1. For the purposes of this discussion I am only considering arguments for a noncognitivist interpretation of legal discourse that are connected to the possibility of a noncognitivist interpretation of moral discourse. There are, no doubt, independent arguments for a noncognitivist interpretation of law, but I am not considering them here primarily because my goal is neither to establish noncognitivism in law nor is it to defeat such an interpretation of legal discourse. My goal is the more modest one of illustrating how arguments for noncognitivism might go and indicating the possible consequences of noncognitivism for claims to legitimate authority.

2. Moore, Michael, Law as a Functional KindGoogle Scholar, in Natural Law Theory: Contemporary Essays (George, Robert P. ed., 1992).Google Scholar

3. This suggests that relationalists who believe that legal discourse is fact-stating (cognitive) cannot accept a noncognitivist account of moral discourse. Because nonrelationalists deny that the truth of legal propositions entail the truth of moral claims, ethical noncognitivism poses no similar dilemma for the nonrelationalist. Whether moral judgments are assertoric or expressive has no direct bearing on the cognitive content of legal discourse. This is one reason for rejecting the common association of legal positivism with moral skepticism. Legal positivists can be justifiably agnostic about the semantics of moral discourse. This is to deny neither that some legal positivists are moral skeptics, nor that some legal positivists are motivated in part by moral skepticism. Still, there is nothing in the nature of legal positivism that calls for a particular semantics of moral discourse. The same cannot be said, of course, for relationalists. That, after all, is the point of the previous discussion.

4. I want to remind the reader that I am not claiming that noncognitivism in fact poses a problem for legal authority. I am only illustrating how such an argument might go, and why it is therefore more important than jurisprudes have so far acknowledged that these issues be addressed in a serious manner.

5. While theorists occasionally gesture towards coherence theories in law, their focus has largely been on coherence theories of knowledge, not truth. To my knowledge, no jurisprude has explicitly defended a coherence theory of truth in law, or at least any version that would be recognizable to a philosopher. Therefore, I will not be discussing coherence theories of truth in law.

6. This fact alone explains in part why I cringe whenever I read an essay in a law review criticizing liberalism because of its commitment, among other things, to abstraction and the correspondence theory of truth. Only law professors who know nothing about the state of philosophy could say things like this. Liberal political theories have no connections whatsoever to theories of the meaning of truth. Why anyone would think that analytic political liberalism would require such a controversial conception of truth is beyond me.

7. The prevailing view is that correspondence theories of truth imply metaphysical realism. There is, however, a minority view, expressed, for example, by Donald Davidson and Richard Rorty, that correspondence theories can be formulated in such a way as to avoid commitment to metaphysical realism. I avoid discussing the subtleties of these issues in this paper for the obvious reason that I am merely trying to set out a framework for jurisprudence, not a complete jurisprudence.

8. In the same way that there is a trivial sense of a world independent of our capacity to represent it-the world of stuff-there is a trivial sense in which objects have a conventional dimension. The names or words we use to refer to objects, e.g., “table,” “chair,” “soap,” are all conventional. We could have used any other name. Although some theorists have been tempted to use this form of conventionality to establish the conventionality of meaning or of meaning facts, no such inference is warranted. The realist view of a world as independent of our capacity to represent it presupposes that the world naturally differentiates itself, not that the words we use to refer to the differentiated stuff of the world cannot depend on human practices or conventions. The conventionality of the words we use has nothing to do with the more important question of whether object differentiation is essentially conventional. Nor is the claim here anything like the “anthropological” one that different cultures have different differentiation schemes, e.g., some see only two colors, whereas others have numerous concepts of white, etc

The anthropological claim for conceptual relativism is interesting, but makes no philosophical point. The ridiculous claim that all discourse is conventional because language is natural, because we could have used other words to refer to the same objects is neither interesting nor philosophical. It represents just another way in which much of the legal theory one sees in law reviews is so philosophically embarrassing, and explains why philosophers have not taken it seriously. The arguments against objectivity, truth, correspondence, etc, that permeate the literature of law reviews are unworthy of college sophomores.

Part of my current mission is to recapture jurisprudence from those who, through ignorance (if not ill will), have threatened to make a laughingstock of both it and those who work in it. The problems of jurisprudence are as difficult as are those in the philosophy of science, and should be treated with the same respect, and pursued by those with sufficient knowledge and skill to move the discussion forward.

9. In short: the world does not come to us already differentiated. Its differentiation is mediated by human practices and beliefs, in particular, views we have about what counts as an object. What counts as an object is determined epistemically: the best account of objects is that which illuminates best the claims we make about objects. Representations are of a differentiated world, but that differentiated world may not be independent of human practices and beliefs in the way in which a correspondence theory of truth requires.

10. Defenders of epistemic or verificationist conceptions of truth do not deny that there are facts. Nor must verificationists about truth deny that facts can be strongly objective in the realist sense. Theirs is an account of the meaning of the truth predicate; their claim is that in asserting that a proposition is true, we cannot be claiming that the fact asserted by the proposition corresponds to the way the world really is. Typically, of course, verificationist theories of truth reject metaphysical realism in favor of a conception of the world as evidence or mind-dependent. Still we have to remember that while a correspondence theorist almost must accept some form of metaphysical realism, being a metaphysical realist does not entail that one is a correspondence theorist about truth.

11. I am hardly competent to settle this debate about truth and its applicability to law, certainly not in the confines of this essay. My more modest ambition is to set out the contours of the debate so that we might have a frame of reference for discussing the role of truth in various jurisprudential theories. Natural lawyers, like Michael Moore, for example, are inclined to treat truth as a matter of correspondence. Their doing so implies a particular conception of what legal facts are and raises epistemic questions about how judges, for instance, are to gain access to those facts. Positivists have not generally taken up the question of what role the truth predicate plays in law, but many of them have been influenced by the later Wittgenstein and probably treat truth as a species of warranted assertability. The interesting case is Dworkin about whom I will have more to say below. I will return both to Dworkin on deflationism (about both objectivity and truth) and to Putnam on truth as warranted-assertability (under suitable conditions).

12. In my view, the way the (legal) world is is independent of our particular representations of it, but not of our capacity to represent it; legal facts are not independent of the beliefs and practices of judges and lawyers, but are not fixed by what those practices at a given moment in time happen to be; and truth in law is not identical with, but is nevertheless continuous with rational justification or warrant. In the final section of this essay, I expand just a bit on some of these claims, in particular, on my views about objectivity. My views are more fully developed in Coleman, Jules L. and Leiter, Brian, Determinacy, Objectivity and Authority, Univ. Penn. L. Rev.Google Scholar

13. The postmodernists are not alone in misunderstanding both pragmatism and the later Wittgenstein, but they should at least be given credit for making the most extravagant claims based on the least understanding.

14. Putnam, Hilary, Are Legal and Moral Values Made or Discovered?Google Scholar

15. For a discussion of the similarities and differences between Putnam and Rorty, see Brian Leiter's essay in this issue. Cf., Leiter, Brian, The Middle Way, 21–31.Google Scholar

16. Admittedly, Dworkin's views play only a minor role in Putnam's piece, but some of his misconceptions are sufficiently widespread among legal and nonlegal theorists alike that attending to them with some care may nevertheless prove to be a worthwhile corrective.

17. In fact, I cannot recall a discussion of vague predicates anywhere in Dworkin's work.

18. Dworkin, Ronald, Hard Cases in Taking Rights Seriously (1978).Google Scholar

19. Dworkin, , Laws Empire (1986).Google Scholar

20. Along with the change in underlying political theory comes a significant change in the role Hercules plays in the overall argument. It is arguable whether Hercules plays an epistemic or metaphysical role in “Hard Cases.” I am inclined to think that he fixes the right answers to legal disputes. The right answer is that which Hercules would give. In my interpretation of the argument, Hercules plays a metaphysical role. Others may be inclined to the view that right answers are fixed in some other way, and that Hercules is simply an extraordinarily reliable tracker of those answers. For those interpreters of Dworkin, Hercules plays an entirely epistemic role. Whatever interpretation of the role Hercules plays in Dworkin's jurisprudence one settles on, it is nevertheless true that Dworkin believes that judges should emulate as best they can the kind of reasoning Hercules would engage in. The reason for that is that the answers Hercules would come to are the correct ones. And judicial authority depends on seeing judges as enforcing right answers to legal disputes.

Things change by the time we get to Law's Empire. Dworkin is still committed to the view that judges should emulate or aspire to decide cases as Hercules would. The reason for their doing so changes, however. Ordinary judges should emulate Hercules, not because his answers are the correct ones-and they are-but because Hercules embodies the ideals of justice, fairness, and due process that ground political obligation, and, indirectly, the legitimacy of legal authority.

21. Dworkin, , Taking Rights Seriously, supra note 19.Google Scholar

22. Rawls, John, A Theory of Justice (1971).Google Scholar

23. I should note that some positivists, myself included, do not believe that moral standards cannot be part of a community's binding legal standards. So we are not committed to the view that appeal to such standards involves discretion. In the postumously published second edition of The Concept of Law, Hart explicitly accepts my formulation of legal positivism, thereby undermining the vast majority of Dworkin's objections of the sort we are currently considering in the text. See, Coleman, Jules L., Negative and Positive PositivismGoogle Scholar, in Coleman, J., Markets, Morals and the Law (1988)Google Scholar and Hart, H. L. A., The Concept of Law (1994).Google Scholar

24. It follows, therefore, that Dworkin, to use Michael Moore's phrase, is a relationalist-as is Michael Moore. Moral truths are always implicated in claims about the law's requirements. The difference between them (on this account) is that for Dworkin the truth-conditional relationship of law and morality results from his theory of law as adjudication and his theory of adjudication as interpretation, whereas for Moore, it follows from his semantic realism.

25. As I noted above, though Dworkin claims that the concept of truth is at the heart of jurisprudence, he has almost nothing to say about its meaning in law. I have ascribed deflationism to him simply because he explicitly advances deflationism about objectivity, and there is little point to being a deflationist about objectivity but not about truth.

26. Its emptiness may not actually matter to the issue before us, since we could have the concept of subjectivity even if nothing turned out to be genuinely subjective. We could, after all, have the concept of bachelor even if all men above the age of majority were married. It is an interesting philosophical question to ask, what conditions are required in order for a coherent concept to exist? Those issues fall far beyond the scope of this essay, however.

27. Coleman, & Leiter, , supra note 13.Google Scholar

28. Again, the analogy with truth is inapt. There is no contrasting and fully developed concept we can contrast with truth. That is, we have no independent, contentful conception of “false.” If we did, we might be able to give a nondeflationist account of truth by contrast with it. On the other hand, we have a contentful, independent conception of subjectivity that allows us to develop a contentful conception of objectivity by contrast with it.

29. I really do not want to rail too much against so much bad philosophy, but I do think we have to point out that the standard deconstructionist simply does not understand any of these subtleties about concepts, their content, the way concepts figure in explanations of our practices, and the like. Take an example like truth. I'm not a deflationist about truth, but it is true that there is no substantive independent concept of the false that can be used as a vehicle for understanding the concept of the true by contrast with it. This is great news for the deconstructionist, for upon noting or learning this she will claim to have deconstructed our concepts of true and false. What nonsense. And what of it? What is the philosophical implication of this deconstruction? That both concepts are empty? That the concepts don't matter in the sense that we should stop talking about assertions as being either true or false? (That's the postmodernist version, isn't it?) That they are subjective? Just what is the implication? Philosophers would not normally speak in terms of a deconstructed concept of truth, but we might well recognize that the concepts of true and false are connected in ways that will not allow us to specify the content of one in terms of the other. But philosophers are leery of drawing anything like a monumental conclusion from this fact. Instead, serious philosophers are likely to treat this fact as a reason for thinking that the concept of truth (and falsity) is best interpreted noncognitively. Then that hypothesis would be tested by asking whether a noncognitive interpretation of truth does justice to our practices, whether it allows us to understand a variety of claims about knowledge, error, and growth, for example, that are central to these practices. Most philosophers ignore these deconstructionist movements for two reasons. Most of the movement picks up but completely misunderstands the point of very complex philosophical arguments; and then it draws conclusions from those arguments that misunderstand the point of philosophy.

30. I should note that I do not recall Dworkin ever using the phrase “deflationist” to characterize his position. That is of no significance, however, since his position is a paradigm of what philosophers refer to as deflationist.

31. Dworkin has a tendency to draw metaphysical conclusions (though he might not regard them as metaphysical) from semantic claims. This is a mistake; the best example of it is his discussion of conventionalism's inability to provide a respectful or plausible account of theoretical disagreement in law. By their own lights, the parties appear to be disagreeing about what the law is, not about what it ought to be. Therefore, we have to posit a conception of what the law is (a metaphysics of the law) that captures their disagreement as they see it. Only then can we provide an explanation of their disagreement that is respectful. But this is surely wrong, and one look at the same debate in moral theory will show us why. In standard moral discourse, disagreeing parties take themselves to be disagreeing about whether a course of conduct, for example, is in fact right or wrong. Does that mean that we must posit a conception of morality that treats moral properties as real in order to make sense of their behavior? Absolutely not. Of course, we might posit real moral properties as a way of making sense of their behavior, but the entire point of noncognitivism in ethics is that we can explain all the relevant aspects of their behaviorincluding even rational and reasoned aspects of their disagreement-without positing moral properties of any sort.

My point is not that Dworkin is wrong to posit a conception of law that explains disagreement in a certain way. Rather, it is that he is wrong to draw the metaphysical inference about law from disagreement itself. The positivist cannot, it is true, explain disagreement by saying that the parties are disagreeing about what the law really is anymore than a noncognitivist can explain moral disagreement by saying that the parties are disagreeing about what is really right or wrong. But both have alternative explanations. The proper metaphysics (of law and morality) will depend on which explanation is, in a suitable sense, the most illuminating of the relevant practice. In fact, just as noncognitivists like Allan Gibbard argue that we can respect disagreement in morals by showing its rationality without being committed to the existence of moral properties, a positivist might argue that we can fully respect theoretical disagreement in law without being committed to the claim that the parties are disagreeing about what the law really is in controversial cases. Absent a total assessment of the coherence and explanatory value of a metaphysics of law, it is simply hasty and unwarranted to draw the inferences that Dworkin does from the mere fact of theoretical disagreement in law.

32. As an aside, notice that the judgment that we are never capable of being outside our practices itself presupposes an archidemean point from which that judgment is made. Similarly, is the claim that we are never capable of occupying a place outside of our practices true? If it is, what makes it true? Is it a fact? What sort of fact: objective, subjective? It cannot be a claim whose truth is determined inside a practice, since it is supposed to be a truth about all of our practices. If the claim should instead be interpreted noncognitively, why should it then given those of us who do not share the attitude toward it that postmodernists apparently do any pause. Isn't there a problem here?

33. I include Sellars because in fact the best arguments for some things postmodernists want to say are in Sellars, because one of Rorty's inspirations is clearly Sellars, and because Sellars is hard to read and my guess is that virtually none of the contemporary theorists who get their philosophy from Rorty (who, by the way, is really quite good in his ability to read and analyze in a serious way analytic philosophy) have ever tried to read Sellars.

34. Negative and Positive Positivism, in J. Coleman, Markets, Morals and the Law (1988).Google Scholar