1. Versions of this paper were also read at a departmental colloquium at Harvard, at Pittsburgh University, and at the annual conference of the Israel Society for the Promotion of Classical Studies at Bar-Ilan University. I am grateful for a great number of critical remarks and suggestions on all these occasions, and especially for the written comments of Professors Brad Inwood and Arthur Madigan. Since the paper is to be published with Prof. Inwood's comments, I have left the main text virtually unchanged, and tried to add some responses in the form of footnotes.
2. e.g. Leo Strauss, Natural Right and History, Chicago 1953, ch. IV, p. 110, p. 146; more recently, K. H. Ilting, Naturrecht und Sittlichkeit, Stuttgart 1983, p. 35, and for Aristotle, W. von Leyden, Aristotle on Equality and Justice, London 1985, pp. 84-90. I have unfortunately not been able to see the article by D. N. Schroeder, Aristotle on Law, Polis IV, 1981, 17-31, who apparently argues that Aristotle was not a natural law theorist. 3. G. Watson, The Natural Law and Stoicism, in: Problems in Stoicism, ed. A. A. Long, London 1971, 216-238.
4. Plato uses the usual formulation in Lgg. X 890A (KaTa (p 6 (T L vs. KŒTà v6pov). I cannot see that either Antiphon or Callicles are advocating the idea that human laws should be modelled on nature's, and so contrary to some scholars - I also do not see them as precursors of the later natural law theory.
5. For a lucid description of this conception of rules, see J. Rawls, "Two Concepts of Rules," Phil. Rev. 64, 1955, 3-32.
6. Despite the promising opening chapter of EN VI (1138b 15-34), Aristotle never explains what is the "right reason" (6p$6g \6yog) that enables the man of practical wisdom to find the mean. He may have thought that since general rules cannot guarantee right action, there could be no general account of this either. If so, he was probably wrong, for even if general rules cannot guarantee the right result, it does not follow that there could not be a general description of it. After all, the superior knowledge of the Platonic ruler that entitles him to correct or ignore the law would presumably itself be general knowledge. The occasional mismatch between rules and desired results cannot be explained by saying that rules are general, results are not. It is rather that there may be indefinitely many ways of reaching a correct result, so that the method that works in most cases, and is therefore laid down as a rule, may not work under unusual circumstances. Exceptions to the rules will be justified if it turns out that some non-standard procedure is more likely to lead to success. The weakness of the law, then, derives not from its generality tout court, but rather from the fact that there often is no unique method for reaching a good or just outcome. Neither knowledge of general rules nor general knowledge of what would be a good result will be sufficient to provide the diagnostic capacities of the man of practical wisdom to which Aristotle appeals - the ability to foresee the consequences of a particular course of action and hence to decide what should be done in a given situation. This seems to be a capacity to correctly apply general knowledge. But the fact that different courses of action may be advisable in different cases does not show that the standard of rightness itself varies with the circumstances.
7. e.g. R. M. Pizzorni, 11 Diritto Naturale dalle origini a S. Tommaso dAquino, Roma, 1978, pp. 19, 20.
8. Several critics have pointed out to me that my preoccupation with discovering the origins of natural law-theory in the development of ethics has led me to neglect the antecedents of Stoic thought in Pre- Socratic or Platonic cosmology. There is no denying, of course, that much of Stoic cosmology, and in particular the idea of a rationally ordered universe, goes back beyond Socrates. The most striking anticipation might seem to be Heraclitus' fr. B114 (DK I p. 176): "Speaking with understanding they must hold fast to what is shared by all, as a city holds to its law, and even more firmly. For all human laws are nourished by the divine one. ft prevails as it will and suffices for all and is more than enough" (tr. Ch. H. Kahn, in: The Art and Thought of Heraclitus, Cambridge 1979, p. 43, with one modification). Obviously, the attempt to base morality on cosmology requires a certain kind of cosmological theory, and in this respect the Stoics had important and influential predecessors. But I would still want to insist that the conception of natural as opposed to human or conventional law, which is meant to address the problem of objectivity, originated with the Stoics (not to mention their attempt to define the good in terms of cosmic order, and thereby to solve the problem of congruence - a project that was not generally taken over by the natural law-tradition). It is not clear what Heraclitus meant by saying that human laws are "nourished by" the divine one. But I doubt that one should read more into this than the thought that the order of a city, its legal system, is but a weak, particular case, or perhaps an analogon, of a wider and stronger cosmic order. This is still far from the claim that moral or legal rules can be derived or measured against natural, objective rules. Tentatively, then, I would say that the elements for a natural law theory could be found in the cosmological tradition; the theory itself was not there.
9. This does not mean that the Stoics thought they could produce a set of moral rules that would admit of no exceptions. Indeed, as Prof. Inwood points out in his comments, the Stoics were quite aware that rules of conduct will be subject to exceptions under unusual circum - stances. Hence their belief that only the sage who knows when and where an exception is justified will achieve perfect virtue. (The last pas - sage he mentions, Plut. Stoic. repugn. 1037E, however, seems to me to be concerned with the correct application of rules rather than with ex - ceptions.) But those exceptions would not be seen as infringements of rules. Rather, if I understand the Stoics aright, they would be justified by appeal to higher-order rules. That is to say, exceptions occur, according to Stoic theory, when there is a conflict of rules, and in such cases one needs some order of priority among rules to decide which rule should override which other. In contrast to Plato and Aristotle, then, the standard of rightness will still be conformity with the law of nature, not some state of affairs to be brought about by correct action.