Understanding Legal Understanding-- Part I
Copyright 1993-1999 Jack M. Balkin. All Rights Reserved.
Understanding Legal Understanding: The Legal Subject and the Problem of Legal Coherence
Similarly, our inquiry into the contributions of "the legal subject" does not imply that the object of legal interpretation has no existence independent of a particular subject's comprehension of it. Our subjectivity contributes to, but does not create the cultural objects we comprehend. This is the dialectic between the subjective and objective aspects of social life--between individual thought, belief, and action on the one hand, and language, ideology, culture, conventions, and social institutions on the other. Culture and cultural objects have meaning only when they are understood by subjects, but their meaning is not dependent on the view of any particular subject. Language, ideology, culture, conventions, and social institutions construct and constitute the individual's subjectivity; yet language, ideology, culture, conventions, and social institutions exist only as instantiated in the thoughts, beliefs, and actions of individuals.(4)
In my view, the relative lack of interest in the legal subject in standard jurisprudential accounts is itself a sociological phenomenon worthy of study. Its key ideological feature is projection: the attribution of features of the self to the objects of understanding.(5) When jurisprudential discussions neglect what the legal subject brings to the object of interpretation, they project the subject's contribution onto the object, thereby manifesting this contribution as a feature or property of law.(6)
This ideological projection has two consequences. First, it misdescribes the nature of the legal system, ignoring its necessary connections to human understanding. Second, and equally important, ideological projection makes the subject's contribution to the legal system invisible. It cuts off possible avenues of inquiry about the "we" who understand the law, thus shielding the subject from intellectual scrutiny.(7)
Indeed, the ideological concealment of the legal subject persists even in recent jurisprudential accounts that are eager to assert the fundamentally interpretive character of law. Consider, for example, Ronald Dworkin's well-known theory of law as integrity. The strength of Dworkin's approach is its premise that legal understanding is an interpretive activity.(8) Thus Dworkin asserts that individual subjectivity must play a significant role in the construction of features of the legal system.(9) However, Dworkin has been unwilling to take this insight to its logical conclusion. If law is truly an interpretive enterprise, we must necessarily be concerned with the ideological, sociological, and psychological features of interpretation and their effects on our internal experience of understanding. This Dworkin has consistently refused to do. He has made clear that jurisprudential critique of a legal interpreter's views may not proceed on sociological, psychological, or ideological grounds. For Dworkin, these are "external" critiques, which can play no role in describing the nature of law.(10) The nature of law is determined by the internal perspective alone--the view of a subject making arguments within the legal system.(11) Hence, one learns nothing about the coherence of the legal system from understanding psychological or ideological needs to see law as coherent or incoherent. Rather, one must make claims about legal coherence and incoherence wholly within the terms of legal discourse by offering legal arguments for why a particular collection of doctrines or a particular justificatory scheme is coherent or incoherent.(12)
Thus, for Dworkin, the legal subject makes her appearance at the beginning of legal theory only to be shielded, protected, isolated, and forgotten thereafter. Once the legal subject has been installed as the interpreter and hence arbiter of the nature of law, psychological and ideological contributions to what she interprets are wholly beyond the bounds of jurisprudential scrutiny. All arguments must be directed instead to the nature of the object she constructs. In this way, Dworkin makes the subject who understands the law as coherent or incoherent invisible and impervious to psychological or ideological inquiry. Ironically, then, Dworkin, an early champion of the interpretive conception of law, has been among the most determined to exclude the legal subject as a subject of jurisprudential study.
Since H.L.A. Hart, jurisprudence has been grounded on the so-called "internal point of view"--the perspective of a participant in the legal system who regards its laws as norms for her behavior.(13) In contrast, I believe that we must ground jurisprudence in a critical perspective, one that employs ideological critique to reflect on our internal experience of law. A critical perspective does not reject the importance of the internal perspective; on the contrary, it makes it a central object of analysis and scrutiny. It recognizes that this perspective constitutes law rather than simply mirrors it and acknowledges the plurality of its forms. A critical perspective takes seriously the contributions of subjectivity to the nature of law; it treats the sociology of knowledge as a full partner in the jurisprudential enterprise. Instead of taking for granted the primacy of the internal viewpoint of participants in the legal system, a critical perspective asks how this internal experience comes about. It recognizes in the internal experience of legal norms an effect whose causes must be unearthed and reflected upon.
To this end, we must recognize how deeply subjectivity affects all forms of legal understanding, not merely those in which we think that a particular individual misunderstands the legal system. It is tempting to describe legal misunderstanding in terms of something additional that a particular subject has brought to the act of comprehension--a certain baggage, whether of ideology or ignorance, that affects (and therefore distorts) the object she comprehends. Yet this characterization misunderstands the nature of misunderstanding because, ironically, it does not go far enough. It merely reinforces the ideological camouflage that makes subjectivity invisible in the ordinary case of legal understanding. Conceding that deviant forms of legal understanding may affect the construction of the object of understanding simultaneously treats these cases as exceptional to the normal, proper case of legal understanding, in which the subject does not intrude upon, distort, or affect the legal object but rather perceives the legal system as it really is. Drawing such distinctions obscures the legal subject's inevitable contribution to the world she understands. Hence, one arrives at jurisprudential projects that argue that they are concerned only with the nature of the law and not the types of mistakes that individuals might make about the legal system. In this way, the contribution of the subject who understands the legal system in the "correct" way is made invisible or irrelevant to jurisprudential inquiry.
In contrast, I insist that jurisprudence must consider how the legal subject constructs the object of her study even in those cases we consider normal or proper instances of legal understanding. The "normal" case of legal understanding is really a deviant case whose deviance has gone unnoticed, projected instead onto the object of legal understanding and given the name of legal reality. Legal understanding is a special case of legal misunderstanding: it is a "misunderstanding" appropriate for the purpose at hand. In both what we call legal understanding and what we assign to legal misunderstanding, the subject has already intruded and brought her fore-understandings, prejudgments, and psychological needs to bear. The subject is already part of the constructed object of interpretation; her invisibility is already reflected in the object's nature.
What difference will a critical or a cultural perspective make to the study of jurisprudence? What new light will be shed on jurisprudential questions by considering the legal subject's contribution to law? The jurisprudential approach I advocate transforms the answers to traditional questions and poses new ones. This Article argues that our subjectivity affects the process of legal understanding in three different ways that conventional jurisprudential discussions have either de-emphasized or ignored.
First, subjects bring purposes to their understanding of law. The study of subjectivity is important to jurisprudence because we must recognize the different contributions we make to the object of understanding when we approach it for different purposes. Legal understanding is not simply the apprehension of preexisting properties of an object. It is a purposive activity of subjects. It is something that we do.(14) To understand as a subject is to understand for a purpose, to participate in a social practice of understanding, and therefore to understand in a certain way that is driven by that purpose and that practice of understanding. We understand law for many different purposes, and we engage in many different activities that we unthinkingly lump together and call "understanding law."
Judgments of legal coherence result from one particular activity of understanding, an interpretive attitude I call rational reconstruction. This is an attempt to see the substantive reason that emanates from legal materials. However, rational reconstruction is not the sole or even the most important form of legal understanding. There are as many different kinds of legal understanding as there are purposes in understanding law. These different forms or activities of understanding are mutually interdependent, but they are not identical.
Jurisprudence has traditionally focused on a single "internal perspective" of participants in the legal system; it thus projects a particular interpretive stance onto the object of its contemplation and describes what it finds as "the nature of law." Nevertheless, because participants engage in a number of activities called "understanding law," the internal perspective is actually a multiple perspective. Instead of a privileged form of legal understanding that reveals the true nature of law, with all other forms of understanding parasitic upon it, I propose an alternative picture: a network of overlapping and interconnected forms of legal understanding, in which hierarchical status and relationships of dependence are provisional and much less clear. Focusing on the legal subject's contribution to legal understanding helps us to separate the various forms of legal understanding and grasp their mutual dependence and differentiation.
Second, subjectivity is important to understanding the nature of law because judgments of legal coherence and incoherence rest upon and are driven by features of the self. They are shaped by the features and sources of our understanding--our preexisting commitments, values, and beliefs, and our knowledge and ignorance of the legal system. Our experience of legal coherence is dynamic rather than static; it changes as we engage in cognitive work to understand legal doctrines and as we encounter new information and new experiences.
Because judgments of coherence and incoherence rest upon the nature of the self, they are also shaped by the self's psychological needs--in particular, by our need to make sense of the world and to see ourselves as rational, well-meaning individuals. Understanding is always shaped by the self's need to preserve belief in its own coherence. Judgments of legal coherence and incoherence are affected by this need. In particular, the urge to reduce cognitive dissonance among our beliefs, actions, and commitments can cause us to see the law as coherent or incoherent. An approach that views coherence as a property of an object unaffected by our subjectivity will wholly miss these effects, for they will already have been projected onto the object of study. Thus we must recognize the contributions of subjectivity to understand how judgments of legal coherence and incoherence become a method of dissonance reduction for lawyers, legal academics, and even professors of jurisprudence.
Third, subjectivity is important to the study of law because legal understanding is also a source of power over the legal subject. In other words, if we want to understand legal understanding, we must recognize not only the effects that our understanding has on the objects we construct, but the effects that the act of understanding has on us. This feedback effect on the subject becomes invisible if we focus solely on the legal object. Jurisprudence usually envisions power only in political terms; it misses the power over the subject that arises from the act of understanding itself. Nevertheless, the process of making the law coherent, of rationally reconstructing it for ourselves, does not leave us untouched. Legal understanding is something that happens to us and changes us. It is a type of receptivity, of vulnerability, which affects us as much as it affects the law we attempt to understand. Legal understanding thus makes the legal subject a locus of ideological power, and the study of jurisprudence is and must properly be a study of this power. Yet the study of legal understanding must also be a humanist endeavor, in which we confront our human capacity for change--for better and for worse--through acts of understanding.
II. Legal Coherence from the Standpoint of the Legal Object
However, before we can begin our discussion of the specific form of understanding involved in judgments of legal coherence, we must first distinguish several different types of coherence. The first is the coherence of a set of factual beliefs, the second is the coherence of a normative system like the law, and the third is the coherence of the world around us. The coherence of factual beliefs is a question of logical or narrative coherence, while the coherence of the legal system is a question of normative coherence. The coherence of the world is prior to both: it is a feature of our existence as understanding beings and rests on our need to see ourselves as rational, morally sensitive people.
A collection of factual beliefs is logically coherent if the beliefs are logically consistent. A more stringent requirement of factual coherence is that the beliefs must be mutually supportive. An example of two mutually supportive beliefs is the belief that John owns a tennis racket and that John likes playing tennis; believing one tends to make the other belief more plausible and vice versa. However, when speaking of factual coherence, I shall, unless otherwise stated, employ the weaker requirement of logical coherence. A set of beliefs can be logically coherent even if they are not mutually supportive, as long as they do not involve a contradiction. In this sense, there is nothing incoherent (at least on the face of things) about believing both that John likes playing tennis and that Sarah is forty-three years old.
Sometimes we use "coherent" in a different and stronger sense to indicate narrative unity or unity of purpose; for example, an account of an event that introduced all sorts of irrelevant (but noncontradictory) facts might be said to lack coherence. This is a question not of logical but of narrative coherence.(15) Thus, although a set of beliefs can be logically coherent in that they are noncontradictory, one can articulate them in a narratively incoherent fashion, so that the audience is confused or does not get the point of the account.
Normative coherence plays a considerable role in my argument; I shall describe it in greater detail momentarily. For now, we can say that normative coherence is the consistency not of logic but of principle. Values and normative judgments are normatively coherent if they employ distinctions and similarities that are principled and reasonable as opposed to those which are arbitrary and unreasonable. Judgments of normative coherence employ a substantive rather than a formal conception of reason. A substantive conception of reason is roughly captured by the statement that someone is acting "reasonably." A formal conception of reason is exemplified by first-order predicate logic. The more formal a conception of reason is, the more indisputable its conclusions but the less relevant they are to practical judgment; the more substantive a conception of reason, the more relevant it is to practical judgment, but the more disputable its conclusions.(16) Because normative coherence relies on a strongly substantive conception of reason, judgments of normative coherence are often contested and contestable.
The coherence of the world around us is special. It can include judgments of logical, narrative, and normative coherence, but it is of an entirely different order. The coherence of the world around us is demanded by our existence as understanding beings. Thus, it is not simply a feature of a particular set of beliefs; it is a feature of our beliefs. It is a goal of the activity of understanding, a global task in which we always engage.
There is a considerable difference between the coherence of a set of beliefs considered abstractly and the coherence of our own beliefs. We have a stake in the latter that we might not have in the former because we have a stake in our belief in ourselves as coherent, rational individuals. Considered in the abstract, the logical, narrative, or normative coherence of a set of beliefs has no personal implications for us. However, when the coherence of our own beliefs is called into question, the inquiry assumes an existential dimension: it becomes a question not about possible relations between abstract propositions, but about ourselves.
Thus, the coherence of the world is the coherence of our own beliefs about the world. If the world is not coherent to us, we must try to make it coherent in order to preserve our own sense of order about ourselves. This task may, in turn, require local judgments of logical, narrative, or normative incoherence. To make sense of the world as a whole, we may have to assume that a particular set of beliefs held by others is self-contradictory, that a particular person is acting arbitrarily, that particular legal doctrines are normatively incoherent, and so on. Hence, accounts of coherence in the social world, or in the actions and beliefs of others, are, at bottom, driven by our need to believe that our own beliefs are ordered, coherent, and rational.(17)
Stricter requirements of normative coherence are possible. For example, Joseph Raz has suggested that coherence requires unity of principle, so that the legal system would be most coherent if we could imagine it as emanating from a single principle.(20) This is a very strong requirement of coherence, and one unlikely to be encountered in any significant part of the law. An alternative view, also suggested by Raz, would require the mutual interdependence of justificatory principles and policies. Just as our beliefs sometimes mutually support each other, the legal system would be coherent if it could be justified by a set of mutually interdependent principles and policies that stood or fell together.(21) Although this requirement is weaker than the first, it is still too strict. It would be remarkable if the many different principles, policies, and purposes underlying the law stood or fell together. The justifications underlying the law are less interdependent than mutually competitive; our hope for a coherent legal system is that the law employs and arbitrates among them in a consistent and principled manner.
Neither of these stricter requirements captures the meaning of legal coherence.(22) Rather, coherence is the principled consistency of the principles and policies underlying the law. The law is coherent if we can view it as emanating from or as explainable by a set of consistent principles and policies.(23) These principles and policies do not have to stem from a single master principle or policy, nor do they have to be mutually interdependent so that if one goes they all go. Nevertheless, they must be mutually supportive to the extent that they are consistent with each other, and any potential conflicts among them must be resolved in a consistent and principled manner.
Note that judgments of coherence can be local or global; they can be about individual doctrines, entire bodies of law, or the legal system as a whole.(24) We can say that the law of owner-occupier liability is coherent but not the law of respondeat superior, for example. If a part of the law is incoherent, then it may make a larger part of the law that includes it incoherent as well, because it produces unresolved conflicts with other parts. Moreover, even if two parts of the law seem coherent in isolation, they might be normatively incoherent when taken together. It is therefore possible that one could see the law as being normatively coherent locally everywhere and yet it would not necessarily follow that the law as a whole was coherent.
(1) The law (or some part of the law) is coherent if we can explain it by a set of consistent principles and policies which, if they were justified, would justify the content of legal doctrine (which may include constitutional, statutory, administrative, and decisional law).
(2) The law (or some part of the law) is coherent if we can explain it by a set of consistent and justified principles and policies which, taken together, justify the content of legal doctrine.(25)
These two positions differ in an important respect. The first says that the law is coherent if there is a set of consistent principles and policies that explains legal doctrine, even if we do not agree with these principles and policies. The second says that the principles and policies that explain existing law must themselves be morally justified. Let us call the first the requirement of hypothetical justification and the second the requirement of actual justification. Obviously, actual justification is a more stringent test of coherence.
A test of actual justification has important disadvantages. Most people who defend the coherence of the law against critical attack would not assume that belief in legal coherence commits them to the view that the law is fully morally justified or that existing legal doctrines cannot be to some extent unjust or oppressive. The question of legal coherence seems to be about something more (or rather something less) than the question of justice. To be sure, one of the reasons we are concerned with legal coherence is that we believe that it has effects on justice. However, this does not mean that coherence and justice amount to the same thing. Procedural fairness also contributes to justice, but it is not the same thing as justice; if substantively evil laws are applied with procedural fairness, they can still lead to very unjust results. In the same way, the requirement of coherence seems to be less demanding than full moral justification.
In contrast, the test of hypothetical justification does preserve the distinction between a coherent legal system and an ultimately just one. It requires only a consistent set of moral principles which, if they were justified, would justify existing legal doctrine. Nevertheless, this theory of legal coherence has complementary difficulties. If the test of actual justification is too stringent, the test of hypothetical justification is too forgiving; it makes the standard of coherence too easy to meet. Because the norms that justify the law do not themselves have to be justified, one could simply make up a series of principles and policies that match the categories and distinctions already present in the law. Thus, if the law denies recovery for economic loss in tort, it can be justified by the moral principle that economic loss should not be recovered in tort. Like a map describing a country on a 1:1 scale, which would be so accurate as to be useless, a coherence account that exactly matches the content of existing doctrine does no justificatory work. It gives us no explanation for why coherence (in the sense of hypothetical justification) is an important or valuable feature for a legal system to have; at the very least it does not serve the goal of justifying the coercive power of the state through law.(26) The test of hypothetical justification proves to be no test at all, for there will almost always be some set of logically consistent principles and policies, no matter how substantively arbitrary or insane, that explains existing legal doctrine.
We thus find ourselves on the horns of a dilemma. A theory of actual justification too readily collapses questions of coherence into those of moral justification and thus guarantees that almost any existing legal system will lack coherence. On the other hand, a principle of hypothetical justification, by divorcing questions of coherence from those of moral justification, seems to do no important justificatory work at all; it threatens to make the issue of coherence tautological.
Considered as a property of a legal object, legal coherence must lie somewhere between the minimal requirements of hypothetical justification and the maximal requirements of actual justification. If any normative principle could serve as a moral principle for purposes of hypothetical justification, hypothetical justification would impose no restraints on the content of legal norms. Hence, there must be some limits on what qualifies as a justificatory principle or policy in an account of legal coherence.(27) Legal coherence must be hypothetical justification according to a particular type of moral principle or policy. For purposes of this discussion, let us call these "bona fide principles and policies." Then we might say that a legal system was coherent if it were hypothetically justified by bona fide principles and policies, and if all conflicts among these principles and policies were in turn resolved through consistent application of bona fide principles and policies.
Merely to state the matter in this way suggests the inherent contestability of questions of legal coherence. Thus, people will disagree over whether a certain principle or policy is bona fide; they will disagree about the proper extension or scope of a bona fide principle or policy; and lastly, they will disagree about whether conflicts between competing policies and principles have themselves been resolved in a principled manner--that is, according to bona fide principles and policies. Indeed, the latter may be the most important form of disagreement in practice. Although people often can agree that certain abstract principles and policies are important, they will often disagree about the consistency of their application and the principled resolution of conflicts between competing principles.
In these disagreements and debates, questions of actual justification necessarily resurface. How does one decide that a policy or principle can form part of a scheme of hypothetical justification or that the balance between competing principles or policies has been properly struck? Although legal coherence does not require that the principles underlying the law be actually justified, questions of moral justification must play an important role in determining the limits of what can constitute a bona fide principle or policy, the proper scope of such principles and policies, and how conflicts among them are to be resolved. Legal coherence always implicates questions of actual justification even though it is not identical with actual justification.(28)
The relationship between bona fide principles and actual justification explains why the test of legal coherence must be more than mere intelligibility or logical consistency, standards that apply to the coherence of a set of factual beliefs or propositions. Factual beliefs are coherent if they are logically consistent or "hang together." The coherence of normative justification differs because of the nature of the subject matter that has to "hang together." There are no formal constraints on the kinds of factual beliefs that can form a coherent set, so long as they are logically consistent. However, not everything can count as a bona fide principle or policy; hence, mere intelligibility or logical consistency cannot be the test of coherence in law. If we cannot see the law as the product of bona fide moral principles and policies, we do not even reach the question of consistency between them.
In arguing that legal coherence is necessarily dependent on questions of actual justification, I am not claiming that questions of coherence must ultimately collapse into those of actual justification so that a legal system cannot be coherent unless it is actually justified. Rather, I am making two interrelated points. First, actual justification is essential to legal coherence because it polices the boundaries of the kinds of principles and policies that "hang together" in a coherent scheme of legal justification. Second, one can only attack the normative coherence of the legal system through arguments about actual justification. We demonstrate legal incoherence by demonstrating the moral arbitrariness and unreasonableness of the justifications offered for legal doctrines. That is why debates about legal coherence are inevitably linked to debates about actual justification.
Nevertheless, although lack of actual justification is a necessary condition for rejecting a proposed justification of legal doctrine, it is not a sufficient condition. What is a sufficient condition? It is the test of a bona fide principle or policy. It is the question of whether or not a principle or policy is so arbitrary or unjust that we must exclude it from a scheme of hypothetical justification. However, there can be no definitive answer to this question, and therefore the question of legal coherence (defined as hypothetical justification by bona fide principles and policies) is essentially contested and contestable. Thus, instead of a unified theory of legal coherence, we have an endless dialectic between the principles of hypothetical and actual justification in which one side defends the consistency of legal doctrine while the other attacks it on the grounds that it is unjust, unreasonable, and arbitrary. Because of the dialectic of hypothetical and actual justification, it is always possible to attack a scheme of hypothetical justification on the grounds that it employs principles or policies that are arbitrary or unjust, or that it arbitrates between competing principles or policies in an arbitrary or unjust fashion. This attack may be unsuccessful or unpersuasive in a particular case; however, it cannot be rejected merely on the grounds that legal doctrines do not need to be just to be coherent. Rather, one must respond that even though legal doctrines are not ultimately just, they are reasonable. Yet the grounds for this judgment of reasonableness depend heavily on beliefs about actual justification. As a result, people may disagree about the coherence of the legal system because of their different views about justice, morality, and politics, even though the test of legal coherence is not one of actual justification.
My thanks to Bruce Ackerman, Akhil Amar, Sandy Levinson, Elizabeth Mertz, Dennis Patterson, Pierre Schlag, and Tom Seung for their comments on previous drafts, as well as to the participants at workshops at Yale Law School, the American Bar Foundation, University of Connecticut School of Law, and Western New England College of Law, where earlier versions of this Article were presented.
1A "subject" is a person who understands something; that something is the "object" of her understanding. Hence a "legal subject" is a person who attempts to understand the law, legal doctrine, and the legal system; these, in turn, are the "legal objects" she apprehends. Concern with the legal subject is thus a concern with how our processes of understanding affect and help constitute the cultural objects we comprehend. One might also use the term "legal subject" to describe how law or legal culture constructs how we think about people--how law and legal culture ascribe particular identities and features to people, defining some characteristics as salient and others as irrelevant. The "legal subject," in this sense, is a subject as seen (and dealt with) through the eyes of the law or legal culture.
2Pierre Schlag has been at the forefront of this trend. In various ways, each of his writings has asked us to reflect on who does the thinking about law. See, e.g., Pierre Schlag, Fish v. Zapp: The Case of the Relatively Autonomous Self, 76 Geo. L.J. 37 (1987); Pierre Schlag, Normative and Nowhere to Go, 43 Stan. L. Rev. 167 (1990); Pierre Schlag, Normativity and the Politics of Form, 139 U. Pa. L. Rev. 801 (1991) [hereinafter Schlag, Normativity and the Politics of Form]; Pierre Schlag, The Problem of the Subject, 69 Tex. L. Rev. 1627 (1991) [hereinafter Schlag, The Problem of the Subject].
Other examples of the trend towards the study of the subject in recent jurisprudence include Stanley Fish, Doing What Comes Naturally: Change, Rhetoric and the Practice of Theory in Literary and Legal Studies (1990); James Boyle, Is Subjectivity Possible? The Postmodern Subject in Legal Theory, 62 U. Colo. L. Rev. 489 (1991); Rosemary J. Coombe, "Same As It Ever Was": Rethinking the Politics of Legal Interpretation, 34 McGill L.J. 603 (1989); Drucilla L. Cornell, Institutionalization of Meaning, Recollective Imagination and the Potential for Transformative Legal Interpretation, 136 U. Pa. L. Rev. 1135 (1988); Meir Dan-Cohen, Responsibility and the Boundaries of the Self, 105 Harv. L. Rev. 959 (1992); Steven L. Winter, Indeterminacy and Incommensurability in Constitutional Law, 78 Cal. L. Rev. 1441 (1990).
Critical Race Theory and Feminist literature raises the problem of subjectivity when it seeks to discover how perspectives of dominant groups are projected onto the social world so that they appear to be intrinsic features of that world. E.g., Martha Minow, Making All the Difference (1990); Patricia J. Williams, The Alchemy of Race and Rights (1991). Other examples of this scholarship argue that the subject's multiple allegiances resulting from her social situation create different and conflicting experiences of social reality, none of which can claim supremacy. See, e.g., Angela P. Harris, Race and Essentialism in Feminist Legal Theory, 42 Stan. L. Rev. 581 (1990).
3See Pierre Schlag, Contradiction and Denial, 87 Mich. L. Rev. 1216 (1989) (reviewing Mark Kelman, A Guide to Critical Legal Studies (1987)). This brief but insightful essay anticipates many of the themes of the present Article.
5For a discussion of ideological projection, see J.M. Balkin, The Mechanisms of Ideology: Bricolage and the Construction of Cultural Software (unpublished manuscript, on file with author); Schlag, The Problem of the Subject, supra note 2, at 1636; Schlag, supra note 3, at 1218; cf. Williams, supra note 2, at 61 (attributions of criminality projected away from whites onto blacks, making white criminality invisible); Christine A. Littleton, Reconstructing Sexual Equality, 75 Cal. L. Rev. 1279, 1310 (1987) (patriarchy locates difference in women rather than in social situations); Frances E. Olsen, The Family and the Market: A Study of Ideology and Legal Reform, 96 Harv. L. Rev. 1497, 1570 (1983) ("The `mystery' of sexuality consists in projecting human qualities separately onto males and females to make each the object of the other.").
10According to Dworkin, "skepticism brought to an enterprise from the outside . . . which engages no arguments of the sort the enterprise requires . . . can make no difference to our own efforts to understand and improve interpretation, art, and law." Ronald Dworkin, A Matter of Principle 177 (1985). Dworkin contends that this sort of "external" skepticism is either irrelevant or impossible. Id. at 176. Only "internal" skepticism is relevant to describing legal practice; this necessarily involves "arguments of the same contested character as the arguments it opposes." Dworkin, supra note 8, at 85-86.
11Dworkin, supra note 8, at 85-86 (despite lawyers' demand for some form of objectivity beyond legal practice, they inevitably "return to their knitting--making, accepting, resisting, rejecting arguments in the normal way"; the "preliminary dance of skepticism" before returning to arguments within legal practice "is silly and wasteful"); see also Dworkin, supra note 10, at 171 ("I have no arguments for the objectivity of moral judgments except moral arguments, no arguments for the objectivity of interpretive judgments except interpretive arguments, and so forth.").
12Dworkin, supra note 8 at 86; see also Dworkin, supra note 10, at 174 ("I have yet been given no reason to think that any skeptical argument about morality can be other than a moral argument, or skeptical argument about law other than a legal argument, or skeptical argument about interpretation other than an interpretive argument."). Skeptical questions about a practice like law are always, in Dworkin's view, questions of interpretation which must be raised and posed within that practice. Id.
13H.L.A. Hart, The Concept of Law 86 (1961) [hereinafter Hart, The Concept of Law]. Hart distinguished the internal perspective from an "external perspective," the point of view of an observer who sees legal rules not as norms that can be correctly or incorrectly followed but simply as producing regularities of behavior. Id. at 87. Other scholars have noted that one does not have to feel bound by legal rules to understand them as norms, to grasp their point as directives, or to argue about what they require. See Neil MacCormick, Legal Reasoning and Legal Theory 291 (1978); Joseph Raz, The Authority of Law 153-54 (1979); Joseph Raz, Practical Reason and Norms, 175-77 (1975). Hart has subsequently acknowledged that this constitutes a third, "hermeneutic" perspective. H.L.A. Hart, Essays in Jurisprudence and Philosophy 14 (1983). Notwithstanding these qualifications, Joseph Raz has noted the general consensus among jurisprudential scholars of the centrality of the internal perspective to accounts of law. See Joseph Raz, The Relevance of Coherence, 72 B.U. L. Rev. 273, 292 (1992) [hereinafter Raz, The Relevance of Coherence].
16A substantive conception of reason is also different from (and richer than) a purely instrumental conception of reason. Whereas the latter arbitrates only over the best means of achieving preexisting ends, the former arbitrates over both means and ends.
18Raz points out that discussions of legal coherence necessarily assume a "base" of legal materials--legal decisions, statutes, rules, and regulations--that must be explained by various principles, purposes, and policies. Raz, The Relevance of Coherence, supra note 13, at 284-85. This base is not completely fixed because the principles, purposes, and policies that explain the base may have a feedback effect that alters our judgment of what is properly considered to be in the base. Id. at 285-86. For example, the question of the proper extent of a particular doctrine or the proper interpretation of a statute or administrative regulation may be affected by our view of what principles, purposes, and policies best justify them. However, the base to be explained must exclude some things because otherwise it would not be clear that it was the law rather than something else whose normative coherence was at stake. Id. at 285.
19Cf. MacCormick, supra note 13, at 106 ("The basic idea is of the legal system as a consistent and coherent body of norms whose observance secures certain valued goals which can intelligibly be pursued all together."). MacCormick's view, like my own, is driven by conceptions of reason and reasonableness that are essentially substantive and practical, but for this reason are also inevitably contested. Id. at 266-74.
20See Raz, The Relevance of Coherence, supra note 13, at 286. Raz calls justification by a single principle "strong monistic coherence." Id. Although he offers alternative definitions of coherence, his basic criterion is that the more unified the principles which explain the law, the more coherent the law is. Id.
22Because the law seems to contain many different principles, policies, and purposes, it is not very difficult for Raz to conclude that the law is not coherent under his definition of coherence. See id. at 295, 310-14. Indeed, because his criteria are so exacting, he concludes that Dworkin's theory of "law as integrity" in Law's Empire is not really a coherence theory at all. See id. at 315-17 (noting that although Dworkin uses the word coherence, his theory of integrity does not require monistic coherence or mutually dependent principles). Although I find Raz's definition too restrictive, much that he says about legal coherence in his sense can be applied to broader conceptions of legal coherence, see id. at 296-97, and I shall try to note the relevance of his arguments to mine as the occasion arises.
23Dworkin distinguishes between principles and policies. See Ronald Dworkin, Taking Rights Seriously 90-100, 113-15, 294-327 (7th prtg. 1980) [hereinafter Dworkin, Taking Rights Seriously]; Dworkin, supra note 8, at 221-24, 243-44. In his view, the former, but not the latter, are primarily relevant to determining the coherence of the legal system, despite the fact that policies play some role in proper interpretation of statutes. See Dworkin, Taking Rights Seriously, at 107-10; Dworkin, supra note 8, at 313 n.1, 338-39. In contrast, I shall assume that a defender of legal coherence may employ both in offering a coherent account of legal materials. Cf. MacCormick, supra note 13, at 259-64 (rejecting distinction for purposes of determinations of legal coherence).
24See Barbara B. Levenbook, The Role of Coherence in Legal Reasoning, 3 Law & Phil. 355, 371 (1984) (arguing that law is only locally coherent); Raz, The Relevance of Coherence, supra note 13, at 311-14 (same).
25Cf. Raz, The Relevance of Coherence, supra note 13, at 290 (modifications of law as coherence thesis). Note that the "consistency" referred to in both definitions is the consistency of principle. Principles and policies must be applied in a consistent manner and all conflicts among competing principles and policies must be resolved in a principled manner. Even if one could articulate a set of principles and policies that explained existing law, the law would not be coherent if these principles and policies were not applied--and potential conflicts between them were not resolved--in a consistent, principled fashion. As a result, in both definitions, the relevant set of principles and policies must include principles and policies used to arbitrate between competing principles and policies and to ensure the consistency of their application.
28Consider, for example, the following two principles, taken from a discussion of tort law in Dworkin's Law's Empire: (1) "No one has a moral right to compensation except for physical injury," and (2) "People have a moral right to compensation for emotional injury suffered at the scene of an accident against anyone whose carelessness caused the accident but have no right to compensation for emotional injury suffered later." Dworkin, supra note 8, at 240. Dworkin easily rejects the first as a justificatory principle because it does not fit existing cases. Id. at 242. However, he rejects the second on the grounds that "it does not state a principle of justice at all." Id. Stated in terms of our previous discussion, Dworkin is saying that the second statement cannot serve as a principle for purposes of hypothetical justification. A system of norms that included it would fail the test of hypothetical justification and would not establish the coherence of the legal system.
Nevertheless, one might ask why the first principle could possibly constitute a principle of justice while the second cannot. My view is that the exclusion of such principles from a coherent scheme of legal justification cannot be on grounds of logical form. It must be on the grounds that the distinctions and similarities drawn by these principles are morally arbitrary. See id. (arguing that the second principle "draws a line that it leaves arbitrary and unconnected to any more general moral or political consideration."). In other words, it must be on the grounds that Dworkin believes these principles fail the test of actual justification.
A similar point applies to resolution of conflicts between principles or policies. There are many ways that we can arbitrate between various principles and policies in a logically consistent manner, and not all of them deserve the title of coherent reconciliations. For example, consider two principles in tort law, the principle that liability should be proportionate to fault, and the principle that persons who have been harmed by the actions of others deserve compensation. Sometimes these principles point in the same direction, but sometimes they conflict, and then we must choose between them. Suppose that we decide that we will resolve the conflict in favor of the fault principle when the plaintiff is less than six feet tall, and in favor of the compensation principle when the plaintiff is six feet or taller. Although there is nothing logically inconsistent about this resolution of the conflict, it seems morally arbitrary. Put another way, conflicts among principles or policies must be resolved by principle, and not by logically consistent but morally arbitrary distinctions and assignments. In deciding how to reconcile competing principles or policies in legal doctrine, we expect that our reconciliation will itself be principled. Thus, when we complain about the reconciliation of two competing principles, we are not complaining about a logical contradiction between them, but about a failure of adequate moral justification for the distinctions drawn and the balances struck.