¹ See Alpers, J., Hercules in bivio (diss. Göttingen 1912)Google Scholar; Waites, M.C., ‘Some features of the allegorical debate in Greek literature’, HSCP 23 (1912) 1–46, esp. 12–19Google Scholar; Galinsky, G.K., The Herakles Theme. The Adaptations of the Hero in Literature from Homer to the Twentieth Century (Oxford 1972) 101–3, 162Google Scholar; Levine Gera, D., ‘Lucian's choice: Somnium 6–16’, in Innes, D., Hine, H. and Pelling, C. (eds), Ethics and Rhetoric (Oxford 1995) 237–50.Google Scholar Further bibliography in Schmid, W. and Stählin, O., Geschichte der griechischen Literatur 1.3 (Munich 1940) 41 n.9Google Scholar (‘eines der einfluβreichsten Stücke der Weltliteratur’). For possible representations in the visual arts of antiquity, see Picard, C., ‘Représentations antiques de l'Apologue dit de Prodicos’, CRAI (1951) 310–22CrossRefGoogle Scholar; idem, ‘Nouvelles remarques sur l'Apologue dit de Prodicos: Héraclès entre le Vice et la Vertu’, RA 42 (1953) 10–41.

² Panofsky, E., Hercules am Scheidewege und andere antike Bildstoffe in der neueren Kunst (Studien der Bibliothek Warburg 18, Leipzig and Berlin 1930)Google Scholar; Reiterer, M., Die Herkulesentscheidung von Prodikos und ihre früh-humanistische Rezeption in der ‘Voluptatis cum Virtute disceptatio’ des Benedictus Chelidonius (diss. Vienna 1957)Google Scholar (non vidi); Harms, W., Homo viator in bivio. Studien zur Bildlichkeit des Weges (Munich 1970)Google Scholar; Reid, J.D. and Rohmann, C., Oxford Guide to Classical Mythology in the Arts, 1300–1990s 1 (Oxford 1993) 527–30Google Scholar; Galinsky (n.1) 198–9, 213–18.

³ Kahn, C., ‘Prodicus’, in Craig, E. (ed.), Routledge Encyclopedia of Philosophy 7 (London and New York 1998) 731–2.Google Scholar

⁴ See, for example, Blass, F., Die attische Beredsamkeit 1 (2nd edn, Leipzig 1887) 30–1Google Scholar (‘Paraphrase’, 30); Alpers (n.1) 16–22; Mayer, H., Prodikos von Keos und die Anfänge der Synonymik bei den Griechen (diss. Paderborn 1913) 8–12Google Scholar (‘steht fest, daβ der Stil nicht prodikeisch ist’, 12); Gigon, O., Kommentar zum zweiten Buch von Xenophons Memorabilien (Basel 1956) 60–3Google Scholar; Untersteiner, M., Sofisti. Testimonianze e frammenti 2 (2nd edn, Florence 1961) 179Google Scholar (‘sebbene lo stile non sia quello di Prodico …, tuttavia, si può ben credere che Senofonte riproduca la sostanza del pensiero di Prodico’); Guthrie, W.K.C., A History of Greek Philosophy 3 (Cambridge 1969) 277Google Scholar (‘We possess at least the content, if not the actual words, of an epideixis of Prodicus’); Ambrose, Z.P., ‘Socrates and Prodicus in the Clouds’, in Anton, J.P. and Preus, A. (eds), Essays in Ancient Greek Philosophy 2 (Albany, NY 1983) 130Google Scholar (‘only a summary of what Prodicus actually wrote’); Russell, D.A. (ed.), An Anthology of Greek Prose (Oxford 1991) 93Google Scholar (Xen. ‘doubtless re-works Prodicus’ treatment extensively'); McKirahan, R.D. Jr, Philosophy before Socrates. An introduction with Texts and Commentary (Indianapolis 1994) 365Google Scholar (‘summarized by Xenophon’). Demont, P., ‘Die Epideixis über die Techne im V. und IV. Jh.’, in Kullmann, W. and Althoff, J. (eds), Vermittlung und Tradierung von Wissen in der griechischen Kultur (Tübingen 1993) 181–209Google Scholar, goes so far as to say (197), ‘spricht sein [sc. Xenophons] Prodikos bekanntlich [!] mit Xenophons Worten über Xenophons Themen’.

⁵ Mayer (n.4) 12 excludes consideration of this passage from his treatment of ‘Die Synonymik des Prodikos’. Similarly, C.J. Classen ignores the evidence of Xenophon in his otherwise excellent treatment of the linguistic studies of Prodicus: ‘The study of language amongst Socrates’ contemporaries' in Classen, C.J. (ed.), Sophistik (Darmstadt 1976) 215–47Google Scholar, at 230–8 (an article that originally appeared in Proceedings of the African Classical Associations 2 (1959) 33–49). de Romilly, J., ‘Les manies de Prodicos et la rigueur de la langue grecque’, MusHelv 43 (1986) 1–18Google Scholar, likewise relies on the evidence of Plato for Prodicus' ‘obsession’ with distinguishing near-synonyms, without mentioning Xenophon. According to Reesor, M.E., ‘The Stoic ἴδιον and Prodicus' near-synonyms’, AJP 104 (1983) 124–33, at 130Google Scholar, ‘Our evidence for Prodicus' theory of language is found in Plato.’

⁶ Compare the Platonic ‘Socrates’ concluding the ‘Defence of Protagoras’ at Theaet. 168c: εἰ δ' αὐτὸς [sc. Protagoras] ἔζη, μεγαλειότερον ἂν τοῖς αὑτοῦ ἐβοήθησεν (the only occurrence of μεγαλεῖος in Plato).

⁷ That these are verbatim transcripts is clear from the fact that the texts are recorded in the Laconian dialect. For a discussion of some of the issues raised by these texts, see Colvin, S., Dialect in Aristophanes and the Politics of Language in Ancient Greek Literature (Oxford 1999) 65–7.Google Scholar

⁸ For similar occurrences, see 4.118–19, 5.18–19, 5.23–4, 5.47 (a treaty, fragments of which survive on stone; see IG I² 86), 8.18, 8.37 and 8.58. Thucydides also quotes the text of three letters (1.128.7, 1.129.3 and 1.137.4) in the course of his excursus on Pausanias and Themistocles, introducing the first two with τάδε. While there are good grounds for doubting the authenticity of these letters to and from the Great King, there is no reason to question Thucydides' belief in their genuineness; see Westlake, H.D., ‘Thucydides on Pausanias and Themistocles – a written source?’, CQ 27 (1977) 95–110, esp. 102–3.CrossRefGoogle Scholar

⁹ See Dover (HCT 4.385): ‘although I see no adequate reason to doubt that Nikias really did write a letter, he certainly did not write what is presented to us here in characteristic Thucydidean idiom’. Similarly Westlake, H.D., Individuals in Thucydides (Cambridge 1968) 190Google Scholar: ‘Its language and style are thoroughly Thucydidean, as is much of its reasoning.’ In his famous programmatic statement at 1.22 Thucydides indicates that the speeches that he records cannot be counted on to preserve τὴν ἀκρίβειαν αὐτὴν τῶν λεχθέντων. The pressing need for ‘eine Geschichte des Citats’, first identified over 100 years ago by E. Norden (Die antike Kunstprosa 1 (Leipzig 1898) 90 n.1), is still felt today, as V. Bers, Speech in Speech. Studies in Incorporated Oratio Recta in Attic Drama and Oratory (Lanham, MD et alibi 1997) 220 n.3, points out.

¹⁰ Nor is Herodotus, who introduces direct speeches indifferently with τοιάδε (e.g. 1.8.2, 1.60.4, 2.173.2), τάδε (e.g. 1.11.2, 1.30.2, 1.41.1) and ὧδε (1.115.2) and concludes them indifferently with ταῦτα (e.g. 1.33, 1.37.1, 1.41.1), τοιαῦτα (e.g. 1.9.1) and ταῦτά κηι (1.98.1, 5.40.2); cf. in particular 6.69.1 ὁ μὲν δὴ τοιαῦτα ἔλεγε, ἠ δὲ ἀμείβετο τοῖσδε and 3.21.1–2 ἔλεγον τάδε … λέγει πρὸς αὐτοὺς τοιάδε. Unfortunately, Bers's perceptive Speech in Speech (n.9) does not concern itself with historical or philosophical texts.

¹¹ See Table 2 (p. 189) in Buckler, J., ‘Xenophon's speeches and the Theban hegemony’, Athenaeum 60 (1982) 180–204.Google Scholar As Buckler notes (188), in some instances the introductory and concluding formulae for the same speech vary between forms like τοιάδε, τοιαῦτα and ὧδέ πως on the one hand and ταῦτα and τάδε on the other. Buckler's detailed and carefully argued paper is concerned only with the speeches in Hell. 6.3–7.5, but his conclusion (204), that ‘Xenophon is subjective and inconsistent in his attitude towards speeches in this portion of the Hellenika’, is surely valid for the Hell. as a whole.

¹² Similarly, in Demosthenes' (18.127) mocking quotation from Aeschines' peroration (3.260), τὰ τοιαῦτα is used in connection with something very close to the ipsissima verba.

¹³ Immediately before this, at 95d, ‘Socrates’ quotes an earlier passage of four lines from Theognis, the first two of which (35–6 West) are also quoted by the Xenophontic ‘Socrates’ at Mem. 1.2.20 and again at Symp. 2.4. This would seem to indicate a predilection on the part of the historic Socrates; for a similar indication concerning a passage from Hesiod, see below, n.48.

¹⁴ The order in which they are quoted (435, 434, 436–8, that is, with two consecutive ‘pentameters’) is found also on an ostracon, P.Berol. 12310 (Viereck, P., ‘Drei Ostraka des Berliner Museums’, in Raccolta di scritti in onore di Giacomo Lumbroso (Milan 1925) 253–5Google Scholar), dating to the second half of the third century BC. Presumably the ostracon derives from a text of Plato, rather than of Theognis; so van Groningen, B.A. (ed.), Théognis. Le premier livre (Amsterdam 1966) 175.Google Scholar

¹⁵ Additionally, a minority of the Homeric manuscripts preserve a variant ἐς τόν (for the Attic ἐς τόν), which is to be preferred: J. Russo in Russo, Fernández-Galiano, M. and Heubeck, A., A Commentary on Homer's Odyssey 3 (Oxford 1992) 28.Google Scholar

¹⁶ L'Homère de Platon (Liège 1949) 207–10.

¹⁷ For the details, and a valuable discussion, see Labarbe (n.16) 101–8.

¹⁸ As Labarbe (n.16) 108 notes, Ion does not correct his interlocutor's lapse.

¹⁹ Burnet, J. (ed.), Plato's Euthyphro, Apology of Socrates and Crito (Oxford 1924)Google Scholarad loc. Further in the same note Burnet says that ‘the formulation of the charge put into the mouth of Socrates in this passage differs considerably’ from the official version.

²⁰ That the version given at Diog. Laert. 2.40 preserves the authentic wording of the indictment is well argued by De Strycker, E. and Slings, S.R. (eds), Plato's Apology of Socrates (Mnemosyne Suppl. 137, Leiden 1994) 84–5.Google Scholar

²¹ For this term, which denotes essentially a lexeme that is neither prepositive nor postpositive, but can occur anywhere in a sentence, see Dover, K.J., Greek Word Order (2nd edn, Cambridge 1968) 12–14.Google Scholar

²² Mem. 2.1.21. This is presumably the same written work referred to by Plato, Symp. 177b: εἰ δὲ βούλει αὖ σκέψασθαι τοὺς χρηστοὺς σοφιστάς, Ἡρακλέους μὲν καὶ ἄλλων ἐπαίνους καταλογάδην συγγράφειν, ὥσπερ ὁ βέλτιστσος Πρόδικος. According to schol. Ar. Nub. 361 (= 84 B 1 DK) the work was entitled Horae: φέρεται δὲ καὶ Προδίκου βιβλίον ἐπιγραφόμενον ῟Ωραι, ἐν ὦι πεποίηκε τὸν Ἡρακλέα τῆι Ἀρετῆι καὶ τῆι Κακίαι συντυγχάνοντα …

²³ 3.104.4–5. For Plato, see the convenient ‘Index of Quotations’ in Brandwood, L., A Word Index to Plato (Leeds 1976) 991–1003.Google Scholar In the orators, apparently, poetic quotations do not occur until after the middle of the fourth century, when Aeschines sets the fashion: Aeschin. 1.128–9, 144–50, 151–2, 2.144, 158, 3.135; Dem. 19.243, 245, 247, 255; Lycurg. Leocr. 92, 100, 103, 107, 109, 132.

²⁴ Isocrates' citations from his own work in his Antidosis are clearly a special case and are not relevant to our concerns here.

²⁵ Prot. 320c–22d, Theaet. 166a–68c, Phdr. 230e–34c. For a recent attempt to argue that the myth in Prot. is genuinely Protagorean, see Morgan, K.A., Myth and Philosophy from the Presocratics to Plato (Cambridge 2000) 132–54CrossRefGoogle Scholar, and note the suggestive confirmatory argument by Fowler, R.L., ‘Herodotos and his contemporaries’, JHS 116 (1996) 86–7.CrossRefGoogle Scholar

²⁶ For an excellent account of the range of possibilities, see Thomas, R., ‘Prose performance texts: epideixis and written publication in the late fifth and early fourth centuries’, in Yunis, H. (ed.), Written Texts and the Rise of Literate Culture in Ancient Greece (Cambridge 2003) 162–88CrossRefGoogle Scholar, esp. 180–5; similarly, eadem, Herodotus in Context. Ethnography, Science and the Art of Persuasion (Cambridge 2000) 254: ‘Prodicus' story of the choice of Heracles, for instance, was often presented … and circulated in a written version, though it cannot necessarily be assumed that this meant Prodicus actually read it out from the text in front of his audience.’ In fact, Aristotle says explicitly (Rhet. 1415b15–17 = 84 A 12 DK) that whenever his audience began to nod off, Prodicus would inject a little something from his 50-drachma lecture. It is, therefore, quite likely that any given oral presentation differed both from the prepared (written?) version and from other oral presentations of ‘the same’ lecture. Compare the discernible discrepancies between Demosthenes' and Aeschines' references to each other's words and the surviving written speeches: Dover, K.J., Lysias and the Corpus Lysiacum (Berkeley and Los Angeles 1968) 168–9.Google Scholar

²⁷ We can assume from the wording of Pl. Apol. 19e that Prodicus was still alive in 399, although we do not know how far beyond that year he lived. There is no reason to have confidence in Philostratus' notice (VS 1.12 = 84 A la DK), that ‘Xenophon the son of Gryllus, when he was a prisoner in Boeotia, used to listen to Prodicus’; see Breitenbach, H.R., RE 9A.2 (1967) 1571–2.Google Scholar

²⁸ So Gigon (n.4) 62. The Platonic ‘Socrates’ in fact claims (Crat. 384b) that he heard Prodicus' one-drachma epideixis (but not the 50-drachma one); refers to a conversation with Prodicus (Phdr. 267b); represents himself as having very often heard (μυρία τινὰ ἀκήκοα) Prodicus distinguishing the meanings of words (Chrmd. 163d); and even considers himself a pupil of Prodicus (Prot. 341a, Meno 96d).

²⁹ Or hear someone give a reading of it, as ‘Socrates’ is said to have heard Zeno read from his writings at the home of Pythodorus (Pl. Parm. 127c)? For statements or implications about ‘Socrates’ reading, see Pl. Phd. 98b (Anaxagoras), Theaet. 152a (Protagoras), Gorg. 462b–c (Polus), Xen. Mem. 1.6.14 (τοὺς θησαυροὺς τῶν πάλαι σοφῶν ἀνδρῶν, οὓς ἐκεῖνοι κατέλιπον ἐν βιβλίοις γράψαντες, which ‘Socrates’ and his friends read together and excerpt).

³⁰ Or rather, a portion of the written text. For, even if the Aristophanic scholiast (n.22) had not told us so explicitly (προσκλῖναι τῦι Ἀρετῦι τὸν Ἡρακλέα καὶ τοὺς ἐκείνης ἰδρῶτας προκρῖναι τῶν προσκαίρων τῆς κακίας ήδονῶν), we would be confident that Prodicus' tale continued on, after the point at which Xenophon breaks off, to include an indication of Heracles' choice between the two women.

³¹ Thus Alpers (n.1) 17: ‘Omnino fabulae Prodiceae, quam dialecto Ionica scriptam fuisse veri simile est, genus dicendi mutare debebat Xenophon, qui Attico utebatur sermone.’

³² For Gorgias and his influence, see Blass (n.4) 1.56, Schmid and Stählin (n.1) 1.3.63 and 96. For an excellent account of Thrasymachus and his oratory, see White, S.A., ‘Thrasymachus the diplomat’, CP 90 (1995) 307–27.Google Scholar It is not clear whether the Attic elements should be purged from the fragments of Diogenes of Apollonia; see Laks, A. (ed.), Diogene d'Apollonie. La dernière cosmologie présocratique (Lille 1983) xv.Google Scholar According to Dover, K., The Evolution of Greek Prose Style (Oxford 1997) 85–6Google Scholar, Zeno of Elea wrote in Attic, but I am aware of no evidence that would support Dover's claim.

³³ Untersteiner, M., The Sophists (English trans., Oxford 1954) 207Google Scholar; cf. Mayer (n.4) 12: ‘für uns steht fest, daβ der Stil nicht prodikeisch ist’ (Mayer's emphasis). Rose, H.J., A Handbook of Greek Literature from Homer to the Age of Lucian (4th edn, London 1950) 258 n.19.Google Scholar The discussion in Gautier, L., La langue de Xénophon (Geneva 1911) 105–8Google Scholar, is somewhat more substantial, but comes to the paradoxical conclusion (a) that the passage is not essentially different from the language and style of Xenophon generally, and (b) that, ‘sciemment, pour des motifs stylistiques, Xénophon a semé ce morceau de termes rares et non attiques’ (107).

³⁴ Editors have generally followed Dindorf's conjectural introduction of θαμά in place of the manuscripts' ἅμα at Cyrop. 8.8.12. But even if this final chapter of Cyrop. is by Xenophon (as is forcefully argued by Levine Gera, D., Xenophon's Cyropaedia. Style, Genre, and Literary Technique (Oxford 1993) 299–300Google Scholar and Mueller-Goldingen, C., Untersuchungen zu Xenophons Kyrupädie (Stuttgart and Leipzig 1995) 262–71)CrossRefGoogle Scholar, the type of pleonasm seen in σύν + ἅμα is not at all uncommon: Mommsen, T., Beiträge zu der Lehre von den griechischen Präpositionen (Berlin 1895) 546–8Google Scholar, and add Soph. fr. 314.76 Radt (Ichn.).

³⁵ The total is 16 if in fact διέσηι (24; the reading of the manuscripts, but obelized by several editors) is correct. The verb διεῖναι is not attested elsewhere, but διαγίγνεσθαι is found in just the required meaning and with the same participial construction (Ar. Av. 45; Thuc. 5.16.1; Xen. Mem. 4.8.4); the future διαγενήσεσθαι, however, is not attested before the time of Epictetus (fr. 25 Schenkl) and Plutarch (Demetr. 49.7).

³⁶ So, for example, Mem. 3.1 has eleven sections and 3.2 has four, so the passage selected begins at the start of 3.3.6.

³⁷ The following are the hapax legomena found for each passage: 1: λίχνος, ὑπερεσθίειν, ἀλαζονικός, ἀμπεχόνη, ἐρασιχρήματος, ἀνδραποδιστής, προσβιβάζειν, διδασκαλικός, νουθετικός, ἐγκυλίνδειν; 2: ἄμμος, κυβερνητικός, ἐφάμιλλος, εὐανδρία, εὐφωνία, ἀρχαιρεσία, κατήκοος, ζημιώδης; 3: ποτέρωθι, ἁπλοίζεσθαι, ἄσκπτος, ἀγράμματος, χαλκεύειν, τεκταίνεσθαι, σκυτεύειν; 4: παράβλημα, συμπαριέναι, δίκροτος, μονόκροτος, τειχύδριον, ἀποτομή, πενθεῖν, εἴσπλους, μικροπολίτης; 5: συναπιέναι, ναυσίπορος, προσομνύναι, εὐθύωρος, δοῦπος; 6: ἐθελόπονος, μειονεξία, διακριβοῦν, γνώρισμα, ἀνίδρωτος.

³⁸ See Dover (n.32) 46–8; O'Sullivan, N., Alcidamas, Aristophanes and the Beginnings of Greek Stylistic Theory (Hermes Einzelschriften 60, Stuttgart 1992) 32–6.Google Scholar

³⁹ Note, e.g., ἡσυχία, αἰδώς, σωφροσύνη, ὑποψία, ἐξουσία, εὐδαιμονία, ἀλήθεια, ἐπιμέλεια, εὐφροσύνη, ἐπιθυμία, φιλία, ἀπόλαυσις, λήθη, μνήμη. In addition, there is a striking number of neuter adjectives used substantivally in this passage (e.g. τὸ ὅν, τὰ τερπνά, τὰ χαλεπά, τὰ καλὰ καὶ σεμνά, τὰ ἀγαθὰ καὶ καλά, τὰ ἡδέα, τὸ χρησιμώτατον, τὰ δέοντα), for which see Solmsen, F., Intellectual Experiments of the Greek Enlightenment (Princeton 1975) 110–25.Google Scholar

⁴⁰ See Long, A.A., Language and Thought in Sophocles. A Study of Abstract Nouns and Poetic Technique (London 1968) 12–14Google Scholar; Lyons, J., Semantics 2 (Cambridge 1977) 442–3.Google Scholar

⁴¹ For this terminology, see Dover (n.32) 26.

⁴² The quotation from Prodicus comprises 976 tokens; the total number of tokens in Xenophon is 312,317. I have used the figure for Xenophon provided by the ‘Perseus Greek Vocabulary Tool’ (http://www.perseus.tufts.edu/cgi-bin/vocab?lang=greek), which gives a slightly more accurate count than that given (321,305) by the ‘Thesaurus Linguae Graecae’, for reasons explained in Berkowitz, L. and Squitier, K.A., Thesaurus Linguae Graecae Canon of Greek Authors and Works (3rd edn, New York and Oxford 1990) xxvii.Google Scholar

⁴³ Heitsch, E., Die Entdeckung der Homonymie (AAWM 1972, 11, Mainz 1972) 28–9Google Scholar; O'Sullivan (n.38) 17–18.

⁴⁴ Rhet. 1407b6–8 = 80 A 27 DK. Diels-Kranz even print Nubes 658–79 as 80 C 3 (‘Imitation’). In their commentaries on Clouds K.J. Dover (Oxford 1968) and G. Guidorizzi (Milan 1996) refer to Protagoras; see also Kerferd, G. B., The Sophistic Movement (Cambridge 1981) 68–9.Google Scholar

⁴⁵ Willink, C.W., ‘Prodikos, “meteorosophists” and the “Tantalos” paradigm’, CQ 33 (1983) 25–33.CrossRefGoogle Scholar At line 361 the chorus names Prodicus as the only one τῶν νῦν μετεωροσοφιστῶν apart from Socrates to whose request they would respond. Aristophanes seems here to be reflecting the existence of a personal relationship between the historical Socrates and Prodicus that he could expect his audience to recognize; see Gomperz, H., Sophistik und Rhetorik (Leipzig and Berlin 1912) 93–6.Google Scholar

⁴⁶ For possible further associations between this passage from Birds and Prodicus, see Nestle, W., ‘Die Horen des Prodikos’, Hermes 71 (1936) 151–70Google Scholar, reprinted in Classen, Sophistik (n.5) 425–51, at 158 = 434 and 162 = 440–1. If Nestle is correct to derive (151–3 = 425–7) the title of Prodicus' work from the personified Horae who are connected with Aristaeus (for whose local cult on Prodicus' native Ceos see Burkert, W., Homo Necans (2nd edn, Berlin and New York 1997) 125–7CrossRefGoogle Scholar), we may see in Prodicus' work an inspiration for Aristophanes' chorus in Clouds, assuming that his Horae predates 423. In Homer (Il. 5.749–51 = 8.393–5) the Horae have charge of the thick cloud that covers the entrance to Olympus.

⁴⁷ So Gomperz (n.45) 101–2 n.225a.

⁴⁸ Critias supports his distinction between ἐργάζεσθαι and ποιεῖν in part with reference to a quotation from Hesiod (Op. 311 ἔργον δ' οὺδὲν ὄνειδος). This quotation comes from the same context as the quotation (Op. 287–92) that introduces the ‘Choice of Heracles’ and, in fact, it is the same quotation with which ‘the accuser’ challenges ‘Socrates’ at Mem. 1.2.56–7 and which ‘Socrates’ explains in a way that implicitly distinguishes between ἐργάζεσθαι and ποιεῖν. It would appear (a) that this Hesiodic passage provided the text on which Prodicus based his sermon (so Nestle (n.46) 164–5 = 443–4; E. Dupréel, Les Sophistes. Protagoras, Gorgias, Prodicus, Hippias (Neuchâtel 1948) 121). and (b) that the historical Socrates was influenced both by the Hesiodic text and the use to which Prodicus put it.

⁴⁹ Obviously there is no need to point out occurrences of εἶναι and γενέσθαι, which Prodicus distinguishes at Pl. Prot. 340b.

⁵⁰ The string continues with πειρατέον εὖ ποιεῖν, θεραπευτέον, ἐπιμελητέον. Other pairs of near-synonyms in close proximity include δοκειῖν and φαίνεσθαι (22), πορίζεσθαι and παρέχειν (25), κερδᾶναι and ὠφελεῖσθαι (25), ἐπίστασθαι and μανθάνειν (28).

⁵¹ Spengel, L., ΣΥΝΑΓΩΓΗ ΤΕΧΝΩΝ sive Artium Scriptores (Stuttgart 1828) 57–8.Google Scholar

⁵² This assumption is made, but not examined, by the authors of the works cited in n.5 above and by Untersteiner (n.33) 213–16; Kerferd (n.44) 70; O'Sullivan (n.38) 17 and others.

⁵³ Blass (n.4) 30–1. Blass is quoted approvingly by Joël, K., Der echte und der xenophontische Sokrates 2 (Berlin 1901) 130–1Google Scholar, who is in turn quoted approvingly by Alpers (n.1) 19–20, Mayer (n.4) 11 and Untersteiner (n.4) 178. In characteristically perverse and polemical fashion (and at astonishing length; see pp. 125–560 [sic!]), Joël attributes the ‘Choice of Heracles’ not to Prodicus or even to Xenophon, but to Antisthenes.

⁵⁴ Mayer (n.4) 22–5; similarly Alpers (n.1) 19–20.

⁵⁵ See Pl. Prot. 358a (ήδύ, τερπνόν, χαρτόν), Arist. Top. 112b22–3 = 84 A 19 DK (χαράν, τέρψιν, εὐφροσύνην), Stob. 4.20b.65 = 84 B 7 DK (ἐπιθυμίαν, ἔρωτα, μανίαν).

⁵⁶ See Rankin, H.D., Sophists, Socratics and Cynics (London 1983) 51Google Scholar; Heitsch (n.43) 24; de Romilly (n.5) 16; eadem, The Great Sophists in Periclean Athens (English trans., Oxford 1992) 235. For the conviction that binary opposition is characteristic of Prodicus' treatment of near-synonyms, see Wöβner, W., Die synonymische Unterscheidung bei Thukydides und den politischen Rednern der Griechen (diss. Berlin 1937) 8–9Google Scholar; Classen (n.5) 231–2 (‘two names … contrasting pairs … two similar words … two at a time’), 234 n.68 (‘there are no certain examples of a tripartite διαίρεσις which can be ascribed to Prodicus’); and especially Dumont, J.-P., ‘Prodicos: De la méthode au système’, in Cassin, B. (ed.), Positions de la sophistique (Paris 1986) 221–32.Google Scholar

⁵⁷ See Taylor, C.C.W., Plato. Protagoras, translated with notes (rev. edn, Oxford 1991) 137–40Google Scholar for a full and detailed discussion of the inconsistencies involved.

⁵⁸ This is supported by the Aristotelian text (Top. 112b22–4) on which Alexander is commenting.

⁵⁹ As far as I am aware, this fundamental point has been noted only by Rijlaarsdam, J.C., Platon über die Sprache (Utrecht 1978) 203.Google Scholar

⁶⁰ According to Hermias, Alexander and Aristotle (n.58), Prodicus prescribed rather the use of ήδονή as the generic term, a practice to which Virtue adheres in her use of the words ἡδονή (once), ἥδεσθαι (once) and ἡδύς (11 times).

⁶¹ Valckenaer (apud Pierson's edn of Moeris) deleted the offending word and Ruhnken (ibidem) transposed it to precede καλοῦσί με. LSJ, following the consensus of ancient lexicographers and scholia, give it a meaning unique to this occurrence in Xenophon: ‘reversely, call something good by a bad name’ (s.v. 1.3, with no adjustment in the 1996 Revised Supplement).

⁶² 2.1.23 ἰέναι τὸν αὐτὸν τρόπον. In this way Prodicus has skilfully made a virtue (so to speak) of necessity; cf. Froleyks, W.J., Der ΑΓΩΝ ΛΟΓΩΝ in der antiken Literatur (diss. Bonn 1973) 135Google Scholar and (for the convention that, in a fictionalized debate, the figure that the author wishes to commend speaks second) 386.

⁶³ Kerferd (n.44) 70; similarly Mayer (n.4) 37; O'Sullivan (n.38) 17; de Romilly, Great Sophists (n.56) 74–5; Rankin (n.56) 50; Reesor (n.5) 130–1.

⁶⁴ See nn.26 and 28 (one drachma and 50 drachmas). The pseudo-Platonic Axiochus (366c) gives figures of 4 obols (διμοίρου) and 2 and 4 drachmas. It is surprising that this author, whose familiarity with the Platonic corpus is so intimate, gives figures incompatible with Plato's. I am reluctant to believe that he has independent evidence and I incline to the view, rather, that he is slyly devaluing Prodicus' performance (which is further devalued by containing Epicurean doctines). The author of this dialogue seems to have had a mischievous sense of humour, one which would have pleased Plato himself; he represents Prodicus (368c) as denigrating agriculture, which appears to have been the main object of praise in Prodicus' Horae (see Nestle (n.46) 168 = 448). Otherwise, the portrait of Prodicus in the Axiochus seems to be an amalgam of what can be found in Pl. Prot. and Xen. Mem.: his epideixis opens with a quotation from Epicharmus (366c; cf. Mem. 2.1.20); it takes place at the house of Callias, the scene of his appearance in Pl. Prot.; ‘Socrates’ will report ταῦτα ἃ μνημονεύσω (366d; cf. Mem. 2.1.21 ὅσα ἐγώ μέμνημαι); the epideixis is concerned with the ‘stages of life’, which may in fact be what the title of Prodicus' Horae refers to; one of those stages is that at which a young man considers τίνα τις τοῦ βίου όδὸν ἐνστήσεται (367a; cf. Mem. 2.1.23 ἀποροῦντα ποίαν ὁδὸν ἐπὶ τὸν βίον τράπηι); the overall theme of the epideixis, the denigration of life, represents a clever appreciation of the ‘underworld’ associations that Plato heaps upon Prodicus at Prot. 315c–16a (in addition to the quotations from the Odyssean katabasis and Prodicus' location in a store-room, note that the buzzing of his voice, τῦς φωνῆς βόμβος, has spectral associations; cf. Soph. fr. 879 Radt βομνεῖ δὲ νεκρῶν σμῆνος). Finally, there is a delightful irony in that the overall theme, which sounds as though it comes from Hegesias ὁ πεισιθάνατος (Diog. Laert. 2.86, 93–6), is put into the mouth of Prodicus, whose ‘Choice of Heracles’ is used by ‘Socrates’ in an (apparently vain) attempt to reform Aristippus, to whose Cyrenaic school Hegesias belonged. For the relationship between Axiochus and Prodicus, see esp. Gomperz (n.45) 105–10.

⁶⁵ Thomas (n.26) 185.

⁶⁶ ‘An advertisement meant to arouse curiosity, sometimes by withholding part of the material information’, Webster's New International Dictionary of the English Language (2nd edn, 1934). Compare Gomperz (n.45) 111 n.241: ‘Im allgemeinen würden wir uns vorstellen, sie [sc. epideixeis] hätten für die Sophisten mehr die Bedeutung einer Reklame für die entgeltlichen Unterrichtskurse gehabt als die einer an sich selbst gewinnbringenden Erwerbstätigkeit.’

⁶⁷ For a similar situation in the medical profession in the fifth and fourth centuries, see L. Dean-Jones, ‘Literacy and the charlatan in ancient Greek medicine’, in Yunis (n.26) 97–121, at 120–1.

⁶⁸ See Willink (n.45) 30–1 for the wealth and ἁβρότης of Prodicus.

⁶⁹ The Orchestra (a place, according to Timaeus' Lexicon, s.v., ἐπιφανής εἰς πανήγυριν) was located in the middle of the Agora, near the Altar of the Twelve Gods and the statue of the Tyrannicides: J. Travlos, Pictorial Dictionary of Ancient Athens (New York 1971) 3 and fig. 29. The two most recent commentators on the Apology differ in their views of what could be purchased there for a drachma: E. Heitsch (Göttingen 2002) 111 n.199, ‘Die Orchestra, bei der man Bücher kaufen konnte …’; De Strycker and Slings (n.20) 308, ‘There is no suggestion whatsoever of the young men buying books.’

⁷⁰ Thomas (n.26) 180.

⁷¹ 337a–c: ἀμφισβητεῖν/ἐρίζειν, κοινός/ἴσος, εὐδοκιμεῖν/ἐπαινεῖσθαι, εὐφραίνεσθαι/ἥδεσθαι; at 340b, however, no explanations are given for the distinctions βούλεσθαι/ἐπιθυμεῖν and γενέσθαι/εἶναι.

⁷² Plato singles out Agathon, for example, as one of those present in the inner sanctum of Callias' house and, while the love-epigram (AP 5.78 = FGE ‘Plato’ III Page) purportedly composed by Plato for Agathon is undoubtedly a product of the early Hellenistic period, it does at least attest to a tradition that, plausibly enough, makes the two men friends. See Ludwig, W., ‘Plato's love epigrams’, GRBS 4 (1963) 59–82Google Scholar, who convincingly shows (68–72) that this poem cannot be an epigram by the philosopher Plato about Plato's relationship with the tragic poet Agathon. Aristotle's notice (n.26) that Prodicus would insert bits from his 50-drachma lecture into his one-drachma lecture could derive only from someone familiar with the contents of both. If Prodicus did not himself make this claim in his writings, a likely source for this information is Aristotle's teacher Plato.

⁷³ Theaet. 184c τὸ δὲ εὐχερὲς τῶν ὀνομάτων τε καὶ ῤημάτων καὶ μὴ δι' ἀκριβείας ἐξεταζόμενον τὰ μὲν πολλὰ οὐκ ἀγεννές; Plt. 261e κἂν διαφυλάξηις τὸ μὴ σπουδάζειν ἐπὶ τοῖς ὀνόμασιν, πλουσιώτερος εἰς τὸ γῆρας ἀναφανήσηι φρονήσεως; cf. Rijlaarsdam (n.59) 134.

⁷⁴ The written version that survives in Xen. Mem. is somewhat truncated, apparently lacking any introduction and omitting the presumable conclusion in which Heracles decided to follow Virtue, but is even so roughly comparable in length to Gorgias' seemingly complete epideixis, In Praise of Helen (82 B 11 DK): the ‘Choice of Heracles’ is 73 percent the length of Gorgias' Helen (although it is only 39 percent as long as Gorgias' Palamedes).

⁷⁵ And ensuring thereby its survival. Later writers know the piece from its appearance in Xen. Mem.: Cicero (Off. 1.118) refers to quod Herculem Prodicus dicit, ut est apud Xenophontem; Philostratus declines to characterize Prodicus' style, ‘since Xenophon provides a satisfactory transcription’ (VS 1.12 p. 496, Ξενοφῶντος αὐτὴν ἰκανῶς ὑπογράφοντος); and Athenaeus even attributes the ‘Choice of Heracles’ to Xen. (510c) or Socrates (544d).

⁷⁶ See 84 B 5 DK and, more fully, Prodicus fr. 5 Untersteiner (n.4); Phld. Piet., part 1, lines 520–5 Obbink; P.Herc. 1428 cols, ii 28–iii 13 and fr. 19 (from part 2 of De Pietate), with extensive discussion in Henrichs, A., ‘Two doxographical notes: Democritus and Prodicus on religion’, HSCP 79 (1975) 93–123Google Scholar, at 107–23, and ‘The atheism of Prodicus’, CronErc 6 (1976) 15–21. The late tradition (schol. Pl. Resp. 600c = Suda π 2365) that Prodicus was condemned to death by the Athenians and forced to drink hemlock for corrupting the young (no mention is made of atheism in this connection) is unreliable and should not be accepted.

⁷⁷ The first mention of the gods hints at an etymology deriving the word θεός from the root of the verb τίθημι (ἧιπερ οἱ θεοὶ διέθεσαν τὰ ὄντα διηγήσομαι μετ' ἀληθείας, 27). The same etymology is found in Herodotus, who says of the Pelasgians: θεοὺς δὲ προσωνόμασάν σφεας ἀπὸ τοῦ τοιούτου ὅτι κόσμωι θέντες … (2.52.1). W. Nestle, Vom Mythos zum Logos (2nd edn, Stuttgart 1942) 507, suggests that Hdt. took this etymology from Anaxagoras, comparing 59 B 12 DK πάντα διεκόσμησε νοῦς, where, however, neither θεός nor τίθημι appears.

⁷⁸ Henrichs, ‘Atheism’ (n.76) 21, from which the following quotation is taken.

⁷⁹ Henrichs, A., ‘The sophists and Hellenistic religion: Prodicus as the spiritual father of the Isis aretalogies’, HSCP 88 (1984) 139–58, at 142Google Scholar, (rightly) finds it hard to imagine how the ‘Choice of Heracles' fitted into a work that rationalistically explained the origins of religious worship. But must we assume that Prodicus' atheism was expounded in his Horae – Galen (84 B 4 DK) attributes a Περὶ φύσεως to him – or even in writing at all? It is true that Philodemus (84 B 5 DK) quotes Persaeus as approving τὰ περὶ 〈τοῦ〉 τὰ τρέφοντα καὶ ὠφελοῦντα θεοὺς νενομίσθαι και τετειμῆσθαι πρῶτον ὑπὸ Προδίκου γεγραμμένα. But we cannot be sure that Persaeus himself referred to ‘what was written by Prodicus’ (as opposed to, e.g., ‘what Prodicus said’). And even if he did refer to ‘writings’, we cannot know how much was explicit and how much was (perfectly justifiable) interpretation. See the sensible remarks of Kahn, Charles, ‘Greek religion and philosophy in the Sisyphus fragment’, Phronesis 42 (1997) 247–62CrossRefGoogle Scholar, at 261: ‘Prodicus' theory admits more than one interpretation … Later doxographers may have correctly diagnosed the atheism latent in his theory, but there is no hint of any legal charge of impiety being lodged against Prodicus. Living in Athens, Prodicus was perhaps also being cautious.’