¹ On the date of Leg., see e.g. Rawson, E. D., ‘The Interpretation of Cicero's “De Legibus’”, ANRW 1. 4 (1973)Google Scholar. 334–56,. at 335–8.

² There is reason to think that Cicero's philosophical position in general altered from the sort of Platonism associated with Antiochus in Rep. and Leg. to Academic scepticism (in the version of his first teacher, Philo of Larisa) in the philosophical encyclopedia of 46–44 B.C. This is the natural reading of Acad. 1. 13, and Cicero is notoriously hostile to the sceptical Academy (whose characteristic methodology is notably absent in Rep. and Leg.) at Leg. 1. 39. Some might feel the hypothesis of a return by Cicero to the philosophical allegiance of his youth after a period of sympathy with Antiochus to be paradoxical. Others will not find this psychological history too surprising; and it needs to be borne in mind that in Rep. and Leg. Cicero is trying to present himself as a major constructive political philosopher: he has no special occasion to consider epistemological questions there, whereas in the encyclopedia he is undertaking a judicious review of all philosophy, which inevitably involves him in more searching reflection on epistemology and methodology. The whole question is to be the subject of a forthcoming paper by J. Glucker.

³ SVF 11. 42 (= Plu., Stoic, rep. 1035 AB) and 1008 (= Etym. Magn. s.v. τελετή). There is some interesting material in P. Boyancé, ‘Cicéron et les parties de la philosophie’, REL 49 (1971), 127–54. But he fails to see that Div. 11. 1–4 clearly follows the standard order: logic, ethics, physics.

⁴ The editors, following the MSS, print plane, ‘manifestly’. Prof. R. G. M. Nisbet suggests the obviously more appropriate plene, comparing Livy xlii. 52. 13: ‘omnia plena cumulataque habere’.

⁵ Burkert, Walter, ‘Cicero als Platoniker und Skeptiker’, Gymnasium 72 (1965), 175–200Google Scholar, at 193–4, supposes that Cicero is repelled by the obscurity of nature. It is a commonplace of his philosophical writings that ϕύσις κρύπτεσθαι ϕιλεῖ. An Academic sceptic, doubtful about the claims of the dogmatist in any area, will quite properly be most sceptical of all about the possibility of discovering the truths of physics: ‘latent ista omnia, Luculle, crassis occultata et circumfusa tenebris, ut nulla acies humani ingenii tanta sit quae penetrare in caelum, terram intrare possit’ (Acad. 11. 122). So (Burkert suggests) Cicero the Academic will avoid physics. Yet (as Burkert oddly fails to note) only a few paragraphs later on the Academica gives an eloquent rationale of why we should do physics, whether we are sceptics or dogmatists (ibid. 127–8); and, of course, ND does include quite a lot of Epicurean and Stoic physics.

⁶ No historian, I have used such authorities as Liebeschuetz, J. H. W. G., Continuity and Change in Roman Religion (1979), ch. 1Google Scholar; Wardman, A., Religion and Statecraft among the Romans (1982), ch. 2Google Scholar; A. Momigliano, ‘The theological efforts of the Roman upper classes in the first century B.C.’, CP 79 (1984), 199–211.

⁷ I owe the judgement that Caecina's publication of the disciplina Etrusca was ‘a major event’ to Elizabeth Rawson (private correspondence). For documentation and discussion of the late Republican writings on divination mentioned in the text see her new book, Intellectual Life in the Late Roman Republic (1985), ch. 20.

⁸ I do not mean that Cicero himself collected the Roman anecdotes, which derive from the well-known historian and lawyer Coelius Antipater, who seems to have had a digression on prophetic dreams in his work on the Second Punic War (cf. ND 11. 8). But it was Cicero's idea to include the material in a work of philosophy.

⁹ Although (as Nick Denyer points out to me) this difference does account for the fundamental contrast between Fat. as an abstract dialectical argument and Div. as a collection and examination of concrete examples. The two subjects demand different methods, as Chrysippus, for example, had already appreciated, to judge from our evidence of his writings on them. Examples are either irrelevant or at any rate have no probative force in a discussion of fate, which requires subtleties of logic not needed in a treatment of divination. So in adopting different methods Cicero will certainly be responding to the dictates of the subject matter and simultaneously reflecting his Greek sources. What the sources can hardly have determined is the idea of writing a Div. and a Fat. in immediate succession, as παρεργα to an ND, with all the further rhetorical possibilities it opens up.

¹⁰ D.L. vii. 149.

¹¹ On Oracles probably followed the same pattern, to judge from 1. 37, 11. 115.

¹² Polybius vi. 10, 12–14; 47, 7–10.

¹³ cf. Rep. 11. 2–3.

¹⁴ cf. Leg. 11. 23, 111. 12.

¹⁵ A policy particularly characteristic of Pyrrhonian scepticism: see Sextus, PH 1. 23–4, with commentary by Burnyeat, M. F., ‘Can the sceptic live his scepticism?’, in Doubt and Dogmatism (1980), ed. Schofield, M., Burnyeat, M. and Barnes, J., at p. 33 n. 26Google Scholar, and J. Barnes, ‘The beliefs of a Pyrrhonist’, PCPS 208 (1982), 1–29, at 14–15. The best evidence known to me that this was also an Academic attitude is Cotta's series of statements in ND 111. 5–9.

¹⁶ This point is elegantly discussed, with his customary learning and authority, by Burkert, W., Gymnasium 72 (1965), 175–6, 183–4.Google Scholar

¹⁷ De Divinatione, ed. Pease, A. S. (1920–1923), pp 12–13Google Scholar.

¹⁸ CP 79 (1984), 209.

¹⁹ Mary Beard, ‘Cicero and divination: the format of a Latin discourse’, above p. 35.

²⁰ CP 79 (1984), 208–9. His main reason for this conclusion seems to be that ‘there is no reply’ to Cotta's arguments against the Stoa—i.e. (if I understand him aright) none penned by Cicero. But the absence of a reply is surely simply a function of Cicero's standard and natural expository procedure in his philosophical dialogues: first he presents the arguments for a thesis, then those against. The spokesman pro is never given the right of reply: to assume that this indicates Cicero's view of the merits of his case would be like thinking that whichever side in court is permitted the last word is deemed to have won. It is true that the case against is often constructed from Academic materials (as in ND 1 and 111, Div. 11, Acad. 11), but Cicero is at pains to insist that the Academy does not require of its adherents commitment to a party line. He is free to take what views seem to him closest to the truth, after listening to what has been said on each side of the question (11. 150; cf. e.g. Acad. 11. 7–8, TD 11. 4–5).

²¹ De Natura Deorum, ed. Pease, A. S. (1955), p. 9Google Scholar; Pease allows that Cicero ‘perhaps makes a shy concession to public curiosity [cf. ND 1. 10] by indicating in in. 95 his support of the more probable views of Balbus’, ibid. n. 4. He discusses the problem further on pp. 33–6 of his edition.

²² Momigliano accepts this last consequence: he denies that Cicero's advocacy of a religion free of superstition at Div. 11. 148–9 is to be taken seriously, But this is not the only place where Cicero quite properly distinguishes the thesis that there are gods from the thesis that there is something in divination: cf. e.g. 1. 10; 11. 41.

²³ Pease, De Divinatione, p. 13.

²⁴ cf. ND 1. 117.

²⁵ cf. e.g. Ogilvie, R. M., The Romans and their Gods (1969), ch. 4, at p. 61.Google Scholar

²⁶ Do Cicero's letters to Caecina (ad Fatn. vi. 5 (? Jan. 45) and 6 (? Oct. 46)) also indicate an inclination to disbelief in divination? He contrasts his own sort of prophecy, based on political experience, with Caecina's Etruscan lore. We may certainly infer that Cicero was not a fervent, committed believer. But the urbanity with which he teasingly develops the τόπος forbids any more decisive deduction.

²⁷ The contrast between iudicium, an assent expressing belief, and approval (probare) of ‘ea quae simillima veri videantur’, a form of acceptance of the appearances which falls short of the commitment to belief, is just one way of expressing the crucial distinction between the Stoic and Academic epistemologies. The contrast or contrasts have been explored in a number of recent publications and I make no attempt in the present paper to contribute to or even reflect the discussion, except in a very general way. See further M. F. Burnyeat, ‘Can the sceptic live his scepticism?’, and G. Striker, ‘Sceptical strategies’, in Doubt and Dogmatism, ed. Schofield et al.; Frede, M., ‘Des Skeptikers Meinungen’, Neue Hefte für Philosophie 15/16 (1979), 102–29Google Scholar; J. Barnes, art. cit. (n. 15); Burnyeat, M. F., ‘The sceptic in his place and time’, in Philosophy in History, ed. Rorty, R., Schneewind, J. B. and Skinner, Q. (1984), 225–54CrossRefGoogle Scholar. In particular I have not tried to get to grips with the way Cicero's Academic scepticism, which permits and requires him to find the truth or something close to it on any given philosophical topic and so to incline either to the case pro or the case contra (e.g. ND 1. 11–13, Acad. 11. 7, Div. 1. 7), differs from the more thoroughgoing practice of ἐττοχή characteristic of Carneades. Cicero's position seems to be inherited from his teacher Philo of Larisa, whose departures from Carneadean scepticism are the subject of Harold Tarrant's new monograph, Scepticism or Platonism? The Philosophy of the Fourth Academy (1985). See also Michael Frede's article, ‘The sceptic's two kinds of assent and the question of the possibility of knowledge’, in Philosophy in History. I think there is still scope for greater clarity about what precisely probatio of ‘ea quae simillima veri videantur’ commits someone to, and about the obscure concept of the veri simile which is crucial to understanding it.

²⁸ Indeed, the sceptical Academy, under Philo's influence, seems to have favoured an epistemology of just this sort (see Tarrant, Scepticism or Platonism?, ch. 3). They presumably took the view that as rationalists the Stoics were straightforwardly committed to a more foundationalist position, and could therefore be required in controversy to justify what they claimed to know: ‘ratione pugnas; patere igitur rationem meam cum tua ratione contendere’ (ND 111. 9–10). But the Stoic position is, as we have seen, more Moorean than this. See further my article ‘Preconception, argument, and god’, in Doubt and Dogmatism.

²⁹ N. C. Denyer, ‘The case against divination’, PCP 211 (1985), 1–10.

³⁰ This is a convenient place at which to comment on the fact that in Leg 11. 32–3 Cicero goes out of his way to raise the philosophical question of the validity of divination and to express his own unequivocal endorsement of the Stoic theory. I take it that this is in contradiction with the view he favours in Div. The most likely explanation of the contradiction is that he has changed his mind on the issue. There is not much to be said for the alternative explanation favoured by some, that Leg. is a work of political theory, in which Cicero adopts the genus civile of the theologia tripertita associated with Varro and some earlier writers, whereas Div., as a work of philosophy, belongs to the genus physicum (so that it would be a kind of category mistake to think Leg. and Div. conflict—they represent two quite different kinds of discourse). For in general there is no evidence that Cicero knew or (if, as is probable, he did know) liked the theologia tripertita—his own distinction in ND in. 5–10 between customary and rational belief does roughly the same sort of job, after all. And in particular there is no ground for treating Leg. 11. 32–3 as not the expression of a properly philosophical view. Cicero has already made provision for augurs in his legal code (11. 31); auspices could be defended on either a utilitarian or a theological theory of augury—it does not matter which from the point of view of the code. So the passage must be counted as a purely philosophical digression, included for the intrinsic interest of the issue, even though its outcome will make no practical political difference. I suggest that when he wrote Leg. Cicero had not yet reflected on Carneades’ arguments against divination: when he did his views, which may never have been deeply rooted, shifted (cf. n. 2 above). Perhaps, therefore, he had been a believer in divination when he wrote the poems Quintus quotes back at him, and when he had the dream about Marius that Quintus cites.