This paper was delivered as the keynote address at the Eleventh Pacific Rim Roman Literature Seminar, ‘Writing Revolution’, held at the University of Natal, Durban, South Africa, June 1997. I should like to thank the participants of that seminar for their responses, its director Professor William Dominik for his invitation and hospitality, and the Human Sciences Research Council of the Centre for Science Development in Pretoria, South Africa, for funding my visit to South Africa as an Overseas Research Fellow.

1. It contains what is probably the last datable allusion in Ovid’s poetry: to the year of Germanicus’ triumph, 17 CE (Fasti 1.63)—despite Herbert-Brown, G.’s bold attempt to substitute the year of the decree of Germanicus’ triumph, 15 CE: Ovid and the Fasti: A Historical Study (Oxford 1994), 186ffGoogle Scholar. The last datable reference in Epistulae is to the year of Graecinus’ suffect consulship, 16 CE (Ex Ponto 4.9). There is considerable counterpoint between Ex Ponto 4 and Fasti; indeed Fasti 1’s proem seems the elaborate fulfilment of Ovid’s elaborate promise of a poetic munus to Germanicus in Ex Ponto 4.8 (esp. 31-36, 65-68).

2. Barchiesi, A., Il poeta e il principe: Ovidio e it discorso Augusteo (Roma/Bari 1994), 79Google Scholar, makes unnecessary concessions to the pre-exilic theorists in his attempt to defend the need to treat the poem as ‘una normale opera letteraria’. Barchiesi’s subtle, wide-ranging but more formalist and more circumspect analysis of Ovid’s poetic discourse in Fasti brilliantly unpacks its literary allusivity and texture but blurs the exilic poem’s political force.

3. The succession dispute (4-8 BCE) resulted in the exile of Augustus’ grandchildren, Agrippa Postumus and Julia. The famine of 5-6 CE was especially severe and was ‘exacerbated by a series of natural disasters (fire, flood, earthquake) as well as by additional taxation and conscription as a result of demands of wars in the north, and there was a threat of serious popular unrest (Dio 55.26-7; also Pliny NH 7.149)’ (Herbert-Brown [n.l above], 223).

4. For a succinct description of the extent of Augustus’ control see Barchiesi (n.2 above), 59f.

5. See Degrassi, A., Inscriptiones Italiae 13.2 (Rome 1963), 141Google Scholar.

6. Because the pontiffs had inserted a leap day every three rather than four years, the calendar was ahead by three days. To correct the error no days were intercalated for the following sixteen years (Macr. Sat. 1.14.13-15).

7. Suet. Aug. 31.2; Dio 55.6.6.

8. Ninety days had to be added to 46 BCE to bring the civil and solar years together: hence Macrobius’ description of it as annus confusionis ultimus (Sat. 1.16.3).

9. On this see Beard, M., ‘A Complex of Times: No More Sheep on Romulus’ Birthday’, PCPS 33(1987), 1-15Google Scholar, esp. 7f.

10. On this see Newlands, C. E., Playing with Time: Ovid and the Fasti (Ithaca, NY, 1995), 27-50.Google Scholar

11. Ganymede’s inclusion is especially unnecessary: his identification with Aquarius was dubious, and the position here is odd: half-way between the true rising of Aquarius on Jan. 22, and apparent rising on Feb 22.

12. On this see Wallace-Hadrill, A., ‘Time for Augustus: Ovid, Augustus and the Fasti’, in Whitby, M., Hardie, P., Whitby, M. (edd.), Homo Viator: Classical Essays for John Bramble (Bristol 1987), 221-30Google Scholar, esp. 228; and Hinds, S., ‘Arma in Ovid’s Fasti’, Arethusa 25 (1992), 81-153Google Scholar, esp. 132ff. Barchiesi’s discussion (n.2 above, 70-73), obfuscates.

13. All translations are my own. In the case of Fasti they are preliminary versions of those due to appear in the Penguin Classics Ovid’s Fasti, translated with introductions, notes and glossary by Boyle, A. J. and Woodard, R. D. (Harmondsworth 1999).Google Scholar

14. The Aeneas group (Aeneas, Anchises, Iulus) in the north exedra is caricatured in the wall painting from a villa near Stabiae, discussed by Zanker, P., The Power of Images in the Age of Augustus, tr. Shapiro, A. (Ann Arbor 1988), 209Google Scholar (fig. 162). The group are represented as apes with dogs’ heads and large phalloi. Ovid was not the only one to treat the imperial iconographic program with less than reverence.

15. Fantham, Elaine, ‘Rewriting and Rereading the Fasti’, Antichthon 29 (1995), 42-59CrossRefGoogle Scholar, misleads when she states (54) that Romulus ‘was neither a Julian nor an ancestor of the Princeps himself. It is true that Romulus died childless, leaving no direct descendants. But he shares with Augustus descent from lulus and Aeneas, a matter which Virgil’s Aeneid parades (1.267-77). In Ennius’ version Romulus’ Aeneas ancestry is even more immediate: he is the grandson of Aeneas via Ilia, Aeneas’ daughter.

16. For a discussion of a Domitianic sculptural representation of the pediment of the restored temple, see Wiseman, T. P., Remus (Cambridge 1995), 146-48.Google Scholar

17. For a discussion see Wallace-Hadrill (n.12 above), 223f.

18. Romulus was directly associated with three festivals: Quirinalia (Feb. 17), Parilia (Apr. 21), Consualia (Aug.), only two in the first six months. But Romulus is extended into other parts of Fasti:: Formation of the Calendar (1.27ff.); Pater Patriae (2.119ff.); Lupercalia (2.359-80, 429-52); Mars’ seduction of Ilia/Silvia (3.11-76); Sabine Rape (3.179ff.); Lemuria (5.451ff.); etc.

19. Fasti 2.143: te Remus incusat, ‘Remus accuses you’. The later re visionary account by the ghost of Remus (Fasti 5.451-80) poses similar problems of belief: Remus’ narrative is as credible as (1) the derivation of Lemuria from Remuria, and (2) the narrator’s source, Mercury, who is not only emphasised as the sole source (5.445-50) but—in a gratuitous passage later in the book—as quintessentially the god of deceivers (Fasti 5.679-92). See further Harries, B., ‘Causation and the Poet’s Authority’, CQ 38 (1989), 164-85CrossRefGoogle Scholar, and CE. Newlands (n.8 above), 68f., 119-21.

20. See Hinds (n.12 above), 144f. For variants on Remus’ death see Wiseman (n.16 above), 9-11.

21. See Fantham, E., ‘The Role of Evander in Ovid’s Fasti’, Arethusa 25 (1992), 155-71Google Scholar, esp. 157.

22. Oddly ignored by Barchiesi (n.2 above, 183-85), who focuses on Erato as narrator and her defence of Claudia’s ‘stile divita elegiaco’. The rewriting of Propertius 1.2.1 at Fasti 4.309f. is telling. But what is told is not (pace Barchiesi) that Claudia has the elegiac lifestyle of a Cynthia but that—like Ovid—she appears to.

23. See Harries (n.19 above), 176f.

24. De Lingua Latina 6.33. He was followed in this by Lucius Cincius who derided those ignorant enough (i.e. Fulvius and Junius) to associate the name of the month with Venus (Macr. sat. 1.12.12). Varro and Cincius noted the absence of any significant festival or sacrifice to Venus in the month and derived the month’s name from aperire. See further Herbert-Brown (n.l above), 85-95.

25. Direct priestly association between the pontifex maximus and Vesta dates from her transference to the Palatine in 12 BCE. For Livy there is no direct association between the chief priest and Vesta. The pontifex maximus resides in the domus publico (not the Regia) adjoining the house of the Vestals and has power over the Vestal Virgins, but they alone have a direct priestly relationship with Vesta, whose eternal flame they alone tended, and whose sacred space was not to be penetrated by a man (non adeunda uiro, Fasti 6.450). Vesta’s cult according to Livy (1.20.5-7) was founded by Numa before the existence of the chief pontificate. See Herbert-Brown (n.l above), 63-81.

26. See Herbert-Brown (n.l above), 98f., developing Weinstock’s thesis.

27. Cf. the echo of Aeneas’ words at Aen. 6.129 in the description of Augustus’ opus and pietas in avenging Julius Caesar (3.709).

28. On this see especially Harries (n.19 above), Newlands (n.10 above), 51-86.

29. This seems to have been Varro’s view too, at least if we are to believe Pliny, who claims to be following Varro in his account of the purpose underlying the establishment of the Floralia: ut omnia bene deflorescerent (‘that everything might shed its blossom well’, NH 18.286)—quoted by Fantham (n.30 below), 51.

30. On Ovid’s treatment of Flora I am indebted to many of the fine observations of Fantham, E., ‘Ceres, Liber and Flora: Georgic and Anti-Georgic Elements in Ovid’s Fasti’, PCPS 38 (1992), 39-56Google Scholar, esp. 49ff.

31. In Book 6 (97-100) he is afraid to choose lest he offend two of the deities.

32. The Ovidian narrator never cites any literary sources (though he mentions on four occasions having consulted ancient calendars and at 3.844, in the discussion of Capta Minerva, he cites an inscription). Contrast the Callimachean narrator who at Aitia fr.75.53-77 says that the story of Acontius and Cydippe comes from the history of Xenomades. See further Newlands (n.10 above), 66f.

33. E.g., the focus on nudity at the Lupercalia together with the Hercules, Omphale, Faunus tale: Newlands (n.10 above), 59ff.

34. The number of authorities interviewed or addressed by Ovid in each book are arranged thus: 1:3; 2:1; 3:3; 4:3; 5:7; 6:9. See Newlands (n.10 above), 79.

35. Note that dissensere is a ‘hapax’ in Ovid.

36. The other goddesses are Juno and Hebe (=Juventas). All three goddesses were associated with the Augustan house: Juno with Livia (Fasti 1.650, etc.); Juventas with Augustus, who restored the goddess’s temple (Res Gestae 19); Concordia with Tiberius and Livia (Fasti 1.645-50), and with Augustus (Suet. Tib. 20).

37. See Harries (n.19 above), 173 and passim.

38. Apollonius ó , succeeded Eratosthenes as librarian at Alexandria.

39. Conte, G.B., Genres and Readers, tr. Most, G.W. (Baltimore and London 1994), 122.Google Scholar

40. Ovid’s formulation is intentionally ambiguous. Barchiesi (n.2 above, 266) errs when he says, ‘sex…totidemque può solo significare dodici.’ Barchiesi has not been alone.

41. Cf. both Minerva’s rupturing of Arachne’s tapestry at Met. 6.131 (rupit), and Ovid’s similar description of Metamorphoses in Tr. 1.7.14: infelix domini quodfuga rupit opus, ‘the work unhappily ruptured by its master’s exile’.

42. Book 3: unsuitability of bellicus Mars for imbellis elegy (has to disarm); Book 4: conjunction of maiord and elegy.

43. The main precursor of Ovid’s Heroides is Propertius 4.3.

44. On this see the influential article of D. Feeney, ‘Si licet etfas: Ovid’s Fasti and the Problem of Free Speech under the Principate’, in Powell, A. (ed.), Roman Poetry and Propaganda in the Age of Augustus (Bristol 1992), 1-25.Google Scholar

45. See esp. Newlands (n.10 above), 146-74.

46. Perhaps the most extraordinary ‘exclusion’ in the pater patriae passage is any mention of whom Augustus was father of.

47. See Holleman, A.W.J., ‘Ovid and the Lupercalia’, Historia 22 (1973), 260ff.Google Scholar

48. The episode is sufficiently disturbing to Fantham (n.15 above, 57) that she decides that Ovid wrote the Priapus’ attempted rape of Lotis in Book 1 to replace it and would have removed the incident in a final version. One thing Fantham and I agree upon is ‘the unsuitability of [Ovid’s] fiction (mendacia uatum) to the dignity of Vesta—and surely to the official standing of the goddess who maintained the generative continuity of Rome’ (57). The resulting destabilisation of ‘official’ images of Romanitas is, however, pace Fantham, far from exceptional in this poem.

49. Mentioned in the Fasti Praenestini, Feriale Cumanum and elsewhere; but omitted in some calendars, e.g. the Tiberian Fasti Verulani.

50. Alluded to at Fasti 2.615f., 5.140. For the Augustan importance of the Lares Compitales, see Phillips, C.R. III, ‘Roman Religion and Literary Studies of Ovid’s Fasti’, Arethusa 25 (1992), 55-80Google Scholar, esp. 65.

51. According to the Fasti Praenestini.

52. Parilia was marked by games ‘to commemorate the announcement of Caesar’s victory at Munda’: see Beard (n.9 above), 9, citing Dio 43.42.3, 45.64.

53. Defined by Quintilian at Inst. Or. 9.2.64 as when ‘from something said something hidden is dug out’, ex aliquo dicto latens aliquid eruitur; among the occasions for its use: ‘if it is not safe enough to speak openly’, si dicere palam parum tutum est, 9.2.66. Cf. Ovid Rem. 359f.: sed tu/ingenio uerbis concipe plura meis (‘you with your intelligence understand more than my words’).

54. The phrase is that of Hinds, S., ‘Generalising about Ovid’, in Boyle, A.J. (ed.), The Imperial Muse: To Juvenal Through Ovid (Berwick, Australia, 1988), 26.Google Scholar

55. On this see esp. Barchiesi (n.2 above), 272-77; Newlands (n.10 above), 211-18, and ‘The Ending of Ovid’s Fasti’ in Boyle, A.J. (ed.), Roman Literature and Ideology: Ramus Essays for J.P. Sullivan (Bendigo 1995), 129-43.Google Scholar

56. Phillips (n.50 above), 63, for example, citing Berger, P., The Sacred Canopy (New York 1967).Google Scholar

57. Poster, Mark, in Groden, M. and Kreiswirth, M. (eds.), The Johns Hopkins Guide to Literary Theory and Criticism (Baltimore and London 1994), 278.Google Scholar

58. Note the revisionary nature of Augustus’ religious revival: editing (including/excluding) Greek and Roman prophetic writings, most especially the Sibylline books; reviving those rites he wished to revive, restoring those temples he wished to restore: see Res Gestae, passim.

59. Contrast the openings of Amores, Ars Amatoria, Metamorphoses, on the one hand, and that of Fasti on the other. Germanicus (in Fasti 1) and Augustus (in Fasti 2) have replaced the fieri ve gods of poetic inspiration.

60. At least Ovid thought him Roman, and emphasises his non-Greekness (Fasti 1.90) in opposition to Callimachus, whose opening of Aitia is recalled. On Fasti’s Callimacheanism Barchiesi (n.2 above) is outstanding.

61. It should be noted that this was not inevitable. The senate decreed in 45 BCE that Quintilis be renamed Iulius, but after the Ides of March the change was ignored until the consul Antony announced that the Ludi Apollinares were to be held Nonis Iuliis. Cicero and the conspirators wished to retain Quintilis (Cic. Att. 16.1.1, 4.1). The public sacrifice on Caesar’s birthday was not performed in 44 or 43, but was made compulsory in 42 by the triumvirs. Octavian’s success made the name stick.