1. See the survey of scholarship by Setaioli, Aldo, ‘Gli “Epodi” di Orazio nella critica dal 1937 al 1972 (con un appendice fino al 1978),’ ANRW 2.31.3 (1981), 1674–1788Google Scholar. Most of the general studies he cites stress the varied nature of the Epodes, and many regard them as experimental, the usual fate of a poet’s early work, which is too often treated as a sign of things to come.

2. There is a convenient table of the dating of these poems by various scholars in Carrubba, Robert, The Epodes of Horace: A Study in Poetic Arrangement (Paris 1969), 16CrossRefGoogle Scholar.

3. The best treatment of the Archilochean aspect of the Epodes is that of Nisbet, R.G.M., ‘Horace’s Epodes and History,’ in Woodman, Tony and West, David, edd., Poetry and Politics in the Age of Augustus (Cambridge 1984), 1–18Google Scholar. Nisbet sees Horace’s adoption of the Archilochean persona as ‘a declaration of his own alienation’ after Philippi, and argues that after he was drawn into Maecenas’ circle the persona is softened and the engagement with politics is muted; Nisbet cites Epode 4 as an instance of the trivialization of the poet’s earlier treatment of war and politics (9). I will argue that a study of poems like Epode 4 in the context of the collection, rather than in terms of chronology, makes this kind of judgment inappropriate. For a table of the ways scholars have grouped the poems, see Carrubba (n. 2 above), 34f.

4. Setaioli (n.1 above), 1689, states that ‘i piu equilibrati interpreti’ of the Epodes agree that in ordering the poems Horace was guided by criteria of metre or his taste for variatio. His criticisms of Carrubba’s elaborate symmetries (1689f.) seem to me justified, but I would add that the assumption, which Carrubba shares with the more ‘balanced’ interpreters, that questions of organization are to do merely with the disposition, rather than the interaction, of poems does not do Horace justice.

5. Horace models the projected emigration from Rome on that of the Phocaeans, under pressure from the Persians, and there are similarities with the account of that event by Herodotus (Hdt. 1.165; see Setaioli, n. 1 above, 1749f.).

6. Fraenkel, Eduard, Horace (Oxford 1957), 43–6Google Scholar.

7. As I will illustrate, Horace repeatedly puns on his name, a fact that has been noted by several commentators. The sexual pun is here continued by constantia, cf. constantior… nervus in Epode 12.19. The precise meaning of lines 15f. is disputed, and some have read offensi instead of offensae. My observations, however, do not depend on anything controversial in the interpretation of these lines. For a discussion of these lines, see Grassmann, Victor, Die Erotischen Epoden des Horaz (Munich 1966), 155–57Google Scholar.

8. Mollis (37) connotes effeminacy, and is used of the pathic (OLD, 15). Perprimal (38), especially in connection with cubilia, may have sexual connotations. For the sexual meaning of premo and its compounds, see Adams, J. N., The Latin Sexual Vocabulary (London 1982), 182fGoogle Scholar. Perprimo itself is used in a sexual sense by Ovid, Ars Am. 1.394.

9. Nisbet (n.3 above), 3, sees a pun on the Greek transliteration of Roma (Rhō mē, ‘strength’) in the juxtaposition of Roma and viribus.

10. For another revision of Vergil, see Nisbet’s remarks (n.3 above), 2f.; on the relation between Epode 16 and Eclogue 4, the relative dating of which has been much discussed, see Setaioli(n.1 above), 1753–61.

11. Comparisons of young girls to skittish fillies that have to be handled with care are common in the ancient world; Anacreon’s is one of the first (335 Page), and Horace himself has one that is close to this at C.3.11.9f., where the recalcitrant Lyde is described as a young filly who metuit tangi ‘fears to be touched’). For palpo in a sexual sense, see Juvenal 10.206. In view of these parallels, I do not think it fanciful to see sexual connotations in the statement that the words of Flaccus will not go through the ear of Octavian except at the right moment.

12. This is the argument of Babcock, C.L. in ‘Omne militabitur bellum. The Language of Commitment in Epode 1,’ CJ 70.1 (1974), 14–31Google Scholar.

13. OLD, intestinum, 1. Compare also Catullus 77, where Rufus, the friend who played Catullus false, is described as intestina perurens (77.3). The Catullan usage strengthens my suspicion that in this poem Horace is ambivalent about his relationship with Maecenas.

14. Horace seems to be parodying the Twelve Tables. Compare Si quis occentavisset sive carmen condidisset … as quoted by Cicero, de Rep., 4.12. The archaic subjunctive, edit also suggests parody. For a later version of the si quis construction specifying a crime, see Livy 29.21.5.

15. The sequel to Jason’s yoking of the oxen is the sowing of the dragon’s teeth, from which the armed men spring up and whom Jason sets fighting among themselves by throwing a stone in their midst; this battle is one of the mythical prototypes of civil war (Lucan, BC 10.549–69).

16. Nagy, Gregory, The Best of The Achaeans (Baltimore and London 1979), 242Google Scholar.

17. Setaioli (n. 1 above), 1703.

18. See Burnett, Anne Pippin, Three Archaic Poets: Archilochus, Alcaeus and Sappho (Cambridge, 1983), 60–5Google Scholar.

19. The ancient tradition of vetulaskoptik that lies behind Epodes 8 and 12 is dealt with in Grassmann (n. 7 above) and the Latin tradition before and after Horace by Richlin, Amy in The Gardens of Priapus: Sexuality and Aggression in Roman Humor (New Haven 1983), 105–116Google Scholar.

20. Another link between the vetula and Canidia is the phrase dente livido used of Canidia at 5.47.

21. The question of the-fictional (or, for some, real) time and place of this poem is one of the canonical ‘problems’ of the Epodes. For a summary of the relevant scholarship, see Setaioli (n.l above), 1716–30; who concludes that Horace speaks from a point when Actium has been won, and that it is to be thought of as having been composed at the site of the battle on board ship, Nisbet’s careful analysis of the poem (n.3 above, 11–17) bears this out, although he sees the poem as a running commentary on the battle, ending with the victory celebration in the evening. My only objection to Nisbet’s interpretation is that he takes the symposiastic motif of the final two lines as symbolic: the release from cares that wine provides is understood by him as a parallel to the release from cares (and foreign bondage) provided by Octavian’s victory. I would rather take the lines in a more conventional sense: Horace is still in a state of anxiety about Octavian and needs to drown it in wine; as Setaioli observes (n.l above, 1724), it was not immediately apparent that Actium was a decisive victory.

22. Maevius is the poet attacked by Vergil at Eclogue 3.90. This reference allows us to glimpse a continuing literary war that runs parallel to the civil war, and one about which Horace could feel more confident.

23. The storm that Horace calls down on Maevius’ ship will break mountain oaks (insurgat Aquilo, quanlus altis montibus/frangit trementis ilices, 10.7f.), and this is recalled in Epode 12, when the member of Amyntas, compared favorably to Horace’s, is described as constantior … quam nova collibus arbor (‘firmer … than a young tree in the hills’, 12.19f.).

24. Richlin (n. 19 above, 113) suggests that ‘invective against vetulae constitutes a sort of apotropaic satire that attempts to belittle and control the power of old women.’