¹ The nature and proper use of persuasion interested Plato throughout his career and we find relevant material in, e.g., the Gorgias, the Phaedrus, the Philebus, and the Republic. Here, however, I cannot undertake the large project of charting this development, but shall confine my attention mostly to the Laws and make only occasional reference to other dialogues.

² Hereafter, Stephanus page numbers occurring without a title refer to the text of the Laws. I have used the Greek text of des Places, É. and Diès, A., Platon, Oeuvres Complètes xi–xii (Paris, 1951–1956)Google Scholar and have quoted, with modifications, the translation of Pangle, T., The Laws of Plato (New York, 1980).Google Scholar

³ Plato repeatedly contrasts the reasons offered in the preludes for doing what the law requires with the fear of sanctions, e.g. 721e, 722e–723a and 853b–d. For the idea that the lawgiver tries to show that the behaviour required by the law is in the best interests of the citizens, see 858d, 891a, 905c and section III.

⁴ As we shall see in more detail below, Plato's practice of persuasion in the Laws involves several important differences from the ethical and political philosophy of the Republic. My main concern here is to explicate the Laws' position, but I shall note from time to time especially significant differences from the Republic.

⁵ In the relevant passages, Plato contrasts πεθειν with: ‘threatening a penalty’ (παπειλσας τν ζημαν 719e9, cf. πειλεῖν σκληρς 885d1 and πειλεῖν 890b5), ‘force’ (βᾳ 722b6) and ‘a tyrannical command’ (τυραννικòν πταγμα 722e7–8).

⁶ βᾳ: Tragica Adespota 402 and Xen. An. 5.5.11. βιζεσθαι: Xen. Cyr. 6.1.34. ναγκζειν: Plato, , Hipparch. 232bGoogle Scholar. ν δλῳ: Soph, . Phil. 102Google Scholar, cf. 612 where πεθειν is spelled out as πεθειν λγῳ. Although persuasion is usually contrasted with force or compulsion, there are instances in which force is described as a kind of persuasion or persuasion as a kind of force, e.g. Aesch. Ag. 385; Dio Cass. 62.16 and Plato, , Soph. 265d7–8Google Scholar. Notoriously, Gorgias puts the decrees of Necessity, seizure by force and persuasion all on a par in his Helen (6–8). Finally, note Pindar's reference at Pyth. 4.219 to the μστιξ Πειθος. There is, however, something of a deliberate paradox in these usages and as Buxton, R., Persuasion in Greek Tragedy (Cambridge, 1982), p. 40Google Scholar, aptly observes, they provide ‘the satisfying frisson of oxymoron’. For an interesting discussion of the relations between πειθώ and βα in Greek literature generally, see Buxton.

⁷ E.g. Horn. Il. 1.132, Od. 2.106 and 14.123: in all of these, persuasion involves deception. Cf. Aesch. Cho. 726 where the goddess Persuasion is described as δολα Πειθώ. Cf. Laws 863b8–9 and England, E., The Laws of Plato (Manchester, 1921), ad loc.Google Scholar

⁸ E.g. Hdt. 8.134; Horn. Il. 24.219, Od. 14.363; Lys. 21.10; Pind. Ol. 2.80 and Laws 909b3–5. The phrase δρα θεοὺς πεθει (e.g. Hes. Frag. 361 Merkelbach–West) was proverbial and is disapprovingly referred to by Plato at Laws 906e6 and Rep. 390e2–3.

⁹ Buxton goes a little too far here: convincing someone by means of an argument is for us a paradigm case of ‘persuading’ someone.

¹⁰ Buxton, op. cit. (n. 6), pp. 48–51. Cf. Chantraine, P., Dictionnaire Étymologique de la Langue Grecque (Paris, 1974), p. 868Google Scholar, who claims that the sense of πεθειν is: ‘“persuader” de toutes les façons, par le raisonnement, les prières, la force, l'argent’. For the noun, πειθώ, LSJ⁹ lists the following meanings: (1) Persuasion as a goddess (note that the goddess Persuasion is contrasted with the goddess Ἀναγκαη at Hdt. 8.111 and with the goddess Bα at Plut. Them. 21), (2) as appellative: persuasiveness, (3) persuasion in the mind, (4) means of persuasion, and (5) obedience. Buxton, op. cit. (n. 6), p. 49, is also helpful here: ‘πειθώ is a member of the class of Greek nouns which end in -ώ, which were … in a high proportion of cases originally proper names designating a quality thought to be typical of the character concerned … Πειθώ is the name given to she who πεθει. Later … the form in -ώ is extended to appellatives as well as proper names … For, although early on [πειθώ] was – it seems – exclusively a proper noun, it later became applied not only to the divinity believed to embody a given quality of behaviour but also to that behaviour itself.’

¹¹ Popper, K., The Open Society and its Enemies⁶ (Princeton, 1971), i.271 n. 10.Google Scholar

¹² Ibid.

¹³ Popper, op. cit. (n. 11), i. 139 and 270 n. 5. In discussing the same passage from the Laws, Versenyi, L., ‘The Cretan Plato’, RMeta 15 (1961), 67–80, pp. 69–70Google Scholar, comments: ‘Unfortunately, even though this sounds as if for the first time enlightenment and rational instruction reared their heads in Plato's city, the impression is deceptive. The preambles simply add persuasive prescription to despotic prescription (722e), compulsion tempered with persuasion to untempered force (722c), and not rational instruction to the force of either mere persuasion or brute violence. Plato's comparison of the doctors' method with those of the legislators is apt. In answer to the children's begging to be treated gently (720a), the kindly doctor sweetens the pill, and talks with his adult patients until he gains their willing consent and continued docility by means of persuasion (720d). Likewise the legislator aims to ensure the quiet, well-disposed – and since well-disposed, docile – acceptance of his laws (723a); and this is what the preambles aim at. In view of this aim is hardly surprising that they consist almost entirely of non-rational persuasive material rather than reasoning’. Versenyi does not make it clear whether he thinks, as Popper does, that this sort of sugar-coating involves extensive use of lies, but he is firm in the claim that such persuasion ‘boils down to nothing but simple habituation, indoctrination, and non-rational persuasion of the citizens’ (p. 68). With regard to Versenyi's comment concerning ‘docility’, cf. n. 23.

¹⁴ Morrow, G., ‘Plato's Conception of Persuasion’, PhR 62 (1953), 234–50, p. 244.Google Scholar

¹⁵ E.g. Morrow, op. cit. (n. 14), p. 243, laments the alleged fact that Plato's methods ‘involve so much attention to the sentiments and make so little use of rational proofs’.

¹⁶ Stalley, R., An Introduction to Plato's Laws (Indianapolis, 1983), p. 43Google Scholar. The remark that ‘it is their literary qualities in which the Athenian Stranger feels so proud’ is mistaken in two respects. First, at 811c–e – the passage to which Stalley refers – the Athenian Stranger is discussing ‘the speeches we have been going through from dawn until the present’ (811c6–8), i.e. all of the first six and a half books of the Laws and not merely the sample preludes just offered. It is simply not plausible to think that the only quality of the entire first half of the Laws of which Plato is proud is its ‘literary quality’. Second, the passage at 811c–e itself has no mention of ‘literary qualities’. It comes during the discussion of whether the young should learn poetry. The Athenian Stranger here complains that the poets ‘have said many things in a noble fashion, but also many things in the opposite fashion’ (811b2–3). This is Plato's old complaint that the ethical views expressed in poetry are frequently false and unedifying. The Athenian Stranger goes on to contrast this with his own speeches in the Laws: ‘compared to most of the speeches that I have learned or heard, in poems, or poured out in prose, these seem to me to be both the most well-measured and especially fitting [μετριώτατοι … κα προσκοντες] for the young to hear’ (811d2–5). Such speeches are ‘fitting for the young’ not because of their elegant prose – not even Plato could think that the text of the Laws possessed, as a whole, many stylistic charms – but because they lead the young to virtue and thus to happiness.

¹⁷ For instances of non-rational πειθώ see, e.g., 798e7 and 933a2–3.

¹⁸ The reference to children is clearly playful and they do not recur in the rest of Passage A or in Passages B and C. Cf. n. 13.

¹⁹ Cf. 857c6–d4 which is quoted in Passage B below. We should not read too much metaphysical import into this distinction: there is no reason to take Plato to be suggesting that the free doctor has some source of knowledge that is completely independent of experience. As England, op. cit. (n. 7), ad loc., rightly notes, here ‘the course of learning and teaching systematically (κατ ϕσιν) [cf. 720e11–721a1 where the good legislator is described as proceeding κατ ϕσιν] in medical schools is opposed to the random picking up by their slave assistants of bits of doctors' skill’. What is important is that the free doctor really does possess knowledge and can thus pass on a reasoned account to the patient. This contrast between μπειρα and the true medical art is reminiscent of Gorg. 463b and 465a; also cf. Laws 938a and Phil. 55e–56b. On the need for theory in medicine, see Hippocrates, De Vet. Med. 20.

²⁰ Cf. Lloyd, G. E. R., Hippocratic Writings (Harmondsworth, 1983), p. 17Google Scholar. Plato, for his own part, would endorse such differences in the treatment of free men and slaves: what a slave deserves, Plato thinks, is a command and not an exhortation (777e–778a). But cf. Epist. VII 331b6–7.

²¹ See 722d–e and 723c–d for Plato's comparison of these προομια to the προομια attached to songs. Plato there puns on νμος as ‘law’ and as ‘song’.

²² 811c–812a. Indeed, the entire education of the citizens constitutes a kind of persuasion.

²³ Given the contemporary pejorative connotation of ‘docile’, it would be a mistake to translate εὐμαθστερον as ‘more docile’. As an inspection of the references in Brandwood, L., A Word Index to Plato (Leeds, 1976)Google Scholar shows, εὐμαθς in Plato never merely means ‘tractable’ or ‘manageable’. It always means ‘good at learning’ or ‘quick at learning’ (cf. the Platonic Defs. 413d8: εὐμθεια εὐϕυΐα ψυχς πρς τχος μαθσεως) and is a trait that Plato thinks is especially distinctive of philosophers (e.g. Rep. 486c and Epist. VII 340d4). The same is true of the two cognate forms (εὐμθεια and εὐμαθα) found in Plato.

²⁴ The present passage (719e–723a) contrasts sharply with parts of the Politicus. Consider, for example, the following passage in which Plato takes as a target ‘the many's’ view that ‘if anyone has anything better than the old law to offer, he must first persuade the state, and then he may make his laws, but not otherwise’ (Pol. 296a7–9). Plato tries to undermine this claim by appealing to an analogy between the lawgiver and a doctor: ‘Suppose a doctor who has a correct grasp of his art [ἔχων δ ρθς τν τχνην] does not persuade, but compels [μ πεθων … ναγκζῃ] his patient – whether it is a child, a man, or a woman – to do what is better, although it is contrary to the written rules. What name shall we give to this use of force? Shall we not call it anything rather than “an unscientific [παρ τν τχνην] and harmful [νοσδες] error”? And a person who was so forced would correctly say anything except that he had been treated harmfully and unscientifically’ (Pol. 296b5–c2; cf. 293a6–c3 and 296d6–297b3). I cannot here provide a full interpretation of the Politicus and it is true that Plato in the Laws still thinks that treating a person in a way that he does not want to be treated can be justifiable if doing so really is in his best interests. Nevertheless in the Politicus passages there is no suggestion that the status of the patient is relevant to his treatment and no hint of the desirability of persuasion. Cf. Pol. 304c10–d2 which insofar as it recommends any kind of persuasion at all, recommends persuading most people simply through ‘edifying stories, but not through teaching’ (δι μνθολογας λλ μ δι διδαχς).

²⁵ I.e. one of the slave doctors. Note, contra Popper, that Plato clearly rejects the slave doctor's point of view.

²⁶ It is worth emphasizing that the analogy here suggests that at least many citizens will come close to possessing the sort of knowledge possessed by the scientific lawgiver.

²⁷ Note that ἥμερος applies here to lawgivers, not the citizens and is connected with a willingness to engage in rational argument. This gives us additional reason to reject Versenyi's interpretation. Cf. 731b–d.

²⁸ We may be concerned that Plato uses νουθετεῖν here, since in the Sophist, Plato characterizes νουθετητικ as ‘the venerable method of our fathers … of sometimes showing anger at their [sons'] errors and sometimes more softly exhorting them’ (Soph. 229e4–230a3). There Plato distinguishes νουθετητικ from the preferred process of elenchus which starts from the assumption that all wrongdoing is involuntary (κοσιον Soph. 230a6) and tries to remove the wrongdoer's false belief and conceit of wisdom by showing him that his opinions are self-contradictory (Soph. 230b ff.). The interpretation of the Sophist passage is controversial, but if the lawgiver of the Laws is engaged in the practice condemned in the Sophist, it might seem to undermine the claim that he is engaged in rational persuasion. But Plato's terminology varies so greatly from dialogue to dialogue that we cannot attribute much significance to the recurrence of νονθετεῖν: (a) νονθετεῖν is paired with teaching at Rep. 399b5 and is recommended for those whose wrongdoing is involuntary (οὐχ κὼν ξαμαρτνειν) at Gorg. 488a2–b1, (b) in the Laws, Plato insists on the claim that all wrongdoing is involuntary (731c–d, 860d) while still recommending νουθετητικ as a way of correcting wrongdoers (Laws 908d7–909a5), and (c) the process recommended for the young atheist in the Laws is in fact the same as the one preferred in the Sophist: giving him an argument and showing him that his beliefs are self-contradictory.

²⁹ Plato states the following terms to the young atheist ‘either to teach [διδσκειν] us that we are not speaking correctly … or, if he is not able to speak better than us, to be persuaded by us and live believing in the gods’ (899c6–d 1). Cf. 891c–d and 905c7–d1 where Plato claims that what the disbeliever needs is an ‘argument’ and appeals to his intelligence: ‘if you should still be in need of some argument [λγου τινς], hearken to us as we speak … if you have any intelligence at all [εἰ νον κα πωσον ἔχεις]’. A few lines later, Plato pronounces himself satisfied with the arguments given: ‘That there are gods and that they exercise supervision over human beings, I would say has been demonstrated by us in no mean fashion’ (905d1–3, cf. 899d1–2).

³⁰ On the question of whether the desire for pleasure is the ultimate cause of the young atheists' disbelief, cf. n. 48.

³¹ Given Magnesia's censorship policies, we might think Plato's emphasis on inquiry is disingenuous. I discuss this point below.

³² Strictly speaking, it is the preludes and not the laws which do the persuading (e.g. 722e7–723b2), but it is convenient to speak of the laws as persuading.

³³ 857d2. Cf. 720c3–5 where one of the complaints made against the slave doctor is that he fails to give an account (λγον διδναι) of the disease to his patient.

³⁴ Plato here is not greatly worried about whether such rational persuasion can engender a false belief. He is confident that the beliefs the laws recommend are true (e.g. 662b) and that the arguments he provides for them are satisfactory (e.g. 905d) although, as is the case throughout all the dialogues, he is willing to admit the possibility of mistake (e.g. 663d).

³⁵ Morrow, op. cit. (n. 14), 238ff. πῳδ literally means ‘song to or over’ and thus comes to mean spell, incantation, enchantment or charm. Cf. LSJ⁹ s.v. πῳδ.

³⁶ E.g. in the Theaetetus, Socrates, after giving an extremely sophisticated statement of Protagoras' ‘secret doctrine’ about perception and change in the sensible world, describes what he is doing as ‘uttering incantations (πᾴδω)’ (Tht. 157c7–d2, cf. 149d1). For other playful uses of πῳδ and its cognates, see Charm. 155e–158b, 175e–176b; Phdo. 77e8 and Rep. 608a4. For a similar playfulness in the Laws, see 900b5. Also, cf. Belfiore, E., ‘Elenchus, Epode, and Magic: Socrates as Silenus’, Phoenix 34 (1980), 128–37CrossRefGoogle Scholar and de Romilly, J., Magic and Rhetoric in Ancient Greece (Cambridge, Mass., 1975), chapter 2.CrossRefGoogle Scholar

³⁷ Songs for children:659e1, 664b4, 671a1 and 812c6. Plato also occasionally applies the term to songs for adults: 665c4, 666c6 and perhaps 944b3. Also cf. 773d6.

³⁸ Nor does it show that Plato thought that all of the citizens' education was a kind of enchantment. We might, however, worry whether children so educated would later be capable of rational reflection and I shall take up this point below.

³⁹ βιζεσθαι … αὐτν 903a10. The use of βιζεσθαι also gives a somewhat playful tone to the passage.

⁴⁰ The preludes often, for instance, appeal to citizens' sense of shame. Also see n. 61.

⁴¹ Cf. 966b1–3.

⁴⁸ E.g. 720d–e in Passage A and 890b5–d5 in Passage C. Also in Passage A, cf. 723a4–7 with 718c8–d7 quoted in the text and in Passage C, note 885c8–e5 and 887e7–888a7.

⁴³ Cf. nn. 23 and 27.

⁴⁴ For the idea that pleasure and the desire for pleasure can destroy or obscure an agent's judgment of what is best all things considered, see 649d. For the different idea that pleasure can cause a change in, and not just a destruction of, the agent's judgment, see 863b6–9.

⁴⁵ Plato often suggests a connection between ‘gentleness’ and the ability to learn and teach. For ἥμερος and μερον, see, e.g., Rep. 410c–e, 554d2, 571c4, 589d2, 591b3; Soph. 230b9,246c9 and Laws 718c8–d7, 765e–766a and, especially, 935a5. For ἠρμα, see, e.g., Phdo. 83a3 and Rep. 476e1, 494d4 and 533d2. For πρᾷος and πρᾳτης, see, e.g., Meno 75d4; Rep. 376c1; Tim. 18a6; Laws 731b–d, 634c8 and perhaps 645a6. But also cf. Tht. 144a–b, although the speaker here is Theodoras and he may not represent Plato's view.

⁴⁶ It might be argued that our differences over how best to foster rational reflection and action show that we have different conceptions of rationality. I cannot try to resolve this issue here, but it should be stressed that both we and Plato would agree on the essential point that rational persuasion involves providing good epistemic reasons for adopting a belief.

⁴⁷ This doctrine in one form or another was an old target for Plato. E.g. both Callicles in the Gorgias and Thrasymachus in the Republic hold variants of this view.

⁴⁸ Although Plato thought that this was the most common explanation of young atheists' beliefs, he did not think that we could explain the atheism of all atheists by reference to their desire for pleasure, e.g. a person with a ‘naturally just disposition’ can still suffer the misfortune of believing that there are no gods (908b4ff., cf. 899dff.). More generally, Plato in the Laws recognizes three causes of wrongdoing (μρτημα, 863a ff.): pleasure and the desire for pleasure, spirited anger and ignorance. Pleasure and spiritedness can cause wrongdoing both by leading the agent to act akratically and by affecting his judgment of what is best.

⁴⁹ Cf. 887a3–8.

⁵⁰ E.g. women magistrates are to enter the homes of young married couples and ‘by admonition and threats stop them from doing anything wrong or foolish’ (784c2–4)! More generally, the lawgiver is to give advice about ‘the many little things that occur in private and in the home’ (788a5–6) and public pressure, by means of praise and blame, is to be brought on the citizens to follow these guidelines.

⁵¹ E.g. in the Republic: (a) the ‘noble lie’ is intended to create false beliefs among the citizens about the basis and workings of their society (Rep. 414b–415d), (b) a great deal of lying and deception will be involved in carrying out various social policies and especially the city's eugenic policy (Rep. 389b–c and 459c ff.), and (c) there is no attempt to provide to those outside of the guardian class a rational understanding of the rules governing their behaviour and surprisingly little effort to provide such an understanding to the non-philosophic auxiliaries. There simply is nothing in the Republic to correspond to the Laws' requirements that (i) the laws or lawgiver provide a rational justification of the laws to those to whom the laws apply, and (ii) the citizens be aware of the fundamental principles that determine the structure of their legal and social system and the reasons for these principles. Cf. Reeve, C., Philosopher–Kings (Princeton, 1988), pp. 208–13.Google Scholar

⁶² Rawls' ‘publicity condition’ requires that citizens know about basic political principles all that they would know if the acceptance of these principles were the result of an agreement. See Rawls, J., A Theory of Justice (Cambridge, Mass., 1971), p. 133Google Scholar and the index entry. Plato's justification for publicity differs, however, from Rawls'. For Rawls, the publicity condition is one of the ‘formal constraints on the concept of right’ (pp. 130–6). For Plato, publicity is justified because it produces the greatest possible happiness for the citizens.

⁵³ I defend the attribution of this justificatory principle to Plato in more detail in a work in progress on the Laws. Relevant passages in the Laws include: 631b, 715b, 718b, 743c, 806c, 863e–864a and 903c. Although ‘happiness’ is now the standard translation of εὐδαιμονα it may have misleading hedonistic connotations. For Plato, εὐδαιμονα is a complex good which has virtue as a necessary component (cf. n. 57). Although pleasure, along with other goods such as knowledge, is a component of happiness, pleasure is not the only or the most important component. On the question of translation, see Kraut, R., ‘Two Conceptions of Happiness’, PhR 88 (1979), 167–97Google Scholar and Vlastos, G., ‘Happiness and Virtue in Socrates' Moral Theory’, Topoi 4 (1985), 3–22, pp. 3–4.CrossRefGoogle Scholar

⁵⁴ On the question of whether Plato would immediately rule out the idea that persuasion could produce knowledge, see Appendix.

⁵⁵ For two recent discussions, see Burnyeat, M., The Theaetetus of Plato (Indianapolis, 1990)Google Scholar, and the radical and, I think, implausible account of Moline, J., Plato's Theory of Understanding (Madison, 1981).Google Scholar

⁵⁶ In the context of the Laws, we might wonder whether the studies recommended for the Nocturnal Council are necessary for ethical knowledge. In the brief and obscure passage that discusses the Nocturnal Council (960b ff.), Plato stresses the importance of knowing how virtue, beauty and the good are each both ‘many and one’. Although the interpretation of this passage is difficult, all that Plato claims, I think, is that this knowledge is necessary for being a good ruler. He does not claim that such knowledge is a necessary precondition for the possession of all other ethical knowledge or that such knowledge is necessary for virtue. Cf. Morrow, G., ‘The Nocturnal Council in Plato's Laws’, AGP 42 (1960), 229–46Google Scholar and Morrow, , Plato's Cretan City (Princeton, 1960).Google Scholar

⁵⁷ In the Laws, Plato holds that virtue is always unconditionally good for its possessor (that is, roughly, that virtue is good for its possessor regardless of what other things he has or lacks). Although Plato recognizes goods other than virtue, he claims that the possession of virtue is at least a necessary condition of other goods, such as health and wealth, being good for their possessor (631b–d). On the question of whether knowledge is necessary for virtue, see n. 59.

⁵⁸ Formally, the possibility is still left open that although rational persuasion fails to produce knowledge, it will be of benefit to an agent who already possesses knowledge and already is virtuous. I consider this possibility in section III below.

⁵⁹ In the Republic, knowledge is required for the possession of any virtue (Rep. 441d–443e), although there Plato also appears to distinguish a lower grade of virtue (πολιτικ ρετ) that requires only true belief (Rep. 430a–c). In contrast, I do not think that Plato in the Laws makes the possession of knowledge a necessary condition for the possession of each and every virtue or that he distinguishes a ‘philosophic’ sort of virtue which requires knowledge and is had only by an elite from a ‘political’ or ‘ordinary’ sort of virtue which does not require knowledge and is the best that a non-philosopher could aspire to. In the Laws, Plato is willing to accept not only knowledge but also some form of true opinion as a ‘leader’ of the virtues (e.g. 688b, 689a–e, 770c–d and 864a–b) and he freely attributes unqualified virtue to those who lack philosophic knowledge (e.g. 641b–c, 770c–771a, 807c–e, 822e–823a and 853b–c). Laws 710a5–8 which refers to ‘popular virtue’ (δημώδης ρετ) is often cited to show that Plato still accepts the Republic's distinction between real or philosophic virtue and some inferior sort of virtue which is what most citizens possess. But δημώδης ρετ at 710a is characterized by Plato as that found in beasts and children: it is a self-restraint with regard to pleasures that might not be directed to the right ends and is directed, at best, by only the crudest forms of true belief. Plato in the Laws never claims that ordinary citizens have only δημώδης ρετ and he expects that their virtue will be directed to the right ends and that they will have some sort of rational justification for their beliefs. Further, δημώδης ρετ is contrasted by Plato himself (689a–e) with having one's moderation directed by the proper sort of true belief, not with a sort of virtue that requires knowledge. Although I cannot offer here a full defence of my claim that knowledge is not required for virtue in the Laws, my arguments above do not require its acceptance. Even if one thinks that most of Magnesia's citizens could only have ‘ordinary virtue’, the Laws breaks sharply with the Republic by attempting to ensure that all citizens have rational justification for their ethical beliefs.

⁶⁰ Since, as we have seen, the preludes are not the only source of rational persuasion, a comprehensive study would also have to examine the entire education of the citizens. And since their education includes reading of the Laws itself, such a study would also have to evaluate the satisfactoriness of all the arguments presented in the Laws.

⁶¹ See, especially, the general prelude to the legal code at 726a–734e. Cf. Silverthorne, M., ‘Laws, Preambles and the Legislator in Plato’, The Humanities Association Review 26 (1975), 10–20, pp. 16–18Google Scholar. Also see Thesleff, H., ‘Studies in the Styles of Plato’, Acta Philosophica Fennica 20 (1967), 1–192Google Scholar. Thesleff distinguishes ten different styles in Plato (pp. 63–80), including the ‘rhetorical’ style which he finds (p. 153) often displayed in the preludes: e.g. 715e–718a, 726a–734e, 823d–824a, 854b–c and 899d–900b.

⁶² E.g. Plato includes in the preludes to the laws concerning murder a doctrine (870d–e, 872e–873a) which he attributes to ‘those involved in the mystery rites’ (870d5–7) and the ‘ancient priests’ (872e1–2) to the effect that a murderer will suffer in a future life exactly the same sort of crime that he has committed (a parricide will murdered by his sons and so on). Also see 854b–c, 873e–874a and 913c1–3.

⁶³ For the possibility that the good of rational persuasion is, in some cases, outweighed by other goods, see section III.

⁶⁴ Plato thinks (853b4–854a3) that few of Magnesia's citizens will need the threat of a penalty and even that attaching penalties to laws is ‘in a certain way shameful’. For an interesting example of such limitations on rational persuasion and of how the legal code is to take them into account, see 933a–e.

⁶⁵ 662b, 663d6–e2 and 664b–c.

⁶⁶ Roughly put, the problem is that the above justification assumes that it is possible to lead a just life even though one is deceived about what the just life is like and what one's reasons are for choosing it and this assumption seems to be inconsistent with Plato's claim in the Laws and elsewhere that virtue requires at least true belief (cf. n. 59). Although this line of argument could be more fully elaborated, it is sufficient to show that Plato will face grave difficulties in justifying general deception of the citizens on ethical questions.

⁶⁷ It does not entail it since it is possible that rational persuasion just happens to coincide with other goods that are truly fundamental, but we have no good reason for supposing that Plato believes that there is such a coincidence.

⁶⁸ A reference to the proto-prelude at 715e7–718a6.

⁶⁹ Cf. 858d–859a (the lawgiver is supposed to teach the citizens what is noble, good and just and what relation these things have to happiness) and 907c–d. The laws may also recommend other goods to the citizens, but it is their recommendation of virtue that Plato thinks is the most significant aspect of their persuasion.

⁷⁰ Cf. n. 57. See, e.g., the general preface to the legal code at 726a–734e. Again, remember that all the Laws, including its advocacy of the claims that (a) virtue is unconditionally good for its possessor and is both necessary and sufficient for happiness (e.g. 660e–661e and 874d), and (b) virtue is a necessary precondition of the goodness of any other good, will be studied by the citizens.

⁷¹ See, e.g., 689a5–e2.

⁷² In the Republic (e.g. 590a–591a), Plato is far more sceptical about the ability of non-philosophers to attain such insight and thus to regulate their own lives. Also, cf. n. 86.

⁷³ It is not sufficient, since for at least some virtues, e.g. moderation, one will also need suitably trained emotions. And Plato may even come close to holding that a unvirtuous person is worse off if he has some grounds for his unvirtuous beliefs (863c–d). I shall leave aside here the question of whether Plato would ultimately be willing to say that the unvirtuous person could have rational grounds for his beliefs.

⁷⁴ Cf. n. 59.

⁷⁵ Ibid.

⁷⁶ Plato's emphasis – e.g. in Passage B – on the servile and degraded status of those who lack rational justification also strongly implies that it is necessary for virtue.

⁷⁷ There are two other options here that can be ruled out. First, one might suggest that being rationally persuaded is good for the agent whether or not he possesses any other good and, in particular, whether or not he is virtuous. But Plato's language in Passages A–C is simply not strong enough to warrant this claim and, in any case, it would be inconsistent with Plato's previous insistence that the possession of virtue is at least a necessary condition of other goods being good for the agent (631b–d). Second, we could hold that knowledge is necessary for ‘high-grade’ virtue, but that an agent lacking knowledge might possess some inferior form of virtue which is of benefit to him. Rational persuasion could be of benefit then because it helps produce this inferior sort of virtue. Such a strategy is unattractive because Plato in the Laws never attributes such an inferior sort of virtue to ordinary citizens (cf. n. 59).

⁷⁸ Since the citizens will read the Laws they will even be aware of the value of certain sorts of advanced learning that Plato requires of the Nocturnal Council. It is an old complaint against utilitarianism that it might require the rulers and the elite of a society to encourage what they know are false ethical beliefs and mistaken forms of ethical reasoning among the rest of the citizens (see, e.g., Smart, J. and Williams, B., Utilitarianism: for and against (Cambridge, 1982), pp. 138ff.)Google Scholar. As we have noted, this would be a justified complaint against the Republic, but not the Laws.

⁷⁹ Nor does he ever provide a clear justification for the Republic's claim that knowledge is necessary for virtue. Cf. Irwin, T., Plato's Moral Theory (Oxford, 1979), pp. 283–5.Google Scholar

⁸⁰ R. Buxton, op. cit. (n. 6), notes a wide ‘range of polarities which may be seen as homologous with πειθώ/βα’ including civilized/uncivilized, inside πλις/outside πλις, mankind/animals and Greeks/barbarians (p. 62), but, surprisingly, does not mention free/slave (nor is there an entry for slaves in the index). Nevertheless, the idea that free men deserve persuasion and slaves deserve compulsion is found elsewhere in Greek literature and social thought. E.g. in Arrian we find that Alexander the Great when in India tried to persuade two Indian wise men named Dandamis and Calanus to accompany him, but that Dandamis refused. ‘So Alexander did not try to force [βισασθαι] him, since he recognized that the man was free. But a certain Calanus, one of the wise men there, was persuaded [ναπεισθναι]’ (Arrian, Anabasis 7.2.). In personal communication, Stephen Halliwell has made the interesting suggestion that we may be able to discover some connections between freedom and rational persuasion in the Athenian democratic tradition.

⁸¹ Note the obvious contrast with the transcendental metaphysics and epistemology of the middle period (e.g. Rep. 500c–d, 518cff., esp. 532c; Phdo. 64a–69e, 78b–84b and Phdr. 249c–e. Cf. Vlastos, G., ‘A Metaphysical Paradox’, in Vlastos, Platonic Studies (Princeton, 1981), pp. 43–57)Google Scholar. It is quite controversial whether Plato continued to hold the transcendental metaphysics and epistemology of the middle period, or abandoned them by the time of the Laws. Although the absence of such an appeal to the special status of the relevant objects of knowledge or to the special value of the psychic states that individuals are in when they cognize such objects does not warrant the conclusion that Plato would reject such an attempt at justification, it seems significant that he does not make such appeals. At the least, the focus of his attention has shifted.

⁸² Cf. n. 24.

⁸³ The doctor–patient analogy breaks down here. The result the law is trying to bring about is that the citizen is virtuous, and rational justification is necessary for virtue; the result the doctor is trying to bring about is health, and it is less plausible to think that the patient must have rationally justified beliefs about his condition and the course of his treatment in order to be healthy. It might help the analogy a bit if we imagine that part of what the patient is taught by the free doctor are methods of preventing future illness, but the slave doctor could also impart at least some of the same information in the form of a command. This breakdown seems to result from Plato's attempt to make a new point by means of an old analogy.

⁸⁴ Once again, we might think that this is still much too weak: a respect for the individual's freedom and for his capacities for rational understanding and action requires that, at least over a certain range of beliefs and actions, we allow the individual to conduct his rational inquiry in the manner that seems best to him and to act in accordance with the rational judgment he reaches.

⁸⁵ And it is not at all clear that Plato has a coherent conception of freedom and a free individual. Just within the Laws, Plato has a great variety of things to say about the nature and value of freedom. In addition to the view I have just sketched, Plato thinks that freedom can be had in excess (e.g. 699e) and praises the Athenians of an earlier generation for being slaves (δολοι) to the laws and to their rulers (698b6 and 701b5–6). We must also remember that the free/slave distinction was a legal distinction in Magnesia and that one's status determined the sorts of treatment one received. E.g. in many cases, although not all, the legal penalties for slaves are harsher. (For lighter penalties for slaves, see 854d–855a; for harsher penalties, see 867c–868c, 914a–c and note the flippant cruelty of 844e–845a.) But it is important to see that many of these differences – including the facts that slaves' schedules are imposed on them without their having a choice and that they receive little education – are connected with the possibility of rational persuasion and rational planning. Even with regard to penalties, Plato emphasizes the idea that slaves will not respond to an argument (e.g. 853e–854a). For an interesting discussion of the ways in which Athens tried to secure for free men the autonomy of their bodies, see Halperin, D., ‘The Democratic Body: Prostitution and Citizenship in Ancient Athens’, in Halperin, One Hundred Years of Homosexuality (New York, 1990), pp. 15–40.Google Scholar

⁸⁶ Although a full analysis of Plato's view of freedom in the Republic is far beyond the scope of this paper, Plato there usually construes ‘freedom’ (λενθερα and its cognates) either as merely a legal or political category or in terms of the ability to choose, either rationally or irrationally, among a range of options and to act on that choice. (Cf. Vlastos, G., ‘The Theory of Social Justice in the polis in Plato's Republic’, Mnemosyne Supplement 50 (1977), 1–40, pp. 25–32, esp. p. 30 n. 103.)Google Scholar Thus Plato typically emphasizes the danger of freedom and stresses the need to limit it (e.g. 562c–563c). In a few scattered passages, we find traces of a more favourable view of freedom, although Plato does not work the idea out (e.g. 405a, 577d and 590c–e, but note 590e3 and 591a2). The connection between freedom and reason goes on to have a long career in the history of ethical and political philosophy. One form of it, for example, plays a central role in Kant's moral philosophy.

⁸⁷ Cf. Murray, J., ‘Plato on Knowledge, Persuasion and the Art of Rhetoric: Gorgias 452e–455a’, Ancient Philosophy 8 (1988), 1–10.CrossRefGoogle Scholar

⁸⁸ Pol. 304c10–d2 recommends persuading most people δι μνθολογας λλ μ δι διδαχς, but the passage seems to allow that the latter would count as persuasion. Cf. Rep. 399b4–7.

⁸⁹ E.g. 720d, 857d–e, 885d–e and 888a.

⁹⁰ See Burnyeat, M., ‘Socrates and the Jury: Paradoxes in Plato's Distinction between Knowledge and True Belief’, PAS Supplement 54 (1980), 172–91, especially pp. 179ff.Google Scholar, for the provocative assertion that Plato in some cases rejects the claim that: If x teaches y that p, then x brings it about that y knows that p. For an opposed, and more plausible view, cf. Barnes, J., ‘Socrates and the Jury: Paradoxes in Plato's Distinction between Knowledge and True Belief’, PAS Supplement 54 (1980), 193–206Google Scholar. Also cf. Lewis, F., ‘Knowledge and the Eyewitness: Plato Theaetetus 201a–c’ CJPh 11 (1981), 185–97Google Scholar. Both Burnyeat (pp. 179 and 190) and Barnes (pp. 196 and 206) cite a number of passages in which Plato appears to accept this claim and all that Burnyeat brings against it is a highly controversial interpretation of the brief passage at Tht. 201a4–c7. And even if Burnyeat were right about the Theaetetus, this would only show that Plato might describe a certain process as an instance of both persuasion and teaching while denying that it produced knowledge; it would not show that Plato is committed to saying that every sort of teaching which could be truly described as persuasion fails to produce knowledge.

I would like to thank Dan Brudney, Alan Code, Stephen Halliwell, Tony Long, Sam Scheffler and Bernard Williams for their comments on earlier versions of this paper.