² In Archilochus' case, his brother-in-law; cf. the testimonium of Tzetzes, who cites the fragment, in West, loc. cit. This makes improbable O. Immisch's suggestion (‘Zu griechischen Dichtern’, Philologus 49 (1890), 193–8)Google Scholar of combining frg. 215 with frg. 196W: ⋯λλ⋯ μ' ⋯ λυσιμελ⋯ς ὦταῖρε δ⋯μναται π⋯θος, as the model for Epod. 11. 1–2.

³ So the verses have most recently been interpreted by Luck, G., ‘An Interpretation of Horace's Eleventh Epode’, ICS 1 (1976), 123Google Scholar, Degani, E.–Burzacchini, G., Lirici Greci (Florence, 1977), p. 5Google Scholar, and Fedeli, P., ‘II V Epodo e i Giambi d'Orazio come Espressione d'Arte Alessandrina’, MPhL 3 (1978), 118Google Scholar; cf. also Schmidt, E. A., ‘Amica vis pastoribus: der Jambiker Horaz in seinem Epodenbuch’, Gymnasium 84 (1977), 413 fGoogle Scholar.

⁴ This has often been pointed out, e.g. Kiessling–Heinze ad loc., Grassmann, V., Die erotischen Epoden des Horaz (Munich, 1966), p. 9Google Scholar, and West, M. L., Studies in Greek Elegy and Iambus (Berlin/N.Y., 1974), pp. 135–6CrossRefGoogle Scholar.

⁵ The fact that ‘me…expetit’ makes sense on its own and ‘urere’ comes much later tells against Axelson's suggestion that ‘me…expetit…urere’ = ‘cupit urere me’ (Ut Pictura Poesis: Studia…P. J. Enk…oblata (Leiden, 1955), p. 48Google Scholar). The sense is ‘Love attacks me above all (for the idea, cf. AP 5. 215. 3–4, ibid. 198 and 98, Prop. 2. 12. 13–20) to burn me’: thus Büchner, K., ‘Dichtung und Grammatik’, Mnemosyne 4. 10 (1957), 30 fGoogle Scholar., and Grassmann ad loc. Also possible, if less vigorous, is Dillenburger's explanation ‘me prae omnibus elegit quem urat’.

⁶ For this aspect of the poem see Leo, F., ‘De Horatio et Archilocho’ in Ausgewählte Kleine Schriften ii (Rome, 1960), 146 ffGoogle Scholar., supplemented by Grassmann, pp. 34 ff. and Luck, , op. cit. pp. 122 ffGoogle Scholar.: cf. also Kirn, B., ‘Zur literarischen Stellung von Horazens Jambenbuch’ (Diss. Tübingen, 1935), pp. 55 ffGoogle Scholar.

⁷ A similar interpretation was proposed by Büchner, K., ‘Die Epoden des Horaz’ in Werkanalysen (Wiesbaden, 1970), p. 60²Google Scholar(‘wenn ich, wie z. B. jetzt, der Liebe verfallen bin’). But Büchner does not develop his argument in any detail, or relate verses 1–2 to the poem as a whole.

⁸ Plüss, Th., ‘Das Jambenbuch des Horaz (Leipzig, 1904), p. 76²Google Scholar was the only person to see that ‘sicut antea’ was in any way significant, although his interpretation of it differs from mine. ‘Übrigens meine ich, wenn sicut hier seine Kraft und Bedeutung wahren solle und wenn V. 11 ff. und 16 ff. so zu verstehen seien, dass dem Sprecher schon vor zwei Jahren seine Verse nichts halfen und ihn verdrossen, dann dürfe man nicht mit Kiessling u. a. verstehen “jetzt anders oder weniger als früher” [sicut antea sc. iuvavit], sondern “jetzt wieder ganz ebensowenig wie früher” [sicut antea sc. nihil iuvavit]’.

⁹ See the discussion of these below.

¹⁰ This ‘remedium’ is hardened into a precept by Ovid, Rem. Am. 462 ffGoogle Scholar.

¹¹ Indeed, on this line of interpretation, 1–2 are little more than an exotic way of saying that Horace is in love.

¹² Quinn, K., ‘Two Crises in Horace's Poetical Career’ AUMLA 5 (1956), 37Google Scholar classifies Horace's supposed claim that love takes away his taste for writing with the familiar idea that love's onset interferes with all other activities; for illustration of the topic Guillemin, (REL 17 (1939), 282 ff.)Google Scholar is referred to. Neither writer points to an instance where falling in love is said to terminate poetic activity. Virgil, Ecl. 10. 62–3Google Scholar, appealed to by Luck, , op. cit. p. 123Google Scholar, is equally unhelpful. Kiessling–Heinze note that Prop. 2. 16. 33–4, which are sometimes compared with 1–2, are in no sense parallel to Horace's verses (as conventionally understood).

¹³ When speaking, as many have done, of elegiac elements in Epode 11, it must however be kept in mind that, of the Roman elegists, only Cornelius Gallus wrote before the composition of the Epode (cf. Grassmann, , p. 36)Google Scholar.

¹⁴ cf. Burck, E., ‘Römische Wesenszüge der Augusteischen Liebeselegie’, Hermes 80 (1952), 182 ffGoogle Scholar.

¹⁵ pp. 37 ff., 90, followed by Fedeli, , op. cit. p. 118Google Scholar.

¹⁶ Appeal to such sources on Epode 11 forms part of his attempt to show that Leo overstressed its background in Elegy and Comedy at the expense of Hellenistic Epigram (pp. 34 ff.).

¹⁷ The first of these represents the poet as proof against the onslaught of love, not vulnerable to it (see Gow, –Page HE ii. 3074 ff.)Google Scholar. The remaining two examples do oppose poetry and love – but in a specialized way. Love is seen overmastering the intellectual, bookish, Callimachean poet – in essence, emotion conquering intellect. The closing line of AP 12. 100, τ⋯ν σοφ⋯ν ⋯ν Μο⋯σαις Κ⋯πρις ἔτρωσε μ⋯νη, illustrates the point. Although the primary meaning of σοφ⋯ς in the context is the technical one of ‘poet’, the epigrammatist is more interested in playing with the original sense of ‘wise’, in order to set up a contrast with its inverse, the power of emotion (represented sense by love). Lines 3–4 of the well-known succeeding epigram (AP 12. 101, Meleager exploit the same notion. Similarly AP 12. 117 (Meleager).

¹⁸ Cf. Ritter (‘versiculos nunc dicit versus epodorum’), Giarratano (‘allude ai giambi’), Villeneuve, and Kiessling–Heinze, ad loc.

¹⁹ Wili, W., Horaz (Basel, 1948), p. 61Google Scholar. The situation in ll, as Wili sees it, is precisely equivaalent to Epode 14, where Horace states that love makes him unable to get on with writing, specifically, his iambi.

²⁰ op. cit. pp. 413 ff.

²¹ Why should Horace say ‘versiculos’ if he meant ‘iambos’? (Contrast, Epod. 14. 6–8Google Scholar ‘deus (sc. Amor) nam me vetat | inceptos, olim promissum carmen, iambos | ad umbilicum adducere’.)

²² Schmidt, sees Epode 11. 1–2Google Scholar as ‘der Anfang…<eines> Neueinsatzes’. For a similar view of the poem, see Quinn, op. cit., and especially Olivier, F., Les Épodes d'Horace in his Essais dans le domaine du monde gréco-romain antique (Geneva, 1963), pp. 111 ffGoogle Scholar.

²³ see Skutsch, O., ‘The Structure of the Propertian Monobiblos’, CPh 58 (1963), 239Google Scholar.

²⁴ For suggested links between Epode 11 and the new Archilochus, see Degani, –Burzacchini, , op. cit. pp. 4 ffGoogle Scholar.

²⁵ Meleager, , AP 12. 49Google Scholar urges such a course on the unhappy lover: Zωροπ⋯τει, δ⋯σερως, κα⋯ σο⋯ φλ⋯γα τ⋯ν φιλ⋯παιδα | κοιμ⋯σει λ⋯θας δωροδ⋯τας Bρ⋯μιος | ζωροπ⋯τει, κα⋯ πλ⋯ρες ⋯φυσσ⋯μενος σκ⋯φος οἴνας, | ἔκκρουσον στυγερ⋯ν ⋯κ κραδ⋯ας ⋯δ⋯ναν; cf. also Asclepiades ibid. 50, Prop. 3. 17. 1 ff.; Tib. 1.2. 1 ff.; [idem] 3. 6. 1 ff.; , Ov.Rem. Am. 805 ffGoogle Scholar.

²⁶ cf. Alciphron, , Epp. 1. 35. 2Google Scholar: κα⋯τοι γε ᾤμην τ⋯ν ἄκρατον ἔσεσθα⋯ μοι παρηγ⋯ρημα ὃν παρ' Εὐφρον⋯ῳ τρ⋯την ⋯σπ⋯ραν πολ⋯ν τινα ⋯νεφορησ⋯μην, ὡς δ⋯ τ⋯ς παρ⋯ τ⋯ν ν⋯κτα φροντ⋯δας διωσ⋯μενος τ⋯ δ⋯ ἄρα ⋯ναντ⋯ως ἔσχεν ⋯νερρ⋯πισε γ⋯ρ μου τ⋯ν ⋯πιθυμ⋯α7nu; ὥστε κλα⋯οντ⋯ με κα⋯ βρυχώμενον ⋯λεεῖσθαι μ⋯ν παρ⋯ τοῖς ⋯πιεικεστ⋯ροιγ⋯λωτα δ⋯ τοῖς ἄλλοις παρ⋯χειν.

²⁷ Past tense. The verb shows that Horace's secret was revealed against his will: cf. ⋯ρ⋯ν ⋯ρνε⋯μενον ⋯μῖν AP 12. 135 quoted n. 30 below.

²⁸ op. cit. p. 147.

²⁹ cf. esp. 1–3: Ἕλκος ἔχων ⋯ ξεῖνος ⋯λ⋯νθανεν ὡς ⋯νιηρ⋯ν | πνε⋯μα δι⋯ στηθ⋯ων, εἶδες, ⋯νηγ⋯γετο, | τ⋯ τρ⋯τον ⋯ν⋯κ' ἔπινε.

³⁰ Οἶνος ἔρωτος ἔλεγχος ⋯ρ⋯ν ⋯ρνε⋯μενον ⋯μῖν | ἤτασαν αἱ πολλα⋯ Νικαγ⋯ρην προπ⋯σεις | κα⋯ γ⋯ρ ⋯δ⋯κρυσεν κα⋯ ⋯ν⋯στασε, κα⋯ τι κατηφ⋯ς | ἔβλεπε, χὠ σφιγχθε⋯ς οὐκ ἔμενε στ⋯φανος.

³¹ For more detailed discussion of these two pieces see Giangrande, G. in L'Épigramme grecque, Fondation Hardt, Entretiens 14 (Geneva, 1968), pp. 120–2Google Scholar.

³² The banquet was traditionally the place where revelations took place about one's love-life; cf. Cairns, F., Hermes 98 (1970), 40 fGoogle Scholar.

³³ e.g. Eratosthenes frg. 36 Powell: Οἶνος…τ⋯ δ⋯ κα⋯ κεκρυμμ⋯να φα⋯νει | βυσσ⋯θεν, Ath. 37e; cf. Theogn. 479 ff.; Ath. 38b.

³⁴ The comparison with Od. 1. 27. 3 ‘verecundumque Bacchum’ is misleading; ‘verecundia’ there relates to restrained behaviour, not to holding one's tongue. Grassmann sees in ‘inverecundus deus…arcana promorat loco’ a sophisticated jest. Bacchus keeps his own ‘arcana’ (viz. ‘orgia’) secret, but is less scrupulous when it comes to others’. He compares Meleager, AP 12. 119. 5 ffGoogle Scholar.: ἦ προδ⋯τας (sc. Bacchus) κἄπιστος ἔφυς τε⋯ δ' ⋯ργια κρ⋯πτειν | αὐδ⋯ν, ⋯κφα⋯νειν τ⋯μ⋯ σὺ ν⋯ν ⋯θ⋯λεις.

³⁵ Seneca makes a significant distinction between ‘lacrimare’, decorous shedding of tears, and ‘plorare’, abandoning oneself without restraint to distress, Epp. 63. 1 ‘Nec sicci sint oculi amisso amico nee Quant; laenmandum est, non plorandum’.

³⁶ cf. Meleager, AP 5. 190. 3Google Scholar and 12. 85. 6: αὐτομ⋯τοις δ' ἄκων ποσσ⋯ ταχὺς φ⋯ρομαι.

³⁷ An intensification of ‘libera consilia’. When advice has no effect, the friends adopt a more censorious approach. For ‘contumeliae’ addressed to a lover, cf. Plaut, . Trin. 641 ffGoogle Scholar.; for the unsuccessful attempts of friends to prise the lover from his affair, cf. Prop. 1. 1. 25–6; 3. 24. 9.

³⁸ Bo, D., Lexicon Horatianum (Hildesheim, 1965–1966)Google Scholar, s.v. ‘ingenium’, is misleading.

³⁹ For the traditional poverty of poets, and the prominence of the theme in Hellenistic and Roman verse, seeCairns, F., Tibullus (Cambridge, 1979), pp. 20–1Google Scholar, with the bibliography there given.

⁴⁰ This is by far the most common meaning of ‘ingenium’ in Horace, : Sat. 1. 4. 43Google Scholar; 1. 10. 63; 2. 1. 67; 2. 1. 75; 2. 6. 15; Epp. 1. 3. 22; 2. 1. 88; 2. 2. 81; Ars Poet. 295; 323; 410; Od. 1. 6. 12; 2. 18. 9. For this interpretation of ‘ingenium’, see Ps.-Acro. ad loc. and Büchner, Epoden, loc. cit.; cf. Leo, , op. cit. p. 149Google Scholar.

⁴¹ But for the general rapaciousness of elegiac puellae, see Lilja, S., The Roman Elegists' Attitude to Women (Helsinki, 1965), pp. 143, 145 f.Google Scholar, and 149 f.

⁴² Seen. 6 supra.

⁴³ See Dawson, C. M., ‘An Alexandrian Prototype of Marathus?’, AJP 67 (1946), 1 ffGoogle Scholar.

⁴⁴ Grassmann, p. 40, rejects any such view; for him, lines 11–12 are an instance of ‘das Thema Geld-Liebe’; he cites as parallels numerous Hellenistic epigrams which handle the complaint of the lover against the beloved's venality.

⁴⁵ See further Petron. 129. 11, Val. Max. 8. 15 praef., Sen. Epp. 7. 7, Mart. 4. 86. 3–5 ‘ut docto placeas Apollinari. | nil…est…candidius benigniusque’, and, for the adjective in both senses, TLL 3. 244. 44 ff.

⁴⁶ Prop. 3. 24. 3–4 ‘noster amor talis tribuit tibi, Cynthia, laudes: | versibus insignem te pudet esse meis’, Tib. 1. 9. 47–8. The elegiac poets often allude to their power to make famous (Prop. 2. 5. 5–6; 2. 25. 3–4; 3. 2. 17–18; 3. 24. 3–4; , Ov.Am. 1. 10. 59–60Google Scholar; 3. 12. 9–10) or celebrate (e.g. Prop. 2. 1. 7–8; 2. 34. 93–4; 3. 24. 3 ff.) the beloved.

⁴⁷ It is possible that vv. 7–8 ‘heu me, per urbem – nam pudet tanti mali – fabula quanta fui!’ demonstrate another respect in which writing poetry proved unsatisfactory for Inachia's lover. Horace's language is remarkably similar to , Ov.Am. 3. 1. 21–2Google Scholar, ‘fabula, nec sentis, tota iactaris in urbe, | dum tua praeterito facta pudore refers’ and Prop. 2. 24. 1–8, ‘“Tu loqueris, cum sis iam noto fabula libro | et tua sit toto Cynthia lecta foro?” | cui non his verbis aspergat tempora sudor? | aut pudor ingenuis aut retinendus amor. | quod si iam facilis spirares, Cynthia, nobis, | non ego nequitiae dicerer esse caput, | nec sic per totam infamis traducerer urbem, | urerer et quamvis non bene, verba darem’ (Text of Bailey, D. R. Shackleton, Propertiana (Cambridge, 1956) pp. 110 ffGoogle Scholar., whom see on the crucial v. 4). Both elegists, like Horace, exploit the motif of pudor in connection with the notoriety of their amor. Both, apropos of the topic of publicizing their unhappy love in verse, offer close parallels to his ‘per urbem…fabula’. Although Horace does not disclose how he became a ‘fabula’, it may be that he is implying the same process of poetic self-betrayal. In that case he offers an additional reason for his initial decision not to write poetry while in love: it makes you notorious.

⁴⁸ For outspoken anger betokening a discidium, cf. Tib. 1.5. 1 ff.

⁴⁹ The slave Olympio is speaking to his master Lysidamus (‘Iuppiter’). The ‘di minores’ are his family, of whom Olympio has just said ‘they all hate me’. Therefore, if you die, says O., once the household passes into the control of your family, who will protect me from the blows I can expect from them? The idea contained in the apodosis (that O. will be exposed to blows) depends for its sense more, if anything, on the ‘quom’-clause than on the protasis.

⁵⁰ ‘Pudor’ is usually explained as ‘my feeling of shame lest I be defeated by an unworthy rival (which keeps me therefore from bowing out of the contest for I.'s affections)’. This cannot be right. What keeps Horace embroiled with Inachia is not his amour propre in face of a rival, but his helpless thraldom to her: witness his involuntary backsliding at 19 ff., and the unspoken implication, in his wish ‘quodsi…inaestuet… libera bilis’ (15–16), that Horace is in a condition of servitium. For other interpretations of this much-discussed phrase see Wickham, Schütz, and Ritter (‘modesty’), Desprez, and Kiessling-Heinze, and further n. 52 below.

⁵¹ For ‘pudor’ = ‘cause of shame’, see OLD 4a s.v. For the noun in this sense applied to behaviour, , Ov.Her. 19. 202Google Scholar ‘nil tua, Cydippe, facta pudoris habent’, Luc. 3. 148–9 ‘Venia est haec sola pudoris degenerisque metus’.

⁵² In view of Epp. 1. 9. 12 ‘depositum…pudorem’, there is something to be said for Lambinus' interpretation ‘my pushing aside of self-respect’ (which I should refer to the disreputable behaviour of the disappointed lover as seen in vv. 7–14, though Lambinus himself did not). However, coming after ‘ventis dividat’ and ‘desinet’, ‘summotus’ surely ought to refer, like them, to a departure from Horace's previous behaviour: in which case ‘summotus’ = ‘submovebitur et’, and so the commentators (including Lambinus in an alternative explanation) have generally construed it.

⁵³ This interpretation founders on ‘impartibus certare’. A lover may ‘certare’ with a rival: it is difficult to see in what sense he could ‘certare’ with a girl who does not spend time with him.

⁵⁴ cf. Pichon, R., Index Verborum Amatoriorum (Paris, 1902)Google Scholar, s.v. ‘par’.

⁵⁵ Orelli rejects this and the preceding suggestion in his useful note ad loc.

⁵⁶ Thus Porphyrio, Lambinus, Desprez, Mitscherlich, Doering, and Tescari.

⁵⁷ e.g. L. Müller (‘me indignis’), Kiessling-Heinze, Giarratano, Turolla.

⁵⁸ Cf. Dillenburger, Schütz, Wickham, Orelli, L. Müller, Giarratano.

⁵⁹ For ‘ingratus’ in this sense he compares Plaut, . Asin. 136Google Scholar ‘ingrata atque irrita esse omnia intellego’ (cf. TLL 7. 1. 1560. 19 ff.). Alternatively, ‘ingrata’ means simply ‘displeasing’. Unhappy love calls for such remedies as wine-drinking, but Horace dislikes the results.

⁶⁰ Epoden, loc. cit. The remarks of Plüss quoted above in n. 8 show that he also understood ‘fomenta’ thus.

⁶¹ A. W. Mair, in the Loebed. ad loc., translates ‘reduce the swollen wound of love’. Gow, –Page, HE 2Google Scholar. 157 and Fraser, P. M., Ptolemaic Alexandria (Oxford, 1972), 2. 545 ffGoogle Scholar. also give ‘reduce’ (in the medical sense).

That we are not dealing here with an attenuated metaphor (LSJ s.v. note κατισχνα⋯νειν…⋯σμ⋯ν Theophr, . de odor. 47)Google Scholar is strongly suggested by the high incidence of medical language in the poem (see Gow-Page and Fraser locc. cit.), and the possibility that Callimachus' addressee Philippus is, like Theocritus' Nicias, a doctor (Edgar, C. C., Zenon Papyri in the University of Michigan Collection (Ann Arbor, 1931), p. 126)Google Scholar.

⁶² Just as, in v. 6, ⋯κκ⋯πτει τ⋯ν φιλ⋯παιδα ν⋯σον, we have a surgical metaphor: ⋯κκ⋯πτω, ‘cut out’, is quite frequent in the Corpus Medicorum Graecorum, e.g. Galen, de atra bile 5. 11Google Scholar: ⋯ταν ⋯κκ⋯ψῃ τ⋯ πεπονθ⋯ς μ⋯ριον ⋯λον ⋯ν κ⋯κλῳ περιτεμὼν χρι τ⋯ν ⋯παθ⋯ν, Soran, . Gynaec. 4. 7. 7Google Scholar: δι⋯ χειρουργ⋯ας ⋯κκ⋯πτειν.

⁶³ κατισχνα⋯νω. Hippocr, . Progn. 23Google Scholar. ἰσχνα⋯νω. Liqu. 3: sea-water ⋯γαθ⋯ν δ⋯ κα⋯ ἰσχνα⋯νειν εὖ [sc.ἔλκεα); Aph. 5. 25: τ⋯ ⋯ν ἄρθροισιν οἰδ⋯ματα… ψυχρ⋯ν πολλ⋯ν καταχε⋯μενον τε κα⋯ ἰσχνα⋯νει de fist. 10: το⋯των τ⋯ν καταπλασμ⋯των…τ⋯…⋯ς ⋯ωυτ⋯ ἔλκοντα ξηρα⋯νει κα⋯ ἰσχνα⋯νει (sc. abscesses); de loc. hom. 38: τ⋯ν φαρμ⋯κων…τ⋯…ἰσχνα⋯νοντα.

⁶⁴ Contrast e.g. Verg, . Aen. 4. 1–2Google Scholar ‘At regina gravi iamdudum saucia cura | vulnus alit venis et caeco carpitur igni’, or the remarkable picture at Ap. Rhod. 3. 761 ff.

⁶⁵ For this meaning, cf. McGann, M. J., Studies in Horace's First Book of Epistles (Brussels, 1969), p. 41²Google Scholar.

⁶⁶ Note the subject of the poem in general, the image of the bee 21 (a symbol of poetic I inspiration, cf. Od. 4. 2. 27 ff.), mention of the ivy crown 25, and the significant positioning of ‘seu condis amabile carmen’ last in the list of Florus' activities.

For ‘fomenta’ referring to poetry, the comment of Porphyrio, on Epp. 1Google Scholar. 1. 10 ‘nunc itaque et versus et cetera ludicra pono’ is also of interest: ‘depositis levibus fomentis animi id est iocis et versibus’.

⁶⁷ For the thought cf. Erbse, H., ‘Dichtkunst und Medizin in Theokrits 11. Idyll’, MH 22 (1965), 232 ffGoogle Scholar.

⁶⁸ For a later inversion of Theocritus' and Callimachus' prescription for curing love cf. Longus l 2. 7 fin.: Ἔρωτος γ⋯ρ οὐδ⋯ν φ⋯ρμακον, οὐ πιν⋯μενον, οὐκ ⋯σθι⋯μενον, οὐκ ⋯ν ᾠδαῖς λεγ⋯μενον, ⋯τι μ⋯ φ⋯λημα κα⋯ περιβολ⋯ κα⋯ συγκατακλιθ⋯ναι γυμνοῖς σώμασι.

⁶⁹ For the pudor-motif in connection with poetry misspent, cf. Tib. 1. 9. 47 ff. ‘quin etiam attonita laudes tibi mente canebam, | at me nunc nostri Pieridumque pudet. | illa velim rapida Vulcanus carmina flamma | torreat et liquida deleat amnis aqua’.

⁷⁰ I am grateful to the editors and an anonymous referee for CQ for suggesting various improvements to an earlier version of this article.