find a mock-inscription for a cage containing a locust:⁶¹ Τάνδε κατ’ εὔδενδρον στείβων δρίος εἴρυσα χειρὶ πτώσσουσαν βρομίας οἰνάδος ἐν πετάλοις, ὄφρα μοι εὐερκεῖ καναχὰν δόμῳ ἔνδοθι θείη, τερπνὰ δι’ ἀγλώσσου φθεγγομένα στόματος. This locust crouching in the leaves of a vine I caught as I was walking in this copse of many trees, so that in a well-fenced home it may make noise for me, chirping pleasantly with its tongueless mouth. Sim. 2 Gow-Page = 25 Fränkel = AP 7.193

Yet how do we know what this epigram is about? To be sure, the ‘locust’ in the translation is supplemented by the translator. In Book 7 of the Palatine Anthology this epigram is surrounded by a series of poems on locusts, but unless Simias’ original epigram book provided a similar context – which is, I admit, a possibility – the reader was expected to figure out the identity of the caught insect by him or herself. The deictic pronoun τάνδε provides a stimulus to the reader’s imagination as it hints only at the female gender of the prey. We understand that the game at play here is Ergänzungsspiel, but at the same time we are reminded of how much this epigrammatic game has to do with riddles;⁶² the question ‘What have I caught’ is clearly implied. The definitive answer comes only with the last line, but it is concealed in a riddle – ‘She sings, although she has no tongue’, which is a description provided by the clearly riddling paradox of the ‘tongueless mouth’ (ἄγλωσσον στόμα). We recognise the familiar convention. It takes a learned reader to solve this riddle; one has to approach the animal world scientifically to find out that whereas most animals do indeed need a tongue to produce sound, some insects have learnt to use other parts of their bodies for this purpose, and specifically ἀκρίδες produce sound by rubbing their hind legs (so Simias’ ‘mouth’ is only a metaphor). All of this information is provided by Aristotle in Book 4 of his treatise Historia Animalium (on ἀκρίδες, see 535b11– 12), a work to which, as I will argue elsewhere, Simias’ epigrams on musical animals emphatically allude. With the epigram on the locust, Simias introduces a wholly new dimension to the old tradition of riddles of Nature – in

 The translation is from Paton , slightly adapted. AP ., an epitaph for a partridge, is a companion epigram; I will extensively discuss this pair elsewhere; for now, see the instructive discussion of the rich Homeric intertext of Simias’ epigrams in Sistakou ,  – .  Kirstein () also reflects, if indirectly, on the link between Ergänzungsspiel and riddles.

When Is a Riddle an Epigram?

165

typically Hellenistic fashion the world of poetry intermingles here with the world of science.⁶³ All in all, it is not surprising that the literary potential of riddles was noticed, and used, by one of Philitas’ peers. A pattern which emerges here suggests that the tradition of the riddle was eagerly explored by poets at the dawn of the Hellenistic age, when they sought new sources of inspiration for their poetic experiments. My next example reflects a further elaboration of this tradition in the poetry of Philitas’ successors. The poem I have in mind is Callimachus’ much-discussed epigram on the nautilus offered to Arsinoe-Aphrodite-Zephyritis:⁶⁴ Κόγχος ἐγώ, Ζεφυρῖτι, παλαίτερον, ἀλλὰ σὺ νῦν με, Κύπρι, Σεληναίης ἄνθεμα πρῶτον ἔχεις, ναυτίλος ὃς πελάγεσσιν ἐπέπλεον, εἰ μὲν ἀῆται, τείνας οἰκείων λαῖφος ἀπὸ προτόνων, εἰ δὲ γαληναίη, λιπαρὴ θεός, οὖλος ἐρέσσων ποσσίν – ἴδ’ ὡς τὤργῳ τοὔνομα συμφέρεται – ἔστ᾿ ἔπεσον παρὰ θῖνας Ἰουλίδας, ὄφρα γένωμαι σοὶ τὸ περίσκεπτον παίγνιον, ᾿Aρσινόη, μηδέ μοι ἐν θαλάμῃσιν ἔθ’ ὡς πάρος – εἰμὶ γὰρ ἄπνους – τίκτηται νοτερῆς ὤεον ἁλκυόνος.

5

10

Κλεινίου ἀλλὰ θυγατρὶ δίδου χάριν· οἶδε γὰρ ἐσθλά ῥέζειν καὶ Σμύρνης ἐστὶν ἀπ᾿ Αἰολίδος. Long ago, Zephyritis, I was a conch, but now you, Kypris, have me, the first votive offering of Selenaia. I who used to sail as a nautilus on the seas, if there was a breeze, stretching out my sail from my own forestays, but if it was calm, radiant goddess, rapidly rowing 5 with my feet – and so my name is fitting to the action – until I fell upon the shores of Iulis, that I might be for you, Arsinoe, your much admired plaything, nor in my chambers any longer as before – for I am without breath – may the egg of the sea halcyon be laid. 10 But show favour to the daughter of Kleinias. For she knows to do good works, and is from Aeolian Smyrna. Call. Ep. 5 Pfeiffer = 14 Gow-Page ap. Ath. 7.318b

 Cf. the collection of essays in Harder et al. .  The translation is from Acosta-Hughes , slightly adapted. On the spacing in the Greek text – which mirrors the presentation of the text in Prescott  – see below.

166

Jan Kwapisz

Unlike the important discussions which focused on an interpretation of this intricate composition and on how it reflects Ptolemaic propaganda,⁶⁵ I am concerned with this epigram’s form and style. It was observed first by L. Sbardella and later by C. Martis that this poem exhibits curious thematic links with the Oyster Riddle; in addition, Martis points out linguistic similarities.⁶⁶ Both sea creatures speak, out of their shells, in the first person; the diction is elaborate and the use of Homeric glosses is a remarkable feature in both cases.⁶⁷ Given these similarities, Sbardella makes a (highly attractive, I think) suggestion that the word παίγνιον which appears in line 8 has a double meaning, as it simultaneously refers to the material object which the poem describes and, as a metapoetic allusion, to the poem itself ⁶⁸ – ‘pur non essendo un γρῖφος’, as he stresses.⁶⁹ I venture to go slightly further in the same direction – perhaps this poem does, after all, have to it some ring of a riddle? In general, the diction of Callimachus’ epigrams is far from the formal elaboration of the paignia we saw above. E. Magnelli observed that the style of Callimachus’ epigrams, unlike the rest of his poetry, is relatively plain and simple; as he puts it with regard to a sample epigram he discusses: No affectation here, no abstruse vocabulary or erudite allusions; rare words are very easy to understand, so that their preciosity is almost disguised … This holds true for almost all of Callimachus’ epigrams.

That is, with one exception; he immediately adds: ‘Only the poem on the nautilus offered to Arsinoe … appears to be a little less plain’.⁷⁰ Although this diagnosis is correct, I do not think that it can be properly applied to the whole poem. As was noticed by D. Fredrick, and before him by H.W. Prescott (whence the quotation), the structure of the nautilus epigram is bipartite: ‘Verses 1– 2 and 11– 12 are complete in themselves and reproduce the simple type of inscription’ (Fredrick labels this part the ‘frame’), whereas ‘verses 3 – 10 [Fredrick’s ‘inset’] are a pedantic inlay

 Gutzwiller , Selden ,  – .  Sbardella , , Martis ,  – .  On the Homeric diction of the nautilus poem, see further Schroeder .  The metapoetic potential of the word παίγνιον is, I argue, further highlighted by the accompanying adjective περίσκεπτον, which is a clear Homeric echo (see Schroeder , ), but also conveys the notion of the poem’s elaboration. Note that παίγνιον clearly has this double meaning in one of Leonides of Alexandria’s isopsephic epigrams ( FGE = AP .), where it is accompanied by περικαλλὲς ἄθυρμα, which closely corresponds with Callimachus’ ‘much admired plaything’.  Sbardella ,  with n. .  Magnelli , .

When Is a Riddle an Epigram?

167

in which the author of the Hypomnemata has allowed his scientific interest to intrude its way into the midst of the votive inscription’.⁷¹ I follow Prescott in graphically separating the ‘inset’ from the ‘frame’ by the spacing in the Greek text. Once this distinction between the two parts of the poem is made, we notice stylistic differences between them. The votive ‘frame’ represents the typical simplicity of Callimachus’ epigrammatic production, whereas it is the ‘inset’ that concentrates all the flamboyance, both lexical and conceptual, which appears so out of place in one of Callimachus’ epigrams. In this ‘inset’, Callimachus plays on the ambiguity of ναυτίλος by characterising the sea nautilus as a sailor. This is a riddling metaphor, and it is extended to the point of creating a real riddle: ‘I am a sailor who sails stretching out his own sail and rowing with his own feet; when I was alive, the sea halcyon laid eggs in my chambers, but now I am a plaything for the queen’. Just as in Simias’ epigram that we saw earlier, this riddle is intended for a learned reader, as the poetic characterisation of the nautilus, at once riddling and scientifically precise, is, as Prescott saw,⁷² strongly reminiscent of the description provided by Aristotle in Historia Animalium (622b5 – 15) – the analogy with Simias’ epigram on the music of the cricket is, therefore, striking. It is in this sense that the word παίγνιον at line 8 is self-referential – for there is a refined and learned poetic paignion that hides, like a halcyon’s egg in a nautilus’ shell, or like a pearl in an oyster, just behind the frame of a simple epigram. According to Fredrick, the structure of the nautilus poem is ‘perhaps a subtle, technopaegnic reflection of the whorls of the shell itself’.⁷³ Be that as it may, we find precisely this sort of ‘technopaegnic reflection’ of another shell in a similar epigram-paignion, one that was composed by Callimachus’ contemporary Theodoridas (probably the same whose fragment we saw in the scholia to the Oyster Riddle!):⁷⁴ – Εἰνάλι’ ὦ λαβύρινθε, τύ μοι λέγε, τίς σ’ ἀνέθηκεν ἀγρέμιον πολιᾶς ἐξ ἁλὸς εὑρόμενος. – Παίγνιον ἀντριάσιν Διονύσιος ἄνθετο Νύμφαις – δῶρον δ’ ἐξ ἱερᾶς εἰμι Πελωριάδος – υἱὸς Πρωτάρχου· σκολιὸς δ’ ἐξέπτυσε πορθμός, ὄφρ’ εἴην λιπαρῶν παίγνιον ᾿Aντριάδων. – Shell, labyrinth of the deep, tell me who found you, a booty won from the gray sea, and dedicated you here. – Dionysius son of Protarchus dedicated me as a plaything for the Nymphs of the grotto. I

   

Prescott ,  – , Fredrick ,  – . Prescott ,  – . Fredrick , . The translation from Paton .

168

Jan Kwapisz

am a gift from the holy Pelorian coast, and the waves of the winding channel cast me ashore to be the plaything of the sleek Nymphs of the grotto. Theodorid. 5 Gow-Page/Seelbach = AP 6.224

Once again, a shell becomes a ‘plaything’, but also a ‘sea labyrinth’ – such a description may be, we feel, loaded with metapoetic significance. Yet another key word in this epigram is σκολιός, ‘winding’ (but also ‘riddling’⁷⁵). How is this poem ‘winding’? The whorls of the shell are, I argue, reflected in the ring composition of how the shell presents itself, since παίγνιον ᾿Aντριάδων of the last line echoes παίγνιον ἀντριάσιν of line 3, and this ‘winding’ repetition is parallelled by the repetition of εἰνάλιε / ἐξ ἁλός in the opening distich.⁷⁶ As a result, this relatively plain epigram proves to be a technopaegnion, or a carefully structured labyrinth of words, as its structure almost graphically illustrates what it says. Yet Theodoridas’ labyrinth has taken us rather far from proper riddles; this poem has no question to ask, and here the ways of paignia and riddles part. My point, however, is precisely this – the career of the clearly riddling class of paignia and the related subclass of the epigram that may be described as the mixing of formal features of riddles and epigrams seems to be, unless the available evidence is misleading, neither long nor particularly fruitful in terms of literary influence. These riddling poems seem to have emerged at the dawn of the Hellenistic age, in the era of experiment which witnessed the naissance of epigrammatic subgenres such as the technopaegnion or the so-called lyric epigram. Just as the lyric epigrams in various experimental metres,⁷⁷ the paignia enjoyed a certain popularity in the third century BC but, for what we know, almost completely disappeared from later literary history. On the one hand, therefore, the influence of the riddle on the epigram should probably not be overemphasised. On the other, however, it is tempting to surmise that once the riddle and the epigram started merging their generic identities after the long coexistence of the two genres in separate spheres of orality and epigraphy, the notion of a poem as a plaything – which is, after all, inherent in all sorts of poetic riddles – played some role in the formation of the Hellenistic literary epigram.

 LSJ s.v.  In my opinion, Gow and Page (, ) miss the point of this epigram when they point out these repetitions as ‘disturbing’.  See West ,  –  and Fantuzzi/Hunter ,  n. .

When Is a Riddle an Epigram?

169

Abbreviations FGE D.L. Page, Further Greek Epigrams: Epigrams before A.D. 50 from the Greek Anthology and Other Sources, not Included in Hellenistic Epigrams or The Garland of Philip, Oxford 1981. LSJ H.G. Liddell / R. Scott / H.S. Jones, A Greek-English Lexicon9, Oxford 1940, Supplementum 1996. SEG Supplementum Epigraphicum Graecum. SH H. Lloyd-Jones / P.J. Parsons, Supplementum Hellenisticum, Berlin / New York 1983. SSH H. Lloyd-Jones, Supplementum Supplementi Hellenistici, Berlin / New York 2005.

Bibliography Acosta-Hughes, B. (2002), Polyeideia: The Iambi of Callimachus and the Archaic Iambic Tradition, Berkeley / Los Angeles / London. Bartol, K. (2013), ‘Versus anacyclici: The Case of P. Sorb. 72v (= adesp. com. fr. 52 PCG)’, in: Kwapisz/Szymański/Petrain 2013, 309 – 319. Berra, A. (2008), Théorie et pratique de l’énigme en Grèce ancienne, Diss. Paris. Besios, M. / A. Kotsonas / Y.Z. Tzifopoulos (2012), Μεθώνη Πιερίας I: Επιγραφές, χαράγματα και εμπορικά σύμβολα στη γεωμετρική και αρχαϊκή κεραμική από το ‘Υπόγειο’ της Μεθώνης Πιερίας στη Μακεδονία, Thessaloniki. Beta, S. (2013), ‘Oracles and Riddles Ambo Fratres: Cultural (and Family) Relations Between Oracula and Aenigmata’, in: J.V. García / A. Ruiz (eds.), Poetic Language and Religion in Greece and Rome, Newcastle, 199 – 206. Bevilacqua, G. / C. Ricci (2012), ‘Obscure inscribere: enigmi e indovinelli epigrafici’, in: Monda 2012, 125 – 150. Bing, P. (1984), ‘Callimachus’ Cows: A Riddling Recusatio’, in: ZPE 54, 1 – 8. — (1986), ‘The Alder and the Poet: Philetas 10 (p. 92 Powell)’, in: RhM 129, 222 – 226. — (2008), The Well-Read Muse: Present and Past in Callimachus and the Hellenistic Poets2, Ann Arbor. — (2009), The Scroll and the Marble: Studies in Reading and Reception in Hellenistic Poetry, Ann Arbor. Bowie, E. (2010), ‘Epigram as Narration’, in: M. Baumbach / A. Petrovic / I. Petrovic (eds.), Archaic and Classical Greek Epigram, Cambridge, 313 – 384. Cameron, A. (1993), The Greek Anthology from Meleager to Planudes, Oxford. Cerri, G. (2005), ‘L’ontano di Filita: soluzione di un enigma e ricostruzione di un percorso critico’, in: QUCC 80, 133 – 139. D’Alessio, G.B. (1990), ‘Aggiunte all’ “Ostrica” (Suppl. Hell. 983 v.3)’, in: ZPE 81, 299 – 303. Day, J.W. (2010), Archaic Greek Epigram and Dedication: Representation and Reperformance, Cambridge. Elmer, D.F. (2005), ‘Helen Epigrammatopoios’, in: CA 24, 1 – 39. Fantuzzi, M. / R. Hunter (2004), Tradition and Innovation in Hellenistic Poetry, Cambridge. Forster, E.S. (1945), ‘Riddles and Problems from the Greek Anthology’, in: G&R 14, 42 – 47. Fredrick, D. (1999), ‘Haptic Poetics’, in: Arethusa 32, 49 – 83. Garriga, C. (1989), ‘Filetas de Cos, Fr. 10K (= 10 Powell)’, in: Lexis 3, 79 – 87.

170

Jan Kwapisz

Gow, A.S.F. / D.L. Page (1965), The Greek Anthology: Hellenistic Epigrams II, Cambridge. Gutzwiller, K.J. (1992), ‘The Nautilus, the Halcyon, and Selenaia: Callimachus’s Epigram 5 Pf. = 14 G-P’, in: CA 11, 194 – 209. — (1998), Poetic Garlands: Hellenistic Epigrams in Context, Berkeley / Los Angeles / London. Harder, M.A. / R.F. Regtuit / G.C. Wakker (eds.) (2009), Nature and Science in Hellenistic Poetry, Leuven / Paris / Walpole, MA. Kirstein, R. (2008), ‘‚Der mitdenkende Leser‘. Überlegungen zum antiken Rätselepigramm’, in: Hermes 136, 466 – 483. Kwapisz, J. (2013a), The Greek Figure Poems, Leuven / Paris / Walpole, MA. — (2013b), ‘Were There Hellenistic Riddle Books?’, in: Kwapisz/Szymański/Petrain 2013, 148 – 167. — (2015), ‘Deciphering Ne Luscinia Segnior’, in: Prace Filologiczne. Literaturoznawstwo 5 (8), 167 – 181. Kwapisz, J. / M. Szymański / D. Petrain (eds.) (2013), The Muse at Play: Riddles and Wordplay in Greek and Latin Poetry, Berlin / Boston. Lasserre, F. (1975), ‘L’élégie de l’huître (P. Louvre inv. 7733 v° inéd.)’, in: QUCC 19, 145 – 176. Leary, T.J. (2014), Symphosius: The Aenigmata, London / New York. Lightfoot, J.L. (2009), Hellenistic Collection: Philitas, Alexander of Aetolia, Hermesianax, Euphorion, Parthenius, Cambridge, MA / London. Luz, C. (2010), Technopaignia: Formspiele in der griechischen Dichtung, Leiden / Boston. — (2013), ‘What Has It Got in Its Pocketses? Or, What Makes a Riddle a Riddle?’, in: Kwapisz/Szymański/Petrain 2013, 84 – 99. Magnelli, E. (2007), ‘Meter and Diction: From Refinement to Mannerism’, in: P. Bing / J.S. Bruss (eds.), Brill’s Companion to Hellenistic Epigram: Down to Philip, Leiden, 165 – 183. Martis, C. (2013), ‘L’enigma del PLouvre inv. 7733 verso: l’epigramma dell’ostrica’, in: SEP 10, 117 – 150. Meyer, D. (2005) Inszeniertes Lesevergnügen: Das inschriftliche Epigramm und seine Rezeption bei Kallimachos, Stuttgart. Miruka, O. (2001), Oral Literature of the Luo, Nairobi / Kampala / Dar es Salaam. Monda, S. (ed.) (2012), Ainigma e griphos: Gli antichi e l’oscurità della parola, Pisa. Ohlert, K. (1886). Rätsel und Gesellschaftsspiele der alten Griechen, Berlin. — (1912). Rätsel und Rätselspiele der alten Griechen2, Berlin. Pappas, A. (2013), ‘The Treachery of Verbal Images: Viewing the Greek Technopaegnia’, in: Kwapisz/Szymański/Petrain 2013, 199 – 224. Parsons, P.J. (1977), ‘The Oyster’, in: ZPE 24, 1 – 12. Parsons, P.J. / H. Maehler / F. Maltomini (2015), The Vienna Epigrams Papyrus (G 40611), Berlin / Munich / Boston. Paton, W.R. (1918), The Greek Anthology V, London / New York. — (1919), The Greek Anthology II, London / New York. Perale, M. (2011), ‘Il catalogo “geografico” di Esiodo: due diversi casi di ricezione nella prima età ellenistica’, in: A. Aloni / M. Ornaghi (eds.), Tra panellenismo e tradizioni locali: Nuovi contributi, Messina, 365 – 389. Pittore, M. (2004), L’ironia negli epigrammi dell’Anthologia Palatina tra manipolazione linguistica e allusività, Alessandria. Potamiti, A. (2015), ‘Γρίφους παίζειν: Playing at Riddles in Greek’, in: GRBS 55, 133 – 153.

When Is a Riddle an Epigram?

171

Prescott, H.W. (1921), ‘Callimachus’ Epigram on the Nautilus’, in: CPh 16, 327 – 337. Reitzenstein, R. (1893), Epigramm und Skolion: Ein Beitrag zur Geschichte der alexandrinischen Dichtung, Giessen. Rotroff, S. (1996), The Missing Krater and the Hellenistic Symposium: Drinking in the Age of Alexander the Great, Broadhead Classical Lecture No 7, University of Canterbury, Christchurch, August 7, 1996. Sbardella, L. (2000), Filita: Testimonianze e frammenti poetici, Roma. Schroeder, C.M. (2012), ‘Homeric Formularity in Callimachus’ Nautilus Epigram (5 Pf. = 14 G-P)’, in: Mnemosyne 65, 295 – 299. Schultz, W. (1912), Rätsel aus dem hellenischen Kulturkreise, Leipzig. Selden, D.L. (1998), ‘Alibis’, in: CA 17, 289 – 412. Sier, K. (2001), ‘Philitas’ Erle oder: Die Kunst des Zitierens (fr. 10, p. 92 Powell)’, in: Philologus 145, 70 – 78. Sila Basak (2006), Bengali Culture and Society through Its Riddles, New Delhi. Sistakou, E. (2007), ‘Glossing Homer: Homeric Exegesis in Early Third Century Epigram’, in: P. Bing / J.S. Bruss (eds.), Brill’s Companion to Hellenistic Epigram: Down to Philip, Leiden, 391 – 408. Spanoudakis, K. (2002), Philitas of Cos, Leiden. Tueller, M.A. (2008), Look Who’s Talking: Innovations in Voice and Identity in Hellenistic Epigram, Leuven / Paris / Dudley, MA. Wachter, R. (2010), ‘The Origin of Epigrams on “Speaking Objects”’, in: M. Baumbach / A. Petrovic / I. Petrovic (eds.), Archaic and Classical Greek Epigram, Cambridge, 250 – 260. Wecowski, M. (2014), The Rise of the Greek Aristocratic Banquet, Oxford. — (forthcoming), ‘Wine and the Early History of the Greek Alphabet: Early Greek Vase-inscriptions and the Symposion’, in: J. Strauss Clay / I. Malkin / Y. Tzifopoulos (eds.), Proceedings of the Conference ‘Panhellenes at Methone: graphê in Late Geometric and Protoarchaic Methone, Macedonia (ca 700 BCE)’, Thessaloniki, June 8 – 10, 2012, Berlin / Boston. West, M. L. (1982), Greek Metre, Oxford. — (1997), ‘When Is a Harp a Panpipe?’, in: CQ 47, 48 – 55. — (2007), Indo-European Poetry and Myth, Oxford. Wilamowitz-Moellendorff, U. v. (1924), Hellenistische Dichtung in der Zeit des Kallimachos II, Berlin.

Giulio Massimilla

The ἀπὸ κοινοῦ Construction of Prepositions as a Feature of the Epigrammatic Style I will investigate the use and the stylistic relevance of a specific syntactic pattern in Greek epigrams. This pattern is usually known as ‘ἀπὸ κοινοῦ construction of prepositions’ and is to be found when, in a sequence of two or more complements introduced by the same preposition, this preposition is implicit in the first complement(s) and becomes explicit only in the following one(s).¹ In the two Homeric poems there seems to be one certain example of the pattern, occurring in the late Doloneia (Iliad 10). After killing Dolon and Rhesus with many Thracians, Diomedes and Odysseus go back to the Greek camp and wash off the sweat (Il. 10.572 f.): αὐτοὶ δ᾽ ἱδρῶ πολλὸν ἀπονίζοντο θαλάσσῃ / ἐσβάντες κνήμας τε ἰδὲ λόφον ἀμφί τε μηρούς (“as for themselves, they washed off the abundant sweat from their shins and the back of their necks and thighs by going into the sea”, transl. B.P. Powell). Here the preposition ἀμφί governs not only μηρούς, but also κνήμας and λόφον. It is doubtful that this σχῆμα ἀπὸ κοινοῦ also appears in a formulaic passage occurring several times in the Odyssey (4.475 f. and elsewhere): ἱκέσθαι (or ἀφίκοιο) / οἶκον ἐϋκτίμενον καὶ σὴν ἐς πατρίδα γαῖαν (“reach your well-ordered household and your native land”, transl. A.T. Murray and G.E. Dimock). Here οἶκον ἐϋκτίμενον could be an accusative governed by the following ἐς together with πατρίδα γαῖαν (according to the σχῆμα ἀπὸ κοινοῦ), but also a simple accusative of motion. Nor can we be sure that the pattern is present in a passage of the Odyssey (12.26 f.) where Circe says that she will tell Odysseus and his men how to have a safe journey: ἵνα μή τι κακορραφίῃ ἀλεγεινῇ / ἢ ἁλὸς ἢ ἐπὶ γῆς ἀλγήσετε πῆμα παθόντες (“in order that through wretched ill-contriving you may not suffer pain and woes either by sea or on land”, transl. A.T. Murray

 Kiefner  has provided the most thorough study of the ἀπὸ κοινοῦ construction (which he calls “Versparung”) as applying not only to prepositions, but also to other parts of the sentence. He rightly concludes that this σχῆμα ἀπὸ κοινοῦ is a stylistic device (pp.  – , “Die Versparung als rhetorisch-dichterische Figur”). Kiefner’s main focus is on Greek tragedy, but he also touches on a wider variety of literary genres, both Greek and Latin, and assembles a useful collection of examples (pp.  – ). For earlier discussions of the ἀπὸ κοινοῦ construction of prepositions, see Lobeck ,  f.; Wilamowitz ,  f.; Kühner/Gerth , .

174

Giulio Massimilla

and G.E. Dimock). Here it is not clear whether ἁλός is a genitive governed by the following ἐπί together with γῆς (in keeping with the σχῆμα ἀπὸ κοινοῦ) or a simple locative genitive. However, it is possible that this Homeric passage had some influence on the later authors adopting the ἀπὸ κοινοῦ construction of prepositions, since, as we will see, they often used this pattern with reference to the land on the one hand, and the sea on the other.² In later times the ἀπὸ κοινοῦ construction of prepositions came to be felt as a mark of high poetic style. Lyric poets, especially Pindar,³ use it quite often. See, for example, Pindar, Nemean 9.21 f. (about the Seven against Thebes): φαινομέναν δ’ ἄρ’ ἐς ἄταν σπεῦδεν ὅμιλος ἱκέσθαι / χαλκέοις ὅπλοισιν ἱππείοις τε σὺν ἔντεσιν (“and the band hastened to go to its manifest ruin with bronze armour and horse-gear”, transl. B.K. Braswell). It is also found in tragic dialogue and song.⁴ See, for instance, Sophocles, Antigone 1176 (the Coryphaeus is asking the Messenger about Haemon’s death): πότερα πατρῴας ἢ πρὸς οἰκείας χερός; (“Was it by his father’s hand or by his own?”, transl. H. Lloyd-Jones). It occurs several times in Hellenistic and late antique poetry.⁵ I will now focus on the use of this construction in Greek epigrams. I will first consider a homogeneous group of votive epigrams dealing with the dedications to Pan by three brothers. Then I will take into account some more occurrences of the pattern in other epigrams.

 Cf. already Timocreon of Rhodes (first half of the th century BC), fr. . –  Page (beginning of a scolion): Ὤφελέν σ᾽ ὦ τυφλὲ Πλοῦτε / μήτε γῇ μήτ᾽ ἐν θαλάσσῃ / μήτ᾽ ἐν ἠπείρῳ φανῆμεν, “Blind Wealth, if only you had appeared neither on land nor on sea nor on the mainland”, transl. D.A. Campbell (μήτ᾽ ἐν γῇ codd. Schol. Ar. Ach. : corr. Brunck; see Martelli ,  n. ). Timocreon’s passage was parodied by Aristophanes, Acharnians  f. ὡς χρὴ Μεγαρέας μήτε γῇ μήτ᾽ ἐν ἀγορᾷ / μήτ᾽ ἐν θαλάττῃ μήτ᾽ ἐν ἠπείρῳ μένειν (μήτ᾽ ἐν γῇ codd.: corr. Bentley; cf. also Ar. Eq. ): see Dunbar ,  f.; Olson ,  f.  See Bossler ,  – ; Kiefner ,  f.  See Barrett , ; Kiefner ,  – .  See Schneider , ; Kaibel ,  f.; Lapp , ; Reed , ; Massimilla , .

The ἀπὸ κοινοῦ Construction of Prepositions

175

1. The ἀπὸ κοινοῦ construction of prepositions in the epigrams on the three brothers Several poems in book 6 of the Palatine Anthology (11– 16 and 179 – 187) deal with the same subject: three brothers (a hunter, a fowler, and a fisherman) dedicate their respective tools to Pan.⁶ The subject must have been very popular, since the earliest epigram in the series is by Leonidas of Tarentum (4th-3rd centuries BC), and the latest by Julian the Egyptian (6th century AD). Nine different poets figure within the series. Here is a list of their names in a very approximate chronological order: 1. 2. 3. 4. 5. 6. 7. 8. 9.

Leonidas of Tarentum (AP 6.13) Antipater of Sidon (AP 6.14) Archias (AP 6.16, 179, 180, 181) Zosimus of Thasus (AP 6.15, 183, 184, 185) Satyrius (AP 6.11) Alexander of Magnesia (AP 6.182) Julius Diocles (AP 6.186) Alpheus of Mytilene (AP 6.187) Julian the Egyptian (AP 6.12)

As a matter of fact, we can only say that Leonidas of Tarentum and Antipater of Sidon (2nd century BC) are the most ancient poets in the group, and that Julian is the latest. We also have good reasons for supposing that the poet Archias who contributed to the three-brothers series is the same Aulus Licinius Archias of Cicero’s speech, whose approximate dates are 120 to 45 BC. By contrast, the identities of the remaining five poets (Zosimus of Thasus, Satyrius,⁷ Alexander of Magnesia, Julius Diocles, and Alpheus of Mytilene) are often elusive, and their dates (both absolute and relative) are extremely uncertain: nevertheless, it seems safe to say that these five were active between the 1st century BC and the 1st century AD.⁸ As far as we know, Leonidas’ epigram was the model poem of the series. Its popularity can also be inferred from its inscription beneath a fresco of three

 On this series of epigrams, see Longo  – ; Laurens ,  –  (= ,  – ); Gutzwiller ,  – ; Penzel ,  – .  For the identity of this poet and the spelling of his name, see Page , .  For the dates of these poets, see Page ,  and  (Zosimus of Thasus),  (Satyrius),  (Alexander of Magnesia), Gow/Page  (),  f. (Julius Diocles),  n.  (Alpheus of Mytilene).

176

Giulio Massimilla

huntsmen in the House of the Epigrams at Pompeii.⁹ Let us read it (AP 6.13 = HE 2249 ff.): Οἱ τρισσοί τοι ταῦτα τὰ δίκτυα θῆκαν ὅμαιμοι, ἀγρότα Πάν, ἄλλης ἄλλος ἀπ’ ἀγρεσίης, ὧν ἀπὸ μὲν πτανῶν Πίγρης τάδε, ταῦτα δὲ Δᾶμις τετραπόδων, Κλείτωρ δ’ ὁ τρίτος εἰναλίων. ἀνθ’ ὧν τῷ μὲν πέμπε δι’ ἠέρος εὔστοχον ἄγρην, τῷ δὲ διὰ δρυμῶν, τῷ δὲ δι’ ἠιόνων. Three brothers have dedicated to you, huntsman Pan, these nets, each from a different chase: Pigres these from fowl, Damis these from beasts, and Cleitor the third from creatures of the sea. In exchange give successful hunting to one in the air, to the other in the woods, and to the third on the shore (transl. K.J. Gutzwiller).¹⁰

The six lines’ length of this epigram became the standard length of the later variations with only two exceptions, i. e. one of the epigrams of Zosimus of Thasus (AP 6.15) and the epigram of Julian the Egyptian (AP 6.12), who both compressed their poems into four lines. The structure of Leonidas’ epigram is very neat. In the first couplet there are the address to Pan and the collective mention of the brothers and of their dedications. The second couplet has the names of the three brothers and the specific indication of their respective hunting trades. The third couplet contains the prayer to Pan to grant successful hunting to each brother in his own field. The later poets sometimes kept close to this structure, but more often varied it in multiple ways. Leonidas’ poem contains all the elements that, with some minor deviations, became standard in the later epigrams dealing with the same subject. These elements are conveniently summarized by Denys Page: “The rules of the game were: (a) that the dedication should be to Pan; (b) that the names of the dedicators should be Pigres, Damis, and Cleitor (only Diocles omits the names); (c) that Pigres should be the fowler, Damis the hunter, and Cleitor the fisherman; (d) that they should be brothers (only Zosimus in 6.184 and 185 omits the motif); (e) that they should dedicate the tools of their trades (only Alpheus dedicates the produce instead of the tools); and (f) that the epigram should end with a prayer for success ‘on land, in the air, and in the sea’ (only Antipater ends with a thank-offering for the past instead of a pray-

 EpGr.  Kaibel. See Gigante ,  – ; Longo  – ,  f.; Gutzwiller , .  Gigante ,  –  (= ,  f.) traces the triple tripartition displayed in this epigram (with reference to the huntsmen, the nets, and the prey) back to Sophocles, Antigone  – .

The ἀπὸ κοινοῦ Construction of Prepositions

177

er for the future; only Satyrius and Zosimus 6.185 have ‘birds, beasts, and fish’ instead of ‘air, land, and sea’)”.¹¹ I will now focus on the poems of this series, where the ἀπὸ κοινοῦ construction of prepositions is used. I will argue that the epigrammatists who wrote on the three brothers treated the ἀπὸ κοινοῦ construction of prepositions as a stylistic feature inherent to their common subject, and vied with each other (and sometimes with themselves) in adopting and varying this syntactic pattern. The construction is already to be found in the epigram of Antipater of Sidon, whose approximate dates are 180 to 100 BC.¹² Antipater was famous for his variations of earlier epigrams, most particularly of Leonidas as is the case here.¹³ Let us read his poem (AP 6.14 = HE 168 ff.): Πανὶ τάδ’ αὔθαιμοι τρισσοὶ θέσαν ἄρμενα τέχνας· Δᾶμις μὲν θηρῶν ἄρκυν ὀρειονόμων, Κλείτωρ δὲ πλωτῶν τάδε δίκτυα, τῶν δὲ πετηνῶν ἄρρηκτον Πίγρης τάνδε δεραιοπέδην. τὸν μὲν γὰρ ξυλόχων, τὸν δ’ ἠέρος, ὃν δ’ ἀπὸ λίμνης οὔποτε σὺν κενεοῖς οἶκος ἔδεκτο λίνοις. To Pan three brothers have dedicated this equipment: Damis a purse net for beasts that roam the mountains, Cleitor these nets for fish, and Pigres this unbreakable neck chain for the feathered birds of the air. For never did they return home with empty nets, one from the woods, one from the air, one from the sea (transl. K.J. Gutzwiller).

Antipater’s reworking of the Leonidean model is very innovative in many ways: for example, Pan is not addressed in the second person, the unspecific word δίκτυα is replaced with three different words describing three specific kinds of snares, and the brothers are mentioned in a different order. Most notably, Antipater’s last couplet does not contain a prayer for the future but a thank-offering for the past.¹⁴ Furthermore, here Antipater changes the Leonidean preposition διά into ἀπό (taking it from Leonidas’ second couplet) and makes use of the σχῆμα ἀπὸ κοινοῦ, since ἀπό governs ξυλόχων, ἠέρος and λίμνης but becomes explicit only before the last noun.¹⁵ Antipater’s innovative use of the σχῆμα ἀπὸ κοινοῦ in his epigram on the three brothers was not lost on Archias, who adopted the ἀπὸ κοινοῦ construction

 Page , .  See Argentieri ,  – .  See Gutzwiller ,  f. and  – .  See Gow/Page  (), ; Longo  – , .  See Gutzwiller , : “Antipater makes the reader work harder by postponing the single appearance of the preposition ἀπό to the third colon”.

178

Giulio Massimilla

of a preposition in one of his four variations of that same subject (AP 6.180). Although as many as five poets named Archias may have contributed to the Greek Anthology, it is very tempting to identify the author of the four epigrams on the three brothers with Aulus Licinius Archias of Antioch, best known to us through Cicero’s famous speech.¹⁶ Licinius Archias may have been personally acquainted with his elder contemporary Antipater of Sidon in Rome,¹⁷ and like Antipater he was admired for his skill in extemporaneous composition, which often involved variation and (as in this case) self-variation. Let us read the epigram of Archias which contains the σχῆμα ἀπὸ κοινοῦ in its last couplet (AP 6.180 = GP 3608 ff.): Ταῦτά σοι ἔκ τ’ ὀρέων ἔκ τ’ αἰθέρος ἔκ τε θαλάσσας τρεῖς γνωτοὶ τέχνας σύμβολα, Πάν, ἔθεσαν· ταῦτα μὲν εἰναλίων Κλείτωρ λίνα, κεῖνα δὲ Πίγρης οἰωνῶν, Δᾶμις τὰ τρίτα τετραπόδων. οἷς ἅμα χερσαίῃσιν, ἅμ’ ἠερίῃσιν ἐν ἄγραις, ᾿Aγρεῦ, ἅμ’ ἐν πλωταῖς, ὡς πρίν, ἀρωγὸς ἴθι. Three brothers, Pan, gave you these symbols of their craft from the mountains, air, and sea: Cleitor these nets for fish, Pigres those nets for birds, and Damis a third set for beasts. Be their helper as before, hunter god, in pursuits on land, air, and water (transl. K.J. Gutzwiller).

Unlike Antipater, Archias imitates Leonidas’ second-person address to Pan, and this pattern will persist in all the other epigrams on the three brothers. The most remarkable aspect of Archias’ epigram is that the basic tripartition into land, air, and sea is not confined (as in Leonidas and Antipater) to the last two couplets, but involves also the first. Furthermore, in the last couplet Archias manages to allude to both Leonidas and Antipater, because he goes back to the prayer of Leonidas, but also points to the thank-offering of Antipater through the parenthetical phrase ὡς πρίν (“as before”). As I said, in the last couplet Archias takes the ἀπὸ κοινοῦ construction of a preposition from Antipater, but in so doing he displays further variations. Firstly,

 See Gow/Page  (),  – , esp.  n. ; Gutzwiller ,  n. . The same applies to another poem ascribed to Archias in the Greek Anthology (AP . = GP  ff.), which is closely related to two epigrams by Leonidas of Tarentum (AP . = HE  ff.) and Antipater of Sidon (AP . = HE  ff.). This group of poems deals with the dedication of spinning and weaving instruments to Athena by three women (three sisters, according to Leonidas and Archias) and provides “a sort of feminine counterpart to the dedication of the three hunters” (Gow/Page  (), ): see Magnelli , ; Argentieri , ; Penzel ,  – .  See Gutzwiller , . Penzel  treats Antipater of Sidon and Archias of Antioch together.

The ἀπὸ κοινοῦ Construction of Prepositions

179

he changes Antipater’s ἀπό into ἐν (as Antipater himself had changed the Leonidean διά into ἀπό). Secondly, and more importantly, he adopts a different kind of σχῆμα ἀπὸ κοινοῦ because here the preposition ἐν is explicit not only in the third, but already in the second complement (ἠερίῃσιν ἐν ἄγραις and ἐν πλωταῖς). We may also note that the σχῆμα ἀπὸ κοινοῦ at the end of the poem is meant to create a sharp contrast with the triple anaphora of the preposition ἐκ in line 1. In the wake of Antipater of Sidon, the epigrammatists dealing with the threebrothers theme came to treat the ἀπὸ κοινοῦ construction of prepositions as an important ingredient of their variations. In this regard, Zosimus of Thasus takes the lion’s share. Nothing is known about this poet and his date. According to Page, the presence of hiatus in the middle of one of his hexameters (AP 9.40.3) suggests that he composed his epigrams between the 1st century BC and the first half of the 1st century AD.¹⁸ The Greek Anthology preserves five poems under his name. Of these poems, as many as four are about the three brothers. Of these, as many as three contain the ἀπὸ κοινοῦ construction of a preposition. Zosimus’ predilection for the three-brothers theme suggests that the first of these three epigrams, which I am now going to discuss, is by him, according to Planudes’ ascription Ζωσίμου, and not by “either Antipater of Sidon or Zosimus”, according to the ascription in the Palatine Anthology.¹⁹ Let us read the poem (AP 6.15 = FGE 404 ff.): Εἰναλίων Κλείτωρ τάδε δίκτυα, τετραπόδων δέ Δᾶμις καὶ Πίγρης θῆκεν ὑπηερίων Πανί, κασιγνήτων ἱερὴ τριάς. ἀλλὰ σὺ θήρην ἠέρι κἠν πόντῳ κἠν χθονὶ τοῖσδε νέμε. The blessed triad of brothers dedicated these nets to Pan: Cleitor his fishing nets, Damis his hunting nets, Pigres his fowling nets. But you grant them good hunting in air, sea, and land (transl. W.R. Paton, adapted).

The four lines’ length is the most notable contribution of this epigram to the game of variations.²⁰ Zosimus manages to squeeze the contents of, respectively, the first two couplets and the third couplet of the Leonidean model into his own first couplet and line 3 until the bucolic diaeresis on the one hand, and into the

 See Page ,  and .  Planudes’ ascription is also favoured by Gow/Page  (),  f.; Page , ; Argentieri ,  f.  See Longo  – , .

180

Giulio Massimilla

final part of his epigram, starting from the end of line 3, on the other hand. In line 4 Zosimus adopts the ἀπὸ κοινοῦ construction of the preposition ἐν like Archias, and uses it similarly, placing ἐν before the last two nouns (πόντῳ and χθoνί). The two poems of Zosimus, which I am now going to discuss, conform to the standard six lines’ length but are quite experimental in their use of the σχῆμα ἀπὸ κοινοῦ. Here is the first of them (AP 6.183 = FGE 408 ff.): Σοὶ τάδε, Πάν, θηρευταὶ ἀνηρτήσαντο σύναιμοι δίκτυα, τριχθαδίης δῶρα κυναγεσίης, Πίγρης μὲν πτανῶν, Κλείτωρ δ᾽ ἁλός,²¹ ὃς δ’ ἀπὸ χέρσου Δᾶμις τετραπόδων ἀγκύλος ἰχνελάτης. ἀλλὰ σὺ κἠν δρυμοῖσι καὶ εἰν ἁλὶ καὶ διὰ μέσσης ἠέρος εὔαγρον τοῖσδε δίδου κάματον. The hunter brothers suspended these nets to you, Pan, gifts from three sorts of chase: Pigres his nets for birds, Cleitor his from the sea, and Damis, the crafty tracker of beasts, his from the land. But you reward their toil with success in wood, sea, and air (transl. W.R. Paton, adapted).

Here Zosimus moves the ἀπὸ κοινοῦ construction of a preposition from its usual place, i. e. the last couplet (containing the prayer), backwards to the second (containing the names of the three brothers and the specific indication of their respective hunting habitats). Furthermore, Zosimus uses the σχῆμα ἀπὸ κοινοῦ in a new and slightly misleading way, since in line 3 πτανῶν is probably a simple genitive (meaning nets “for birds”), while ἁλός and χέρσου are governed by ἀπό (meaning nets “from the sea and the land”). We may note that, by contrast, in the final prayer the three hunting habitats, besides reversing the order of line 3, are each provided with its own preposition: first ἐν, second its prosodic alternative εἰν, and third, with a more radical change, διά. A similar but not identical structure is found in the other poem of Zosimus that I am now going to discuss (AP 6.184 = FGE 414 ff.): Τρισσὰ τάδε τρισσοὶ θηραγρέται, ἄλλος ἀπ’ ἄλλης τέχνης, πρὸς νηῷ Πανὸς ἔθεντο λίνα, Πίγρης μὲν πτανοῖσιν ἐφεὶς βόλον, ἐν δ’ ἁλίοισι Κλείτωρ, ἐν θηρσὶν Δᾶμις ἐρημονόμοις. τοὔνεκα, Πάν, τὸν μέν τε δι’ αἰθέρος, ὃν δ’ ἀπὸ λόχμης, τὸν δὲ δι’ αἰγιαλῶν θὲς πολυαγρότερον.

 I insert δ᾽ between Κλείτωρ and ἁλός, as suggested by Stadtmüller , .

The ἀπὸ κοινοῦ Construction of Prepositions

181

The three huntsmen, each from a different craft, dedicated these three nets in Pan’s temple, Pigres who set his nets for birds, Cleitor who set his for sea fish, and Damis who set his for the beasts haunting the wilds. Therefore, Pan, make them more successful, the one in the air, the other in the thicket, and the third on the beach (transl. W.R. Paton, adapted).

Here too, Zosimus moves the ἀπὸ κοινοῦ construction of a preposition backwards to the second couplet, this time applying it to all three hunting trades and placing ἐν before the last two members of the sequence (ἁλίοισι and θηρσὶν … ἐρημονόμοις). Here too, in the final prayer the three hunting habitats are listed in a different order from lines 3 f., and are each provided with its own preposition: first διά, second ἀπό, and third διά again. The ἀπὸ κοινοῦ construction of a preposition is also to be found in an epigram on the three brothers, that both the Palatine Anthology and Planudes ascribe to a poet named Alexander. The Palatine Anthology ascribes it more specifically to “Alexander of Magnesia”. It is the only epigram preserved under this name in the Greek Anthology. We know nothing about the poet and his date.²² According to Page, “style and subject point to the period covered by Philip’s Garland (90 BC–AD 40), the later rather than the earlier half”.²³ Let us read the poem (AP 6.182 = FGE 7 ff.): Πίγρης ὀρνίθων ἄπο δίκτυα, Δᾶμις ὀρείων, Κλείτωρ δ’ ἐκ βυθίων σοὶ τάδε, Πάν, ἔθεσαν, ξυνὸν ἀδελφειοὶ θήρης γέρας, ἄλλος ἀπ’ ἄλλης, ἴδρι τὰ καὶ γαίης, ἴδρι τὰ καὶ πελάγευς. ἀνθ’ ὧν τῷ μὲν ἁλός, τῷ δ’ ἠέρος, ᾧ δ’ ἀπὸ δρυμῶν πέμπε κράτος ταύτῃ, δαῖμον, ἐπ’ εὐσεβίῃ. Pigres dedicated to you, Pan, his nets for birds, Damis his for mountain beasts, and Cleitor his for those of the deep: a common gift from the brothers for their luck in the various kinds of chase to you who are skilled in the things of land and sea alike. In return for which, and recognising their piety, o god, give one dominion in the sea, the other in the air, the third in the woods (transl. W.R. Paton, adapted).

In the first two couplets Alexander blends the contents which were separate from each other in the Leonidean model, i. e. the address to Pan and the collective mention of the brothers and of their dedications on the one hand (Leonidas’ lines 1 f.), and the names of the three brothers and the specific indication of  Some scholars have suspected that the ascription in the Palatine manuscript might be either mistaken or incomplete, and it has been thought that Alexander Aetolus (th-rd centuries BC) might be the author of the epigram (= dubious fragment  Magnelli). Magnelli ,  –  is righly sceptical about this conjecture.  Page , . Magnelli ,  f. prefers the st century BC.

182

Giulio Massimilla

their respective hunting trades on the other hand (Leonidas’ lines 3 f.). We may also note the refined use of the prepositions in Alexander’s first couplet: first ἀπό, second ἀπό understood, third ἐκ. In the wake of Antipater of Sidon, Alexander reserves the ἀπὸ κοινοῦ construction of a preposition for the last couplet. Although his last couplet contains a prayer for the future, not a thank-offering for the past as in Antipater’s poem, in point of style Alexander keeps very close to Antipater, since he too makes use of an ἀπὸ κοινοῦ construction of the preposition ἀπό, that becomes explicit only before the last noun (ἀπὸ δρυμῶν).²⁴ Alexander’s line 5 is strongly reminiscent of Antipater’s corresponding line also for the forms of the three pronouns (τῷ, τῷ, ᾧ) and the metrical positions of the three members of the sequence. To be sure, Alexander takes care to reverse the order chosen by Antipater: instead of “woods, air, and sea”, he has “sea, air, and woods”. The last poem on the three brothers that I am going to discuss is one of the twelve epigrams ascribed to Alpheus of Mytilene in the Greek Anthology.²⁵ We have no other information about this poet. Gow and Page “are very willing to concede that Alpheus is probably a post-Augustan author”.²⁶ Here is the epigram (AP 6.187 = GP 3536 ff.): Πανὶ κασιγνήτων ἱερὴ τριάς, ἄλλος ἀπ’ ἄλλης, ἄνθετ’ ἀπ’ οἰκείης σύμβολον ἐργασίης, Πίγρης ὀρνίθων, ἁλίων ἀπομοίρια Κλείτωρ, ἔμπαλιν ἰθυτόνων Δᾶμις ἀπὸ σταλίκων. ἀνθ’ ὧν εὐαγρίην τῷ μὲν χθονός, ᾧ δὲ διδοίης ἐξ ἁλός, ᾧ δὲ νέμοις ἠέρος ὠφελίην. To Pan a pious trinity of brothers dedicated a token from his proper trade, each from his own: Pigres of birds, Cleitor a portion of fish, Damis again from his upright hunting-stakes. Give in return good hunting, to the one from the land, to the other from the sea, to the other grant help from the sky (transl. D.L. Page).

The macrostructure of the poem keeps quite close to the Leonidean model, but there are some interesting variations in point of detail. First of all, the brothers do not dedicate their nets but their produce.²⁷ Besides, in both the second and third couplets Alpheus separates the first two members of the sequence from the third. In lines 3 f. the simple genitives ὀρνίθων and ἁλίων, accompanying  See Magnelli , .  The ascription of this specific epigram to “Alcaeus”, inserted by the corrector of the Palatine manuscript in place of “Alpheus”, may be disregarded: see Gow/Page  (), .  Gow/Page  (),  n. .  See Gow/Page  (), .

The ἀπὸ κοινοῦ Construction of Prepositions

183

ἀπομοίρια, are replaced by the prepositional phrase ἰθυτόνων … ἀπὸ σταλίκων. In lines 5 f. the three members are rearranged in chiastic order, and the first two share the hyperbatic clause εὐαγρίην … διδοίης, that in the third is replaced (again chiastically) by νέμοις … ὠφελίην. Like Antipater and Archias, Alpheus reserves the ἀπὸ κοινοῦ construction of a preposition for the last couplet, but uses it in a still different way, since ἐξ is implicit in the first complement (χθονός), explicit in the second (ἐξ ἁλός), and again implicit in the third (ἠέρος).²⁸ To sum up, the passages I have discussed so far suggest that the epigrammatists who wrote on the three brothers regarded the ἀπὸ κοινοῦ construction of prepositions as a stylistic feature inherent to their common subject and as a testing ground to prove their skill in variation.

2. The ἀπὸ κοινοῦ construction of prepositions in other epigrams I will now discuss some occurrences of the ἀπὸ κοινοῦ construction of prepositions in epigrams that do not belong to the three-brothers series. I start from passages by poets I have already mentioned. The first epigram I have selected is again by Antipater of Sidon.²⁹ Here too, we have a triple dedication, this time to Apollo: a woman called Phila dedicates a lyre, and two men called Sosis and Polycrates dedicate a bow and nets.³⁰ Let us read the poem (AP 6.118 = HE 496 ff.): Ἁ φόρμιγξ τά τε τόξα καὶ ἀγκύλα δίκτυα Φοίβῳ Σώσιδος ἔκ τε Φίλας ἔκ τε Πολυκράτεος. χὠ μὲν ὀιστευτὴρ κεραὸν βιόν, ἁ δὲ λυρῳδός τὰν χέλυν, ὡγρευτὴς ὤπασε πλεκτὰ λίνα. ἀλλ’ ὁ μὲν ὠκυβόλων ἰῶν κράτος, ἁ δὲ φέροιτο ἄκρα λύρας, ὁ δ’ ἔχοι πρῶτα κυναγεσίας. The lyre, the bow, and the intricate nets are dedicated to Phoebus by Sosis, Phila and Polycrates. The archer dedicated his horn bow, the singer her lyre, the hunter his twisted nets.

 Gow/Page  (),  think that ἠέρος could also be a simple genitive attached to ὠφελίην (“the air’s assistance”). I find this unlikely.  The epigram is ascribed to “Antipater” (without ethnic) in the Greek Anthology, but we can be sure that its author is Antipater of Sidon (not Antipater of Thessalonica) because it occurs in a Meleagrian context in the Palatine Anthology: see Gow/Page  (), ; Argentieri , .  As Gow/Page  (),  remark, it is not clear which of the two men dedicates which object. On this epigram, see Argentieri , ; Penzel ,  – ; Massimilla ,  n. .

184

Giulio Massimilla

Let the one be unsurpassed in archery, let her be supreme in playing the lyre, and let the other be first among huntsmen (transl. W.R. Paton, adapted).

The structure of this epigram is very similar to that of the epigrams on the three brothers, including the one of Antipater himself (AP 6.14). Here too, Antipater uses the ἀπὸ κοινοῦ construction of a preposition, but he moves it backwards to the first couplet in connection with the names of the three dedicators, omitting the preposition ἐκ before the first name (Σώσιδος) and placing it before both the second and the third (ἔκ τε Φίλας ἔκ τε Πολυκράτεος). It seems that Antipater of Sidon had a predilection for the ἀπὸ κοινοῦ construction of prepositions³¹ since he used it also at the beginning of another of his dedicatory epigrams (AP 6.111.1 f. = HE 476 f.):³² Τὰν ἔλαφον Λάδωνα καὶ ἀμφ’ Ἐρυμάνθιον ὕδωρ νῶτά τε θηρονόμου φερβομέναν Φολόας … The hind that used to graze about the Ladon and the waters of Erymanthus and the heights of Pholoe, home of wild beasts … (transl. W.R. Paton, adapted).³³

Here the preposition ἀμφί, like ἐξ in Alpheus’ poem on the three brothers (AP 6.187.5 f.), is implicit in the first complement (Λάδωνα), explicit in the second (ἀμφ’ Ἐρυμάνθιον ὕδωρ), and again implicit in the third (νῶτα … θηρονόμου … Φολόας). Leonidas of Tarentum himself may have used the ἀπὸ κοινοῦ construction of a preposition, if he is the author of an epigram (AP 7.449) that was wrongly attached to the preceding one (bearing the ascription Λεωνίδα Ταραντίνου) in the Palatine Anthology.³⁴ The poem containing the σχῆμα ἀπὸ κοινοῦ is in fact

 See Kaibel ,  – ; Waltz , ; Penzel , .  Also this epigram is ascribed to “Antipater” (without ethnic) in the Greek Anthology, but its Meleagrian context in the Palatine Anthology guarantees that its author is Antipater of Sidon (as is also suggested by the mention in it of historical figures dating from the th-nd centuries BC): see Gow/Page  (),  – ; Argentieri ,  f. and . This poem is closely related to Leonidas of Tarentum, AP . = HE  ff.: see Argentieri ,  f.  As regards the place-names, see Gow/Page  (), : “Pholoe is a range east of Olympia and on the border between Elis and Arcadia; Erymanthus a river which flows south from the mountain of the same name and follows the line of Pholoe, presently falling into the Alpheus; Ladon might be the Arcadian tributary of the Alpheus somewhat to the east of Erymanthus, but is more probably the less famous Ladon which flows westward from Pholoe and falls into the Peneus”.  Neither epigram appears in the Planudean Anthology. It is worth mentioning that Stadtmüller , , in pointing out the similarity between AP . and Antipater of Sidon, AP .. f.

The ἀπὸ κοινοῦ Construction of Prepositions

185

a variation (or, as it seems, a self-variation) of the preceding. It celebrates the success achieved in many fields by a Cretan named Pratalidas, and although it is included in book 7 of the Palatine Anthology, it does not even pretend to be an epitaph like the preceding poem. Let us read the epigram (AP 7.449 = HE 2028 ff.): Πραταλίδᾳ παιδεῖον Ἔρως πόθον, Ἄρτεμις ἄγραν, Μοῦσα χορούς, Ἄρης ἐγγυάλιξε μάχαν. πῶς οὐκ εὐαίων ὁ Λυκάστιος, ὃς καὶ ἔρωτι ἆρχε καὶ ἐν μολπᾷ καὶ δορὶ καὶ στάλικι; Eros gave to Pratalidas success in the pursuit of his boy loves, Artemis in the chase, the Muse in the dance and Ares in war. Must we not call him blessed, the Lycastian³⁵ supreme in love and song, with the spear and the hunting-stake? (transl. W.R. Paton, adapted).

Here the four gods mentioned in the first couplet preside over the four kinds of success listed (in a different order) in the second. In lines 3 f. the first two nouns, ἔρωτι and μολπᾷ, are both governed by the preposition ἐν, that becomes explicit only before the second. By contrast, the last two nouns (δορί and στάλικι) are best taken as simple instrumental datives.³⁶ Also Zosimus of Thasus comes into play again. As I said, the Greek Anthology preserves five poems under Zosimus’ name, four of which are about the three brothers. We have seen that, of these four, as many as three contain the ἀπὸ κοινοῦ construction of a preposition. Zosimus must have been extremely fond of this σχῆμα ἀπὸ κοινοῦ since he uses it also in his fifth, and last, extant epigram. This is an epideictic poem on a soldier saved from drowning by his shield, a subject that Zosimus shares with other contemporary and later epigrammatists.³⁷ Zosimus makes the shield speak in the first person. The poem begins as follows (AP 9.40.1– 3 = FGE 426 – 428): Οὐ μόνον ὑσμίνῃσι καὶ ἐν στονόεντι κυδοιμῷ ῥύομ’ ἀρειτόλμου θυμὸν ᾿Aναξιμένους, ἀλλὰ καὶ ἐκ πόντου … Not only in combats and in the groaning din of battle do I protect the spirit of valiant Anaximenes, but in the sea too … (transl. W.R. Paton, adapted).

(discussed above), doubts the ascription Λεωνίδα Ταραντίνου and thinks that Antipater of Sidon may be the author of AP ..  Lycastus was a small town in the neighbourhood of Cnossus.  See Gow/Page  (), .  Cf. Julius Diocles, AP . = GP  ff.; Leonidas of Alexandria, AP . = FGE  ff.; Theon of Alexandria, AP ..

186

Giulio Massimilla

Here the ἀπὸ κοινοῦ construction of the preposition ἐν, that governs both ὑσμίνῃσι and στονόεντι κυδοιμῷ, is particularly remarkable because it applies to three typically Homeric words. I will now consider some passages by poets I have not mentioned so far. It is noteworthy that several of these poets use the ἀπὸ κοινοῦ construction of prepositions with reference to the land-sea polarity. As I said at the beginning, this polarity, that is also at work in the land-sea-air triad throughout the series of epigrams on the three brothers, may have been regarded as particularly appropriate for the σχῆμα ἀπὸ κοινοῦ under the influence of the Homeric phrase ἢ ἁλὸς ἢ ἐπὶ γῆς (Od. 12.27). We can start from a funerary epigram ascribed to Meleager in the Palatine Anthology, and to “Antipater” (without ethnic) in the Planudean (AP 7.470). The majority of scholars favour Meleager’s authorship, but the ascription to Antipater of Sidon has been defended too.³⁸ Perhaps, given Antipater’s fondness for the ἀπὸ κοινοῦ construction of prepositions, the presence of this pattern in the poem could provide one more reason to credit Planudes’ ascription. The epigram is in the form of a dialogue between a passer-by and the dead, a philosopher called Philaulus. The σχῆμα ἀπὸ κοινοῦ occurs in lines 3 f. = HE 4732 f.: – Ἔζησας δὲ τίνα στέργων βίον; Φ. Οὐ τὸν ἀρότρου οὐδὲ τὸν ἐκ νηῶν, τὸν δὲ σοφοῖς ἕταρον. – “What sort of livelihood did you choose to follow?” – Ph. “Not plowing nor sailing, but I was a companion of wise men” (transl. K.J. Gutzwiller).

The preposition ἐκ governs both ἀρότρου and νηῶν, and the whole phrase Οὐ τὸν ἀρότρου / οὐδὲ τὸν ἐκ νηῶν articulates the opposition between land and sea. The same polarity (in reverse order) is again expressed by means of the ἀπὸ κοινοῦ construction of a preposition (this time ἐν) in the first couplet of a funerary epigram for an old fisherman, which the Greek Anthology ascribes to an otherwise unknown poet called Addaeus of Mytilene (AP 7.305.1 f. = GP 47 f.):³⁹

 See Stadtmüller , ; Argentieri ,  f.  See Gow/Page  (),  and . Stadtmüller ,  proposes emending the ascription ᾿Aδδαίου Μιτυληναίου to ᾿Aλφειοῦ Μιτυληναίου. In this case, the author of AP . would be the same Alpheus of Mytilene who wrote one of the epigrams on the three brothers (AP . discussed above). Yet despite the parallels pointed out by Stadtmüller between AP . and some of Alpheus’ epigrams (including AP .), there is no compelling reason to alter the name ᾿Aδδαίου.

The ἀπὸ κοινοῦ Construction of Prepositions

187

Ὁ γριπεὺς Διότιμος, ὁ κύμασιν ὁλκάδα πιστήν κἠν χθονὶ τὴν αὐτὴν οἶκον ἔχων πενίης … Diotimus the fisherman, whose boat was his trust amid the waves, and the same on land the abode of his poverty … (transl. D.L. Page).

Here the words κύμασιν … κἠν χθονί straightforwardly articulate the opposition between sea and land: Diotimus was so poor that the boat he used for fishing when he was at sea became his home when he was on land (at the end of the poem we read that the same boat served as funeral-pyre after Diotimus’ death). A more refined formulation of the contrast between land and sea, this time displaying the ἀπὸ κοινοῦ construction of the preposition εἰς, is to be found at the end of a funerary epigram by the rather obscure poet Isidorus of Aegae.⁴⁰ The deceased, who speaks in the first person, is a farmer turned sea-trader who died because a violent blast of wind capsized the ship he was travelling on. He regrets having foolishly quit farming for trade, and bitterly remarks (AP 7.532.5 f. = GP 3901 f.): οὐκ ἄρ᾽ ἁλωάς αὑτὸς ἐπιπνείει κεἰς ὀθόνας ἄνεμος. They are different winds, I see, that blow on threshing-floors and on sails (transl. D.L. Page).

As the dead man has learnt at his own expense, the mild winds that helped him to do the winnowing on land are not the same as the furious winds that killed him at sea. The ἀπὸ κοινοῦ construction of a preposition also refers to the land-sea polarity in an epideictic epigram of the late antique poet Arabius Scholasticus (6th century AD). A plesasant suburb, blooming with vegetation and close to the sea, speaks in the first person and praises its own wealth (AP 9.667.3 f.): τερπνὰ δέ μοι γαίης τε καὶ ἐξ ἁλὸς ἄλλοθεν ἄλλος καὶ γριπεὺς ὀρέγει δῶρα καὶ ἀγρονόμος. Both the fisherman and the farmer offer me pleasing presents from land and sea (transl. W.R. Paton, adapted).

Here the preposition ἐξ governs both γαίης and ἁλός. The corresponding grammatical subjects are arranged chiastically in the following line (γριπεὺς … ἀγρο-

 See Gow/Page  (), .

188

Giulio Massimilla

νόμος). This passage, with its mention of different dedicators and habitats, has much in common with the epigrams on the three brothers. The ἀπὸ κοινοῦ construction of prepositions, as used by some other poets so far unmentioned in this paper, has no reference to the land-sea opposition. In a funerary epigram for Sophocles, Simias of Rhodes (4th-3rd centuries BC) writes that Sophocles was often awarded with the Acharnian (i. e. Attic) ivy wreath as a sign of his preeminence in both chorus and dialogue (AP 7.21.3 f. = HE 3282 f.):⁴¹ … πολλάκις ὃν θυμέλῃσι καὶ ἐν σκηνῇσι τεθηλώς βλαισὸς ᾿Aχαρνίτης κισσὸς ἔρεψε κόμην. … whose locks the curving ivy of Acharnae often crowned, blossoming in the orchestra and on the stage (transl. W.R. Paton, adapted).

We may note the polarity underlying the phrase θυμέλῃσι καὶ ἐν σκηνῇσι, where the two nouns refer respectively to chorus and actors.⁴² In a funerary epigram by Theodoridas of Syracuse (3rd century BC) the poet’s voice addresses a man shipwrecked and drowned off the north-east coast of Cyprus near the small islands called “the Keys of Cyprus”, at the end of the bay of Cyprian Salamis (AP 7.738.1– 3 = HE 3554– 3556): Κληῖδες πόντου σε καὶ ἐσχατιαὶ Σαλαμῖνος, Τίμαρχ᾽, ὑβριστής τ᾽ ὤλεσε Λὶψ ἄνεμος νηί τε σὺν φόρτῳ τε. The “Keys” of the sea and the outskirts of Salamis and the violent south-west wind destroyed you, Timarchus, with your ship and cargo (transl. W.R. Paton, adapted).

Here the preposition σύν governs both νηί and φόρτῳ.⁴³ In an epigram praising Octavian Augustus, Crinagoras of Mytilene (1st century BC-1st century AD) writes that the emperor’s fame will follow him everywhere, even if he goes to the Hercynian forest in Germany or to the African promontory of Soloeis and to western Libya (AP 9.419.1– 4 = GP 1935 – 1938):

 On the ascription Σιμμίου Θηβαίου, covering both AP . and  in the Palatine Anthology, Gow/Page  (),  comment: “They are not very likely to be by the only Simias of Thebes known to us [i. e. the follower of Socrates appearing in Plato’s Phaedo], and the ascription to S. of Rhodes, though uncertain, seems the most likely solution”. See also Stadtmüller , xxvi.  See Gow/Page  (), .  Lobeck ,  thinks that σύν could also be taken as following νηί in anastrophe: if so, νηί τε σύν would be equivalent to σύν τε νηί. I find this less likely (see also Seelbach , ).

The ἀπὸ κοινοῦ Construction of Prepositions

189

Κἢν μυχὸν Ὀρκυναῖον ἢ ἐς πύματον Σολόεντα ἔλθῃ καὶ Λιβυκῶν κράσπεδον Ἑσπερίδων Καῖσαρ ὁ πουλυσέβαστος, ἅμα κλέος εἶσιν ἐκείνῳ πάντῃ. Though Caesar the Most August should journey to the depths of Hercynia’s forest or outermost Soloeis and the fringe of the Libyan Hesperides, glory shall go with him everywhere (transl. D.L. Page).

Apparently the preposition ἐς in line 1 governs both the phrase μυχὸν Ὀρκυναῖον and the hendyadys πύματον Σολόεντα … καὶ Λιβυκῶν κράσπεδον Ἑσπερίδων, but we cannot rule out that μυχὸν Ὀρκυναῖον is a simple accusative of motion.⁴⁴ The ἀπὸ κοινοῦ construction of prepositions is also to be found in some inscribed epigrams, as a mark of high style. I conclude by considering one of these passages. An epigram from Anazarbus in Cilicia, dating from the 2nd or 3rd century AD, praises the bravery of an equestrian officer and of his aging son, who has followed in his father’s footsteps (SGO vol. 4, 19/17/04.4 f.): ἴκελα γὰρ στρατιαῖς, ἴκελα δ᾽ ἐν πολέμοις καὶ ῥέξαν καὶ ἔτλησαν ὅτ᾽ ἤνθεον. For they did and endured the same things in expeditions and the same things in wars, when they were in the bloom of life (my translation).

Here the σχῆμα ἀπὸ κοινοῦ, with the preposition ἐν governing both στρατιαῖς and πολέμοις, suits well the idea of continuity from father to son that pervades the epigram. To sum up, the passages I have discussed in the second part of this paper suggest that the ἀπὸ κοινοῦ construction of prepositions was regarded as appropriate for the epigrammatic style: it was used as a tool to achieve both brevity and refinement. We have seen too that this σχῆμα ἀπὸ κοινοῦ was perceived as particularly suitable for expressing the land-sea polarity. Furthermore, it seems that a predilection for the ἀπὸ κοινοῦ construction of prepositions was also a matter of personal taste, as in the case of Antipater of Sidon and Zosimus of Thasus.⁴⁵

 See Rubensohn , .  In my discussion of the ἀπὸ κοινοῦ construction of prepositions in Greek epigrams I have not taken into account some passages that seem to be best understood without reference to the σχῆμα ἀπὸ κοινοῦ: Antipater of Sidon, AP .. f. = HE  f. Τὴν καὶ ἅμα χρυσῷ καὶ ἁλουργίδι καὶ σὺν ἔρωτι / θρυπτομένην (here, as Gow/Page  (),  note, “the meaning of σὺν ἔρωτι is not quite plain, but the anaphora of καί suggests that σύν is equivalent to a second ἅμα”: see also Penzel ,  f.); Archias, AP .. = GP  καὶ θυμέλῃσι καὶ ἐν

190

Giulio Massimilla

Abbreviations EpGr. Kaibel Kaibel, G. (1878), Epigrammata Graeca ex lapidibus conlecta, Berlin. FGE Page, D.L. (1981), Further Greek Epigrams, Cambridge. GP Gow, A.S.F./Page, D.L. (1968), The Greek Anthology. The Garland of Philip and Some Contemporary Epigrams, 2 vols., Cambridge. HE Gow, A.S.F./Page, D.L. (1965), The Greek Anthology. Hellenistic Epigrams, 2 vols., Cambridge. SGO Merkelbach, R./Stauber, J. (1998 – 2004), Steinepigramme aus dem griechischen Osten, 5 vols., Stuttgart/Leipzig/Munich.

Bibliography Argentieri, L. (2003), Gli epigrammi degli Antipatri, Bari. Barrett, W.S. (1964), Euripides. Hippolytos, Edited with Introduction and Commentary, Oxford. Bossler, C. (1862), De praepositionum usu apud Pindarum, Darmstadt. Dunbar, N. (1995), Aristophanes. Birds, Edited with Introduction and Commentary, Oxford. Gigante, M. (1971, 20112), L’edera di Leonida, Napoli. — (1979), Civiltà delle forme letterarie nell’antica Pompei, Napoli. Gow, A.S.F./Page, D.L. (1965), The Greek Anthology. Hellenistic Epigrams, 2 vols., Cambridge. — (1968), The Greek Anthology. The Garland of Philip and Some Contemporary Epigrams, 2 vols., Cambridge. Gutzwiller, K.J. (1998), Poetic Garlands. Hellenistic Epigrams in Context, Berkeley/Los Angeles/London. Kaibel, G. (1882), ‘Sententiarum liber secundus’, in: Hermes 17, 408 – 424. Kiefner, G. (1964), Die Versparung. Untersuchungen zu einer Stilfigur der dichterischen Rhetorik am Beispiel der griechischen Tragödie (unter Berücksichtigung des σχῆμα ἀπὸ κοινοῦ), Wiesbaden.

πολέμοισιν (simple locative dative followed by ἐν + dative expressing an outward circumstance); Antiphilus, AP .. = GP  ἠελίῳ κἠν νυκτί (simple locative dative, meaning ὑπ᾽ ἠελίῳ, followed by ἐν + dative as a temporal complement); anonymous epigram, AP .. = FGE  Πυθοῖ κἠν Ἰσθμῷ (simple locative dative of a proper name followed by ἐν + dative of a common name as a locative complement). In Crinagoras, AP .. f. = GP  f. Θάρσει καὶ τέτταρσι διαπλασθέντα προσώποις / μῦθον καὶ τούτων γράψαι ἔτι πλέοσιν, Reiske’s γράψαι ἔτι (where the Palatine manuscript has γραψα ἐνι: the epigram is not in the Planudean Anthology) is better than Rubensohn’s γράψαι ἐνὶ (see Gow/Page  (), ). The σχῆμα ἀπὸ κοινοῦ has been introduced by conjecture into the text of some other epigrams: Simias, AP .. = HE  Μουσῶν ἄμμιγα καὶ Χαρίτων (Hecker proposed κἀκ Χαρίτων); Leonidas of Tarentum, AP .. = HE  ἠέρα καὶ νεφέλας (Meineke proposed κεἰς νεφέλας); Meleager, AP .. f. = HE  f. Τερπνὸς μὲν Διόδωρος, ἐν ὄμμασι δ᾽ Ἡράκλειτος, / ἡδυεπὴς δὲ Δίων, ὀσφύϊ δ᾽ Οὐλιάδης (Graefe proposed emending Τερπνὸς to Στέρνοις).

The ἀπὸ κοινοῦ Construction of Prepositions

191

Kühner, R./Gerth, B. (1898), Ausführliche Grammatik der griechischen Sprache, vol. 2.1: Satzlehre, Hannover/Leipzig. Lapp, F. (1965), De Callimachi Cyrenaei tropis et figuris, Bonn. Laurens, P. (1989, 20122), L’abeille dans l’ambre. Célébration de l’épigramme de l’époque alexandrine à la fin de la Renaissance, Paris. Lobeck, C.A. (18663), Sophoclis Aiax, Berlin. Longo, O. (1986 – 1987), ‘Leonid. AP VI 13 e la sua fortuna (Cacciatori, uccellatori, pescatori)’, in: Museum criticum 21 – 22, 277 – 302. Magnelli, E. (1999), Alexandri Aetoli testimonia et fragmenta. Introduzione, edizione critica, traduzione e commento, Florence. Martelli, M.F.A. (2008), ‘“O cieco Pluto”. L’invocazione di Timocreonte al dio della ricchezza (fr. 731 PMG = Campbell)’, in: Annali della Facoltà di Lettere e Filosofia dell’Università degli Studi di Milano 61, 263 – 276. Massimilla, G. (2010), Callimaco. Aitia, libro terzo e quarto. Introduzione, testo critico, traduzione e commento, Pisa/Rome. — (2011), ‘Tre tipi di armonia musicale nella poesia ellenistica (Honest. AP IX 250, Antip. Sid. APl 220, Call. fr. inc. sed. 669 Pf.)’, in: Eikasmós 22, 171 – 185. Olson, S.D. (2002), Aristophanes. Acharnians, Edited with Introduction and Commentary, Oxford. Page, D.L. (1981), Further Greek Epigrams, Cambridge. Penzel, J. (2006), Variation und Imitation. Ein literarischer Kommentar zu den Epigrammen des Antipater von Sidon und des Archias von Antiocheia, Trier. Reed, J.D. (1997), Bion of Smyrna. The Fragments and the Adonis, Edited with Introduction and Commentary, Cambridge. Rubensohn, M. (1888), Crinagorae Mytilenaei epigrammata, Berlin. Schneider, O. (1856), Nicandrea, Leipzig. Seelbach, W. (1964), Die Epigramme des Mnasalkes von Sikyon und des Theodoridas von Syrakus, Wiesbaden. Stadtmüller, H. (1894), Anthologia Graeca epigrammatum Palatina cum Planudea, vol. 1, Leipzig. — (1899), Anthologia Graeca epigrammatum Palatina cum Planudea, vol. 2.1, Leipzig. Waltz, P. (1906), De Antipatro Sidonio, Bordeaux. Wilamowitz-Moellendorff, U. von (18952), Euripides. Herakles, vol. 2, Berlin.

Style in Literary Epigram a) Sepulchral Style

Egbert J. Bakker

Archaic Epigram and the Seal of Theognis

Words are immortal; stone is not. At least that is the contention of those who make poems out of words. The beginnings of the theme are implicit, without poetry overtly asserting its superiority, when in the Iliad Nestor indicates to his son Antilochos the turning-point in the horse-race in the funeral games for Patroclos that is about to begin: a “sign” whose original function is unclear; it may have been ἤ τευ σῆμα βροτοῖο πάλαι κατατεθνηῶτος ἢ τό γε νύσσα τέτυκτο ἐπὶ προτέρων ἀνθρώπων either the tomb of some mortal who has been dead for a long time, or it was made as a turning-post for men who lived before; Il. 23.331– 2

If this σῆμα is indeed someone’s tomb, it has failed in that function: not only can it not provide everlasting fame to the deceased; it is not even unambiguously recognizable as a tomb any longer.¹ But epic also provides templates for the opposing claim, the everlasting durability of monuments and their stone:² Menelaos builds a cenotaph for Agamemnon, “so that his kléos could never be extinguished” (ἵν᾽ ἄσβεστον κλέος εἴη, Od. 4.84); Hector envisages that the tomb of the adversary he will slay in the upcoming duel will be a conduit for his kléos through the speech of the sailor who perceives it (Il. 7.86 – 91); and the Achaeans build Achilles a tomb “large and perfect, on a prominent headland on the broad Hellespont, so that it is visible from afar for people on the sea, those who live now as well as those who will live after” (Od. 24.80 – 4). When much later the monuments begin to speak through the epigrams inscribed on them, the claim becomes more specific and the stakes go up. The most elaborate and principled statement about material everlasting presence is the so-called Midas epigram, ascribed in the biographical tradition to Homer:³

 On this σῆμα, see also Ford ,  – ; Grethlein , ; cf. Il. . – .  See Redfield , , ; Grethlein ,  – ; Steiner , .  Vit. Hom. Herodot.  –  Allen (the text printed here); Cert.  –  Allen; cf. Peek ; AP ..; Plat. Phdr. d.

196

Egbert J. Bakker

χαλκῆ παρθένος εἰμί, Μίδεω δ’ ἐπὶ σήματι κεῖμαι· ἔστ’ ἂν ὕδωρ τε ῥέῃ καὶ δένδρεα μακρὰ τεθήλῃ ἠέλιός τ’ ἀνιὼν λάμπῃ, λαμπρά τε σελήνη, καὶ ποταμοί γε ῥέωσιν ἀνακλύζῃ δὲ θάλασσα, αὐτοῦ τῇδε μένουσα πολυκλαύτου ἐπὶ τύμβου ἀγγελέω παριοῦσι Μίδης ὅτι τῇδε τέθαπται. I am a bronze maiden, and have been placed on the tomb of Midas. As long as water flows and tall trees bloom, the sun rises and shines as well as the bright moon, as long as rivers flow and the sea washes up against the shore, remaining in this self-same place on the much-bewailed tumulus I will pronounce to all who pass by that Midas lies buried here.

Reading this epigram as the composition of Cleoboulos of Lindos,⁴ Simonides takes issue with it in a poem that specifically questions the durability of the stone that must have been the material support of the inscription:⁵ τίς κεν αἰνήσειε νόωι πίσυνος Λίνδου ναέταν Κλεόβουλον, ἀεναοῖς ποταμοῖσ’ ἄνθεσι τ’ εἰαρινοῖς ἀελίου τε φλογὶ χρυσέας τε σελάνας καὶ θαλασσαίαισι δίναισ’ ἀντία θέντα μένος στάλας; ἅπαντα γάρ ἐστι θεῶν ἥσσω· λίθον δὲ καὶ βρότεοι παλάμαι θραύοντι· μωροῦ φωτὸς ἅδε βουλά. Who in his right mind would praise Kleoboulos who dwells in Lindos? This is the man who placed the force of the stele against ever-flowing rivers, the spring-flowers, the gleam of the Sun and of the golden Moon as well as the whirls of the sea. Everything is less than the gods. A stone even the hand of man can shatter to pieces: this is a fool’s thought. Simon. PMG 581

It is ironic, of course, that Cleoboulos’ epigram did gain the more fluid immortality of words: as attested throughout antiquity by the numerous quotations of it— and their numerous textual differences—it became detached from its original inscribed locus on stone. The monument itself, by contrast, is more than unrecognizable; in vanishing completely it has fared worse even than Nestor’s turningstone. Simonides himself, too, has lent his voice to stone in verses that became equally “delapidarized”, as his famous Thermopylae epigram can attest: we

 Diog. Laert. ., where both the epigram and Simonides’ poetic critique are quoted.  On this poem and the Midas epigram, see also Ford ,  – .

Archaic Epigram and the Seal of Theognis

197

would not have known it if not for the literary tradition, which has yielded multiple versions of it:⁶ ὦ ξεῖν᾽, ἀγγέλλειν Λακεδαιμονίοις ὅτι τῇδε κείμεθα τοῖς κείνων ῥήμασι πειθόμενοι Stranger, report to the Lakedaimonians that here we lie, obeying their words

These epigrams have lost their stone support, which made them lose their textual integrity, but also, paradoxically, ensured their endurance as disembodied literature.⁷ A further difference between poems and stones is that poems can travel, whereas monuments cannot; poems can go in active search of their audiences, whereas monuments are stationary; they have to wait, passive and immobile, until someone comes to them or happens upon them. Pindar’s statement is probably the best-known expression of this principle: οὐκ ἀνδριαντοποιός εἰμ’, ὥστ’ ἐλινύσοντα ἐργάζεσθαι ἀγάλματ’ ἐπ’ αὐτᾶς βαθμίδος ἑσταότ’ I am no sculptor, to craft statues that stand unmovable on the same pedestal. Pind. Nem. 5.1

But this movability is not without risk in some cases, as we shall see later, and some poems may need a symbolic “lapidarization.”

The stone and its voices In spite of its claims as to its natural durability, then, stone can be seen by some poets as at a disadvantage in the confrontation with the poetic word. Still, as this chapter will argue, the durable and unmovable monumentality of inscribed epigram is the benchmark by which written texts are measured and the standard to which

 See Baumbach ,  –  (“Entlapidarisierung”); cf. Baumbach, Petrovic, and Petrovic ,  – .  Some textual variants of the Midas epigram: . om. . Μίδεω: Μίδου, Μίδα; κεῖμαι: ἧμαι, . ῥέῃ: νάῃ. . λάμπῃ: φαίνῃ. . γε ῥέωσιν: πλήθωσι; ἀνακλύζῃ: περικλύζῃ. . πολυκλαύτου ἐπὶ τύμβου: πολυκλαύτῳ ἐπὶ τύμβῳ. . ἀγγελέω: σημανέω. Variants of the Thermopylae epigram recorded by Strabo (..; cf. Baumbach, Petrovic and Petrovic ,  n. . ἀγγέλλειν: ἀπάγγειλον. . ῥήμασι πειθόμενοι: πειθόμενοι νομίμοις.

198

Egbert J. Bakker

they aspire. The archaic inscribed epigram is more than just the model for epigram as a literary genre in Hellenistic times and beyond; it sets the standard for writing itself, even for the very notion of authorship. I will explore this idea on the basis of the awareness of epigram that is apparent in the elegiac corpus of the Theognidea. I start by emphasizing the by now well-understood fact that reading and writing in archaic contexts involves voice, the physical voice of both writer and reader and of the object or stone itself. The idea that writing cannot exist without voice is evident in the fact that the writing on the earliest inscribed objects is construed as the voice of that object. One of the best known “speaking objects” (where the object has actually been preserved, not just the inscription) is the inscribed Mantiklos figurine from Delphi (first quarter of the 7th c. BC), now in the Boston Museum of Fine Arts: Μάντικλος μ᾽ ἀνέθεκε ϝεκαβόλοι ἀργυροτόξσοι τᾶς {δ}δεκάτας· τὺ δὲ, Φοῖβε, δίδοι χαρίϝετταν ἀμοιβ[άν] Mantiklos has dedicated me to the far-hitter of the silver bow as part of the tithe; and you, Phoibos, give pleasing compensation.

The text evinces a formulaic grammar for dedications, with fixed slots for the names of the dedicator and the divinity around the structural element μ᾽ ἀνέθηκε placed before the trochaic caesura.⁸ The object speaks, but it evokes the presence of both the god and the dedicator and in reading the epigram the reader reperforms the original act of dedication.⁹ The practice of giving “voice” to a material object through a written inscription is particularly prominent and common in the case of early dedicatory inscriptions.¹⁰ But some of the early sepulchral monuments speak for themselves as well. In addition to the Midas epigram quoted above, we can think of the following two instances, both early sixth-century pieces, from Corcyra and Methana, resp: στάλα Ξενϝάρεος τοῦ Μhείξιός εἴμ᾽ ἐπὶ τύμῳ I am the stele of Xenwares the son of Meixis on his tomb. CEG 146; 575 – 50 BC; Corcyra

 See also examples in the list provided by Trümpy ,  – . The second hemistich of line  (δίδοι χαρίϝετταν ἀμοιβ[άν]) is Homeric (Od. .); or perhaps we should venture the opposite possibility that the single Homeric occurrence picks up a sacrificial or dedicatory formula in actual religious practice (e. g., CEG  – ). See also Wachter , .  On dedicatory inscriptions as reperformances of original acts of dedication, see Day  (discussion of the Mantiklos text at Day ,  – ).  Wachter ,  –  argues that the practice of endowing objects with “speech” originated with dedications.

Archaic Epigram and the Seal of Theognis

199

Εὐμάρες με πατὲρ ᾿Aνδροκλέος ἐντάδε σᾶμα ̃ ποιϝέσανς καταέθεκε, φίλο μνᾶμα hυιέος ἐμεν Eumares, father of Androcles made and set me up here as a monument, to be a memorial of his dear son. CEG 137; 600 BC, Methana

The “speech” of these oggetti parlanti has been considered in animist terms, as if they are magically endowed with speech.¹¹ This conception is countered by Jesper Svenbro (1993) with a typically poststructuralist conception of the “speaking subject.” All that counts for Svenbro is the monument’s or object’s ego-centric “hereness”, a presence that goes at the expense of the writer, who, as Svenbro puts it, effaces himself in the act of writing.¹² The first-person assertion cannot be separated from the object/monument; but the very act of reading the inscription—the reader lending his voice to its statement—removes the link between the object’s egō and the voice. Svenbro’s conception of reading is imbued with a typically poststructuralist resistance to the “metaphysics of the voice”, to the “conviction that the first person necessarily implies an inner life and a voice.”¹³ But we do not have to suppress any “inner life” of the speaking object or to resort to primitive animism to see the votive objects and sepulchral monuments of archaic Greece as having an agency that is specified by a “voice.”¹⁴ Mantiklos’ figurine is an enduring agent in the exchange of charis between its dedicator and Apollo. There is intended “synchronic” reciprocity, a relation that is projected, diachronically, onto the future, and the inscription says so. Further, voice in the context of archaic Greek reading is more than making up for the absence of a “writer.” It is acting out a role—or roles—encoded in the writing. In this way it is not so much the writer’s absence that is compensated for as his presence perpetuated: reading the inscription is reperforming the original act of dedication or consecration. “Writer” in fact, is a reductive term in this connection. Mantiklos’ and Eumares’ “writing” in the epigrams just presented is subsumed under the much more essential role of dedicating and burying. And the reader is there to reconfirm the “writer” in this original role, the action that resulted in the physical “hereness” of the object or monument in question. I argue, then, that Svenbro’s “anthropology of reading” does not take into account the cultural and cognitive aspects of the writing and reading of archaic epi   

Burzachechi . Svenbro ,  – . Svenbro , , citing (n. ) Derrida’s La voix et le phénomène. On the agency of material objects, see Gell .

200

Egbert J. Bakker

gram as an encoding and a decoding of a role and a voice. But there are also more straightforward linguistic problems with his conception. The category of speaking objects is not as large as Svenbro’s understanding of the inscriptions would have it, and we may ask what it is, linguistically, that constitutes a “speaking object.” The first-person markers (the personal pronoun ἐγώ and the first-person form of the verb) must refer to the physical object. This qualifies the Midas epigram as a speaking object (χαλκῆ παρθένος) as well as the Xenwares stele (στάλα … εἰμ᾽) as well as the Eumares tomb (με … σᾶμα), and of course the Mantiklos figurine. But it disqualifies, for example, the well-known Phrasikleia statue: σε̃μα Φρασικλείας· κόρε κεκλέσομαι αἰεί, ἀντὶ γάμο παρὰ θεο̃ν τοῦτο λαχο̃σ᾽ ὄνομα The tomb of Phrasikleia. “I will always be called ‘girl’, having received this name as share from the gods instead of marriage.” CEG 24; Attica, mid-6th c.

Punctuating differently (not reading a full stop after the proper name), Svenbro translates the first line as “I, Phrasikleia’s sêma, shall always be called girl”, thus assuming an appositional relationship between σε̃μα and κεκλέσομαι. The biographical details (as opposed to references to the speaker as artifact), however, strongly argue against a “speaking object” interpretation. It seems more fruitful to read the inscription (after the proper name) as the speech of the dead girl, embodied by the stone and performed by the reader. Next to the voice of the object, then, the stone can carry the voice of the deceased, which puts the reader in what we could call a mimetic role.¹⁵ A third voice comes into play when we realize that Svenbro’s “ego-centric” interpretation of a number of inscriptions that do not feature the first-person pronoun or a first-person verb hinges on a questionable interpretation of the proximal deictic ὅδε that is a defining feature of many of the sepulchral and dedicatory inscriptions. For Svenbro this demonstrative pronoun is used in archaic Greek in a way that is at odds with English or French syntax, in that it is semantically so closely related to the first grammatical person as to take over its syntactic properties. Thus, for example, in the following case, Δϝεινία τόδε [σᾶμα], τὸν ὄλεσε πόντος ἀναι[δές] This the tomb of Dweinias, whom the ruthless sea destroyed. CEG 132; Corinth, mid-7th c. BC

 On the deceased speaking, see Vestrheim ,  – ; Tsagalis , ; compare also the Thermopylae epigram quoted above.

Archaic Epigram and the Seal of Theognis

201

Svenbro would translate “I am the tomb of Dweinias”, as if the mere occurrence of a τόδε or ὅδε (‘this/me here’) equals εἰμί ‘I am’ when the verb is absent.¹⁶ It is true that there is an affinity between the proximal deictic pronoun ὅδε and the first grammatical person, just as there is one between οὗτος and the second, and between ἐκεῖνος and the third.¹⁷ This “objective” function, however, is only part of the meaning of the Greek demonstratives and crucially leaves out the cognitive-pragmatic dimension. The pronoun ὅδε is associated with “me” not in a stable, objective (i. e., semantic or lexical) link, but typically insofar as that what is close to “me” (the speaker), either physically or perceptually, is not necessarily close to “you” (the hearer); by an easy cognitive extension, “close” can become “known” or “familiar”, unknown or unfamiliar to an interlocutor and hence potentially newsworthy or salient.¹⁸ The speaker, then, who refers to himself with ὅδε does so, not because the pronoun inherently means “me”, but because he presents himself as something noteworthy to his interlocutor. This is particularly clear in one of the literary examples that Svenbro cites (Odysseus addressing Telemachus): ἀλλ᾽ ὅδ᾽ ἐγὼ τοιόσδε, παθὼν κακά, πολλὰ δ᾽ ἀληθείς, ἤλυθον εἰκοστῷ ἔτεϊ ἐς πατρίδα γαῖαν Me here (ὅδ᾽ ἐγώ), such as I am, after suffering hardship and many wanderings I have come in the twentieth year to my fatherland. Od. 16.205 – 6

Odysseus surely constitutes the most noteworthy presence that ever was before Telemachus’ eyes.¹⁹ The cognitive understanding of ὅδε may alert us to the fact that the monument or object and its epigram are part of a dynamic communicative exchange. The monument is not “ego-centric” and hence monologic, but instead intensely reader-oriented and dialogic. The epigram provides the answer to a projected question on the part of a prospective reader: “Whose tomb are you/is this?” “Who set up this stele?” “Who dedicated you/this?” The question usually remains implicit—as in historiographical narrative, when Herodotus writes, for ex Svenbro , .  Kühner-Gerth , , also referenced by Svenbro.  Bakker ,  – .  The neglect of the pragmatic and cognitive dimension is equally apparent in the case of ἐκεῖνος, which Svenbro (, , ) treats as the marker of its referent’s absence. But physical absence may be mental presence, and ἐκεῖνος is frequently used to mark a “continuous” topic, i. e., it tracks a referent that had been established earlier in the flow of discourse. This happens, for example in Simonides’ Thermopylae epigram cited above; cf. CEG .

202

Egbert J. Bakker

ample, “Cyrus spoke thus” (ἔλεγε τάδε, 1.210) in answer to the reader’s implicit question “What did Cyrus say?” We might read Xenwares’ stele as an answer to an implied question like this. Its voice is that of the monument qua answer to the question; but since the answer is received in reading, the reader becomes part of it and we can equally see the epigram as an assertion on the part of the reader based on the information provided by the epigram. The answer and its performance coincide in the act of reading. In some (often cited) cases the question is made explicit, so that the inscription frames itself as an explicit answer, with a “dialogue” between the reader and the object or monument. Reading epigrams like this is performing yourself, staging yourself as reader: πᾶσιν ἴσ᾽ ἀνθρόποις hυποκρίνομαι hόστις ἐ[ροτ]ᾶι hός μ᾽ ἀνέθεκ᾽ ἀνδρο̃ν· ᾿Aντιφάνες δεκάτεν To all people I give the same response, whoever it is who asks who of men has set me up as votive offering: Antiphanes as a tithe. CEG 286; Acropolis, 490 – 480 BC αὐδὴ τεχνήεσσα λίθο, λέγε τίς τόδ᾽ [ἄγαλμα] στῆσεν ᾿Aπόλλωνος βωμὸν ἐπαγλαΐ[σας]. Παναμύης υἱὸς Κασβώλλιος, εἴ μ᾽ ἐπ[οτρυνεις] ἐξειπε̃ν, δεκάτην τήνδ᾽ ἀνέθηκε θε[ῶι]. Expertly crafted voice of the stone, tell who has placed this statue here, lending splendor to Apollo’s shrine? Panamues son of Casbollis, if you do urge me to speak up, as a tithe he set this for the god. CEG 429; Halicarnassos, 475 BC

In the first case the reader’s question (“Who dedicated you?”) is taken for granted and subsumed in the inscribed statement; the “response” proper does not come until the second half of the pentameter in a contextually conditioned elliptic utterance (the central dedication verb has been cleverly displaced into the framing question): ᾿Aντιφάνες δεκάτεν. The stone’s answer is “equal” to all people: the epigram’s permanence has “sealed” the dedicator’s original act as an egalitarian offering. In the second case the reader’s question is represented directly, as he addresses the stone’s αὐδή as an interlocutor in its own right. The reader’s question is prompted by his curiosity at seeing the agalma and hence logically comes before the reading of the inscription. This means that the reader is staged qua reader by the very text he is reading.²⁰

 On these two epigrams, see also Tueller ,  – ; the αὐδὴ τεχνήεσσα the voice rich in

Archaic Epigram and the Seal of Theognis

203

The monumental aorist We see, then, that Svenbro’s ego-centric monuments and objects are in reality dialogic; they expect, even assume, the keen attention of readers, who will want to know the details: whose dedication is this? Who set up this tomb? Such questions may lead us back to the concept that goes hand in hand with the monument’s purported ego-centricity in Svenbro’s conception: the absence of its writer. In a trivial sense any writer is absent for the act of the reading of his writing to be performed successfully and meaningfully. But in the case of archaic inscribed epigrams we may wonder whether the dedicators of objects and erectors of tombs really think of themselves as being absent when the act of reading their statement is performed. In this regard there is a linguistic element that stands out next to the proximal deictic pronoun as a reader- or reading-oriented signal, and that is the stereotypical aorist verb, ἀνέθηκε in the case of dedications (as in the examples presented earlier) and ἐπέθηκε (sometimes κατέθηκε or simply θῆκε) or ἔστησε in the case of funeral monuments, of which two examples:²¹ σε̃μα πατὲρ Κλέ〚β〛ολος ἀποφθιμένοι Ξενοφάντοι θε̃κε τόδ᾽ ἀντ᾽ ἀρετε̃ς ἐδὲ σαοφροσύνες This tomb his father Kleoboulos for deceased Xenophantos has set up, in recognition of his excellence and moderation. CEG 41; Attica, 530 – 20 BC σε̃μα Κύλον παίδοι ἐπέθεκε{ν} θανότοι με̃μα φιλεμοσύνες The tomb Cylon has set up for his two dead sons, A memorial to his affection. CEG 32; Attica, ca. 530 BC

It would be too simple to say that this is simply a past verb, reporting on a past action by someone, the writer or commissioner, who is absent when the act of reading is performed.²² The temporal deixis here is more complex and interesting than that. The aorist verbal stem not only signals that the verb’s transitivity, the

(i. e., enabled by) τέχνη, I would take with the subjective genitive as belonging to the stone, rather than to the reader (pace Svenbro ,  – ).  For a discussion of the aorist in archaic dedicatory and sepulchral epigrams, see also Bakker ,  – ; for the meaning of the aorist at large, see Bakker ; ,  – .  Svenbro , : “Soon the writer would no longer be there; he became the third person by virtue of the fact that he wrote.”

204

Egbert J. Bakker

transfer of energy it denotes, has been completed; the transfer of energy has resulted in a physical object that has been able to withstand the tooth of time until the moment of its reading. The temporal reference of the verb bridges the temporal gap between the past act of erecting the monument or dedicating the object and the present act of contemplating it and reading the inscription.²³ We could even say that the act denoted by the “monumental aorist” is not completed until the reader allows the monument to say that so-and-so has erected it—or, conversely, until the monument prompts the reader to say this. In other words, the inception of the action—When exactly did Antiphanes’ act of dedication take place? When did Xenophantos die and did Kleoboulos erect the tomb? —is less important than that a reader reads about that moment in his own here and now, each time anew, as long as the monument is standing. The epigram is strongly reader-oriented, and the tomb meant to be, drawing on Thucydides’ phrase, a μνῆμα ἐς ἀεί.²⁴

Competing for kléos We saw that Agamemnon’s and Achilles’ tombs in the Iliad lend kléos to the deceased as memorials conspicuous in the landscape. Similarly the tombs of Xenwares and Dweinias communicate through their inscriptions (quoted above) their identity, and hence their kléos to posterity. The monumental aorist complicates this picture, since the act of creating the monument may be thought of as deserving kléos as well—and in some cases is explicitly said to have been performed with an eye toward kléos for the performing agent.²⁵ The creator of a funeral monument is in this regard in a different position from someone who dedicates an object to a god in a sanctuary. The god does not need the surrogate immortality that kléos provides and his or her name does not even have to be mentioned, depending on the context of the dedication. Antiphanes’ inscription, quoted above, does not specify any god, and Panamyes’ mentions only “the god”, the name of Apollo being implied after the mentioning of the god’s sanctuary.

 The actual crafting of the monument or the statue (and act subsumed under the more general and fundamental act of ordering the monument to be erected) is sometimes (though not always) referred to in the imperfect ἐποίει (e. g., CEG —cf. Bakker , —, , ), which presents the act as located in the past.  Thucydides draws on the temporal reference of what I call here the “monumental aorist” in the first sentence of his work; see Bakker ,  – ; : .  E. g., CEG , .

Archaic Epigram and the Seal of Theognis

205

In the case of the sepulchral monument and epigram the situation is different. The sepulchral monument’s creator, the subject of the monumental aorist, is in the information structure of the inscription no more than one of the two newsworthy items next to the name (and sometimes biography) of the person buried. In many cases the creator’s name, often the dead person’s parent, is presented as at least as important as that of the dead person, as in the case of the Kleoboulos epigram just cited. And in some cases the name of the person buried is not even provided, which in a real sense defies the monument’s purpose as a “sign” pointing to the person deceased and thereby prolonging his life. This happens in the case of the Cylon epigram just cited.²⁶ There is a competition, then, between the grave monument’s beneficiary and its creator; for both the tomb is a venue for kléos. This tension comes out clearly in what may be narratologically the most complex epigram in the archaic and classical material, the Timarete (or Mnesitheos) inscription from Eretria (midfifth century), which contains at least three voices: χαίρετε τοὶ παριόντες· ἐγὸ δὲ θανὸν κατάκειμαι. δεῦρο ἰὸν ἀνάνεμαι· ἀνὲρ τίς τε̃ι δε τέθαππται; ξε̃νος ἀπ᾽ Αἰγίνες, Μνεσίθεος δ᾽ ὄνυμα· καί μοι μνε̃μ᾿ ἐπέθεκε φίλε μέτερ Τιμαρέτε τύμοι ἐπ ᾿ ἀκροτάτοι στέλεν ἀκάματον, hάτις ἐρεῖ παριôσι διαμερὲς ἄματα πάντα· Τιμαρέτε μ᾽ ἔσστεσε φίλοι ἐπὶ παιδὶ θανόντι Greetings, passers-by! This is me here, lying dead. Come here and read. “What man is buried here?”— “A stranger from Aegina. Mnesitheos is his name”— and my dear mother Timarete has set up this monument on the top of the mound, an imperishable stele that will say continuously for all days to come the following words: “Timarete has set me up upon her dear son, who has died.” IG XII 9.285 = CEG 108; Eretria, 450 BC

This epigram consists of two elegiac couplets framed by one hexameter at the beginning and two at the end. The final line is typologically similar to the first line of the Cylon epigram just cited: the name of the monument’s creator as subject of an aorist verb and the designation “child” in the dative. The monument is “speaking” in this line,²⁷ but in the earlier parts of the epigram we encounter two  The names of the dead sons may of course have been elsewhere on the monument, on statues, for example.  Not for Svenbro ,  – , who reads the beginning of the previous line as οὗ τις ἐρεῖ παριοῦσι, to be translated as “where someone will say.”

206

Egbert J. Bakker

more voices. Just like Phrasikleia, the person buried, Mnesitheos, speaks, addressing the passers-by, the readers, who in the act of reading will perform his voice. The dead man’s voice is mediated both through the monument, the stele mentioned later (possibly a statue as well), and through the voice of the reading passer-by. The text, in fact, sets up the reader qua reader, since the exhortation can of course be received only when the reader is already reading. The question ἀνὲρ τίς τε̃ι δε τέθαππται; I take as a direct question on the part of the reader (or rather, staged for him), who in voicing his curiosity becomes the third voice in the epigram. The reader’s question of course logically precedes the answer found in the act of reading, but the conceit is that it is subsumed in that act. The monument, in other words, does not oppose itself to the passer-by and his question; just like Antiphanes’ dedication cited earlier, it encompasses the passer-by. It does not merely invite the act of reading; it embodies the act of reading, and so the reader is part of the monument. The answer to the reader’s question continues to be in the dead man’s voice (line 3), who adds supplemental information, the identity of the monument’s commissioner, the dead man’s “dear mother”, Timarete. In this way the dead man’s voice stages the voice of his own monument, which takes center stage in the last line of the text, Τιμαρέτε μ᾽ ἔσστεσε φίλοι ἐπὶ παιδὶ θανόντι, which as we saw could well serve as a one-hexameter inscription in its own right. The monument becomes in the end a “speaking object” like Eumares’ monument for his son Androcles cited earlier (p. 199). This complex text thus contains three voices and is a compound containing two simpler types, the “This is the grave of so-and-so” type and the “So-and-so has set me up/set this up” type. Both the name of the deceased and that of the commissioner are highlighted and the monument’s commemorative function pertains to both.

Placing the seal We saw that statue of Phrasikleia points to the untimely death of the young girl who died before marriage; the bronze maiden on Midas’ tomb points to Midas. But the statue of Mnesitheos (if that is what it is; or else the stele carrying his voice) points to Timarete, who ordered the construction of the tomb and its stele. She is the monument’s “author”, but her name comes to be activated only through the activation of her son’s name. The two are locked in mutual dependency: Mnesitheos owes to his mother Timarete the tomb that bears his name; Timarete cannot impart her name to posterity without her son’s tomb. In calling Timarete an “author”, I do not mean to imply that she actually composed the text of the epigram, but inasmuch as she is responsible for the

Archaic Epigram and the Seal of Theognis

207

monument as its commissioner she is its “writer” or author: the work bears her name. The situation envisaged, with a given feature of a “work” pointing to its “author”, aligns the Mnesitheos-Timarete tomb with a celebrated—and vexed— question in archaic Greek poetics, the so-called “seal” of Theognis. As is well-known, the name of Theognis as author is acted out in the poem that according to many stood at the head of a written original collection of the songs of “Theognis” that developed over time into the Theognidea we now possess:²⁸ Κύρνε, σοφιζομένωι μὲν ἐμοὶ σφρηγὶς ἐπικείσθω τοῖσδ’ ἔπεσιν, λήσει δ’ οὔποτε κλεπτόμενα, οὐδέ τις ἀλλάξει κάκιον τοὐσθλοῦ παρεόντος· ὧδε δὲ πᾶς τις ἐρεῖ· ‘Θεύγνιδός ἐστιν ἔπη τοῦ Μεγαρέως· πάντας δὲ κατ’ ἀνθρώπους ὀνομαστός. ἀστοῖσιν δ᾽ οὔπω πᾶσιν ἁδεῖν δύναμαι. οὐδὲν θαυμαστὸν Πολυπαΐδη· οὐδὲ γὰρ ὁ Ζεὺς οὔθ᾽ ὕων πάντεσσ᾽ ἁνδάνει οὔτ᾽ ἀνέχων. Kurne: let a sphrēgis be placed by me, as I perform my trick, in/with these very lines; their theft will never go unnoticed, nor will anyone change them for the worse with the excellent one [/element] present; and so will everyone speak: “It’s the lines of Theognis of Megara; he is named among all men.” But my townsmen, I am not yet able to please them all. Nor is this surprising in any way, son of Polypaos: not even he himself, Zeus, is able to please all alike, either by providing rain or by withholding it. Thgn. 19 – 26

This elegy is among the most discussed in the corpus of archaic Greek poetry, especially among scholars of previous generations.²⁹ But instead of returning directly to the most discussed items (e. g., the meaning of σοφιζομένῳ or σφρηγίς), let us first pay attention to the ways in which it borrows the language and concepts of inscribed epigram. The thematically central verb ἐπικεῖμαι is an element of sepulchral epigram. We encountered it in the Midas epigram (Μίδεω δ’ ἐπὶ σήματι κεῖμαι ‘ on the tomb of Midas that I sit’), and the verb is attested elsewhere in the same function on sepulchral monuments (e. g., μνῆμα τόδ᾽ ἔστ᾽ ἐπὶ σάματι κείμενον ‘this memorial here sitting on the tomb’, CEG 83, Athens, mid-5th century). It is of course the passive-intransitive counterpart of the ubiq-

 On the history of the collection, see West ; Bowie .  On this poem, see the extensive discussions (with extensive bibliography) by Selle ,  –  and Colesanti ,  – . The latest engagements in English with the seal and its interpretive problems are Ford , Edmunds , and Bakker, forthcoming.

208

Egbert J. Bakker

uitous monumental aorist ἐπέθηκε ‘has placed upon’ which we also find in the Mnesitheos-Timarete epigram. If the active aorist is the performance of the perpetuation of the monument in the act of reading, then the passive third-person imperative ἐπικείσθω can perhaps be seen as the performance of the original act of consecration of the grave: “Let the marker hereby be placed upon the grave.” An actual attestation of the similar original act of dedication (to be perpetuated by the aorist ἀνέθηκε) occurs in Plato’s Symposium, when Agathon dedicates his speech to Eros: οὗτος … ὁ παρ᾽ ἐμοῦ λόγος … τῷ θεῷ ἀνακείσθω (Pl. Symp. 197e6 – 7): “Let hereby this speech of mine be dedicated to the god.”³⁰ It appears, then, that the σφρηγίς sits on “these verses here” (τοῖσδ’ ἔπεσιν) —note the proximal deictic—as a statue or a stele on a tomb, a σῆμα. The reaction of the anonymous listener in the future (ἐρεῖ) is not only a conscious allusion to the sailor who “reads” the tomb of Hector’s victim from afar (Il. 7.89 – 91) and pronounces (ἐρέει) Hector’s kléos, but also to the reader of the epigram, cast in the role of passer-by or of interlocutor of the speaking object. The reader of epigram is the recipient of kléos, but also its source, inasmuch as reading is a redistribution of language, as well pointed out by Svenbro in a discussion of the reading-verb ἀνάνειμαι in the Mnesitheos-Timarete epigram.³¹ In the same way, all the recipients of Theognis’ poetry—participants in symposia all over the Greek world—will propagate the poet’s fame by “reading” out the recognition of Theognis’ poetry that they have just experienced. But what is it that makes everyone recognize these verses as the epē of Theognis? This feature of the poetry is both something that makes it recognizable, identifiable, but also something that fixes it, so that “these verses” come to do what Antiphanes’ dedicatory inscription says it does: giving all people an equal “answer”, preventing the verses from saying different things to different people because of changes made to them. The idea of sphragis fits both functions: a sphragis can act as signet or label identifying something as someone’s property; and it can be a seal, preventing tampering with or stealing the content of a container.³² The latter is quite explicitly addressed in the second and third lines of the poem. I propose to see in the prevention of theft and change that the seal purportedly will effect a symbolic “lapidarization.” The suppression of the poetry’s developing variants through an authoritative written collection is seen in terms of the perma-

 On the performative properties of third-person passive imperatives, see further Bakker, forthcoming.  Svenbro ,  – . Of course, if Svenbro’s reading of οὗ τις for ἅτις in l.  of the Mnesitheos-Timarete epigram holds, there would be another strong parallel for Theognis’ sphrēgis poem.  See, e. g., Woodbury .

Archaic Epigram and the Seal of Theognis

209

nence of inscribed epigram, the creation of a text that “says equal things to all people”, not different things to different people in different places.³³ As to what actually constitutes the sphrēgis, I side with those who have seen in it the vocative Κύρνε, which occurs in every genuine Theognis song.³⁴ The later practice of literary sphragis, an author’s signature embedded in the texture of the work, would seem to favor the proper name Θεύγνιδος (ἐστὶν ἔπη), but that expression of recognition seems more the reaction, the consequence of the observed presence of the seal, not the seal itself. Kurnos is the putative ever-present recipient of the poetry, but in being always there he comes to seal the poetry through the very presence of his name.³⁵ The idea of Κύρνε as the seal is in itself not new, but the epigrammatic connection adds a new dimension. In terms of the performance of Theognis’ elegies, the boy whose presence is conjured by the vocative constitutes the ideal audience, but in terms of a written corpus the vocative becomes the poetry’s typifying marker. It is placed “on” the poetry, as its most conspicuous feature, directing the listener to who “authored” the songs. In other words, the collection of sympotic elegiac Theognis songs comes to function as Kurnos’s tomb, created for him by Theognis, whose fame is mediated by his poetic creation, just as Timarete’s fame is channeled through her dead son Mnesitheos’s tomb that she has commissioned. Just like Timarete and her son, Theognis and “Kurnos” rely on each other for their fame. We might perhaps even go further and align young Kurnos with the dead son buried by a father who “signs” the grave in order to selfmemorialize. The interdependence between Theognis and Kurnos comes out in more detail when we read the sphrēgis poem in conjunction with the elegy in which Kurnos’s fame is announced. This song, replete with Homeric and epigrammatic allusions, is often assumed to have been placed at the end of the elegiac sylloge, along with the sphrēgis poem bookending the collection:³⁶ σοὶ μὲν ἐγὼ πτέρ’ ἔδωκα, σὺν οἷσ’ ἐπ’ ἀπείρονα πόντον πωτήσῃ, κατὰ γῆν πᾶσαν ἀειρόμενος ῥηϊδίως· θοίνῃς δὲ καὶ εἰλαπίνῃσι παρέσσῃ ἐν πάσαις πολλῶν κείμενος ἐν στόμασιν,

240

 In Bakker, forthcoming I go further and argue that fixation goes beyond textuality and includes the poetry’s audience and setting.  An attempt at bibliographical exhaustiveness is made in Selle ,  n. .  I argue that τοὐσθλοῦ παρεόντος in the poem’s third line (Thgn. ) applies both to textual integrity and to the integrity and stability of the poetry’s audience (“with the noble one present,” taken as a genitive absolute with conditional force).  See Selle ,  – .

210

Egbert J. Bakker

καί σε σὺν αὐλίσκοισι λιγυφθόγγοις νέοι ἄνδρες εὐκόσμως ἐρατοὶ καλά τε καὶ λιγέα ᾄσονται. καὶ ὅταν δνοφερῆς ὑπὸ κεύθεσι γαίης βῇς πολυκωκύτους εἰς ᾿Aίδαο δόμους, οὐδέποτ’ οὐδὲ θανὼν ἀπολεῖς κλέος, ἀλλὰ μελήσεις ἄφθιτον ἀνθρώποισ’ αἰὲν ἔχων ὄνομα, Κύρνε, καθ’ Ἑλλάδα γῆν στρωφώμενος, ἠδ’ ἀνὰ νήσους ἰχθυόεντα περῶν πόντον ἐπ’ ἀτρύγετον, οὐχ ἵππων νώτοισιν ἐφήμενος· ἀλλά σε πέμψει ἀγλαὰ Μουσάων δῶρα ἰοστεφάνων. πᾶσι δ’, ὅσοισι μέμηλε, καὶ ἐσσομένοισιν ἀοιδή ἔσσῃ ὁμῶς, ὄφρ’ ἂν γῆ τε καὶ ἠέλιος. αὐτὰρ ἐγὼν ὀλίγης παρὰ σεῦ οὐ τυγχάνω αἰδοῦς, ἀλλ’ ὥσπερ μικρὸν παῖδα λόγοις μ’ ἀπατᾷς.

245

250

To you I have given wings, with which to the boundless sea you will fly, lifted over the entire earth, easily; at feasts and banquets you will be present, at all of them, placed in the mouth of many; and with their pipes of clear sound the young men lovely in decorous fashion beautiful and clear songs they will sing you; and when in the depths of the gloomy earth you will go into the much-wailing abode of Hades, you will never, not even in death, lose your kleos; no, you will be on the mind of the people, each time, with a name unwilting, Kurne, roaming all over the land of Hellas and through the islands, crossing the barren sea rich in fish, not riding on horseback; your vehicle instead will be the splendid gifts of the Muses crowned with violets, and to all those who care, even men of the future, song you will be, equally , as long as earth and sun . But I, I do not even get the slightest respect from you; no, as if I am a small boy you are deceiving me with idle talk.

This poem reads as a reflection on the effects of the sphrēgis, after it has been allowed, performatively, to do its work in and through the songs contained in the collection. It announces the addressee’s eternal renown, the kléos that is the compensation for death. There is in ll. 245 – 6 a clear allusion to Achilles’ universal and eternal fame ὥς σὺ μὲν οὐδὲ θανὼν ὄνομ᾽ ὤλεσας, ἀλλά τοι αἰεὶ πάντας ἐπ᾽ ἀνθρώπους κλέος ἔσσεται ἐσθλόν, ᾿Aχιλλεῦ Thus not even in death did you lose your name; no, forever to all the people will your kleos be noble, Achilles. Od. 24.93 – 4

Archaic Epigram and the Seal of Theognis

211

(Note that πάντας ἐπ᾽ ἀνθρώπους is probably an intertext of πάντας δὲ κατ’ ἀνθρώπους ὀνομαστός, the assertion of Theognis’s fame in the sphrēgis poem.) But these lines are addressed to Achilles in Hades after his death and in the context of a description of his conspicuous tomb on the shore of the Hellespont. Theognis’ written elegiac corpus promises to be just such a tomb for Kurnos, a monument of Panhellenic scope and renown. The elegy provides clear references to funerary epigram. κείμενος (l. 240) means being on the lips of many symposiasts, but cannot fail to evoke the idea of lying buried, and hence being on the lips of all the readers of the epigram on one’s stele.³⁷ And in addition to the physical details of death in 243 – 4 we notice the epigrammatic indication of monumental eternity in 252 ὄφρ’ ἂν γῆ τε καὶ ἠέλιος, which is reminiscent of the Midas epigram, possibly in direct allusion. However, the poem’s addressee does not seem to be interested in soaring over the high seas in the eternal life provided by poetry. The two last lines of the poem contrast a desirable situation in the future with a less desirable situation in the present (note that the sphrēgis poem does the same). Against the backdrop of the metapoetic promises about eternal renown and panhellenic fame the fickle and unpredictable behavior of the speaker’s ἐρώμενος can be seen in terms of anxiety about textual fluidity and variance, the kind of thing that can happen, as we saw, to “delapidarized” epigram. But conversely, from the point of view of the fickle youth addressed, this is the stance that equates everlasting fame with death, poetry with epigram, and inscribed, speech-endowed stones with the silence of the grave. The Theognis collection contains, among other reflections on death, a speaking voice that can be read as the expression of this idea, and thus as Kurnos’ answer to the fame elegy: ἥβῃ τερπόμενος παίζω· δηρὸν γὰρ ἔνερθεν γῆς ὀλέσας ψυχὴν κείσομαι ὥστε λίθος ἄφθογγος, λείψω δ ᾿ ἐρατὸν φάος ἠελίοιο, ἔμπης δ᾿ ἐσθλὸς ἐὼν ὄψομαι οὐδὲν ἔτι. Enjoying my youth, I romp and play happily. There is time enough for me to lie under the earth, deprived of my soul, speechless like stone // just like an unspeaking stone, as I leave the lovely light of the sun, and noble as I may be, I will not be able to see. Thgn. 567– 70

This is a sympotic anti-epigram, a denial of the tomb and a carpe diem – style celebration of life over death; stone is not just introduced as a benchmark of

 For κεῖμαι in sepulchral epigrams, see, e. g., CEG  (κεῖμαι τε̃ιδε θανόν), , , ; cf. κείμεθα in Simonides’ Thermopylae epigram (above) and Thgn.  κείσομαι (below).

212

Egbert J. Bakker

voicelessness (“speechless like stone”), but also, pointedly, as critique of the αὐδὴ τεχνήεσσα of the inscribed stele (“like a speechless stone”), whose muteness under the eternal sun is the silence of death. Inscribed epigram thus serves in a number of ways as a model for the body of written poetry bearing the name of Theognis. Its inherent writtenness, thingness, purported permanence, and the possibility to channel the fame of its commissioner provides a natural template for the notion of sphragis, an author’s signature on a work thought to have reached immutable permanence in being written. Equally important as the name-recognition of Theognis himself is Kurnos’ fame, which is presented as the kleos conferred by poetry. From the perspective of the body of Theognidean poetry, which is highly conscious of its own writtenness, poetic fame is conceived as a monument, a tomb, and Theognis’s elegiac poetry as equivalent to an epigram. But Theognidean poetry does participate in the programmatic poetic critique of inscribed monuments and their purported everlasting durability. It does so in the voice of the symposiast who prefers words and song to stone, and whose celebration of life is tantamount to the rejection, not only of the tomb itself, but also of what is written on it.

Abbreviations CEG P.A. Hansen (ed.) (1983), Carmina Epigraphica Graeca, Berlin.

Bibliography Bakker, E.J. (1997), ‘Verbal Aspect and Mimetic Description in Thucydides’, in: E.J. Bakker (ed.), Grammar as Interpretation: Greek Literature in its Linguistic Contexts, Leiden, 7 – 54. — (2006), ‘Contract and Design: Thucydides’ Writing’, in: A. Rengakos / A. Tsakmakis (eds.), Brill’s Companion to Thucydides, Leiden, 109 – 29. — (2007), ‘Time, Tense, and Thucydides”, in: CW 100, 113 – 22. — (2010), “Pragmatics: Speech and Text’, in: E.J. Bakker (ed.), A Companion to the Ancient Greek Language, Malden, 151 – 67. — (forthcoming), ‘Trust and Fame: The Seal of Theognis’, in: E.J. Bakker (ed.), Authorship and Greek Song: Questions of Authority, Authenticity, and Performance, Leiden. Baumbach, M. (2000), “‘Wanderer, kommst du nach Sparta …”: Zur Rezeption eines Simonides-Epigramms’, in: Poetica 32, 1 – 32. Baumbach, M. / A. Petrovic / I. Petrovic (eds.) (2010), Archaic and Classical Greek Epigram, Cambridge.

Archaic Epigram and the Seal of Theognis

213

Bowie, E. 1997. ‘The Theognidea: A Step Towards a Collection of Fragments?’, in: G.W. Most (ed.), Collecting Fragments–Fragmente Sammeln, Göttingen, 53 – 66. Burzachechi, M. (1962), ‘Oggetti parlanti nelle epigrafi greche’, in: Epigraphica 24, 3 – 54. Colesanti, G. (2011) Questioni teognidee. La genesi simposiale di un corpus di elegie, Rome. Day, J.W. (2010), Archaic Greek Epigram and Dedication, Cambridge. Edmunds, L. (1997), ‘The Seal of Theognis’, in: L. Edmunds / R.W. Wallace (eds.), Poet, Public, and Performance in Ancient Greece, Baltimore, 29 – 48. Ford, A. (1985), ‘The Seal of Theognis: The Politics of Authorship in Archaic Greece’, in: T.J. Figueira / G. Nagy (eds.), Theognis of Megara: Poetry and the Polis, Baltimore, 82 – 95. — (1992), Homer: The Poetry of the Past, Ithaca. — (2002), The Origins of Criticism: Literary Culture and Poetic Theory in Classical Greece, Princeton. Gell, A. (1998), Art and Agency: An Anthropological Theory, Oxford. Grethlein, J. (2008), ‘Memory and Material Objects in the Iliad and the Odyssey’, in: JHS 128, 27 – 51. Peek, W. (1955), Griechische Vers-Inschriften: Grab-Epigramme, Berlin (repr. Chicago 1988). Redfield, J. (1975), Nature and Culture in the Iliad: The Tragedy of Hector, Durham. Selle, H. (2008), Theognis und die Theognidea, Berlin. Steiner, D.T. (2001), Images in Mind: Statues in Archaic and Classical Greek Literature and Thought, Princeton. Svenbro, J. (1993), Phrasikleia: An Anthropology of Reading in Ancient Greece, Ithaca. Trümpy, C. (2010), ‘Observations on the Dedicatory and Sepulchral Epigrams, and their Early History’, in: Baumbach / Petrovic / Petrovic, 167 – 79. Tsagalis, C.C. (2008), Inscribing Sorrow: Fourth-Century Attic Funerary Epigrams, Berlin. Tueller, M.A. (2010), “The Passer-By in Archaic and Classical Epigram”, in: Baumbach / Petrovic / Petrovic, 42 – 60. Vestrheim, G. (2010), ‘Voice in Sepulchral Epigrams: Some Remarks on the First an Second Person in Sepulchral Epigrams, and a Comparison with Lyric Poetry’, in: Baumbach / Petrovic / Petrovic, 61 – 78. Wachter, R. (2010), ‘The Origin of Epigrams on “Speaking Objects”’, in: Baumbach / Petrovic / Petrovic, 250 – 60. West, M.L. (1974), ‘On the History of the Theognidean Sylloge’, in: Studies in Greek Elegy and Iambus, Berlin, 40 – 71. Woodbury, L. (1952), ‘The Seal of Theognis’, in: Studies in Honour of Gilbert Norwood, Toronto, 20 – 41.

Michael A. Tueller

Words for Dying in Sepulchral Epigram A grave inscription can fulfill its function very simply; in the U.S. the stereotypical grave marker is inscribed “here lies”, followed by the name of the deceased. Nothing more is needed to know that the person is dead. From the beginning, however, Greek sepulchral epigram tended to be more explicit. Throughout antiquity, about half of Greek sepulchral epigrams—both inscribed and literary— tended to include some verb, or at least a metaphorical phrase, meaning “die” or “kill.”¹ While euphemism was always possible, it was rarely desirable: the trend, rather, was toward more narrative epigrams, which tended to increase the attention paid to the death itself. This paper will examine the words, and a few longer locutions, used to denote death in sepulchral epigram. Due to the great number of different expressions involved, this paper does not attempt an exhaustive catalog: the rarest cases it passes over.² We will trace the major developments in the use of these expressions from their earliest beginnings in inscribed verse epigram through the end of the period embraced by the Greek Anthology.

1. ὄλλυμι The oldest epigrams express death with the word ὄλλυμι and its compounds. The primary function of this word is to express death—a function it performs independently of the grammatical construct in which it is employed. A few examples should suffice to prove the point:

 Even when there is no explicit word for death, sepulchral epigram has many ways of clearly indicating that it memorializes a dead person. Eventually, the verb κεῖμαι (much like the stereotypical English expression “here lies”) became sufficient; many epigrams also mention the presence of a corpse (νεκρός/νέκυς), euphemize the death with πάσχω, or give a brief narrative of how the death occurred. Although these indicators of death are worthy of study in their own right, this paper focuses only on words and phrases meaning, or equivalent to, simply “die” or “kill.”  The most significant omitted expressions include: words for taking, usually involving θάνατος or underworld divinities or places; words for burial used to mean death (i. e. the death assimilated to the burial); words meaning “fall” as a metaphor for “die”; words for snuffing out and loss of breath (except in one case); and verbs of completion (e. g. “completed the course of life”). As will be noted in the conclusion, some other metaphorical expressions receive somewhat light treatment.

216

Michael A. Tueller

στε͂θι καὶ οἴκτιρον Κροίσο | παρὰ σε͂μα θανόντος hόν | ποτ᾽ ἐνὶ προμάχοις ὄλεσε | θο͂ρος Ἄρες. Stand and take pity by the tomb of dead Croesus, whom furious Ares once killed in the front lines. CEG 27, c. 540 – 30? BC, Attica παρι|ὼν στῆ|θί τε κἀ|ποίκτιρον· | στήλη καὶ τό|δε σῆμα θυ|γατρός, ἣ προ|λιπο͂σ᾽ ἥβης ἄν|θος πάτροθε|ν ὤλετο μονο|γενής. As you pass by, stop and take pity. This is the pillar and marker of a daughter, who left the flower of her youth when she died, her father’s only child. CEG 174B, c. 475 – 450 BC, Sinope [εἴτε ἀστό]ς τις ἀνὲρ εἴτε χσένος | ἄλοθεν ἐλθὸν Τέτιχον οἰκτίρα|ς ἄνδρ᾽ ἀγαθὸν παρίτο, ἐν πολέμοι | φθίμενον, νεαρὰν hέβεν ὀλέσαν|τα. ταῦτ᾽ ἀποδυράμενοι νε͂σθε ἐπ|ὶ πρᾶγμ᾽ ἀγαθόν. Whether a townsman, or a stranger coming from elsewhere, one must pass by after pitying Tetichus, a good man, who died in war, losing the prime of his youthful life. After having pronounced these laments, go forth for a good work. CEG 13, c. 575 – 550? BC, Attica τλέμονες, hοῖον ἀγο͂να μάχες τελέσαντες ἀέλπ[το] | φσυχὰς δαιμονίος ὀλέσατ᾽ ἐμ πολέμοι, . . . . O wretches, what a hopeless battle you prosecuted when you lost your blessed lives in war …. CEG 5(ii).1– 2, 447/446 BC, Attica Γναθίο, | το͂ σφυχὲ ὄλετ᾽ ἐ[ν δαΐ], | hιερός εἰμι | το͂ hεροιάδο. Of Gnathias, whose life was lost in battle, I am sacred, the son of Heroiades. CEG 47(i).3 – 4, c. 525 – 500? BC, Attica

These inscribed epigrams all date from near or before 450 BC. According to Liddell and Scott,³ we have a few different meanings of the word here—active meanings “to destroy” and “to lose”, and middle meanings that function more like passives than usual, meaning “to perish” and “to be lost.” Certainly when we translate ὄλλυμι we must take these definitions into account, but they really say very little about how ὄλλυμι works. After all, Greek would not, in this period, use the word to denote the loss or demise of just anything: the only thing that can be lost or perish with this word is some equivalent for the life or the person  LSJ ὄλλυμι.

Words for Dying in Sepulchral Epigram

217

himself. We must understand, then, that the word means death; its actual translation is secondary. In the active meaning “destroy” the word also has a very limited set of subjects. At first the most common subject is the one seen in CEG 27 above—Ares.⁴ Only divine beings or notionally divine forces may play this role. Other epigrams, some of which may be quite early, assign the action to Fate or Time.⁵ Such abstract divinities, however, can be difficult to separate from human action, or even from the deceased himself. An epigram on Philitas cited by Athenaeus (Anonymous 134 FGE) uses ὤλεσε to say that Philitas died a victim of his own thoughts and words; another (Anacreon, AP 7.263) uses the word to tell of a sailor slain by his own πόθος—his longing to return home. In other words, these two both died as a result of their own personality defects: divine forces have become highly personalized. At other times the divine forces that defeat the deceased are more easily envisioned. The earliest use of ὄλλυμι in sepulchral epigram is the following: Δϝεινία τόδε [σᾶ|μα], τὸν ὄλεσε π|όντος ἀναι[δές]. This is the marker of Deinias, whom the shameless sea destroyed. CEG 132, c. 650? BC, Corinth

Here the sea steps in to play the fatal divinity. The sea and its components—wind and wave, and the like, continued to be popular as subjects of ὄλλυμι in the Hellenistic period.⁶ This theme was well enough known that it could be played ironically. Antipater of Thessalonica (AP 7.625) tells of a sailor who, though tested by many a sea, is slain by the shallows (ἆ πόσον ὕδωρ / ὤλεσε τὸν τόσσῳ κεκριμένον πελάγει) when he takes an apparently drunken tumble from the prow of his moored ship. In the 6th century CE, Julian the Prefect brings both strands together (AP 7.586.1– 2). On the one hand, he denies that sea and wind destroyed (ὄλεσσε) the deceased; but then he blames a different abstract divinity—ἔρως. And this is not even the usual ἔρως, but rather a lust for profit. The intransitive use of ὄλλυμι, meaning “to die”, traveled a different path. Perses, one of the very earliest hellenistic epigrammatists, innovates here in a way that has gone unnoticed: ὤλεο δὴ πρὸ γάμοιο, Φιλαίνιον, οὐδέ σε μάτηρ Πυθιὰς ὡραίους ἄγαγεν εἰς θαλάμους

 It is also seen in CEG .  E. g. AP ., .  E. g. AP ., , , , .

218

Michael A. Tueller

νυμφίου· ἀλλ᾽ ἐλεεινὰ καταδρύψασα παρειὰς τεσσαρακαιδεκέτιν τῷδ᾽ ἐκάλυψε τάφῳ. You perished before your marriage, Philaenium, and your mother Pythias did not bring you in your prime to the chamber of your bridegroom; but piteously she gouged her cheeks and veiled you with this tomb at the age of fourteen. Perses, AP 7.487

The epigram begins with ὤλεο. This may not seem much different from earlier epigrammatic appearances of the 2nd aorist middle of ὄλλυμι, such as the ὤλετο found in CEG 174B above. The difference, however, is significant. Prior to Perses, that precise form had, as far as we know, only been used once in Greek, in Iliad 24.725—the first line of the first lament for Hector, the capstone of the Iliad. Perses thus apostrophizes the deceased, and brings her case vividly to the mind, while at the same time ennobling her by an allusion to a great lament. It is also worth noting that the allusion is specifically to funerary lament, not its echoes in the mouth of later visitors to a tomb. The remoteness of a tomb thus slips away, and the reader is plunged into the immediacy of the funeral. And who is the person so honored? A fourteen-year-old girl. On the other hand, perhaps this allusion does not elevate the deceased, but rather depreciates the poetic lament. Anyte implies as much in her epigram 10 HE, when she applies ὤλεο, again prominently positioned as the first word in the epigram, to a dog—just a puppy—who had died of snakebite. From then on, the form ὤλεο and its compounds become a commonplace in sepulchral epigram, both serious and ironic, both book-epigram and real inscribed epitaphs.⁷

2. φθίω The next verb for death to appear in the inscriptional record is φθίω; its earliest attestation appears in this epigram (which was also cited above for its use of ὄλλυμι): [εἴτε ἀστό]ς τις ἀνὲρ εἴτε χσένος | ἄλοθεν ἐλθὸν Τέτιχον οἰκτίρα|ς ἄνδρ᾽ ἀγαθὸν παρίτο, ἐν πολέμοι | φθίμενον, νεαρὰν hέβεν ὀλέσαν|τα. ταῦτ᾽ ἀποδυράμενοι νε͂σθε ἐπ|ὶ πρᾶγμ᾽ ἀγαθόν.

 Strangely, other than in epigram, the word’s only attestations in all of Greek literature remain the Iliad line and specific citations of it.

Words for Dying in Sepulchral Epigram

219

Whether a townsman, or a stranger coming from elsewhere, one must pass by after pitying Tetichus, a good man, who died in war, losing the prime of his youthful life. After having pronounced these laments, go forth for a good work. CEG 13, c. 575 – 550? BC, Attica

Φθίω and its compounds invariably appear in the aorist middle. The great majority of the time, throughout the history of Greek epigram, it takes participial form, and subordinates itself to other ideas in the epigram. (In CEG 13, for instance, both φθίμενον and ὀλέσαντα are subordinate to the request for mourning in the first couplet.) The following is typical: παιδὸ̣ς ἀ̣π̣οφ ̣ θιμένοιο Κλ̣εο̣ί ̣το το͂ Μεν|εσαίχμο μνε͂μ᾽ ἐσορο͂ν οἴκτιρ᾽ὸς καλὸς | ὂν ἔθανε. Take pity as you look on the marker of Cleoetus (deceased), son of Menesaechmus. He was good as he died. CEG 68, c. 500? BC, Attica

Here the participle is closely associated both with the idea of the μνῆμα and with grief for the deceased. This is not saying much. In fact, the participle φθίμενος is nearly always redundant in inscribed sepulchral epigram; the epigrams almost always give other indications that they are about a dead person, either by directly stating that death has occurred, or, more commonly, by a mention of a μνῆμα for the person or by indications of lament. Why, then, is the word used so frequently, appearing 30 times in the Carmina Epigraphica Graeca collection?⁸ We may suspect an epic borrowing here; the word is common in Homer, including many uses of the participial form. The participle is found most memorably in two statements made by Achilles:⁹ νῦν δ’ ἵνα καὶ σοὶ πένθος ἐνὶ φρεσὶ μυρίον εἴη παιδὸς ἀποφθιμένοιο, τὸν οὐχ ὑποδέξεαι αὖτις οἴκαδε νοστήσαντ’, ἐπεὶ οὐδ’ ἐμὲ θυμὸς ἄνωγε ζώειν οὐδ’ ἄνδρεσσι μετέμμεναι … .

 In addition to these  participial uses, there is one appearance of the finite verb κατέφθιτο (CEG ), on which see below.  Gregory Nagy (, chapter ) analyzes the linguistic characteristics of the word, and connects it to fertility. In accordance with his scheme, the prominence of the word in epigram may also indicate that the memorial is an artificial grant of immortality when natural mortality has caused the body to wither. This reasoning, however, cannot explain the extraordinary prominence of the participle above all other forms of the word.

220

Michael A. Tueller

But it turned out that you too would have infinite grief in your heart for your dead son, whom you will never again welcome home after his return journey, since my heart does not command me to live or to be among men … . Il. 18.88 – 91 ἤδη γὰρ Πηλῆά γ’ ὀΐομαι ἢ κατὰ πάμπαν τεθνάμεν, ἤ που τυτθὸν ἔτι ζώοντ’ ἀκάχησθαι γήραΐ τε στυγερῷ καὶ ἐμὴν ποτιδέγμενον αἰεὶ λυγρὴν ἀγγελίην, ὅτ’ ἀποφθιμένοιο πύθηται. As for Peleus, I think that he is already quite dead—or if perhaps he is still barely alive he is in pain, both from hateful old age and from continual waiting for sad news of me, when he will learn that I am dead. Il. 19.334– 337

In these passages, as happens occasionally in the Iliad, Achilles contemplates his death; but here he does more: he imagines the reaction of his parents to his death. In the case of his father, whom Achilles cannot address directly, the situation is inevitably much like a sepulchral epigram: the family hears a message of death in the absence of the one who has died. The use of the participle in inscribed sepulchral epigram, then, likely recalls these passages to mind, and the pain involved in the death of a family member. Still, in nearly every case φθίμενος is not strictly necessary as an indicator of death; other parts of the epigram bear that semantic weight more firmly. There are exceptions, however: τέκνον ἐμῆς θυγατρὸς τόδ᾽ ἔχω φίλον, ὅμπερ ὅτε αὐγὰς ὄμμασιν ἠ|ελίο ζῶντες ἐδερκόμεθα ̃ ἐχον ἐμοῖς γόνασιν καὶ νῦν φθίμενον φθιμένη ᾽χω. I hold this dear child of my daughter. While alive I used to hold him on my knees as we beheld the sun’s rays with our eyes, and now I, deceased, hold him, deceased. CEG 89(ii), c. 410 BC, Attica

Here the strongest marker of death is the participles φθίμενον and φθιμένη. (Although the verb ἔχω, repeated three times, is an indicator of a tomb, it is not a very strong one.) This epigram is not very formulaic; its use of φθίω is carefully engineered to build to the striking image of the two embracing corpses at the end. In fact, this epigram in part establishes a new pattern for φθίμενος. It carefully differentiates two phases of existence—while the two people were alive and after they died. By marking the boundary of death, it allows the narrative of life to continue past it.¹⁰  This trend is continued by epigrams such as CEG  and AP .. See also CEG .

Words for Dying in Sepulchral Epigram

221

Narrative considerations also motivate the few occurrences of φθίω as a finite verb. The first example in sepulchral epigram is the following: παῖδά τοι ἰφθίμαν Δαμαινέτου ἅδε Kρατίσταν, | ᾿Aρχεμάχου δὲ φίλαν εὖνιν ἔδεκτο κόνις, | ἅ ποθ’ ὑπ’ ὠδίνων στονόεντι κατέφθιτο πότμωι, | ὀρφανὸν ἐμ μεγάροις παῖδα λιποῦσα πόσει. This dust received Cratista, the comely child of Damaenetus, and the dear wife of Archemachus. She met her fate in the groans of childbirth, leaving an orphaned child at home for her husband. CEG 576, c. 350? BC, Attica

The first couplet here has given all the necessary information, but the second gives a brief story of Cratista’s death in childbirth. The same is true for all later uses of a finite verb φθίω: they are never simple statements that death has occurred, but always occur in narratives with some level of detail about the death.¹¹ Occasionally the participle will play this role as well.¹² Over time however (as we will see below), θνῄσκω became the default verb for use in death narratives; φθίω, already largely redundant, declined in prominence. Some examples of its use are found in literary epigram, but they do not innovate beyond the inscribed examples already cited. In the Hellenistic period it was gradually abandoned. While it appears twice in Anyte’s small oeuvre and four times in Leonidas of Tarentum, Callimachus uses it only once, and most of his contemporaries not at all. A third of its appearances in the Garland of Meleager are in poems that are attributed, whether rightly or wrongly, to pre-hellenistic poets, or in anonymous poems that give the appearance of great age. While Antipater of Sidon uses the word four times, Meleager uses it only once, and its decline continues from there. In the entire Cycle of Agathias it occurs only five times.

3. θνῄσκω No other word indicates death more clearly than θνῄσκω. It makes its appearance at essentially the same time as φθίω, and in very much the same way:¹³

 AP ., ; ..  E. g. AP ..  This example has already been seen above, as an example of the use of ὄλλυμι.

222

Michael A. Tueller

στε͂θι καὶ οἴκτιρον Κροίσο | παρὰ σε͂μα θανόντος hόν | ποτ᾽ ἐνὶ προμάχοις ὄλεσε | θο͂ρος Ἄρες. Stand and take pity by the tomb of dead Croesus, whom furious Ares once killed in the front lines. CEG 27, c. 540 – 30? BC, Attica

Here we see the genitive participle θανόντος, much as we saw ἀποφθιμένοιο in CEG 68, and it is again closely associated with the σῆμα and with grief. Again, it is redundant—if it were omitted, it would not be missed. Unlike φθίω, however, θνῄσκω soon begins to appear frequently as a finite verb. The earliest example is in CEG 68 itself, which is worth a second look: παιδὸ̣ς ἀ̣π̣ο̣φθιμένοιο Κλ̣εο̣ί ̣το το͂ Μεν|εσαίχμο μνε͂μ᾽ ἐσορο͂ν οἴκτιρ᾽ὸς καλὸς | ὂν ἔθανε. Take pity as you look on the marker of Cleoetus (deceased), son of Menesaechmus. He was good as he died. CEG 68, c. 500? BC, Attica

The word ἔθανε here is typical of other early uses. While it is grammatically primary in its clause, the primary focus of the clause is not on it: it is on Cleoetus’s goodness as he died. While goodness is very generic, epigrams such as the following are more typical: Προκλείδα{ς} τ̣όδ̣ε σᾶμα ̣ κεκλ|έσεται ἐνγὺς ὁδοῖο, | hὸς περὶ τᾶς αὐ̣το͂ γᾶς | θάνε μα⟨ρ⟩νάμενος. This will be called the marker of Procleidas near the road. He died battling in the vicinity of his land. CEG 142, c. 475 – 450? BC, Acarnania

This epigram reports a more specific kind of goodness: the deceased died in battle. Again, although θάνε is the finite verb, the emphasis is clearly on the combat rather than the death. Both of these two uses—in which θνῄσκω, whether as participle or finite verb, is subordinate to other ideas—continue with very little change or diminution through to the Cycle of Agathias, though many other uses are added to them. One interesting variation may be found in AP 7.720, where Chaeremon replaces the participle “battling” with a phrase that indicates the results of the battle—redrawing borders. As Gow and Page point out, it is possible that Chaeremon

Words for Dying in Sepulchral Epigram

223

makes this adaptation in order to emphasize that the deceased was buried in the very land he had helped to acquire.¹⁴ In the wake of the Persian Wars, θνῄσκω began to expand its repertoire of functions: ἐνθάδ᾽ ἄνωρος | ἐὼν ἔθανον, | θῆκαν δέ μ᾽ ἑταῖ|ροι νόσφι τοκέ|[ων]· τῆλ᾽ ὢ πατρίδ|[ο]ς ἡμετέρης κεῖ|[μ]αι ἄλαστα π⟨α⟩θὼ|ν ἠέλιον προλι|πών, οὐδὲ τοκέω|ν ἐγ χερσὶ θανὼν | ᾿Aΐ{ι}δην θεραπεύω. αὐτὰρ ἔμοιγ᾽ ὄν|υμ᾽ ἐστὶ Κόβων Σάκ|εω υἱός. ἁπάντων πρ|εσχύτατος παίδων ἐς ἔ|ρεβος κατέβην. I died an untimely death and my companions placed me here, far from my parents. Oh! I lie far from our homeland; I left the light of the sun after suffering unforgettably; I serve Hades without even dying in my parents’ arms! In any case, my name is Cobon, son of Saces; I, the oldest of all their children, descended into the darkness. CEG 171, c. 475 – 450? BC, Egypt

Here we see both the finite verb and the participle of θνῄσκω closely attached to statements that are clearly intended to generate an emotional response in the reader. The most common such statements are the two seen here: dying too young, and dying far from one’s home and friends.¹⁵ While it might seem possible to connect these sentiments to verbs such as φθίω and ὄλλυμι, in fact they tend strongly to occur with θνῄσκω. And we do not find only negative sentiments: we also find exactly the reverse, a death that is fortunate because it occurred after a long full life, or in the arms of friends.¹⁶ The most unusual feature of θνῄσκω is the weighty implications it attaches to tense. In the corpus of archaic and classical inscribed epigram, time is governed by fairly strict rules: present time can be used only of things that are present at the time and place of reading. In the epigrams of mid-fourth century Attica, however, θνῄσκω begins to show the one significant exception to this rule: it may appear in the present tense even when its intended time is clearly past:¹⁷ ἐννέα ἐτῶν ἐβίων δεκάδας, θνείσκω δὲ γεραιός, | σωφροσύνην δὲ ἤσκησα, ἔλιπον δὲ εὔκλειαν ἀμεμφῆ.

   

Gow / Page , .. Dying too young: CEG , . Dying far from home/friends: CEG (ii). Dying at a good old age: CEG , . Dying among friends: CEG . There are some other, insignificant, exceptions (Tueller , ).

224

Michael A. Tueller

I lived nine decades, and I am dying an old man. I achieved prudence, and left a blameless fame. CEG 531(ii), c. 350? BC, Attica

All of the verbs here are set in the past, and all are aorist except for θνείσκω, which is in the present despite the fact that past time is its clear setting. It seems likely that this peculiar exception is made in order to bring the death more vividly to the mind of the reader. This phenomenon is almost exclusively a feature of inscribed epigram, and of the classical period. Only a few authors of the Greek Anthology seem to have grasped it. Leonidas once invokes it, in AP 7.163, and then it lies dormant until Meleager, who was a close observer and imitator of even distant predecessors. Meleager only uses θνῄσκω in the present tense; in the following example, he shows that he knows exactly what he is doing: τὸν ταχύπουν, ἔτι παῖδα συναρπασθέντα τεκούσης ἄρτι μ᾽ ἀπὸ στέρνων, οὐατόεντα λαγὼν ἐν κόλποις στέργουσα διέτρεφεν ἁ γλυκερόχρως Φανίον, εἰαρινοῖς ἄνθεσι βοσκόμενον. οὐδέ με μητρὸς ἔτ᾽ εἶχε πόθος, θνῄσκω δ᾽ ὑπὸ θοίνης ἀπλήστου πολλῇ δαιτὶ παχυνόμενος. καί μου πρὸς κλισίαις κρύψεν νέκυν, ὡς ἐν ὀνείροις αἰὲν ὁρᾶν κοίτης γειτονέοντα τάφον. I was a swift-footed, long-eared rabbit, still a child, just snatched away from my mother’s breast; Phanion, with her sweet skin, loved and nurtured me in her lap, and fed me with spring flowers. I was no longer in the grip of longing for my mother, but I am dying of obesity, from overeating at her great buffet. She hid my corpse by her couch, to see in her dreams my grave ever serving as neighbor to her bed. Meleager, AP 7.207

Just as in CEG 531, we see here a series of verbs that all refer to the past, and they are all in true past tenses except θνῄσκω. After Meleager, the phenomenon rapidly dies out.¹⁸ The perfect tense also has a particular significance: ἄσβεστον κλέος οἵδε φίλῃ περὶ πατρίδι θέντες κυάνεον θανάτου ἀμφεβάλοντο νέφος· οὐδὲ τεθνᾶσι θανόντες, ἐπεί σφ᾽ ἀρετὴ καθύπερθε κυδαίνουσ᾽ ἀνάγει δώματος ἐξ ᾿Aΐδεω.

 The only later example I find is Diogenes Laertius, AP ..

Words for Dying in Sepulchral Epigram

225

These men wrapped their dear homeland in unquenchable glory, and donned the dark cloud of death. Though they died, they are not dead, since their outstanding excellence retrieves them from the house of Hades. Simonides, AP 7.251

Though the dating of this epigram, as with all epigrams assigned to Simonides, cannot be completely certain, it seems likely that it should be attached to a Persian War battle, possibly Plataea.¹⁹ This date would make it the earliest use of the perfect of θνῄσκω in epigram, and it is typical of the trend that follows. Here, τεθνᾶσι, as we would expect, represents the permanence of death, as opposed to the simple fact of dying, which we see in θανόντες. Epigrams that use the perfect nearly always make this focus explicit, usually by comparing the state of being dead to the state of being alive, either with some kind of hope, as here, or with resignation.²⁰ This trend lasts through the Garland of Meleager; in the Garland of Philip, it has dissipated, and in epigrams of later generations, the perfect and the aorist of the word lose all distinction, and can freely be substituted for one another.²¹ It is also with the Garland of Meleager that θνῄσκω undergoes the change that is most significant for its continued survival in epigram—the reason that it persists where φθίω drops away. This change may be seen in an epigram by Leonidas of Tarentum, AP 7.295. The emotional effects of this epigram will be discussed below; here it is only necessary to note that its use of ἔθαν(ε) in line 7 is utterly unremarkable: the word serves no function beyond moving forward the narrative of the death. Θνῄσκω at this point broke free of the connections discussed above, and became the ordinary word used to say “he died” when the story called for it. Callimachus (see below) would complain about the word’s constant employment, but its sheer usefulness won out.

4. κτείνω Given that death in battle was the fate most deserving of memorialization in epigram, we would expect that κτείνω would be prominent. In fact, it makes its appearance relatively late, and tentatively. We first see it in a famous epigram from the Persian War:

 Page , .  See AP ., ; ..  In AP ., , ,  the perfect is used with aorist force; AP . uses the aorist with perfect force.

226

Michael A. Tueller

μνῆμα τόδε κλεινοῖο Μεγιστία, ὅν ποτε Μῆδοι Σπερχειὸν ποταμὸν κτεῖναν ἀμειψάμενοι, μάντιος, ὃς τότε κῆρας ἐπερχομένας σάφα εἰδὼς οὐκ ἔτλη Σπάρτης ἡγεμόνας προλιπεῖν. This is the marker of famed Megistias, whom the Medes once killed after crossing the river Spercheius. He was a seer, and though he knew his doom clearly at the time, he did not dare to leave Sparta’s leaders behind. Simonides, AP 7.677

We see immediately how κτείνω differs from ὄλλυμι, the other word we have seen so far that can take the meaning “to kill.” Whereas ὄλλυμι ennobled death in battle as the work of Ares or Fate, here the agent is the Persians. The fact that Megistias died at their hands does nothing to ennoble his death; we must assume that Simonides composed the epigram in the way that he did because he was keen to name the Persians—that is, to lay the blame for Megistias’s death on them. Certainly, given the connotations of ὄλλυμι that we have just seen, he could not make the Persians the subject of ὤλεσαν. The verb κτείνω thus becomes a logical choice— and its first appearance in epigram is an accusation, nested in a participial phrase that shows the Persians transgressing boundaries. Κτείνω proved its usefulness here, but Simonides may have been a bit too far ahead of his time. The word may have seemed, to its early readers, to be too pedestrian, though Herodotus, it is true, has no complaint, and believes that Simonides was acting out of good will.²² In any case, κτείνω does not appear in epigram again for more than a hundred years; then Aristotle uses it in an epigram for his friend Hermias, who was killed in 342: τόνδε ποτ᾽ οὐχ ὁσίως παραβὰς μακάρων θέμιν ἁγνὴν ἔκτεινεν Περσῶν τοξοφόρων βασιλεύς, οὐ φανερᾷ λόγχῃ φονίοις ἐν ἀγῶσι κρατήσας, ἀλλ᾽ ἀνδρὸς πίστει χρησάμενος δολίου. This man was once, impiously, in transgression of the pure ordinance of the blessed ones, killed by the king of the bow-bearing Persians. He did not beat him openly with a spear in murderous contests, but employing the pledge of a treacherous man. Aristotle 1 FGE

Here it is even more clear that the context is accusation. Again it is a Persian who does the killing, and he does so in violation of almost all that is holy. Κτείνω then goes on to be quite popular in hellenistic epigram, but it does not lose its connotation. Sometimes its subject is simply enemies, but its most  Hdt. ..

Words for Dying in Sepulchral Epigram

227

common subject is murderers, though predatory animals, cruel family members, pirates, and the like also make an appearance.²³ The two verbs, ὄλλυμι and κτείνω, seem to have divided up the work of killing: ὄλλυμι covered divine, often indirect or obscure, causation; κτείνω covered the efficient cause. In that light, the few times that divinities are the subject of κτείνω take on additional nuance. In AP 7.617, for instance, Zeus kills Orpheus—but he does not do so indirectly; rather, he sends a lightning bolt that strikes him down. Even more striking is the following: σῆμα τόδ᾽ Αἴαντος Τελαμωνίου, ὃν κτάνε Μοῖρα, αὐτοῦ χρησαμένα καὶ χερὶ καὶ ξίφεϊ. οὐδὲ γὰρ ἐν θνητοῖσι δυνήσατο καὶ μεμαυῖα εὑρέμεναι Κλωθὼ τῷδ᾽ ἕτερον φονέα. This is the marker of Ajax son of Telamon, whom Fate killed, employing his own hand and sword. Though she was eager, Clotho could not find anyone else among mortals to be his murderer. Anonymous, AP 7.148

Here Fate kills Ajax. The Fates would seem to be the paradigmatic case of divine agency, for which ὄλλυμι would be warranted—and elsewhere Fate is found as the subject of ὤλεσεν.²⁴ But Ajax’s death presents problems. In fact, the second couplet of the epigram can be seen as a justification, not simply for the fact that Ajax’s death was a suicide, but also for the use of κτείνω in the first couplet. Fate is not literally the direct cause of Ajax’s death: Ajax himself was. But nevertheless Fate took more direct action then usual. Why? Because no other efficient cause could be found. The second couplet implies that the subject of κτείνω should ordinarily be sought among mortals—only when this is impossible must the gods move on their own.

5. Euphemism and metaphor So far, we have looked at words that clearly mean death. But poets could take a more circuitous route to the subject. The territory of death metaphor is vast, and it will be impossible here to make more than a cursory exploration of it. Nevertheless, a few significant landmarks deserve notice.

 Murderers: AP .,  – , . Predatory and venomous animals: AP ., , , . Family members: AP ., , . Thieves and pirates: AP ., .  AP ..

228

Michael A. Tueller

Epigram employs phrases on the order of “descending into the darkness” and “leaving the light of the sun” as early as the second quarter of the 5th century.²⁵ These travel metaphors are ineffective as euphemisms, since they obscure nothing. “The darkness” is clearly Hades, and it is often labeled as such in the very same poem. This implies that soothing tender feelings is not the motivation for their use. Simonides, however, does seem to have consolatory intentions in his adaptation of this metaphor in AP 7.251, which was seen above. There the darkness, rather than being a domain one enters, is more similar to clothing. Even though death is named explicitly—the men “don the dark cloud of death”—Simonides consistently aims to diminish its effects. He does this by counterbalancing the cloud of death with an opposite clothing in the first line; the men also wrap their homeland in glory. In the second couplet, Simonides’ consolatory purpose becomes explicit. He states plainly that the men, though they have died, are not dead; and when he mentions travel and Hades, the direction is the opposite of what one would expect for a sepulchral epigram: the excellence of these men rises from Hades triumphant. Leonidas of Tarentum is particularly creative in his use of metaphor to speak of death. He rarely hints at a consolatory purpose, however. Rather, his metaphors are crafted to connect the metaphor of death to the surrounding narrative. Let us begin with a simple example: Θῆριν τὸν τριγέροντα, τὸν εὐάγρων ἀπὸ κύρτων ζῶντα, τὸν αἰθυίης πλείονα νηξάμενον, ἰχθυοληϊστῆρα, σαγηνέα, χηραμοδύτην, οὐχὶ πολυσκάλμου πλώτορα ναυτιλίης, ἔμπης οὔτ᾽ ᾿Aρκτοῦρος ἀπώλεσεν, οὔτε καταιγὶς ἤλασε τὰς πολλὰς τῶν ἐτέων δεκάδας· ἀλλ᾽ ἔθαν᾽ ἐν καλύβῃ σχοινίτιδι, λύχνος ὁποῖα, τῷ μακρῷ σβεσθεὶς ἐν χρόνῳ αὐτόματος. σῆμα δὲ τοῦτ᾽ οὐ παῖδες ἐφήρμοσαν οὐδ᾽ ὁμόλεκτρος, ἀλλὰ συνεργατίνης ἰχθυβόλων θίασος. Theris, older than old, made his living from the success of his lobster-traps, and swam more than a gull; a fish-raider, a netsman, a scallop-diver, he was unaccustomed to many-oared navigation. And yet Arcturus did not destroy him, neither did a squall drive away his many decades, but he died in a reed hut, like a lamp, snuffed out of natural causes—by the long duration of his life. This marker was constructed not by his children or his wife, but by the assembled guild of fishermen. Leonidas of Tarentum, AP 7.295

 The earliest example of both is found in CEG , cited above.

Words for Dying in Sepulchral Epigram

229

Leonidas spends two couplets setting up our expectations. Theris is a fisherman who has seen every variety of danger at sea. But then, the third couplet tells us, he did not die—notice that the word is a compound of ὄλλυμι—from divine forces like star or storm. Leonidas then, in the fourth couplet, abruptly shifts the scene. Now we are in a grass hut—a far more peaceful setting than we have seen so far. Inside such a hut we would expect to find a lamp, and this lamp represents the way the man died—with the gentle guttering of the flame as it exhausts the last of its fuel. His comrades in fishing then return at the end to build his tomb. One might detect some consolation here, but in fact that consolation is not a feature of the “snuffing” metaphor itself, but of the facts of the man’s death: he died at home. The metaphor of being snuffed by time, like a lamp, parallels the peaceful interior passing of Theris. Other epigrams of Leonidas are not at all reassuring. In AP 7.504, the deceased, another fisherman, reaches the end of his fated thread as he thrashes about amidst his fishing lines. In AP 7.662, a toddler tastes death. While the expression “tasting death” is not uncommon in English, it virtually never appears in Greek. Leonidas’s metaphor is thus very striking: this child, who, as we can imagine, puts everything in his mouth, does so also with death. In AP 7.273, Callaeschrus “slipped out of life” just as easily as his ship slips through the water. Later epigrammatists, while not entirely shy of deploying metaphor, tend to restrict its range much more than Leonidas. Metaphors of sleep, in fact, outnumber all other metaphors. Sleep and death are brothers. There is a family resemblance: the body is inactive when subject to either. This inaction is the theme when the euphemism of sleep is approached for the first time: Πόντου ἀπ᾽ Εὐξείνου Παφλαγὼν μεγάθυμος ᾿Aτώτας | ἧς γαίας τηλοῦ σ〚σ〛ῶμ᾽ ἀνέπαυσε πόνων. | τέχνηι δ᾽ οὔτις ἔριζε· Πυλαιμένεος δ᾽ ἀπὸ ῥίζης | εἴμ᾽, ὃς ᾿Aχιλλῆος χειρὶ δαμεὶς ἔθανεν. Great-hearted Atotas, a Paphlagonian from the Euxine Sea, rested his body from toil far from his land. No-one rivaled him in his craft. I am from the root of Pylaemenes, who died when mastered by Achilles’ hand. CEG 572(ii), after c. 350? BC, Attica

Atotas here rests his body from toil—a result that sounds entirely positive, especially for a miner, whose toils were presumably difficult indeed. In fact, given that this metaphor is applied here for the first time, the epigram runs the risk of unclarity: is Atotas actually dead, or is his shift just over? Only the fact that the epigram is inscribed leads one to conclude that the cessation of labor is permanent.

230

Michael A. Tueller

Miners generally did not enjoy any retirement; they stopped working only when they died. But if retirement is a real possibility, even a permanent end to labor could be ambiguous. Consider the following epigram: ἡ Μαραθη[νὴ τῆιδ]ε φίλας ἀνέπαυσε Κ̣ [ ̣ ̣ ̣ ]α̣ ̣ σὺν γήραι [χαλεπῶ]ν χεῖρας ἀφ᾽ ἱστοπόδ[ων, ὀγδωκοντ[αέτις μέ]ν̣, ἔτι κρέξαι δὲ λιγε[ίαι κερκίδι λε[πταλέον] στήμονα δυναμ̣έ̣νη· χαιρέτω ἐκ [καμάτ]ων ὁσίη γυνή, ἥτι ̣ς ̣ ἀ̣τ[̣ ρ]ύ̣[τ]ω̣ [ι ζωῆι θυγατέρων πέμπτον ἐπεῖδε θέρος. Marathian C…a here rested her own hands, weighed down with age, from the troublesome loom. She was eighty, but still able to strike the slender thread with the humming shuttle. A farewell from labor to the pious woman, who saw a fifth crop of daughters in her tireless life. Posidippus 45 Austin-Bastianini

This poem, which appears after Leonidas has inaugurated the tradition of retirement-epigrams, toys with the reader: this woman has put aside her shuttle, but still seems pretty spry at first. Only with the final couplet do sepulchral formulas clearly intrude. With χαιρέτω and the count of the generations she has seen, we infer that the woman must be dead. The metaphor of sleep was not always so unclear; with the proper phrasing, it could be anchored unmistakably in death. Callimachus finds a winning formula: Κρηθίδα τὴν πολύμυθον, ἐπισταμένην καλὰ παίζειν, δίζηνται Σαμίων πολλάκι θυγατέρες, ἡδίστην συνέριθον ἀείλαλον· ἡ δ᾽ ἀποβρίζει ἐνθάδε τὸν πάσαις ὕπνον ὀφειλόμενον. Crethis, full of stories, an expert player, is often sought by the daughters of the Samians— their sweetest workmate, a chatterbox. But she slumbers here the sleep due to all. Callimachus, AP 7.459

By using two words for sleep, one as the object of the other, and then following up by observing that this is “the sleep due to all”, Callimachus euphemizes without the possibility of misunderstanding.²⁶ It is no surprise that this formula, sometimes with very little variation, is adopted by later epigrammatists.²⁷

 See also his AP ., “the foreordained sleep.”  AP ., , , , .

Words for Dying in Sepulchral Epigram

231

Callimachus, for all his notorious prickliness and irony, habitually expresses himself delicately when it comes to death. We might well see him as an advocate for this kind of euphemism: τῇδε Σάων ὁ Δίκωνος ᾿Aκάνθιος ἱερὸν ὕπνον κοιμᾶται. θνῄσκειν μὴ λέγε τοὺς ἀγαθούς. Here Saon of Acanthus, son of Dicon, “sleeps the sacred sleep”: don’t say that the good “die.” Callimachus, AP 7.451

This poem has long been seen as parallel to Simonides’ consolation above (AP 7.251): Callimachus is saying that Saon, because of his goodness, will, in a sense, never really die. These sentiments are surely present, but Callimachus is doing something more here. The quotation marks I have introduced into the translation above are intended to highlight the significance of the imperative λέγε. Callimachus is making a point about our habits of speech: we should not say that good people die; we should euphemize. Again, as in AP 7.459, Callimachus’s euphemism is not susceptible to misinterpretation—“sleeping the sacred sleep” can only be death—but actual death is simply not something we should say, at least in the case of the good. Nevertheless, metaphor is often more interesting when it opens the gate to ironic ambiguity. The following poem wrenches the reader’s emotions: τὸν τριετῆ παίζοντα περὶ φρέαρ ᾿Aρχιάνακτα εἴδωλον μορφᾶς κωφὸν ἐπεσπάσατο· ἐκ δ᾽ ὕδατος τὸν παῖδα διάβροχον ἅρπασε μάτηρ σκεπτομένα ζωᾶς εἴ τινα μοῖραν ἔχει. νύμφας δ᾽ οὐκ ἐμίηνεν ὁ νήπιος, ἀλλ᾽ ἐπὶ γούνοις ματρὸς κοιμαθεὶς τὸν βαθὺν ὕπνον ἔχει. Three-year-old Archianax, playing by a cistern, was drawn in by the mute image of his form. His mother snatched him from the water soaking, uncertain whether he had any portion of life. The toddler did not pollute the nymphs, but fell asleep on his mother’s knees, and has deep sleep. Posidippus or Callimachus, AP 7.170

A toddler playing by a cistern is an image to disturb any parent. When Archianax falls, we are relieved to hear that his mother snatched him out, even though she does so with the verb ἁρπάζω, which is often used of Hades; and we too wonder if she has arrived in time. The final couplet, however, is deeply unsatisfying. The child, we read, did not pollute the nymphs—certainly, then, he did not die in the water. But does that mean that he did not die, or that he did not die there? The answer is that he “has the deep sleep on his mother’s knees.” Unlike the two Cal-

232

Michael A. Tueller

limachus epigrams above, this is not clearly “the sleep that is due to all” or “the sacred sleep”: it is “the deep sleep.” Here a great interpretive weight seems to rest on the definite article. It would be much different to say that the child “is having a deep sleep”; that could perhaps be the kind of sleep a child has subsequent to a sudden traumatic experience: first he is shocked, then he wails until he is exhausted, and then he sleeps. But “the deep sleep” seems more definite and permanent. On the other hand, the article could simply refer to the kind of sleep that is appropriate to a three-year-old; or the article might not matter at all: the simple presence of two words for sleep in succession might be enough to suggest death. Or the presence of the word “deep”, recalling the depth of the water, could imply that he drowned even after he was pulled out. Given the usually high degree of control exercised by Callimachus (or Posidippus, if the poem is his), it is unlikely that the ambiguity here is due to drifting poetic attention. More probably we see here an intentional failure of euphemism. The anxiety with which the narrative builds leads perfectly to irresolution, in which the reader feels, not only that she must know how the story ends, but that she has already been told how it ends, but just can’t quite see it clearly. This inability to see clearly when death is the subject is itself a large part of the point.

6. Conclusion This paper can give only a sampling of the ways in which sepulchral epigram spoke of death. It has barely touched on metaphors of travel, and has omitted a number of locutions involving darkness or underworld divinities, as well as a host of interactions with the noun θάνατος, among others. Nevertheless, a few conclusions can be drawn. Death, Epicurus said, is nothing to us.²⁸ He has his own reasons, but one reason is that no-one has ever told us what death is like. Despite its obvious significance, it is an empty signifier, ready to accept any meaning we put into it. This may most easily be seen in the word φθίω, which, for all its literal meaning, adds nothing to most of the epigrams in which it appears; it can be, and eventually is, omitted without loss. We can also see the emptiness of death in the fact that a death in battle can be framed in several different ways: when A kills B and the verb “kill” is ὄλλυμι, then A ennobles B. When, on the other hand, A kills B and the verb is κτείνω, then B is not the focus of our attention at all; the point is that A is blameworthy. Or yet again, when B dies in battle with the verb θνῄσκω,

 D.L. ..

Words for Dying in Sepulchral Epigram

233

his death may be so vivid as to escape the bounds of time, or it may be logically subordinated to the greatness of his deeds in life. Either way, the reader sees only his efforts, and no gods or enemies are even mentioned. And yet all of these are the same death. Leonidas takes the emptiness of death as a cue to reframe it completely with each new narrative, giving each character a death that matches the circumstances in which the death happened. For Leonidas, no two deaths are the same. Callimachus, on the other hand, finds the void of meaning in death profoundly disturbing: it is best to call it by the name of something we understand. Yes, death happens, but let us not say so: a double dose of sleep will get us through the night.

Abbreviations CEG FGE GP HE

Carmina Epigraphica Graeca. Further Greek Epigrams (Page 1981). The Garland of Philip (Gow/Page 1968). Hellenistic Epigrams (Gow/Page 1965).

Bibliography Baumbach, M. / A. Petrovic / I. Petrovic (2010), Archaic and Classical Greek Epigram, Cambridge. Gow, A.S.F. / D.L. Page (1965), The Greek Anthology: Hellenistic Epigrams, 2 vols. Cambridge. Gow, A.S.F. / D.L. Page (1968) The Greek Anthology: The Garland of Philip and Some Contemporary Epigrams, 2 vols. Cambridge. Nagy, G. (1979), The Best of the Achaeans: Concepts of the Hero in Archaic Greek Poetry, Baltimore. Page, D.L. (1981), Further Greek Epigrams, Cambridge. Tueller, M.A. (2011), “The Passer-By in Archaic and Classical Epigram”, in: Baumbach / Petrovic / Petrovic 2010, 42 – 60.

Style in Literary Epigram b) Philosophical Style

Benjamin Acosta-Hughes

A Little-Studied Dialogue: Responses to Plato in Callimachean Epigram For Marco Fantuzzi commilitoni mei

Over a decade now I have found myself, though occupied with other topics, coming back to possible echoes of Plato in Callimachus. This is usually due to reading a Plato text again and recognizing something distinctly (if distantly) familiar from Callimachean epigram, or, on careful re-reading of a Callimachean epigram, recognizing something that might be a Platonic citation, possibly varied or recast, in the context of a short poem in elegiac couplet – an epigram. At issue are particularly Callimachus’ homoerotic epigrams, and this is perhaps unsurprising – Plato composed masterful treatments on homoerotic, or more precisely pederastic love, which treat the nature of the relationship in the abstract and attempt to put forward a definition of an erotic bond that was a feature of Greek culture. In truth, the construction of pederastic love as a discursive philosophical topic may well be due to Plato, as the discussion of Achilles and Patroclus at Symp. 179e5 – 180b5. And unlike most other epigram authors of what is popularly defined as the Hellenistic era,¹ Callimachus only treats of homoerotic love in his extant erotic epigrams.² The model is a logical one, sure, not unlike Apollonius’ recollections of Herodotus in his own epic of adventure into the unknown and treatment of cultural alterity. Yet this does not quite explain the ap-

 The term ‘Hellenismus’ was originally coined by the German historian Johann Gustav Droysen ( – ), and defined in the preface (Vorrede) of the first () edition of his Geschichte des Hellenismus (p. ). For recent discussion of this definition and of subsequent (mis)interpretation of Droysen’s term see Bastianini , Canfora . As a historical term it is used to define the period between Alexander’s death ( BC) to the death of Caesarion and the submission of the last Hellenistic kingdom, Ptolemaic Egypt, to Rome ( BC). In other respects the term is problematic: in the case of poetry the cultural span is a much longer one.  Ep.  Pf. ( GP) is a partial exception, in that the epigram treats of the love of one Callignotus, once sworn to a girl named Ionis, now transferred to a male eromenos (line : νῦν δ’ ὁ μὲν ἀρσενικῷ θέρεται πυρί). It is well worth noting in this context that the theme of lovers’ oaths not reaching the ears of the gods is paralleled at Pl. Phileb. c , Symp. b-c. In poetry the trope appears as early as Hes. fr.  MW, from a scholion to this same passage of the Symposium.

238

Benjamin Acosta-Hughes

parent frequency of Platonic recollections in what is, ultimately, the rather small corpus of Callimachus’ extant epigrams. And recollections of Plato are not limited to the erotic epigrams, even though they are there more obvious. There are some sixty-three extant Callimachean epigrams (63 is contested) and a handful of epigram fragments.³ Of the extant epigrams, thirteen have homoerotic love as their theme, and it is in these poems that most of the Plato recollections occur,⁴ though there are again other epigrams implicated in this dialogue as well. One of these last, Ep. 23 Pf. (53 GP), happily provides the testimonium that Callimachus treats Platonic material in his epigrams – the poem centers on what is essentially a (mis)reading of Plato’s Phaedo, and it’s central figure, Cleombrotus, is one of two notable absences at Socrates’ death (Phaed. 59c3 – 4).⁵ The other, Aristippus, is recalled in another Callimachean epigram (Ep. 20 Pf. = 32 GP). Proper names are themselves one feature of this discourse: of the proper names that feature in Callimachus’ epigrams a surprising number have parallels in Platonic dialogues. The number is slightly higher if, with the Platonic dialogues, we include Xenophon’s Symposium, and if, alongside Callimachus’ epigrams, we can include the Iambi. In some cases the parallel is fairly obvious: Theatetus (Ep. 7 Pf. = 57 GP), Menexenus (Ep. 44 Pf. = 9 GP), Critias (Ep. 12 Pf. = 43 GP; Ep. 55 Pf. = 16 GP), Euthydemus (Iambus 3), Ion (Iambus 13) happen to all be the names of Platonic dialogues. Of these Ep. 7 Pf. might make for an interesting first case study. Ἦλθε Θεαίτητος καθαρὴν ὁδόν. εἰ δ’ ἐπὶ κισσόν τὸν τεὸν οὐχ αὕτη, Βάκχε, κέλευθος ἄγει, ἄλλων μὲν κήρυκες ἐπὶ βραχὺν οὔνομα καιρόν φθέγξονται, κείνου δ’ Ἑλλὰς ἀεὶ σοφίην. Theatetus has gone on a pure road. If the path itself does not lead him to your ivy, Bacchus, the heralds will voice the names of others for but a short moment. His wisdom Greece will voice forever.

It has long been assumed in the scholarship on this epigram that Theatetus is a contemporary of Callimachus, apparently a poet who turns to some other poetic

 Callim. Ep.  Pf. ( GP) is often contested on ’stylistic’ grounds, but should not be excluded primarily due to the epigram having heterosexual love as its subject (and so not being “Callimachean”). That the poem is by Rufinus (so the APl.) remains however possible.  For earlier treatments of several of these poems and their associations with Plato texts see Acosta-Hughes , ; Acosta-Hughes/Stephens .  See further Acosta-Hughes/Stephens ,  – .

A Little-Studied Dialogue: Responses to Plato in Callimachean Epigram

239

form than drama.⁶ Enrico Livrea has further posited, expanding on an earlier suggestion by i.a. Gow-Page,⁷ that Theatetus is an Alexandrian poet, perhaps of Cyrenean origin, author of several epigrams in the Palatine Anthology (AP 6.357 = 1 GP, AP 7.727 = 3 GP, AP 7.449 = 4 GP, AP 7.499 = 4 GP, AP 7.444 = 5 GP) and two cited by Diogenes Laertius (4.25 = 2 GP, 8.48 = 6 GP).⁸ It is worth recalling here that Callimachus in his epigrams shares with Plato the artistic gesture of including contemporary figures in his epigrams, as e. g. famously Aratus in Ep. 27 Pf. (56 GP). Yet it is also typical of Callimachus’ poetry that any word or phrase may bear more than one allusion, a feature of this polymath artist. Hence one might posit an echo of another Theatetus, the central figure of Plato’s dialogue of the same name. Is there further support for this suggestion in Callimachus’ epigram? There is. While σοφία at line 4 is frequently associated from Pindar onwards with ‘poetry’, as it clearly is here, it also bears the meaning ‘wisdom’. And the key topic of the Theatetus is, indeed, the definition of ‘wisdom’ (143d-145e). In this way Callimachus’ epigram can be understood as beginning and ending with proper nouns that are the title and the subject of Plato’s dialogue. Are there other features in the poem that might be said to support this suggestion? καθαρὴν ὁδόν certainly has clear Pindaric parallels in Ol. 6 and in Isthm. 5,⁹ and in both cases the metaphor of turning onto a pure road can be read as an aesthetic decision in one poetic direction over another or others.¹⁰ Does it also have Platonic resonance? There is the phrase in the Theatetus again (208c6) τὸ διὰ στοιχείου ὁδὸς ἐπὶ τὸ ὃλον, ‘the road through the elements

 Cf. GP HE ..  GP HE .. See Livrea ,  –  = Livrea ,  – .  There is a second epigrammatist of this name, Theatetus Scholasticus, three of whose epigrams (AP ., . and .) derive from the Cycle of Agathias. Two epigrams in the Planudean Anthology (A.Pl , ) are also ascribed to him. See Livrea  and .  Cf. Ol. . –  ὦ Φίντις, ἀλλὰ ζεῦξον ἤ- | δη μοι σθένος ἡμιόνων | ᾇ τάχος, ὄφρα κελεύθῳ τ’ ἐν καθαρᾷ | βάσομεν ὄκχον, ἵκωμαί τε πρὸς ἀνδρῶν | καὶ γένος· κεῖναι γὰρ ἐξ ἀλ- | λᾶν ὁδὸν ἁγεμονεῦσαι | ταύταν ἐπίστανται, στεφάνους ἐν Ὀλυμπίᾳ | ἐπεὶ δέξαντο· ‘O Phintis, yoke for me the strength of mules, quickly, that I may drive my chariot in a pure road, and arrive at my family’s race. For those mules know beyond others to lead on this road, and have received crowns at Olympia.’ Isth. . –  εἰ δὲ τέτραπται | θεοδότων ἔργων κέλευθον ἂν καθαράν, | μὴ φθόνει κόμπον τὸν ἐοικότ’ ἀοιδᾷ | κιρνάμεν ἀντὶ πόνων. ‘If someone has turned along the pure road of god-given works, do not begrudge to blend with your song the suitable praise in return for toils’.  Cf. D’Alessio ,  n. , La Penna . On the ‘path’ as metaphor in Pindar see Steiner ,  – , and in Pindar and Callimachus see Asper ,  – , on Callim. Ep.  Pf. see esp.  – .

240

Benjamin Acosta-Hughes

to the whole’.¹¹ What is of particular interest here is the change of direction, a selection of a new focus of dialogue. A perhaps more intriguing parallel is Tim. 53c1– 2 τῶν κατὰ παίδευσιν ὁδῶν, ‘ways of education’ where again there is the sense of selection.¹² The term κέλευθος (line 2), one of great programmatic value in both Pindar and Callimachus, does not occur in Plato. κέλευθος is a poetic word until the end of the Classical age: the thought is Platonic, but dressed in poetic language, typical of Callimachus’ appropriation of Plato. And it is fairly reasonable for a reader who knows even a little about Plato and the poetic arts to note that Theatetus’ choice in Callimachus’ epigram appears to be one away from drama to another type of ‘wisdom’ or ‘poetry’, a choice that can be understood as reflecting Plato’s own conceptualization of ‘good poetry’.¹³ The addressee of Callimachus’ epigram, Ep. 44 Pf. (9 GP), Menexenus has also a name that parallels the title of one of Plato’s dialogues. Ἔστι τι ναὶ τὸν Πᾶνα κεκρυμμένον, ἔστι τι ταύτῃ ναὶ μὰ Διώνυσον πῦρ ὑπὸ τῇ σποδιῇ. οὐ θαρσέω· μὴ δή με περίπλεκε· πολλάκι λήθει τοῖχον ὑποτρώγων ἡσύχιος ποταμός. τῷ καὶ νῦν δείδοικα, Μενέξενε, μή με παρεισδύς οὗτος σιγέρπης¹⁴ εἰς τὸν ἔρωτα βάλῃ. There is, by Pan, hidden, there is there, yes, by Dionysus, fire under the ash. I have not courage. Do not embrace me. Often a quiet river escapes notice as it eats away at a wall. Wherefore even now I fear, Menexenus, lest this creeping thing slipping within me cast me into love.

Menexenus is the name of three figures in Plato’s Dialogues: Socrates’ interlocutor in the dialogue of the same name; Socrates’ son Menexenus; the companion of the boy Lysis in Plato’s Lysis. The first is a young man with an interest in pol-

 Pl. Theat. c – : {ΘΕΑΙ.} Ὀρθῶς ὑπέμνησας· ἔτι γὰρ ἓν λοιπόν. τὸ μὲν γὰρ ἦν διανοίας ἐν φωνῇ ὥσπερ εἴδωλον, τὸ δ’ ἄρτι λεχθὲν διὰ στοιχείου ὁδὸς ἐπὶ τὸ ὅλον· τὸ δὲ δὴ τρίτον τί λέγεις; Cf. Plat. Phileb. b  –  οὐ μὴν ἔστι καλλίων ὁδὸς οὐδ’ ἂν γένοιτο; Polit. a δύο τινὲ καθορᾶν ὁδὼ τεταμένα φαίνεται… .  Pl. Tim. c –  ἀλλὰ γὰρ ἐπεὶ μετέχετε τῶν κατὰ παίδευσιν ὁδῶν δι’ ὧν ἐνδείκνυσθαι τὰ λεγόμενα ἀνάγκη, συνέψεσθε.  E. g. the classic division at Rep.  – .  I follow D’Alessio ,  n.  in accepting Bentley’s σιγέρπης, ‘creeping secretly’, a rare term defined by Hesychius as λαθροδάκτης, ‘biting secretly’. The text is corrupt, the AP ms. reads οὗτος οσειγαρνης, as well as βάλλῃ for βάλῃ.

A Little-Studied Dialogue: Responses to Plato in Callimachean Epigram

241

itics, to whom Socrates recites Aspasia’s Epitaphios Logos. ¹⁵ Socrates’ son Menexenus appears at Apol. 34d and Phaed. 116b. It is the last figure, the Menexenus of the Lysis, that may be of particular interest here. When Socrates and his young companion Ctesippus enter the wrestling-school (Lys. 207e) they perceive the beautiful Lysis, object of Hippothales’ affection. The young man, however, is at first too shy to approach them, until his close friend Menexenus leaves off his game and comes over to greet them (207b). Only then does Lysis approach them and sit by Menexenus.¹⁶ Menexenus in Plato’s dialogue is thus the proximate aition of Socrates’ dialogue with the beautiful young Lysis, and because the enamored Hippothales is hidden and blushing in the background, Menexenus effectively leads Lysis to Hippothales’ love. In Callimachus’ epigram, Menexenus is the addressee, and also apparently the figure whose erotic appeal the speaker seeks to avoid (line 3 μὴ δή με περίπλεκε). Callimachus’ epigram consists of three couplets, each featuring an image of stealth and the hidden. The metaphor of the first couplet, love concealed as ember under ash, finds a parallel in Callimachus’ fifth Iambus (lines 23 – 26).¹⁷ Strikingly, following closely upon this parallel in Iambus 5 is an allusion at lines 31– 33 to Plato’s Theages, another Platonic (or pseudo-Platonic) text centered on the education of a young man.¹⁸ The image of the second couplet of Ep. 44 Pf., of water wearing away stone, finds perhaps its most familiar articulation in Lucretius (RN 1.313 stilicidi casus lapidem cavat), though it occurs in a number of contexts in both Aristotle and Plato.¹⁹ The compound ὑποτρώγω is

 The authenticity of this piece has been often doubted in the past, but it is now generally accepted as genuine, based in large part on two references to the speech in Aristotle’s Rhetoric: b  and b . Guthrie ,  –  has a brief selection of mid-th century views of the dialogue. See now also Trivigno ,  – .  See Lys. d  – : Μενεξένῳ μὲν γὰρ δὴ πάντων ὢν τυγχάνει. In much of the dialogue that follows Lysis and Menexenus are treated as not just age-mates, but as a pair. N.b. e. g. the duals in e and d.  Callim. Iamb.  (fr.  Pf.)  – : τὸ πῦρ δὲ τὠνέκαυσας, ἄχρις οὐ πολλῇ | πρόσω κεχώρηκεν φλογί, | ἀλλ’ ἀτρεμίζει κἠπὶ τὴν τέφρην̣ ọἰ ̣[χ]ν̣εῖ, κοίμησον, ‘But the fire which you have set alight, as long as it has not progressed further with much flame, but lies still and dwells in ashes, quench it.’  Callim. Iamb.  (fr.  Pf.)  – : ἐγὼ Βάκις τοι καὶ Σίβυλλα [καὶ] δάφνη | καὶ φηγός. ἀλλὰ συμβαλεῦ | τᾤνιγμα, καὶ μὴ Πιτθέως ἔχε χρείην. Cf. Plat. Theag. d  – : Socr. Εἴποις ἂν οὖν μοι τίνα ἐπωνυμίαν ἔχει Βάκις τε καὶ Σίβυλλα καὶ ὁ ἡμεδαπὸς ᾿Aμφίλυτος; Further on the Platonic resonances in Iamb.  see Acosta-Hughes/Stephens ,  – .  Cf. Aristot. Phys. b – : οὔτε γὰρ αὐξάνεσθαι οὔτε φθίνειν οἷόν τε τε συνεχῶς, άλλ’ ἔστι καὶ τὸ μέσον. ἔστι δ’ὄμοιος ὁ λόγος τῷ περὶ τοῦ τόν σταλαγμὸν κατατρίβειν καὶ τὰ ἐκφυγόμενα τοὺς λίθους διαιρεῖν.

242

Benjamin Acosta-Hughes

exceedingly rare: it occurs only twice elsewhere, once at Xenophanes 12.3,²⁰ where the context is symposiastic, and once in Xenophon’s Symposium, a text with which Callimachus’ epigrams share a number of affinities.²¹ Here the context is both symposiastic and homoerotic, and the contextual similarity gives some support to a link between the two. σιγέρπης in the epigram’s final line is of course exceedingly rare, but παρεισδύς is suggestive. The present participle παρεισδύνων occurs only in the fragments of the pro-Macedonian author Demades,²² both times of the slippage of unjust argument into the mind; the past participle παρεισδύς occurs in the masculine nominative singular only at Diog. Laert. 11.142.10, of the fourth century Eretrian philosopher Menedemus, who was also active as an ambassador to several of the Diadochoi. On a more general level, the epigram’s speaker urges Menexenus to maintain a distance, lest the speaker become infatuated – it is by his own example of coming to sit next to Socrates (Plat. Lys. 207b) that the figure Menexenus in Plato’s dialogue draws the boy Lysis over to Socrates, and so to the infatuated though concealed gaze of Hippothales.²³ Xenophon’s Symposium, like the Lysis, the Charmides and especially like Plato’s dialogue of the same title, is another Socratic text where gazing upon a beautiful young man is central to what then evolves in part as a result of the presence of young male beauty. Here Autolykus, like Agathon, is the figure whose victory occasions the gathering: his beauty has a mesmerizing effect on the older men present not unlike that of Charmides in Plato’s dialogue of the same name (Charmides is himself a figure in the Symposium).²⁴ Of particular interest here is the  Xenoph. Fr. . –  ( DK) πὰρ πυρὶ χρὴ τοιαῦτα λέγειν χειμῶνος ἐν ὥρηι | ἐν κλίνηι μαλακῆι κατακείμενον, ἔμπλεον ὄντα, | πίνοντα γλυκὺν οἶνον, ὑποτρώγοντ’ ἐρεβίνθους· | ’τίς πόθεν εἷς ἀνδρῶν; πόσα τοι ἔτε’ ἐστί, φέριστε; | πηλίκος ἦσθ’, ὅθ’ ὁ Μῆδος ἀφίκετο;’  Xen. Symp. .: Μηδαμῶς, ἔφη, ὦ Σώκρατες. εἰς μὲν γὰρ μάχην ὁρμωμένῳ καλῶς ἔχει κρόμμυον ὑποτρώγειν, ὥσπερ ἔνιοι τοὺς ἀλεκτρυόνας σκόροδα σιτίσαντες συμβάλ-λουσιν· ἡμεῖς δὲ ἴσως βουλευόμεθα ὅπως φιλήσομέν τινα μᾶλλον ἢ μαχούμεθα.  Demad. frr. . de Falco: οὕτως ἄδικος παρεισδύνων λόγος εἰς τὰς τῶν δικαστῶν γνώμας οὐκ ἐᾷ δι’ ὀργὴν συνορᾶν τὴν ἀλήθειαν; Demad. fr. .. de Falco: οὕτως ἄδικος παρεις-δύνων λόγος εἰς τὰς τῶν δικαστῶν γνώμας οὐκ ἐᾷ δι’ ὀργὴν συνορᾶν τὴν ἀλήθειαν.  Here the verbal expressions are ᾔει παρακαθιζησόμενος and συμπαρεκαθεζετο respectively.  Xen. Symp. .  – : Αὐτόλυκος μὲν οὖν παρὰ τὸν πατέρα ἐκαθέζετο, οἱ δ’ ἄλλοι, ὥσπερ εἰκός, κατεκλίθησαν. εὐθὺς μὲν οὖν ἐννοήσας τις τὰ γιγνόμενα ἡγήσατ’ ἂν φύσει βασιλικόν τι κάλλος εἶναι, ἄλλως τε καὶ ἂν μετ’ αἰδοῦς καὶ σωφροσύνης, καθάπερ Αὐτόλυκος τότε, κεκτῆταί τις αὐτό. πρῶτον μὲν γάρ, ὥσπερ ὅταν φέγγος τι ἐν νυκτὶ φανῇ, πάντων προσάγεται τὰ ὄμματα, οὕτω καὶ τότε τοῦ Αὐτολύκου τὸ κάλλος πάντων εἷλκε τὰς ὄψεις πρὸς αὐτόν· ἔπειτα τῶν ὁρώντων οὐδεὶς οὐκ ἔπασχέ τι τὴν ψυχὴν ὑπ’ ἐκείνου. οἱ μέν γε σιωπηρότεροι ἐγίγνοντο, οἱ δὲ καὶ ἐσχηματίζοντό πως. πάντες μὲν οὖν οἱ ἐκ θεῶν του κατεχόμενοι ἀξιοθέατοι δοκοῦσιν εἶναι· ἀλλ’ οἱ μὲν ἐξ ἄλλων πρὸς τὸ γοργότεροί τε ὁρᾶσθαι

A Little-Studied Dialogue: Responses to Plato in Callimachean Epigram

243

prominent speaker and γελοτοποιός Philippus, who is something of a foil to Socrates in the text, as the burlesque in this dialogue contrasts with the serious. Callimachus Ep. 46 Pf. (3 GP) is a poem that has a number of associations with Plato, and I have written on these earlier.²⁵ Ὡς ἀγαθὰν Πολύφαμος ἀνεύρατο τὰν ἐπαοιδάν τὠραμένῳ· ναὶ Γᾶν, οὐκ ἀμαθὴς ὁ Κύκλωψ· αἱ Μοῖσαι τὸν ἔρωτα κατισχναίνοντι, Φίλιππε· ἦ πανακὲς πάντων φάρμακον ἁ σοφία. τοῦτο, δοκέω, χἀ λιμὸς ἔχει μόνον ἐς τὰ πονηρά τὠγαθόν· ἐκκόπτει τὰν φιλόπαιδα νόσον. What an excellent charm Polyphemus discovered for one in love. Yes, by Ge, the Cyclops was no uneducated guy. The Muses, Philippus, slim love down. Poetry is the effective remedy for everything. This, I think, and hunger, is the sole good against evil: for it cuts out the boy-loving disease.²⁶

The Philippus who is the addressee of Callimachus Ep. 46 Pf. (3 GP) has been widely assumed to be a doctor-friend of the poet, based in large part on the analogy of Nicias, the doctor-poet who is the addressee of three of Theocritus’ Idylls (11, 13, 28), and particularly Id. 11 on Polyphemus’ love for Galatea. Callimachus’ epigram and Theocritus’ eleventh Idyll clearly reflect one another: not only are there the thematic parallels (and clearly mirrored differences), there are a number of verbal parallels as well. Gow-Page, following Edgar 1931: 126, refer to a physician active in Alexandria mentioned in a papyrus from the Zenon Archive of 240. Edgar here is, however, only offering a hypothesis, it is not more than that.²⁷

καὶ φοβερώτερον φθέγγεσθαι καὶ σφοδρότεροι εἶναι φέρονται, οἱ δ’ ὑπὸ τοῦ σώφρονος ἔρωτος ἔνθεοι τά τε ὄμματα φιλοφρονεστέρως ἔχουσι καὶ τὴν φωνὴν πρᾳοτέραν ποιοῦνται καὶ τὰ σχήματα εἰς τὸ ἐλευθεριώτερον ἄγουσιν. ἃ δὴ καὶ Καλλίας τότε διὰ τὸν ἔρωτα πράττων ἀξιοθέατος ἦν τοῖς τετελεσμένοις τούτῳ τῷ θεῷ. ἐκεῖνοι μὲν οὖν σιωπῇ ἐδείπνουν, ὥσπερ τοῦτο ἐπιτεταγμένον αὐτοῖς ὑπὸ κρείττονός τινος.  Acosta-Hughes ,  – .  φιλόπαιδα νόσον at line  of Callimachus’ epigram is somewhat paralleled by Plat. Rep. d  – : Ἄλλῳ, εἶπον, ἔπρεπεν, ὦ Γλαύκων, λέγειν ἃ λέγεις· ἀνδρὶ δ’ ἐρωτικῷ οὐ πρέπει ἀμνημονεῖν ὅτι πάντες οἱ ἐν ὥρᾳ τὸν φιλόπαιδα καὶ ἐρωτικὸν ἁμῇ γέ πῃ δάκνουσί τε καὶ κινοῦσι, δοκοῦντες ἄξιοι εἶναι ἐπιμελείας τε καὶ τοῦ ἀσπάζεσθαι.  Edgar ,  – , on the figure of the doctor Philippus in P. Mich.  lines  – : ἐπετέλεσεν δὲ μάλιστα Κα- | φισοφῷν ὁ Φιλίππου τοῦ ἰατροῦ υἱός, ‘but the most effective was Kaphisophon son of Philippos the doctor’ [trans. Edgar]. ‘Further I think there is good reason for identifying the father (who no doubt resided in Alexandria, for otherwise Philon would not have mentioned him) with the Phillippos to whom Kallimachos addressed his epigram on

244

Benjamin Acosta-Hughes

While this is certainly a logical direction of inquiry, names can bear more than one allusion. Unlike most of the characters in Xenophon’s dialogue, Philippus is not one of the καλοὶ κ’ἀγαθοί: indeed he is originally not admitted by the slave-boy as he is not ἄριστος (which has many associations, but in this context is not high-born, rather he is a jester) and because he ‘brings nothing’. He lives by his ability to elicit laughter, and without laughter he must go hungry (15 – 16). He also does not partake of the erotic dialogue, his concern being rather his stomach: as the opening lines of the dialogue note, the interest here is the acts of the καλοὶ κ’ἀγαθοί at play. There are a number of parallels between Callimachus’ epigram and Xenophon’s text, which is now generally assumed, following Huss 1999, to be a lighthearted work (as is, of course, at a certain aesthetic level, the theme of Polyphemus in love). Another epigram that may have associations with one or more Platonic texts is Ep. 43 Pf. (13 GP). Ἕλκος ἔχων ὁ ξεῖνος ἐλάνθανεν· ὡς ἀνιηρόν πνεῦμα διὰ στηθέων (εἶδες;) ἀνηγάγετο, τὸ τρίτον ἡνίκ’ ἔπινε, τὰ δὲ ῥόδα φυλλοβολεῦντα τὠνδρὸς ἀπὸ στεφάνων πάντ’ ἐγένοντο χαμαί· ὤπτηται μέγα δή τι· μὰ δαίμονας οὐκ ἀπὸ ῥυσμοῦ εἰκάζω, φωρὸς δ’ ἴχνια φὼρ ἔμαθον. The stranger’s wound went unnoticed. How dreadfully he drew breath from his breast – did you see? When he drank the third time, the roses shedding their petals from the man’s garlands were all on the ground. Truly he was very burned. By the spirits I don’t liken this from his manner. As a thief I know the tracks of a thief.

This poem bears a close resemblance to a sympotic epigram of Asclepiades (Sens 2011, 121– 22; Fantuzzi/Hunter 2004, 338 – 42). One Platonic dialogue that may be implicated here is again the Lysis, as Marco Fantuzzi has observed (2012, 341): the passage, Lysis 204b5-c2, is specifically about recognition of one in love, and, interestingly, comes just before the section discussed regarding Ep. 28 above. That the same portion of the Lysis is used twice in Callimachus’ epigrams is revealing, and suggests a very careful reading of the Lysis, often now considered a rather minor dialogue, against not only Callimachus’ epigrams, but in juxtaposition with Hellenistic poetry more generally.

the medicinal powers of song and poverty. (…) For naturally the poet [sc. Callimachus] would submit his remedy to a physician, just as Theorkritos dedicated his poem on the same theme to Nikias, ἰατρὸν ὲόντα; naturally also he would address a native of Kos in Doric; this is just the kind of allusiveness in which Kallimachos delighted.’

A Little-Studied Dialogue: Responses to Plato in Callimachean Epigram

245

This is the passage at hand (204b1-c2): Καὶ ὃς ἐρωτηθεὶς ἠρυθρίασεν. καὶ ἐγὼ εἶπον· Ὦ παῖ Ἱερωνύμου Ἱππόθαλες, τοῦτο μὲν μηκέτι εἴπῃς, εἴτε ἐρᾷς του εἴτε μή· οἶδα γὰρ ὅτι οὐ μόνον ἐρᾷς, ἀλλὰ καὶ πόρρω ἤδη εἶ πορευόμενος τοῦ ἔρωτος. εἰμὶ δ’ ἐγὼ τὰ μὲν ἄλλα φαῦλος καὶ ἄχρηστος, τοῦτο δέ μοί πως ἐκ θεοῦ δέδοται, ταχὺ οἵῳ τ’ εἶναι γνῶναι ἐρῶντά τε καὶ ἐρώμενον. And he [sc. Hippothales], when asked grew red, and I said, ‘O Hippothales son of Hieronymos, no longer say this, whether you love someone or no. For I know that you not only are in love, but that you are already far along in love. In other things I am of little worth and useless, but this somehow is given me from God, that I am swiftly able to recognize one in love and one beloved.

As in the case of the Callimachus epigram, at issue is a concealed ailment, and one somehow gifted to recognize the presence of the love. The symposiastic context of Callim. Ep. 43 Pf. encourages one to look to Plato’s dialogue of that title, and this turns out to be a very successful choice. Alcibiades arrives at the end of the Symposium (n.b. τὸ τρίτον ἡνίκ’ ἔπινε at Callim. Ep. 43.3), and the initial description of his arrival figures both a garland of flowers and ribbons and a reference to the depth of his voice (Symp. 212d3-e2): Καὶ οὐ πολὺ ὕστερον ᾿Aλκιβιάδου τὴν φωνὴν ἀκούειν ἐν τῇ αὐλῇ σφόδρα μεθύοντος καὶ μέγα βοῶντος, ἐρωτῶντος ὅπου ᾿Aγάθων καὶ κελεύοντος ἄγειν παρ’ ᾿Aγάθωνα. ἄγειν οὖν αὐτὸν παρὰ σφᾶς τήν τε αὐλητρίδα ὑπολαβοῦσαν καὶ ἄλλους τινὰς τῶν ἀκολούθων, καὶ ἐπιστῆναι ἐπὶ τὰς θύρας ἐστεφανωμένον αὐτὸν κιττοῦ τέ τινι στεφάνῳ δασεῖ καὶ ἴων, καὶ ταινίας ἔχοντα ἐπὶ τῆς κεφαλῆς πάνυ πολλάς, And not much later we heard the voice of Alcibiades in the courtyard, very drunk and shouting loudly, asking where Agathon was and ordering that he be led to Agathon. Then the flute-player supporting him and some of his followers led him in, and he stood at the doors, crowned with a thick garland of ivy and violets, and with many ribbons upon his head.

Alcibiades takes off the ribbons and places the garland on Agathon’s head (213b2– 3), and then, on discovering Socrates at his other side, places some on his head as well (213d7-e5). Then, under the excuse of drunkenness, he launches upon the narrative of his attempt to seduce Socrates (217e3 – 4): τὸ δ’ ἐντεῦθεν οὐκ ἄν μου ἠκούσατε λέγοντος, εἰ μὴ πρῶτον μέν, τὸ λεγόμενον, οἶνος ἄνευ τε παίδων καὶ μετὰ παίδων ἦν ἀληθής, but hence you would not have heard me speaking, unless first, as the saying goes, wine without boys and with boys were true,

246

Benjamin Acosta-Hughes

Alcibiades utters a Greek anticipatory figure of the Latin in vino veritas that is based on an allusion to Alcaeus fr. 366 V., οἶνος, ὧ φίλε παῖ, καὶ ἀλάθεα, the very allusion with which Asclepiades opens his poem, Ep. 18, which is the poem closely associated with this Callimachean epigram.²⁸ Alcibiades then goes on to tell of his attempt to seduce Socrates. At the very opening of this narrative, Alcibiades compares himself to one who has bitten by a poisonous snake (the adder), whose experience can only be understood by those who have been similarly afflicted (Symp. 217e6 – 218a7): ἔτι δὲ τὸ τοῦ δηχθέντος ὑπὸ τοῦ ἔχεως πάθος κἄμ’ ἔχει. φασὶ γάρ πού τινα τοῦτο παθόντα οὐκ ἐθέλειν λέγειν οἷον ἦν πλὴν τοῖς δεδηγμένοις, ὡς μόνοις γνωσομένοις τε καὶ συγγνωσομένοις εἰ πᾶν ἐτόλμα δρᾶν τε καὶ λέγειν ὑπὸ τῆς ὀδύνης. ἐγὼ οὖν δεδηγμένος τε ὑπὸ ἀλγεινοτέρου καὶ τὸ ἀλγεινότατον ὧν ἄν τις δηχθείη – τὴν καρδίαν γὰρ ἢ ψυχὴν ἢ ὅτι δεῖ αὐτὸ ὀνομάσαι πληγείς τε καὶ δηχθεὶς ὑπὸ τῶν ἐν φιλοσοφίᾳ λόγων, οἳ ἔχονται ἐχίδνης ἀγριώτερον, νέου ψυχῆς μὴ ἀφυοῦς ὅταν λάβωνται, καὶ ποιοῦσι δρᾶν τε καὶ λέγειν ὁτιοῦν… And further the suffering of one bitten by an adder overcame me as well. For they say I think that someone who has experienced this is not willing to say what it was like except to those who have been bitten, as only these will recognize the experience, and will pardon his doing and saying anything from the pain. I have been bitten by something more painful and in the most painful place where one can be bitten – in the heart or soul or whatever one should call it, struck and bitten by philosophical arguments, which have a more savage bite than the adder, when they overcome a young man who is not unintelligent, and make him say and do anything at all…

φωρὸς δ’ ἴχνια φὼρ ἔμαθον, as a thief, I know the tracks of a thief. The last line of Callimachus’ epigram becomes the clearer through the allusion – even εἰκάζω now takes on a slightly different shading, as Alcibiades’ likeness of his own condition to one bitten by a snake is, of course, an analogy.

 Asclepiades Ep. : Οἶνος ἔρωτος ἔλεγχος· ἐρᾶν ἀρνεύμενον ἡμῖν ἤτασαν αἱ πολλαὶ Νικαγόρην προπόσεις· καὶ γὰρ ἐδάκρυσεν καὶ ἐνύστασε καί τι κατηφὲς ἔβλεπε, χὠ σφιγχθεὶς οὐκ ἔμενε στέφανος. Wine is a test of love: though he kept denying to us that he was in love. Nicagoras’ many toasts revealed the truth. For he cried and hung his head and had a very downcast look. And the garand bound to him did not remain on his head. (trans. A. Sens) The same ‘motto’ occurs at the beginning of Theocr. Id. 29 (see Acosta-Hughes 2010, 110 – 113). Alcaeus was particularly famed in Antiquity for his pederastic poetry, an aspect of Alcaeus that we can no longer see in his very fragmentary extant work, which derives mostly from his stasiotic poetry. On the Alexandrian edition of Alcaeus see Liberman 2002, XL-LXXXVI; AcostaHughes 2010, 134– 40.

A Little-Studied Dialogue: Responses to Plato in Callimachean Epigram

247

There may be yet a third Plato text implicated in this same epigram, which makes for a strikingly implicated poem, one that alludes to multiple well-known Platonic passages. Φώρ (line 6) is not a common poetic word (this would be κλέπτης), indeed it is not a common word period. For Callimachus to use this so emphatically in the epigram’s final line suggests that there is a clue here that his audience is meant to understand. The general take on this final line is that the phrase is a sententia (e.g. GP HE 2.168 e.g.), although such a Greek ‘saying’ is unknown.²⁹ There is, however, a remarkable use of the term in the first book of Plato’s Republic, and again the setting is a paideutic one (and one also with homoerotic shading).³⁰ At Rep. 334a1-b6 Socrates and the young Polemarchus come to the very unusual point of argument that justice might be a form of thievery: ᾿Aλλὰ μὴν στρατοπέδου γε ὁ αὐτὸς φύλαξ ἀγαθός, ὅσπερ καὶ τὰ τῶν πολεμίων κλέψαι καὶ βουλεύματα καὶ τὰς ἄλλας πράξεις; Πάνυ γε. Ὅτου τις ἄρα δεινὸς φύλαξ, τούτου καὶ φὼρ δεινός. Ἔοικεν. Εἰ ἄρα ὁ δίκαιος ἀργύριον δεινὸς φυλάττειν, καὶ κλέπτειν δεινός. Ὡς γοῦν ὁ λόγος, ἔφη, σημαίνει. Κλέπτης ἄρα τις ὁ δίκαιος, ὡς ἔοικεν, ἀναπέφανται, καὶ κινδυνεύεις παρ’ Ὁμήρου μεμαθηκέναι αὐτό· καὶ γὰρ ἐκεῖνος τὸν τοῦ Ὀδυσσέως πρὸς μητρὸς πάππον Αὐτόλυκον ἀγαπᾷ τε καί φησιν αὐτὸν πάντας ἀνθρώπους κεκάσθαι κλεπτοσύνῃ θ’ ὅρκῳ τε. ἔοικεν οὖν ἡ δικαιοσύνη καὶ κατὰ σὲ καὶ καθ’ Ὅμηρον καὶ κατὰ Σιμωνίδην κλεπτική τις εἶναι, ἐπ’ ὠφελίᾳ μέντοι τῶν φίλων καὶ ἐπὶ βλάβῃ τῶν ἐχθρῶν. But then the same man would be a good guard of a military camp who also is good at stealing the deliberations and other affairs of the enemy? Yes indeed. And of what one is a clever guard, of this he is also a clever thief? It would seem so. If then the just man is clever at guarding silver, he is also clever at stealing it? So, said he, the argument shows. Then the just man is shown to be some sort of thief, it seems, and you run the risk of having learned this from Homer. For Homer treats Odysseus’ grandfather on his mother’s side, Autolykus, with affection, and says he excels over all men in stealing and perjury. It would seem that that justice, according to you and Homer and Simonides, is some sort of thievery, however with an eye to the benefit of friends and the harm of enemies.

Let us pause to consider the passage for a moment. Polemarchus risks having learned (μεμαθηκέναι