^* Parts of this paper have been heard as lectures at the Center for Hellenic Studies, Washington; the Society for the Promotion of Hellenic Studies, London; Columbia University, New York; Yale University; Dartmouth College, New Hampshire; and Wellesley College, Mass. I am grateful to the many scholars who offered suggestions and criticisms on these occasions, to the generations of pupils who have engaged in helpful discussion, and to the Journal's referees.

¹ Editions: West, M. L., Iambi et elegi graeci ante Alexandrum cantati, 2 vols (Oxford 1971–1972)Google Scholar and Theognidis et Phocylidis fragmenta (Berlin/New York 1978)Google Scholar; Gentili, B. and Prato, C., Poetarum elegiacorum testimonia et fragmenta (Leipzig 1979)Google Scholar. For a bibliography of editions, commentaries and discussions up to 1977 (for elegy in general, and for Callinus, Tyrtaeus, Mimnermus, Solon and Xenophanes) see Gentili–Prato xiv–xliv. For commentary on Archilochus see Tarditi, G., Archiloco (Rome 1968)Google Scholar; on Theognis van Groningen, B. A., Théognis: le premier livre (Amsterdam 1966)Google Scholar and Vetta, M., Teognide, libro secondo (Rome 1980)Google Scholar. Important discussion of the genre and of individual passages is to be found in West, M. L., Studies in Greek elegy and iambus (Berlin/New York 1974)CrossRefGoogle Scholar (hereafter ‘Studies’).

The basic study of sympotic poetry remains Reitzenstein, R., Epigram und Skolion (Giessen 1893)CrossRefGoogle Scholar. Note also Gentili, B., ‘Epigramma ed elegia’, Fond. Hardt xiv (1968) 39–41 Google Scholar; G. Giangrande, ‘Sympotic literature and epigram’, ibid. 91–174; Tsagarakis, O., Self–expression in early Greek lyric, elegiac and iambic poetry, Palingenesia xi (Wiesbaden 1977)Google Scholar—based on his 1966 Munich dissertation Die Subjectivität in der griechischen Lyrik; Rösler, W., Dichter und Gruppe (Munich 1980)Google Scholar and reviewing Tsagarakis (op. cit.) in Gnomon lxxx (1980) 609–16Google Scholar; Aloni, A., Le muse di Archiloco (Copenhagen 1981)Google Scholar; Vetta, M., Poesia e simposio nella Grecia antica (Rome/Bari 1983)Google Scholar—a collection of essays with introduction by Vetta and a reprint of his ‘Un capitolo di storia di poesia simposiale (per l'esegesi di Aristoph. Vesp. 1222–48)’, Dialoghi di archeologia ix–x (1977) 242–66Google Scholar; Burnett, A. P., Three archaic poets (London 1983)Google Scholar. The above are cited by author's name only. See also Sympotica, ed. O. Murray (forthcoming: papers of a symposium on the symposium held in Oxford in 1984). References to fragments of the elegists are to West's edition.

² On Theognis see West 65–71.

³ In his edition West signals 10 complete poems in the Theognidea: 39–52; 183–92; 257–60; 263–66; 373–400 (with some lacunae); 467–96; 511–22; 667–82; 861–4; 959–62. I doubt 263–6, but would tentatively add 757–64; 1231–4; 1387–90. Outside the Theognidea West accepts another 9, but of these only Solon 13 is of any length (76 lines); Simonides 6 and Sophocles 4 veer towards epigram; and the riddles of Cleobulina 1 and 2, and sententiae of Demodocus 2, 3, 4 and 5 belong to distinctive classes of poetry. West would consider adding Archilochus 5 and 13: 5 I doubt, but 13 may (with Solon 13) be our only guide to the shape of an early elegy outside the Theognidea.

⁴ For discussion of the metrical form cf. West 9–10; id., Greek Metre (Oxford 1982) 35–46 Google Scholar. For a subtler view of the implications of shared vocabulary cf. Aloni 80 ff. 102 ff.

⁵ Dover, K. J., ‘The poetry of Archilochus’, Fond. Hardt x (1964) 190–4Google Scholar, unfortunately misinterpreted by Campbell, D. A., Greek lyric poetry (London and New York 1967) 169 Google Scholar.

⁶ West 13–14.

⁷ Campbell, D. A., ‘Flutes and elegiac couplets’, JHS lxxxiv (1964) 63–8CrossRefGoogle Scholar, followed by Rosenmeyer, T. G., ‘Elegiac and elegos’, Cal. Stud. Class. Ant. i (1968) 217–31Google Scholar, whose arguments against the singing of early elegy do not persuade me (nor, apparently, Herington, J., Poetry into drama [Berkeley/Los Angeles 1985]Google Scholar whose balanced discussion [31–6] and clear statement of the evidence [192–3] I only saw when this paper was in final draft). Neither Campbell nor Rosenmeyer satisfactorily explains Theognis 239 ff.

⁸ But these lines' interpretation is much disputed. See van Groningen; Campbell (n. 7) 347–9; West 149 f.

⁹ For a careful discussion, perhaps reading too much into these few words, see Aloni 31–48. I do not attempt to give a bibliography of the many discussions of the ‘I’ in personal poetry: see however most recently Rösier, W., ‘Personale reale o personale poetica’, Quad. Urb. Cult. Class, xix (1985) 131–44Google Scholar.

¹⁰ West 11–12: contra Campbell (n. 7) 64, Rosenmeyer (n. 7) 221 n. 23.

¹¹ West 12.

¹² Ibid. 10–13.

¹³ For Mimnermus 14 see below section IV; for an apparently exhortatory scrap of Archilochus, Peek, W., ZPE lix (1985) 13–22 Google Scholar.

¹⁴ Sympotica (forthcoming, cf. n. 1)

¹⁵ Φιλόχορος δέ φησιν κρατήσαντας Λακεδαιμονίους Μεσσηνίων διὰ τὴν Τυρταίου στρατηγίαν ἐν ταῖς στρατείαις ἔθος ποιήσασθαι, ἂν δειπνοποιήσωνται καὶ παιωνίσωσιν, ᾄδειν καθ’ ἓνα <τὰ> Τυρταίου. κρίνειν δὲ τὸν πολέμαρχον καὶ ἆθλον διδόναι τῷ νικῶντι κρέας.

Philochorus FGrH 328 F 216 = Athenaeus xiv 29, 630F;

καί περι τούς ἄλλους ούδένα λόγον ἔχοντες περὶ τούτου οὓτω σφόδρα ἐσπουδάκασιν, ὣστε νόμον ἔθεντο, ὃταν ἐν τοῖς ὅπλοις ἐκστρατευόμενοι ὦσι, καλεῖν ἐπί τήν τοῦ βασιλέως σκηνὴν ἀκουσομένους τῶν Τυρταίου ποιημάτων ἃπαντας, νομίʒοντες οὓτως ἂν αὐτοὺς μάλιστα πρό τῆς πατρίδος ἐθέλειν ἀποθνῄσκειν

Lycurgus, in Leocr. 107.

¹⁶ West II suggests Archilochus 1, 2 and 5. It is not clear to me whether he is also suggesting Theogn. 887–8 and 1043.

¹⁷ Cf. Dover (n. 5) 205 ff.; Rösler (n. 9). Some good points (amid much obscurity) emerge from Tsagarakis' discussion, especially his introduction (1–9).

¹⁸ ἳππος ἐγὼ καλὴ ἀεθλίη, ἀλλὰ κάκιστον ἄνδρα φέρω, καί μοι τοῦτ’ ἀνιηρότατον. πολλάκι δὴ ’μέλλησα διαρρήξασα χαλινὸν φεύγειν ὠσαμένη τὸν κακὸν ἡνίοχον. Theogn. 257–60

ἐχθαίρω κακὸν ἄνδρα, καλνψαμένη δὲ πάρειμι, σμικρῆς ὄρνιθος κοὔφον ἔχοντα νόον. ibid. 579–80

οἳ με φίλοι προδιδοῦσι καὶ οὐκ ἐθέλουσί τι δοῦναι ἀνδρῶν φαινομένων. ἀλλ’ ἐγὼ αὐτομάτη ἑσπερίη τ’ ἔξειμι καὶ ὀρθρίη αὖτις ἔσειμι, ἧμος ἀλεκτρυόνων φθόγγος ἐγειρομένων. ibid. 861–4.

I am not persuaded by the theory that 257–60 should follow 261–6 and be taken as the utterance of the girl ( Cataudella, Q., Riv. cult, class, e. med i [1960] 7–20 Google Scholar = Intorno ai lirici greci [Rome 1972] 120–38Google Scholar; Davison, J. A., CR ix [1959] 1–5 Google Scholar = From Archilochus to Pindar [London 1968] 281–8Google Scholar): but even were it correct, it would explain only one of the Theognidean passages.

¹⁹ Cf Anacreon 417P. Male sexual partners could also be seen as horses, cf. Theogn. 1249–52, 1267–70, with Vetta ad locc.

²⁰ ἔμε δείλαν, ἔlμε παίσlαν κακοτάτων πεδέχοισαν Alcaeus 10LP ἐκ ποταμοῦ ̉πανέρχομαι πάντα φέρουσα λαμπρά Anacreon 385P

²¹ Page, Sappho and Alcaeus (Oxford 1955) 185 Google Scholar. The allegorical interpretation is generally accepted, e.g. by Rösler 126–34, Burnett 150–3. That neither Archilochus' nor Alcaeus' song was composed or sung on a real ship was observed briefly by W. J. Slater in his stimulating and suggestive paper ‘Symposium at sea’, HStClPh lxxx (1976) 161–70Google Scholar. This was published only after I had reached my own view of the poems, and I do not here reproduce the abundant evidence that he marshalls for a persistent habit of describing sympotic activity in nautical terms. The support Slater's arguments offer to my case is all the greater because Archilochus 4 was not one of his primary witnesses but was only introduced allusively (168).

²² Burnett 39 also sees the occasion as ‘pretended’, taking it to be ‘a storm at sea’, but does not discuss phylake. Gerber, D. E., in a judicious discussion, ‘Archilochus, fr. 4 West: a commentary’, Ill. Class. Stud. vi 1 (1981) 1–11 Google Scholar, suggests (11 n. 6) that the poem ‘was recited at some convivial gathering’ but may have been composed (or presented by Archilochus as composed) while he was on watch. Vetta xiv–xv saw the importance of ξεινοι and δεῖπνον but thought the occasion military, after a victory, and that Archilochus transposes to it the atmosphere of the symposium. No discussion of the problem in Campbell (n. 5), Aloni or Podlecki, A. J., The early Greek poets and their times (Vancouver 1984)Google Scholar. West 11 supposed the ship to be beached; Gerber 3 argued that it is anchored. The imitation of this poem by Synesius in epist. 32 Hercher = 45 Garzya, demonstrated by Garzya, A., ‘Una variazione Archilochea in Sinesio’, Maia x (1958) 66–71 Google Scholar, repr. with additions in id., Studi sulla lirica greca (Florence 1963) 161–9, may indicate that Archilochus' ship was at sea rather than beached (though Garzya does not seriously consider the latter possibility) but it cannot tell us that the song was actually first sung on a ship, only that this is the situation Archilochus evoked in the poem, pace Nisbet, R. G. M., ‘Horace's Epodes and history’, Poetry and politics in the age of Augustus, edd. Woodman, A. J. and West, D. A. (Cambridge 1984) 200 Google Scholar n. 62. Nisbet's suggestion that it is engaged in a blockade may overestimate the capacity of pentekonters: Garzya himself wanted to dilute the sense of φυλακῇ to ‘sorvegliar di notte’ (n. 12), doubtless because he was attracted by the notion that the ship was a merchant vessel carrying wine (surely ruled out by θοῆς line 5).

²³ On Dionysius Chalcus (with discussion of his penchant for strained metaphor) cf. Garzya, A., Riv. Fil. Class, xxx (1952) 193–207 Google Scholar, repr. in Studi (n. 22) 91–102. It is tempting to see Dionysius' σέλματα as a reference to the κλῖναι of symposiasts (or are we to imagine some symposia as involving seated, not reclining participants?): if so, then Archilochus' αέλματα could also so allude. For the story from Timaeus see Athenaeus ii 5, 37 B-D (=FGrH 566 F 149). For discussion of these and many other instances of sympotic activity viewed in nautical terms see Slater (n. 21).

²⁴ I see no grounds for drawing a sharp distinction between the extent to which a song might evoke a situation that was not actual in its first and in subsequent performances.

²⁵ Gentili, B., RFIC xciii (1965) 129–34Google Scholar; Bossi, F., Quad. Urb. Cult. Class, xxxiv (1980) 23–7Google Scholar.

²⁶ Theoc. 7. 65–6:τόν πτελεατικόν οῙνον ἀπὸ κρατῆρος ἀφυξῶ πὰρ πυρί κεκλιμένος, κύαμον δέ τις ἐν πυρί φρυξεῖ. Note also τρύγα in line 70, cf. τρυγὸς in Archilochus 4.6.

²⁷ So Rankin, H. D., Emerita xl (1972) 469–74Google Scholar. A simple gesture would suffice to convey to the audience that the singer was simultaneously reclining on a couch and leaning on his spear.

²⁸ The full text of Plutarch, Solon 8.2 ( = Solon 1) runs: ἐσκήψατο μέν ἔκστασιν τῶν λογισμῶν, καί λόγος εỉς τήν πόλιν ἐκ τῆς οỉκίας διεόθη παρακινητικῶς ἔχειν αủτόν. ἐλεγεῖα δέ κρύφα συνθείς καί μελετήσας ὥστε λέγειν ἀπὸ στόματος, ἐξεπήδπσεν εỉς τὴν ἀγοράν ἄφνω, πιλίδιον περιθέμενος, ὄχλου δέ πολλοῦ συνδραμόντος ἀναβάς έπί τόν τοῦ κήρυκος λίθον ἐν ὠδῇ διεξῆλθε τήν ἐλεγείαν, ἧς ἐστιν ἀρχή … Polyaenus i 20.1; Diogenes Laertius i 46 (where Solon actually has a herald read out his poem!).

²⁹ The allusions in Demosthenes are at 19.252 (the prohibition and Solon's performance of the elegies) and 255 (the πιλίδιον—I owe this point to B. M. W. Knox). Note that whereas Plutarch and Diogenes Laertius talk as if the elegy were recited or read, Demosthenes and Polyaenus know that the elegies were sung (ᾗδε).

³⁰ Lefkowitz, M. R., The Lives of the Greek poets (London 1981) 40 Google Scholar with n. 3. I doubt, however, if the erection of a statue of Solon by the Salaminians can be connected with a misunderstanding of ἀντ̉ ἀγορῆς to mean ‘before the market place’, as she suggests. Fr. 10 is quoted by Diogenes Laertius in connection with an anecdote that seems to be a doublet of the Salamis story, concocted to explain the term μανίην. The hypothesis that in the Salamis Solon put his words into the mouth of a fictitious character (Podlecki [n. 22] 123) is neither necessary nor helpful. More useful (but accepting the ancient tradition) Tedeschi, G., ‘Solone e lo spazio della communicazione elegiaca’, Quad. Urb. cult, class, x (1982) 41 ffGoogle Scholar.

³¹ Cf. Theogn. 783–8, beginning, ἧλθον μὲν γὰρ ἔγωγε καὶ εἰς Σικελήν ποτε γαῖαν. This form of poem is the corollary of the prosphonetikon, such as we find in Archilochus 24, Alcaeus 350 LP: cf. Cairns, F., Generic composition in Greek and Roman poetry (Edinburgh 1972) 18 ff.Google Scholar; Russell, D. A. and Wilson, N. G., Menander Rhetor (Oxford 1981) 328 fGoogle Scholar.

³² At least as far down Athenian society as elegy was sung—a question unanswerable in the present state of our evidence. West 16 postulates a similar process, cf. Reitzenstein 49 n. 1. For a similar argument that 4c is addressed to Solon's hetairoi (but identifying ἡμεῖς with Solon's group) cf. Vetta xvii.

³³ 12.

³⁴ τὴν ἐλεγείαν … ἐν ἧ πρὸς ἑκατέρους ὑπὲρ ἑκατέρων μάχεται καί διαμφισβητεῖ, Athenaion Politeia 5 = Solon 4a.

³⁵ Ibid. = Solon 4c:

ὑμεῖς δʼ ἡσυχάσαντες ἐνὶ φρεσὶ καρτερὸν ἦτορ, οἳ πολλῶν ἀγαθῶν ἐς κόρον [ἠ]λάσατε,

ἐν μετρίοισι τί̣θ̣ε̣σ̣θ̣ε μέγαν νόον οὔτε γὰρ ἡμεῖς πεισόμεθʼ, οὔθʼ ὑμῖν ἄρτια τα[ῦ]τ̣’ ἔσεται.

³⁶ Aristotle Pol. 1296a18.

³⁷ We have a fragment of another poem in which Solon put a speech into another's mouth, 33. The introduction of two speakers was already part of Archilochus' technique, cf. 23, 172–81, 196A.

³⁸ 12–13.

³⁹ For discussion, Davison and Cataudella (cited n. 18); van Groningen ad loc., 452 ff; West 152.

⁴⁰ West 18 may thus be too rigorous in suggesting that elegy is never used for sexual narratives and fantasies. Theogn. 993 ff. comes close to sexual fantasy. For such songs in other genres cf. Archilochus 196A; Ar. Acharn. 271–5.

⁴¹ For girls ministering at symposia cf. Theogn. 1002, perhaps also 1211 ff. The girl in 1211 ff. is certainly servile, but the context not certainly sympotic; at 1002 the context is sympotic, and that the girl is Laconian probably shows she is a slave. For vases cf. Langlotz, E., Griechische Vasen in Würzburg (Munich 1932)Google Scholar no. 483 fig. 152 ( = Fehr, B., Orientalische und griechische Gelage [Bonn 1971] no. 409 cf. p. 101 Google Scholar); Boardman, J., Athenian black figure vases (London 1974) 210 Google Scholar; id., Athenian red figure vases (London 1975) 219; F. Lissarague in Sympotica.

⁴² Even the association of aulos music with mourning is first documented by vases for the sixth century and literary testimony for the fifth, cf. Reiner, E., Die rituelle Totenklage bei den Griechen (Tübingen 1983) 67–70 Google Scholar. Homer does not mention the aulos at Il. xxiv 720.

⁴³ For arguments leading to the same conclusion, Gentili 59–63.

⁴⁴ Plut. quomodo aud. poet. 6 = Mor. 23 b ( = Archilochus 9, 10–11): ὅταν δὲ τὸν ἄνδρα τῆς ἀδελφῆς ἠφανισμένον ἐν θαλάττῃ καὶ μὴ τυχόντα νομίμου ταφῆς θρηνῶν λέγη μετριώτερον ἂν τὴν συμφορὰν ἐνεγκεῖν εἰ κείνου κεφαλὴν καί χαρίεντα μέλεα Ἥφαιστος καθαροῖσιν ἐν εἵμασιν ἀμφεπονήθη, τὸ πῦρ οὕτως, οὐ τὸν θεὸν προσηγόρευσε.

⁴⁵ Id. 12 = Mor. 33ab ( = Archilochus 11).

⁴⁶ Archilochus 13 (perhaps complete).

⁴⁷ For analysis of the ways 1–2 may be interpreted see Kamerbeek, J. C., ‘Archilochea’, Mnem. xiv (1961) 1 ffCrossRefGoogle Scholar. Burnett (46), ‘Pericles, our wailing grief will not be blamed by any burgher at his feast’, gives too little weight to the main verb τέρψεται.

⁴⁸ Campbell 396; Page, D. L., Further Creek Epigrams (Cambridge 1981) 295 Google Scholar (denying that it is a complete poem); contra Gentili 62–3, seeing it as an elegiac epigram comparable to a skolion.

⁴⁹ Plut. Cim. 4.10 ( = Archelaus 1): καί δυσφορήσας ἀποθανούσης, εἴ τι δεῖ τεκμαίρεσθαι ταῖς γεγραμμέναις ἐπὶ παρηγορίᾳ τοῦ πένθους ἐλεγείαις πρὸς αὐτόν …

⁵⁰ Ps. Plut. Consolatio ad Apollonium 106b (= Antimachus t7): ἀποθανούσης γὰρ τῆς γυναικὸς αὐτῷ Λύδης, πρὸς ἣν φιλοστόργως εἶχε, παραμύθιον τῆς λύπης αὑτῷ ἐποίησε τὴν ἐλεγείαν τὴν καλουμένην Λύδην, ἐξαριθμησάμενος τὰς ἡρωικὰς συμφοράς, τοῖς ἀλλοτρίοις κακοῖς ἐαυτοῦ ποιῶν λύπην.

⁵¹ Pausanias x 7.5–6. West's interpretation (in his edition s.n. ‘Echembrotus’) of Pausanias' date as 586 is right as against Rosenmeyer's (above n. 7) of 582.

⁵² West 5. The list of Pythionikai had been drawn up by Aristotle and Callisthenes, cf. Pfeiffer, R., History of Classical Scholarship i (Oxford 1968) 79–80 Google Scholar.

⁵³ D. L. Page ‘The elegiacs in Euripides’ Andromache' in Greek Poetry and Life (Essays presented to Gilbert Murray) 206–30.

⁵⁴ Ps. Plut. de musica 1132c: ὁμοίως δὲ Τερπάνδρῳ Κλονᾶν, τὸν πρῶτον συστησάμενον τοὺς αὐλωδικοὺς νόμους καί τὰ προσόδια, ἐλεγείων τε καί ἐπῶν ποιητὴν γεγονέναι. The comparison with Terpander shows that Heraclides Ponticus (here being quoted) thought that Clonas used the aulodic nomes to set his poetry to music. Rosenmeyer (n. 7) 222–6 is excessively sceptical in his questioning of the tradition in the de musica and consequent denial that these early poets composed in the elegiac metre.

⁵⁵ Ibid. 1134a. Sacadas' victories were at the Pythia of 586, 582 and 578 BC, cf. Paus. x 7.4 and (citing discussions of the nomos Pythios) Frazer ad loc.

⁵⁶ Suda iii 522.21 A: Ὄλυμπος Μαίονος, Μυσός, αὐλητὴς καί ποιητὴς μελῶν καί ἐλεγείων. Schol. Ar. Equit. 9a: ὁ δὲ Ὄλυμπος μουσικὸς ἦν, Μαρσύου μαθητής. ἔγραψε δὲ αὐλτικοὺς καί θρηνητικοὺς νόμους.

⁵⁷ In the sense lament, Ap. Rhod. ii 782; Lucillius, Anth. Pal. xi 135.3; Hadrian in W. Peek, Griechische Versinschriften 2050.5; in the metrical sense ‘a poem in elegiacs’, Callim. fr. 7.13; Apollonidas epigr. 26 ( = Anth. Pal. x 19) 5; Meleager epigr. 1( = Anth. Pal. iv.1) 36; Pollianus, Anth. Pal. xi 130.3; G. Kaibel, Epigrammata Graeca 1000.

⁵⁸ Etym. Magn. 326.46 = 935 Gaisford s.v.‘ἔλεγος’. θρῆνος ὁ τοῖς τεθνεῶσιν ἐπιλεγόμενος εἴρηται δὲ παρὰ τὸ ἓ ἓ λέγειν ἐν τοῖ τάφοις. Cf. Suda ii 241 no. 774A s.v. ‘ἔλεγος’. θρῆνος. ἀπὸ τοῦ ἒ ἒ λεγ́ειν. ἢ οἱ πρὸς τὸν αὐλὸν ᾀδόμενοι θρῆνοι and Marius Plotius Saccrdos, Gramm. Lat. vi 509.31 Keil: elegiacum metrum dictum est quod εε sonat interiectioncm flautis.

⁵⁹ Cf. R. Pfeiffer (n. 52) 61–2.

⁶⁰ Ibid. 53–4.

⁶¹ For the date c. 425 see Euripides, Andromache, ed. Stevens, P. T. (Oxford 1971) 19 Google Scholar.

⁶² 2–3; cf. already Dover (n. 5) 187 ff.

⁶³ Critias 4:

καὶ νῦν Κλεινίου υἱὸν Ἀθηναῖον στεφανώσω Ἀλκιβιάδην νέοισιν ὑμνήσας τρόποις· οὐ γάρ πως ἧν τοὔνομ' ἐφαρμόӡειν ἐλεγείῳ

νῦν δ' ἐν ἰαμβείῳ κείσεται οὐκ ἀμέτρως Pherecrates fr. 153. 6–7:

ὁ δ' ἄχθεται αὐτος ὁ θύων τῷ κατακωλύοντι καὶ εὐθὺς ἔλεξ' ἐλεγεῖα.

⁶⁴ Geissler, , Die Chronologie der altattischen Komödie (Berlin 1925)Google Scholar wanted a late date because he believed Philoxenos to be mentioned in fr. 145 Edmonds quoted by Ps. Plut. de mus.; so too Koerte in RE xix.2 (1938) 1989–90 (at earliest 410 BC). But as Edmonds insisted Philoxenus is not named, and his date of 418 BC ( Fragments of Attic Comedy i [Leiden 1957] 263 Google Scholar n. a on fr. 145) better suits a dramatist not otherwise attested after 415 BC.

⁶⁵ Thuc. i 132.2–3 (Pausanias on the tripod dedicated at Delphi) ἠξίωσεν ἐπιγράψασθαι αὐτὸς ἰδίᾳ τὸ ἐλεγεῖον τόδε … τὸ μὲν οὗν ἐλεγεῖον οἱ Λακεδαιμόνιοι ἐξεκόλαψαν … Ion of Samos in ML 95(c) line 5: ἐχσάμου ἀμφιρύτου τεῦξε ἐλεγεῖον∶ Ἴων. West 20 n. 30 suggests a date of 404 BC but Meiggs and Lewis (291) think it may have been composed (as it was certainly inscribed) in the fourth century.

⁶⁶ As West 7 n. 8.

⁶⁷ Ibid. 6–7.

⁶⁸ On threnoi cf. Harvey, A. E., ‘The classification of Greek lyric poetry’ CQ v (1955) 168–172 Google Scholar; Alexiou, M., The ritual lament in the Greek tradition (Cambridge 1974) 11 ffGoogle Scholar.

⁶⁹ 7.

⁷⁰ Fr. 122.14.

⁷¹ PMG 890.4—perhaps not even an allusion to the term skolion—εὐθὺν χρὴ τὸν ἑταῖρον ἔμμεν καὶ μὴ σκολιὰ φρονεῖν.

⁷² For cases where it nevertheless was so used see n. 57.

⁷³ The aulos was used to accompany a wide range of choral song and dance, cf. Huchzermeyer, H., Aulos und Kithara in der griechischen Musik bis zum Ausgang der Klassichen Zeit (Emsdetten 1931) 23–5, 38 ff.Google Scholar; Webster, , The Greek chorus (London 1970)Google Scholar passim. Its use to accompany solo performances is not so well attested, cf. Huchzermeyer 25 f., 55 n. 222.

⁷⁴ Cf. Huchzermeyer (n. 73) 13 with n. 54; Bowra, C. M., Early Greek elegists (Cambridge Mass. 1935) 6 Google Scholar. The Armenian link is treated as doubtful by Schere, A., ‘Die Sprache des Archilochos’, Fond. Hardt x (1964) 90 Google Scholar and Chantraine, P., Dict. et. de la langue grecque (Paris 1968–1967) 334 Google Scholar s.v. ἔλεγος, but neither rejects it, and both admit the possiblity of a borrowing from Phrygian, cf. Hommel, H., Rh. Mus. lxxxviii (1939) 194 Google Scholar. Martin West has suggested to me that the Armenian vowel e in ełegn always derives from a diphthong, so that only with difficulty can it be related to the Greek ε. But (as I am informed by Prof. C. J. F. Dowsett, to whom I am grateful for this help) whereas in ełegn the first e is short, the second long, both es are short in its (irregular) genitive ełegan, the adjectival ełegneay, and one form of the diminutive ełegnik. The irregularities lead Prof. Dowsett to suspect that ełegn is a loanword in Armenian.

⁷⁵ For the hypothesis that the (developed) aulos came from Phrygia cf. Huchzermeyer (n. 73) 35–7.

⁷⁶ For Olympus cf. n. 56. For Marsyas e.g. Herodotus vii 26; Xen. Anab. i 2.8; Paus. ii 7.9.

⁷⁷ 1134a: ἐν ἀρχῇ γὰρ ἐλεγεῖα μεμελοποιημένα οἱ αὐλῳδοὶ ᾗδον· τοῦτο δὲ δηλοῖ ἡ τῶν Παναθηαίων ἀναγραφὴ ἡ περὶ τοῦ μουσικοῦ ἀγῶνος (though cf. Rosenmeyer's scepticism, above n. 54).

⁷⁸ 13. By contrast Bowra, C. M., ‘Xenophanes, fr. 3’ CQ xl (1944) 119–26Google Scholar (repr. as ‘Xenophanes on the luxury of Colophon’ in On Greek margins [Oxford 1970] 109–21Google Scholar) ad fin. recognised the variety in scale and character of early elegy and suggested that Mimnermus' Smyrneis might have parallels in the work of Semonides, Panyassis and Xenophanes. Jacoby, F. ‘Some remarks on Ion of Chios’, CQ xli (1947) 4 Google Scholar n. 6 ( = Abhandlungen zur griechischen Geschichtsschreibung [Leiden 1956] 149 Google Scholar n. 27) notes that Semonides may have written a verse ktisis and that Panyassis certainly did, but he regards the work ascribed by Diogenes Laertius to Xenophanes as a forgery, cf. n. 99.

⁷⁹ ibid. 74. Two books are attested by Porphyrio on Horace Epist. ii 2.21: Mimnermus duos libros luculent(is vers)ibus scripsit.

⁸⁰ Call. Aet. fr. i 11–12:

τοῖν δὲ δυοῖν Μίμνερμος ὅτι γλυκύς, αἱ κατὰ λεπτὸν

[ῥήσιες] ἡ μεγάλη δ' οὐκ ἐδίδαξε γυνή.

Colonna, A., ‘Mimnermo e Callimaco’, Athenaeum xxx (1952) 191–5Google Scholar was the first to set out a case for seeing the Smyrneis as Mimnermus' tall lady, following an idea of Corte, F. Della, ‘La Nannò di Mimnermo’, Atti dell' Acc. Ligur. iii (1943) 1–11 Google Scholar. He saw that the Smyrneis should be a ktisis poem like that of Xenophanes (n. 100), although he may go too far in suggesting that Smyrna actually claimed foundation by an Amazon on the basis of Strabo xii 550C: εἴη γὰρ ἂν λέγων τὴν ὑπὸ τῶν Αἰολέων καὶ Ἰώνων οἰκισθεῖσαν ὒστερον, πρότερον δ' ὑπὸ Ἀμαӡόνων· καὶ ἐπωνύμους πόλεις τινὰς εἷναί φασι, καὶ γὰρ Ἔφεσον καὶ Σμύρναν καὶ Κύμην καὶ Μύριναν. Given the more complex story told by Strabo at xiii 633–4 (n. 82) οἰκισθεῖσαν here may mean no more than ‘settled’.

⁸¹ Alan Cameron has suggested to me that there is equal wit in alluding to the personal name Nanno with its dwarf-overtones by the expression ‘tall lady’. That leaves unexplained the appropriateness of referring to the Smyrneis by the phase αἱ κατὰ λεπτὸν [ῥἠσιες].

⁸² The story as told by Strabo xiv 633–4C is complicated. Ephesus was originally called Smyrna, after an Amazon who captured it, and there was still a part of Ephesus in Strabo's day called Smyrna. The inhabitants of this part of Ephesus marched on the later site of (Old) Smyrna, ejected the Leleges who were its occupants, and after later expulsion themselves by Aeolians recovered it with the help of Colophonians. How much of this goes back to an early source is debatable—it does not tally with Herodotus i 150, cf. Podlecki (n. 21) 58, and for full discussion see Klügmann, , ‘Ueber die Amazonen in den Sagen der kleinasiatische Städte’, Philol. xxx (1870) 524–56Google Scholar. I am not persuaded by the sceptical argumenta e silentio of P. Devambez, ‘Les Amazones et l'Orient’, RA (1976) 265–80, that Amazons were first associated with foundation of Asia Minor cities in the mid-fifth century: we have too few vases and monuments to make them compelling, and Devambez has to play down Pindar's reference (ap. Paus. vii 2.7, fr. 174 Snell-Maehler) and leave unexplained that of Hecataeus F 226 (cited 275 n. 1) which secures the association for the end of the sixth century. A seventh-century poet could have given the Amazon Smyrna some coverage in his ktisis, and Callimachus traded on this allusively.

⁸³ Paus. ix 29.4 = Mimnermus 13.

⁸⁴ P. Univ. Mediolan. 17 col. ii 26 (p. 83 Wyss): ]σ[υνάγε]ιν δμω[ῆ]ις ἐνδέξεται (Antim. fr. 180). ἀντὶ τοῦ ἐπ[ιτ]άξηι Μίμνερμ[ος] δ[' ἐν] τῆι Σμυρνῃἴδι· ( = Mimnermus fr. 13a)

ὣς οἳ πὰρ βασιλῆος, ἐπε[ί ῥ'] ἐ[ν]εδέξατο μῦθọṿ,

ἤ[ϊξ])ᾳν κοίληι[ς ἀ]σπίσι φραξάμενοι.

⁸⁵ Solon 13, also opening with an invocation to the Muses, is a poem of only 76 lines. But the invocation is not a standard poetic call for inspiration—Solon adapts that motif and prays to the Muses to give him olbos.

⁸⁶ Cook, J. M., Charisterion eis A. K. Orlandon i (1965) 148–52Google Scholar, cf. West 74, Podlecki (n. 21) 60 (who leaves the question open).

⁸⁷ For the motif οἶοι νῦν βροτοί εἰσιν cf. Griffin, below pp. 37–8.

⁸⁸ It is also possible that the whole fragment is from a speech within a battle narrative: note the similarity to Agamemnon's speech to Diomedes, Il. iv 370 ff. (cited by Campbell).

⁸⁹ We do not have enough hexameter poetry attempting the same task (e.g. Eumelus?) to act as a control on our few remains of this sort of elegy.

⁹⁰ Strabo xiv 634 = Mimnermus 9. These lines do not obviously support Strabo, but cohere better with Herodotus i 150. The phrase ἀργαλέης ὕβριος ἡγεμόνες seems odd in Mimnermus' own mouth: is this too from a speech within narrative? (So already Gentili 67, following Tsagarakis' 1966 dissertation arguing from the implication of the first person plural: but cf. n. 95.)

⁹¹ Suda iv 610.5 A: Τυρταῖος Ἀρχεμβρότου, Λάκων ἤ Μιλήσιος, ἐλεγοποιὸς καὶ αὐλητής… ἔγραψε πολιτείαν Λακεδαιμονίοις καὶ ὑποθήκας δι' ἐλεγείας καὶ μέλη πολεμιστήρια, βιβλία ε'. It is possible, but unlikely, that δι' ἐλεγείας is to be taken with both πολιτείαν and ὑποθήκας.

⁹³ The phrase ὑπὸ τὸν Μεσσηνιακὸν πόλεμον in Aristotle is easier to refer to the first (eighth century) war than to that of the mid-seventh century (which Aristotle could be classifying as an ἀπόστασις rather than πόλεμος). But some take the reference to be to the seventh-century war and Tyrtaeus' own time, e.g. Cartledge, P. A., Sparta and Lakonia (London 1979) 127 Google Scholar (putting the stasis after Hysiae, 669 BC), 134; Murray, O., Early Greece (Hassocks 1980) 163 Google Scholar. Podlecki (n. 21) 104–5 does not make it clear which war he thinks is in question.

⁹⁴ The round number of 20 arouses suspicion. But that it is a conventional figure does not make it less interesting that Tyrtaeus should give it at all.

⁹⁵ Cf. Tsagarakis 23 with n. 9: but we cannot insist, as he does, that ‘if the poet follows traditional usage’, the first person plural requires us to take the speaker as a participant in the event. Podlecki (n. 21) 103 gives the Eunomia/Politeia a similar shape, but takes the exhortation of 2 to be from a ‘patriotic exhortation’ and seems to treat the whole poem as exhortatory.

⁹⁶ It is possible that some or all of fragments 19–23a are from battle scenes in the Eunomia rather than from exhortatory elegies—their assignation to the latter category is based on future first person plural (19.11, 12; 20.15) and third person plural verbs (19.16, 18, 20; 23.13; 23a.20), but the specific references to enemies (Messenians 23.6; Arcadians and Argives 23a.15) find no parallel in the manifestly exhortatory poems, and the reference to the three Dorian tribes as fighting separately (19.8) causes problems if the context is Sparta of the mid-seventh century: cf. Podlecki (n. 21) 97 (himself opting for the seventh century). That all are from speeches set within a battle narrative is of course improbable, but some may be.

⁹⁷ Suda iv 363.IA Σιμωνίδης. Κρίνεω, Ἀμοργῖνος, ἰαμβογράφος. ἔγραψεν ἐλεγείαν ἐν βιβλίοις β', ἰάμβους. γέγονε δὲ καὶ αὐτὸς μετά ọ' καὶ υ' ἔτη τῶν Τρωικῶν. ἔγραψεν ἰάμβους πρῶτος αὐτὸς κατά τινας; and 360.7 τῆς Ἀμοργοῦ ἐστάλη καὶ αὐτὸς ἡγεμὼν ὑπὸ Σαμίων. ἔκτισε δὲ Ἀμοργὸν εἰς τρεῖς πόλεις, Μινῴαν, Αἰγιαλόν, Ἀρκεσίνην. γέγονε δὲ μετὰ υς' (υọ' Ε) ἔτη τῶν Τρωικῶν. καὶ ἔγραψε κατά τινας πρῶτος ἰάμβους, καὶ ἄλλα διάφορα, ἀρχαιολογίαν τε τῶν Σαμίων. On the problem of Semonides' date and works see now Pellizer, E., ‘Sulla cronologia, la vita e l'opera di Semonide Amorgino’, Quad. Urb. Cult. Class. xiv (1983) 17–28 Google Scholar.

⁹⁸ It is conceivable that the archaiologia was a forgery, or that one or both books of elegiac poetry consisted of short elegies. But that so much as one, far less two books of short elegies by Semonides were circulating in antiquity is improbable given that the remains of only one elegiac poem are known to us, and that is more often ascribed to Simonides of Ceos than to the poet of Amorgos. The identity of the archaiologia with the two books of elegies, taken as certain by Bowra (n. 78) 120, is more cautiously treated as a guess that may well be right by H. Lloyd-Jones, Females of the species (Park Ridge 1975). Jacoby had been characteristically sceptical (FGrH IIIB [Leiden 1955] Kommentar, Notes to p. 456, n. 8a) about the equation of the archaiologia with the elegy, but begs a number of questions when (456) he says it ‘scheint sich speziell auf die kolonisation von Amorgos bezogen zu haben. Der titel ist sicher spät und deckt vielleicht nicht mehr als eine elegie etwa vom umfang von Tyrtaios' Eunomia oder Solons Salamis’.

⁹⁹ D. L. ix 20: ἐποίησε δὲ καὶ κολοφῶνος κτίσιν καὶ τὸν εἰς Ἐλέαν τῆς Ἰταλίας ἀποικισμὸν ἔπη δισχίλια. (The work is taken to be in hexameters by Campbell 332; Fraenkel, H., ‘Xenophanesstudien’, Hermes lx (1925) 174–5Google Scholar (although Fraenkel saw the analogy with Panyassis' Ionica). If what has been suggested above about the content of Mimnermus' and Tyrtaeus' poetry is correct Xenophanes was not innovating in giving ‘dem Epos einen modernen Inhalt’. Jacoby (n. 78) was persuaded by the suggestion of Hiller, E. ‘Die literarische Tätigkeit der sieben Weisen’, Rh. Mus. xxxiii (1878) 518–29Google Scholar, that the ascription by Diogenes Laertius of a 2000 line work to Xenophanes was due, like that of works of specified length to the seven sages by Diogenes, to the mendacious περὶ ποιητῶν of Lobon of Argos (2nd cent, BC?: cf. West, M. L., The Orphic poems [Oxford 1983] 60 Google Scholar n. 85); Hiller's suggestion was also accepted by Croenert, W., ‘De Lobone Argivo’, χάριτες Leo (Berlin 1911) 123–45Google Scholar. However, poetic activity, both in hexameters and in elegiacs, is securely attested for Xenophanes, and this could be a reliable tradition deriving from the pinakes of Callimachus (which Lobon is thought to have been spoofing).

¹⁰⁰ Bowra (n. 78) 121 had already noted that the term ἔπη did not prevent this being an elegiac poem. For early elegists' use of the term epe to describe their work cf. West 7. For a second century AD use meaning no more than ‘lines of writing’ (LSJ) cf. Lucian de hist, conscr. 28. For a contested use cf. n. 112.

¹⁰¹ Elea was occupied not by Colophonians but by Phocaeans, Herodotus i 167, but Xenophanes may have left Colophon when it was captured by Harpagus (id. 169: 545 or 540 BC) and his movements thereafter are uncertain. Fr. 8 claims 67 years of wandering which started at the age of 25. For a discussion of the chronology of his life and work cf. Steinmetz, P., ‘Xenophanesstudien’, Rh. Mus. cix (1966) 13–73 Google Scholar, esp. sec. I Zur Datierung 13–34.

¹⁰² E.g. Bowra (n. 78) 120, H. Fraenkel (n. 99) 174. See Matthews, V. J., Panyassis of Halikarnassos (Leiden 1974) 11–19 CrossRefGoogle Scholar (chronology and life) 26–31 (the Ionica) for full discussion: he attributes frr. 24, 25 and 29 to the Ionica, but at no point discusses how the Alexandrians might have divided the 7000 lines into books. Panyassis is not, however, mentioned in West's edition or Studies.

¹⁰³ For Simonides' Salamis see West's edition ii 112–13.

¹⁰⁴ See von Blumenthal, A., Ion von Chios (Stuttgart and Berlin 1939)Google Scholar; F. Jacoby (n. 78) 4–7. The arguments are (1) The work is cited among Ion's prose works in the list given by the scholiast on Ar. Pax 835. But this scholion is clearly conflating more than one source, and its phraseology does not prove that the ktisis was grouped with the prose works: ἔγραψε δὲ καὶ κωμῳδίας καὶ ἐπιγράμματα καὶ παιᾶνας καὶ ὕμνους καὶ σκόλια καὶ ἐγκώμια καὶ ἐλεγεῖα, καὶ καταλογάδην τὸν πρεσβευτικὸν λεγόμενον, ὃν νόθον ἀξιοῦσί τινες, καὶ οὐχὶ αὐτοῦ. φέρεται δὲ αὐτοῦ καὶ κτίσις καὶ κοσμολογικὸς καὶ ὑπομνήματα· καὶ ἄλλα τινά. (2) It is called a σμγγραφή by Paus. vii 4.8: this certainly suggests a prose work, but the term ξυγγραφεύς is occasionally found of writers of verse works, e.g. Dio Pr. 12.5 ξυγγραέας ἡδίστους ἐμμέτρων καὶ ἀμέτρων λόγων. (3) The only extant verbatim fragment is in prose, fr. 19, from Etym. Magn. and Orion s.v. ‘λόγχας’. λόγχας τὰς μερίδας Ἴωνες λέγουσιν. Ἴων ἐν χίου κτίσει· ἐκ τῆς Τέω λόγχης λόγχας ποιεῖ πεντήκοτα. The MSS offer ποιεῖν, which was corrected to ποιεῖ ν' by Gaisford (1615 = 569.34). But the present tense ποιεῖ is not appropriate to a historical narrative (other fragments of Ion's works show no penchant to the historic present), and we ought to consider πόρε πεντήκοντα which would be easy palaeographically. If Ion wrote πόρε, then he wrote a hexameter:

ἐκ τῆς Τέω λόγχης λόγχας πόρε πεντήκοντα (with ). For πόρε in this sedes cf. Il. i 72; vii 146; ix 663; x 546; xv 441; xvi 690; xxii 472; xxiii 92; Od. iv 130; v 321, 372; vi 228; x 394, 494; xi 282. For πεντήκοντα at the end of a hexameter Il. ii 719; v 786; xi 679; Od. ix 35, 48; xvi 247.

If the text of fr. 19 were secure, the matter would be clear: unfortunately it is not, but the possibility of a verse ktisis must be kept open. If the ktisis was in verse, then the pentameter quoted by Plutarch Thes. 20.2 (= Ion fr. 29) could belong to it:

(Chios) τόν ποτε Θησεΐδης ἔκτισεν Οἰνοπίων

Jacoby thinks the line indicates an allusive reference and ‘shows clearly that the elegy … did not narrate the foundation of Chios at length’. That seems to go too far. So summary a reference would not be out of place in an opening sequence (which in turn would be more likely to be known to Plutarch or his source).

¹⁰⁵ Ancient testimony on sympotic practice gives an important place to equality (or at least comparability) of contributions. Note that in the literary elaboration of Plato's Symposium the longest speech (that of Socrates) occupies less than 400 lines of the OCT.

¹⁰⁶ As proposed by Wade-Gery and supported with several arguments by West, M. L., Hesiod. Theogony (Oxford 1966) 44–5Google Scholar.

¹⁰⁷ W. Rösler has argued (in a paper forthcoming in Sympotica) that some of the fragments I have been considering (Tyrtaeus 2, 4 and 5; Mimnermus 9 and 14) belong to a sort of sympotic elegy that focussed on the past achievements of the community. Although that is a satisfactory explanation of these fragments, the additional evidence I adduced seems to me to favour rather my reconstruction.

¹⁰⁸ E.g. Eumelus' Korinthiaca: surviving fragments (in Kinkel) relate only to the ‘mythical’ period. For Eumelus' date cf. Janko, R., Homer, Hesiod and the hymns (Cambridge 1982) 231–3.Google Scholar Martin West suggests to me that Asius and Hegesinous' Atthis (Kinkel ep.gr.fr. 208) may be added.

¹⁰⁹ Plato Ion 532a-b—ΣΩ. Οὐκοῦν σὺ φῂς καὶ Ὅμηρον καὶ τοὺς ἄλλους ποιητάς, ἐν οῖς καὶ Ἡσίοδος καὶ Ἀρχίλοχός ἐστιν, περί γε τῶν αὐτῶν λέγειν, ἀλλ' οὐχ ὁμοίως, ἀλλὰ τὸν μὲν εὖ γε, τοὺς δὲ χεῖρον;—ΙΩΝ. καὶ ἀληθῆ λέγω.—ΣΩ.Οὐκοῦν, εἔπερ τὸν εὖ λέγοντα γιγνώσκεις, καὶ τοὺς χεῖρον λέγοντας γιγνώσκοις ἄν ὅτι χεῖρον λέγουσιν.—ΙΩΝ. Ἔοικέν γε.— I do not think we can infer from this passage (as does Podlecki [n. 21] 52) that Archilochus' poetry was actually part of Ion's repertoire as a rhapsode.

¹¹⁰ Dio Pr. lx.1 (= Archilochus 286): φασὶ γὰρ οἱμὲν τὸν Ἀρχίοχον ληρεῖν ποιοῦντα τὴν Δηιάνειραν ἐν τῷ βιάӡεσθαι ὑπὸ τοῦ Κενταύρου πρὸς τὸν Ἡρακλέα ῥαψῳδοῦσαν, ἀναμιμνήσκουσαν τῆς τοῦ Ἀχελῴου μνηστείας καὶ τῶν τότε γενομένων, ὥστε πολλὴν σχολὴν εῖναι τῷ Νέσσῳ ὅτι ἐβούλετο πρᾶξαι. Cf. frr. 287–9: narrative of the deeds of Heracles would be appropriate entertainment at a festival associated with Heracles, whose cult was later important in Thasos (as noted by Podlecki [n. 21] 520).

¹¹¹ Archilochus 304, 305.

¹¹² Notopoulos, J. A., ‘Archilochus, the aoidos’, TAPhA xcvii (1966) 311–15Google Scholar. One text he cites that might seem to point to hexameter poetry (as indeed he takes it) is Theocritus epig. 21.5–6 (of Archilochus): ὥς ἐμμελής τ' ἐγένετο κἠπιδέξιος ἔπεά τε ποιεῖν πρὸς λύραν τ' ἀείδειν. But ἔπεα can denote any dactylic verse.

Note that we have no evidence for historical narrative in Archilochean elegiacs of the sort I argue to be represented by the Smyrneis and Eunomia. For recent history Archilochus seems to have created poems of considerable length in trochaic tetrameters, to judge from the scale of treatment and continuous narrative implied by the surviving (largely epigraphic) excerpts (88–106, and some perhaps of 107–15).

¹¹³ For a discussion of the relation between public banquets and symposia cf. P. Schmitt Pantel ‘Repas sacrificiel et symposion: deux modèles de banquets civiques’ in Sympotica.