I.

I am not now, nor ever was, a member of the Critical Legal Studies (CLS) movement. But I want to consider the adequacy of the answer Ronald Dworkin gave almost twenty years ago to an objection (to his conception of law as integrity), which he attributed to CLS though he himself referred to it in more general terms as “internal skepticism.”¹ I want to consider the adequacy of his response, not out of any desire to revive the intellectual fortunes of CLS and certainly not as a complaint that Professor Dworkin failed to do justice to CLS criticisms during the brief lifetime of the movement. CLS was never particularly interested in Dworkin, nor he in them.² But they had a common interest in what I would like to call “the background elements” of a legal system—the principles and policies that lie in back of the rules and texts that positivists emphasize. Ronald Dworkin and Duncan Kennedy, for example, are both theorists of the legal background. But what they say about it is very different. Dworkin thinks recourse to the background affords the resources necessary for legal decision in cases where the foreground law is disputed or indeterminate. Kennedy thinks the background is so riven with contradiction as to be capable of offering spurious support for everything and determinate support for nothing in legal reasoning.³ So I am interested in what one might think of as Dworkin's descriptive optimism—his view that the background elements (of a legal system like that of the United States) are capable of bearing the weight of determinate argument that he wants to assign to it in his theory of law.

(p.156) I will argue that, in order to answer the CLS critique, a defender of law as integrity has to place considerable emphasis on what Professor Dworkin regards as the constructive side of his argument. Instead of saying that the legal background is coherent, Dworkin has to say that it is capable of being made coherent at the hands of a sufficiently resourceful interpreter. Unfortunately the response cannot rest there. Dworkin's constructivism might licence an ingenious and versatile manipulation of existing legal materials: we take the parts of the (often contradictory background) that we like or approve, and we use them to add some sort of doctrinal credibility to the positions we are aiming at. But it is not clear that this fits the justification that Dworkin furnished for his method in Law's Empire, for that justification requires a party to argue, not merely that there are legal materials which he can make use of in support of his brief, but that he is using background materials in a way that keeps faith with the network of mutual commitments that makes us the community we are.⁴ Unless the Dworkinian advocate can make such an argument, he has nothing with which he can resist the urgings of a pragmatist, to the effect that we should not bother with the background material at all, once the foreground has been shown to be indeterminate, but we should make our case directly for whatever is likely to promote the social good. To rebut that pragmatist position—which Dworkin acknowledges has to be rebutted to make room for his conception of law as integrity⁵—it is not enough to show that we can make something coherent and attractive out of the legal background. The Dworkinian advocate has to show that we must attempt to make something coherent and attractive out of the legal background. And what counts as a good faith attempt has to be responsive to the ideas lying behind that “must.” My worry is that the constructivism that Dworkin appeals to, in order to evade the CLS critique, succeeds (if it does) only at the cost of cutting loose of the idea that the community is in some sense already committed to a coherent and principled position which the Dworkinian lawyer has a responsibility to unearth.

So: my aim in this paper is to explore the extent to which Professor Dworkin is put to a hard choice between the agile and discerning constructivism he needs to respond to CLS, on the one hand, and the integrity thesis about the commitments of the community which he invokes to justify the claim that making coherent sense of the existing legal materials, foreground and background, is something we are morally required to do. I think Dworkin really is confronted with a dilemma here, and I am not sure that a way through it can be negotiated. If he hangs on to the integrity position, he makes it harder to respond to the skepticism of CLS. But if he weakens the integrity requirement or loosens its connection to what is to count as an appropriate mode of legal argument, then he leaves himself defenseless against the pragmatist position that—clever though it is—the constructivism of Dworkinian legal argument serves no useful purpose.

(p.157) As I said in my opening paragraph, I am not exploring Dworkin's response to internal skepticism in order to revive the moribund fortunes of CLS. The Crits can look after themselves.⁶ I am exploring their critique and his response because they highlight a connection between two different aspects of Dworkin's jurisprudence, usually dealt with and criticized—if they are criticized—separately. On the one hand, there is the question of whether his constructive or interpretive method actually delivers the goods for judicial reasoning. On the other hand, there is a philosophical question about the ultimate justification for his insistence that we delve into the legal background rather than give up and resort to an overtly pragmatic approach when foreground law runs out or reveals itself to be indeterminate. The theory of integrity is supposed to give us our answer to the second question: we must persist with the background in order to keep faith with the commitments of our community. But what if the method for delivering the goods in hard cases—creative constructive theory-building—turns out not to be a way of keeping faith with our community? What if it makes a mockery of the idea of communal commitments, which law as integrity bows to? Then the two parts of Dworkin's jurisprudence—the clever constructivism and the portentous theory of integrity—sail past each other, and it will be hard to hang on to a sense of the whole. We need to take this prospect very seriously; and if there is an adequate Dworkinian response to the CLS challenge we must ensure that it does not fall at this hurdle.

II.

The key to Dworkin's jurisprudence is and has always been a view about what law involves besides a heap of enacted rules (constitutions, codes, and statutes).⁷ Dworkin argues that as well as the enacted rules, which we see in the foreground of legal analysis, a working legal system also comprises background policies and principles, either those associated with particular enactments—as when we say the policy of the Sherman Act is to foster competition—or those that pervade the body of the law as a whole—as when we talk about “the well established policy of English law of imposing a more extensive liability on intentional wrongdoers than on careless defendants,“⁸ or as when we invoke the maxim that “[n]o one shall be permitted to profit by his own fraud.”⁹ Dworkin believes that it is appropriate to (p.158) invoke these principles and policies in legal argument and indeed that it is inappropriate not to. Not only are they important legal resources to be put to use in hard cases where the foreground law is indeterminate, but they are also capable of standing against foreground law (as the principle last-mentioned above stood against the New York state legislation relating to wills in Riggs v. Palmer).¹⁰ Indeed a case may be made that background principles and policies have to be assumed in order to explain why the foreground law has the force it does.¹¹ Certainly any jurisprudence that neglects these background elements is inadequate.

There is a lot in Professor Dworkin's account of the legal background that I do not want to dispute. In various places, he distinguishes between policies and principles, as different types of background element: he associates “policy” with social goals and “principle” with norms or reasons that command a particular distribution of benefits, freedoms, and responsibilities.¹² I have no quarrel with that, nor with the arguments that Dworkin rests on that distinction. I accept Dworkin's view that background principles (in the narrower sense) are capable of sustaining claims of right.¹³ I also see little to quarrel with in what he says about the logical character of the contribution that principles and policies make to legal argument: he talks of the dimension of “weight,” which distinguishes their contribution from the contribution made by rules.¹⁴ Finally, I have no quarrel with what Professor Dworkin says about the kind of presence that principles and policies have in a legal system. He argues—convincingly in my view—that their status as law “lies not in a particular decision of some legislature or court, but in a sense of appropriateness developed in the profession and the public over time.”¹⁵ (I think, too, that he is right to point up the difficulty this poses for H.L.A. Hart's concept of the rule of recognition.) None of these matters is in dispute between Dworkin and the Crits. In fact, although CLS is certainly less optimistic than Professor Dworkin is about the contribution that these background elements can make to determinate legal reasoning, their criticisms nevertheless presuppose that the legal background is at least as important for jurisprudence as Dworkin claims it is. If anything they are less sympathetic (p.159) than Dworkin is to the positivist conception that he is attacking,¹⁶ for they believe that a focus on purely foreground elements may radically underestimate various forms of pathology endemic to (say) the modern American legal system.

In his early work, Dworkin imagined that the background elements of a legal system might present themselves as discrete norms, which could be identified and individuated by judges and advocates and spoken about almost in the way that enacted rules are spoken about. Thus we might talk about the principle that no man is to profit from his own fraud, and distinguish it from other principles such as the principle that statutes are not to be given retroactive effect (if a non-retroactive interpretation is possible) or the principle that intentional wrongdoing is worse than equally harmful negligent wrongdoing. Even though background principles are not enacted as texts and even though they have no canonical formulation, still we do sometimes give a principle a verbal formulation and make it the focus of legal argumentation, asking how it should be interpreted, and drawing out its implications for particular cases. Judges and advocates do this all the time. They make assertions about what the general policy of the law is or what principles pervade it; then they go on to use these assertions as premises for arguing about the disposition of the cases in front of them. This gives the impression of a sort of two-step process involved in the use of background elements in legal argument, analogous to the two-step process that is typically used in arguing about textual rules. In step (1), we identify the norm in question—by the recognition and citation of a validly enacted text (in the case of a rule) or by the more diffuse argument used to discern the presence of a policy or principle in the law. Then, in step (2), we argue for a particular disposition of the case in front of us, on the basis of the bearing which we think the rule we have cited or the principle or policy we have identified has on the facts of the case. This is a perfectly lucid model and, as I have said, it is characteristic of much of what lawyers and judges say about principles and policies.

Its disadvantage is the way it assimilates principles and policies to more discrete forms of legal provision, such as statutory rules, as though the identification, individuation, and citation of the former were in the end not much different from the identification, individuation, and citation of the latter.¹⁷ In fact, there are huge differences. It's not just that rules characteristically have a canonical formulation which principles and policies lack; it is also that rules have (as it were) a fixed level of generality, whereas the tendency of principles and policies is always to expand the sphere of their application, something which has to do with their operating more like reasons than like enacted provisions. If there is any limit on the generality of a policy or principle, it is a matter of the esthetics or lucidity of its application to the (p.160) case with which we are primarily concerned. We state the principle so that its application to the case at hand is patent, rather than a matter of further interpretation. This suggests that the distinction between steps (1) and (2) is somewhat misleading, and that we might do better to think more directly about the relation between the legal background and the instant case.

And that is the model suggested in some parts of Law's Empire. The idea there is that in any given case, a judgment for the plaintiff, say, is to be defended by showing that it coheres better with existing legal materials than a judgment for the defendant would. “Coheres better with” is understood as what Professor Dworkin calls a post-interpretive claim. Counsel for the plaintiff offers a justificatory theory which he says makes sense of all or most of the existing legal materials—prior decisions, established doctrine, legislation, and so on—and he shows that that theory would also justify a decision in favor of his client.¹⁸ The defendant offers a contrary case, and counsel for the plaintiff will argue that the theory offered by the defense is either less attractive in itself than the one that he has presented or covers less of the existing materials than his does, or both. Of course, it is unlikely that either of the theories on offer will stand in a justificatory relation to all the existing legal materials. Some materials may be simply irrelevant—too far from the issues in question in terms of the concerns they might be thought to embody, even under the most abstract characterization. And some may have to be treated as outliers, relevant but incapable of being reconciled with the theory being put forward. Also, it is unlikely that either theory—plaintiff's or defendant's—will amount to (what anyone regards as) a perfect theory of justice. Even allowing for outliers, both theories will have to accommodate decisions (precedents, enactments) that their proponents might not support if they were setting out their favored account of justice on a blank slate. Still, plaintiff will argue that the theory the defendant invokes to make the existing law look coherent is further from an ideal theory of justice than the theory he (the plaintiff) invokes, and defendant will return the compliment. So they will argue back and forth on this dimension of justice as well as on the dimension of how much of the legal record their rival theories cover.

When we put the matter in this way, the distinction between background and foreground elements in a legal system—between principles and rules—is seen as a sort of façon de parler. What the existing law really amounts to is nothing more than a mass of decisions—by constitution framers, by legislators, and by thousands of earlier judges. Relying on background elements in current legal argument is not really a matter of introducing a different set of provisions, viz. legal principles as opposed to legal rules. It is rather a particular way of working with that existing mass of decisions, though it may also involve self-conscious reference to the recorded (p.161) efforts of previous decision-makers (earlier judges, for example), to work with the mass of previous decisions that confronted them in just the same sort of way.¹⁹

So there are these two ways of characterizing Dworkin's approach: (i) his earlier characterization in terms of norms called principles, i.e. norms that were comprised in the law just like rules—only in the background of the law rather than in the foreground; and (ii) his later approach which talked of rival theories put forward by those who were working with the existing law to justify a current decision. I don't really think they are different, or that any difference between the two really matters. A principle (or a policy) is nothing much more than a theory (or an element of a theory) about what makes sense of the existing law, expressed directly in normative terms.²⁰ We can say, in the voice of the theorist (or the current advocate) that the best interpretation of the existing law—the theory that covers most of it and makes it look good—is that it does not permit people to profit from their own fraud. Or we can “point to” or “cite” a principle—“No one is to profit from his own fraud”—more or less conscious of the fact that this in effect is something we read into the law when we commit ourselves to the theory just mentioned. The latter character-ization will seem particularly appropriate, when the mass of materials we are confronting includes holdings by other judges which interpret the materials that confronted them in just the way that we interpret the materials that confront us: those holdings will enable us to point to something like “authority” for the principles we are relying on.²¹ The alternative characterization will be more useful, however, in cases where we have no such authority, but are striking out on a new interpretive path, because we have a theory which we think better explains the existing law (and generates a result for our client) than previous efforts at interpretation do. Dworkin seems happy to move back and forth between these models in Law's Empire. Sometimes he says—in line with characterization (ii)—that a judge “must choose between eligible interpretations [theories] by asking which shows the community's structure of institutions and decisions—its public standards as a whole—in a better light.”²² And sometimes he says—in line with characterization (i)—that his method “asks judges to assume… that law is structured by a coherent set of principles.”²³ As I say, I don't think anything much turns on this distinction;²⁴ I mention it only in order to make it clear that I think the two amount more or less to the same thing, and that the critical points I am going to develop can be applied to Professor Dworkin's jurisprudence in both its phases.

(p.162) III.

Much of the CLS critique of Dworkin's work is superficial:²⁵ for example, CLS scholars charge that Dworkin's jurisprudence is “elitist,” that he is insufficiently concerned with “the deprived and disadvantaged in society,” that he is in denial about the connection between law and politics, that his approach to justice is too abstract, or that he “exhibits a profound antipathy for common consciousness among the citizens.”²⁶ Dworkin himself, however, has identified a line of CLS argument that poses a considerable threat to his enterprise.

The threat presents itself as follows. The law we confront in a modern democracy is not the work of a single author. We do not live in a one party state, nor do we even live in a unitary state. Law-makers of various persuasions coexist at various levels or in the various tributaries of our legal system, and contribute to the growth of the law from their diverse moral and political perspectives. Moreover, we live in a society whose law is the repository of its checkered political history. When one party succeeds another in electoral competition, it does not attempt to wipe the legal slate clean and begin all over again; instead the latest generation of statesmen work incrementally to make such modest changes in the law as their priorities dictate and their political circumstances permit, leaving the remainder—the work of the predecessors—to coexist comfortably or uncomfortably with whatever changes they have managed to make.²⁷ The law in fact is a patchwork of provisions, the work of a multitude of authors living and dead, with diverse and conflicting political agendas, and diverse and conflicting commitments of principle.

In the midst of this patchwork—in the midst of this mess, we might say—Dworkin's lawyer is supposed to be able to find sufficient coherence to assert credibly that the law is pervaded by principles which favor his client's case and to argue credibly against any similar but opposing claim made on behalf of his opponent. But why should we assume that this is possible? As Dworkin acknowledges, nothing in the way the law was produced guarantees that the lawyer or the judge will succeed in finding a coherent interpretation of it.²⁸ Or if he can find a coherent interpretation of the patchwork, nothing in the way that the law was produced guarantees that his success precludes similar success for his opponent: nothing guarantees that the law has a shape amenable to a unique coherent justification of principle.²⁹ Without that guarantee, or at any rate without a reasonable prospect (p.163) of a unique coherence result, the Dworkinian style of legal argument is just a recipe for the reproduction of indeterminacy.

The message of Critical Legal Studies is sometimes taken to be that law is systematically biased towards the interests of the ruling class. If that were so, then there might possibly be some hope for Dworkin's jurisprudence: the advocate or judge hoping to dredge up some principled coherence from the law might expect to find in it a more or less coherent version of ruling class ideology. But that has always been a superficial reading of CLS. The most thoughtful CLS scholars have emphasized the contradictory and conflict-ridden aspect of our law. Law, says Roberto Unger, “is the product of real collective conflict, carried on over a long time, among many different wills and imaginations, interests and visions.”³⁰ With this sort of provenance, any given body of legal doctrine is bound to be messy, rich in compromises, exceptions and contradictions.

Warring solutions to similar problems will coexist. Their boundaries of application will remain uncertain. Interests and ideals favored in some domains will be discounted in others for no better reason than the sequence in which certain decisive conflicts took place and the relative influence enjoyed by contending parties of opinion at the time. Intellectual fashions will join with preponderant interests to produce results that neither interests nor fashions alone would have allowed us to predict. Defeated or rejected solutions will remain, incongruously, in the corners of the law as vestiges of past approaches …³¹

One would have to be naively optimistic—to the point of some sort of Hegelian faith in the cunning of reason³²—to believe that these processes yield a single determinate structure of principled reasons.

Indeed the criticism may be drawn even more tightly than that. In any piece of high profile litigation, the parties will represent or say they are representing not only their own interests but competing visions of what law has to offer in society. But given our checkered political history, it is incredible that there will not already be layers in the archeological midden of law to support each of the visions propounded by the rival parties. Plaintiff may be able to find liberal strains in the law, corresponding to the times and the places where liberal principles have predomin-ated in the law-making processes; and defendant will be able to find conservative strains in the law corresponding to the times the law-making processes have been (p.164) wrested away from their liberal opponents. All that is there, waiting in the law for the rival parties to come along and raid in their latest forays as litigators.

Duncan Kennedy offers the tightest version of this view.³³ He believes the conflicting tendencies in the law can be resolved into two opposed visions of the individual and society—an individualist vision and an altruist vision—and that these two visions are represented haphazardly in the law that confronts us as the highwater marks of the ascendancy in law-making of first one tendency and then of the other tendency.³⁴ Hence, what is or appears to be at stake for a given individual engaged in litigation will usually present itself as one or other of these tendencies, and Professor Kennedy's position is that the body of law that any litigation draws upon, and that the parties look to for support for their rival positions, is also riven with the contradictory appearance of both positions. From the advocates’ point of view, the result is an uncomfortable awareness that there is enough in the law for his opponent to make an argument of principle every bit as convincing (or as unconvincing) as his own. And from the judge's point of view, there is the sense that his decision might go either way: the judge may be predisposed to approach a case in the individualist mode, but may find himself suddenly, as Kennedy puts it, “shifting modes,“³⁵ attracted suddenly by the lines of available background argument that might sustain legal argument for the side that represents the altruistic tendency. Either way, whether the judge shifts modes or not, or whatever the direction in which the shift takes place, he will find no difficulty in constructing an argument of exactly the sort, drawing on exactly the sort of elements, that Dworkin recommends. For Kennedy is also a student of the legal background: the only difference is that his more realistic scrutiny of it has not convinced him, as Dworkin is convinced, that it yields to only one compelling argument of principle in difficult cases.³⁶

Dworkin, as I said, acknowledges the importance of this objection. He calls it a form of internal skepticism³⁷—to be distinguished from external or philosophical (p.165) skepticism about the very ideas of value and coherence, which he thinks is a distraction in jurisprudence³⁸—and insists that it has to be taken seriously.³⁹ It is not out of the question, he acknowledges, that the existing law could be so riddled with inconsistency as to be unamenable to the sort of analysis he proposes. After all, we know the foreground law is inconsistent, if not in its direct normative implications,⁴⁰ then in the spirit that informs it—in the purposes, principles and policies that must be assumed to make sense of it. The Dworkinian judge “knows that legislative supremacy gives force to some statutes that are inconsistent in principle with others… But he assumes that these contradictions are not so pervasive and intractable… that his task is impossible.” The question now is: can that assumption be defended?

IV.

Dworkin's first pass at the objection is to deny that the extent of contradiction is as great as the CLS scholars make it out to be. “The literature of critical legal studies announces rather than defends these claims [about pervasive contradictions], as if they were self-evident.”⁴¹ Pierre Schlag thinks this is just summary dismissal of CLS criticism,⁴² but he is wrong and wrong, I suspect, for two reasons. He is wrong, first, because Dworkin does not rest here on a flat denial. Indeed, he goes on to mobilize other arguments, which I will consider in a moment. And Schlag is wrong, secondly, because there may be more to the factual disagreement about the extent of contradiction than meets the eye.

I shall return to the second point in Section VIII. But for the moment, I would like to explore the first point, the one that does not involve Dworkin simply disputing the extent of contradictions. After all, Dworkin is hardly in a position to deny the existence of something like contradiction or incoherence in the law. On the contrary, the background presupposition of his embrace of integrity is that the settled law of any community has had multiple authors—framers, legislators, judges, etc.—with quite radically differing visions of justice. (If there were no conflict between the various views about justice expressed in the enacted law, there would be nothing for the Dworkinian ideal of integrity to do.)⁴³ So any attempt to rebut internal skepticism simply by saying there are no contradictions, or by saying that whatever contradictions there are, are low-level and unimportant, (p.166) would undermine the point of his preoccupation with integrity. Of course, opinions about the extent and pervasiveness of such contradictions as there are in the law might vary. And in the end, that's what it may come down to. But my initial suspicion is that Dworkin has not wanted to rely simply on the sort of limp response that H.L.A. Hart gave to the critique of positivism put forward by the Legal Realists—“Well, it's not really as bad as all that.”⁴⁴ That would make his jurisprudence contingent and precarious, and its applicability to real-world legal systems perhaps unacceptably conditional.

V.

So what else does he say? Dworkin's second response is to argue that this sort of CLS skepticism neglects an important philosophical distinction between competing principles (such as autonomy and mutual concern) which may figure in a single view about justice, and contradictory principles (such as equality and inequality) which cannot possibly be combined in one coherent conception.⁴⁵ Referring to a tort law example he uses in Chapter 7 of Law's Empire, Dworkin says there is no contradiction in recognizing both a principle of collective solidarity and a principle of individual fault-based responsibility: “on the contrary, any moral vision would be defective if it wholly disowned either impulse.… No general interpretation that denied either one would be plausible; integrity could not be served if either were wholly disavowed.”⁴⁶ The challenge, he says, is to articulate a principled balance between them in the interpretive theory that we use to justify our particular legal conclusions.

Now, what Dworkin says here is right, as far as it goes. Most moral theories, including theories of justice, do have elements that stand in tension with one another; and it is the mark of moral maturity not to regard this as a logical defect, but rather to see it as an inevitable feature of our coming to terms with the plurality of values and principles.⁴⁷ But does pointing this out meet the objection? I fear it does not.

CLS scholars like Kennedy deny that the individualist and altruist elements they discern in the legal background are merely prima facie principles or the sort of (p.167) in-tension ingredients that one would find combined or balanced in a well-worked-out moral theory.⁴⁸ What they claim to find in the background of our law are not rival tendencies but incompatible visions or incompatible theories of what we owe to others and to the collective life of our society.⁴⁹ Each of these theories or visions already balances egotistical against other-regarding influence. So, for example, on Duncan Kennedy's account, the individualist element in private law is not just a vector, exerting an egoistic influence on the final social solution. It already represents a determination of how self-regarding and other-regarding elements might be balanced. Kennedy observes that “individualism is sharply distinct from pure egotism, or the view that it is impossible and undesirable to set any limits at all to the pursuit of self-interest,“⁵⁰ and it is distinct precisely in this point that it already represents an attempt—the individualist attempt—to balance selfish and other-regarding tendencies. Something similar, Kennedy says, is true of what he calls altruism:

[T]he altruist is unwilling to carry his premise of solidarity to the extreme of making everyone responsible for the welfare of everyone else. The altruist believes in the necessity and desirability of a sphere of autonomy or liberty or freedom or privacy within which one is free to ignore both the plights of others and the consequences of one's own acts for their welfare.⁵¹

Once again, the altruism one finds in the legal background is not just a solidaristic or communitarian tendency; it is already a view about how such a tendency should be balanced against the egotistical tendencies that compete with it. Each of the rival views about justice in society that Kennedy claims to discern in the background is already a view about how competing principles such as autonomy and mutual concern should be ranked and weighed within a single conception. In that sense, each does contradict the other's weighting and ranking.

Dworkin ought to have no difficulty seeing this. After all, the presupposition of his concern with integrity is that the authorship of enacted law reflects a diverse (p.168) array of views about justice, not merely a diverse array of impulses (some selfish, some communitarian) which would later have to be organized into a single theory of justice by the Dworkinian interpreter.⁵² It is not a case of one statute's having been passed by an autonomy faction and another statute's having been enacted by the party of mutual concern, and of its never having occurred to anyone except the judge (now faced with the two statutes together) that a suitably complex position could accommodate both principles.

I have not said anything so far about Kennedy's even more skeptical claim, that (whether we regard individualism and altruism as competing balances or as competing principles to be balanced) no metaprinciple of balance can be defended as objectively correct. Kennedy says:

[I]t is futile to imagine that moral and practical conflict will yield to analysis in terms of higher level concepts. The meaning of contradiction at the level of abstraction is that there is no metasystem that would, if only we could find it, key us into one mode or the other as circumstances “required.”… [W]e cannot “balance” individualist and altruist values… except in the tautological sense that we can, as a matter of fact, decide if we have to. The imagery of balancing presupposes exactly the kind of more abstract unit of measurement that the sense of contradiction excludes.⁵³

This claim is supposed to cut off an approach that the Dworkinian lawyer might take in the face of what we have said so far. Instead of saying that he is trying to balance conflicting tendencies whose presence in the law does not reflect earlier attempts to balance them, the Dworkinian lawyer could say that he is purporting to replace the existing—and admittedly contradictory—balancings embedded in the law by individualists and altruists, with a true balance, one that would be neither individualist nor altruist, but just correct. Kennedy wants to pre-empt this by announcing that the idea of a true balance is simply out of the question.

To the extent he considers it, I think Dworkin regards this as an expression of external skepticism about right answers in regard to balancing, and as such he rejects it out of hand.⁵⁴ True, it may not be possible for the Dworkinian lawyer to demonstrate that the new balancing—with which he proposes to supersede the contradictory balancings embedded in the law by individualists and altruists—is objectively correct. But non-demonstrability is not the same as falsity or futility.

I think this last part of Kennedy's critique is properly dismissed in this way. But whether the dismissal ultimately helps Dworkin is a question we must postpone until after we have examined the merits—in jurisprudence—of the constructivist approach that we are now imagining the Dworkinian lawyer undertaking.

(p.169) VI.

According to Professor Dworkin, the CLS or internal skepticism critique labors under another misapprehension. The fact that various elements in our law have conflicting inspirations does not mean that we cannot construct a theory that resolves the contradictions into some sort of attractive coherence. After all, Dworkin's judge or lawyer is not simply reporting the provenance of the legal materials he is dealing with: “He tries to impose order over doctrine, not to discover order in the forces that created it.”⁵⁵ The interpretive task is not to find out the purpose or intention with which past legal decisions were actually made; it is to make something good and coherent of the past decisions whether that good or coherent understanding corresponds to what the decision-makers in the past had in mind or not.

There is a quick version of this Dworkinian response and a longer version. The quick version might go as follows. Dworkin might want to use this point about constructivism to contest Professor Kennedy's assumption that there are elements in the existing legal record which just are individualist and that they are contradicted by elements in the existing legal record which just are altruist, as though all the elements come to us ready labeled in that way. In fact, Dworkin might want to say, what we face is an array of decisions: say, the enactment of statute S₁, a decision for the plaintiff in case C₁, a decision for the defendant in case C₂, the enactment of statute S₂, and so on. Calling S₁ and C₁ altruistic and S₂ and C₂ individualistic, and asserting therefore that the array of decisions is riddled with contradictions is itself already an interpretation of the materials. Dworkin might protest therefore that Kennedy is rigging the game by helping himself to these characterizations, as though they were part of the preinterpretive specification of the materials.

This is a fair point, as far as it goes, and I shall return to it at the very end of the paper. Some of what Kennedy says about altruism and individualism is that they are characteristics of legal argument, not of legal materials: they are “two opposed attitudes that manifest themselves in debates about the content of private law rules,” and “they are helpful in the general task of understanding why judges and legislators have chosen to establish or enact particular private law doctrines.”⁵⁶ It looks, then, as though a Dworkinian lawyer might resist these characterizations—leaving it open to himself to characterize the various mater-ials in other ways—and thus resist the premise that he is confronted with inherently contradictory materials. This strategy makes particular sense with regard to legislative materials. We should be wary of any attempt to treat particular legislators’ purposes as though they were canonical and on a par with the text of the statute they enacted. Dworkin has been a fervent opponent of this mode of (p.170) statutory interpretation,⁵⁷ and he must be allowed the advantage of that here. So Dworkin's interpreter need not accept that S₁, say, is an altruistic statute; he may want to fit it into his theory in a different way, which is at odds with what its original sponsors thought.

With case law, though, the situation is more complicated and less helpful to Dworkin. I don't think he would (or should) want to associate his constructivist approach to integrity with the claim that all we have to interpret is the actual holding of a case—e.g., appellant loses, plaintiff has to pay $40,000 damages—and that we can ignore the opinion produced by the court explaining that outcome, as though it were the case-law analogue of legislative history. Unless one wants to adopt the position of an extreme legal realist—like Jerome Frank⁵⁸—we will usually regard judicial opinions accompanying decisions as part of the record we are required to interpret and not as extraneous to it. Of course, those opinions are themselves partly interpretive of the materials that their authors confronted. The judicial opinion accompanying the decision in case C₃ will be in part an interpretation of the decisions (and the opinions) in cases C₂ and C₁; so the judge in C₄ reading the opinion in C₃ will be reading the opinion of someone who was trying to do just what he is doing (so far as C₁ and C₂ are concerned). What the judge is interpreting is thus in part a set of interpretations. The activity of interpretation is recursive: one interprets the past interpretations of others for they too are part of the community's record. So one interprets not just outcomes but modes of decision, patterns of argument, emergent doctrines, and so on. And for this, it is not inappropriate to say, as Kennedy and others sometimes say, that they are individualist or that they are altruist. One just has to read them, and it may be very difficult—a Herculean task, in fact—to read the altruism out of the decisions that one wants to reconcile with one's own consistent individualist interpretation or vice versa.

So there is no refuge for the constructivist in this quick gambit. The contradictions among the materials he confronts may be non-negotiable. If this is the case, then Dworkin's constructive interpreter has to adopt a different and less glamorous strategy. He has to pick and choose from among the existing contradictory decisions the ones that his interpretation proposes to fit, and discard the rest. Confronting a given set of legal materials, Dworkin's judge or lawyer will try to construct a coherent theory—or state an appropriate principle or set of principles—that implicates as many of them as possible. Since there are contradictions (as opposed to mere tensions) in the materials, no doubt a significant number of them will not be covered by the theory or the principles he proposes. But this doesn't mean he cannot come up with a theory or a set of principles that fits a significant portion of them. (Indeed, if the materials are pervaded by contradiction, as Kennedy asserts, then one would expect Dworkin's interpreter should be able to come up with a theory that fits roughly half.)

(p.171) Now, a mischievous interpreter might take a different strategy: he might come up with something purporting to be an interpretive theory, which was as inconsistent as the patchwork of materials it purported to fit. But that would just show bad faith—the bad faith that many opponents of CLS accuse its practitioners of harboring: they want to make the law look bad. “Nothing is easier or more pointless,” says Dworkin, “than demonstrating that a flawed and contradictory account fits as well as a smoother and more attractive one.” But to sustain his internal skepticism, the critic “must show that the flawed and contradictory account is the only one available.”⁵⁹ The internal skeptic is not entitled to say that the interpretive enterprise is doomed to failure unless he has tried as hard as he can: “[he] must claim to have looked for a less skeptical interpretation and failed.”⁶⁰

Actually this last point won't do, so far as the burden of proof is concerned. Internal skepticism is not refuted by a showing that a non-contradictory interpretation—a non-contradictory theory or a consistent set of principles—is available. For of course one can select a non-contradictory subset from any mass of inconsistent propositions: “p” on this sort of constructivism, would be a consistent interpretation of {p, ˜p}, and similarly individualism (or, alternatively, altruism) would be a consistent interpretation—on this constructivist approach—of the law as Professor Kennedy portrays it. After all, it's a matter of basic logic that one can make a case for anything on the basis of contradictory premises. So the onus cannot be on the CLS skeptic to show that no case can be made. What must be shown, in order to refute the internal skeptic, is either that only one such coherent interpretation can be constructed or that a credible case can be made that one of the available coherent interpretations of the contradictory materials is better than any of the others.⁶¹ Too often, I think, Dworkin's formulations neglect the forensic adversarial context of legal argument: it is not enough to show that a lawyer can come up with a legal argument; what he comes up with must be capable of refuting and displacing the legal argument that his opponent is likely to come up with as well.⁶²

Of course, all this needs to be leavened with some acknowledgment of whatever is legitimate in Dworkin's argument about right answers and demonstrability: there can, he insists, be right answers to legal questions, even though there is no acknowledged method of demonstrating that an answer is right.⁶³ I will assume (p.172) that this is roughly correct,⁶⁴ and that the case for internal skepticism cannot be merely that Dworkin has offered no algorithm for correctness in legal argument.

Still, one is inclined to say: something must count as tending to show—or there must be something one can do which counts as attempting to show—that the argument one is making is better than any opposing argument one can imagine being made by an opponent in the same case, drawing constructively upon the same array of materials. I mean something must count as that, apart from one's consciousness that one's own argument can be made or that it just is the argument one is making. This is because one is also conscious—confronted as one is, with this mass of contradictory materials—that one's opponent's argument can be made as well.

I get the impression that Professor Dworkin is sometimes prepared to say that this is a misplaced demand, analogous to what he regards as the unhelpful demand that for a person who believes that moral propositions are objectively true or false, something must count as trying to show that a given moral proposition is objectively correct, over and above the mere assertion and re-assertion of that proposition.⁶⁵ At other times, however, he characterizes the phenomenology differently, in a way that suggests that someone wedded to an interpretive argument may sensibly regard himself as having reasons for thinking the argument he is wedded to is correct. The characterization I have in mind is set out, in the first instance, in regard to interpretive claims about literature:

Someone just converted to a new reading of Paradise Lost, trembling with the excitement of discovery, thinks his new reading is right, that it is better than the one he has abandoned, that those yet uninitiated have missed something genuine and important, that they do not see the poem for what it really is. He thinks he has been driven by the truth, not that he has chosen one interpretation to wear for the day because he fancies it like a necktie. He thinks he has genuine, good reasons for accepting his new interpretation and that others, who cling to the older view he now thinks wrong, have genuine, good reasons to change their minds.⁶⁶

Now, I have no doubt that sometimes legal argument may have this character: trembling with excitement, a jurist comes up with a new interpretation of existing doctrine, a sense of something that others have missed, and of their having good reasons to abandon their old tired reading in favor of one's own. Sometimes it may feel like that (though there would still be philosophical questions about what to infer from such phenomenology). But is this how we expect it to be for most cases? Is this phenomenology—the trembling excitement of discovery—likely to be characteristic of legal work in the context of materials that are as contradictory as the (p.173) CLS scholars say our existing American law is? For remember: Dworkin is offering the constructive aspect of his argument as a way of finessing the CLS claim about contradictions. Never mind how conflicted the law is, he says, never mind how riddled it is with contradictions; one can always come up with some sort of constructive argument for one's position. So the question is—if the legal materials are as conflicted (and are known to be as conflicted) as, say, Duncan Kennedy thinks they are⁶⁷—would one expect the process of constructive argument that Dworkin promotes to have the sort of phenomenological flavor of “discovery” and “good genuine reasons” that characterizes our imagined student's new interpretation of Paradise Lost? Given the knowledge that anyone faced with a set of contradictions has about the opportunities that inconsistent premises offer for a multitude of arguments in opposite directions, I think this is most unlikely.

We are contemplating the possibility that faced with an array of legal materials as contradiction-ridden as Kennedy and other CLS scholars take them to be, a constructive argument interpreting the law in favor of one side or the other in litigation will be able to satisfy some very modest threshold of “fit.” Crudely, it will make sense of roughly half of the materials it considers, and its proponent will be conscious that any remotely competent work done on behalf of the other party will also satisfy that threshold of fit, since it makes sense of the other half of the (conflicting) legal materials under consideration. In these circumstances, should we expect either advocate to make his argument with a burning consciousness that he is right about the law and his opponent wrong? Should we expect the judge, who entertains both arguments and understands the condition of the possibility of each of them, to be conscious of a basis for discriminating between them?

Dworkin remains remarkably upbeat, as he faces the prospect of this even rivalry between two equally viable theories purporting to make “sense” of the same mess of contradictory legal decisions. What he says, at least about the judge's situation, is this:

Hard cases arise, for any judge, when his threshold test [of fit] does not discriminate between two or more interpretations of some statute or line of cases. Then he must choose between eligible interpretations by asking which shows the community's structure of institutions and decisions—its public standards as a whole—in a better light from the standpoint of political morality. His own moral and political convictions are now directly engaged.⁶⁸

The idea is that a judge, faced with two arguments that “fit” (in this quixotic sense) an equal quantity of the contradictory legal materials, will resolve the issue between them by considering which is superior on the merits, that is, on grounds of justice or other substantial values. And that will be the tie-breaker.⁶⁹

(p.174) To understand what this position amounts to, we need to recall that on the CLS account the legal materials are not just contradictory in some technical sense. They are contradictory as between the main views about justice (and other substantial values) already held by diverse factions in society, and also as between the main values that are likely to be at stake in any piece of litigation to which it is supposed that law as integrity should be applied. We need to remember, too, that this is the hypothesis that Dworkin is supposed to be responding to, and in connection with which the passage just quoted above is supposed finally to refute the internal skeptic's critique.

Let me put it very schematically: suppose Kennedy is right and there are two recognizable moral positions, Ind (for individualism) and Alt (for altruism) in the law—contradicting one another, but spread throughout the settled law in an array of decisions in numbered cases {Ind ₁,Alt ₂,Ind ₃,Alt ₄,Ind ₅,Alt ₆, …}. And now once again, the old adversaries face off against one another in yet another case—case 7, say. So the two sides set about constructing their arguments, following the constructive method that Dworkin recommends to practitioners of law as integrity. The individualist party cites a subset of the established cases {Ind ₁,Ind ₃,Ind ₅} and he puts cases 2, 4, and 6 aside. He is conscious, of course, that his opponent is likely to cite an equally sized subset {Alt ₂,Alt ₄,Alt ₆}, putting cases 1, 3, and 5 aside. So what does the individualist party propose to clinch his argument? He says that his argument with its support in half the case law is better than his opponent's, because it is committed to Ind and because Ind is, as he believes, a better moral theory than Alt. The reason that clinches the matter on his account—the reason that is supposed to be capable of clinching the matter on Dworkin's account—is simply that in this conflict between two moral views that have come to pervade the law, he has aligned himself with the better one!

I very much doubt whether Dworkin will be embarrassed by this (though I will show shortly that he ought to be). The point of the critique implicit in my schematic example tends to be obscured by the fact that when Dworkin says that the judge's “own moral and political values” are engaged in the choice among rival theories that fit the existing legal materials equally well, we are not reminded that those very values are likely to be implicated on one side of the contradictions that pervade the law (on the CLS account). He sometimes writes as though the contradictory legal materials—leading to the impasse of fit—are one thing, and the values the judge brings to his task are another. I don't mean that Dworkin's argument rests on this misapprehension. He would probably bite the bullet and say that the judge has no choice but to rely on his own views about justice, even when his own views about justice are exactly the views that are represented (and contradicted) in the diverse legal materials facing him. But the possibility is not one that he brings to our attention. More important, he doesn't draw to our attention the fact that the attorneys for the parties⁷⁰ are already committed to one side or the other in (p.175) this debate; they are litigating the issue precisely because they are partisans of Ind and Alt respectively. So the values they appeal to when they try to clinch their argument are, in some sense, question-begging arguments. They assume the very point at issue and they then claim that this assumption breaks the tie between their own argument and that of their opponents!

VII.

An even more serious difficulty with Professor Dworkin's constructivist response to the internal skeptic is the question of its compatibility with the underlying argument for integrity in Law's Empire.

Briefly: Dworkin's judge (and hence Dworkinian attorneys) are supposed to be preoccupied with the interpretation of the existing legal materials because, like all of us, they have an obligation rooted in political morality to keep faith with the commitments of their community. This, according to Dworkin, is why it is worth persisting in trying to make sense of the existing legal record, tainted as it is with contradiction (if the skeptics are right). Absent that sense of obligation, the sens-ible thing to do in legal practice would be to abandon wrestling with the messy array of existing law and simply settle (and argue) cases pragmatically on the basis of what will be best for a just and prosperous future. But if the skeptic is right, if the legal materials are as contradictory as he says they are, then there is no coherent record to keep faith with, and therefore no ground for the integrity-based case against pragmatism to stand on. There may be clever moves—constructivist moves—that a Dworkinian lawyer can make which look impressive and which, at a stretch, have the feel of interpreting even the messy record that the internal skeptic postulates. But those moves do not and cannot engage the sort of integrity values that the interpretive exercise is ultimately answerable to. In other words, they have no point or at least they don't have the sort of point that integrity is supposed to give them. Quite the contrary: these constructivist moves in the face of contradictory materials make a mockery of the communitarian values underpinning integrity. That, in the end, is the seriousness of the challenge posed by internal skepticism: the challenge is not merely that Dworkin is giving Hercules an impossible task; it is that the task given to Hercules makes no sense, if the legal world is as the internal skeptic says it is.

There are several phases in this argument. The first is Professor Dworkin's own acknowledgment that in the end his own jurisprudence stands or falls with the refutation of pragmatism. He phrases the point in terms of legal rights—that is, rights which are generated by legal principles (or rights which are the upshot of successful interpretive theories).

The pragmatist thinks judges should always do the best they can for the future, in the circumstances, unchecked by any need to respect or secure consistency in principle with what other officials have done or will do.… [Pragmatism] rejects what other conceptions (p.176) of law accept: that people can have distinctly legal rights as trumps over what would otherwise be the best future properly understood.… I do not say this in any triumphant way. The fact that a true pragmatist rejects the idea of legal rights is not a decisive argument against that conception. For it is not self-evident that the idea of legal rights is attractive. Or even sane.⁷¹

Indeed, as Dworkin goes on to say, it is quite easy to make the idea of legal rights seem foolish. If a judge is convinced that the doctrine in a line of cases has proved itself unjust, why should he accept the argument that a party has a legal right that the injustice be perpetuated? Maybe if the party can show that he would suffer more on account of his reliance on the injustice than his opponent would gain were it corrected, or maybe if it can be shown that society is better off with this doctrine as a basis for social predictability—maybe then there are reasons for accepting his argument. But those arguments are themselves pragmatic and they don't affect the underlying point that it seems foolish to worry away at difficult precedents, tangled doctrine, or obscure legislation to find out what the law really says, when one suspects it will be unjust and when it is clear anyway that no one is really relying on any particular interpretation and that interpretive argument contributes nothing to predictability. Dworkin is adamant in Law's Empire that conventionalism, as a theory of law, falls at this hurdle—conventionalism cannot withstand the pragmatist challenge⁷²—and he cannot consistently confront law as integrity with a lower bar than this.

So a jurisprudence of integrity must be capable of explaining why we ought to refuse the pragmatist invitation to decide cases simply in a way that is best for the future. A jurisprudence of integrity must explain why it is not foolish to worry away at difficult precedents, tangled doctrine, or obscure legislation in a quest for a solution (to the case in front of us) which is consistent with what our community has already committed itself to. And the account of legal argument that it gives must connect with that explanation.

Now, Dworkin's answer to the pragmatist challenge has to do with the conditions for political legitimacy. Decisions by judges and other legal officials will be enforced, and there is a question about the legitimacy of their enforcement and the obligation that citizens have to defer to them—particularly those on the losing side in the decisions. Professor Dworkin pursues a long and interesting argument suggesting that legitimacy depends on the extent to which the obligation to accept legal outcomes can be presented as a form of associative or communal obligation.⁷³ Associative obligation, he says, can be sustained only among the members of a community “who share a general and diffuse sense of members’ special rights and responsibilities from or toward one another.”⁷⁴ Delving into the legal record for principles or attempting to produce an interpretive theory (p.177) which gives an attractive and coherent account of what has been done among us already in the name of our association—both of these are ways of trying to show what that “diffuse sense of members’ special rights and responsibilities” is. And connecting those principles or that theory with the decision one is arguing for in a particular case is a way of showing that parties in the case have an associative obli-gation to abide by that decision, because they are already in a sense committed to it.

Dworkin argues that such a strategy has a huge legitimacy-advantage over the pragmatist approach. In aiming to make society better for the future, the pragmatist naturally commands the support of those who accept his vision of social justice and the general good. But it is hard to see why anyone with a different or opposed vision of social justice or the general good should accept the decisions of a pragmatist official. Integrity, however, makes a different sort of claim—one rooted ultimately in reciprocity. Even when we disagree with one another about justice or the general good, we can see that some among us have been benefitted by decisions applying principles of a certain sort, and we can appreciate the fairness of allowing others to be benefitted by those principles, even when we disagree with them, because we think that their consistent and open-ended application across a diverse array of cases helps to establish us as a community of principle.

So: this is the case that Dworkin needs in order to make sense of modes of legal argumentation that grapple—in this non-pragmatic way—with the obscure implications of the set of decisions our community has already committed itself to. Such arguments present our association with one another as a community of principle and they use that as a basis for saying that a decision, now, that flows from the principles we have unearthed has legitimacy even for those who would be inclined in the abstract to oppose it. To claim this legitimacy advantage, we have to be willing to view the existing record of our community in a certain light—“to see and enforce it as coherent.”⁷⁵ A judge who accepts integrity must be prepared to treat the parties who come before him in this spirit: “They are entitled, in principle, to have their acts and affairs judged in accordance with the best view of what the legal standards of the community required or permitted at the time they acted, and integrity demands that these standards be seen as coherent, as the state speaking with a single voice.”⁷⁶ And it is in this spirit that Professor Dworkin undertakes the enterprise which we have been discussing in Sections II through VI.

Let us return now to the internal skepticism of CLS and to the constructivism that Dworkin uses to respond to it. The internal skeptic denies that it is possible to view the record of a community like ours as one in which a certain set of coherent principles have consistently determined legal outcomes. The record of our community, he says, does not disclose any consistent “sense of members’ special rights and responsibilities from or toward one another,” diffuse or otherwise. The internal skeptic denies therefore that any resolution of an instant case can made (p.178) legitimate by showing that it flows from such a set of coherent principles. The record of our community, he says, is one of incoherence, in which the principles necessary to sustain some decisions are contradicted by the principles necessary to sustain others. There is nothing, says the internal skeptic, corresponding to Dworkin's description of a consistent set of principles presupposed by all or most of the existing legal materials, from which we could infer legitimate answers to the legal questions that currently confront us. Professor Dworkin may say that “associative obligations can be sustained among people who share a general and diffuse sense of members’ special rights and responsibilities from or toward one another, a sense of what sort or level of sacrifice one may be expected to make for another.”⁷⁷ But Duncan Kennedy's point is precisely that we are conflicted about what we owe one another; we don't share a single conception; we share contradictory ones, if we share anything at all. Again, Dworkin may say that “[l]aw as integrity asks judges to assume, so far as this is possible, that the law is structured by a coherent set of principles… and it asks them to enforce these in the fresh cases that come before them, so that each person's situation is fair and just according to the same standards.”⁷⁸ But the skeptical case is that this is not possible, and that this instruction simply cannot be carried out, because the existential assumption fails: law is not structured by a coherent set of principles.

For myself, I find it difficult to see how any form of versatile constructivism can displace or mitigate this verdict. A Dworkinian judge or lawyer might make an attempt to establish a record that is consistent as far as it goes, in the sense of identifying principles which offer a coherent explanation of some of the extant decisions. But the internal skeptic thinks that any such attempt will be easy to discredit. Whatever decisions it explains, it will leave an equal number of decisions unexplained (treated as outliers). So any such interpretation (in Dworkin's own words) will “show the record of the community in an irredeemably bad light, because proposing that interpretation suggests that the community has characteristically dishonored its own principles.”⁷⁹

Notice that this is not just a matter of there being rival theories available, and no demonstrable way of settling the issue between them. We can accept Professor Dworkin's claim that “people are entitled to a coherent and principled extension of past political decisions even when judges profoundly disagree about what this means”⁸⁰ and his view that “consistency in principle [can] be important for its own sake,… [even] when it is uncertain and controversial what consistency really requires?“⁸¹ But those maneuvers do not save the position here. For what we face, if the internal skeptic is correct, is not a dispute between two viable interpretations which unfortunately cannot be settled in a demonstrable way; what we face are two rival interpretations, each of which is discredited by the fact that it seems to show a very considerable part of the community's record in a very bad light.

(p.179) The situation is not saved by a formulation which Dworkin sometimes uses, in which law-as-integrity is understood in terms of an ethic of trying to portray the law as coherent, rather than in terms of a duty to uncover the coherence that is actually there. I mean formulations like the following:

We want our officials to treat us as tied together in an association of principle, and we want this for reasons that do not depend on any identity of conviction among these officials, either about fit or about the more substantive principles an interpretation engages. Our reasons endure when judges disagree, at least in detail, about the best interpretation of the community's political order, because each judge still confirms and reinforces the principled character of our association by striving in spite of the disagreement, to reach his own opinion instead of turning to the usually simpler task of fresh legislation.⁸²

These formulations are quite common in Law's Empire and elsewhere. Dworkin says that integrity is upheld “when people in good faith try to treat one another in a way appropriate to common membership in a community… and to see each other as making this attempt, even when they disagree about exactly what integrity requires in particular circumstances.”⁸³ He says that “[w]e gain even through the attempt”⁸⁴ and that “[l]aw's empire is defined by attitude, not territory or power or process”⁸⁵ and “[l]aw as integrity consists in an approach, in questions rather than answers.”⁸⁶ But not everything that one does to advance one's own position through legal argument can count as a good faith attempt of this kind. Nor are the conditions for a good faith attempt only subjective—such as that the lawyer is doing the best for integrity that he can, or whatever. The attempt must be something which it makes sense to embark on, and the critical or skeptical case is that, given what participants must know about the pervasively contradictory character of the existing legal materials, picking and choosing from among them those that suit one's purposes can hardly be described as a way of trying to display one's allegiance to their underlying coherence.

VIII.

Since the end of Section IV, I have been working on the assumption that the legal background in a society like ours is about as riddled with contrary principles as the CLS skeptics say it is. I proceeded on this assumption in order to see whether Professor Dworkin had anything convincing to say in response to their skepticism apart from a flat denial of its factual premise. I think we have concluded that he has not. The distinction between tensions and contradictions does not do the trick and neither does the move to constructivism nor the insistence that non-demonstrability is compatible with objectivity.

(p.180) That leaves the flat denial. At the end of Section IV, I was inclined to doubt that Dworkin wanted to rest his jurisprudence on this. It sounded flabby and unconvincing, and appeared to make the whole enterprise contingent and precarious. But maybe that was premature, and if we turn around the argument that we have just been developing (in Section VII), we may be able to see why. I said that constructive interpretation by a Dworkinian advocate cannot count as a bona fide attempt to keep faith with the commitments of his community if it is known to the advocate that the record of the community is one of pervasive and systematic contradiction. If the advocate knows that roughly half the decisions embody the moral proposition that members of the community have very limited obligations to one another and roughly half of the decisions embody the moral proposition that members of the community have very extensive obligations to one another, he cannot with a straight face pull out just the materials in the latter half and argue on that basis that his opponent has a particular obligation to his client because this is what we—the community—have always been committed to. But suppose his knowledge of the extent and pervasiveness of the contradictions is not so clear. After all, the illustration that we used in Section VI—an array of decisions in numbered cases {Ind ₁,Alt ₂,Ind ₃,Alt ₄,Ind ₅,Alt ₆,}—was a ludicrously over-simplified version of the legal record that most advocates are likely to be confronted with. Ludicrously over-simplified, and ludicrously over-clarified: the tens of thousands of extant legal decisions that we find in the law reports and the statute books present a much more equivocal and tangled record than this. Confronted then with real-world legal materials, the Dworkinian advocate will have in the back of his mind the possibility that he will not be able to make a determinate case—or, as I emphasized in Section VI, the possibility that his opponent may be able to make a case as determinate as the one that he makes—but he may think it worth trying to see whether he can do any better, for integrity, than that. If it seems to him that the prospect of success here is not out of the question, then there is no reason to say—as I said about the overly simple case—that his attempt just makes a mockery of integrity.

So the factual disagreement between Dworkin and the skeptics may come down to this: Are the contradictions so clear and so pervasive that it is evidently not worth trying to see whether a particular set of principles (or a particular interpretation) fits the existing law? “Not worth trying” is the key predicate in this formulation. It does not mean and it should not be taken to mean that there is no legal argument that we can come up with. If we are prepared to pick and choose, we can always come up with something, even if or (as I said earlier on the basis of elementary propositional logic) especially if we know the materials are inconsistent. “Not worth trying” must refer to the conditions that would make the attempt important or worthwhile; that is, it must refer to the general idea of integrity. It must not be out of the question that our argument or our principles fit a very significant portion of the legal materials, and it must not be out of the question that this is not the case for our opponent's argument. If the constructive strategy that organizes our attempt rests on our having to say, even implicitly, that “the (p.181) community has characteristically dishonored its own principles,“⁸⁷ then it is not an attempt worth making.

I think, in the end, this is where Dworkin should take his stand against the Crits. He should say (and he does say): it is not clear up front that attempts to argue in the mode of law-as-integrity are doomed to failure. If it were clear, we should have no reason to resist the siren charms of pragmatism: forget the existing law; ask instead what's best for the future; and take one's chances on the legitimacy issue. But sometimes legal argument looks promising, and when it does we are obliged to make the attempt (and the theory of integrity explains why). The cruder versions of CLS jurisprudence were easy to discredit: it simply was not so clear that law is systematically biased in favor of one social class as some CLS scholars said it was. Law is more complicated than that. The more sophisticated versions of CLS that I have been considering in this paper are certainly less easy to dismiss, and they deserve as Dworkin says to be taken seriously.⁸⁸ But perhaps the least convincing thing about them is the simplicity of their thesis about contradiction: there are these two positions (individualism and altruism) and the existing legal materials divide clearly and evenly between them. So long as we stick with that simple picture, the futility of law-as-integrity is evident. But complicate and muddy it somewhat, with more terms to the contradictions and arrays of less easily legible materials, then the appearance of futility recedes. And at that stage, as Dworkin says,⁸⁹ there is nothing to do but try, for nothing else will reveal whether an attempt can succeed while keeping faith with the motivations behind integrity.

The conclusions I have reached are perhaps not quite what Dworkin would want, for two reasons. First, I have insisted that what one is trying to make is a determinate legal argument, and that means one that excludes or promises to exclude the prospect of a similar argument by one's opponent. To repeat: nothing is easier than for a constructive jurisprudence to come up with something that looks like a legal argument. The test is whether there is a reasonable prospect that one can show that nothing but this argument will keep faith with the commitments of our community. Secondly, I do think that Dworkin is simply wrong to think that this burden can be discharged by using one's own moral and political convictions to break ties.⁹⁰ If the only thing to differentiate plaintiff's argument from defendant's argument, as a matter of law, in plaintiff's eyes is that plaintiff is an altruist and—as the plaintiff thinks—altruism is right (as a theory of justice), then plaintiff is arguing as a pragmatist. He is not arguing in good faith in the mode of law-as-integrity since, ex hypothesi, he concedes that the rest of his argument (the legal part) fails to discharge the burden that integrity imposes and what he appeals to as a tie-breaker has no connection with integrity, or nothing that would differentiate the stance of a defender of law-as-integrity from a defender of pragmatism. The moral is that the use of the tie-breaker, too, must satisfy the conditions of integrity. It can't just be wheeled out on an ad hoc basis to rescue an argument that would fail at the bar of integrity without it. (p.182)