1. W. Shakespeare, Hamlet 2.2.560-617 (Signet ed.). Quotations are from 561-70, 574 and 616f.

2. Ibid. 2.2.547-54.

3. Aristotle Poetics 1451a.36ff.—for Aristotle, the essential difference between poetry and history.

4. The text used is that of Mayer, Roland, Tacitus: Dialogus de Oratoribus (Cambridge 2001Google Scholar) with some variants as noted. Translations unless otherwise indicated are my own.

5. That these examples hardly help Aper’s moral case has long been recognised. See Fantham, E., Roman Literary Culture: From Cicero to Apuleius (Baltimore & London 1996), 194Google Scholar.

6. On the conventionality of this opening gambit (the fictional supposition that a friend has asked the question to which the work will provide an answer), see Mayer (n.4 above) ad loc. with bibliography there cited. But this very conventionality only serves to highlight the fact that the first third of the work concerns itself with something else. It is clear from the narrative that when Aper and Julius Secundus went to Materaus’ house it was not with the intention of having a debate about why oratory has declined. That Tacitus should produce this conversation as a response to Fabius Justus’ question is deliberately provocative; provocative too is the way in which the reader is forced to try and reconcile what is said before Messalla’s entrance at chapter 14 and what is said after. Cf. Bartsch, S., Actors in the Audience: Theatricality and Doublespeak from Nero to Hadrian (Cambridge MA 1994), 98-101CrossRefGoogle Scholar.

7. Martin’s claim that ‘It is because of oratory’s decline that Maternus has transferred his talent and energies to the writing of poetry’ (Martin, R., Tacitus [London 1981], 61Google Scholar) puts the cart before the horse. When Maternus talks about lucrosae huius et sanguinantis eloquentiae usus recens, ‘the recent employment of this gain-getting and blood-stained eloquence’) at 12.3 (see pp.l28f. below) he is not talking about a decline in rhetorical skill but in moral standards, and the Contrast is not between oratory past and present but between the world as it is and the golden age where there was no need of rhetoric at all. Cf. Mayer (n.4 above), 36 with n.92; Goldberg, S.M., ‘Appreciating Aper: The Defence of Modernity in Tacitus’ Dialogus de Oratoribus’ CQ 49 (1999), 224-37CrossRefGoogle Scholar, at 229.

8. Almost as if to counter this perception, Tacitus as narrator describes Maternus as concitatus et uelut instinctus (‘excited and as if inspired’, 14.1) and has Secundus describe his diction to Messalla as poetarum quam oratorum similior oratio (‘a way of speaking more like that of poets than of orators’, 14.2). The fact remains that Maternus’ speech is a highly rhetorical set piece that explicitly counters the points made by Aper (in spite of Williams’s assertion that it ‘is in no sense a direct reply’—Williams, G., Change and Decline: Roman Literature in the Early Empire [Berkeley/Los Angeles/London 1978], 28Google Scholar—or Luce’s ‘nor do the speeches show in any systematic way where opposing views are in error’: Luce, T.J., ‘Reading and Response in the Dialogus’, in T.J. Luce and A.J. Woodman [eds.]. Tacitus and the Tacitean Tradition [Princeton 1993], 11-38Google Scholar, at 28). Indeed Luce rather argues against himself when he compares Aper’s and Maternus’ arguments on fame in this opening pair of speeches. We should also note that Maternus himself describes his defence of poetry in legal terms (cotidianum hoc patrocinium defendendae aduersus te poeticae exerceo, ‘I engage in daily advocacy against you in defence of the art of poetry’, 4.1). See also pp.l36f. below.

9. To give a recitatio was less dangerous than to publish; adverse comment could be put down to praua interpretatio (‘malicious construction’) as Secundus puts it at 3.2 which the published version would then put right. Indeed, the recitatio functioned as a way of getting ‘comments and criticisms’ which could then be taken into account when preparing the editio. See Dupont, F., ‘Recitatio and the Reorganization of the Space of Public Discourse’, in T. Habinek & A. Schiesaro (eds.), The Roman Cultural Revolution (Cambridge 1997), 44-59Google Scholar, at 48-50, where Maternus’ case is discussed. But Maternus will not heed the implied warning in Secundus’ questions at 3.2 to cut out controversial passages; with him it is ‘publish and be damned’—and quickly, too (maturare libri huius editionem festino, ‘I’m hurrying to get the publication of this book out of the way’, 3.3). Cf. Bartsch (n.6 above), 89f.

10. The text is disputed. Mayer (n.4 above ad loc.) following Stroux, J., ‘Vier Zeugnisse zur römischen Literaturgeschichte der Kaiserzeit I: Maternus, Redner und Dichter’, Philologus 96 (1931), 338-49Google Scholar, places a full stop after tragoediarum; Maternus’ part in the fall of Vatinius would on this reading be introduced as something he achieved in causis agendis (‘in pleading cases’) as opposed to today when his reputation is for tragedy rather than oratory. See also Barnes, T.D., ‘The Significance of Tacitus’ Dialogus de Oratoribus’, HSCP 90 (1986), 225-44Google Scholar, at 239; Devreker, J., ‘Curiatius Maternus’, in F. Decreus & C. Deroux (eds.), Hommages à Jozef Veremans (Brussels 1986), 101–08Google Scholar, at 102f. However, as Mayer acknowledges, the full stop is ‘undeniably somewhat abrupt’, and it would seem more germane to Maternus’ argument to make this a triumph of tragedy rather than oratory (so Bartsch [n.6 above], 200-02); that surely is the appropriate way to deal with someone described as studiorum…sacra profanantem (‘profaning the sacred rites of intellectual pursuits’). Given the description of Vatinius at Annuls 15.34 (corpore detorto, facetiis scurrilibus, ‘[a man of] misshapen body and scurrilous wit’), a recitatio of a drama based on Iliad 2 with Maternus playing Ulysses to (an implied) Vatinius’ Thersites would be a suitably potent putdown—and one that would appeal to Nero’s theatrical interests. Less convincing though in some ways neater is the thesis of Duret, who argues that the play in question is the Domitius mentioned by Aper at 3.4 (Duret, L., ‘Dans l’ombre des plus grands II’, ANRW II.32.5 [1986], 3152-3346Google Scholar, at 3209); but if this were so, given that it had been thus disparagingly introduced by his opponent, one would expect Maternus to identify it. Another textual problem lies in im<perante> Nerone (L. Müller and M. Haupt), which Mayer has also incorporated into his text. The MSS vary between in Nerone and in Neronem. The former has been taken to refer to another praetexta written by Maternus, the Nero (so Bartsch [n.6 above], 104). This (like the Octavia) could not have been written until after Nero’s death, and would almost certainly have to be a Flavian work: i.e., it would not have been composed and recited until only a year or two before the time at which the Dialogus is set. This would make Maternus’ literary career extremely short and so not particularly worth celebrating; and in any case, what sort of achievement for recitatio would it have been to bring down an odious favourite of Nero when the protection of the princeps was no longer available? The reading in Neronem (favoured by Duret, op. cit. 3208) seems to me to make most sense. Taken with potentiam it would mean ‘power over Nero’; such a construction is attested for potestas (see OLD s.v. 1), and would not be inconsistent with Ann. 15.34 or with Nero’s propensity to fall under the influence of morally worthless characters. Maternus’ achievement would lie in the fact that he managed to destroy Vatinius during Nero’s reign, not after it; and it would show that right from the outset Maternus had both the moral courage and the political savvy to challenge the system and get away with it. (The incident may well have been recorded in the lost part of the Annals.)

11. Saleium nostrum, egregium poetam uel, si hoc honorificentius est, praeclarissimum uatem (‘our friend Saleius, an excellent poet—or, if this is the more honorific title, an absolutely outstanding bard’); the bombast and hyperbole in honorificentius and praeclarissimum index the tone of Aper’s remark. Maternus’ use of it is another example of the way in which he responds to specific words and ideas raised by Aper (cf. n.8 above).

12. On the relationship between Tacitus’ Maternus and Plato’s Socrates see Rutledge, S.H., The Literary, Cultural, and Historical Background of Tacitus’ Dialogus de Oratoribus (Diss. Brown 1996), 19-28Google Scholar. However to say as Rutledge does that ‘Tacitus bases Maternus’ character largely on Socrates’ (28) seems to me to be overstating it.

13. Marvell, Andrew, ‘The Garden’ line 16, from Miscellaneous Poems (London 1681CrossRefGoogle Scholar).

14. The reading of H; other MSS have pa(l)lantem. Mayer (n.4 above) adopts Bottischer’s fallacem. Since Maternus is speaking poetically (cf. Secundus at 14.2), surely he should be allowed a poetic phrase or two. famam paltentem is certainly striking, but no more so than Persius’ pallentis mores (‘pallid habits’, Sat. 5.15) or even Virgil’s pallentes morbi (‘pallid diseases’, Aen. 6.275).

15. dulces…Musae, Geo. 2.475; insanum forum, Geo. 2.502 (I owe the latter to Philip Hardie). The ideas and images in this passage might all be seen as deriving from Georgics 2. But while Aper may sneer at those who affect a preference for Lucretius over Virgil (23.2), Maternus’ devotion to poetry is likely to have led him to both. The phrase remotum a…curis is a clear echo of Lucretius’ semotum a curis (DRN 1.51).

16. Gallus is offered the reeds used by the ‘old Ascraean’ (= Hesiod) to compose a song about a particular nemus et lucus. For the imagery of this passage and the significance of Gallus’ failure to separate himself from the world of Mars and Amor in Eclogue 10, see Boyle, A.J. (ed.), The Eclogues of Virgil (Melbourne 1976), 24fGoogle Scholar. and 27-30.

17. Cf. Fantham (n.5 above), 195: ‘Perhaps [Maternus’] most convincing argument is for the innocence of his poetry and the pleasure it affords him—or would be if the works that he names were not all politically colored.’

18. On the possible content of the Domitius see Syme, R., Tacitus (Oxford 1958), 110Google Scholar; Mayer (n.4 above) ad 3.4.

19. The fabula praetexta had a political function right from its inception. In the third and second centuries BCE it celebrated imperium (see O’Neill, P., ‘Triumph Songs, Reversal and Plautus’ Am-phitruo’. Ramus 32 [2003], 1-38Google Scholar, at 17, with bibliography cited at 32 n.57). As with annates, what was celebratory under the republic took on a very different function under the principate. Cf. Flower, H.I., ‘Fabulae praetextae in Context: When were Plays on Contemporary Subjects Performed in Republican Rome?’, CQ 45 (1995), 170-90CrossRefGoogle Scholar.

20. See Boyle, A.J., Tragic Seneca: An Essay in the Theatrical Tradition (London & New York 1997), 52Google Scholar: ‘The moral blackness of [Seneca’s] Thyestes seems absolute.’ That this is Seneca’s reflection on the world in which he found himself seems inescapable; inescapable too the fact that in Atreus we see that kind of exercise of absolute power without moral restraint that Tacitus was to portray so tellingly in the Annals. On the link between Seneca’s Atreus and Nero, see Calder, W.M. III, ‘Secreti loquimur: An Interpretation of Seneca’s Thyestes’, Ramus 12 (1983), 184-98CrossRefGoogle Scholar (although I do not agree with the author’s conclusion that Atreus is presented as a model for Nero to follow rather than one of what Nero had become). Further remarks on the Thyestes theme in n.44 below.

21. Cf. Bartsch (n.6 above), 82 and 89.

22. existere is Muretus’ emendation for the MSS ex his.

23. This is of course a Ciceronian idea (cf. Cic. Rep. 1.68, 3.23, with an acknowledged debt in the earlier case to Plato’s discussion of democracy at his Rep. 562c-563e—see next paragraph). See Köstermann, E., ‘Der Taciteische Dialogus und Ciceros Schrift De Re Publico’, Hermes 65 (1930), 396-421Google Scholar, at 415-17. The subtext here is that the Rome of Cicero’s time was undergoing the Platonic slide from democracy to tyranny.—On the Ciceronian flavour of the Dialogus in general see Luce (n.8 above), 12f.; Fiocchi, L., ‘La cornice drammatica del Dialogus de Oratoribus’, in L. Bertelli … G. Garbarino (eds.), De tuo tibi: Omaggio degli allievi a halo tana (Bologna 1996), 287-302Google Scholar, at 288-91,294, 299f.

24. See esp. Plato Rep. 389e-390a, 390e, 391c-d.

25. The notion of rhetoric as a weapon goes back to Plato’s Gorgias 456c-457c; here Maternus is taking up Aper’s use of the Nicostratus analogy at 10.5 (if you are built to succeed in violent sports, then you have a duty to participate in them). Maternus’ use of this as criticism of rhetoric is another Socratic position he adopts (see Rutledge [n.12 above], 23f.). For a general discussion of the use of this image in the Dialogus see Murphy, J.P., ‘Tacitus on the Education of the Orator’, ANRW 11.33.3 (1991), 2284-97Google Scholar, at 2292f.

26. That this occurred shortly before the dramatic date of the Dialogus gives added point to having Aper mention this confrontation: opposition is dangerous, and one should not seek to match one’s own sapientia against that of the sapientissimus. Cf. Martin (n.7 above), 60. Bartsch (n.6 above, 109f.) aptly draws attention to the fact that Helvidius is linked to Cato and Brutus by Eprius Marcellus at Hist. 4.8.3; to produce a Cato at precisely this time could thus even more readily be seen as oppositional political commentary. Further, Maternus is given a Helvidian persistency: not just Cato but also Thyestes…. As Suetonius’ account shows, it was the repetitive nature of Helvidius’ denigrations that finally led Vespasian to lose patience (cf. also Dio 66.12). And he too was warned (Arrian Epictetus 1.2.19-21).

27. That delatores continued to function in Vespasian’s reign may be inferred from the fact that Titus put a conspicuous end to their activities (Suet. Diu.Tit. 8.5, Dio 66.19.3). For a discussion of their role under Vespasian see Winterbottom, M., ‘Quintilian and the vir bonus’, JRS 54 (1964), 90-97Google Scholar, at 93f. See further n.39 below.

28. On this see Penwill, J.L., ‘Expelling the Mind: Politics and Philosophy in Flavian Rome’, in A.J. Boyle & W.J. Dominik (eds.), Flavian Rome: Culture, Image, Text (Leiden 2003), 345-68Google Scholar, at 345f. and 353-58.

29. Williams (n.8 above), 35-45. Luce (n.8 above, 23) aptly terms this position ‘chronological schizophrenia’.

30. A strong argument for irony is put by Köhnken, A., ‘Das Problem der Ironie bei Tacitus’, MH 30 (1973), 32-50Google Scholar, who concludes (49): ‘Im Dialogus wird die Schleier der Ironie in den Worten des Maternus durchsichtig, wenn man seine zweite Rede vor dem Hintergrund derersten sieht.’ Cf. Borzsak, S., ‘Le “Dialogue” de Tacite et le “Brutus” de Cicéron’, BAGB 1985, 289-97Google Scholar, at 297: ‘…ni Gudeman ni beaucoup d’autres n’ont senti et compris l’ironie latente de Tacite.’ Those who argue that it should be taken at face value include Martin (n.7 above, 64 and 66) and Goldberg (n.7 above), 236f. For Goldberg ‘Maternus’ words…should not be isolated from their cultural context…There is no reason for Maternus, early in Vespasian’s reign, to be any less sincere in expressing these sentiments than Pliny was early in the reign of Trajan.’ But neither should Maternus’ words be isolated from their literary context, and Goldberg seems to me not adequately to address the question he himself poses (loc. cit.): ‘How can the man whose outspokenness provided the very occasion for this discussion now speak so submissively?’

31. Barnes (n.10 above), 243. Barnes identifies the Maternus of this dialogue with the Spanish consular and governor of Syria mentioned in AE 1973.283 and the so-called ‘sophist’ Maternus who was put to death in 90 or 91 for composing a speech against tyrants as a declamation exercise. This is strongly and convincingly argued against by Duret (n.10 above), 3206f.

32. Rudich, V., ‘Accommodation to Corrupt Reality: Tacitus’ Dialogus de Oratoribus’, AncW 11 (1985), 95-100Google Scholar; quotations from 98. See also id., Political Dissidence under Nero: The Price of Dissimulation (London & New York 1993), 24Google Scholar.

33. Bartsch (n.6 above, 114-19) in a detailed discussion of Köhnken (n.30 above) argues that ‘doublespeak’ is ‘a more informative concept than irony’ (115) in interpreting and accounting for the discrepancy between Maternus’ second speech and the first, in that doublespeak by definition has two possible meanings, whereas irony has a surface ‘false’ meaning and a submerged ‘true’ one. As will become apparent, I have considerable sympathy with this view.

34. Some have taken issue with this view, pointing to perceived inconsistencies between the positions put by Messalla and Quintilian. See e.g. Rutledge (n.12 above), 61-83, who argues that Messalla does not believe that Quintilian’s educational program is capable of being established in the current intellectual and social climate. This does not however mean that Messalla disagrees with Quintilian’s basic principles; what it does show is that he is pessimistic about their implementation. See Martin (n.7 above), 61; Barnes (n.10 above), 238; and in particular the detailed arguments of CO. Brink, , ‘Quintilian’s De Causis Corruptae Eloquentiae and Tacitus’ Dialogus de Oratoribus’, CQ 39 (1989), 472-503CrossRefGoogle Scholar, at 484-91, and ‘History in the “Dialogus de Oratoribus” and Tacitus the Historian’, Hermes 121 (1993), 335-49Google Scholar, at 342-45.—The irony of course is that Messalla is himself an instance of what he claims is now impossible: a young man with egregious rhetorical skills.

35. Cf. Costa, C.D.N., ‘The “Dialogus”’, in T.A. Dorey (ed.), Tacitus (London 1969), 19-34, at 26Google Scholar.

36. It is not a sufficient explanation simply to say that it is ‘a literary convention frequently employed in ancient dialogues’ to have an unexpected visitor arrive (Barnes [n.10 above], 228). In the archetypal case of Plato’s Symposium, Alcibiades gives his reasons for coming to Agathon’s house (Symp. 212e); so too with Catulus at Cic. De Or. 2.3.13-14, and by implication the staggered arrivals of Furius, Rutilius, Laelius, Mummius, Fannius, Scaevola and Manilius at Cic. Rep. 1.17-18 (see 1.14 cum…familiarissimique eius [sc. Scipionis] ad eum frequenter per eos dies uentitaturos se esse dixissent, ‘since his [Scipio’s] most intimate friends had said that they would be coming on several occasions to visit him over that time’). Mayer (n.4 above, 37) comments, ‘It is perhaps the only artistic weakness in Tacitus’ plotting that it [Messalla’s arrival] is not motivated in any way’ and seems content to leave it at that. I think we are meant to ask the question. And the claim that it would be embarrassing to have Regulus’ half-brother present when Maternus is attacking the use to which delatores put their rhetorical skills (so e.g. Fiocchi [n.23 above], 294: ‘L’aspra requisitoria di Matemo contro i delatori…sarebbe apparsa indelicata se pronunciata alia presenza di Messalla’; cf. Williams [n.8 above], 29) is also not a sufficient answer; the question is why does Messalla come to Maternus’ house at all.

37. It may be objected that Messalla’s generally favourable treatment in the Histories argues against his being given a negative role of this kind in the Dialogus. But in fact there is no inconsistency; on the contrary, the notices of him in the Histories complement the way he is presented here nicely. The key passages are Histories 3.9 and 4.42. In the former, Messalla is described as Claris muioribius, egregius ipse et qui solus ad id bellum artis bonus attulisset (‘a man of famous ancestry, himself outstanding and the only one who would have brought a sense of culture to that war’). egregius is certainly high praise, but not undeserved for a man who stepped into the breach to command a legion when still in his early 20s (as 4.42 [quoted below] demonstrates, he was still under 25 in the following year); the important thing to note is that this precocious talent is employed in the loyal service of Vespasian. The clause that concludes the sentence is seriously misinterpreted by translators: apart from the fact that qui with the pluperfect subjunctive does not represent an unqualified statement of fact, bonus artes does not mean ‘honourable pursuits’ (Moore [Loeb]), ‘honest purpose’ (Hadas [Modern Library College]), ‘element of integrity’ (Wellesley [Penguin]) or ‘honesty’ (Fyfe/Levene [Oxford World’s Classics]); it means ‘cultural pursuits’ (OLD s.v. ars 6a). Messalla is a learned man, exactly the sort of person who would stand up for the ‘classics’ in a debate such as the one in which he participates in the Dialogus. It is not inconsistent with such a sensibility that he is the source of the tear-jerking (and Lucanesque) story of the soldier who killed his own father at 3.25. 4.42 is even more interesting: magnam eo die pietatis eloquentiaeque famam Vipstanus Messalla adeptus est, nondum senatoria aetate, uusus profratre Aquilio Regulo deprecari….igitur Messalla mm causam neque reum tueri, sed periculis fratris semet opponens flexerat quosdam (‘On that day Vipstanus Messalla won a great reputation for familial responsibility and eloquence, as he dared though not of senatorial age to intercede on behalf of his brother Aquillius Regulus….So Messalla neither addressed the charge nor defended the accused, but by placing himself in his brother’s dangerous situation had changed the view of some [of the senators].’). Messalla’s courageous display of pietas in defending what the context clearly shows to be a person of indefensible behaviour and in so doing running against the tide of senatorial opinion (recalling Lausus’ attempt to protect his equally repulsive father Mezentius in Aeneid 10 with clear potential for a similar outcome) argues a closeness and sense of familial obligation outweighing any other moral or personal considerations. His loyalty to his half-brother coupled with his track record of service to Vespasian combine to make it eminently reasonable that Maternus should regard him with suspicion. He has no reason to believe that Messalla is his friend. Nor do we really have reason to believe that Messalla was a personal friend of the youthful Tacitus, either; Syme offers no evidence for his assertion that he was (Syme [n.18 above], 108).

38. Contrast the reaction to the arrival of Laelius at Cic. Rep. 1.19, where Furius specifically states that there is no need to change the topic of conversation. It is not until 1.31 that Laelius argues that the participants should shift their attention to a different issue. Similarly when Alcibiades arrives unexpectedly in Plato’s Symposium, Eryximachus gives him the opportunity to give a speech in praise of Love just as the rest have been doing before he arrived. These parallels highlight the way in which Messalla is excluded from the initial topic of conversation in the Dialogus. It is not he who chooses to change the subject (as Alcibiades and Laelius do) but Aper.

39. That Messalla’s relationship to Regulus may be the chief explanation for this has not been generally recognised. It is briefly noted as an element in the situation by Murgia, C.E., ‘The Date of Tacitus’ Dialogus’, HSCP 84 (1980), 124Google Scholar n.69, and by Bartsch (n.6 above), 115, but neither develops the point. Goldberg (n.7 above, 228) in fact argues against it, asserting that ‘his [Regulus’] career as a delator ended in 70….After Nero’s death, Regulus only spoke for the defence in criminal cases…’ But even if this were so, there is no reason to be confident in the aftermath of the Helvidius case that Regulus might not be thinking of resuming his former profitable career. We should recall, too, the speech with which Curtius Montanus responds to Messalla’s defence of Regulus at Histories 4.42, which strongly argues that he [Regulus] should be eliminated before he grows yet more powerful, and concludes with the ominous words an Neronem extremum dominorum putastis?…non timemus Vespasianum; ea principis aetas, ea moderatio: sed diutius durant exempla quam mores…optimus est post malum principem dies primus (‘Do you imagine that Nero will be the last of the tyrants?…We do not fear Vespasian—consider this emperor’s age and his moderation—but precedents have a longer life than character….The best day after a bad emperor is the first.’). The clear implication is: seize the moment, because our own experience shows that things will only get worse. As indeed they did at the very next meeting, where in the presence of Domitian and Mucianus, the cases against the delatores lapsed: patres coeptatum libertatem, postquam obuiam itum, omisere (‘The senators abandoned their new-found liberty immediately someone opposed it’, Hist. 4.44). For Maternus and his friends that optimus dies is now long past, and Regulus is still in Rome.

40. On the exaggerated nature of Maternus’ praise of Vespasian, see Kohnken (n.30 above), 47.

41. Costa (n.35 above, 29), following Syme (n.18 above, 116 with nn.4 and 5), notes that ‘Maternus’ words non de otiosa el quieta re loquimur…sed est magna ilia et notabilis eloquentia alumna licentiae [“We are not speaking of a quiet and peaceful thing..but that great and remarkable eloquence is the foster-child of uncontrolled freedom”, 40.2] are a clear echo of a passage in the Brutus: pads est comes otique socio et iam bene constitutae ciuitatis quasi alumna quaedam eloquentia [“Eloquence is the companion of peace, the partner of tranquillity, and, as it were, the foster-child of an already well-ordered state”, 45]’ [Costa’s translations] and interprets this as Maternus ‘correcting’ Cicero. (See also Martin [n.7 above], 63f.; Borzhák [n.30 above], passim.) That no doubt is the message intended for Messalla. But alluding to this passage also draws attention to its context: Cicero’s sentence is framed by images of tyranny, in that just before it he claims that those who are ‘shackled under the lordship of kings’ (regum dominatione deuinctis) have no cupiditas dicendi (‘desire to speak’), while immediately after it he quotes Aristotle’s view that rhetorical theory originated in Sicily ‘after the removal of the tyrants’ (sublatis in Sicilia tyrannis). An interesting colouring for the upcoming representation of a people in obsequium regentis paratos (‘ready to be led by their ruler’) under the leadership of sapientissimus et unus (‘the one most wise’) (41.3-4, quoted p. 127 above. And when Maternus goes on seemingly to contradict Cicero’s assertion with the words quae in bene constitute ciuitatibus non oritur (‘which [sc. eloquence] does not arise in well-ordered communities’, 40.2), we might think of the semantic gulf that divides the Flavian concept of bene constituta ciuitas from the Ciceronian. What price the mixed constitution now?

42. In a short but interesting article Alan Cameron suggests on the basis of numerous other examples of the dialogue form that the Dialogus de Oratoribus is set a few days beofre the death of the principal character and host (Cameron, A., ‘Tacitus and the Date of Curiatius Maternus’ Death’, CR n.s. 17 [1967], 258-61Google Scholar). He points to inter alia the fact that Tacitus carefully estabishes the dramatic date (17.3) and Maternus’ reference to his funerary statue at the end of his first speech (13.6). Cameron supposes that Maternus was in fact denounced [by Regulus?] and that he met his end through enforced suicide—in other words, that this author met the same fate as those who wrote politically charged portraits of Cato (Lucan) and Atreus (Seneca) in the reign of Nero. Cameron is followed by Williams (n.8 above), 34; Murgia (n.39 above), 102 and 122; Rudich (‘Accommodation’ n.32 above), 98; Duret (n.10 above), 3207; Mellor, R., Tacitus (New York & London 1993), 17Google Scholar and 114; Luce (n.8 above), 24; Bartsch (n.6 above), 104-06. Contra Devreker (n.10 above), 104f.; Barnes (n.10 above), 240.

43. On Tacitus’ disapproval of histrionic self-matyrdom, see Agricola 42.4; cf. Penwill (n.28 above), 360-62.

44. On Varius’ Thyestes see Leigh, M., ‘Varius Rufus, Thyestes and the Appetites of Antony’, PCPS 42 (1996), 171-97Google Scholar. Leigh argues that the characterisation of Atreus in this play would have been designed to accord with the negative portrayal of Antony in Octavian’s propaganda; but it wasn’t Antony who murdered the child… At 185-87 Leigh plausibly maintains that every Atreus or Thyestes produced at Rome had a political subtext as ‘the fundamental paradigm for anti-tyrannical discourse in tragedy’. The question is who is the tyrant against whom the discourse is directed.

45. Was Ovid’s Medea likewise a political work? Was the domination of Medea, so telling a feature of Seneca’s version, read or readable as a comment on the domination of Livia, the femina princeps of Pont. 3.1.125? It is of course impossible to know—but given that all the other plays of Maternus whose titles are mentioned have political ramifications, it would be odd if his Medea did not.

46. C. Day Lewis’ translation of Virgil’s obscuris uera inuoluens (Aen. 6.100, of the Sibyl’s ‘second Achilles’ prophecy).

47. Duret (n.10 above), 3211: ‘On n’y parle que d’accusations, et la plaisanterie prend un caractère sinistra.’

48. Mayer (n.4 above) ad loc. observes the ‘urbanity’ of this laughing closure, citing parallels from Ciceronian and other dialogues. That of course is part of the Ciceronian aura with which Tacitus has invested this work (cf. Mayer 27-31 and nn.23 and 41 above); but the issue here is what they are laughing about.

49. Cf. Boyle (n.16 above), 19f.

50. Cato is said to be deo plenus (‘filled by [the] god’, 9.564), just as Phemonoe is plena…Phoebo (‘filled by Phoebus’, 5.186f.) when she gives her prophecy to Appius. Indeed, Cato’s speeches are often described in vatic terms: cf. 2.285 (arcano sacras reddit Cato pectore uoces, ‘From his mystic breast Cato gave forth holy words’) and 9.255 (erupere ducis sacro de pectore uoces, ‘Words burst forth from the leader’s [sc. Cato’s] holy breast’). The difference is that the god who inspires Cato is the god within (tacita quern mente gerebat, ‘which he carried in his silent mind’, 9.564), the Stoic guiding principle, , and his oracular utterances are uncompromisingly republican and uncompromisingly moral, not the product of quasi-madness like Phemonoe’s. The role of the uates now is to draw a morally debased world back to traditional uirtus—both by what one is and what one writes.

51. The date is calculated on what Aper says at 17.3: …ac sextam iamfelicis huius principatus stationem qua Vespasianus rempublicam fouet (‘…and now this fortunate principate’s sixth period of duty in which Vespasian is nourishing the state’). This is in fact the last in the series of reigns which Aper lists, going back to Augustus’ first (suffect) consulship in 43 BCE. This he does not in order to tell Messalla what the date is but to establish that those we term antiqui are not really all that ‘ancient’: after all, Cicero only died 120 years ago (or according to three MSS, 110). All MSS agree that Aper gives Augustus 59 years, which has dutifully been emended to 56; but the problem then is that the figures don’t add up to 120. So it is assumed that Aper is simply ‘rounding up’ (so Mayer [n.4 above] ad loc). Easier perhaps to assume that Aper has made a mistake in his calculation of the length of Augustus’ reign. Either way there seems little justification for making the dramatic date exactly 120 years after 43 (i.e. 77 CE) as argued by Beck, M., ‘Das dramatischer Datum des Dialogus de Oratoribus: Überlegungen zu einer in Vergessenheit geratenen Streitfrage’, RhM 144 (2001), 159-71Google Scholar; despite the fact that statio is not elsewhere in Latin used to mean ‘year’, the context here surely makes it impossible for it to mean anything else. Having used annus three times already in this sentence, Aper is simply employing a piece of rhetorical uariatio which helps highlight the compliment to the current princeps.

52. Syme(n. 18 above), 63.

53. Syme (n.18 above), 109.

54. The parallels are interesting. Both Agricola and Julius Africanus were of Gallic extraction (Agricola, Tac. Agr. 4.1; Africanus, Ann. 6.7.4 [of his father]), and both succeeded in being ‘great men under bad emperors’ (Agr. 42.4), Agricola under Domitian and Africanus under Nero. To Africanus is attributed that magnificent put-down of Nero on the occasion of Agrippina’s murder: rogant te, Caesar, Galliae tuae ut felicitatem tuam fortiter feras (‘Caesar, your Gallic provinces ask that you bear up bravely under your good fortune’, quoted by Quintilian at IO 8.5.15). As an example of doublespeak, this is as good as anything Maternus offers in his second speech in the Dialogs.

55. So e.g. Dudley, D.R., The World of Tacitus (London 1968), 72Google Scholar.

56. Cf. Mayer (n.4 above), 18: ‘This sounds not unlike what Tacitus himself is to do as a historian.’ Cf. also Duret (n.10 above), 3212 (though the judgment on Tacitus is perhaps a little harsh): ‘Maternus n’est pas l’interprète de Tacite mais il incarne l’idéal élevé que l’historien conçoit en secret, se sachant trop faible pour jamais y atteindre; non point que c’est Tacite: ce que Tacite voudrait etre.’ Mayer also suggests (op. cit., 7f.) that Tacitus became disillusioned with forensic oratory after the prosecution and conviction of Marius Priscus for extortion in 100 CE. For Juvenal the knowledge that Marius was living high on the hog on the proceeds in comfortable banishment while the province he had fleeced received not a sestertius in compensation (quid enim saluis infamia nummis?, ‘Who cares about loss of reputation provided he keeps the cash?’, Juv. Sat. 1.48; for Marius as instance see 49f.) was one of the things that turned him to writing satire; for Tacitus it turned him to another way of exploiting the potential of the famous dead (Sat. 1.170f.): the writing of history. Better to be a Virgil than a prosecutor: the Virgil of the Aeneid, with its way of writing Rome under pax et principatus (cf. Köhnken [n.30 above], 48f.; see further p. 139 below).

57. And so presumably Tacitus must have had the Annals project in mind at the time the Dialogus was composed. This suggests a later rather than earlier date for the work. The consulship year (102 CE) of the dedicatee, Fabius Justus, has always been a strong candidate, but arguments have been advanced for as early as the spring of 97 (Murgia, CE., ‘Pliny’s Letters and the Dialogus’, HSCP 89 [1985], 171-206Google Scholar, followed by Barnes [n.10 above], 230-32 and 243f.) or as late as 107 (Syme, R., ‘The Friend of Tacitus’, JRS 47 [1957], 131-35Google Scholar, at 135 n.36). Murgia’s argument rests on a series of what he claims to be verbal echoes of the Dialogus in Book 1 of Pliny’s Letters, but none of these is sufficiently close or remarkable to be convincing (so Brink, CO., ‘Can Tacitus’ Dialogus be Dated? Evidence and Historical Conclusions’, HSCP 96 [1994], 251-80Google Scholar). In Letter 1.6, on which Murgia principally bases his case, Pliny is much more likely to be taking a shot at his old professor, Quintilian, contradicting his assertion that there are too many distractions in the countryside to engage in literary creation (see esp. IO 10.3.22-24). Letter 9.10 on the other hand (as Murgia notes, op. cit., 178; cf. also Luce [n.8 above], 14 n.16) clearly alludes to the Dialogus: itaque poemata quiescunt, quae tu inter nemora et lucos commodissime perfici putas (‘and so there is no progress on the poems which you think are so easy to complete among “the woods and the groves”’, 9.10.2). This plus the fact that the liber mentioned in 8.7 as having been sent by Tacitus for Pliny’s comment sounds much more like the Dialogus than the Histories (so Syme lot: cit.) pushes the balance of probabilities towards the later date.

58. Cf. Bartsch (n.6 above), 106.

59. And not all that dissimilar, either. While Aristotle draws a sharp distinction between poetry and history (see p. 122 above with n.3), Quintilian sees them as very close: historia…est…proximo poetis et quodammodo carmen solutum et…ad memoriam posteritatis et ingenii fantam componitur (‘history is very close to [what you get in] the poets, a kind of poem in prose, and is aimed towards providing a record for posterity and a reputation for talent’, IO 10.1.31). Fame apud posteros (‘among those to come’) is very much part of Maternus’ argument for the superiority of his literary pursuits over Aper’s at 12.4-5; cf. also 13.4-5 and the perpetuitas famae (‘eternity of fame’) achieved by Pomponius Secundus. Cf. Mellor (n.42 above), 114. On the affinity of poetry and history cf. Segal, C.P., ‘Tacitus and Poetic History: The End of Annals 13’, Ramu’s 2 (1973), 105-26Google Scholar, in which the Aeneid and the Annals are classed together as 'Rome's two great poetic histories’ (105).

60. Cf. Rudich (‘Accommodation’ n.32 above), 100: ‘Tacitus offered his own solution to the problem of accommodation. He chose to become the kind of person that he claims he was during the debate between Aper, Messalla and Maternus: a Witness.’ In my view, the choice had already been made when the Dialogus was written (or at least when it was being prepared for publication)—cf. n.57 above.

61. Virgil in fact is the poet whom Maternus holds up against Aper’s Saleius Bassus, not Pomponius Secundus as argued by Bartsch (n.6 above), 120f. At 9.3 Aper dismisses Saleius’ verses as pulchri (‘pretty’) and iucundi (‘pleasant’); for Quintilian he is the poet who never grew up (uehemens et poeticum ingenium Saleii Bassi fuit, nee ipsum senectute maturuit, ‘Saleius Bassus’ talent was poetic and full of verve, but it did not mature with age’, IO 10.1.90). It is the kind of well-constructed, entertaining but essentially harmless stuff that wins government hand-outs (the grant of half a million sesterces, which Aper lauds as an example of Vespasian’s liberalitas 9.5 in tones reminiscent of social scientists talking about grants given for research in the humanities). Virgil’s work did mature and he won the respect of the entire nation, a respect equal to that of the princeps himself (13.2). To write in such a way that enables one both to remain in the good graces of the princeps and to maintain one’s integrity is the ideal to which Maternus aspires; and that is what Virgil achieved in the Aeneid.

62. The case for an anti-Augustan sub-text in the Aeneid is forcefully and convincingly made in Boyle, A.J., The Chaonian Dove: Studies in the Eclogues, Georgics and Aeneid of Virgil (Leiden 1986), 85-176Google Scholar. Cf. also Penwill, J.L., Two Essays on Virgil: Intertextual Issues in Aeneid 6 and Georgics 4 (Bendigo 1995), 5-26Google Scholar.

63. That Octavian (as he then was) tried to save Cicero in the conference at Bononia makes no difference to the eventual outcome; indeed, the cynical horse-trading in human lives that took place on this occasion makes the final decison all the more morally repugnant. So Plutarch: ‘Thus they [the triumvirs] let their anger and fury take from them the sense of humanity, and demonstrated that no beast is more savage than man when possessed with power answerable to his rage’ (Plut. Cic. 46.4, tr. Clough). Mayer (n.4 above, 41) notes the significant connection between the time of Cicero’s death and the establishment of the princtpate, but important though that is the question of who was responsible for Cicero’s death is of at least equal importance.

64. See nn.23,41 and 48 above.

65. This article arises out of a paper first given at the Pacific Rim Roman Literature Seminar at the University of Natal, Durban, in June 1997. I would like to thank those who heard it on that and subsequent occasions for their comments and criticisms. Special thanks to Philip Hardie for his detailed comments on an earlier draft.