1. Le refers to Ronald Dworkin, Laws Empire (1986). Page numbers in parentheses are from the 1986 edition.

2. Except that interpretations through performance (of music or a play, etc.) display rather than explain the meaning of what they interpret

3. Hart, H.L.A., Postcript, The Concept of Law 239 (1994).Google Scholar

4. In one of the best studies of Hart's work, D.N. MacCormick has described Hart's internal point of view, reliance on which was central to his methodological innovation, as hermeneutic. See MacCormick, Neil, H.L.A. Hart 37–40 (1981)Google Scholar. I remember a conversation with Hart in which it was clear that he saw nothing wrong with the description. He was more ambiguous about the attractiveness of the word.

5. His other argument, consisting in a new and challenging account of the nature of interpretation, shows not that other theorists did not see their accounts as explanations of the meaning of, i.e., as interpretations of, social practices, but that they did not share his understanding of interpretation. I will not discuss Dworkin's own account of interpretation in the present article.

6. By the time the book was published, Dworkin was aware of the fact that Hart and others did not think of themselves as explaining the meaning of “law.” Nevertheless, he persisted in thinking that that is exactly what Hart was doing. Cf. Dworkin, , Law's Empire, supra note 1, at 418 n.29.Google Scholar

7. Hart, , The Concept of Law. supra note 3, at 246.Google Scholar

8. When I remarked that interpretation is the explanation of the meaning of its object, I used “meaning” broadly to include nonsemantic meaning. Explaining the meaning of words (“bachelor” means an unmarried male, etc.) is never an interpretation, and explaining the literal meaning of sentences—only given some special circumstances.

9. In that narrow sense of “semantics,” one needs more than semantics to answer questions of content. That when he said “I wish I were dead” he meant that he is very unhappy, and cannot see a way out, is not something we can learn from the meaning of the words or the sentence uttered (by itself) nor from rules for its use (alone).

10. In part, the same approach was supported by Wittgenstein's reflections on the variety of language games. For Hart's comment on those years, see Essays in Jurisprudence and Philosophy 2–3 (1983).Google Scholar

11. Hart, H.L.A., The Ascription of Responsibility and Rights, 49 Proc. Aristotelian Socy 171 (1948/1949)CrossRefGoogle Scholar, later disavowed by him.

12. Hart, H.L.A., Definition and Theory in Jurisprudence, reprinted in Essays, supra note 10.Google Scholar

13. My claim is not that Hart's analysis of legal statements and utterances is incompatible with belief in the objectivity of value and of morality. It is that the plausibility of the analysis depends on the rejection of the objectivity of value and morality. Once their objectivity is admitted there is no reason for accepting Hart's analysis rather than the view that legal statements and utterances are just like all other statements.

14. See Stavropoulos, N., Objectivity in Law 129–36 (1996)Google Scholar. For a conflicting view, see Kress, K., The Interpretive Turn 97 Ethics 834, 855ff (1987).CrossRefGoogle Scholar

15. To judge from conversations with him, and from a draft of an unpublished reply to Hart's Postscript

16. I should make it clear that this is not Dworkin's reason for regarding the question as a semantic one. I will come to his reason later.

17. Even when we count words with several meanings as several words, one for each meaning.

18. Perfect command of a concept implies being able to use it correctly in all possible circumstances. But not only is that a condition which in fact few achieve, it gives rise to theoretical difficulties. One who has perfect command of a concept can make mistakes in its application or use. But the boundary between a mistake about the concept and a mistake about its application is vague, as is its theoretical nature.

19. These remarks about the difference between philosophical explanations of concepts and the conditions for having concepts are consistent with and parallel C. Peacocke's distinction between possession and attribution conditions for concepts. & see his A Study of Concepts 27–33 (1992).Google Scholar

20. To be precise, his claim was that they do not habitually obey the commands, i.e., laws, of such institutions. That conditi'on, strictly understood, would mean that they are not disposed to obey, do not have a habit of obeying.

21. Two laws are constitutionally continuous if either they derive their validity from the same authorizing norm (directly or indirectly), or one of them is a basic norm and the other derives from it.

22. Most of the countries that gained their independence from Britain and France in the post-WWII years became independent without break of constitutional continuity. On the other hand, most countries absorb breaches in constitutional continuity without much effect on their identity. In Britain the loss of the Great Seal in 1688 and the House of Lords' Practice Statement of 1966 are sometimes mentioned as examples.

23. Strictly speaking, this sentence is false. Dworkin says (Le 45)Google Scholar that “I shall call the semantic sting the argument I have just described,” and that is “the argument that unless lawyers and judges share factual criteria about the grounds of law there can be no significant thought or debate about what the law is” (Le, 44)Google Scholar. I believe that I follow most readers of Dworkin in taking the sting to refer not to the argument which he finds mistaken, but to his own argument, which is meant to refute die mistaken argument, and exhibit its absurdity. In any case, the quotation above (Le, 44)Google Scholar cannot refer to the issue of how to explain die meanings of words or of concepts since it incorporates the claim that the law is identified by factual criteria, which is not part of that dispute.

24. For example, “Semantic theories suppose diat lawyers and judges use mainly the same criteria … in deciding when propositions of law are true or false; diey suppose diat lawyers actually agree about die grounds of law” (Le, 33)Google Scholar. Dworkin nowhere limits the rules to conditions of application in the narrow sense, i.e., conditions under which statements of the form “it is the law that …” are true. Even though his discussion gravitates in that direction, we should remember diat it is meant to apply to an explanation by reference to shared rules that provide criteria for the meaning of the word.

Like all explanations, criterial explanations are successful to the degree that they respond to the interests diat prompted diem.

25. The circularity in this characterization can be easily eliminated by making clear that die identity of the concept is determined by the existence of a population that uses a concept with criteria for correct use that are correct described by the explanation.

26. Wittgenstein's rule-following argument in Philosophical Investigations (1953), Putnam, 's twin-earth arguments in The Meaning of “Meaning,” and Is Semantics Possible? in H. Putnam, Mind, Language and Reality 215, 139 (1975)CrossRefGoogle Scholar. See also Kripke, S., Naming and Necessity (1980)Google Scholar. Burge, , Individualism and the Mental, 4 Midwest Stud. Phil. 73 (1979)CrossRefGoogle Scholar, and Other Bodies, in Thought and Object 97Google Scholar (Woodfield, ed., 1982)Google Scholar. My remarks in the sequel are not meant to relate to all aspects of their arguments. In particular, they do not relate to Putnam's conclusions regarding the “world-involving” aspects of natural kind words. Nor is there reason to expect my comments to be entirely in line with the arguments of these philosophers.

27. Note that there is no implication here that a linguistic community can share criteria only if there is someone who knows them completely. The criteria may be shared by the linguistic community even if no single person knows all aspects of them.

28. Or in British English, or in some dialect, depending on the language I am using at the time. It may be undetermined.

29. Burge has tried to show how radical and unappealing such a chang will have to be. See, e.g., his Individualism and Psychology, 95 Phil. Rev. 3 (1986)Google Scholar

30. Burge defines individualism as “the view that if one fixes those non-intentional physical and functional states and processes of a person's body whose nature is specifiable without reference to conditions beyond the person's bodily surfaces, one has thereby fixed the person's intentional mental states and processes in the sense that they could not be different intentional states and processes from the ones that they are.” See Burge, , Cartesian Error and the Objectivity of Perception, in Subject, Thought, and Context 117Google Scholar (Pettit, & McDowell, eds. 1986)Google Scholar

31. Naturally, they cannot make just any mistake. To be people who use the rule they must have some notion of what the criteria are.

32. See Finnis, John, Natural Law and Natural Rights (1980), ch. 1.Google Scholar

33. For my own explanation, see, among other places, Ethics in The Public Domain (rev. ed. 1990), chs. 9, 10.Google Scholar

34. In the Postscript, supra note 3, at 246, Hart writes that one mistake which led Dworkin to belief in his semantic sting argument is his conflation of meaning with criteria of application. (See also The Concept of Law at 160.) I do not know how Hart understood that distinction. Possibly my remarks regarding the relative independence of interlinked concepts is relevant to an elucidation of such a distinction. However, in the absence of an explanation of his distinction it is impossible to evaluate this reply by Hart.

35. The same is true of Dworkin's interpretive disputes about the meaning of a concept. They too presuppose that at least one of the parties has an incomplete understanding of the concept.

36. Do we use one concept of “promise” or two when saying “I promise to get you the book by tomorriw,” and when saying “I'll break your bones if you don it again, and that's a promise”? Most philosophers who have written on promises take it for granted that two concepts are involved. I believe that that position is not based on an analysis of the common understanding of the concept (s) of a promise, but on theoretical reasons they have for drawing the boundaries between promises and threats.

37. Reprinted as The Model of Rules—I, in Taking Rights Seriously (1977).

38. Note that that task is wider than the explanation of the nature of law, which is confined to essential features of the law, features without which it would not be law.

39. Since published as Gavison, R., Issues in Jurisprudence: the Influence of H.L.A Hart (1987).Google Scholar

40. Hart, , Postcript, supra note 3, at 247.Google Scholar

41. Legal rules are typically expressed by normative propositions (assigning rights, liabilities, responsibilities, etc.) or propositions setting conditions for the application of normative propositions. While often legislative measures are not formulated in such terms, their meaning is expressed by them and we can say that the legislation endorses the propositions or makes them into law.

42. You may say that not every explanation which improves understanding is an interpretation. But every explanation that improves the understanding of a phenomenon with a meaning, or a content, is an interpretation, for interpretation is an explanation, or display, of the meaning of what is explained. Here what is explained are social practices and institutions that constitute the backbone of legal systems. Their meaning is illuminated by jurisprudential explanations.

43. Though in all these instances an interpretation of this concept or that may be part of the case for the interpretation of the play, historical events, or law.

44. Similar litigation occurred in Rhodesia after it unilaterally declared independence. See discussions of these in Eekelaar, J., principles of Revolutionary LegalityGoogle Scholar and Finnis, J.M., Revolutions and Continuity of Law, in Oxford Essays in Jurisprudence (2nd series, A.W.B. Simpson ed., 1973).Google Scholar

45. Though, importantly, we do know that it has to accommodate the existence of “theoretical disputes” about the law.

46. In these comments I wish neither to endorse nor to deny the view that courts' decisions always represent the state of the law at a time just prior to their decision, as the courts believe it to be.

47. I can see plenty of reasons why they should not penalize people for violation of laws which were not in effect in the country, and so on. Such facts have a moral bearing on the issue of the justice of these courts. The theoretical conclusion that the system they operate is or is not a legal system seems to have little bearing.

48. Unless, of course, that system refers them to the writings of jurisprudence as setting a test of validity in it.

49. On the Nature of Law (The Kobe lectures of 1994), 82 Archiv Rechts & Sozialphilosophie 1–25 (1996).Google Scholar

50. Hart, , Postscript, supra note 3, at 241.Google Scholar