No CrossRef data available.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 16 February 2009
In a recent article Brian Leiter concluded that a useful normative theory of adjudication is impossible. A normative theory of adjudication would be a theory that, among other things, identified the moral and political norms that judges ought to follow in determining the law for any particular legal dispute. Letter's elegant and subtle argument, stripped to its bones, runs as follows: Philosophers of law regard a correct normative theory of adjudication as being dependent upon an antecedent descriptive theory. The dependence here, as Leiter describes it, is of a very strong sort and unique among philosophical theories: Any normative theory, to be acceptable, cannot depart from the actual practice of judges and lawyers. Consequently, the content of the normative side of the theory is simply to “continue to do what you've been doing,” supplemented, perhaps, by Holme's injunction to do it more selfconsciously and explicitly.
1. Leiter, Brian, Heidegger and the Theory of Adjudication, 106 Yale L.J. 253 (1996).CrossRefGoogle Scholar
2. Leiter is in fact somewhat ambivalent in his conclusions. He states firmly early in the article that “[t]o the extent that the theory of adjudication is in the business of description, the theory of adjudication is out of business.” Id. at 261. Much later in the article, he says categorically of the theory's normative ambitions: “In other words, because normative theory of adjudication is parasitic upon accurate description, the impossibility of complete description necessarily handicaps prescription. If what judges are doing in deciding cases is, in significant part, dependent upon the horizon of intelligibility created by the [Heideggerian] Background, and if we have no way of articulating the type of understanding constituted by the Background, then, on the normative side, we can do no better than an empty gesture.” Id. at 276. Just a page later, however, in a footnote, Leiter observes, “Of course, it remains possible, and indeed likely, that we can still articulate norms for decision making that speak to those aspects of judicial decision that are not coextensive with the Background competence. Such norms are also likely to be quite important.” Id. at 277, n. 87. We address here the stronger position, though it makes litde diiference to our critique as we argue both that Leiter is wrong about the relationship between description and prescription in the theory of adjudication and misunderstands the implications of Heidegger for normative theory.
3. Id. at 255–61.
4. Id. at 276–77.
5. Id. at 254.
6. Ronald Dworkin, Law's Empire 225 (1986).Google Scholar
7. See Sunstein, Cass, Legal Reasoning and Political Conflict (1996).Google Scholar
8. See, e.g., Gerhardt, Michael J., A Tale of Two Textualists: A Critical Comparison of Justices Black and Scalia, 74 Boston L. Rev. 24 (1994)Google Scholar, and Eskridge, William N. et al. Statutory Interpretation as Practical Reasoning, 42 Stan. L. Rev. 321 (1990).CrossRefGoogle Scholar
9. See generally Posner, Richard A., Economic Analysis of Law (3d ed. 1986).Google Scholar
10. Leiter, supra note 1, at 282Google Scholar (emphasis in original). Leiter's proposal to pursue any descriptive theory at all may appear perverse given his conclusion about the impossibility of descriptive theory. Leiter appears to be suggesting that his critique applies only to descriptions that include mentalisuc elements, and not those that are purely behavioristic. It is not clear from his argument that any such distinction is sound. It is nonetheless surely perverse to rely on Heidegger to pave the way to a behavioristic or non-normative account of social practices inspired by Quine, as we discuss in note 80 infra.
11. Dworkin, , supra note 6, at 141.Google Scholar
12. The legal doctrine of estoppel is generally invoked to prevent a party from denying the truth of an earlier statement, which was uttered when another, innocent party has relied on the statement to his detriment See Black, 's Law Dictionary 851 (6th ed. 1990)Google Scholar. The same idea is evident in the notion of “frustration of investment-backed expectations,” a part of the Supreme Court's takings jurisprudence since Penn Central Transportation Co v City of New York, 438 U.S. 104 (1978).Google Scholar
13. Russell, Bertrand, A History of Western Philosophy 606 (1945).Google Scholar
14. Searle, John R., The Construction of Social Reality 27–29 (1995).Google Scholar
15. Quine, W.V., Ontological Relativity and Other Essays 27 (1969)Google Scholar. Time has not tempered Quine's behaviorism. In a relatively recent essay in the Times Literary Supplement, Quine wrote:
The purpose of words, people used to say, is to communicate ideas. But if we know what ideas people mean by their words, we know it only by observing how they and others use the words. Oh, we can ask them, but then we just get more words. Words are what we have to go on; ideas are as may be. Verbum sapienti satis est.
Quine, W. V., Words Are All We Have to Go On, Times Lit. Supplement, 07 3, 1992, at 8Google Scholar
16. Leiter, , supra note 1, at 282.Google Scholar
17. Hare, R.M., Freedom and Reason v (1963).Google Scholar
18. Hart, H.L.A., The Concept of Law 17 (2d ed 1994)Google Scholar
19. Gibbard, Allan, Wise Choices, Apt Feelings 7 (1990)Google Scholar
20. Leiter, , supra note 1, at 256–57.Google Scholar
21. Hart, , supra note 18, at 101–2.Google Scholar
22. Leiter, , supra note 1, at 256 n. 8.Google Scholar
23. See, e.g., Llewellyn, Karl, The Common Law Tradition: Deciding Appeals 121–26 (1960)Google Scholar, and Jerome Frank, Law and the Modern Mind ch 4 (1930)
24. Professor Madry was confronted by a classic example of this phenomenon early in his career. During a U.S. District Court clerkship, Madry was assigned to write a draft opinion for an appeals court case on which the judge sat by designation. The judge indicated his initial position on the case, which turned out after some research to be completely unsupported by fairly clear Supreme Court precedent. Weeks of argument ensued and were finally ended when the judge said to the effect, “I don't care what the law is, When I was sworn in, the chief judge at the time, my mentor, told me to always do the just thing, and then make the law fit.”
25. Leiter, , supra note 1, at 257Google Scholar. There is a trivial sense in which current practice is substantially correct, but it has little bearing on Leiter's argument as he frames it. In developing normative theories, we routinely disregard possible but extremely unlikely facts as irrelevant to our considerations. For example, thinking about the morality of murder, we do not consider the possibility that tomatoes are slumbering human souls and who therefore deserve the same regard as other humans. There is nothing impossible about the suggestion, but it is outlandish and doesn't deserve any serious consideration. If we understand “correct,” then, as being based on a reasonable set of facts, because current practice likely is based substantially on facts that we would all acknowledge as relevant, it is also thereby substantially correct We differ only on the details of those facts or their implications for how we ought to behave. Leiter, however, is suggesting an even stronger kind of correctness where even in its details, current practice is correct.
26. Id. at 258.
27. Id. at 259.
28. Holmes, Oliver Wendell Jr., The Path of the Law, 10 Harv. L. Rev. 465–67 (1897).Google Scholar
30. Id. at 467.
30. Id. at 469.
31. Id. at 467.
32. Id. at 468.
33. Id. at 469.
34. Moore, Michael S., The Need for a Theory of Legal Theories: Assessing Pragmatic Instrumentalism, 69 Cornell L. Rev. 988, 989 (1984)Google Scholar. See also at 993 (“These [core] theories [of adjudication, legislation, and citizenship] are both descriptive and normative, telling us how judges, legislators, and citizens do and should perform their roles”).
35. Leiter, , supra note 1, at 259.Google Scholar
36. Moore, , supra note 34, at 993.Google Scholar
37. Id.
38. Id. at 1005.
39. See Leiter, , supra note 1, at 259Google Scholar citing Coleman, Jules, Negative and Positive Positivism, 11 J. Legal Stud. 139 (1982)CrossRefGoogle Scholar; Raz, Joseph, Legal Positivism and the Sources of Law, in The Authority of Law (1979)Google Scholar; and Lyons, David, Principles, Positivism and Legal Theory, 87 Yale L. J. 415 (1977).CrossRefGoogle Scholar
40. The exception, and it is a minor one, is Coleman, who appends a rather lengthy final footnote to his article in which he criticizes Dworkin's early theory of adjudication. Even here, however, Coleman s arguments have more to do with certain inconsistencies in Dworkin's own arguments. He is not himself offering a theory of adjudication or critiquing Dworkin's arguments in a way that implicates the kind of methodology that Leiter suggests is endemic to a theory of adjudication. See id. at 163–64.
41. See Raz, , supra note 39, at 48–52.Google Scholar
42. Dworkin, , supra note 6, at 239Google Scholar. Though the quoted portion concerns common-law adjudication, Dworkin applies the same principles to all areas of law.
43. Id. at 240–48.
44. Id. at 255.
45. Id. at 256.
46. Id. at 52 (emphasis in original). Dworkin also says elsewhere that “fitting what judges did is more important than fitting what they said.” Id. at 248. This degree of subjectivity in the determination of laws confronts Dworkin with a dilemma he does not appear to recognize. Law's Empire is framed largely as an account of what makes propositions of law true. Dworkin wants to insist that propositions of law are true when supported by a constructive interpretation of the authoritative legal literature, most emphatically not only when they enjoy a certain historical pedigree, like having been enacted, adopted, or announced by an authorized legal agency. Reformulating his theory of adjudication into a theory of truth, Dworkin writes, “According to law as integrity, propositions of law are true if they figure in or follow from the principles of justice. fairness and procedural due process that provide the best constructive interpretation of the community's, legal practice.” Id. at 225. Giventhe strong subjective element involved in construcuve interpretation, however, there cannot be one single true statement of the law but potentially as many true Statements of law as there are interpreters. This would involve a very odd notion of truth and would in any event defeat Dworkin's apparent preference for a public standard for criticizing judicial opinions. See id. at 11.
47. Id. at 139.
48. Id. at 65 (emphasis added).
49. Id. at 44. Dworkin means by “grounds of law” other propositions the acceptance of which made propositions of law true. Thus, identifying grounds of law is giving a theory of truth for law. Because Dworkin insists that grounds of law include more than historical facts, embracing all that would be consulted in a constructive interpretation, his theory of adjudication dovetails with his theory of truth. For positivista, who see the grounds of law as constituted by historical facts, the theory of truth for law is distinct from the theory of adjudication, and all that Dworkin regards as grounds for truth beyond historical facts, positivists might regard as possible sources of law. One could, therefore, be a positivist with regard to a theory of legal truth and still embrace constructive interpretation as an accurate descriptive theory of adjudication.
50. Id. at 44.
51. Id. at 45.
52. 115 S. Ct. 1624 (1995).
53. Dworkin, , supra note 6, at 62–63 (emphasis in original).Google Scholar
54. Id. at 64.
55. Searle, John R., The Construction of Social Reality 23 (1995).Google Scholar
56. Id. at 25–26.
57. Dworkin, , supra note 6, at 63–64Google Scholar
58. Id. at 64. Dworkin's only evidence of the kind of substantial disgreement that he relies should be taken as representative and not just eccentric. The mere presemce in any society of eccentrics surely cannot be taken as evidence of the lack of agreement or acceptance necessary to constitute a social practice.
59. Id. at 68.
60. Id. at 73.
61. Id. at 69.
62. Id. at 68.
63. Id.
64. Id. at 32.
65. Id. at 68 and 73.
66. W. at 50.
67. We will see that this way of putting the issue is misleading.
68. It should be pointed out that interpreting Heidegger is quite tricky. And there are multiple readings of what Heidegger is up to. However, this reading of Heidegger is shared by Leiter. Both of us, as is much of the community, have been influenced by the important work of Hubert Dreyfus. See Hubert Dreyfus, Being-in-the-World (1991). We will argue that Leiter mistakes some of the implications of his own interpretation of Heidegger.
69. From a Heideggerian point of view, there seem to be two sufficient conditions for paralysis. It would be sufficient to paralyze action if the criteria for relevance were either conscious or rule-governed. The former case is well-known from the everyday experience of observing the paralyzing effects of hyper self-consciousness. The latter case is the problem of the infinite regress of rules. If our mental life is governed by rules, won't you need further rules to determine which rule to use, ad infinitum? From a Heideggerian perspective, the fact that there are two sufficient conditions for paralysis should not be surprising if consciousness has a rule structure.
70. In artificial intelligence research, this is called the “frame problem”; i.e., for an agent to act effectively, the irrelevant should be invisible to the agent. Noticing the irrelevant paralyzes behavior. Yet the irrelevant has to be noticed in the sense that the only way to know whether something is relevant or irrelevant is by looking at it The “frame problem” is, how do you know something is irrelevant without looking at it? Or in computer terms, how do you program a computer to ignore something without programming the computer to look at it to know to ignore it?
This problem has not been solved. Computer programs to date are confined to extremely circumscribed situations so that the frame problem does not appear. But the goal of AI is to create an intelligent computer, i.e., a computer that can function in the relatively open situations such as those that we function in. In such situations, the frame problem occurs for computers. An excellent introduction to the frame problem can be found in Dennett, Daniel, Cognitive Wheels: The Frame Problem in Artificial IntelligenceGoogle Scholar, in The Robot's Dilemma: The Frame Problem in Artificial Intelligence (Pylyshyn, Zenon ed., 1987).Google Scholar
71. This is doubly controversial. Many people believe that our behavior is rule-governed by non-conscious, and hence non-introspectable, rules. That is obviously the premise behind much AI research, which is trying to program a computer with those rules. And it is the premise behind much social science research. Second, it is an open question whether science must describe human behavior in terms of rules at all. Psychologists are currently developing non-rule-based scientific (computational) explanations for human behavior (viz., parallel distributed-processing models). See, e.g., Connections and Symbols (Pinker, Steven & Mehler, Jacques Eds., 1988).Google Scholar
72. Leiter mistakenly claims that according to Heidegger the background can't be made explicit, because of its vastness, the “sheer magnitude” of the job of articulating the background whose “complexity is staggering” (Leiter, , supra note 1, at 267–68).Google Scholar That is not correct. It is not in us to make explicit; i.e., no amount of introspection can reveal it—even infinite introspection.
Further, the character of the background prevents it from being made explicit There is an infinite number of renderings of the background. This is because it does not consist in well-defined, individuated objects. The background is a mass concept, not a count concept. So, according to Heidegger, there is an indefinite number of ways to individuate the background given our projects and interests.
73. This is highly controversial. The most common view in cognitive psychology is that the criteria for relevance does consist of rules in our head; i.e., stored as neural events in our brains.
74. Leiter, , supra note 1, at 261 (emphasis added).Google Scholar
75. It turns out that the leading American commentator on Heidegger, not surprisingly, is also the leading American critic of artificial intelligence, Hubert Dreyfus. See Dreyfus, Hubert, What Computers Cant Do: The Dmits of Artificial Intelligence (1979).Google Scholar
76. To put this in more of a Heideggerian language, it is not a mistake to adopt the present-at-hand stance. It is not avoidable. It is part of everydayness. The mistake is not to see the present-at-hand in the larger context of being-in-the-world.
77. This includes, according to Heidegger, positivist social scientists and artificial intelligence researchers, who do not see that their actual everyday work violates their philosophic pronouncements.
78. It should be pointed out that Heidegger is a conservative writing in the defense of authority, not a radical, skeptic, or anarchist. By insisting that rules are contextual, he is blocking the possibility of total or radical reform: if we cannot grasp the full sense of our rules, radical reform would be an illusion and attempts at radical reform will always carry in its wake unintended consequences. Similar arguments can be found in Michael Polanyi, Personal Knowledge: Towards a Post-Critical Philosophy (1962), and Michael Oakeshott, Experience and Its Modes (1985). These are twentieth-century reworkings of arguments defending tradition over reason on the grounds that reason operates within a background tradition that reason cannot fully analyze.
79. This section is compatible with Heidegger's work, but it is not Heideggerian.
80. Leiter, , supra note 1, at 281.Google Scholar
81. We have not discussed Leiter's proposal to provide a norm-free, Quinean descriptive theory of adjudication because this proposal is, in the article we are discussing, nothing more than that, a proposal for future work. Nonetheless, Leiter's apparent argumentative strategy, begun in the present work, of using Heidegger as a prelude to a Quinean project is very odd. Leiter essentially utilizes Heidegger's thesis, that a complete description of our practices is impossible, to argue for Quine's thesis that normative projects (such as normative theories of adjudication or normative epistemology) are nonstarters. This is a problematic strategy. Quine does not need nor want Heidegger's help. Quine has powerful, serious, controversial, and un-Heideggerian arguments for his position. And Heidegger's argument leads to some very un-Quinean conclusions.
Leiter's proposal appears to elide this mismatch. Indeed, Leiter seems to be attempting to shape Heidegger to fit into a Quinean mold. But it cannot be done. Quine is the arch-naturalist, who believes that science explains everything (and if it can't be explained by science, it is not worth explaining). Heidegger is the arch-anti-naturalist, who believes that science cannot explain anything of importance (and when it does offer explanations, it distorts or loses the phenomenon in question). According to Quine, the world has no values, just physical happenings: physical events following physical events. Physics explains all. Values and ethics are treated as relics of a bygone religious era. But in the world described by Heidegger, physics explains little, because for Heidegger our existence is thick with value. The basic level of existence is the ready-to-hand; i.e., our functional relationships with the equipment around us—that is, our being-in-the-world. Or to put that differently, we live in a world of value and normativity. And only in our moments of abstraction and reflection does the world appear Quinean to us. It is not how we live.
Who is right, Quine or Heidegger? That is difficult to say. But it is clear that neither would want to be used to defend the other. More importantly, it is not clear how either could be used to defend the other. Heidegger's argument in a nutshell leads to the conclusion that we are all engaged in normative projects all of the time. It is not that we cannot judge; it is that we cannot help but to judge. Normative projects, such as the normative project of a theory of adjudication, are unavoidable. You could not create a non-normative theory of adjudication. To attempt to derive from that the Quinean conclusion that normative projects are a nonstarter seems to be—a nonstarter.