André, J., ed., Ovide: Contre Ibis. Collection Budé (Paris, 1963).Google Scholar

Brink, C. O., ed., Horace on Poetry 2. The Ars Poetica (Cambridge, 1971).Google Scholar

Ellis, R., P. Ovidii Nasonis Ibis (Oxford, 1881).Google Scholar

Lloyd-Jones, H. and Parsons, P., edd., Supplementum Hellenisticum (Berlin/New York, 1983).CrossRefGoogle Scholar

Nisbet, R. G. M. and Hubbard, M., A commentary on Horace. Odes Book II (Oxford, 1978).Google Scholar

Watson, L., ARAE: The curse poetry of antiquity (Leeds, 1991).Google Scholar.

1. For the distinction see Cicero, Off. 1.104, where two categories of jesting are distinguished – illiberal jesting and its urbane counterpart; on the history of the distinction in antiquity see Bramble, J. C., Persius and the programmatic satire (Cambridge, 1974) 190ffCrossRefGoogle Scholar. The distinction is crucial to Horace's famous defence against the charge of malice in S. 1.4 (cf. ‘laedere gaudes’, inquit, ‘et hoc studio prauus facis’, 79–80), on which see Dickie, M., ‘The disavowal of invidia in Roman iamb and satire’, in Cairns, F., ed., Papers of the Liverpool Latin Seminar 3 (1981) 185–93Google Scholar with Hunter, R. L., ‘Horace on friendship and free speech’, Hermes 113 (1985) 486–90Google Scholar. Cf. Bramble, op. cit., 200 on Tr. 2.563 and 565: ‘more statements than disclaimers – if anything, iambic is envisaged rather than satire – these passages nonetheless derive from the sanctioned contrast between venomous and humane wit’.

2. On various kinds of innuendo in Tristia 2 see Wiedemann, T., ‘The political background to Ovid's Tristia 2’, CQ n.s. 25 (1975) 264–71CrossRefGoogle Scholar, Scott, K., ‘Another of Ovid's errors?’, CJ 26 (1931) 293–6Google Scholar, Claassen, J.-M., Poeta, exsul, votes: a stylistic and literary analysis of Ovid's Tristia and Epistolae ex Ponto (diss. Stellenbosch, 1986) 307–14Google Scholar, Focardi, G., ‘Difesa, preghiera, ironia nel II libro dei Tristia di Ovidio’, SIFC 47 (1975) 86–129Google Scholar, Vulikh, N. V., ‘Ovid and Augustus’, VDI 103 (1968) 151–60Google Scholar and ‘La révolte d'Ovide contre Auguste’, LEC 36 (1968) 370–82Google Scholar.

3. Cf. Hinds, S. E., ‘Generalising about Ovid’, Boyle, A. J., ed., Ramus: Critical Essays in Greek and Roman Literature 16 (1987) 25–6Google Scholar on Ovid's use of the ‘hermeneutic alibi’ of verbal ambivalence to evade the charge of deliberate flippancy in his treatment of Augustus in Metamorphoses 15.

4. Who was Ibis? For various conjectures see Elis xix–xxvii, La Penna xvi–xix and André xxiv–xxvi. But cf. Housman, A. E., ‘The Ibis of Ovid’, JPh 35 (1920) 316Google Scholar = Diggle, J. and Goodyear, F. R. D., edd., The classical papers of A. E. Housman (Cambridge, 1972) 1040Google Scholar: Ibis ‘is much too good to be true’, born all too conveniently on the dies Alliensis (219–20) and in the African wilderness (221–2); ‘nor does a man assail a real enemy, the object of his sincere and lively hatred, with an interminable and inconsistent series of execrations which can neither be read nor written seriously’. Like Wilkinson, L. P., Ovid recalled (Cambridge, 1955) 355Google Scholar, I incline to Housman's view that Ibis is a fiction; but for an impartial summary of opinion on the whole controversy see Watson, 130–1.

5. So Housman (1920) 317–18 = (1972) 1041, followed by Wilkinson (n. 4) 356–7 and Kenney, E. J., The Cambridge history of classical literature II: Latin literature (Cambridge, 1982) 454CrossRefGoogle Scholar.

6. On the militia amoris motif see McKeown, J. C., Ovid: Amores. A commentary on Book I (Liverpool, 1989) 258–9Google Scholar with further bibliography.

7. See Zipfel, K., Quatenus Ovidius in Ibide Callimachum aliosque fontes inprimis defixiones secutus sit (diss. Leipzig, 1910) 9Google Scholar, Rostagni, A., Ibis. Storia di un poemetto greco (Florence, 1920) 8Google Scholar, Kolar, A., ‘Inwieweit ist Ovids Ibis von der Ibis des Kallimachos abhängig?’, Berl. Ph. W. 15 (1933) 1244Google Scholar, Cahen, E., Callimaque et son œuvre poétique (Paris, 1929) 70Google Scholar.

8. Perrotta, G., ‘Studi di poesia ellenistica vi. L'Ibis di Callimaco’, SIFC n.s. 4 (1926) 147Google Scholar.

9. See OLD s.v. 7a for examples.

10. La Penna on Ibis 54 concedes this point even though on the strength of Am. 2.17.21–2 he writes of modo at Ibis 56 ‘non è impossibile intenderlo nel senso di “metro”’. The other example he cites, rudem praebente modum tibicine Tusco (A.A. 1.111) is unhelpful because modus there clearly refers not to poetic metre, but to musical measure in accompaniment to Etruscan dance (see Hollis, A. S., Ovid: Ars amatoria Book I (Oxford, 1977) 54Google Scholarad loc.).

11. See La Penna on 55–6 for examples; while he does not entirely rule out modo in the sense of ‘metre’ (cf. n. 10 above), he favours the sense of ‘manner’, ‘mode’.

12. Both papyri derive from the same roll. For the former see Supplementum Hellenisticum fr. 970, pp. 478–81; for the latter see Huys, M., Le Poème élégiaque hellénistique P. Brux. Inv. E. 8934 et P. Sorbonn. Inv. 2254 (Papyri Bruxellenses Graecae vol. II.22; Brussels, 1991)Google Scholar for which reference I am indebted to Watson 260. Barns, J. W. B. and Lloyd-Jones, H., ‘Un nuovo frammento papiraceo dell'elegia ellenistica’, SIFC 35 (1963) 219ffGoogle Scholar. suggest Phanocles as author of P. Sorbonn. 2254; Huys, op. cit. favours Hermesianax. In either case the fragments would predate Callimachus and possibly Moero as well (see Watson 166), thus representing the early development of the Hellenistic (elegiac) curse-tradition.

13. So La Penna xliii. On the (doubtful) possibility that the Arai formed part of the Chiliades see conveniently Watson 81 n. 103.

14. For bibliography on the metrical issue see Watson 79 n. 92; for a summary of scholarly debate on the broader relationship between the two Ibis poems see Watson 79 n. 86.

15. Cf. Horace, , Ars 73–86Google Scholar: after generically delineating the epic hexameter (73–4), elegy (75–8), iambics (79–82) and lyric (83–5), Horace inveighs against ignorance of generic divisions. Why should one be called a poet if one fails to observe generic and stylistic propriety (descriptas … uices … operumque colores, 86; for definition of these terms see Brink 171 ad loc.)? Why, on these Horatian criteria, should Ovid be hailed as a proficient poet when he fails to observe generic propriety in the Ibis? Mitigation for Ovid's seeming breach of generic protocol will be argued for below; but the Horatian passage serves to underline the generic novelty which the Ibis represents.

16. On the former see Knox, P. E., Ovid's Metamorphoses and the traditions of Augustan poetry, Camb. Phil. Soc. Suppl. 11 (Cambridge, 1986) 9ffGoogle Scholar. and 23nn. 1–2 and 9–10 for further bibliography on the Callimachean credentials of Met. 1.1–4; but cf. Kovacs, D., ‘Ovid, Metamorphoses 1.2’, CQ n.s. 37 (1987) 460–2CrossRefGoogle Scholar for isolated opposition to the prevalent view of Ovid's Callimachean adherence. On Callimachean influence on the Fasti see Kenney (n. 5) 428–30.

17. So Herter, H., ‘Literatur zur hellenistischen Dichtung aus den Jahren 1921–35’, Bursians Jahresb. 255 (1937) 178Google Scholar.

18. La Penna xxvii–xxix classifies the Ibis as a deuotio, detecting in lines 97–106 the human sacrifice standard to the ritual. But cf. Watson 208–9, arguing that since by Ovid's time the verb deuoueo was regularly used in the general sense of laying curses on someone (see OLD s.v. 3), the term cannot be restricted to the limited sense of ritualistic deuotio. The influence of defixiones or curse-tablets has also been detected in the Ibis. See Zipfel (n. 7) 5–27 with La Penna xx–xxix for general discussion of the defixio tradition; but cf. Watson 194–216, arguing against anything more than the ‘marginal influence’ of defixiones on the Hellenistic literary curse-tradition.

19. See Watson 114–15 for a lengthier statement of this formulation with nn. 254–5 for a summary of divergent scholarship.

20. On the disputed etymological origins of elegy as a song of lament (cf. Am. 3.9.3–4, Tr. 5.1.5–6) see conveniently Hinds, S. E., The metamorphosis of Persephone: Ovid and the self-conscious muse (Cambridge, 1987) 103 and 160Google Scholar n. 13 for further bibliography.

21. See Watson 103–13 (‘Obscurity and function’) and 168–77 (‘Obscurity and learning’).

22. So e.g. Perrotta (n. 8) 156, Cahen (n. 7) 70, 73, La Penna xli.

23. So Schneider, O., Callimachea II (Leipzig, 1873) 278ffGoogle ScholarPubMed., but immediately countered by Riese, A., ‘Zur Beurteilung von Ovidius' und Kalimachos' Ibis’, Jahrb.f.kl.Phil. 109 (1874) 377–9Google Scholar.

24. Cf. Ibis 51, 639, where Ovid terms his curse poem of 644 verses a mere libellus. For the word as a neoteric diminutive signalling the nature of a book's poetic content and not necessarily its size, see Quinn, K., Catullus: The poems² (London, 1973) 89Google Scholar on 1.1. For exiguus denoting generic humility as opposed to higher callings see Horace, , Ars 11Google Scholar with Brink 167 ad loc., Prop. 3.9.35–6, 4.1.59–60, Ovid, , Fasti 2.3–4CrossRefGoogle Scholar, 6.22, Tr. 2.329–30, 531–2, Pont. 3.3.33–4. On programmatic ‘smallness’ as basic to Hellenistic poetics see Cairns, F., Tibullus: A Hellenistic poet at Rome (Cambridge, 1979) 21Google Scholar and n. 93 for additional bibliography.

25. Rostagni (n. 7) 7–54, arguing further that since lines 299–300 refer to an event of the year 214 B.C. (the death of the Asian king Achaeus), the Greek poem which Ovid translated was the post-Callimachean work of an Alexandrian imitator.

26. ‘Review: A. Rostagni, Ibis’, CR 35 (1921) 67–8Google Scholar = (1972) 1049–51–a blistering attack on Rostagni's whole thesis.

27. La Penna li–liv lists the coincidences.

28. On Ovid's possible use of non-Callimachean sources see La Penna lvi–lxxi.

29. And iambic invective is not on the same elevated generic level as tragedy (cf. grandius … opus, Am. 3.1.70). For grandis marking an implicit contrast between elevated tragedy and lowly invective, cf. R.A. 375–8 (grande sonant tragici, 375) and Horace, , Ars 79–80Google Scholar (grandes … coturni, 80). Note also Callimachus, fr. 112.9 Pf., , where πεζόν anticipates the Iambi. For this interpretation of πεζόν, supported by Horace's later use of pedestris of verse (musa pedestris, applied to his satires, Sat. 2.6.17; cf. Ep. 2.1.251, Ars 95), see dayman, D. E., ‘Callimachus' Iambi and Aitia’, ZPE 74 (1988) 277–8Google Scholar; see also 288 n. 3 for the scholarly history of this interpretation and also for proponents of the now discredited view that at fr. 112.9 Callimachus alludes to his prose writings.

30. In line 7 I follow most editors in reading euitamus amictu, thereby rejecting euitabimus as given in the MSS; but the latter is not out of place (cf. dabis, 8), though a conjecture such as Baehrens' acta is needed to retain it. I remain uncertain about my choice of reading, and no conclusions will be based upon it. For discussion of the difficulty see Fordyce, C. J., Catullus (Oxford, 1964) 404–5Google Scholar.

31. Quinn (n. 24) 455 and Fordyce (n. 30) 403 are so inclined, but without insisting on the point. Cf. Macleod, C. W., ‘Catullus 116’, CQ n.s. 23 (1973) 308CrossRefGoogle Scholar = Collected essays (Oxford, 1983) 185Google ScholarPubMed: ‘it [116] has all the air of being a prelude to the other poems directed at him [Gellius]’.

32. Macleod (1973) 308 = (1983) 185.

33. Macleod, ibid. Cf. Schmidt, E. A., ‘Catulls Anordnung seiner Gedichte’, Philologus 117 (1973) 233CrossRefGoogle Scholar and Forsyth, P. Y., ‘Comments on Catullus 116’, CQ n.s. 27 (1977) 352–3CrossRefGoogle Scholar, taking carmina … Battiadae (116.2) to be a cross-reference to 65.16 by which Catullus links the beginning and end of his elegiac book.

34. Cf. Ellis, R., A commentary on Catullus (Oxford, 1877) 397Google Scholar: ‘I see no reason to doubt that this epigram was written after the others.’

35. Cf. Nemeth, B., ‘To the evaluation of Catullus 116’, ACD 13 (1977) 28–30Google Scholar, detecting in Catullus' imagery of single combat at 116.4 and 7–8 an echo of his references to iambics at 36.5 and 40.2.

36. But cf. Newman, J. K., Roman Callimachus and the modification of the Alexandrian sensibility (Hildesheim, 1990) 67Google Scholar, detecting Archilochean influence in tela (116.7) on the grounds that tela ( = βέλη) ‘presupposes the derivation of iambos from βάλλω found, e.g., in the Etymologicum Magnum (p. 463–27)’. But see n. 37 below.

37. Cf. Wiseman, T. P., Catullus and his world: a reappraisal (Cambridge, 1985) 186Google Scholar, citing Callimachus, fr. 112.9 as a tempting point of comparison for 116.8, but then qualifying the parallel on the grounds that ‘Catullus has already given us iambics in the Callimachean style (especially scazons in the manner of Hipponax) in the collection to which he is now saying farewell. They were as finely polished as the rest of his poems, so Catullus can hardly be announcing that genre with the breaking of “Callimachean” norms which lines 3 and 8 so conspicuously represent’ (on this last point see Macleod (1973) 306–7 = (1983) 183–4). Wiseman goes on to suggest (186ff.) that instead of iambics Catullus threatens to assail Gellius in iambic senarii and the genre of Laberian mime.

38. For πεζόν denoting Callimachus' transition to the Iambi see n. 29 above. On Callimachus' deliberate fusion of two distinct genres, elegiac and iambic, see Clayman (n. 29) 277–86. Cf. Knox, P. E., ‘The epilogue to the Aetia’, GRBS 26 (1985) 59–65Google Scholar, arguing that the epilogue originally concluded Books I–II and was then moved to the end of IV when the work was expanded; by the time of that revision the promise of Iambi would already have been met, rendering the epilogue in the revised version of the Aetia more a statement of metrical versatility than of literary intent. If this is so, then at Ibis 643–4 Ovid modifies the Callimachean nuance by making bellicose intent his priority, not metrical versatility.

39. On this change of ethos see Hutchinson, G. O., Hellenistic poetry (Oxford, 1988) 49ffGoogle Scholar.

40. Cf. Clayman, D. L., Callimachus' Iambi, Mnemosyne Suppl. 59 (Leiden, 1980) 58Google Scholar: ‘Callimachus' approach to invective is not so obviously blunt as Hipponax's. It is nevertheless true that Callimachus' Iambi are full of personal abuse directed at named or more probably pseudonamed individuals and that this abuse sometimes has an obscene character.’

41. For further discussion see La Penna vii–xii and André vi–viii.

42. See n. 4 above.

43. Cf. André vi, identifying Ibis with the unnamed enemy (nescioquis) depicted at Tr. 1.6.13–14 and stating without argument that the Ibis predates Tr. 1.6. But even if André were right, his supposition that Tr. 1.6 is not necessarily an early poem (‘Trist., 1.6 … passe, peut-être à tort, pour un des premiers poèmes’) still allows the Ibis to follow Tr. 1.8, 3.1, 4.9 and 5.8 in chronological and thematic order.

44. Cf. Brink 174 on 89–118: ‘The key term throughout is appropriateness, τὸ πρέπον or the like.’

45. For the distinction between Ovid's private character and public persona cf. Tr. 1.9.59–60 and Pont. 2.7.47–50. The claim is frequently made from Catullus onwards: cf. Cat, 16.5–6, Apul. Apol. 11, Mart. 1.4.8, 11.15.13, Pliny, Ep. 4.14.4–5. Could Ovid be resorting to a Catullan trick to exculpate himself in Tristia 2?

46. Brink 482: ‘H's psychology demands this close link between utterance and the emotion felt. Unless the emotion is genuine, diction will not sound true (108–10).’ Cf. p. 188: ‘His ]H.'s] own concern is to link appropriateness of emotional styles with real emotion – a doctrine of poetic sincerity, a rarish thing in ancient literary criticism.’ But on the problem of gauging artistic/rhetorical sincerity and on the difference between genuine emotion and ‘genuinely’ simulated emotion, see Rudd, N., ‘Theory: sincerity and mask’, in Lines of enquiry: studies in Latin poetry (Cambridge, 1976) 145–81, esp. 171ffCrossRefGoogle Scholar.

47. On the programmatic construction in Tr. 1.1 of this correspondence between Ovid's exilic persona, his shabby book and defective verse, see Williams, G. D., ‘Representations of the book-roll in Latin poetry: Ovid, Tr. 1, 1, 3–1 and related texts’, Mnemosyne 45 (1992) 178–89CrossRefGoogle Scholar.

48. See further Brink 117 and 484 on Ars 32–7 and Horatian ‘wholeness’.

49. But for discordant hints in Ovid's seemingly effusive praise of Augustus here see Harries, B., ‘Causation and the authority of the poet in Ovid's Fasti’, CQ n.s. 38 (1989) 166–7Google Scholar.

50. On the programmatic treatment of Mars as inermis at the start of Fasti 3 see further Hinds, S. E., ‘Arma in Ovid's Fasti–Part 1: Genre and mannerism’, Arethusa 25 (1992) 88–90Google Scholar.

51. Zipfel (n. 7) 5–27 argues that defixiones formed Ovid's main source in lines 67–250; but cf. n. 18 supra and see contra Watson 200–8, where each of Zipfel's arguments is set out and countered.

52. On the owl as a bird of ill-omen in Latin literature see La Penna ad loc.

53. Cf. M. 5.404 (ferrugo applied to the reins of Pluto's chariot) and Aen. 6.303 (ferrugineus used of Charon's boat). For further examples see La Penna on 233, adding Culex 273.

54. For black as a colour of mourning cf. Juv. 10.245, Stat. Silv. 2.1.19 and 5.1.19, and see Toynbee, J. M. C., Death and burial in the Roman world (London, 1971) 46Google Scholar.

55. Cf. also candor in the literary sense of stylistic lucidity (see OLD s.v. 5); so too candidus (see OLD s.v. 9, with further examples cited by Peterson, W., Quintilian: Book X (Oxford, 1891) 68Google Scholar on 10.1.73). Ovid's diction in the Ibis lacks Alexandrian σκοτεινότης (on his ‘insipid and colourless diction’ see Watson 174–5), but his affected obscurantism nevertheless deprives the poem of stylistic candor. Could a fundamental irony of the Ibis be that Ovid defends his claim to moral candor (cf. 8) in a medium which incongruously lacks stylistic candor?

56. For examples of latro applied to a crude oratorical manner see OLD s.v. 5b. For caninus applied to rabid oratory see the examples cited by La Penna on 232.

57. For Homeric and later examples of κύων as an insult see LSJ s.v. II with Lilja, S., Dogs in ancient Greek poetry (Helsinki, 1976) 21–5Google Scholar and index s.v. ‘“Dog” as term of abuse’. For canis as an insult in Roman comedy see Lilja, , Terms of abuse in Roman comedy (Helsinki, 1965) 33Google Scholar; see elsewhere Horace, Epod. 6.1, Petr. 74.9 with OLD s.v. canis 2a and b.

58. The milk of wild animals (cf. 229) conventionally nurtures infants who will develop savage characters. Virgil's Camilla is a conspicuous example (Aen. 11.571); cf. Theocr. Id. 3.15–16, Aen. 4.367, Ovid, Tr. 1.8.43–4, 3.11.3 etc.

59. On which see Dickie (n. 1) 207 n. 64 with Lilja (n. 57) (1976) 105, 115–16.

60. Cf. Dickie (n. 1) 195–203, arguing that in Epode 6 Horace's representation of himself as the noble dog (cf. 5–6) confronting a malicious adversary (canis, 1) amounts to a programmatic denial that he, a new Archilocus or Hipponax (cf. 13–14), is an iambographer driven by wanton malice. He attacks only deserving targets.

61. Cf. Cicero, , Balb. 57Google Scholar, Horace, Carm. 4.3.16, Ovid, Tr. 4.10.123–4, Pont. 3.4.74, Seneca, , Phaed. 492–3Google Scholar etc.

62. For inuidia as unjust in attacking the undeserving see the examples cited by Dickie (n. 1) 204 n. 13; for cowardly inuidia see Dickie 202 with 204 n. 78.

63. Cited by Dickie (n. 1) 201 in connection with canine inuidia in Epode 6.

64. Note also liuent rubigine dentes (776), where the teeth of Invidia are suitably stained with the tartar of spite – an Ovidian variation on her ater dens. Martial 12 pr. 14 and Statius, S. 1.3.103 emulate Ovid in associating liuor with rubigo.

65. ferrugo itself carries the figurative connotation of envy; cf. Laus Pis. 107 (animus … mala ferrugine purus), Aus. 417.62.

66. See Nisbet and Hubbard 339–40 on Horace, Carm. 2.20.4 for extensive examples of the motif.