¹ The tempora are: (1) up to A.D. 14, (2) to early 23, (3) to early 29, (4) to late 31, (5) to March 37.

² The Annals of Tacitus (Cambridge, 1972), i. 37–40Google Scholar.

³ Cornelius Tacitus: Annalen (Heidelberg, 1963), i. 38Google Scholar.

⁴ Goodyear 38, 40.

⁵ Martin, R., Tacitus (London, 1981) 105, 139 43Google Scholar; Luce, T. J., ‘Tacitus' conception of historical change’, Past Perspectives: Studies in Greek and Roman Historical Writing, ed. Moxon, I. S., Smart, J. D. and Woodman, A. J. (Cambridge, 1986), pp. 152–7Google Scholar.

⁶ ‘Postremo suo tantum ingenio utebatur', CQ 24 (1974), 312ff., esp. 316–17CrossRefGoogle Scholar. He explains dissimulatio as one of Tacitus' rhetorical devices for blackening Tiberius' character. It should be added that since Tacitus has chosen to present Tiberius in tyrannical terms, his characteristics will be the exact opposites of those of the good man or ideal ruler (Cic. Off. 2.44 ‘nullum obscurum potest nee dictum eius esse nee factum’; Plin, . Pan. 83.1Google Scholar ‘nihil tectum, nihil occultum patitur”).

⁷ ‘The Question of Character-development: Plutarch and Tacitus’, CQ 33 (1983), 469ff., esp. 481–7CrossRefGoogle Scholar. Further bibliography on the whole question will be found in most of the works mentioned.

⁸ Gill 482, cf. 484 ‘It seems clear that Tacitus' account is intended to show Tiberius' character did not change during his rule but was only more clearly revealed’.

⁹ Luce 155, cf. 156 ‘I do not doubt, let me reaffirm, that he viewed the emperor's ingenium as perverse and unchanging’.

¹⁰ Koestermann ad loc. (‘indem er seiner wahren Natur freien Lauf liess’), Martin 105, Gill 485 (‘The emphasis in suo ingenio is on real character’, his italics), Goodyear 39. In the commentary of Draeger, A. and Heraeus, W.⁷ (Leipzig-Berlin, 1914) we read ‘seitdem er…nur den Eingebungen seines Naturells folgte’Google Scholar, and in that of Nipperdey, K. and Andresen, G.¹¹ (Berlin, 1915) ‘sich aller Selbstbeherrschung entledigte, sich vollständig gehen liess’Google Scholar. The Loeb editor, Jackson, J. (London-Cambridge, MA, 1937), renders ‘follow his own bent’Google Scholar; the Penguin translator. Grant, M. (rev. repr., Harmondsworth, 1974) has ‘he expressed only his own personality’Google Scholar; in the Budé editions we have ‘il se laissa aller au penchant de sa nature’ (Goelzer, H., Paris, 1938)Google Scholar and ‘il ne suivait plus que le penchant de sa nature’ (Wuilleumier, P., Paris, 1975)Google Scholar. Knoche, U., ‘Zur Beurteilung des Kaisers Tiberius durch Tacitus’, Gymn. 70 (1963), 213Google Scholar says ‘gibt sich der Kaiser, wie der Historiker es darstellt, seiner wahren Natur…hin’ (cf. 216); Haussler, R., Tacitus und das historische Bewusstsein (Heidelberg, 1965), p. 322Google Scholar collects various examples of (suo) ingenio uti/uiuere and τῇ ϕύσει χρῆσθαι.

¹¹ Its omission by Gill 485 (above, n. 10) is particularly striking.

¹² This interpretation of Tacitus' words allows solam a role in articulating the argument of the passage: the usual interpretation is ‘only Augustus had a/the mind capable of such a great burden’, although words for ‘only’ are regularly omitted in Latin (see e.g. Kenney on Lucr. 3. 144). My general thesis receives better support from the former rendering (see OLD solus 4) but is in no way embarrassed by the latter.

¹³ See, e.g., Levick, B., Tiberius the Politician (London, 1976), p. 148CrossRefGoogle Scholar.

¹⁴ These partnerships have been fully discussed by Kornemann, E., Doppelprinzipat und Reichsteilung im Imperium Romanum (Leipzig-Berlin, 1930)Google Scholar and Grenade, P., Essai sur les origincs du principal (Paris, 1961)Google Scholar.

¹⁵ See further Purcell, N., ‘Livia and the Womanhood of Rome’, PCPS 32 (1986), 78–105Google Scholar.

¹⁶ Tiberius' solitudo, hinted at in A.D. 21 (3.31.2 ‘longam et continuam absentiam paulatim meditans’), is a motif of Book 4, being urged on him by Sejanus at 41.3 and twice emphasised at 67. 1–2 when he withdraws to Capri. Apart from some Greek intellectuals, Tacitus names only Cocceius Nerva and Curtius Atticus as the emperor's companions on his departure from Rome in 26 (4.58.1): the former starved himself to death in 33 (6.26.1–2), the latter had already been eliminated by Sejanus at some earlier point (6.10.2). For Sejanus himself see below, pp. 204–5 and n. 40.

¹⁷ E.g. Luce 153 ‘a peculiar conflation: three years and a whole book intervene between these two [deaths]’. Knoche 214–15 seems to me not to come to terms with the problem at all. See further below, p. 201 and n. 24.

¹⁸ So too Luce 156, but he sticks to the traditional view of Tiberius' ingenium (above, n. 9). Tacitus regularly uses mores = ‘behaviour’ rather than ‘character’, e.g. 1.54.2, 4.13.2, 13.2.1, H. 4.44.2.

¹⁹ Martin 142.

²⁰ The five casus are: (1) proscriplum…secutus, (2) ubi domum…amore erat, (3)sed maxime…declinans, (4)dein…annis, (5)mox…obtinuil. For the five tempora see above, n. 1.

²¹ That is, Tacitus has (as often) employed two different colores; for their relevance to historiography see Wiseman, T. P., Clio's Cosmetics (Leicester, 1979), pp. 7–8, 26Google Scholar.

²² For this use of sub with a person see Brink, C. O., Horace on Poetry: Epistles Book II (Cambridge, 1982), pp. 426–7Google Scholar.

²³ These two are of course the converse of the colores just mentioned.

²⁴ Martin 105, cf. Goodyear 39, Gill 482. Martin's later statement (141 ‘There is no indication in Tacitus or anywhere else that Tiberius' behaviour was motivated by regard or fear of either Germanicus or Drusus’) is a good illustration of this retrospective inference: Tiberius' alleged regard or fear of the two men is nowhere mentioned in the obituary.

²⁵ Martin 141, cf. Luce 154 (below, n. 36).

²⁶ Goodyear 1.240–1.

²⁷ For the problem see, e.g., Martin 140–2. Knoche expressed the view (a) that the breaks between Books 3/4, 4/5 and 5/6 correspond to the three central turning-points of the obituary at A.D. 23, 29 and 31; and (b) that these breaks represent the ‘Fortfall bestimmter Hemmungen’ (216). While(a) is not in dispute, as will be seen, (b) founders both on Tacitus' reference to Germanicus and Drusus (above, p, 201 and n. 24), which Knoche has been compelled to fudge (above, p. 200 and n. 17), and on the absence of Sejanus from Tacitus’ summary of 23–9 (see Section III above).

²⁸ This interpretation, if accepted, meets the complaint of Luce 153 that there is no difference between the two periods: ‘concealment and artful simulation of virtue remained the same…The obituary says, in effect, that Tiberius was essentially the same under Germanicus, Drusus and Livia: that is, from the start of the Annals to the beginning of the fifth book.’

²⁹ For Tacitus' use of ‘self-correction’ (reprehensio) see Luce 154—5.

³⁰ Any assessment of the remaining cases during this period must take into account the fact that on five other occasions there were acquittals or the charges were dismissed (4.13.2, 29.1, 36.3) and that four of the convictions involved adulterous couples (4.42.3, 52.3). Even the savaging of a suicide's property is obliquely expressed (20.1 ‘saeuitum tamen in bona’), although it admittedly evokes from Tacitus the comment ‘ca prima Tiberio erga pecuniam alienam diligentia fuit’.

³¹ For this use of aut see OLD s.v. 6b; the corrective function of the word at 4.1.1 is in my view supported by the paronomasia saeuire…saeuientibus. The technique would be typical of Tacitus (above, n. 29).

³² In his note on Tacitus' use of aut Goodyear rightly comments that ‘the shade of meaning varies from passage to passage’ (1.8.2n.).

³³ The saeuientes are primarily the delatores and/or Sejanus’ henchmen. Whether they include Sejanus himself depends upon whether one thinks he is able simultaneously to be, as Tacitus says he is, their initium et causa.

³⁴ There is nothing unusual in the fact that these two periods correspond to the programmatic statement at 4.1.1 in reverse order. The Odyssean and Iliadic halves of the Aeneid similarly correspond to the first two words of the epic in reverse order (see, e.g., Bloch, A., ‘Arma virumque als heroisches Leitmotiv”, MH 27 (1970), 207)Google Scholar; and analogous devices are employed by Thucydides at more than one point in his first book.

³⁵ It is worth noting that although Tacitus has often implied Tiberius' saeuitia before now (e.g. 1.4.3–5, quoted on p. 205), his authorial imputations are restricted to very rare and specific occasions (e.g. 1.53.3).

³⁶ Martin 139, cf. Luce 154 ‘The role that Sejanus plays in the fourth stage of the obituary is also odd… This is the only place in the obituary in which Sejanus appears, and…his presence somehow prevented Tiberius' loathsome character from its full emergence.’

³⁷ The context at 5.3.1 is as follows: ‘nam incolumi Augusta erat adhuc perfugium. quia Tiberio inueteratum erga matrem obsequium neque Seianus audebat auctoritati parentis antire; tune uelut frenis exsoluti proruperunt, missaeque in Agrippinam ac Neronem litterae quas pridem allatas et cohibitas ab Augusta credidit uulgus.’ It is important to be clear (a) that this is the only certain passage in Tacitus’ main narrative where the idea of a restraining influence is expressed, (b) that the restraint is not exclusive to Tiberius (as scholars often seem to assume) but is applicable to Sejanus as well. In my view the second of these points detracts from the value of 5.3.1 as a support for the traditional view of the obituary. But, however that may be, it should be noted that the passage does not contradict my own earlier argument that the persons mentioned in the obituary, including Livia (cf. 4.57.3). had been collaborators of Tiberius and are presented as such in the obituary: 5.3.1 is typical of Tacitus (see above, p. 202) in that it simply reveals a different aspect of Livia from that portrayed in the obituary. The passage therefore resembles 4.1.1 on Sejanus as initium el causa and 4.6.2–4 on the general excellence of the years 14–23 (above, pp. 202–203: all three passages provide information which supplements but does not contradict statements made in the obituary.

³⁸ See, e.g., Koestermann ad loc. Similarly oblectis and timuit in the penultimate sentence are picked up by pudore and metu respectively in the final one.

³⁹ In the same way at 1.59.1 (‘ut quibusque bellum inuitis aut cupientibus erat, spe uel dolore accipitur’) inuitis and cupientibus correspond to dolore and spe respectively; or at H. 2.40.3 (‘lit cuique audacia uel formido, in primam postremamue aciem prorumpebant aut relabebantur’) audaeia and formido correspond to primam and postremam etc. Martin 141 takes dilexit and timuit chronologically: ‘At what point Tiberius ceased to love Sejanus and began to fear him is not clear.’

⁴⁰ It is clear that Sejanus could not have exercised such influence if he had not been Tiberius' socius laborum. This is particularly true of the period after Tiberius had settled on Capri in 26, since it was by virtue of his position that Sejanus then realised his ambition of controlling the access of others to the emperor while at the same time maintaining regular access himself (4.41.2, 67.3, 71.3, Dio 58.4.9). But the same is also true of the earlier years, during which he is described by Tacitus as a man who capitalised on his role as socius laborum, pretending to help the emperor while all the time furthering his own aims through a cruel series of trials and assassinations. This indeed is one of the principal themes of Book 4 as a whole, and the statement at 4.59.2 is as good a summary of the position as any: ‘quamquam exitiosa suaderet, ut no n sui anxius cum fide audiebatur. ‘Sejanus’ manipulation of Tiberius was an intrinsic element of his role as socius laborum, a role which he had indeed manipulated Tiberius into bestowing in the first place. Thus it cannot be objected that Sejanus' presentation in the obituary conflicts with the explanation of his presence there as Tiberius' socius or adiutor (above, p. 202).

⁴¹ Kristol, Irving, Encounter 6 (05 1956), 86Google Scholar.

⁴² Despite Goodyear on 1.4.1 there is no actual conflict between these two passages since 1.4.3–5 is in oratio obliqua (nor, perhaps surprisingly, is this one of those places in Tacitus where it is not immediately obvious whether he is speaking in propria persona or not).