¹ I am grateful to Hugh Lloyd-Jones and to Nicholas Richardson for commenting on a draft of this paper; the blame is, of course, still mine. The second part of the paper develops points made briefly in the first section of my forthcoming book, The Poetics of Greek Tragedy.

² This term is perhaps potentially misleading. M. Gagarin has pointed out that δίκη does not mean ‘justice’ in a broad sense in WD, but is restricted to ‘“law”, in the sense of a process for the peaceful settlement of disputes’ (CPh 68 [1973], 81Google Scholar). This is, I think, correct; but the conclusion which Gagarin draws (‘WD is not a treatise about morality or justice, but rather about prosperity and the necessity of an effective legal process to help achieve it’) is distorted, since Hesiod is clearly concerned with a much wider range of moral issues (fraternal loyalty, respect for parents, for ξένοι, etc: see 182–8, 327–35). I would prefer to say, therefore, that δίκη in its restricted sense is for Hesiod an exemplary case for right social behaviour in general. It is this general ethic, and not δίκη to which I here apply the term ‘justice’.

³ West, M. L., Hesiod, Works and Days (Oxford, 1978), p. 37Google Scholar. Is there actually a verbal connection? νείκεα…⋯φέλλοις would be an odd phrase to apply to mere observation; I do not think, therefore, that the words are meant to be, as West suggests, a transitional equivocation: rather, they convey obliquely a significant new piece of information (namely, that Perses is an active litigant as well as an observer).

⁴ West finds this ‘artificial’, and suggests that, had it been in his mind from the start, Hesiod would have ‘described the bad Eris in 14ff. more in terms of Perses’ way of life, in terms of νείκεα rather than πόλεμος' (pp. 36–7). But it is surely quite natural to introduce the bad Eris in general terms, indicating the full range of her activity before an application is made to the particular circumstances of the poem; and the common term (δ⋯ρις) secures a measure of continuity.

⁵ West finds a number of difficulties in the presentation of Perses. (i) Those found in lines 11–41 I have discussed in the text. (ii) There is no demonstrable inconsistency between 35ff. and 394ff. (West, pp. 35, 38); for example, Perses may have squandered his unjust gains, appealed to his brother for assistance, and threatened further litigation on being rebuffed. (iii) A number of West's remarks (pp. 36, 39–40) seem to presuppose that the protasis of a conditional must reflect the circumstances of that conditional's purported utterance; I find this very strange. The real difficulty, it seems to me, is precisely the opposite one to that which worries West: not in producing a coherent account of the circumstances consistent with all the data of the poem, but in selecting among the many mutually exclusive accounts which the data fail to exclude. That Hesiod gives the background in so cursory a manner might indicate that the situation is a real one, which Hesiod expected his audience to be familiar with; and/or that the question is not of great importance for understanding the poem (which few, I imagine, would deny).

⁶ West, pp. 42–6.

⁷ Lord quotes a Yugoslav bard who liked to think a new song over for a day before performing it, but implies that most would be able to sing it without such preparation (The Singer of Tales [Cambridge, Mass., 1960], pp. 26–7Google Scholar); but this is when they have heard the song from another singer, and so have assimilated prior to their own performance a clear conception of what has to be said and in what order. Lord does give an example of a song genuinely improvised in a stronger sense (pp. 286ff., n. 3); but it is a miserable specimen, and produced in very exceptional circumstances. West's account of Works and Days seems to assume that a poem might be produced in much the same way under more normal conditions, and indeed that the poet would then willingly reproduce it and preserve it in writing; that is quite a different matter. (On the scope for premeditation in ‘oral composition’, see West, 's careful statement in I Poemi Epici Rapsodici non-Omerici e la Tradizione Orale, ed. Brillante, C., Cantilena, M. and Pavese, C. O. [Padua, 1981], pp. 62–3Google Scholar.)

⁸ For the narrative technique by which such details as the existence of the πίθος are withheld until they become essential, see Fraenkel, E., Aeschylus, Agamemnon (Oxford, 1950), p. 805Google Scholar. The extreme compression of the narrative, which leaves the nature and provenance of the πίθος implicit, is presumably a generic feature of small-scale epos of this kind; the technique was admired and imitated by many Hellenistic poets.

⁹ E.g. West, p. 49.

¹⁰ It is impossible to know whether it was Hesiod or a predecessor who interpolated the heroes into the sequence of metallic γένη. If it was Hesiod, it is unlikely to have been (as some suppose) because he felt obliged to reconcile the myth with the epic tradition; the syncretising urge was surely not so powerful in a poet content to juxtapose these two incompatible aitia. The insertion makes artistic sense in WD: the justice of the heroes throws the injustice of the iron men into relief, as the ease of their final state does the iron men's adversities. So, rightly, Verdenius, in Hésiode et son Influence (Entretiens sur l'Antiquité Classique 7, Geneva, 1962), pp. 130–2Google Scholar.

¹¹ ‘The meaning is obvious: the weak are at the mercy of the strong. The common people already understand this, but Hesiod makes his fable simple and clear for the kings so that they too will understand’ (Gagarin, op. cit., p. 92 n. 58). I fail to see: (i) why the kings are supposed to be ignorant of this obvious truth; (ii) why, if they were, Hesiod should have thought it helpful to enlighten them; (iii) why the kings need telling ‘clearly and simply’ — are they so dim? Contrast WD 202.

¹² Od. 14.459–517 perhaps suggests the acuteness that could be required of the audience of an ainos; Odysseus does not even warn his hearers that a covert intention is involved.

¹³ Thus the interpreted ainos does not imply that the hawk was wrong: only that the kings would be wrong in acting like it, the principles of judgement applicable to hawks being inapplicable to men (or at any rate, to men who are not at war); this disarms the objections of Welles, C. B., GRBS 8 (1967), 17–19Google Scholar. I should add that I do not wish to read the ainos as an exact allegory of Hesiod's situation; it suffices that ⋯οιδός (208) hints clearly to the kings that there are points of comparability such that the story has important implications for their own dealings with Hesiod.

¹⁴ I take the progression ‘thirty thousand guardians — Dike — Zeus’ to be a designed escalation; contrast West: after 264 Hesiod ‘is unable to make a coherent continuation. There follows a mere dribble of additional thoughts…’ (p. 50).

¹⁵ The unusually elaborate address to Perses in 286 confirms that a major new subsection of the poem is opening.

¹⁶ Hesiod adds that it is easier when one has arrived (290–2); I take this to be a remark made in passing and designed to soften the deterrent impact of the difficulty of the right path. (It is not very convincing; Hesiod does not really envisage a point at which one could relax from the perpetual round of toil.)

¹⁷ West, p. 229 (on 287–92).

¹⁸ Hesiod adds that idleness makes one unpopular with gods and men, as a factor accentuating the tendency of the idle to impoverishment (303–10). That the gods' disapproval is damaging to prosperity is obvious; that human disapproval is damaging may be less so: but see 342–55.

¹⁹ Admittedly a less striking one. I take ὧδ' ἔρδειν in 382 retrospectively.

²⁰ West, p. 45 (although he is more grudging about it than these quotations suggest); West's discussion of [B] is as helpful as his comments on [A] were unhelpful.

²¹ That between 695 and 617 (see West, p. 326, on 695) scarcely less so: the ‘season’ for an annual operation like ploughing is not very like the ‘season’ for marriage.

²² To allay alarm, I had better say at once that I do not think it can. But some have entertained the possibility; e.g., Friedländer, (in Hesiod, ed. Heitsch, E. [Wege der Forschung 44, Darmstadt, 1966], p. 237)Google Scholar: ‘Der “Schifferkalender” ist das letzte was man mit Gewißheit dem Hesiod zuschreiben darf… Von dem, was nun noch folgt, wüßte ich nicht, wie man den Beweis des hesiodischen Ursprungs erbringen wollte.’

²³ See, most recently, Janko, R., Homer, Hesiod and the Hymns (Cambridge, 1982), pp. 99–100Google Scholar, with notes. On the ‘sphragis’ device in general, see Kranz, W., Studien zur antiken Literatur (Heidelberg, 1967), pp. 27–78Google Scholar; the reality of the device does not, of course, depend on the questionable derivation of the term from Theognis 19.

²⁴ See West, p. 347. The difficulties which he notes there and in the subsequent commentary are perhaps as striking as the positively Hesiodic turns of phrase that he detects, some of which are paralleled as closely outside as within Hesiod: the best examples that he gives are ψεύδεα θ' αἱμυλίους τε λόγους (789; cf. 78) and ⋯υτροχαλῷ ⋯ν ⋯λωῇ (806; cf. 599, and contrast the Homeric ⋯υκτιμένῃ); his reference to 817 neglects the anomalous addition of πολυκλήιδα. See further Solmsen, F., TAPA 94 (1963), 293–320Google Scholar.

²⁵ West, pp. 333–4.

²⁶ Hesiodos Erga (Berlin, 1928), p. 129Google Scholar.

²⁷ Most striking is the genitive form of the formula in 718 (cf. Th. 33); but note also ὡραῖος (695; cf. 32, 307, 617, 630, 685), ἔνδοθι οἴκου (733; cf. 523, 601).

²⁸ For example, I find unconvincing Solmsen's theory (HSCP 86 [1982], 30–1Google Scholar) that 618–45 and 646–94 are alternatives; as I pointed out above, the functions of the two autobiographical insets are quite different (West's suggestion of authorial interpolation would be a more defensible way of accounting for the difficulties here). But for a minor concession to this way of thinking, see n. 33. It might be worth adding that West has already argued that the poem's original ‘prospect’ did not extend beyond 381 (pp. 44–5): the speculation which I have aired here modifies this view by treating [B] not simply as an extension of [A], but as an authorial interpolation into (or possibly as a replacement for part of) an originally fuller version of [A]; unfortunately West's arguments for excluding [B] from the original prospect are not compelling.

²⁹ See Callimachus on Aratus (Ep. 27.1 Ἡσιόδου τό τ’ ἄεισμα) and Vergil on himself (G. 2.176 Ascraeumque cano …carmen). A theoretical niche for didactic poetry is found in Diomedes (482.30–483.3 Keil), which Pöhlmann, (ANRW 1.3 [Berlin-New York, 1973], pp. 825–35)Google Scholar antedates to the early Hellenistic period — plausibly, in view of the resurgence of didactic poetry at this time among self-conscious and articulate poets. I should stress, however, that articulateness is not necessary; genres need not be explicitly distinguished in a culture: implicit recognition of the distinction, manifesting itself in different dispositions to respond, suffices.

³⁰ He does so implicitly, Memmius being no more his sole audience of address than Perses in WD; but in Lucretius' case, a good deal of contextual information being available, we can be more confident in identifying the poem's implicit claims.

³¹ This subdivision of formal into final and purely formal didactic poetry is not meant to exclude more delicate discriminations. See, for example, Effe, 's useful study, Dichtung und Lehre: Untersuchungen zur Typologie des antiken Lehrgedichts (Zetemata 69, Munich, 1977)Google Scholar; my ‘final’ embraces his ‘sachbezogen’ and ‘transparent’ categories, my ‘purely formal’ his ‘formal’ and the ‘spielerisch-parodistisch’ class that he treats separately as a Sonderform. (The ‘transparent’ class is that in which the formal didactic programme — Stoff — does not coincide with the final-didactic intent — Thema: e.g., Vergil's Georgics, to the extent that that poem is final-didactic at all; Effe places Aratus in this category — mistakenly, in my judgement.)

³² West, p. 53.

³³ Probably less: 576–7 and 578–84 look like alternative versions.

³⁴ Cf. Rowe, C. J., JHS 103 (1983), 128–9Google Scholar.

³⁵ Stokes, M. C., Phronesis 7 (1962), 4CrossRefGoogle Scholar; cf. pp. 33–7 and West, , Hesiod, Theogony (Oxford, 1966), pp. 379–83Google Scholar. I do not wish to get drawn too far into the debate over the authenticity of this episode here, but I note three points. (i) The poet of WD evidently had no aversion to doublets (as his two aetiological logoi prove); so it would be no surprise if he took the opportunity which a reduplication of the theomachy-motif would provide for reworking some impressive material. (ii) In that case, we might expect a greater striving for effect in the second passage; whether we should say ‘strain’ rather than ‘strive’ is a matter of subjective judgement, and I do not think that we know enough about contemporary taste in such things to warrant athetesis on aesthetic grounds. (I do not, in any case, wholly share the widespread distaste for this episode. Typhoeus is a good monster: an anthropomorphic body sprouting a hundred serpentine heads is a formidable conception — far more so than the conventional representation of Typhoeus in visual art, for which see West's note on Th. 306; since the plethora of heads is his most remarkable feature, the anaphora of κεɸαλή in 824–30 is not without point.) (iii) Those who, like West (in his note ad loc.) or Solmsen, (Gnomon 40 [1968], 328)Google Scholar, are willing to explain the difficulties in the context of Th. 139–53 by assuming a later addition by Hesiod himself cannot infer the inauthenticity of the Typhoeus episode from its neglect in 881, since the same explanation would be possible there. (West's own explanation of that neglect, on p. 381, will not do if one stresses, as I have done, the climactic importance of the episode.)

³⁶ W. Stroh criticises more traditional readings of the passage and offers an interpretation similar to that outlined here (‘Hesiods lügende Musen’, Studien zum antiken Epos, ed. Görgemanns, H. and Schmidt, E. A. [Meisenheim am Glan, 1976], pp. 85–112Google Scholar); but I would not concur with all his arguments and conclusions. His paper is far more convincing, however, than the reply by Neitzel, H. (Hermes 108 [1980], 387–401)Google Scholar. Of Neitzel's argument against Stroh (p. 389) I observe: (i) that it seems to turn on taking Th. 27 as if spoken by Hesiod propria persona while attributing Th. 28 to the Muses; (ii) that one of its premises is false (if I said, for example, of this article that it contains falsehoods which resemble truth, I would not necessarily mean that I had wittingly included falsehoods; I might mean only that I was sure to have made some errors, and that those errors must resemble the truth at least sufficiently for me to have mistaken them for the truth); and (iii) that there are in any case many conceivable reasons why one might include even witting untruths in one's utterances, to which Neitzel's few dismissive words (‘ein rhetorisches Spiel’) do little justice.

³⁷ Mnemosyne 25 (1975), 235Google Scholar; he refers to Th. 229, WD 78, 789.

³⁸ I assume that the beauty of the song consists in the attractiveness of its content no less than in its style or form; it sometimes seems to be forgotten in these discussions that veracity is not the sole excellence of content (cf. nn. 36 and 43).

³⁹ This story is not told by Homer in his own persona, it is true; but it is attributed by him to a bard, so that its implications about what we can, in Homer's eyes, expect from poets must stand.

⁴⁰ As Pindar observes: N. 7.20–3 (cf. O. 1.27–32). He expresses disapproval here because he is committed in the context to magnifying Ajax's κλέος, and therefore to diminishing that of Odysseus, Ajax's chief rival; what he would say on the matter when not under such constraint is a matter for conjecture. (We should be wary of appeals to such passages as O. 1.35ff. in connection with Hesiod and Homer, since Pindar seems to have been influenced in them by post-Hesiodic and post-Homeric philosophical critiques, such as that found in Xenophanes frr. 10–12.)

⁴¹ See Walcot, P., Ancient Society 8 (1977), 1–19Google Scholar, on Odysseus and ‘the art of lying’.

⁴² Macleod, C., Collected Papers (Oxford, 1983), p. 4Google Scholar.

⁴³ Contrast (e.g.) Verdenius, W. J., in The Sophists and their Legacy, ed. Kerferd, G. B. (Hermes Einzelschriften 44, Wiesbaden, 1981), p. 122Google Scholar. I take it that μορɸ⋯ ⋯πέων covers attractiveness of content as well as of form: cf. Od. 8. 166–77 (170–3 recall Th. 80–93), where I suspect that the lack of μορɸή attributed to Euryalus' speech resides not so much in its untruth (although Odysseus counters the slur in the most effective way possible by showing it to be untrue) as in its personal offensiveness; true or false, that is not the kind of thing that one ought to say to guests.

⁴⁴ It may be relevant that the verb which Eumaeus uses (θέλγειν) so often connotes some kind of deception: e.g. Il. 14.214–17, 21.276, 604, Od. 14.384, 16.194–5. Songs are meant to be βροτ⋯ν θελκτήρια; but the lying tales of unscrupulous wanderers are another matter.

⁴⁵ The consequent question of the kinds of pleasure that an audience might expect to derive from Hesiodic poetry goes beyond the limited scope of this paper; I hope to raise it in a subsequent and more comprehensive study of Greek attitudes to poetic ⋯δονή.