Skip to main content Accessibility help

Herbert Hart and the Semantic Sting

  • Timothy A. O. Endicott (a1)


Even to disagree, we need to understand each other. If I reject what you say without understanding you, we will only have the illusion of a disagreement. You will be asserting one thing and I will be denying another. Even to disagree, we need some agreement.



Hide All

1. Dworkin, Ronald, Law's Empire 45 (1986). I will refer to this work as LE in the text, with page numbers in parentheses.

2. I conclude that criteria (in this special sense) are more or less complete from Dworkin's assumption that there can be no interesting disagreement about the requirements of the law between people who share criteria. That conclusion is also supported by Dworkin's suggestion that a theorist who suffers from the semantic sting will say that, in borderline cases, “people speak somewhat differently from one another. So lawyers may use the word ‘law’ differently in marginal cases …” (LE, 39). The suggestion is that two people who share criteria cannot disagree about an application of a word.

3. Hart, H.L.A., The Concept of Law 246 (2d ed. with Postscript; Bulloch, Penelope A. & Raz, Joseph eds., 1994). I will refer to this work in the text as Concept, with page numbers in parentheses.

4. Cf. Concept, 17: “Its [the book's] purpose is not to provide a definition of law, in the sense of a rule by reference to which the correctness of the use of the word can be tested.” See also Concept. 209.

5. Cf. the reference at Concept, 297 to Wittgenstein, Ludwig, Philosophical Investigations (1953), sees. 208–38, At sec. 242 Wittgenstein writes, “If language is to be a means of communication there must be agreement not only in definitions but also (queer as this may sound) in judgments.…”

6. Hart does claim that the officials of a legal system share criteria of legal validity (and his theory, though not his account of the meaning of the word “law,” is vulnerable if that claim cannot be supported). But those criteria are not semantic; see text below.

7. There is a further objection to the view that the notion of a rule of recognition commits Hart to a criterial semantics for the word “law.” A rule of recognition does not provide “criteria” in the Dworkinian, sense (see supra note 2): The tests for the truth of propositions of law provided by a rule of recognition lack the completeness of “criteria” (see Gh. VII.4 of Hart, , The Concept of Law).

8. Here too Hart could agree with Wittgenstein: “…if a person has not yet got the concepts, I shall teach him to use the words by means of examples and by practice.—And when I do this I do not communicate less to him than I know myself.” Wittgenstein, , supra note 5, at sec. 208.

9. In Taking Rights Seriously (1977), Ch. 2.

10. In Natural Law and Natural Rights (1980), John Finnis gives a valuable elaboration of this aspect of Hart's approach, associating it with Aristotle's notion of “pros hen homonymy,” and with Joseph Raz's approach to the extension of theoretical terms (at 9–11).

11. More precisely, the best conception will yield a determinate outcome in any case unless there is a de (which is inconceivable) or unless the preinterpretive materials which a conception must fit are too meager (which does not happen in any real legal system). For a recent discussion, see Dworkin, Ronald, Objectivity and Truth: You'd Better Believe It, 25 Phil. & Pub. Aff 87, 136–38 (1996).

12. The distinction Hart refers to is at Concept, 160; Hart discussed the same distinction in an introduction he wrote for Perelman, Chaim, The Idea of Justice and the Problem of Argument, vii, viii (1963): “Justice is a concept of complex structure within which we should distinguish a constant formal element and a varying material element. This distinction might be presented in terms used in recent English moral philosophy as one between the constant definition of justice and the varying criteria for its application.”

13. Cf. Taking Rights Seriously, supra note 9, at 136: “If the Court finds that the death penalty is cruel, it must do so on the basis of some principles or groups of principles that unite the death penalty with the thumbscrew and the rack.”

14. I do not mean that you can know English, or the meaning of words like “epic” or “cruel” without knowing anything else.

15. Note that Hart begins The Concept of Law with some crazy claims about law (from Llewellyn, Holmes, Gray, Austin, and Kelsen; Concept, 1). He does not dismiss them as senseless—he describes them as seeming “strange and paradoxical” and says that “such statements are both illuminating and puzzling: they are more like great exaggerations of some truths about law unduly neglected, than cool definitions” (Concept, 2).

16. I think that this claim about their debate is consistent with what Dworkin says about the nature of concepts, since “the contrast between concept and conception is … a contrast between levels of abstraction at which the interpretation of the practice can be studied” (LE, 71). “Political philosophy thrives … in spite of our difficulties in finding any adequate statement of the concept of justice” (LE, 93). He implies that we could say the same about the concept of law.

17. So we could not have a “pivotal” dispute about whether someone with no hair is bald, or about whether someone 7′6″ high is tall. I could still disagree with you even in a paradigm case, however, by making a crazy claim. There is no sharp distinction between pivotal disputes and disputes in which one party is making a crazy claim.

18. This claim becomes important in assessing Dworkin's theory if there is no answer to some questions of the application of vague expressions. For an argument to that effect (and a brief discussion of Dworkin's claim that abstract concepts are not vague), see Endicott, T.A.O., Vagueness and Legal Theory, 3 Legal Theory 37 ( 1997).


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed