¹ This article is based on papers read at the Universities of Cambridge and Reading, the University Colleges of Cardiff and Swansea, and the University of North Carolina at Chapel Hill. I am grateful to the audiences at these occasions for their helpful comments and criticisms, and particularly to Michael Crawford and Philip Stadter.

² Ogilvie, R. M., The Romans and their Gods (London, 1970), 18Google Scholar: ‘The psychology of the Romans was based on the assumption that a man's character is something fixed, something given to him at birth. Nothing could ever alter that character or the actions which flowed from it’. Cf. Ogilvie, , A Commentary on Livy (Oxford, 1965)Google Scholar, note on 3. 36. Syme, I. R., Tacitus (Oxford, 1958), 421Google Scholar: ‘It was the way of thought of the ancients to conceive a man's inner nature as something definable and immutable’. Cf. Goodyear, F. R. D., The Annals of Tacitus (1–6), I (Cambridge, 1972), 37–40Google Scholar, Martin, Ronald, Tacitus (Berkeley and Los Angeles, 1981), 105Google Scholar, and Alexander, W. H., ‘The Tacitean“Non Liquet” on Seneca’, University of California Publications in Classical Philology 14. 8 (1952), esp. 352–77Google Scholar.

³ e.g. Dihle, Albrecht, Studien zur Griechischen Biographie (Göttingen, 1956), 76 ff.Google Scholar; he modifies his statement somewhat on pp. 81 ff.

⁴ Martin, loc. cit., and Alexander, op. cit. 352 ff., cite Theophrastus' Characters in support of their views. But, apart from the relative triviality of this work, its psychological presuppositions are unclear. It is not apparent whether the ‘characters’ are intended to be one-dimensional caricatures (types of personality dominated by one overriding trait), or simply collections of the behavioural ‘marks’ or ‘tokens’ (charaktēres) of a given defective trait, e.g. distrustfulness or superstition. These little sketches give no indication of the origin of the traits, whether innate or acquired. Cf. Jebb, R. C., The Characters of Theophrastus ² (London, 1909), 22–3Google Scholar, Körte, A., ‘Charakter’, Hermes 64 (1929), 77–8Google Scholar, Allport, Gordon, Personality (New York, 1937), 56–8Google Scholar, Fortenbaugh, W. W., ‘Die Charaktere Theophrasts’, Rh.M. 118 (1975), 62–82Google Scholar.

⁵ For euphuïa, see Arist, . EN 3. 5. 17–18Google Scholar, 1114b 12 and context, 10. 9. 3, 8 and 6. 13. 1. In Stoic theory, all men have a natural capacity for excellence at birth, though this often fails to develop fully; see e.g. SVF 1. 566, 3. 214, and cf. Long, A. A. in Problems in Stoicism, ed. Long, (London, 1971), 184Google Scholar, and ‘The Stoic Concept of Evil’, PQ 18 (1968), 336 ffGoogle Scholar. On innate temperament see n. 6 below.

⁶ See, e.g., PI. Prt. 323–7 (much of PI.R. Books 2–4 is concerned with this question; cf. Gill, C., ‘Plato and the Education of Character’, in Plato's Concept of Education, Paideia Special Issue, 1983, ed. Simmons, George)Google Scholar; Arist, . Pol. 7. 12. 6–7Google Scholar,EN 2. 1,10.9.6 ff. (cf. Burnyeat, M. F, ‘Aristotle on Learning to Be Good’, in Essays on Aristotle's Ethics, ed. Rorty, Amélie Oksenberg [Berkeley and Los Angeles, 1980], 69 ff)Google Scholar; cf. Plu. Mor. 2a ff. On the interplay between innate temperament and other factors, see PI. Ti. 86b–87b, Epicur, . On Nature (text in Epicuro, Opere, ² ed. Arrighetti, G. [Turin, 1973], 34. 25–30)Google Scholar, Lucr. 3. 288–322, Sen. Ira 2. 19 ff.

⁷ See, e.g. Arist, . EN 3. 5Google Scholar(cf. Irwin, T. H., ‘Reason and Responsibility in Aristotle’, in Essays on Aristotle's Ethics, ed. Rorty, , 117–55)Google Scholar; for Epicurus' views, see D.L. 10. 133, ref. in n. 6 above, and Rist, J. M., Epicurus (Cambridge, 1972), 90 ff.Google Scholar, Furley, David, Two Studies in the Greek Atomists (Princeton, 1967), 227–36CrossRefGoogle Scholar; for the Stoics, see Long, , in Problems in Stoicism, 173ff.Google Scholar, I and Margaret Reesor, E., Stough, Charlotte, in The Stoics, ed. Rist, J. M. (Berkeley and Los Angeles, 1978), 187–202, 202–31Google Scholar.

⁸ For the ‘choice of lives’ theme, associated with arrival at adulthood, see Guthrie, W. K. C., A History of Greek Philosophy III (Cambridge, 1969), 277–8, 366–7Google Scholar, PI. R. 358–68, 617d ff. (cf. R. 550b, 553 b–d, 572d; the psychic types choose their way of life on becoming adult); Plu. Mor. 12c; Cic. Off. 1. 115 ff. and cf. refs. in n. 46 below.

⁹ See further Rabbow, Paul, Seelenführung (Munich, 1954)Google Scholar, Farrington, B., in Lucretius, ed. Dudley, D. R. (London, 1965), 24 ff.Google Scholar, McGann, M. J., Studies in Horace's First Book of Epistles, Collections Latomus, 100 (Brussels, 1969), esp. 33 ff.Google Scholar, Rudd, Niall, Lines of Enquiry: Studies in Latin Poetry (Cambridge, 1976), 158–9, 163–4CrossRefGoogle Scholar. Cf. also Aristotle, refs. in n. 61 below.

¹⁰ Of the two authors 1 am most concerned with here, Plutarch's philosophical background needs no elaboration. Tacitus is not a philosopher in this sense (passages such as Ann. 6. 22 on fate and free will are exceptional) but would be familiar with the main lines of Stoic and Epicurean moral theory. In any case, as I argue below, there is a good deal of overlap between philosophical and popular thinking on character-development in this period. Cf. n. 23 below.

¹¹ Here I simply sketch this distinction, which I explore more fully in a forthcoming book. This distinction is partly based on the historical development and associations of the terms ‘character’ and ‘personality’, partly on a divergence of perspectives on human nature I find in contemporary thinking.

¹² It is typical of the character-viewpoint to draw a distinction between temperamental qualities, which are a product of the innate, psycho-physical constitution and developed, ethical character-traits. Immanuel Kant, for instance, distinguishes between ‘temperament’, i.e. ‘what nature makes of man’ (e.g. being melancholic or phlegmatic) and ‘character’, i.e. ‘what man makes of himself’, Anthropology, tr. Dowdell, V. L. (Carbondale and Edwardsville, 1978), 195–207, esp. 203Google Scholar. Cf. also refs. in n. 67 below.

¹³ The development of these sorts of concerns in the modern period is brought out well by Trilling, Lionel in Sincerity and Authenticity (Oxford, 1972)Google Scholar.

¹⁴ Examples of personality-terms are melancholic, choleric, introvert, extravert, stolid, volatile, aggressive, defensive, stable, neurotic. In the personality-viewpoint, there is not the same concern to distinguish innate temperamental from developed personality-traits as in the character-viewpoint (cf. n. 12 above); indeed the two types of traits are very closely related. See further Allport, op. cit., Eysenck, H. J., The Structure of Human Personality (London, 1953)Google Scholar, Harré, Rom, ed., Personality (London, 1976)Google Scholar.

¹⁵ The character-viewpoint is also interested in psychological change, which it describes in terms of the development of good or bad character-traits.

¹⁶ e.g. Dihle, op. cit. 76. Curiously enough, Dihle gives no examples of the type of modern biography he is describing; he refers only to the Entwicklungsroman ‘Grüne Heinrich’. Cf. Goodyear, op. cit., who contrasts ancient historiography with ‘the eighteenth-nineteenth-century novel’ in this respect, p. 37, n. 3. Cf. Alexander, op. cit. 355: ‘Character in human life as we view it is something developmental. It is not a thing of fixed nature … but a developmental result, definable at any given time as being a certain something, and again at the next point of review perhaps capable of being exhibited as a certain quite different something’(his italics). This seems to me very overstated; cf. n. 19c below.

¹⁷ cf. Woolf's, Virginia appraisal of Strachey's work in Collected Essays IV (New York, 1967), 221–8, cf. 229–35Google Scholar. For an earlier stage in the debate, see Brake, Laurel, ‘Judas and the Widow’, Prose Studies 4. 1 (1981), 39–54Google Scholar; see further Mendelsen, Edward, ‘Authorized Biography and its Discontents’, in Studies in Biography, ed. Aaron, Daniel (Cambridge, Mass., 1978), 9 ffGoogle Scholar.

¹⁸ Erikson, Erik, Young Man Luther (New York, 1958)Google Scholar. Cf. Shelston, Alan, Biography (London, 1977), 10–13, 62 ffGoogle Scholar. For questions about the notion of personality, in relation to biography, see Clifford, James, ‘“Hanging up Looking Glasses at Odd Corners”: Ethnobiographical Prospects’, in Studies in Biography, ed. Aaron, , 41 ffGoogle Scholar. Modern psychological approaches have been applied to the subjects of ancient biographical and historiographical writing, e.g. Marañón, Gregorio, Tiberius, The Resentful Caesar, tr. Wells, W. B. (London, 1956)Google Scholar, Thiel, J. H., Kaiser Tiberius (Darmstadt, 1970)Google Scholar.

¹⁹ Alexander, op. cit., actually finds in Strachey's work a ‘Tacitean’ sense of ‘the persistent strength of … innate characteristics’ (365 and context). This seems oddly inconsistent with his previous description of modern conceptions of character (cf. 355 ff. and n. 16 above); does he regard Strachey as somehow pre-modern or untypical?

²⁰ Suetonius, of course, like Tacitus, includes accounts of sexual activities; but these are presented with implicit or explicit moral judgments. Contrast Marañón's psychological treatment of ‘Tiberius’ love-life', op. cit. 36–60.

²¹ Momigliano, Arnaldo in The Development of Greek Biography (Cambridge, Mass., 1971)Google Scholar points to various signs of interest in ‘individuals’ and ‘personalities’ in Classical Greek and Hellenistic biography, e.g. in his conclusion, pp. 101–4. But he is not there contrasting ‘personality’ and ‘character’ as I am here, and is thus not dealing with the same issue. I am, in any case, concerned here only with the later period, esp. the early Empire.

²² Plu. Aem. 1 (sometimes printed as a preface to Tim., e.g. in the Loeb), Demetr. 1. 3–6. Cf. Russell, D. A., ‘On Reading Plutarch's Lives’, G&R 13 (1966), 140 ff.Google Scholar, Plutarch (London, 1972), 100 ffGoogle Scholar., Gossage, A. J., ‘Plutarch’, in Latin Biography, ed. Dorey, T. A. (London, 1967), 49–51Google Scholar, Wardman, Alan, Plutarch's Lives (London, 1974), 18 ffGoogle Scholar.

²³ cf. Dihle, op. cit. 60 ff., Leo, Friedrich, Die Griechisch-Römische Biographie (Leipzig, 1901), 188 ffGoogle Scholar. I am not assuming ‘that Plutarch as a biographer thought in Aristotelian terms under the influence of previous Peripatetic biographers’, the view criticised by Momigliano, in Second Thoughts on Greek Biography (Amsterdam, 1971), 251Google Scholar. Rather, I think that Plutarch shared certain general attitudes of post-Aristotelian Greek morality, notably in emphasising ‘the importance of continuous moral choice and … habit in shaping the character of a man’ (Momigliano, ibid.).

²⁴ Russell, , Plutarch, 142Google Scholar, Pelling, C. B. R., ‘Plutarch's Adaptations of his Source Material’, JHS 100 (1980), 127–40, esp. 138–9CrossRefGoogle Scholar; cf. de Lacy, Phillip, ‘Biography and Tragedy in Plutarch’, AJPh 73 (1952), 159–71Google Scholar.

²⁵ Liv. Praefatio, 10: ‘hoc illud est praecipue in cognitione rerum salubre ac frugiferum, omnis te exempli documents in inlustri posita monumento intueri; inde tibi tuaeque rei publicae quod imitere capias, inde foedum inceptu, foedum exitu. quod vites’. Tac, . Arm. 3. 65. 1Google Scholar: ‘praecipuum munus annalium reor, ne virtutes sileantur, utque pravis dictis factisque ex posteritate et infamia metus sit’. Cf. Ann. 14. 64. 3, Hist. 1. 3. 1. Sallust's concern with private and public (ethical) mores is clear inCat. 4 ff. The use of historical figures as moral exempla is pervasive in Rome; cf. Earl, Donald, The Moral and Political Tradition of Rome (Ithaca, 1967), 30 ff., 74 ff.Google Scholar; Litchfield, H., ‘National Exempla Virtutis in Roman Literature’, HSCP 25 (1914), 1–71Google Scholar.

²⁶ Ann. 4. 33. 2: ‘quia pauci prudentia honesta ab deterioribus … discernunt’, cf. 4. 33. 4: ‘malefacta … gloria ac virtus’.

²⁷ Wardman studies these features of Plutarch's approach at length, op. cit., chs. 2–4.

²⁸ Mor. 2a–b. The essay is thought to be spurious; but there seems nothing un-Plutarchian in this particular, rather traditional, formulation; cf. n. 6 above.

²⁹ See, e.g. Mor. 80e ff., 452f ff., refs. in n. 30 below, and cf. n. 9 above, 61 below.

³⁰ Mor. 97 a, cf. 52a–lb, 52f–53a; cf. Russell, , Plutarch, 93–6Google Scholar. Aristotle too associates stability with excellence (see EN 2. 4. 3, 1104b 34, βεβαίως κα⋯ ⋯μετακιως, EE 7. 5. 2, 1239b 12–15, κα⋯ ⋯ άуαθ⋯ς μέν ὅμοιος άε⋯ κα⋯ οὐ μεταβάλλκται τ⋯ ἦθος, ό δ⋯ ϕα⋯λος κα⋯ ⋯ ἄϕρων οὐθ⋯ν ἔοικεν ἔωθεν κα⋯ έσπέρας).

³¹ Mor. 472c; cf. 465f–466a, Cic. Off. 1. 110, Hor. Epi. 1. 7.98, Sen. Tran. 7.2. For the history of this theme, see McGann, , op. cit. 10 ffGoogle Scholar.

³² 551 e. In 552c he says that μεуάλαι ϕύσεις eventually ε⋯ς τ⋯ μόνιμον κα⋯ καθεστηκ⋯ς ἧθος ⋯λθεῖν, but only after ⋯ν σάλῳ διαϕέρονται and when their innate nobility arrives at λόуου α⋯ ⋯ρετ⋯ς συνερу⋯ν ήλικίαν (552 d). Cf. Russell, , Plutarch, 85–7Google Scholar; the neat distinction between phusis and ēthos made here is not consistently maintained (see n. 61 below). The theme of metabolē of character among rulers, and the question of its causes, is well established, as is clear from Polybius. In 9. 22. 7–26. 11, Polybius, discussing Hannibal, considers that his real phusis was obscured by the pressure of circumstances and the advice of associates. In the case of Philip V, he gives a different analysis: δοκεί τ⋯ μ⋯ν ⋯уαθ⋯ ϕ⋯σει περ⋯ αύτόν ὺπάρξαι, τ⋯ δ⋯ κακ⋯ προβαίνοντι κατ⋯ τ⋯ν ήλικίαν ⋯πιуενέσθαι, 10. 26. 8. Cf. Walbank, F. W., A Historical Commentary on Polybius (Oxford, 1967), notes on 9. 23. 10,10. 26. 7–10Google Scholar. For Plutarch's analysis of Philip V, see Arat. 51. 4, discussed below, and n. 64 below.

³³ Brut. 1. The passage recalls Plato's account of excellence as psychic harmony produced by education, R. 410 c ff., cf. esp. (Plu.)ὥσπερ τ⋯ ψυχρήλατα τ⋯ν ξιϕ⋯ν, σκληρ⋯ν ⋯κ ϕύσεως κα⋯οὐ μαλακ⋯ν ἔΧων κτλ and PI.R. 411alO–bl, ὥσπερ σίδηρον ⋯μάλαξεν κα⋯ Χρήσιμον⋯ξ … σκληρο⋯ ⋯ποίησεν.

³⁴ Cor. 1. 2–4, Mar. 2; cf. 45–6, esp. 46. 1–4. Cf. Russell, , ‘Plutarch's Life of Coriolanus’, JRS 53 (1963), 21–8, esp. 27Google Scholar.

³⁵ Dion 4. 2–3; for the comparison with the younger Brutus in this respect, see Dion 1.

³⁶ Alc. 2. 1, cf. 4. 1 ff.

³⁷ 16. 6, 23. 5–6; cf. Russell, , Plutarch, 117, 122–4Google Scholar, and‘Pluarch, “Alcibiades” 1–16’, PCPhS 12 (1966). 37 ffGoogle Scholar.

³⁸ This seems implicit in Ale. 16. 1–6; and perhaps also in Demetr. 1. 7 (which refers to PI. R. 491 e and context, a discussion of the crimes committed by ψυχ⋯ς … εὐϕυεστάτας deprived of good παιδαуωуία); cf. also Mor. 552c–d. This idea may also underlie Plutarch's remark in Ale. 2. 1, τ⋯ δ' ἦθος αὐτο⋯ πρ⋯ς αύτ⋯ κα⋯ μεταβολ⋯ς ⋯πεδείξατο. But his point here is more probably that his ēthos did not have sufficient strength to remain consistent in life–s contingent vicissitudes (in spite of what he says in 23. 5–6); cf.Russell, , ‘Plutarch, “Alcibiades” 1–16’, 38Google Scholar.

³⁹ Plutarch exhibits strongly the tendency common ancient moral theory to conceive a good ēthos as one which is effective in controlling its emotions or pathē; cf. dihle, op. cit. 64–9, Babut, Daniel, Pluarque et le Stoïcisme (Paris, 1969), 318 ffGoogle Scholar., and discussion below.

⁴⁰ A few examples: Sen. Contr. 1. 8. 5, Juv. 14. 1 ff., Cic. Att. 10. 11, Quint. 1. 2. 6–8, 1. 3. 6ff., esp. 1. 3. 12–13, 2. 2. 1 ff., esp. 5–8, Ter. Ad. 414 ff. Plin. 3. 3, Hor. C. 3. 6. 16 ff. See further Bonner, Stanley, Education in Rome (London, 1977)Google Scholar, index entries ‘moral standards’, ‘moral training‘,

⁴¹ e.g. Plaut, , Trin. 301 ffGoogle Scholar,. Most. 133 ff., Hor. Sat. 1. 4. 103 ff., Epi. 1. 2. 55–71, Cic. Off. 1. 1–2, 1. 93–123.

⁴² Ogilvie, ref. in n. 2 above.

⁴³ cf. Earl, Donald, The Political Thought of Sallust (Cambridge, 1961), 60 ffGoogle Scholar. None the less, Earl repeats the usual formula, ‘Sallust seems to have been incapable of conceiving a slowly developing … character’, p. 87.

⁴⁴ Cael. 10, 41, 42–3, 76–7.

⁴⁵ 13. 2. 2. In Nero's case, this attempt at moral guidance is ineffective and ineffectual (‘si virtutem aspernaretur, voluptatibus concessis’). But we should not generalise this case to the belief that all attempts at moral guidance are ineffective, because of the determining power of innate character (as Alexander seems to, op. cit. 358 ff.). Nero's combination of innate, or at least already developed, bad character (13. 2. 1,‘abditis adhuc vitiis’) and absolute power, given to him at seventeen, is not a common one.

⁴⁶ Other examples: Plin. 4. 2. 1, ‘erat puer acris ingenri, sed ambigui, qui tamen posset recta sectari, si patremnon referret’; Pers. 5.34–5, on adopting the toga virilis, ‘cumque iter ambiguum est et vitae nescius error / diducit trepidas ramosa in compita mentes…’; Hor. AP 161–3, ‘imberbis iuvenis…cereus in vitium flecti’.

⁴⁷ cf. refs. in nn. 40–1. Whether the Romans are ever interested in personality, not character, is too large a question to raise here; but in connection with the developing adult, the concern with (moral) character is certainly much more prominent.

⁴⁸ See the examples discussed above.

⁴⁹ e.g.Alex. 2, Demetr. 4,Ale. 2, Lys. 2.

⁵⁰ cf. refs. in nn. 28–32 above.

⁵¹ Dihle points out, op. cit. 63, that in these childhood anecdotes Plutarch tends to use phusis not ēthos to denote the child's ‘character’, thereby indicating its incomplete development. This is perhaps right, though the phusis/ēthos distinction is not consistently maintained, and phusis is also used for adult character; cf. below, esp. n. 61.

⁵² For examples, seeDaitz, S., ‘Tacitus' Technique of Character Portrayal’,AJPh 81 (1960), 30–52, esp. 36 ffGoogle Scholar; as Daitz notes, 40, many of his sketches are quasi-obituaries. The fact that a person can be introduced into the narrative with his characteristics neatly itemised should not be taken as evidence that he enters into life with his character already defined.

⁵³ cf. n. 25 above.

⁵⁴ cf. Russell, , Plutarch, 102–3Google Scholar. This concern is very obvious in Plutarch's comparisons; cf. Erbse, H., ‘Die Bedeutung der Synkrisis in den Parallelbiographien Plutarchs’, Hermes 84 (1956), 398–424Google Scholar. The predominant connotation of ēthos / mores in rhetorical treatises is an ethical one; see e.g. Arist, . Rh. 1. 2, 1. 9Google Scholar, Cic, . De Or. 2. 182–4Google Scholar, Quint. 6. 2. 8, 11, 13, 18.

⁵⁵ Whether or not the adult moral agent is directly responsible for the development of his own vices, he is held liable to judgment, to praise and blame, for the virtues and vices he possesses. This is a standard feature of ancient and modern views of moral character; cf. refs. in n. 7.

⁵⁶ e.g. Sallust's Jugurtha, Plutarch's Brutus the younger, Dion, Coriolanus, etc. Cf. discussion above.

⁵⁷ op. cit. 50 ff. His account reflects Freud's theories about the importance of infantile I experience, particularly sexual experience.

⁵⁸ The ‘character-viewpoint’ lays much more stress on education and conscious self-modification, the ‘personality-viewpoint’ on more complex, less fully conscious processes of change. Dihle has similar observations, op. cit. 78, though not made in these terms. Cf. also Misch, Georg, A History of Autobiography in Antiquity, tr. Dickes, E. W. (London, 1950), vol. 1, 291–2Google Scholar. Of course, not all modern biographers are as psychoanalytic and ‘personality-centred’ as Erikson; for discussion, see refs. in nn. 17–18 above.

⁵⁹ cf. refs. in nn. 29, 32 above, and Wardman, , op. cit. 132–6Google Scholar.

⁶⁰ Russell, , ‘On Reading Plutarch's Lives’, 144–7Google Scholar, Dihle, , op. cit. 84–7Google Scholar, Hamilton, J. R., Plutarch, Alexander: A Commentary (Oxford, 1969), xxxviii–xxxixGoogle Scholar.

⁶¹ Both Russell, ibid. 147, n. 2, and Dihle, op. cit. 63–4, admit that this distinction is not consistently maintained. Examples include Demetr. 2. 3 (τ⋯ ἦθος ⋯πεϕύκει),Dion 8. 2 and Numa 3. 5 (ϕύσει…τ⋯ ἦθος), Arist. 2 (ϕύσις…⋯ν ἤθει), Lys. 23. 2 (ϕύσεις used of developed, adult characters; contrast the σχηδεία/ϕύσις distinction inLys. 2. 2),Aral. 49. 1, ή δ᾽ ἔμϕυτος κακία, τ⋯ν παρ⋯ ϕύσιν ρχηματιαμ⋯ν…διέϕαινεν αὺτο⋯ τ⋯ ἦθος cf.Cato 1. 2, discussed by Dihle, op. cit. 63. Dihle finds the origin of Plutarch's alleged belief in an unchangeable phusis in Aristotle, op. cit. 84–6. But in Aristotle's ethical writings, which are the most relevant to this question, while he thinks phusis makes a contribution to the developed ēthos, he does not, to my knowledge, assert that the phusis-elemenl is unalterable. Indeed, he states in the Politics (7. 12. 7) that ἔνια δ⋯ οὺθ⋯ν ⋯ϕελος ϕ⋯ναι֗ τ⋯ у⋯ρ ἔθη μεταβαλεῖν ποιεῖ…πολλ⋯ у⋯ρ παρ⋯ τοὺς ⋯θισμοὺς κα⋯ τ⋯ν ϕύσιν πε=σιν5;ύ πράτουσι διά τ⋯ν λόуον, ⋯⋯ν πεισθ⋯σιν ἔΧειν βέλτιον. Cf. EN 2.9, 4, advice on correction of natural defects, σκοπεῖν δ⋯ δεῖ πρ⋯ς ἃ κα⋯ εὐκατάϕοροί ⋯σμεν ἄλλοι у⋯ρ πρ⋯ς ἅλλα πεϕύκμεν…εỉς τοὐναντίον δ' έαυτοὺς ⋯ϕέλκειν δεῖ.

⁶² Sulla, 30. 5:ώς τ⋯ ἤθη μένειν μ⋯νειν οὐκ ⋯ώσαις ⋯π⋯ τ⋯ν ⋯ξ ⋯ρχ⋯ς τρόπων κτλ, Sert.. 10. 4: ϕύσεις χρηστ⋯ς…οὐκ ⋯δύνατον…συμμεταβαλεῖν τ⋯ ἦθος cf. Russell, ibid. 146.

⁶³ See e.g. Them. 2. 5, Lys. 2. 2, Cor. 1. 2, Mar. 2. 1, for phusis in this sense.

⁶⁴ cf. Arat. 49. 1, which also, however, touches on the corrupting effect of power, a theme developed in Sulla, 30. 5: ⋯πε⋯ δ⋯ τ⋯ς τύΧης εὐροούσης ⋯παιρόμενος τοῖς πράуμασι πολλ⋯ς μ⋯ν ⋯νέϕυε κα⋯ μεуάλας ⋯πιθυμίας, ήἔμϕυτος κακία, τ⋯ν παρ⋯ ϕύσιν σχηματισμ⋯ν ⋯κβια⋯ομέη κα⋯ ⋯ναδύουσα, κατ⋯ μικρ⋯ν ⋯πεуύμνου κα⋯ διέϕαινεν αὐτο⋯ τ⋯ ἦθος. Plutarch scarcely mentions the evil influence of Demetrius (50. 3–4), which Polybius sees as crucial to the corruption (4. 77. 1–4, 5. 12, 7. 11–14, 10. 26–7).

⁶⁵ In Sert. 10. 3, he considers a possibility (which he does not accept) that Sertorius' phusis was not exhibited in his life but concealed; but it is the phusis of the developed adult he is concerned with, whether it is exhibited or concealed.

⁶⁶ Note MΜ⋯ριος μ⋯ν οὖν…οὺ μετέβαλε τῇ ⋯ξουσία τ⋯ν. Σύλλας δ⋯ κτλ…εἴτε κίνησίς ⋯στι κα⋯ μεταβολ⋯ ϕύσεως ύπ⋯ τύΧης… (Sulla, 30. 4–5). Σερτώριος… ϕύσει δ⋯ ⋯νέκπληκτος ὢν παρ⋯ τ⋯ δειν⋯ κα⋯ μέτριος εὐτυχαν ⋯νεуκεῖν; cf. τ⋯ν ϕύσιν, ϕύσεις χρηστ⋯ς (Sert. 10. 2–4). It is the qualities exhibited in adult lives Plutarch is discussing. Perhaps he thinks that such qualities have their roots (at least in part) in innate qualities, but that idea is neither explicit nor relevant here. This usage is not unique; see, e.g.,αί ϕιλότιμοι ϕύσεις(Lys. 23. 2). When Plutarch uses the phrase παρά ϕύσιν to describe temporary political expedients(Mar. 28. 1, Luc. 6.2, Per. 7.2), he is surely not ascribing to people innate political inclinations but life-long, if interrupted, political attitudes.

⁶⁷ The link between ⋯ρετή and προαίρεσις goes back at least as far as Aristotle; cf. n. 69 below. This is also a feature of modern moral thinking; see, e.g., Hunt, Lester, ‘Character and Thought’, APQ 15 (1978), 177–86Google Scholar, Taylor, Charles, ‘Responsibility for Self’ in The Identities of Persons, ed. Roily, Amélie Oksenberg (Berkeley and Los Angeles, 1975), 281–99Google Scholar.

⁶⁸ See, e.g. Plb. 5. 12. 7, 9. 22. 10 (most rulers are forced to act παρ⋯ τ⋯ν αύτ⋯ν προαίρεσιν; the issue in question is whether τ⋯ς ϕύσεις of rulers are concealed or revealed by circumstances), 16. 28. 5–6. Cf. Plu. Aral. 48. 3, ⋯…προαίρεσις αὺτο⋯ κα⋯ τ⋯ ἦθος.

⁶⁹ See also the combination τ⋯ς ϕύσεις αύτ⋯ν τ⋯ς διαθέσεις Dem. 3. For this aspect of Plutarch's thought cf. Leo, op. cit. 188 ff., Dihle, op. cit. 60 ff., Wardman, op. cit. 107 ff.; see further von Fritz, K., ‘Die Bedeutung des Aristoteles für die Geschichtsschreibung’, Fondation Hardt, Entretiens vol. IV (Vandoeuvres-Geneve, 1956), 104–5Google Scholar.

⁷⁰ cf. refs. in n. 69, Cic, . Off. 1. 66 ffGoogle Scholar., 1. 102 ff., 3. 100, and discussion below.

⁷¹ A few examples: Aesch, . A. 750–6Google Scholar, Hdt. 3. 80. 3, Thuc. 3. 45. 4, PI. Prt. 339 ff. (Simonides' poem), Lg. 875 a–c.

⁷² e.g. Cic, . Tusc. 3. 21Google Scholar, ‘summa potentia summaque fortuna sed ignarum quern ad modum rebus secundis uti conveniret’ (Cicero ascribes the criticism to Theophrastus); Liv. 9. 18. 1–4, esp. ‘nondum merso secundis rebus, quarum nemo intolerantior fuit … ex habitu novae fortunae novique, ut ita dicam, ingenii quod sibi victor induerat’. See further Hamilton, , op. cit. lx–lxiiGoogle Scholar; Arrian's response to these criticisms is discussed by Stadter, Philip A., Arrian of Nicomedia (Chapel Hill, 1980), 103–14Google Scholar.

⁷³ cf. refs. in n. 71 above, esp. Hdt. 3. 80. 3 and PI. Lg. 875 a–c.

⁷⁴ This leads Plutarch to raise the general question whether phusis can change; and phusis here is surely used as a general term for ‘character‘ (the developed character of the adult Sulla) and not in contrast to the terms ēthē and tropoi; cf. n. 66 above.

⁷⁵ Sert. 10.4; cf. τῷ δαίμονι συμμεταβαλεῖν τ⋯ ⋯θος; Plutarch here adapts Heraclitus' dictum, ἦθος ⋯νθρώπῳ δαίμων (fr. 118), which seems to be one of the earliest statements of the ‘ethical’ view of the relationship between character and circumstances.

⁷⁶ The phrase recalls the Platonic picture of complete ⋯ρετή as a unification of the psyche under reason, guaranteeing psychological stability (PI. R. 442–4), an idea elaborated by Aristotle and the Stoics; cf. refs. in nn. 30, 33 above.

⁷⁷ In Sert. 10. 4, all the terms (κατ⋯ λόуον συνεστ⋯σαν…προαιρέσεις κα⋯ ϕύσεις…συμμεταβαλεῖν τ⋯ ἦθος), in combination, contribute to Plutarch's analysis of the precise nature of Sertorius' incompleteness of excellence in character. The analysis given of Sertorius would also fit Sulla (with success, instead of failure, showing up incompleteness of excellence), although Plutarch does not articulate it so fully there. In Sulla tyrannical power is said to alter his emotional tenor (ϕιλόуελως ⋯κ νέου…κα⋯ πρ⋯ς…χα⋯να κα⋯ ⋯πάνθρωπα), as well as his political attitudes.

⁷⁸ Our evidence for the recognition of this disparity goes back to Sen. Clem. 1. 6, Pliny, , HN 14. 144Google Scholar (‘in senecta iam severo atque etiam saevo’), 34. 62 (‘quamquam imperiosus sui inter initia principals’), and is strongly marked in , Suet.Tib. 41Google Scholar(cf. 61 and Gaius 6. 2), , Tac.Ann. 4Google Scholar. 1. 1, 6. 51. 3. The ancient accounts and their problems and inconsistencies are brilliantly analysed by Pippidi, D. M., Autour de Tibère (Bucharest, 1944), 11–87Google Scholar; cf. Balsdon's, J. P. V. D. review in JRS 36 (1948), 168 ffGoogle Scholar. Bibliography is reviewed by Garzetti, Albino, ‘Sul Problema di Tacito e Tiberio’, RSI (1955), 70–80Google Scholar. Cf. Syme, Ronald, Tacitus (Oxford, 1958), 420 ff., 688 ff.Google Scholar, Seager, Robin, Tiberius (Berkeley and Los Angeles, 1972), 255 ffGoogle Scholar.

⁷⁹ D.C. 58. 23. 1: Tiberius appeals to the senate not to give Gaius untimely honours μ⋯…⋯ξοκείλῇ (cf. 58. 12. 6, Sejanus ⋯ξεϕρόνησεν by excessive honours). Plb. 4.48.11: ⋯παρθε⋯ς τοῖς εὺτυχήμασι παρ⋯ πόδας ⋯ξώκειλε; 7.1. 1, πλο⋯τον περιβαλομένους ⋯ξοκεῖλαι ε⋯ς τρνϕ⋯ν κα⋯ πολυτέλειαν; 18. 55. 7, τυχὼν…μεуάλης ⋯ποδοχ⋯ς κα⋯ περιουσίας…εἰς ⋯σέλуειαν⋯ξώκειλε κα⋯ βίον ⋯συρ⋯. , Plu.Mar. 2. 3Google Scholar, ύπ⋯ θυμο⋯ κα⋯ ϕιλαρχίας ⋯ώρου κα⋯ πλεονεξι⋯ν ⋯παρηуορήτων εỉς ὠμότατον κα⋯ ⋯уριώτατον у⋯ρας ⋯ξοκείλας; 45. 6,ε⋯ς ἂτοπον ⋯ξοκεῖλαι παρακοπήν; Brut. 1. 1.σκληρ⋯ν ⋯κ ϕύσεως κα⋯ οὐ μαλακ⋯ν ἔΧων ύπ⋯уου τ⋯ ἦθος ἄΧρι παιδοϕονίας τ⋯ν κεκτημένον ε⋯ςἔτερον ἦθος. See further, LSJ⁹, s.v. ⋯ξοκέλλω, and for the theme, refs. in n. 71 above.

⁸⁰ Suet. Gaius 6. 2 (change at the death of Germanicus), Tib. 61. 1 (change at the death of Sejanus); cf. Tib. 42. 1, 57. 1 (‘saeva ac lenta natura ne in puero quidem latuit’), 59. 1, 62. 1. This theme is well embedded in Tacitus' account; see esp. Ann. 1. 4. 3, 5. 3, 6. 51. 3. It has been suggested that the idea of the deleterious effect of the removal of external restraints was transferred by Tacitus from a political to a personal context, Klingner, F., Studien zur griechischen und römischen Literatur (Zürich, 1964), 658Google Scholar. Perhaps so; though the corrupting effect of absolute power is an old theme (cf. n. 73 above).

⁸¹ The inconsistency of this suggestion with his unfavourable character-sketch of Tiberius in 57. 1 is brought out by Pippidi, op. cit. 82–4. Elsewhere Dio simply notes change in behaviour but attempts no further explanation, e.g. 57. 19. 1 (absence of rivals, again), 58. 28. 5.

⁸² Convulsus, ‘his state of mind was undermined’ (cf.Oxford Latin Dictionary, 2e), is close in meaning to ⋯ξοκ⋯λλειν. For discussion of Arruntius' remark, see Hands, A. R.,‘Postremo suo tantum ingenio utebatur’, CQ n.s. 24 (1974), 312–17, esp. 314–15Google Scholar.

⁸³ cf. refs. in n. 80 above and n. 98 below.

⁸⁴ See refs. in n. 78 above, esp. Pippidi, , op. cit. 50–66Google Scholar; cf. Syme, , op. cit. 420–2Google Scholar, Seager, , op. cit. 262Google Scholar.

⁸⁵ See refs. in n. 2 above.

⁸⁶ Examples in Tacitus: Hist. 1. 50. 4 (Vespasian) ‘solusque omnium ante se principum in melius mutatus est’ (noted by Goodyear, op. cit. 40). Cf. Ann. 3.69.2 ‘multos in provinces contra quam spes aut metus de illis fuerit egisse: excitari quosdam ad meliora magnitudine rerum, hebescere alios’.

⁸⁷ A further issue for Plutarch was whether excellence of character can be lost through bodily illness (Per. 38; cf. Wardman, , op. cit. 46–7, 137)Google Scholar or old age (Fab. 26. 4) or over-relaxation and decadence (comparison of Cim. and Luc, 1–4).

⁸⁸ According to Dio, 57. 23. 4, there were speculations later about Tiberius' having gone mad ⋯ξεστηκ⋯ναι…αøτοȗ τ⋯ν ϕρεν⋯ν), though he does not say whether this was thought to be a result of absolute power (cf. refs. in n. 79 above).

⁸⁹ Ann. 4.1.1, ‘cum repente turbare fortuna coepit, saevire ipse aut saevientibus vires praebere. initium et causa penes Aelium Seianum’; cf. 4.6.1, ‘Tiberio mutati in detenus principatus initium ille annus attulit’; 4. 20. 1, ‘ea prima Tiberio erga pecuniam alienam diligentia fuit’ 6. 1. 1., T ‘saxa rursum et solitudinem mans repetiit, pudore scelorum et libidinum, quibus adeo indomitis.1, exarserat, ut more regio pubem ingenuam stupris pollueret’; 6. 19. 1–2, ‘magnitudinem pecuniae…sibimet Tiberius seposuit inritatusque suppliciis cunctos…necari iubet’. (On the inconsistency between these passages and other parts of Tacitus' account, see Pippidi, , op. cit. 57–66.)Google Scholar For refs. in Polybius, see n. 64 above; for the bad influence on character of an evil adviser I as a continuing theme, see D.C. 73. 1–2 (Commodus), 78. 11. 5 (Caracalla) (cf. Crook, J., Consilium Principis [Cambridge, 1955], 26, 76 ff., 81 ff.)Google Scholar. On avaritia, libido, crudelitas, superbia as stock tyrannical faults, see Dunckle, J. R., ‘The Rhetorical Tyrant in Roman Historiography: Sallust, Livy, Tacitus’, CW 65 (1971), 12–20, esp. 18–20Google Scholar; cf. ‘The Greek Tyrant and Roman Political Invective of the Late Roman Republic’, TAPA 98 (1967), 151–71Google Scholar.

⁹⁰ Plb. 7. 13. 7 (καӨ⋯πε…⋯γγευσ⋯μενοσαἳματοσ ⋯νӨρωπεíου…οὐ λ⋯κοσ ⋯ξ⋯νӨρωπου…⋯λλ⋯τ⋯ραννοσ ἐκ βασιλ⋯ωσ); he refers to PI. R. 565d–e; cf. 571–5, 588c–591e. Cf. Tac. Irefs. in n. 89 above, esp. Ann. 4. 1. 1, saevire, 6. 19. 2, ‘inritatusque suppliciis.…necari iubet’ and context; cf. 6. 39. 2. Cf. Walker, B.The Annals of Tacitus (Manchester, 1952), 210–11Google Scholar.

⁹¹ Ann. 6.6. 1–2 (cf. Suet. Tib. 67); cf. Pl. Grg. 524d–e (the idea of conscious guilt is not present in Plato's passage). For Tiberius as deliberately evil, see discussion below.

⁹² cf. refs. in n. 2 above.

⁹³ cf. Hands, op. cit. esp. 314, n. 6, 317. Hands thinks Tacitus' method of portrayal drew on rhetorical advice, such as that, in accusing a man whose previous life is good, one should argue ‘eum ante celasse, nunc manifesto teneri, etc’ (auct. ad. Her. 2. 5, Hands, 313); but he denies that rhetorical preconceptions compelled Tacitus to adopt this approach. The question which interpretations of Tiberius were available before Tacitus has been much discussed (see e.g. the scholars cited in n. 78 above and March, F. B., The Reign of Tiberius [Oxford, 1931], pp. 233 ff.)Google Scholar; certainty is impossible, but I incline to Pippidi's view, op. cit. 66 ff., that Tacitus' portrait is in many respects an original and fully thought-out conception. I think it likely that he considered, but rejected, alternative psychological models, such as the good man corrupted by power (of which Polybius ‘Philip V, Lucius Arruntius’ Tiberius, and Plutarch's Sulla, are variations; cf. D.C. 57. 13. 6), even though hints of ‘the tyrant's decline’ colour his own account (nn. 89–90 above). Indeed, his own account is a kind of sustained, if implicit, argument against such a view. Tacitus' reasons for forming his interpretation have been much discussed; the Domitian-parallel is often noted, e.g. Walker, , op. cit. 206Google Scholar.

⁹⁴ e.g. Ogilvie, , ‘The Romans and their Gods’, 18Google Scholar, Alexander, , op. cit. 355–6Google Scholar, Daitz, , op. cit. 32–3Google Scholar.

⁹⁵ Strictly speaking, the words do not actually specify that the superbia is innate in Tiberius. They state that superbia was ‘an old and ingrained’ characteristic of the Claudian family. It was well known that family traits, both good and bad, could be transmitted from one generation to another by a variety of means, including family stories (of the kind Suetonius reports, Tib. 1 ff.; cf. Plb. 6. 53–4 for a famous report of this means of transmission). Thus the comment does not specifically assert genetic transmission of this trait to Tiberius, although this is probably implied.

⁹⁶ Saevitia, while associated syntactically with superbia, is not presented as innate in the same way.

⁹⁷ Cf. Goodyear, , op. cit. nn. on 1.4.4, pp. 121–2Google Scholar. Cf. D.C. 58.23.1, Tiberius warns the senate, in the case of Gaius, not to ⋯καíροισ τιμαîσ αὐτΌν, μ⋯ καì ⋯ξοκεíλ⋯ ποι, ⋯παíρῃ. For superbia and saevitia as tyrannical traits, cf. n. 89 above.

⁹⁸ See, e.g., Ann. 1. 7. 7, 1. 10. 7, 1. 11. 2, 1. 24. 1, 1. 33. 2, 1. 76. 4, 2. 28. 2, 2. 42, 3. 2. 3, 3. 3. 1, 3. 44. 4, 4. 57, 4. 71. 3, 6. 24. 3, 6. 50. 1, 6. 51. 3.

⁹⁹ The significance for Tiberius of this period is highlighted, e.g. at Ann.2. 42. 2 ff. (‘Archelaus…invisus Tiberio. quod eum Rhodi agentem nullo officio coluisset, etc.’); 4. 57. 2 (‘Rhodi secreto vitare coetus, recondere voluptates insuerat’). The events are referred to again in 6. 51. 1–2.

¹⁰⁰ That is, nature, family upbringing, youthful contact with society; cf. nn. 40–6 above.

¹⁰¹ cf. n. 99 above.

¹⁰² That is should we see the kind of contrast between ingenium and mores some scholars find in Plutarch between phusis and Ḕthos or ḔthḔ Cf. n. 60 above.

¹⁰³ e.g. Quint. 2. 8 (cf. Oxford Latin Dictionary, senses 4 and 5). In Tac. Ann. 14. 19, where one man (contrasted with another) is said to be ‘ut par ingenio, ita morum diversus’, this is the intended sense, esp. oratorical talent.

¹⁰⁴ See examples in Oxford Latin Dictionary, senses 1 a and b. There are interesting and pointed uses of the term in Plaut. Stich. 126 (‘Edepol vos lepide temptavi vostrumque ingenium ingeni) and context, Most. 135 (‘postea quom immigravi ingenium in meum’), but in neither case does the point seem to be to stress the innate factor.

¹⁰⁵ cf. Liv. 3. 36.1, ‘ille finis Appio alienae personae ferendae fuit. Suo iam inde vivere ingenio coepit, novosque collegas…in suos mores formare’, and Ter. Ad. 69–71, ‘malo coactus qui suom officium facit, / dum id rescitum iri credit, tanu'sper pa vet; / si sperat fore clam, rursum ad ingenium redit’. Cf. Plutarch's use of phusis sometimes for developed character, whether revealed or concealed, e.g. Sert. 10. 3, perhaps Aral. 51. 3, and cf. n. 66 above.

¹⁰⁶ cf. n. 25 above.

¹⁰⁷ Contrast Suetonius' ‘flashback’ (Tib. 57) to Tiberius ‘saeva ac lenta natura…in puero’. In the case of other people in the Annals, Tacitus sometimes comments on such points, e.g. 1. 12. 4, Asinius Gallus ‘patris ferociam retineret’, 2. 43. 2, ‘insita ferocia patre Pisone’, but not in the case of Tiberius. Here, Tacitus is much more concerned to show that the middle-aged Tiberius had, in secret, the same qualities that he claims the aged Tiberius revealed. This interpretation seems designed to offer defence against any suggestion that Tiberius, like Fabius (Plu. Fab. 26.4), might have declined through old age. For the idea of aged infirmity of judgment, see Ann. 1. 7. 7, senili adoptione inrepsisse; Tiberius, Tacitus insists, triumphed over physical infirmity, 6. 50. 1, ‘iam Tiberium corpus, iam vires, nondum dissimulatio deserebat’.

¹⁰⁸ Sometimes, indeed, the saevitia ‘breaks out’ or ‘flashes out’ in spite of his guard (e.g. 1. 4. 3 ‘indicia saevitiae quamquam premantur erumpere’, 1. 74. 4, ‘exarsit…rupta taciturnitate’ etc.; cf. Walker, , op. cit. 62–5, 91, 159)Google Scholar. Elsewhere, however, the saevitia is itself deliberated on, and is only revealed when Tiberius thinks it safe. See refs. in n. 98 above, esp. 4. 57.1–2, ‘saevitiam ac libidinem, cum factis promeret, locis occultantem…et Rhodi secreto vitare coetus, recondere voluptates insuerat'; cf. 1. 4. 4, ‘ne aliud quam iram et simulationem et secretas libidines meditatum’, and Goodyear, , op. cit. 124Google Scholar, note on meditation: ‘The overtone “trained himself in (for)” may accompany the main sense “practised”’; 4. 71. 3, ‘nullam atque…Tiberius…ex virtutibus suis quam dissimulationem diligebat…gnarus lentum in meditando, ubi prorupisset, tristibus dictis atrocia facta coniungere’; cf. 6. 24.

¹⁰⁹ If he can in addition promote such speculation, as he seems to do in the phrase insita…superbia (1. 43), so much the better; this point does not seem to be repeated elsewhere, e.g. in 6,51,3.

¹¹⁰ e.g. 4. 32. 2, ‘truci vultu’; 2. 29. 2, ‘immoto…vultu’; 3. 44. 4, ‘neque…vultu mutato’; 2. 28. 2, ‘non vultu alienatus, non verbis commotior (adeo iram condiderat)’; 2. 29. 2, ‘recitat Caesar, ita moderans, ne...etc.’; 3. 15. 2, ‘Tiberium sine miseratione, sine ira, obstinatum clausumque vidit’; 1.11.2, ‘suspensa semper etobscura verba’; 3. 51.1, ‘solitissibiambagibus’; 4. 31. 2, ‘velut eluctantium verborum’; 4. 52. 3, ‘audita haec raram occulti pectoris vocem elicuere, corruptamque etc.’. Plutarch's range of psychological data (in theory and practice) is discussed by Pelling, op. cit. 135 ff.

¹¹¹ Syme, discussing Tacitus' reproduction of Tiberius' style of speaking, says, ‘He is reproducing a personality, with its characteristic manner’, op. cit. 701, but he does not take this to denote absence of moral judgment of the man's character; cf. pp. 420 ff.

¹¹² Note his self-description as an acute observer and judge of emperors (Ann. 4. 33. 2), an attitude doubtless fostered by experience of the imperial court; cf. Walker, op. cit., ch. ix, Syme, , op. cit. 417–19Google Scholar.

¹¹³ Tacitus suggests that Tiberius' character, as well as his actions and emotions, is, to a large extent, ‘up to him’, a product of (perverse) choice; cf. refs. in nn. 108, 98, 7, 67, above.