In an article in an earlier issue of this journal Vivienne Gray sought to challenge my claim that Xenophon's account of Prodicus' narrative concerning the Choice of Heracles (Mem. 2.1.21–34) represents ‘a very close approximation to Prodicus’ actual wording'.Footnote ¹ Since that time, Gray's article has been cited approvingly by Louis-André Dorion and David Wolfsdorf, both of whom consider that Gray has settled the matter, at least as far as the linguistic aspect of my argument is concerned.Footnote ² In view of this, I feel it necessary to point out that Gray's argument is fundamentally flawed. In addition, I should like to take this opportunity to show that the arguments of Dorion and Wolfsdorf, which are aimed at other aspects of the case that I made for Xenophon's faithfulness to Prodicus' text, are likewise invalid.

In my 2004 article I demonstrated (pp. 131–2) that the passage in which Xenophon's Socrates recounts Prodicus' narrative contains a significantly large number of vocabulary items not attested elsewhere in the works of Xenophon, testing that claim by examining six passages of comparable length selected at random. The statistics showed (p. 132) that ‘the frequency (15) of hapax legomena in the “Choice of Heracles” is at least 50 percent higher than the largest number and more than twice the average number in the comparable passages’. Gray (p. 427) admits that the statistics ‘do look impressive’, but she alleges that ‘there are even more [sc. hapax legomena] in other passages, such as Memorabilia 3.10, which has around eighteen occurrences in under 900 words’. In a footnote she gives a list of nineteen [sic] words. Three of those words, however, occur twice in the passage (ζωτικός, ἄρρυθμος, ἀνάρμοστος), which would disqualify them from consideration according to most criteria by which hapax legomena are defined. A further three words are in fact attested elsewhere in the Xenophontine corpus.Footnote ³ This leaves the passage, contrary to Gray's claim, with only twelve hapax legomena, 20 per cent fewer than in the Choice of Heracles. Inasmuch as the passage chosen by Gray consists of Socrates' discussions of technical matters with a painter, a sculptor and a manufacturer of breastplates, it is difficult to see how this constitutes, as Gray claims (p. 427), a better comparison with the Choice of Heracles ‘than the more objective comparisons’ that were randomly selected by me, which on average display fewer than half the number of hapax legomena found in the Choice of Heracles.

Gray does not explain the basis on which she selected Memorabilia 3.10 for comparison, nor does she provide citations for the ‘other passages’ of which 3.10 is given as an example, so that we cannot know whether they, too, contain discussions of technical matters. In any event, Memorabilia 3.10 exhibits fewer hapax legomena than the Choice of Heracles, the content of which is by no means technical in nature. Thus, so far from undermining my case, Gray seems to have produced further support for it. For if a passage of a technical nature contains fewer hapax legomena than another passage of comparable length, the latter passage is demonstrably anomalous. Moreover, Gray appears not to have understood the point of my demonstration. She begins her discussion (pp. 426–7) by quoting Léopold GautierFootnote ⁴ in support of her assertion that the number of rare words in the Choice of Heracles is compatible with Xenophontine authorship, ‘because his regular style is precisely to employ rare words’. Gray goes even farther in her conclusion (p. 435), saying, ‘There is no evidence that Prodicus’ vocabulary contained rare words, but plenty of evidence that Xenophon's did.' This statement is, as far as it goes, perfectly correct: inasmuch as the Choice of Heracles is the only surviving text of reasonable length that claims to represent what Prodicus said, and inasmuch as it is precisely the authorship of this text that is sub judice, it cannot legitimately be counted as evidence for Prodicus' style. With regard to Xenophon, Gray's article provides a useful typology (pp. 427–9) of rare words in his writings.Footnote ⁵ But Xenophon's, or Prodicus', fondness for using rare words is irrelevant to my argument. The point of examining hapax legomena was to show that the Choice of Heracles exhibits a significantly higher concentration of words attested only once in Xenophon than is found in comparable passages, the implication being that the concentration found in the Choice of Heracles suggests that these are likely to be words not in Xenophon's vocabulary. The passages chosen for comparison provided a baseline against which to measure a passage whose authorship is in question. Indeed, if Xenophon's style is in fact characterized by a fondness for using words that occur elsewhere only rarely, my demonstration that the Choice of Heracles significantly exceeds Xenophon's normal practice is all the more telling.

David Wolfsdorf (n. 2) challenges a different aspect of the case made in my 2004 article. I had claimed that the Choice of Heracles accurately reflects the way in which Prodicus distinguishes among words that are often used with no apparent difference in meaning (pp. 134–40). I made the further claim that Xenophon faithfully represents Prodicus' intent with regard to a particular feature of the passage, namely that the personifications of Virtue and Vice differ in the care with which they distinguish among near-synonyms, Virtue being consistent and Vice being careless in her use of words. This was intended (pp. 136–7) to account for the apparent difference between this passage's use of words referring to pleasure (ἡδονή, τέρψις, χαρά, εὐφροσύνη) and the testimonia of later writers, who explicitly tell us by what criteria Prodicus distinguished among these very words. Wolfsdorf examines in great detail the evidence preserved by the fifth-century a.d. Platonic commentator Hermias, presenting the first English translation, not only of the testimonium pertaining to Prodicus, but of a good deal of the context in which it appears.Footnote ⁶ From this examination he concludes that there is reason to believe that Hermias' testimony, despite dating from a time nearly 1,000 years after the lifetime of Prodicus, is reliable. Regardless, however, of Hermias' accuracy in recording the thought of Prodicus, Wolfsdorf's understanding of Hermias' text is seriously undermined by a faulty translation of the relevant portion of the Greek. Here are Hermias' Greek (238.22–239.2 Couvreur = 250.24–251.1 Lucarini and Moreschini) and Wolfsdorf's translation:

By taking καλῶν as the genitive plural of the adjective καλόν (‘of fine things’), Wolfsdorf has removed any possible construction for the accusative τέρψιν. He makes good the deficiency that he has created by (apparently) supplying a word which he translates ‘saying’. But Wolfsdorf provides no parallels that would encourage us to believe that Hermias can leave such a word to be understood. In any event, there is no need to supply such a word. The word is καλῶν, which is not the genitive plural of the adjective meaning ‘fine’, but rather the participle of the verb καλῶ, agreeing with Πρόδικος.Footnote ⁸ What Hermias in fact says is:

Nor is this a minor matter. On the contrary, the conviction that καλῶν is to be construed as a genitive plural, defining the types of pleasures designated by the terms τέρψις, χαρά and εὐφροσύνη, is fundamental to Wolfsdorf's argument. According to Wolfsdorf ([n. 2], 140–2 and [n. 8], 12–13), Prodicus is here concerned only with what Wolfsdorf terms ‘refined pleasures’, that is, pleasures whose objects are ‘fine things’ (καλά; similarly Mayhew [n. 2], 127–8). This is, according to Wolfsdorf, in contrast to ‘vulgar or base pleasures’, generally referred to simply as ἡδοναί, which Prodicus does not bother to define. In fact, Wolfsdorf refers to ‘the absence of ἡδονή’ from the list of terms that Prodicus defines. While it is true that Prodicus does not, in the testimonium from Hermias, define ἡδονή, it is certainly not absent from the discussion. On the contrary, the terms τέρψις, χαρά and εὐφροσύνη are all explicitlyFootnote ¹⁰ defined as species of ἡδονή.

Wolfsdorf next turns to a discussion ([n. 2], 143–4) of the way in which these same terms for ‘pleasure’ are used in Prodicus' Choice of Heracles, saying appropriately that this passage would ‘appear to have been the occasion that motivated him to draw and deploy distinctions between base and refined pleasures’. Noting that all of the terms referred to in Hermias' testimonium appear also in the passage from Xenophon,Footnote ¹¹ Wolfsdorf goes on to say,

But it had been precisely my point (pp. 136–7) that Prodicus makes Virtue speak in a consistent and ‘correct’ manner, whereas Vice uses words indiscriminately and, further, that Xenophon has faithfully preserved this crucial aspect of the piece.Footnote ¹³ Indeed, I had used the very examples adduced by Wolfsdorf in making the case, saying (p. 136), ‘Virtue's usage … is actually consistent with the definitions of χαρά that Alexander [Aphrodisiensis, in Top. 112b21] (εὔλογος ἔπαρσις) and Hermias (ἡ τῆς ψυχῆς ἡδονή) attribute to Prodicus’. For surely the pleasure that the virtuous young feel when they are praised by their elders is not auditory delight, but joy in their soul. Thus, taking καλῶν in Hermias' testimonium as the participle from the verb καλῶ not only comports with normal Greek usage but also allows that testimonium to remain consistent with the evidence of Xenophon.

We turn now to the article by Louis-André Dorion, which takes a position diametrically opposed ([n. 2 (2008)], 88, 106) to my own. According to Dorion, so far from representing the language, or even the thought, of Prodicus, the Choice of Heracles is in every respect the work of Xenophon alone, and represents ‘in fact a sort of epitome of the major ethical positions upheld by Socrates and by Xenophon’.Footnote ¹⁴ Of course, if Z. Philip Ambrose is right to see the influence of Prodicus' Choice of Heracles on Aristophanes' Clouds and Cratinus' Pytine (423 b.c.), there can be no question of Xenophon's having created the piece in the fourth century.Footnote ¹⁵ Be that as it may, Dorion refers to Gautier and Gray (p. 91) and repeats their assertions that it is characteristic of Xenophon to employ unusual items of vocabulary and that the language and style of the Choice of Heracles are indistinguishable from those of the rest of Xenophon's writing. The main concern of Dorion's article (pp. 94–5), however, is to document exhaustively and for the first time the numerous ethical correspondences between the Choice of Heracles and the rest of Xenophon, in contrast to me, who never (Dorion's emphasis) concern myself with the moral aspects of the piece.Footnote ¹⁶ According to Dorion ([n. 2], 96), the main topic of the piece – which is never mentioned in my article – is enkrateia, the importance of which for Xenophon's Socrates, in contrast to the Socrates of Plato, is well known. One would not guess from Dorion's discussion that enkrateia and words derived from it never appear in the Choice of Heracles, either. To be sure, the account of the Choice of Heracles is introduced by Socrates as part of his effort to urge Aristippus to practise enkrateia (last mentioned in 2.1.7). But given that Xenophon's Socrates regards enkrateia as the very ‘foundation of virtue’ (τὴν ἐγκράτειαν ἀρετῆς εἶναι κρηπῖδα, Mem. 1.5.4), the absence of enkrateia and related words from a piece in which Virtue herself plays a starring role is curious. Dorion had shown that he is sensitive to matters of vocabulary, by devoting an article to explaining why the word does not occur in the early dialogues of Plato and only sporadically in his later works.Footnote ¹⁷ If the absence of a word – which is frequent in the Memorabilia – from the early dialogues of Plato needs to be explained, surely some explanation of the absence of that same word from the Choice of Heracles is called for. The most reasonable explanation would seem to be that the Choice of Heracles is not by Xenophon but by Prodicus.

There are two additional serious problems with the argument of Dorion, according to whom the ideas reflected in the piece are not those of Prodicus. In the first place, demonstrating that the ideas in the piece are consistent with those of Xenophon cannot prove that they are not (also) those of Prodicus, whose influence on Xenophon may have been as profound as it was, according to Xenophon himself, on Socrates. In the second place, why would Xenophon jeopardize his own credibility as a transmitter of the teachings of Socrates by attributing ideas to Prodicus that some of his readers would know were not those of the sophist? The Choice of Heracles is explicitly said, not only by Xenophon but by others as well, to exist in written form.Footnote ¹⁸ Anyone who had access to the written text could easily convict Xenophon of gross misrepresentation. At the time the Memorabilia was published, Isocrates, who is sometimes said to have been a pupil of Prodicus (Dion. Hal. Isoc. 1; Plut. Mor. 836f), was still alive. It is likely that Plato, too, was still alive at that time, and perhaps also Aristippus, the man to whom, according to Xenophon (Mem. 2.1.18, 34), Socrates recounted Prodicus' story. In a work in which Xenophon seeks to set the record straight regarding the doctrines of his beloved teacher, who was unjustly condemned to death, why would he falsify the record by foisting those doctrines onto someone else, particularly in a case where the falsification could be so readily exposed?

Dorion makes an attempt ([n. 2], 107) to address this issue, but not only is his answer unsatisfactory, it also underlines the seriousness of the first problem, raised at the start of the previous paragraph. According to Dorion, Prodicus is the only sophist who is described as ‘wise’ in the corpus of Xenophon's writings.Footnote ¹⁹ He continues:

Dorion goes on to give other examples of Xenophon's concern to ‘show that Socratic discourse is in agreement with the values cherished by the polis’. But if the values expressed in the Choice of Heracles are in reality those cherished by the polis, and if Xenophon's purpose is to show that the ideas of men like Prodicus ‘correspond closely to those of Socrates’, there are no grounds for denying that they were held and upheld also by Prodicus. And if that is the case, there is no reason not to assign the Choice of Heracles to Prodicus, as is in fact done by Xenophon. By acknowledging the conventional character of the ideas expressed in the Choice of Heracles, Dorion is fully in agreement with my own conclusion (pp. 140–1), that the piece was designed to avoid controversy and was intended to appeal ‘to the conventional and old-fashioned (and wealthy) fathers of the boys likely to be Prodicus’ customers'.

Of course, showing that objections to the case presented in my 2004 article are variously invalid does not constitute proof that my argument has merit. But the foregoing has been designed to show that critics will need to muster more substantial arguments if they seek to reverse Mayhew's assessment ([n. 2], 204), that ‘Sansone has provided reasons to think that it [i.e. Mem. 2.1.21–34] is a fairly close approximation, which often employs the language of the original’.