1. I have insisted on this pluralism in a paper entitled ‘Methodological issues in the comparison between East and West’ to be published by Pedilavium. Not only are there important differences between each main domain of inquiry, for example, mathematics, astronomy, medicine, and between each period, in both China and Greece, but one must also recognise the diversity of what passes as ‘mathematics’ ‘astronomy’ and so on, in each society.

2. Such a view influences the analyses offered by Nakamura, H., The ways of thinking of Eastern peoples (1960)Google Scholar. The idea of a radical contrast, on this score, between the West and China goes back at least as far as Voltaire, whose Au Roi de la Chine dates from 1771.

3. Cf. Lloyd. ‘Methodological issues’ (n. 1).

4. Those conventionally categorised as ‘philosophers’ before Plato are, in any event, unlikely to have considered themselves as all engaged in the same one inquiry. (‘wise man’) and (‘Sophist’) were, and remained, terms with fluctuating and indeterminate reference: cf. Lloyd, G. E. R., The revolutions of wisdom (1987) 83–102Google Scholar.

5. Heraclitus,frr. 40, 42, 56, 57, 106.

6. Two such points relate to the choice of the Boundless, rather than any determinate substance such as Thales' Water, as principle (see Aristotle, , Physics 204b24–9Google Scholar) and to the answer to the question of why the earth is at rest (Aristotle, , On the heavens 294a28–b1, 295b10–16)Google Scholar.

7. Where Anaximander fr. 1 had described the relations between certain cosmic factors as governed by order (they pay the penalty, , to one another for their injustice, ), Heraclitus fr. 80 insisted that ‘justice’. , is ‘strife’, .

8. With Empedocles fr. 17.26, for instance, compare Parmenides fr. 8.50ff. But Empedocles agrees with Parmenides, as does Anaxagoras, that nothing comes to be from not-being. With Parmenides fr. 8.6ff., compare Empedocles fr. 8 and Anaxagoras fr. 17.

9. As Jaeger, W., Aristotle: fundamentals of the history of his development (1948)Google Scholar made much of, Aristotle sometimes includes himself among those holding the Platonist views he criticises, as for instance he does in the first book of the Metaphysics 990b8–9, 22–4.

10. Thus On regimen in acute diseases sets out to describe matters of great importance that are not understood by doctors (ch. 3, II 238.8–244.6 Littré), and On regimen refers repeatedly to the discoveries the author has made, which ‘none of my predecessors attempted to understand’ (III ch. 69, CMG I 2.4, 200.28–30). For further examples, see Lloyd, , Revolutions (n. 4) ch. 2, 57–70Google Scholar.

11. The evidence is set out and discussed in Lloyd, G. E. R., Methods and problems in Greek science (1991) ch. 17, 400–2Google Scholar.

12. See Cuomo, S., ‘Tradition and innovation in Pappus of Alexandria’, Ph.D. dissertation, Cambridge, 1994Google Scholar.

13. This is discussed in Lloyd, , Revolutions (n. 4) ch. 2, 101–8Google Scholar. As Manuli, P., ‘Lo stile del commento’, in Formes de pensée dans la collection hippocratique, edd. Lasserre, F. and Mudry, P. (1983) 471–80Google Scholar, stressed, it is important not to underestimate the extent to which, in Greece, the commentary may be used as a vehicle for the expression of original ideas. In China, a similar point holds also, for example, of Liu Hui's commentaries on the mathematical classic, the Nine chapters of the mathematical art.

14. Lloyd, Methods (n. 11) ch. 17, 416.

15. On the various strata, their composition and dates, in the Zhuangzi compilation, see Graham, A. C., Chuang-tzu: the Seven Inner Chapters (1981)Google Scholar and Disputers of the Tao (1989) 172–4, 306–11Google ScholarPubMed.

16. Zhuangzi 33, at lines 16–17, 34–5, 42–3, 55 and 63–4 in the Harvard–Yenching edition (1949).

17. The differences between Ming Jia and Greek Sophists are well brought out by Reding, J.-P., Les Fondements philosophiques de la rhétorique chez les sophistes grecs et chez les sophistes chinois (1985)Google Scholar, though he continues to write of Chinese sophists.

18. Qian, Sima. Shi Ji 130Google ScholarPubMed: 3289.

19. Qian, Sima. Shi Ji 130Google ScholarPubMed: 3289.

20. Hanfeizi 50: 1080, trans. Graham, , Disputers (n. 15) 273Google Scholar.

21. Lü Shi Chunqiu 15/8, trans. Graham, , Disputers (n. 15) 214–15Google Scholar, modified.

22. As in ‘Robber Zhi’, Zhuangzi 29Google Scholar.

23. Zhuangzi 33, at lines 27–31.

24. However, as will be noted below, p. 34, the term for ‘Confucian’ is often used loosely.

25. Hanfeizi 50: 1080.

26. As Hui Shi is in Zhuangzi 33, at lines 69–87, cf. Graham, , Disputers (n. 15) 75–82 and 82–95Google Scholar.

27. Mencius 6A/6, and the chapter in Xunzi entitled ‘Our nature is bad’, ch. 23, discussed by Graham, , Disputers (n. 15) at 117 and 244–51Google Scholar.

28. See C. Cullen, Astronomy and mathematics in ancient China: the Zhou Bi Suan Jing (forthcoming).

29. See Sivin, N., ‘Cosmos and computation in early Chinese mathematical astronomy’, T'oung Pao 55 (1969) 1–73CrossRefGoogle Scholar.

30. See Wagner, D. B., ‘Liu Hui and Tsu Keng-chih on the volume of a sphere’, Chinese Science 3 (1978) 59–79Google Scholar, and ‘Doubts concerning the attribution of Liu Hui's commentary on the Chiu-Chang Suan-Shu’, Acta Orientalia 39 (1978) 199–212Google Scholar.

31. Qian, Sima, Shi Ji 105Google ScholarPubMed: 2810–11. On the relations between book learning and practical experience in Chunyu Yi's account of his own training, see Sivin, N., ‘Text and experience in classical Chinese medicine’, in Knowledge and the scholarly medical traditions, ed. Bates, D.Google Scholar (forthcoming). The Canggongzhuan is also the subject of a forthcoming study by E. Hsu.

32. I have discussed Galen's opportunism in citing previous authorities in favour of his own theories in ‘Scholarship, authority and argument in Galen's Quod animi mores’, in Le Opere psicologiche di Galeno, edd. Manuli, P. and Vegetti, M. (1988) 11–42Google Scholar, and his eagerness to rebut the charge of in Methods (n. 11) ch. 17, 400 and n.8.

33. Cf. also the contrast drawn between Platonic philosophy and conventional astronomy by Proclus in the Proem to Hyp. 4.15–6.5.

34. There is a similar reference to ‘100 Jia’ in Sima Qian 130: 3320.

35. Sivin, N., ‘On the word “Taoist” as a source of perplexity. With special reference to the relations of science and religion in traditional China’, History of Religions 17 (1978) 303–30CrossRefGoogle Scholar.

36. See especially Nylan, M. and Sivin, N., ‘The first Neo-Confucianism: an introduction to Yang Hsiung's “Canon of Supreme Mystery” (T'ai hsuan ching, c. 4 B.C.)’, in Chinese ideas about nature and society, edd. LeBlanc, C. and Bladen, S. (1987) 41–99Google Scholar.

37. See Graham, , Disputers (n. 15) 379–81Google Scholar.

38. See especially Keiji, Yamada, ‘The Formation of the Huang-ti nei-ching’, Acta Asiatica 36 (1979) 67–89Google Scholar, and Keegan, D., ‘The Huang-ti nei-ching’, Ph.D. dissertation, University of California, Berkeley (1988)Google Scholar.

39. A number of instances are discussed in N. Sivin, ‘Cosmos’ (n. 29), and Yi-Long, Huang, ‘Court divination and Christianity in the K'ang Hsi era’, Chinese Science 10 (1991) 1–20Google Scholar discusses the continuation of this right down to the debates with the Jesuits in the seventeenth century.

40. Sivin, ‘Text’ (n.31).

41. N. Sivin, personal communication.

42. Sivin, ‘Text’ (n. 31) at n. 27, on Ling Shu 48.1. Cf. also Keegan, , Huang-ti (n. 38) 233–8Google Scholar.

43. However, Xunzi criticises many ru as worthless. See Graham, , Disputers (n. 15) 236Google Scholar, who further points out that Xunzi's position is in part a reaction to the challenge of those later classed as ‘Legalists’.

44. Qian, Sima. Shi Ji 105Google ScholarPubMed: 2794–6, 2815–16.

45. I am grateful to Professor David McMullen for clarification on this point.

46. See, for example, Mencius 3B/9, Zhuangzi 33 at lines 30, 74, 79, 86. Graham, , Disputers (n. 15) 167Google Scholar. has, however, argued for a positive role for ‘argumentation’, , bian, among the Mohists, but this in the sense related to the cognate, , ‘discriminate’, applied to distinguishing right from wrong. In the Mohist Canon (A 74) ‘argumentation’ is defined as ‘disputing over the converse. To win in argumentation is to fit the fact.’

47. Aristotle, , Topics 8Google Scholar, systematises these rules of procedure: see, for example, Moraux, P., ‘La joute dialectique d'après le huitième livre des Topiques’, in Aristotle on dialectic, ed. Owen, G. E. L. (1968) 277–311Google Scholar.

48. On Coans and Cnidians, see Smith, W. D., ‘Galen on Coans versus Cnidians’, Bulletin of the History of Medicine 47 (1973) 569–85Google ScholarPubMed, and Lonie, I. M., ‘Cos versus Cnidus and the historians’, History of Science 16 (1978) 42–75, 77–92CrossRefGoogle ScholarPubMed. On the Hellenstic medical ‘sects’, see von Staden, H., ‘Hairesis and heresy: the case of the haireseis iatrikai’, in Jewish and Christian self-definition, vol. III, edd. Meyer, B. F. and Sanders, E. P. (1982) 76–100, 199–206Google Scholar, and Herophilus: The art of medicine in Ptolemaic Alexandria (1989) 22–6Google Scholar, and Frede, M., Essays in ancient philosophy (1987)Google Scholar chh. 12–15, cf. Lloyd, , Revolutions (n. 4) 158–67Google Scholar.

49. For Theophrastus, see Diogenes Laertius 5.36, for Euclid of Megara, D.L. 2.106, for Zeno of Citium, D.L. 7.2 (where Timocrates is reported to have said that Zeno was also taught by Xenocrates for ten years: since Zeno was only about twenty when Xenocrates died, this may be thought unlikely), and for Chrysippus, D.L. 7.179, 183–1.

50. Thus in the Academy, in the first century B.C., there was a bitter dispute between Philo and one of his ex-pupils, Antiochus, in which Philo stood for the view of Plato that had been dominant since Arcesilaus, but Antiochus sought to revive the ‘Old Academy’. The reaction of another who had been taught in the Academy, namely Aenesidemus, was to move away to inaugurate his own brand of neo-Pyrrhonian scepticism. The chief primary sources are Cicero, 's Academica, I.13, 43–6Google Scholar, Photius, , Library 169b18–170b3Google Scholar, discussed by Long, A. A. and Sedley, D. N., The Hellenistic philosophers, vol. I (1987) 438–49, 468–73CrossRefGoogle Scholar.

51. Thus the founder of the so-called Empiricist sect of medicine, in the mid-third century B.C., was Philinus, who had started out as a Herophilean: on the various affiliations of other Herophileans, see von Staden, , Herophilus (n. 48) 445–71Google Scholar.

52. On the subject of philosophical allegiance, see Sedley, D. N., ‘Philosophical allegiance in the Greco-Roman world’, in Philosophia Togata, edd. Griffin, M. and Barnes, J. (1989) 97–119Google Scholar.

53. One extra factor that no doubt influenced the headship passing first to Speusippus was that he was Plato's nephew. Xenocrates, however, was no relation of either Plato or Speusippus, nor was it the general rule that the headship of philosophical or medical sects stayed within a family.

54. With D. L. 7.179 on Chrysippus (), compare D.L. 7.167 on Ariston, Herillus and Dionysius (). Dionysius, however, was also called (‘the renegade’) and Ariston is described as a with his own followers. Long, and Sedley, , Hellenistic philosophers (n. 50) 359Google Scholar. go so far as to speak of Ariston's ‘heresy’, though that English term now has stronger connotations than those of the original Greek .

55. On the Academy, see above n. 50. As for the Stoa, it is customary to distinguish the so-called Middle Stoics, Panaetius and Posidonius, and Later Stoics, such as Seneca, from the three original heads, Zeno, Cleanthes, Chrysippus and their contemporaries. A note of caution is, however, in order, since for the Middle Stoics, especially, we generally rely on secondary sources, who can certainly not be counted on to be reliable on questions to do with agreements and disagreements between later and earlier members of the school.

56. This is particularly true of the Sceptics in so far as they advocated suspension of judgement by way of an argument from – the notion that for every ‘dogmatic’ view about underlying reality there is an opposing view of equal strength.

57. Although many philosophers, Plato and Aristotle especially, fought hard to distinguish true philosophy from legal and political debate, some, including Aristotle himself, recognised the similarities and what the former might owe to the latter. At On the heavens 294b7–13, in the middle of a discussion of his predecessors’ view on why the earth is at rest, Aristotle points out that it is a common habit we all share, to relate an inquiry not to the subject-matter, but to our opponent in argument. Arguing on both sides of a question, in utramque partem, was practised extensively in both philosophy and medicine – and its respectability and advisability were much debated: see, for example, Plutarch, , On Stoic self-contradictions 1035f–1036dGoogle Scholar.

58. Sivin, N., ‘Yin yang and the five phases’, in Lloyd, G. E. R. and Sivin, N., Tao and LogosGoogle Scholar, forthcoming.

59. See Sivin, ‘Cosmos’ (n. 29).

60. See Sivin, ‘Text’ (n. 31) at n. 7.

61. Han Fei's study concentrates on the psychological aspects of persuasion. No attempt is made to give a formal analysis of the modes of argument as such: there is no equivalent to Aristotle's survey of the enthymeme and the paradigm, the counterpart in rhetoric, so he claims, to syllogism and induction. See further Levi, J., ‘L'art de la persuasion à l'époque des Royaumes Combattants’, Extrême-Orient Extrême-Occident 14 (1992) 49–89CrossRefGoogle Scholar, and Lloyd, G. E. R., ‘The agora perspective’, Extrême-Orient Extrême-Occident 14(1992) 185–98CrossRefGoogle Scholar.

62. This list could, of course, be much extended. At the same time Greek dedications to those who wielded little or no political power are also common. This would seem to be the case with the person to whom Ptolemy dedicates both the Syntaxis and the Tetrabiblos, the otherwise unknown Syrus.

63. The recipients of the Ptolemies' patronage included poets and literary critics among many others, as much as, perhaps even more than, those engaged in one or other branch of the inquiry concerning nature: see Fraser, P. M., Ptolemaic Alexandria, 3 vols. (1972)Google Scholar. Their motivations were less a matter of seeking to foster science, than of boosting their own prestige, less a matter of sponsoring brilliant scientists, than one of sponsoring brilliance of any type.

64. See, for example, Griffin, M. T., Seneca: a philosopher in politics (1976) 202Google Scholar, Schofield, M., The Stoic idea of the city (1991)Google Scholar ch. 4. Similarly Dio Chrysostom combines advice to the emperor Trajan with a continued adherence to at least certain aspects of his Cynical philosophical ideals: see Desideri, P., Dione di Prusa: un intellettuale greco nell' impero romano (1978)Google Scholar ch. 5.

65. D.L. 2.66, 67, 70, 78.

66. D.L. 6.38.

67. See, for example, Pritchett, W. K. and Neugebauer, O., The calendars of Athens (1947)Google Scholar, van der Waerden, B. L., ‘Greek astronomical calendars and their relation to the Athenian civil calendar’, JHS 80 (1960) 168–80CrossRefGoogle Scholar, Meritt, B. D., The Athenian calendar in the fifth century (1928)Google Scholar and The Athenian year (1961).

68. See Philo, , Belopoeica 50.20–26Google Scholar, cf. Marsden, E. W., Greek and Roman artillery: technical treatises (1971)Google Scholar ch. 4.

69. In the classical period the possibility of implementing political proposals on the occasion of the founding of new colonies may have influenced such writings as Plato's Laws (if not also his Republic). The actual founding of Thurii in 443 B.C. is an enterprise in which Protagoras is reported to have been involved. However, to counterbalance this, we should bear in mind that the debate, reported in Aristotle, , Politics 1, 1253b14–5b15Google Scholar, on whether slavery is ‘natural’ or ‘conventional’, is one that remained at the purely theoretical level, and had no impact nor outcome in terms of the policies recommended or adopted with regard to slaves in Greek antiquity.

70. Outside those fields, too, the insistence on the importance of ‘philosophy’ sometimes reflects a bid to upgrade a particular calling, by representing it both as disinterested (that is, not driven by the profit motive) and as based on a sound theoretical understanding. Both the doctor Galen and the architect-engineer Vitruvius represent the best exponents of their two callings as the philosophically inclined ones. Thus even with domains that certainly do have important practical applications, the orientation towards the theoretical in some Greek and Roman writers is pronounced.

71. I am most grateful to all those who participated in the discussion of this paper, and especially for detailed written comments and observations from David McMullen, Nathan Sivin and Robert Wardy. They have been responsible for numerous corrections and improvements, not that they can be held in any way liable for the theses here expressed.