Michigan Law Review
May, 1992

1992 Survey of Books Relating to the Law
VI.  Legal Philosophy

*1520 CONCEPTIONS OF VALUE IN LEGAL THOUGHT

LOVE'S KNOWLEDGE.  BY MARTHA C. NUSSBAUM.  NEW YORK:  OXFORD UNIVERSITY PRESS.
1990.  PP. XIV, 403.  $42.50.

Richard H. Pildes [FNa]
Copyright © 1992 by the Michigan Law Review Association; Richard H. Pildes
 

  Martha Nussbaum, [FN1] formally trained in ancient philosophy, is increasingly finding an audience among those concerned with contemporary law and policy.  But Nussbaum herself has long defied conventional academic boundaries, and her recent book, Love's Knowledge, continues this refreshingly unorthodox approach.  In this collection of essays, she begins in her home territory, sharpening conflicts between Aristotle and Plato in ways that link their concerns to ours, then turns literary critic to vindicate the Aristotelian position through readings of James, Proust, and Beckett, and finally concludes with her own efforts at crafting literary fiction.  All this in the service of arguing that much contemporary moral, political, literary, and economic theory shares a disturbing family resemblance:  in different ways, for different reasons, these disciplines have increasingly distanced themselves from both the difficulty and richness of pressing ethical conflicts.  Not afraid to conceive the task of theory in traditional terms - as the effort to find practical guidance for concrete personal and public choices - Nussbaum relentlessly argues that contemporary academic disciplines are failing in this task.  For a legal audience, the question is whether legal thought and practice have much to learn from this multifront assault on the academic citadels.  I believe that they do.

  Nussbaum believes the place to begin revitalizing contemporary intellectual thought is, perhaps paradoxically, with Aristotle.  In both method and substance, Nussbaum is a committed Aristotelian (though as an Aristotelian, it is central to her views that method and substance are mutually defining and constituting).  But her Aristotelianism hardly sustains a nostalgic polemic against modernity in the way, for example, that Alisdair MacIntyre enlists Aristotle. [FN2]  Instead, through *1521 Nussbaum's interpretive skills, Aristotle's thought becomes compelling for contemporary concerns.  For she reveals an Aristotle and Plato in profound conflict over competing ethical visions - competing visions that endure and turn out to frame the central divide in contemporary thought across several disciplines.  Her lament is that views that can be seen as descendants of Plato's are winning the academic and cultural day.

  Against the ascendancy of these views, Nussbaum constructs what might be called a neo-Aristotelian alternative.  In an essay philosophically central to the book, she argues that this alternative consists of four mutually reinforcing conceptions about how values and rational choice among them are best understood. [FN3] First, we experience ourselves and our public institutions as struggling to make choices involving radically distinct values.  We should accept this experience and acknowledge that these values cannot be reduced to some ultimate, single value; to respect them properly means to recognize their independent intrinsic worth. Hence, rational choice must be understood as choice among "incommensurable" values.  Second, we should accept a moral methodology in which deliberation is understood as the internal interpretation and revision of existing social practices and ethical judgments. [FN4]  Third, reason and passion should not be seen as antagonists, but often as necessary sustaining conditions for each other; *1522 just as certain affective experiences depend upon the existence of particular beliefs, reasoned deliberation often must be anchored in truths partly known through less intellectualized responses of outrage or offense. Emotions respond to cognition and often embody the most deeply rooted ethical commitments necessary to deliberating well.  Finally, as Aristotle asserts, good decisionmaking cannot involve a kind of rule formalism in which rankings of different values are taken as settled before we confront a setting of actual choice.  Particularly for law, Aristotle claims that the "discernment" of correct choice rests with the proper "perception" of each new particular setting that a decisionmaker confronts.  Framed in legal terms, we might interpret this quality of perceptiveness as entailing at least two commitments:  (a) that judging should not be a process of mechanically applying past rules to new contexts, but of interpreting rules so that their underlying purposes are most faithfully served in each new context; (b) more radically, that judging should remain open to the possibility that new patterns of conflict can bring to light reasons to reinterpret and revise the structure of values reflected in pre-existing rules.  Against rule formalism, Aristotle emphasizes the priority of particulars to general rules and the inability of any exhaustive general framework to provide a complete, fixed account of the hierarchy of values appropriate for deliberation. [FN5]

  To discover in Aristotle a source of perspectives on many of the debates that preoccupy contemporary legal theory is intriguing. Taken as a whole, this Aristotelian vision provides a distinct conception of how we might understand rationality and choice - a conception that stands against efforts to extend to areas like law or ethics visions of rationality common in the natural sciences.  In a contrapuntal variation, a companion essay argues that Plato's thought exemplifies the competing aspiration. [FN6]  According to Nussbaum, Plato *1523 believed that unless ethics could be made a "real science" (techne), capable of orderly systematization of values in a way that would "save our lives" from the experience of conflict and uncertainty, it did not deserve our attention (pp. 106-13).  To provide such a science of conflict resolution, Plato rejected each element in the Aristotelian vision. In the clash between these visions lies the roots of much contemporary disagreement, including, I hope to show, disagreements in legal theory and practice.

  Of what value is pursuing these disagreements to their ancient philosophical sources?  Unreflective traditionalism, however common a feature of practical legal argument, is not the reason.  Still, some contemporary debate in law, particularly the revival of rule formalism, boldly asserts that "law is rules" - and hints that any other conception of law corrupts traditional rule- of-law ideals. [FN7]  Against these accounts, Aristotle might now be offered as a subversive figure. [FN8]  But there are better reasons.  For in Nussbaum's sympathetic unveiling of *1524 each, we discover that Plato and Aristotle pursued these issues to a depth and subtlety that, even to modern sensibilities, is striking.

  Each element in the Aristotelian conception of rationality makes a claim with important correlates for legal reasoning.  But for legal thought, turning to Nussbaum's account is most promising with regard to two of these elements:  the incommensurability of values and the appropriate role of emotion in forming justified beliefs. For legal theory, issues surrounding rule formalism not only are best addressed as distinctly legal ones, given the institutionalized, bureaucratic setting of modern law, but have, since legal realism, been the subject of some of the most incisive thinking among legal scholars. [FN9] And arguments that the method of justifying normative claims ought to be the internal interpretation of existing practices, rather than the search for some external, "more fundamental" point of view, have recently been elaborated in political theory, for example, by Michael Walzer; [FN10] in legal theory by Ronald Dworkin (though Dworkin initially scorned Walzer's approach before his volte face embrace of it); [FN11] and in epistemology by Richard Rorty.  [FN12] In contrast, the nature of emotions and their relation to rational thought is an area legal scholarship has just begun to explore, [FN13] and Nussbaum's essays on Beckett (pp. 286-313) *1525 and Proust (pp. 261-85) offer a provocative starting point for those interested.  But in this review, I plan to concentrate instead on Aristotle's final claim:  that deliberative practices, in areas like legal decisionmaking as well as moral choice, ought to recognize a plurality of incommensurable values.  For I hope to show that in Nussbaum's hands, Aristotle's arguments provide a lever for unearthing the deepest stakes in much contemporary legal controversy.

  Because I will focus on only one aspect of this over 400-page book (and will concentrate on extending Nussbaum's arguments to new areas, rather than describing her own approach in detail), I should first say a few words about the rest of it.  Love's Knowledge manifests an integrity of intellectual commitment that accounts for Nussbaum's movement from ancient philosophy to a series of critical literary essays, which engage in a genre of moral critique familiar from the days before literary theory turned to more aesthetic concerns.  Nussbaum challenges the style of current philosophical inquiry as too spare and suggests that literary and dramatic texts, with their evocative, richly portrayed settings of conflict, should be moved to the center of ethical analysis.  She would undoubtedly agree with Cynthia Ozick's view that "metaphor is one of the chief agents of our moral nature, and that the more serious we are in life, the less we can do without it." [FN14]  Unlike Love's Knowledge, though, much current philosophical work on method and justification consists of repeated clarion calls for pragmatic reasoning, or local narratives, or exploration of the stabilized cultural matrices implicitly structuring the interpretation of doctrine - without moving more than a few paragraphs beyond exhortation to example.  Nussbaum admirably follows where her critique leads; she takes on the more difficult task of seeking to exemplify, through the labor of detailed readings, the specific kinds of insights literature might offer to ethics.  This makes for considerable repetitiveness, because many of the essays revisit similar philosophical themes.  But it also accounts for the twin achievements of Nussbaum's confrontation with contemporary literary and philosophical theory.  For she succeeds in sketching a style of philosophical inquiry that avoids lapsing into the kind of esoteric refinement of technical debates that seems to be an increasing byproduct of the division of labor in today's academy. [FN15]  At the same time, she reclaims moral conflict and ethical choice as the heart of much literature and, hence, of reading well. *1526 What emerges is an approach to philosophy that is engaged with the most richly imagined of practical settings, and an approach to reading attuned to the most profound questions of meaning and value.

 
I

 
  The view that the diverse things we appear to value can, in fact, be rationalized in terms of some single value is central to Platonic thought, classical and modern utilitarianism, much of the economic analysis of law, and other central techniques of modern social science.  Against this view, Nussbaum seeks to revive the Aristotelian argument that the things we value are (or are best understood as) irreducibly diverse. [FN16]  If the Platonic view is right, rational choice can be guided by the simple instruction to maximize value, for all choices would ultimately *1527 involve choosing greater or lesser quantities of the same value. But if choices involve values that cannot be compared with each other along a single scale of value, rational choice cannot be a process of maximizing.  Ultimately, then, we will have to confront the question of what rational choice might be if it does not entail maximizing a single, uniform value.

  Understanding rationality is thus at stake in how we conceive values.  But rationality, the animating Enlightenment ideal, is now under siege from several quarters.  Some challenges attack the concept of rationality itself, claiming that it can only serve to mask relations of power.  Others challenge not the concept but specific conceptions of it now prevailing.

  Rejecting Plato and acting on the view that values are often incommensurable poses the second sort of challenge.  It asks that we replace ways of thinking about rationality that inform Platonic thought, utilitarianism, and current "rational choice" models dominant in the social sciences with other ways of understanding what it means to choose rationally.  The question of how values and rational choices among them should best be understood implicates the public policies we should collectively adopt, the legal norms courts should choose, and the way we might resolve many types of conflicts, including moral ones, in our lives.

  The claim that we do and should see values as incommensurable can be interpreted in a number of different ways.  First, we often speak in evaluative terms that employ not quantitative but qualitative contrasts between values: we distinguish between noble and ordinary pursuits, or between ideals and interests, or between justice and welfare.  This language suggests that we appreciate some values not merely by *1528 considering them "better" than others, but by acknowledging that they are of a distinct (we might say more elevated) quality.  The way these kinds of theoretical commitments are expressed in concrete practice is by refusing to let "higher values" be sacrificed for "lower" ones; we treat intentional discrimination on the basis of race as wrong, even where it might be efficient. [FN17]  What constitutes an unacceptable sacrifice is a matter of how we (those whose interpretations are legitimately authoritative) interpret the values and depends upon understandings typically not capable of being reduced to formulaic rules.  But we are dealing with values that stand in a hierarchical relation to each other, rather than ones reducible to differing magnitudes of some single value.  These situations, which appear to involve qualitative contrasts, we might describe as implicating hierarchically incommensurable values. [FN18]  Recognizing some values to be of higher quality than others need not lead to an abstraction or absolutism in which they mechanically "trump" all competing values.  Respecting higher values depends, instead, on refusing to compromise them in circumstances and for reasons that would express contempt for their defining principles.

  There is a second way, capable of being formulated with greater analytic precision, in which values might be considered incommensurable.  We might face situations in which we do not think either of two options is better than the other, but we also do not think the options of equal value.  Suppose I must decide between a career as a legal academic or as a government lawyer.  The attractions of each seem powerful, but in different ways and along distinct dimensions.  Both are appealing; hence, the language of better or worse feels out of place.  In *1529 theory, we can test this phenomenology by asking a further question.  Suppose I start off undecided between the two options, and one of them is then enhanced in some way I value as an unambiguous improvement (I learn the job as a government lawyer will involve more independence and will pay significantly more than initially described).  If I still remain unable to view this alternative as the better one, then incommensurable values seem involved.  The options are not simply of equal value (in which case I might flip a coin) because an unambiguous improvement in one would have then turned it into the clearly "better" one.  But it has not.  We might describe these situations as ones involving radically incommensurable values. [FN19]

  Finally, when collective decisions are at issue, as in political contexts, no shared consensus may exist even as to the very meaning and understanding of what is at stake.  We are now a long way from early 1960s "end of ideology" claims that politics in America is essentially instrumental debate over how to reach ends that are the subject of widespread consensus. [FN20] But the diametrically opposed Weberian vision, that politics involves the clash of warring ideologies and ultimate ends, fails to capture yet more subtle conflicts that have become increasingly apparent with the rise of pluralism (as both social fact and political ideal).  For between instrumental techniques and final ends, there emerge disagreements over the prior question of what policy choices mean.  Different groups, acting against the background of different political, moral, religious, or other traditions, may view the same choice as carrying very different meanings.  Does state toleration of hate speech, or Nazi marches, express the very essence of mature democratic freedom - or a misconceived understanding of what genuine democracy means? Often there are no more encompassing, shared social understandings available to resolve these disagreements.  In situations like this, we might say we confront interpretive incommensurabilities of values. [FN21]

  *1530 The claim that values are incommensurable can thus be made more precise by distinguishing between hierarchical, radical, and interpretive ways incommensurabilities might arise. But thus far I have tried only to describe the way we talk about choices; I have not tried to show that this language is justified. The possibility of self-deception remains.  Nor have I suggested how Nussbaum's revival of Aristotle might help in deciding the "truth" about this question - and just what kind of truth we might understand ourselves to seek regarding a question of this sort.  To address these questions from the perspective of legal decision, we first need a firmer sense of what is at stake.

 
II

 
  Denial of the view that values are incommensurable has been, I believe, one of the defining characteristics of the past generation of legal thought. Approaches to justifying and criticizing legal norms have blossomed in recent years, but a number of important approaches implicitly concur in rejecting the understanding of values as incommensurable.

  If one were looking for a single moment when this rejection achieved the status of widely shared conviction in legal thought, an appropriate focal point would be Roscoe Pound's famous 1943 article A Survey of Social Interests.  [FN22] There Pound sought to write a final eulogy for the free contract, private property ideology of the Lochnerera Supreme Court by deconstructing jurisprudences that rested on the concept of legal "rights."  According to Pound, the language of rights, as expressed in legal decisions, only produced confusion and induced courts to prejudge social conflicts unreflectively.  Once a conflict was framed in the conventional judicial categories of rights versus general policy goals, "our way of stating the question may leave nothing to decide." [FN23]  Pound argued that courts should abandon this discourse of "rights" and "public policies" and instead understand all legal and policy choice to involve conflicts among nothing more than competing interests, all qualitatively the same.  For it was critical to the judicial method Pound urged that, " w hen it comes to weighing or valuing claims or demands with respect to other claims or demands, *1531 we must be careful to compare them on the same plane." [FN24]  Only through rationalizing legal disputes in terms of a single, uniform entity - individual interests, capable of being measured and weighed in units of greater or lesser quantities - could legal decision be disciplined by reason.

  Pound's arguments, though pursued with his distinctive historical and comparative law sophistication, were not novel.  They reflected the culmination of developments in legal thinking that had begun at least fifty years earlier. Oliver Wendell Holmes had initiated this style of thought in 1894, with his protolegal realist challenge to existing understandings of central tort law categories.  The article he published that year, Privilege, Malice, and Intent, [FN25] offered a vision of private law as inevitably involving conflicts among competing, equally legitimate interests; to resolve these conflicts, courts necessarily had to make fine-grained policy judgments concerning what Holmes called "questions of degree." [FN26]  By reframing judicial decision this way, Holmes rejected the traditional view that legal reasoning fundamentally involved a process of characterization - of fitting a case to its correct type (either by analogy from preexisting categories or deduction from preexisting concepts).  Instead, courts, like legislatures, were primarily making judgments along a continuum of competing interests; they could only aim for a socially optimal point of conflict accommodation. [FN27]

  Reconceptualizing legal theory in these ways had the intent and effect of undermining the view that legal categories should, at times, be understood to reflect qualitative distinctions among values and interests.  If we historicize Pound's writings and similar texts of legal realism, the motivation for this attack is easy enough to understand.  Pound self-consciously wrote with an eye toward delegitimating the ideology of the Lochner Supreme Court.  That ideology had been constructed upon a foundation of natural rights and common law categories in which particular rights, namely those to private property and freedom of contract, were conceived in absolute terms.  That is, incommensurability - of particular rights with social welfare - and qualitative distinctions among values - between freedom of contract, for example, and other public policy interests - lay at the heart of the Lochner judicial framework.  [FN28] Not surprisingly, then, one weapon in *1532 an overall assault on this ideology was the effort to undermine the absolute character of "higher" legal values, an absolutism that completely subordinated all competing public poicy concerns to existing legal "rights."

  Insofar as these critiques influenced legal doctrine, their net effect was undoubtedly beneficial.  They opened the way to a legal system more accommodating of political responses to the economic transformations of the twentieth century.  But this is also the moment when the sense of incommensurable values in legal reasoning began to dissolve.  Realists did not necessarily conceive their work in terms of competing conceptions of values or as an all-out assault on the incommensurability of values, though Pound surely did.  And while the Lochner era might have been challenged in terms of the particular legal understandings central to it (how is freedom of contract best understood?), Pound raised the stakes and essentially attacked incommensurability per se.

  Whatever the general intent behind these shifts in legal thought, the result was a gradual decline in the role of qualitative reasoning [FN29] in law.  In legal practice, the concrete manifestation of this decline was the emergence of the balancing test, which began around 1910 and became dominant around 1930 in both public and private law. [FN30]  In place of qualitative reasoning, the balancing test suggested a quantitative weighing of clashing, equally legitimate, interests.

  The image of balancing, though, always remained ambiguous. Taken more literally, it might portray legal rationality as the identification of the value-maximizing outcome among competing interests, all conceived as being restatable in terms of some single, uniform, underlying value.  Taken more metaphorically, however, it might do no more than signal the relevant competing interests in a process of seeking some "accommodation" among them - a process of context-specific judgment that might leave room for qualitative distinctions. [FN31]  Whether "balancing" language was to be understood literally or metaphorically *1533 thus remained uncertain.  It was left to two contemporary perspectives on legal decision to push these early developments down the path that most directly challenges the view of legal values as incommensurable.

  The first of these perspectives is law and economics.  While significant variations exist among the methods and questions law-and-economics scholars address, a considerable segment of normative work in the field implicitly rests upon rejection of the view that values should be understood as incommensurable. Approaches that seek to resolve legal conflicts by recasting them as problems in maximizing some single value (whether it is preference satisfaction, utility, wealth, or efficiency) share this form.  It is easy enough to find broad statements suggesting this rejection, such as the audacious extremes to which Richard Posner, in his early work, was willing to go in intimating that justice might simply be equivalent to efficiency:  "Would the objection to medical experimentation on convicts remain unshaken were it proved that the social benefits of such experiments greatly exceeded the costs?" [FN32]  But the casual brazenness of statements like this suggests that their main aim is to taunt and tease, rather than to be taken too literally; and Posner himself has since retreated a bit from these histrionic claims. [FN33]  From the perspective of public policy, far more significant are the subtle, less provocative ways the rejection of incommensurability continues to form an implicit, unexamined foundation for myriad, microscopic applications of this approach.  For that purpose, I want to consider Louis Kaplow's intriguing effort to develop a unified theory, based in economic thought, for judging when those disadvantaged by changes in government policy ought to receive *1534 compensating relief. [FN34]

  This is the problem of legal transitions - the fair distribution of the burdens that inevitably accompany even desirable policy reforms - and has long been considered among the most difficult ethical and legal questions in a dynamic society. [FN35]  Kaplow insightfully notices how pervasive the problem is; it arises when tax laws are changed, the common law is altered, the government imposes new regulatory constraints on uses of property, or new policies interfere with the performance of previously arranged private contracts.  Traditionally, when these questions have arisen as a matter of legal doctrine (as in the question whether government interference with use or possession of existing property rights constitutes a taking, for which the Constitution requires just compensation), courts have asked whether reasonable reliance on the status quo or legitimate expectations of stability have been disturbed.  If so, government must respect the values of reliance and expectations and pay compensation for their violation.

  But as Kaplow understands these norms, they are incoherent and ought to be recast in terms of the value of efficiency.  As he says:  "economic analysis ... and ... criticism of appeals to reliance and expectations demonstrate that many of the usual justifications for these protections are without merit and that these protections are themselves inefficient."  [FN36]  Kaplow's thesis is that "uncertainty concerning government policy is analytically equivalent to general market uncertainty." [FN37] Viewed in this way, changes in government policy ought to be treated as simply one more potential risk that existing interest holders must factor into their calculus when making initial decisions as to which courses of action to pursue.  For the most part, we leave people to self-protection against these sorts of uncertainties, which they seek through private insurance markets and, where possible, by diversifying their holdings.  When government acts, Kaplow urges, we ought to do the same.  Moreover, doing so generates efficient incentives to future owners; when investments are next made, investors will know to discount the possibility of future changes in government policy into their initial investment strategies and willingness to pay.  The practice of government payment of compensation distorts efficiency-enhancing market signals.  Thus, judicial focus on reliance and expectations should be replaced with a focus on efficiency, understood as a scheme of risk allocation and incentive creation, in which individuals "bear all *1535 real costs and benefits of their decisions." [FN38]  In the absence of exceptional circumstances, private markets best provide that scheme.  In practice, this means no form of compensation should be granted or required in situations such as a local government's decision to ban all building on undeveloped land in order to preserve the views of existing homeowners.  In addition, courts should abandon many of the central distinctions that have emerged over many years of judicial effort to give content to the kinds of expectations and reliance entitled to protection.

  This is necessarily an extremely crude sketch of a subtle and nuanced argument.  But I think this summary does reveal fairly the way in which the basic structure of arguments like these, common to much economic analysis of law, display at their foundation the complete rejection of any notion of incommensurability.  Risk to existing holdings is treated as a single- dimensional variable. Risks may vary in degree, but all risks are treated as the same in kind and as having the same meaning to those burdened.  From the perspective of the disadvantaged individual, it makes (or, perhaps, ought to make) no difference whether a loss results from private market forces or public policy changes; whether the property in question is personal rather than commercial property, or, as Hobhouse put it, "property for use" rather than "property for power"; [FN39] whether justifications such as welfare- maximization or moral purposes are offered for the change in public policy; or whether the government has physically taken a parcel of property or only regulated its use.  According to this sort of analysis, all pose the same analytic problem.  As Kaplow puts it, " a  private actor should be indifferent as to whether a given probability of loss will result from the action of competitors, an act of government, or an act of God." [FN40]  The situations may vary in the degree of burden imposed, but these differences do not reflect any qualitatively significant distinctions among the types of settings in which public change occurs. [FN41]

  *1536 Note how qualitative distinctions get radically displaced in this kind of analysis. [FN42]  First, the very structure of the problem, as historically conceived, dissolves.  While legal doctrine and public policy have traditionally recognized a fundamental distinction between publicly imposed disruptions of existing distributions and private ones, in this analysis all sources of disruption become equivalent.  Second, a similar equivalence is created between contexts in which the byproducts of changes in public policy benefit as well as burden individuals. For following through on the implications of his analysis, Kaplow suggests that if government must pay compensation for losses associated with policy transitions, symmetry (that is, the erasure of contextual distinctions) *1537 requires that government be entitled to recoup benefits from individuals who indirectly profit from these transitions. [FN43]  Thus, a fundamental reworking of transition policy in general, and its application in specific contexts such as the takings area, is urged.  I do not seek to evaluate these changes for now, but rather to call attention to the way that rejection of values as incommensurable is ultimately the foundation for this sort of economic effort to displace traditional legal analysis.

  Perhaps surprisingly, a similar stake emerges in understanding certain strands of critical legal studies (CLS), the second important contemporary effort to pursue the challenge to incommensurable values that legal realism implicitly initiated. Particularly in early CLS work, which revived the realist insight that legal decisionmaking cannot rest solely on the internal elaboration of doctrine through a distinctly legal logic, the view of values as incommensurable came under attack.  As a representative example, consider Mark Kelman's effort to show that substantive criminal law doctrine rests on underlying "interpretive constructs" that, rather than internal "legal analysis" itself, determine the way criminal law categories are concretely applied. [FN44]

  According to Kelman, these constructs reflect the ways legal analysis implicitly characterizes, frames, or defines the (relevant) context of a social disruption.  For purposes of criminal law, "events" are not sets of hard, physical facts of objective data, which we first gather and then evaluate through the legal norms of criminal responsibility.  Instead, because multiple ways of characterizing facts are often available, an "event" can be given definition and meaning only through interpretive categories.  For example, Kelman points to how complex the often unnoticed element of time can be in criminal cases.  Depending on whether we place a narrower or broader time frame on events, different conclusions will follow regarding central criminal law concepts like voluntariness or mens rea.  Thus, whether an epileptic is guilty of negligent homicide, when his car kills a pedestrian after he has a seizure that renders him unconscious, depends on whether we define "the event" as the moment of collision (no voluntary action on defendant's part) or as the moment the defendant chose to drive the car knowing his risk of seizures (a voluntary action).

  As a result, Kelman argues that the qualitative distinction between voluntary and involuntary action, which is conventionally invoked in many social as well as legal settings, does no real work in resolving *1538 cases like this.  What is decisive is the initial characterization of the event: whether the interpretive template applied is a temporally narrow or broad one. Once this frame is fixed, conclusions as to voluntariness follow.  Kelman then observes that the background interpretive constructs pervasive in the criminal law are applied differently - or, as Kelman puts it, inconsistently - in different contexts.  For example, relatively broader time frames are implicitly invoked to permit certain kinds of defenses that look back beyond the discrete moment of "the criminal event," such as duress, subjective entrapment, provocation, or insanity.  But for purposes of other criminal law principles, such as the hostility to strict liability or the way voluntary act requirements are understood, much narrower time frames are used to define "the event."  As a result, Kelman asserts that background interpretive constructs of the criminal law, while decisive, are "arational"; rest on "nondefensible interpretations";  [FN45] and, most provocatively, cannot be justified because they reflect "unresolvable inconsistency." [FN46]

  Noticing that the application of legal concepts implicitly assumes certain background cultural understandings incisively reveals the potential contingency of existing legal approaches.  But this sort of critique also has several problems, which can be described in at least two different ways.  On one reading, Kelman can be understood to demand that there be some single conceptual framework that means the same thing in different settings and can be neutrally applied across all contexts to characterize how time, voluntariness, and other aspects of an event should be understood. With such a framework in place, criminal law would apply a "unitary" concept of time or voluntariness and insure that these concepts are treated "consistently" despite differences in the specific nature of an alleged crime.  Alternatively, we might read Kelman to be arguing that the background interpretive frameworks informing criminal law are themselves incommensurable and, for that very reason, are necessarily "indefensible," "arational," and "unresolvably inconsistent."  To link these two descriptions, we might say Kelman's assertion is that in the absence of a single, commensurable framework for describing all events, we are left to vacillate between competing characterizations, with no justifiable way of rationalizing our assignments of criminal responsibility. Indeed, Kelman argues that many conflicts in criminal law reduce to conflicts between more intentionalistic and deterministic accounts of action - and that the choice between these (in general or in any specific context) is a choice between fundamentally incommensurable social theories [FN47] *1539 (here we are dealing with what I earlier called "interpretive incommensurabilities of value" [FN48]).

  In terms of the concerns in this review, these views are all troubling because they seem to deny the possibility of understanding values as appropriately incommensurable in some contexts.  On the one hand, they suggest that the presence of incommensurable values makes rational choice between values impossible; on the other, they assume that rational attributions of criminal responsibility can be made only if we have a single commensurable interpretive framework that can be applied, mechanically, in all settings. This style of critique, then, threatens to turn into a comprehensive assault against incommensurable values.  To be sure, the identification of arational cultural constructs giving content to legal categories is not directly presented as this sort of attack.  Yet here, as in many places, intent matters less than effect.

  As a specific example, consider Kelman's analysis of the "imperfect self- defense" doctrine.  This doctrine defines circumstances in which a defendant's culpability is reduced because he is considered guilty of merely negligent, as opposed to intentional, conduct.  As Kelman sees it, applications of the doctrine veer arationally between broader and narrower definitions of the relevant "event."  He offers two allegedly contradictory examples.  In the first, the defendant has a sincere, but objectively unreasonable, belief that she is being attacked with deadly force, thus requiring her to use deadly force in self-defense; considerable authority supports treating this defendant as guilty of only negligent homicide. [FN49]  Kelman then contrasts the decision of military courts to hold Lt. William Calley guilty of intentional murder for his deliberate shooting of Vietnamese villagers.  Calley argued that he believed his actions were excused because his superiors had lawfully ordered the shooting; even if this belief were wrong, and the order had been unlawful, he argued that he should be guilty at most of negligent homicide.  The military courts rejected this defense, holding that Calley had committed intentional murder:  no justifiable excuse exists when "the superior's order is one which a man of ordinary sense and understanding would, under the circumstances, know to be unlawful ...." [FN50]

  Kelman suggests that we should see these cases as far more similar than the difference in outcomes suggests.  He argues that these differences can be reached only by breaking up the events and their temporal frame in radically different ways across the two contexts.  In the first, we unify the events into a single narrative so that the defendant's *1540 state of mind at the moment of killing is understood against the background of her earlier formed, negligent belief.  To justify the second outcome, Kelman argues we must break up the sequence of events and isolate Calley's state of mind at the moment of killing, at which point he acts deliberately to take life.  The judgment of intentional homicide, implies Kelman, treats Calley ahistorically, as if he had suddenly arrived on the scene just at that moment and decided to shoot the villagers.  The only "explanation" for the different results in the cases is that "[w]e unify when we want to account for but deny that we are looking at the background of an intentional act; we disjoin and focus on the 'second' incident when we want to obliterate the past altogether." [FN51]

  I suspect many readers will share my sense that there is nothing disconcerting about the different way criminal responsibility is understood in these two contexts.  Kelman assumes that a rational system of moral or criminal responsibility must define "an event" in the same way across all settings.  But persuasive reasons, themselves grounded in moral or instrumental concerns, might well exist for not doing so.  To decide whether such reasons are present, we must look at the way these background interpretive constructs are applied and debate whether they themselves embody certain normatively appropriate commitments - or whether they are subject to a more carefully targeted set of social and political (but not for that reason arational) critiques.

  Consider the reasons we might distinguish Kelman's imperfect self-defense cases (as with all classic criminal law problems, a significant literature exists in this area, and I will suggest only a brief intuitive justification). Initially, one case seems to involve mistakes of fact, the other mistakes of law, although from Kelman's description one cannot say for sure; but if so, this long-standing distinction in criminal law might itself embody persuasive justifications (Kelman does not address the distinction or whether convincing reasons support it).  But I do not want to rest on what some might consider this "technical" point (though the existence of persuasive reasons for the distinction would make the point itself a normative, rather than a "technical," one).

  For other, more important moral and political reasons are at stake.  The  Calley case can be understood to express the distinct duties criminal law ascribes to those who occupy a specific social role, that of a military subordinate.  What is importantly different about the two events is not how broadly or narrowly in time they are defined, but the reasons we might want to recognize distinct duties in the two settings.  After Nuremberg, sincere belief in the legality of orders authorizing the deliberate, avoidable killing of civilians is simply not an acceptable excuse for the avoidance of responsibility.  Criminal law often imposes *1541 distinct duties associated with specific social roles. [FN52]  Here, I think many will see convincing reasons - which themselves reflect important moral and policy concerns - for imposing the duty in Calley.  The excuse of "just following orders" is one with which we have all too much experience by now; in the late twentieth- century, the imposition of a duty not to commit war crimes and not to follow orders commanding them hardly seems difficult to justify.  In contrast, the social problem of actions motivated by negligent beliefs as to self-defense is significantly different, with its own distinct history, distribution of consequences, likelihood of being strategically invoked, and more.  These make for moral and policy distinctions that provide rational reasons for accepting and rejecting the negligently "mistaken sincere belief" defense in the two settings or, in Kelman's terms, for going back in time to take into account the source of the belief in one setting while refusing to do so in the other. Given the politically, pragmatically, and, most important, morally significant differences in the contexts, why would anyone consider it rational to treat them the same?

  Kelman's problems, here, I think are two.  First, he assumes implicitly that American criminal law must rest on Kantian conceptions of moral responsibility.  Criminal responsibility cannot turn on circumstances external to the freely willing subject, but must respond to actual mental states and objective conduct.  Calley is no different from the negligent self-defender, because at the moment they act, both believe, sincerely but wrongly, that their deliberate taking of a life is justified.  But this Kantian perspective is not and cannot be the foundation for criminal law. Responsibility (moral, civil, or criminal) is always imposed externally to reflect social and political judgments about the kinds of obligations individuals in particular communities owe to each other. [FN53]  As a result, criminal responsibility is not tied solely to internal states of mind and objective conduct, things which *1542 the individual can, in theory, control and choose; socially imposed standards of conduct necessarily play a significant role in defining responsibility.  These standards are what give content to the background interpretive constructs Kelman notices.  Second, recognizing that responsibility is imposed from the outside, based on social and political judgments of the obligations reciprocally owed (judgments that may change as public debate alters conceptions of duties, as in the wake of Nuremberg), does not make our sense of responsibility "arational" or "indefensible."

  To generalize the point, notice that this style of critique undermines the possibility of treating values as incommensurable. The qualitative distinctions central to treating certain values as hierarchically or radically incommensurable are reflected in just the distinctions between contexts that Kelman's critique seeks to undermine.  Some of the qualitative distinctions central to criminal law resonate through much of social and moral interaction, such as the distinction between acts and omissions, or between intentional and unintentional conduct, or between voluntary and involuntary action.  Within criminal law, these are the categories Kelman means to call into question.  But because this challenge is not aimed at transforming any specific application of existing doctrine, it becomes all-encompassing - within criminal law and in its broader implications.  If the existence of interpretive frameworks is, in itself, understood as "nondefensible," then qualitative distinctions, whose existence depends on framing contexts differently, cannot be sustained.  From a pragmatic perspective, distinctions like those between higher and lower values are meaningful only insofar as our practices embody them.  The way we do so is by embedding distinctions like this into the very interpretive constructs Kelman challenges.  In other words, Kelman may be right that qualitative distinctions, such as that between the wrongfulness of voluntary compared to involuntary action, take on color only against larger background interpretive frameworks.  But that alone provides neither an argument for abandoning all such frameworks (or any specific framework) nor for declaring them (or any specific one) unwarranted.  Until we look at specific contexts, and decide what purposes criminal law ought to serve in them, we cannot know whether the incommensurabilities reflected in practice ought to be retained.  The charge of "inconsistency" itself rests on an unexamined baseline (we might say an interpretive construct) in which the desirability of commensurability is unquestioned.  But we need reasons, which Kelman does not offer, for privileging commensurability in this way.

  As with the legal realism of Pound, if Kelman's CLS work is interpreted in historical context, it is easy to understand the motivation for what turns out to be a sweeping assault on incommensurable values.  Kelman's arguments are motivated by the view that existing legal categories and ideology contribute both directly and through their reifying *1543 tendencies to maintaining present injustices.  But rather than targeting specific legal understandings that might be thought to work this way, Kelman's emphasis on the arationality of "interpretive constructs" per se leads to a full-blown assault on incommensurable values. Just as Pound's rejection of Lochner led to the demand that all rights and interests be considered on the same plane, Kelman's resistance to libertarian liberalism generates a perspective that undermines all qualitative distinctions among values.

  To the extent Kelman and Kaplow can be taken to represent CLS and law and economics more generally, an interesting convergence thus emerges.  Both seem to start with the view that values ought to be commensurable; both appear to reject the recognition within legal norms of qualitative distinctions among values and contexts.  An even clearer example of this convergence emerges in Kelman's own rejection of socially conventional conceptions of causation when making legal attributions of causal responsibility.  According to Kelman, "Coase seems correct in arguing that distinctions between natural and legal harms are not particularly relevant in deciding what liability rules ought to govern." [FN54]  This comment comes in response to Richard Epstein's argument that legal norms should reject Coasian joint causation approaches because they are an "artifact" not consistent with what Epstein called "natural facts" [FN55] (what in a more pragmatic vocabulary might instead be called the conventions embedded in existing social practices). Kelman's response seems to suggest that conventional conceptions of causation, which entail qualitative distinctions among contexts, such as acting and omitting to act, are not just contestable in difficult cases, but irrelevant to the legal assignment of cause. Just like Kaplow's approach to transition policy, Kelman's approach to criminal law depends upon denying the importance of qualitative or moral distinctions between contexts.

  Yet CLS and law and economics seem so clearly motivated by different objectives that this convergence might seem (to advocates as well as observers) not just puzzling, but disconcerting. Moreover, despite their efforts to reject incommensurability, both approaches ultimately cannot avoid falling back on it.

  For example, when Kelman's critical eye shifts from criminal law to law and economics, he tellingly exposes numerous ways actual behavior reveals the importance of qualitative distinctions between contexts - contexts Coasian economists instead assume individuals will treat as equivalent. [FN56]  The distinctions Kelman observes are ones people *1544 make, but Coasian theory would reject, between categories such as spending of dollars already in hand (realized income) versus dollars to be received through some future course of conduct (opportunity income), or distinctions between events consumers psychologically consider closed and those they consider open (the consumer's disinclination in some settings to equate sunk costs and future costs).  But these distinctions parallel or sometimes reproduce at least some of the qualitative distinctions in criminal law - such as those between acts and omissions, or those that bound off an event as that specific event and no other - that Kelman's criminal law critique views as the product of arational interpretive constructs.  Indeed, what gives these constructs their structure is likely to be the very kind of psychological and social phenomenon, deeply embedded in existing practices, that Kelman in his critique of Coase identifies. [FN57]

  Similarly, Kaplow argues that individuals should largely protect their own reliance and expectation interests through private market mechanisms of insurance and diversification.  But this does not eliminate reliance and expectation questions so much as displace them one level up the ladder of generality:  Should individuals be able to rely on the existing insurance markets and the stability of the other holdings into which they diversify?  The efficiency analysis seems to produce an infinite regress and is thus ultimately indeterminate; it is always possible to argue that individuals should have discounted the possibility of further and further levels of change into their initial investment strategies.  If that prospect is troubling, reliance and expectation interests will have to be understood and vindicated on their own terms; otherwise, the circle cannot be broken.

  Perhaps we are witnessing something that might be called the "Formalization of Critique."  With respect to the Formalization of Law in the late nineteenth century, the very perfecting of legal formalism paradoxically proved its undoing.  As legal categories became more general and systematically organized around formalism's key concepts, this process exposed contradictions and tensions that made legal formalism more vulnerable.  Perhaps something similar is taking place with certain styles of contemporary scholarship:  the perfection of abstract systems of analysis, and their application across more areas of doctrine, may be laying bare their most troubling aspects.  If theorists like Kaplow and Kelman find it more difficult in practice to abandon incommensurability than their theoretical discussions suggest, perhaps we ought to reconsider the possibility that qualitative distinctions matter.

  *1545 This is where Martha Nussbaum's work reenters the story.  For through the contest between Plato and Aristotle over whether values ought to be understood as commensurable or qualitatively distinct, two central questions emerge.  First, just what kind of question is the question whether values "are" commensurable or not?  Second, what exactly is at stake for social practices, like legal decisionmaking, in the decision whether to regard certain values as radically or hierarchically incommensurable with others?  Putting incommensurability back at the center of postrealist legal debates requires that these questions be the first addressed.

 
III

 
  The first question might be interpreted as a metaphysical one. We could understand it as seeking after the "true" ontological status of values; as asking whether they are "in fact" commensurable or not. [FN58]  But surprisingly enough, Nussbaum suggests that Aristotle - and even Plato - instead conceived this question in terms that have a distinctive air of twentieth-century pragmatism about them.  Both ask the question in the following terms:  Which of the alternative ways of understanding values ought we to adopt in light of the personal and social consequences that would follow?  Notice that this question cannot be answered without embracing some at least thinly substantive understanding of the good (not available in interpretations of liberalism that make it dependent on comprehensive neutrality):  an understanding that, on certain questions, we are prepared to judge some kinds of lives to be more fulfilling than others for humans.  [FN59]

  For those interested in practical choice - in the concreteness of actual legal, political, and moral settings - this pragmatic approach puts the question of incommensurable values in its most urgent and interesting form. Pragmatism of this sort keeps reflection from becoming mired in semantic strategies of classification or in theoretical analyses so remotely abstract they fail to generate useful guidance for resolving practical conflicts. I consider this the central, though not fully appreciated, contribution of pragmatism:  clarifying the terms in which it is most useful to care about questions like whether values are incommensurable.  Framing this question in terms of how alternative answers to it will influence social practices and our self-understandings, *1546 rather than in terms of metaphysical truths, gives those of us interested in law and norms the most compelling reasons for caring about Kaplow's and Kelman's rejection of incommensurable values.

  That brings both Aristotle and Plato to the second question: What is at stake in acting on the view of values as commensurable or not?  For Plato, as perhaps for Kaplow and Kelman, the reason we should assume commensurability of values is that we can then lead a life free of certain disturbing pains and intolerable conflicts.  In a provocatively dramatic passage, Nussbaum quotes Plato as arguing that embracing a commitment to treating values as commensurable is necessary to "sav[e] our lives" (p. 106).  This may contain a bit of rhetorical flourish, but it points to the breadth of changes Plato believes - rightly, I think - would follow from understanding all values to be generally commensurable and acting accordingly.  Plato recognizes that doing so means more than simply adopting a particular method of decision. Instead, belief and emotion are connected here, for Plato seeks to show that adopting certain beliefs, such as commensurability of values, will transform the very passions and feelings we experience.  Certain beliefs are necessary to make certain emotions possible: fear, love, grief, regret (modern techniques of cognitive psychotherapy, ofcourse, rest on this very view).  Changing the relevant background beliefs will necessarily transform our experience in ways that alter or dissolve these feelings.  This leads to the profound pragmatic motivation for Plato's arguments that we ought to embrace commensurability. For what turn out to be at stake are not just decision strategies, but the textures of experience - indeed, of ourselves.

  The choice over how to conceive values could not be understood more deeply.  Nussbaum brings to the surface the way in which transforming emotional experience motivates Plato's aspiration to understand values as commensurable. Indecisiveness, weakness of will, and neurotic conflict will dissolve if we come to view values as commensurable.  Through this lens, every choice involves comparisons along a single dimension of value; however much existing social conventions and our own reactions lead us to feel and talk as if choices implicate higher and lower or radically distinct concerns, we should distrust this experience and treat the language of qualitative distinctions as deceptive.  Instead, we should interpret choices as always involving greater and lesser quantities of the same thing.  If we manage to adopt this view, these experiences will disappear; few people feel conflicted or suffer weakness of will when asked whether they want five dollars or five hundred.  Plato does not suggest that this is the way values and choices are actually perceived, nor does he suggest that experiences like conflict and weakness of will are nonexistent.  He recognizes the radically transformative aspect of his proposal:  that is its very point.  As Nussbaum observes, "[t]he most astonishing claim implied by this argument is that the acceptance of the qualitative singleness and homogeneity *1547 of all the values actually modifies the passions, removing the motivations we now have for certain sorts of irrational behavior" (p. 111).

  But the transforming scope of this vision is not limited to irrational behavior, or rather, not just to experiences we would willingly concede to Plato are irrational.  Some of the most disruptive, surprising, enabling emotional experiences - grief, love, regret - will also dissolve with the commitment to commensurability.  And this is for the good, according to Plato, for he locates these destabilizing experiences in the realm of the irrational. Intense passion depends upon a belief in the uniqueness of that to which the passion is attached.  Commensurability denies this uniqueness.  It asks that we try to see what we value in the particular as mere instantiations of the general; that we actively engage in emotional modification with the aim of coming to experience all that we value as a single, homogenous quality.

  Consider how adopting this belief would transform experiences like regret.  The sense of regret often accompanies making difficult choices that one nonetheless believes to be right; in certain settings, such as the choice between a life of contemplation and a life of action, it might well follow whichever choice is made.  The best explanation for regret is that the options are experienced as incommensurable, so that any choice involves a qualitatively distinct kind of loss.  Any outcome entails the sacrifice of some genuine value.  But if we could accept commensurability, making the right choice would always mean that we have maximized the single value common to all options. There will thus be no experience of sacrifice; we will have chosen more over less, an occasion not for regret but for celebration of our rationality.  Plato urges accepting commensurability precisely because this "proper stance" toward the world will save us not just from regret, but from grief at certain kinds of losses, the intensity of love, and other "disturbing" emotions.

  Plato is not naive about the kinds of change his vision requires.  He recognizes that beliefs cannot necessarily be altered through pure reflection alone; his implicit social theory acknowledges that ideas, social practices, and the economic organization of society reciprocally influence and constitute each other.  Transformations in specific institutions and forms of culture may be necessary before it is plausible that commensurability of values might be embraced.  Hence the proposals in the Republic for communal childrearing, elimination of private property (indeed, of a private realm altogether  [FN60]), and his hostility toward literature, which cultivates in form and substance the sense of uniqueness of persons, commitments, and values.  Carried to the extreme, this is what a thoroughgoing commitment to commensurability, *1548 not just at the margins, but across the entire range of value and choice, would mean.

  This original impulse to understand rational choice as maximizing some single value shared across seemingly diverse goods is thus differently motivated than in more contemporary analogues. In these modern analogues to Plato, such as Benthamite utilitarianism or law and economics, critics have often seen the overweening influence of scientific models of truth.  According to these critiques, what drives these modern versions of Plato's rationalism is the contemporary allure of the natural sciences; the effort to understand and guide human behavior wraps itself in epistemological models of truthseeking that imitate those thought to guide the natural sciences. [FN61]  Although one must be careful to avoid reading modern conceptions of science back into Plato's language, [FN62] Plato self-consciously acknowledges seeking a "science of measurement"; and Nussbaum argues that the allure of fifth-century scientific conceptions of rationality partly motivates Plato's philosophy (p. 107).  But the modern criticism reveals distinctively modern preoccupations; it assumes that methodological considerations alone, specifically imitative flattery of the methods of the natural sciences, are the principal motivation behind efforts to find a "science" to guide human choice.  It is therefore more interesting in some ways to discover that for Plato this desire itself was only instrumental, since less intellectual, more substantive concerns drove it.  The crucial Platonic perception is that rational choice, conceptions of value, and self-understanding are mutually linked.  Ultimately, Plato's aim was a way of approaching choice that would transform experience to diminish aspects he found disturbing. [FN63]

  This, then, is the underlying battleground in the question of how values are best conceived.  Aristotle frames the clash in similar terms, recognizing the quality of individual self-experience and the character of social relations to be the ultimate stakes involved in pragmatically choosing how values should be understood. In rejecting Plato's understanding of value and choice, these were the very concerns that led Aristotle to embrace the particular four elements that make up his alternative vision, sketched at the start of this review, of rational deliberation. [FN64]

  *1549 Aristotle began from the principle that it is the good for human beings, not some abstracted, idealized being (p. 391), that ethics should seek.  Hence his belief that moral inquiry should aim at interpretively grasping and helping to perfect the best aspects of existing practices, rather than striving to transform human beings into something radically different. From this belief followed his internal understanding of critique, one committed to rational criticism of existing practices from within. And this understanding, in turn, required that the phenomenology of choice - the way people actually experience values and decisions - be taken seriously.  Efforts to avoid the difficulties posed by treating values as sometimes incommensurable fail to acknowledge the way social relations and individual experiences depend upon appreciating values in certain ways.  Positing values to be commensurable, as Plato did, and then justifying this with a thinly sketched vision of some ethereal, wholly remade human being, should, in Aristotle's view, be exposed not as sophistication, but as a strategy of avoidance.  And as Nussbaum nicely elaborates upon Aristotle's words, "[e]vasiveness is not progress" (p. 60).

 
IV

 
  The danger of evasion is ultimately what is troubling about the arguments of Kaplow and Kelman and, through them, of certain strands in both law and economics and CLS.  Returning to the ancient conflict between Aristotle and Plato helps provide a language for suggesting what has been evaded, and for expressing why we ought to care.

  In dissolving norms of reasonable reliance and justified expectations into ones of efficient risk spreading, Kaplow's economic approach to legal transitions assesses the former norms in the metric of the latter.  Indeed, the analysis may go beyond treating these values as commensurable, for Kaplow's view actually seems to be that traditional norms like reliance and expectation are circular and have no substantive content.  But whether Kaplow is understood as reducing one set of values to another, rationalizing them in terms of some single higher value, or arguing that one set is empty, the same question remains.

  That question is not what an efficient transition policy should be.  It is, rather, what is at stake in conceptualizing this question as one of efficiency, or of efficiency alone.  We might say it is whether the choice of transition policies has cultural as well as economic consequences. [FN65]  By cultural consequences, I mean how government transition policy defines and creates people's social understandings concerning the nature *1550 of their relations to each other, to "property," and to government itself.  The decision whether to compensate is, in part, a decision about the nature of the underlying property right being asserted. Compensation expresses respect for the legitimacy of some claims; its denial expresses rejection of other claims of entitlement.  An important aspect of transition policy, therefore, is to give meaning to "property" itself.

  Consider an extreme example, in part because the extremes often most starkly reveal the elemental structure of legal problems.  When national abolition of slavery was first discussed during the Civil War, many Union policymakers, including Lincoln, [FN66] considered whether slaveowners ought to receive compensation for what was, under existing positive law, "property." By the end of the war, this suggestion was inconceivable, for emancipation had come increasingly to define the very meaning and purpose of the war and its sacrifices. [FN67]  Compensation would have been a moral outrage; the war had come to stand, in part, for the principle that slavery was wrong - and had always been so, the positive law notwithstanding.  The decision of what transition policy to adopt was thus not a technical, subsidiary issue, but central to defining the meaning of both slavery and the Civil War itself.

  This expressive role of transition policy is an inextricable aspect of all compensation decisions, including constitutional interpretations of the Just Compensation Clause.  These decisions reflect and create social understandings about which policy changes interfere with existing investments morally important enough to be considered "property."  That is part of the role of concepts like "reasonable reliance" and "distinct, investment-backed expectations," [FN68] and that is why, interpreted against a history of prior applications, they are not empty, or circular, or placeholders for some other concept, like efficiency.  But compensation decisions participate in more than simply defining property as an initial matter of formal entitlement.  Not all formally equivalent entitlements are the same; not all forms of "property" warrant the same legal treatment for compensation or other purposes.  Even after initial allocations of rights in property have been made, decisions about compensation for government action continue to express and define the qualitative character of different types of property (as well as the nature of the government's reasons for acting).  That, too, is the office of concepts like reasonable reliance and legitimate expectations.  Some property interests merit more security against public action than others; some reasons for government policy changes are qualitatively more compelling than others.  Even after formal entitlements have been defined initially, a central role of remedies for government's violation of them - of public compensation practices - continues to be *1551 giving content and meaning to those entitlements. [FN69]  That is not a role that purely economic approaches to transition policy sufficiently appreciate. [FN70]

  Decisions about how the costs of change should be distributed thus have two dimensions, not one.  These decisions will have specific instrumental consequences; they have immediate distributional consequences and will affect future investment decisions directly.  But they also express public attitudes toward the very consequences intentionally being brought about. These attitudes toward the intended or foreseeable effects of policy will also affect the relationship individuals perceive themselves as having with government.  We might say that public actions create social understandings in addition to producing outcomes, or that part of the outcome an action produces is the social meaning it has.  Here, as elsewhere, "thought, motive, and action make one another what they are." [FN71]  The world of meaning creation is certainly no less important than the world of production, for social relations are defined and sustained through the meanings that actions express.  But more significantly, these worlds should not be conceived as two separate domains; social understanding and productive processes, such as investment decisions, mutually shape and influence each other.

  This is no less true for public policy changes than other actions.  Transition policies must therefore be assessed partly in terms of how they will be perceived, what they will express to those affected, and what kinds of social and political relations they will help constitute.  *1552 In every setting, one aspect of public policy is the sort of understandings and relations we want to construct to evaluate and guide action in that setting. In some settings, we might want actors to understand rational choice, or fairness, or reasonable reliance, in terms of economic efficiency. Repeated commercial interactions between large-scale economic organizations of relatively equal economic power, or between such entities and the regulatory actions of the government, might be one such arena.  But in other areas, there might be convincing reasons we want these norms to be understood in intrinsic terms, or in terms other than economic efficiency.

  The questions this choice raises are not easy to address. Whether public policy explicitly addresses them or not, though, the resulting rules of the game will always shape our understanding of what our aims are and how we ought to pursue them.  These questions about the proper terms of interaction in different domains are, I believe, precisely the ones traditional doctrine attempts to get at through efforts to define and secure legitimate expectations and reasonable reliance.  And as a cautionary warning to the development of alternative approaches, these questions cannot be addressed by isolating any particular norm - be it reliance, equality, or fairness - and attempting to find its meaning standing alone. Actions take on meaning only against a larger web of historical practices and understandings. [FN72]

  Once these effects of transition policies are kept in mind, the possibility of profound conflicts between instrumental and expressive consequences must be appreciated.  Efficiency-enhancing public policies can be self-defeating if, through the meanings they express about government's attitude toward individuals, they undermine the conditions necessary for sustaining social cooperation and trust. [FN73]  David Hume *1553 recognized this long ago in making it central to his theory of property [FN74] and, hence, justice.  [FN75]  According to Hume, the primary purpose of rules of property is not maximization of net social production.  Instead, property rules protect security of expectations and stability for the purpose of creating and sustaining the conditions that make fruitful cooperative associational existence possible at all. Life in community, offering numerous types of advantages, depends upon rules of property that express commitment to sustaining social coexistence on terms of fair, mutually respectful cooperation. These rules reflect social conventions that have emerged as to what kinds of relationships people can expect from each other - and when government acts, what kinds of collective obligations and responsibilities define the relationship between individuals and government.  For Hume, that is all justice is:  the social conventions that need to be respected in order to sustain the psychological and sociological basis for continuing, healthy social coexistence. [FN76]  As Hume concluded, "'Tis very preposterous, therefore, to imagine, that we can have any idea of property, without fully comprehending the nature of justice ...." [FN77]

  If the dynamic effects of public policies regarding compensation are likely to be both cultural and economic, any sensible approach to transition policies will have to take both into account.  We might still conclude in the end that treating the relevant values as commensurable best reflects the appropriate concerns.  But this involves a choice that can sensibly be reached only with full appreciation of all the relevant consequences.  And as Aristotle and Plato, though opposed in vision, both recognize, the crucial choice is the very first one:  whether to treat the values in conflict, such as fairness and efficiency, as qualitatively the same.  Once that is done, the problem becomes the technical (though not unimportant) one of determining how to maximize the single value now understood to be at stake.  But a central concern for theories of rational choice must be deliberating over which problems ought to be treated as ones of single-dimension maximization.  That entails considering the way making values commensurable is likely to transform not only discrete decisional outcomes, but more significantly, the texture of our lives and social relations - as Plato hoped and Aristotle feared.

  Most troubling about Kaplow's theory are not necessarily the policy outcomes to which it might lead, but that these outcomes would emerge without these critical questions of value having been considered.  *1554 Of course, the debate between Plato and Aristotle arises in the context of a comprehensive choice to adopt one view of values or the other in general, while legal transition policy, in contrast, implicates the cultural consequences at the margins of treating particular values, like reliance and security of expectations, as commensurable with other, discrete values in specific contexts.  But there is no reason to think cultural consequences matter any less at the margins than economic ones.  By illuminating what is at stake in wholesale commitments to commensurability, Aristotle and Plato reveal what must be attended to in every setting where questions of competing values must be confronted.

  To note these considerations is not to prescribe any particular compensation practice.  My aim here is only to point out central consequences that any compensation practice ought to address.  But from this perspective, perhaps radical changes in existing compensation practices ought to begin with a presumption against them if an important point of these practices is to vindicate expectations generated by existing conventions.  To overcome this presumption, any new approach to compensation practices must take account of the cultural consequences it is likely to produce.  We need persuasive, pragmatic reasons for suddenly equating certain values that law and policy have long considered intrinsically incommensurable in important ways.

  The same can be said for Kelman's assertion that the background interpretive constructs that give content to criminal law norms are arational and nondefensible.  As a strategy for finding creative openings to undermine existing criminal law policies, there is much to be said for this style of argument.  By highlighting ambiguities in the way "the event" is framed temporally and spatially, Kelman's analysis identifies points of vulnerability on which advocates or policy reformers might seize to press for creative change.  But by asserting that these background cultural or moral understandings are arational, Kelman's arguments undermine the foundation for treating values as incommensurable. [FN78]  Unlike Kaplow, Kelman does not offer any specific alternative policies that might follow from his implicit rejection of the incommensurability of values.  But even though Kelman's rejection is not as explicit as Kaplow's, it is more sweeping; Kelman believes the distinctions that pervade substantive criminal law rest throughout on nondefensible interpretive frameworks that, in effect, mark out certain values as incommensurable with others.  And as with Kaplow, it is troubling that Kelman does not address the social and political consequences of assuming that values ought generally to be *1555 commensurable. [FN79]

  The concerns that Aristotle expresses about abandoning action-guiding understandings concerning the incommensurability of values ought to be especially disturbing for CLS scholars.  Much of CLS work self-consciously aims to revitalize faith in collective action to redress existing inequities. Kelman himself remarks that one of the principal goals of CLS is to counter the tendencies of conservative law and economics by establishing that progressive reforms are capable of bringing about their intended consequences: "Resuscitating the near-instinctive sense of outrage at gross inequality, selfishness, and the glorification of anticommunitarian exclusiveness requires faith that efforts to rectify these injustices are not, a priori, fanciful and unreasonable." [FN80] But this resuscitation is likely to require more than belief in the instrumental efficiency of public policy.  It might require, as well, faith in the moral baselines upon which this very sense of outrage and injustice frequently depend.  As Nussbaum's revival of Aristotle suggests, it might require conviction that certain values are sometimes not to be treated as commensurate with others.

 
V

 
  The tensions described in this review are not new, nor have they gone unnoticed in legal scholarship.  The conflict between treating values as commensurable or recognizing qualitative distinctions among them - hierarchical and radical incommensurabilities - resonates with conflicts others have described in different terms. Some years ago, for example, Bruce Ackerman, seeking to untangle "takings" jurisprudence, discerned "two fundamentally different ways of thinking about law, each of which has roots in our present legal culture." [FN81]  Ackerman called these the views of the Ordinary Observer and the Scientific Policymaker and diagnosed legal culture as schizophrenically *1556 riven between them. [FN82] I, too, think contemporary legal and policy analysis is divided between competing paradigms of thought, and I locate the foundation of this divide in competing ways of conceiving values and "rational" choice.  But Ackerman saw the alternative frameworks he described as fundamentally incompatible and noncomparable; his principal aim was to unlock the structure of this conflict, rather than to offer a basis on which it might be resolved.  By focusing on incommensurable values and the social relations that different conceptions of value construct, I have tried to develop the stakes in this choice between competing conceptions of value and to suggest the considerations in terms of which this choice ought to be made in every context the conflict arises.

  Similarly, arguments against permitting market transactions in certain goods, such as reproductive capacity when traded through transactional structures of paid surrogacy relationships, share certain features with the approach I have suggested here.  These arguments, which Margaret Radin has developed most extensively in legal theory, are often framed in terms of avoiding the "commodification" of certain goods or making them "market-inalienable."  [FN83]  On this view, the social relationships within which goods are produced, exchanged, or distributed affect the way in which we understand and appreciate those goods; mechanisms of production, exchange, and distribution do not stand wholly outside the goods involved and are not neutral with respect to them.  Instead, in the terms I have used here, these mechanisms have an expressive dimension and influence the way we understand how the culture we participate in values different goods.  Legal norms that refuse to permit people to sell parts of their bodies, or to sell themselves into slavery, are one way we manifest the difference between widgets and persons.  We value the former in terms of exchange and use value; we value the latter by respecting the dignity and integrity of distinct persons.  Barriers against commodification create different ways of valuing different goods and, in doing so, seek to define qualitatively distinct, incommensurable values that characterize how we aspire to treat different goods.

  *1557 Arguments of this sort against universal commodification thus reflect concerns similar to those I have identified about whether values ought generally to be treated as commensurable. [FN84]  But concerns about commodification are a subset of the universe of concerns I have tried to describe. Critics of commodification focus on one specific hierarchical incommensurability they seek to maintain:  that between the domain of the market and the domain of the personal, or between the values expressed through market-exchange and gift-exchange.  They seek to preserve the particular boundary between the market and its ways of valuing from other spheres of social interaction. [FN85]

  By contrast, the emphasis on incommensurable values asserts a more pervasive value pluralism.  Social interaction and public policy, at their best, implicate a complex set of distinct values that cannot be captured in a single, sharp dichotomy between market and nonmarket realms.  Focusing on the broader commitment to incommensurable values extends the concerns underlying anticommodification arguments in two directions.  First, this focus offers a more expansive view of the number of distinct domains of value we might choose to recognize.  For example, one wayof understanding constitutional rights and of reconstructing their foundation in the postrealist world is to view rights as incommensurably higher values than ordinary public policy objectives.  On the view of incommensurability I have sketched here, this would not mean denying all tradeoffs between rights and public policies, but rather recognizing that we have to deliberate about the meaning of specific rights as intrinsic objects of value; conflicts between rights and policy should be resolved by trying to interpret the right and respect its integrity. Similarly, we need to consider whether government action implicates certain values differently than does private action, and, if so, whether certain of those values - such as reliance and fairness - ought to be understood as incommensurable with other relevant concerns, such as insuring efficient allocation of private investment resources.  In many domains beyond the market and the personal, we might believe it important to draw boundaries between distinct types of values.

  Second, this focus on incommensurable values expands upon commodification *1558 concerns because it highlights how complex interaction is even within supposedly unitary domains like "the market."  Within the sphere of commercial exchange, we might still recognize (positively or normatively) distinctions between higher and lower values:  honesty, integrity, or intrinsic norms of professional duty and fair dealing are often values we want to respect in commercial relations without reducing the values these norms embody to the marketplace's dominant values associated with preference satisfaction.  For both private and public decision, we need to be reminded of the multiple, perhaps incommensurable, values that ought to inform deliberation and action.  [FN86]

  In this review, I have tried to suggest that many existing practices, social or legal, are imbued with commitments to treating values as incommensurable.  I have also argued that competing ways of conceptualizing situations of choice - whether the values involved ought to be seen as commensurable or not - must be resolved partly in terms of the kinds of social and political relations that the choice will build or erode.  And I have wondered whether legal respect for norms - such as fairness, reliance, security of legitimate expectations, or voluntariness - may be necessary to sustain the social foundations of community, social foundations that are preconditions to our ability to pursue collective goals, such as enhancing net social production or changing the economic circumstances that make crime more likely.  Focusing on the specific types of collective relations that are created and sustained by treating values as incommensurable provides a way of understanding why certain legal norms, like fairness or reliance, might best be understood in terms independent of values such as efficiency.

  Martha Nussbaum notes that in Plato's Laws, the Stranger asserts that one of the barriers to desirable political practice is that many people lightly believe different values to be commensurable when they really are not so (p. 123).  He calls this failure to confront the issue of commensurability "a condition not human but more appropriate to certain swinish creatures," and declares himself "ashamed not only on my own behalf, but also on behalf of all Greeks." [FN87]  If that condition threatens important aspects of contemporary legal thought, perhaps it is time to confront in more depth how we should best understand values and what it means to make "rational" choices among them. *1559 Martha Nussbaum's work not only poses the right questions, but with compassion and insight tellingly reveals just how much is at stake.

 
[FNa]. Assistant Professor of Law, University of Michigan.  A.B. 1979, Princeton; J.D. 1983, Harvard. - Ed.  My gratitude to Elizabeth Anderson, Guyora Binder, David Charny, Heidi Feldman, Don Herzog, Greg Keating, Larry Kramer, James Krier, Jeff Lehman, and Don Regan.

 
[FN1]. University Professor and Professor of Philosophy, Classics, and Comparative Literature, Brown University.

 
[FN2]. See ALISDAIR MACINTYRE, AFTER VIRTUE 238 (1981).  In other work, Nussbaum argues that applying Aristotle's political thought in contemporary circumstances supports a kind of politics associated with European-style social democracy.  Martha C. Nussbaum, Aristotelian Social Democracy, in LIBERALISM AND THE GOOD 203 (R. Bruce Douglass et al. eds., 1990).  She also distinguishes the substance of her Aristotelian liberalism from the kind of contemporary liberalism, grounded in Kant, that finds expression in the work of theorists like John Rawls.  See, e.g., JOHN RAWLS, A THEORY OF JUSTICE (1971).  Two main points of distinction are Nussbaum's views that political theory must rest upon some relatively "thick" conception of what is good for human beings and that this good must not be understood solely as a matter of abstract moral reasoning powers, but in terms of the institutional, material, and educational resources needed to make possible this substantive conception of human flourishing. Martha C. Nussbaum, Non-Relative Virtues:  An Aristotelian Approach, in XIII MIDWEST STUDIES IN PHILOSOPHY 32, 38-39 (Peter A. French et al. eds., 1988) [hereinafter Nussbaum, Non-Relative Virtues].

 
[FN3]. Reducing Nussbaum's vision to a set of axioms, while necessary for a brief review, clashes with the spirit of her book, which seeks to establish its claims not primarily through deductive axioms subject to logical "proof" but by exemplifying a certain sensibility that it implicitly asks readers to experience before judging.

 
[FN4]. See MARTHA C. NUSSBAUM, THE FRAGILITY OF GOODNESS 240 (1986), where Aristotle is quoted as offering an account of philosophical method not too distinct from Rawlsian reflective equilibrium:
    Here, as in all other cases, we must set down the appearances and, first working through the puzzles, in this way go on to show, if possible, the truth of all the beliefs we hold about these experiences; and, if this is not possible, the truth of the greatest number and the most authoritative.  For if the difficulties are resolved and the beliefs are left in place, we will have done enough showing.
Id.  (translating ARISTOTLE, NICOMACHEAN ETHICS book VII at 1145b).
  The proper scope of this internal criticism, that is, what it takes as its object, makes Aristotle more of a universalist than those contemporary pragmatists who believe political justification can only meaningfully take place within the practices and norms of particular communities.  Although Aristotle's views are too complex to capture here, he believed simultaneously in a single objective account of the human good and in a virtue-based ethics according to which virtue is relative to the social conventions that define appropriate conduct in diverse spheres of human activity.  See Nussbaum, Non- Relative Virtues, supra note 2, at 39.

 
[FN5]. Pp. 54-105.  To familiar charges that this sort of contextual approach to decisionmaking runs the danger of lapsing into an "absolutely empty 'situation ethics,"' see, e.g., HILARY PUTNAM, REALISM WITH A HUMAN FACE 194 (James Conant ed., 1990), Nussbaum responds with arguments analogous to Ronald Dworkin's conception of law as internal interpretive reconstruction:
    [T]he perceiver who improvises morally is doubly responsible: responsible to the history of commitment and to the ongoing structures that go to constitute her context; and especially responsible to these, in that her commitments are forged freshly on each occasion, in an active and intelligent confrontation between her own history and the requirements of the occasion.
[p. 94]

 
[FN6]. Two qualifications of what follows are in order.  First, viewing Plato as committed to conceiving values as commensurable depends heavily upon arguments of Socrates in The Protagoras. But as Nussbaum acknowledges elsewhere, in later dialogues Plato distances himself somewhat from these arguments.  NUSSBAUM, supra note 4, at 121; see also TERRENCE IRWIN, PLATO'S MORAL THEORY:  THE EARLY AND MIDDLE DIALOGUES (1977).  Second, one must be careful not to assume that the epistemological understandings of "science" that animated Plato's aspiration for a "real science" of ethical reasoning are the same as those that characterize the contemporary natural sciences.  Apart from the dangers of anachronism, there are linguistic issues of concern as well. The Greek word techne, which Plato uses, Nussbaum variously interprets as "real science" (p. 107); "an orderly systematization of practice in some area that would yield increased control over the ungoverned aspects of human existence" (p. 107); and "human art of science," NUSSBAUM, supra note 4, at 89.  But Nussbaum does argue that in late fifth-century Greece, the time of Plato's youth, the aspiration to scientific knowledge, conceived as the ability to number, order, and measure, exerted powerful intellectual and social influence.  Id. at 89-121.

 
[FN7]. See, e.g., Antonin Scalia, The Rule of Law as a Law of Rules, 56 U. CHI. L. REV. 1175, 1176, 1182 (1989) (enlisting Aristotle in support of rule formalism); see also, e.g., Morrison v. Olson, 487 U.S. 654, 733 (1988) (Scalia, J., dissenting) ("Today's decision ... is ungoverned by rule, and hence ungoverned by law.").  Beneath these kinds of claims might be either of two distinct, somewhat competing, accounts of the historical development of law.  In one version, law is progressively rationalized from the personal exercise of equitable discretion (Justice Scalia offers the example of Louis IX of France, rendering justice under an oak tree after Mass) to a body of highly articulated, public standards knowable in advance of decision and enforced through the impersonal operation of an institutionalized bureaucracy.  On the second account, at some more recent moment, such as the rise of the twentieth- century welfare and regulatory state, law "reverted" to more ad hoc, discretionary, open-textured and nonrule-like forms.  For debate over whether the welfare state and the rule of law are incompatible, see FRIEDRICH A. HAYEK, THE ROAD TO SERFDOM (1944) and Harry W. Jones, The Rule of Law and the Welfare State, 58 COLUM. L. REV. 143 (1958).  For a historical sociology of law that incorporates both these accounts, see ROBERTO M. UNGER, LAW IN MODERN SOCIETY (1976).

 
[FN8]. Justice Scalia believes that adhering to Aristotle "is a pretty good place to stand," and states "I stand with Aristotle" in the view that legal norms should be understood and applied as a system of rules.  Scalia, supra note 7, at 1182.  But the central theme of Nussbaum's work is that Aristotle rejected this view of law and ethics and stood for precisely the opposite understanding.  Thus, Aristotle, as quoted by Nussbaum, argued that indeterminacy was the essence of practical affairs and that "[t]he error is not in the law or in the legislator, but in the nature of the thing, since the matter of practical affairs is of this kind from the start," (p. 70); as a result, he argued that one who makes decisions according to antecedently fixed general principles not reinterpreted to fit specific contexts is like an architect who tries to use a straight ruler on the curves of a fluted column. Instead, says Aristotle, he should do what the builders of Lesbos did:  measure with a flexible strip of metal, called the Lesbian Rule, that "bends to the shape of the stone and is not fixed." Id.  With respect to statutory interpretation, Aristotle is known for rejecting the kind of plain meaning approach of Justice Scalia and, instead, advocating an equitable method of interpretation (epiekeia) in which the judge is to elaborate statutes in terms of their underlying purposes and more general equitable considerations.  P. 69; see Note, Intent, Clear Statements, and the Common Law:  Statutory Interpretation in the Supreme Court, 95 HARV. L. REV. 892, 896 & n. 35 (1982) (discussing Aristotelian concept of epiekeia).

 
[FN9]. See, e.g., FREDERICK SCHAUER, PLAYING BY THE RULES:  A PHILOSOPHICAL EXAMINATION OF RULE-BASED DECISION-MAKING IN LAW AND IN LIFE (1991) (Reviewed in this issue by Professor Mark V. Tushnet. -Ed.); John Dewey, Logical Method and Law, 10 CORNELL L.Q. 17 (1924); Duncan Kennedy, Form and Substance in Private Law Adjudication, 89 HARV. L. REV. 1685 (1976); Duncan Kennedy, Freedom and Constraint in Adjudication:  A Critical Phenomenology, 36 J. LEGAL EDUC. 518 (1986).

 
[FN10]. See MICHAEL WALZER, SPHERES OF JUSTICE (1983).

 
[FN11]. Compare RONALD DWORKIN, What Justice Isn't, in A MATTER OF PRINCIPLE 214 (1985) (arguing that Walzer's interpretation of existing conventions cannot provide a legitimate justification for political and moral choices) with RONALD DWORKIN, LAW'S EMPIRE (1986) (justifying law as an "interpretive concept"). Dworkin does now assert that the political and legal theory judges use to bring interpretive coherence to the existing, authoritative primary legal materials must be evaluated along two distinct dimensions of "fit" and "substance." "Fit" reflects the commitment to bringing integrity to the system of legal norms by rendering them, to the extent possible, internally coherent; but "substance" reserves some space for rejecting outcomes that internal reasoned elaboration alone would produce when those outcomes violate some external conception of justice, validated in some way independent of existing legal norms.  But this escape clause for substantively unjust outcomes does no significant work in Dworkin's examples or the application of his theory, and Dworkin does not address the possible tension between outcomes that "fit" and "substance" might recommend.  Internal integrity, or fit, is thus clearly the primary focus of Dworkin's conception of the morality of legal decision, at least in contemporary England and the United States.

 
[FN12]. See, e.g., RICHARD RORTY, CONTINGENCY, IRONY, AND SOLIDARITY (1989); RICHARD RORTY, PHILOSOPHY AND THE MIRROR OF NATURE (1979).

 
[FN13]. Ultimately, the turn in legal scholarship to historical narrative or "storytelling," see, e.g., Symposium, Legal Storytelling, 87 MICH. L. REV. 2073 (1989), rests on implicit claims about the legitimate causal connection between emotions and justified beliefs, or the absence of a sharp distinction between these realms of experience, so that the evoking of "emotional" responses to depictions of social experience (particularly those not within the self-experience of the average consumer of legal scholarship) makes the resulting changes in belief normatively justifiable.  But though this connection between emotions and justified beliefs is critical to both the psychology and the normative grounding for these developments in scholarship, legal scholarship has just begun to explore the connection.  For one example, see Martha L. Minow & Elizabeth V. Spelman, Passion for Justice, 10 CARDOZO L. REV. 37 (1988).

 
[FN14]. CYNTHIA OZICK, METAPHOR & MEMORY 270 (1980).

 
[FN15]. As James Conant has put it, we seem to be "at a point in the history of our culture when so many of philosophy's official practitioners have come to accept the idea that compromising their original sense of excitement and hope is simply an inevitable part of the cost of the professionalization of their subject."  James Conant, Introduction to PUTNAM, supra note 5, at xv, 1xxiv.

 
[FN16]. Among the tools of modern social science Nussbaum attacks is social choice theory, the most prominent development of which is known as Arrow's Theorem.  Pp. 64-65.  Social choice theory studies the formal properties of collective decisionmaking processes; Arrow's Theorem suggests that these processes, such as democratic politics, cannot simultaneously produce "rational" outcomes and follow minimally "fair" decision procedures.  See KENNETH J. ARROW, SOCIAL CHOICE AND INDIVIDUAL VALUES (2d ed. 1963); Kenneth J. Arrow, Value and Collective Decision-Making, in PHILOSOPHY AND ECONOMIC THEORY 110 (Frank Hahn & Martin Hollis eds., 1979); cf. Richard H. Pildes & Elizabeth S. Anderson, Slinging Arrows at Democracy:  Social Choice Theory, Value Pluralism, and Democratic Politics, 90 COLUM. L. REV. 2121, 2124 (1990). Nussbaum is right to note that the formal models employed in social choice theory lose their explanatory power if we view values as incommensurable in the Aristotelian sense.  But I believe she wrongly locates the point at which the field of social choice theory becomes incompatible with the view that values ought be treated as incommensurable.
  Nussbaum believes this incompatibility arises when social choice theorists demand that collective decision processes, to be fair, satisfy a condition known as the "independence of irrelevant alternatives."  This condition states that how an individual ranks two options against each other should depend only on the relative value (or disvalue) of those two options, and not on any other possible options.  The independence condition states that the way in which a person compares the value of any two particular options is not affected by the presence or absence of other options; if I prefer A to B, then when choosing between those two I ought always to choose A over B, no matter what other options might be around.
  Nussbaum suggests that conceiving of rational choice in this way asks us to deny the reality of experiences like tragic choices, in which, no matter what we do, we must violate some principle or obligation we hold dear.  She believes that "a consequence" of the independence condition is that the decisionmaker must "regard it as irrelevant that all the available options are hideous by comparison" to what would be possible under other circumstances.  P. 64.  But this is not what the independence condition requires or seems likely to produce.  The condition does not entail anything about whether the options we are choosing among should be viewed as good or bad, whether our decision will generate considerable benefits or minimize horrific losses, or whether we view ourselves as maximizing utility or minimizing disutility.  The condition only requires that if A is a better choice than B, all things considered, then A ought to remain better than B regardless of whether other possible options (C, D, E) are present or not. The relative ranking of A and B should be unaffected by the presence or absence of other possibilities.  Pildes & Anderson, supra, at 2132 n.34.  There are reasons to believe that the independence condition is not an appropriate condition to demand of rational decisionmaking, and many actual collective decision procedures (such as logrolling) violate this condition.  But Nussbaum's concern - that the condition will hide from us the fact that in some circumstances all our choices are bad - is not among these reasons.
  Nussbaum's instincts are right, though, because a different but central tenet of social choice theory is fundamentally inconsistent with the Aristotelian view of values as incommensurable.  This is the way the theory (and economic theory more generally) defines "rational" choice:  choices are "rational" only insofar as they demonstrate consistency, in the sense that if I prefer A to B, and B to C, I must also prefer A to C.  In these theories, rationality must embody transitivity among choices.
  But transitivity is a coherent demand only on the assumption that values are commensurable.  If the values in options A, B, and C can be reduced to a single value that can be compared along a common metric, then the failure to choose A over C would be irrational in that we would be failing to maximize value in our choice.  If I am asked to choose the tallest person in a group, then if I conclude Frank is taller than Bob and Bob taller than David, I must conclude that Frank is also taller than David.  But if we perceive the options to involve incommensurable values, the requirement of transitivity loses its relevance.  A, B, and C would then involve qualitatively distinct values, and rational choice would mean respecting the intrinsic values among different options, rather than maximizing some single value.  If admissions decisions among students with similar records were made through pairwise comparisons, a law school might rationally prefer a student who will diversify the student body to one with great academic promise, the one with academic promise to one with public service potential, and the one with public service potential to the one who would diversify the school.  Because there is no single value defining "the best" potential students from among those otherwise competent, rational choice here does not need to embody transitivity.  Thus, the constraints that social choice theory imposes on collective decisionmaking must be rejected if the Aristotelian conception of values is accepted, not because the independence condition is inconsistent with this conception, but because the very definition of rationality in the models would be flawed.  For an extended critique of the view that transitivity ought to be viewed as a necessary condition of rational choice, see id. at 2145-66.

 
[FN17]. I purposefully choose intentional discrimination to avoid more complex concerns about how discrimination more generally should be understood. With respect to the core concept of discrimination, we define it by debating the moral character of the relevant practice in question, rather than by asking whether it is efficient.  Thus, we neither permit discrimination where it is efficient nor generally define it as only those practices that are inefficient. For example, use of some criteria as proxies for others, such as educational level for predicted workplace competence, often reflects efficient stereotyping or rational statistical generalization; but where race is the proxy, its use in similar fashion is proscribed even where doing so might be equally "efficient" generalizing. See generally CHRISTOPHER JENCKS, RETHINKING SOCIAL POLICY: RACE, POVERTY, AND THE UNDERCLASS 40-46 (1992) (distinguishing five kinds of economic discrimination).  It is possible that the outer periphery of what constitutes discrimination - the cases at the current margin of debate - might properly be defined in terms of inefficient employment practices.  But even were that so, it would neither (a) suggest the core concept of discrimination is to be understood in similar terms nor (b) explain why we single out this particular inefficiency for legal prescription.  Thus, when current antidiscrimination law is challenged as being inefficient in theory, see Richard A. Epstein, Two Conceptions of Civil Rights, 8 SOC. PHIL. & POLY. 38 (1991), theoretical counterargument that antidiscrimination law might promote, or not be inconsistent with, efficiency is appropriate.  See Cass R. Sunstein, Why Markets Don't Stop Discrimination, 8 SOC. PHIL. & POLY. 22, 23 (1991).  But it remains important to remember that these ought not be the ultimate terms in which to understand what is wrong with discrimination.

 
[FN18]. This conception of incommensurable values is developed in 2 CHARLES TAYLOR, PHILOSOPHICAL PAPERS 230-48 (1985).  "It is this dimension of qualitative contrast in our moral sensibility and thinking that gets short shrift in the utilitarian and formalist reductions."  Id. at 240.

 
[FN19]. This type of incommensurability is explored in JOSEPH RAZ, THE MORALITY OF FREEDOM 321-57 (1986).  Radical incommensurability does not mean that the options cannot be compared in any way at all, but rather that the comparisons along common dimensions cannot be commensurated into a judgment of "total worth."

 
[FN20]. See DANIEL BELL, THE END OF IDEOLOGY (1960).

 
[FN21]. This is the conception of incommensurability that, presented in an extreme form, is the heart of Alisdair MacIntyre's nostalgic critique of modernity.  MACINTYRE, supra note 2, at 62-78. MacIntyre argues that collective moral discourse and rational choice is impossible in conditions of modern societies precisely because people interpret conflicts against a background of distinct, rival normative traditions that are themselves incommensurable.  For MacIntyre, the absence of a single, unified, shared moral tradition - what less charitably might be called a totalistic moral environment - in modern pluralistic societies makes moral conflict pervasively unresolvable.  One need not accept the view that insurmountable differences in interpretation are distinctly pervasive in modern societies to acknowledge their relevance to some issues, such as, perhaps, abortion (which MacIntyre appears to treat as the characteristic issue of modern politics).
  To the extent we find ourselves riven with internal conflict over the very definition of what should be recognized to be at stake when facing certain choices - conflicts stemming from the claims of different normative traditions (religion, political liberalism, the self-understandings of subcultural groups) on our identity and loyalty - the problem of interpretive incommensurabilities might exist within individuals as well.
  For a lengthier development of several different ways to understand claims that values are incommensurable, see Pildes & Anderson, supra note 16, at 2145- 66.

 
[FN22]. The article was first published in 1921 as Roscoe Pound, A Theory of Social Interests, 15 PAPERS & PROC. AM. SOC. SOCY. 16 (1921), but received significant attention among legal scholars when it was revised and reprinted much later as Roscoe Pound, A Survey of Social Interests, 57 HARV. L. REV. 1 (1943) [hereinafter Pound, Survey].  The earlier date of Pound's initial pursuit of these themes reflects the centrality of rights critiques to the legal realism of the 1920s.

 
[FN23]. Pound, Survey, supra note 22, at 2.

 
[FN24]. Id.

 
[FN25]. Oliver W. Holmes, Privilege, Malice, and Intent, 8 HARV. L. REV. 1 (1894).

 
[FN26]. Id. at 8.

 
[FN27]. Not surprisingly, given the nascent state of legal realism at the time it was written, Holmes' article actually occupies a more ambivalent position than often recognized on the question of incommensurable values.  Even as he argued that legal decision involved only questions of degree and conflicting interests, Holmes resisted a thoroughgoing utilitarian ethic and, in passages like the following, continued to recognize qualitative distinctions among values:  "The gratification of ill-will, being a pleasure, may be called a gain, but the pain on the other side is a loss more important.  Otherwise, why allow a recovery for a battery?"  Id. at 5 (emphasis added).

 
[FN28]. As is most clearly described in Duncan Kennedy, Toward an Historical Understanding of Legal Consciousness:  The Case of Classical Legal Thought in America, 1850-1940, 3 RES. L. & SOC. 3, 10-11 (1980), and implicit in Cass R. Sunstein, Lochner's Legacy, 87 COLUM. L. REV. 873 (1987).

 
[FN29]. The essentially qualitative orientation of prerealist, late nineteenth-century legal consciousness has also been variously described as an "on-off," or "spheres of action," or "deductive," or "boundary"-oriented conception of law.  See generally Al Katz, Studies in Boundary Theory: Three Essays in Adjudication and Politics, 28 BUFF. L. REV. 383 (1979); Kennedy, supra note 28; Gary Peller, The Metaphysics of American Law, 73 CAL. L. REV. 1151, 1199 (1985); Roscoe Pound, Liberty of Contract, 18 YALE L.J. 454 (1909).

 
[FN30]. See, e.g., Pennsylvania Coal Co. v. Mahon, 260 U.S. 393, 413-16  (1922) (Holmes, J.); International News Serv. v. Associated Press, 248 U.S. 215, 247-48 (1918) (Holmes, J., concurring); Hudson County Water Co. v. McCarter, 209 U.S. 349, 355 (1908) (Holmes, J.).  After emerging in this period, the balancing test came to dominate constitutional decisionmaking by the late 1930s.  See T. Alexander Aleinikoff, Constitutional Law in the Age of Balancing, 96 YALE L.J. 943, 948 (1987).

 
[FN31]. See, e.g., Steven Shiffrin, The First Amendment and Economic Regulation:  Away from a General Theory of the First Amendment, 78 NW. U. L. REV. 1212, 1249 (1983) ("[B]alancing is nothing more than a metaphor for the accommodation of values.").

 
[FN32]. RICHARD POSNER, ECONOMIC ANALYSIS OF LAW 23 (2d ed. 1977).  Read closely, this specific rhetorical question of Posner's might be taken to suggest not that values such as dignity or bodily integrity and efficiency are fully commensurate but that these values be viewed as lexically ordered in a qualified way.  Two goods are lexically ranked if no amount of the higher good will be traded for any amount of the lower good.  In a more complex version of this concept, Rawls argues, for example, that basic rights should be understood as lexically prior to income but only after a certain threshold level of income is reached.  RAWLS, supra note 1, at 542.  Posner's formulation - that the benefits must "greatly exceed" the costs, rather than simply being greater - might suggest that bodily integrity should have a qualified lexical priority over net social welfare.  But the clear import of examples like these, in the context of Posner's work as a whole, is to suggest greater commensurability of values than existing legal and social norms recognize.

 
[FN33]. In the most recent edition of his treatise, Posner deemphasizes his efficiency argument regarding forced participation in medical experimentation. RICHARD A. POSNER, ECONOMIC ANALYSIS OF LAW 25 (3d ed. 1986).  He also acknowledges "there is more to justice than economics," id. at 25-26, and makes other subtle modifications to his earlier views.  Compare POSNER, supra note 32, at 23 (2d ed.) ("[T]here is probably more to notions of justice than a concern with efficiency ....") with POSNER, supra, at 25 (3d ed.) ("[T]here is more to notions of justice than a concern with efficiency.").  For an analysis of Posner's evolving views, see, in this issue, David A. Logan, The Man in the Mirror, 90 MICH. L. REV. 1739 (1992) (reviewing RICHARD A. POSNER, CARDOZO:  A STUDY IN REPUTATION (1990)).

 
[FN34]. See Louis Kaplow, An Economic Analysis of Legal Transitions, 99 HARV. L. REV. 509 (1986).

 
[FN35]. Among the best theoretical treatments is that on expectations and justice in HENRY SIDGWICK, THE METHODS OF ETHICS 264-94 (7th ed. 1981), first published in 1874 as the changes associated with the early industrialization of England were being absorbed in politics and thought.

 
[FN36]. Kaplow, supra note 34, at 565.

 
[FN37]. Id. at 520.

 
[FN38]. Id. at 529.

 
[FN39]. L. T. Hobhouse, The Historical Evolution of Property, in Fact and in Idea, in PROPERTY:  ITS DUTIES AND RIGHTS 3, 9-10, 21-22 (2d ed. 1922); see Margaret J. Radin, Property and Personhood, 34 STAN. L. REV. 957 (1982).

 
[FN40]. Kaplow, supra note 34, at 534 n.70.

 
[FN41]. Kaplow does briefly address more traditional fairness concerns that might explain some of the qualitative distinctions central to existing transition practices and legal doctrine. Id. at 576-82.  The implications of this discussion for his larger analysis, however, are unclear.  On the one hand, he tentatively claims that concerns for fairness or other principles of justice are largely reflected in the economic analysis; motivations for concern with fairness turn out to be "much akin" to those an economic approach embodies, these two approaches have "congruent implications," and hence no "independent principle of justice," such as fairness, is necessary to rationalize and justify transition policy.  Id. at 577 & n.197.  On the other hand, he is not prepared to assert "an identity" between ethical concerns and the economic analysis, id. at n.197, and notes that, contrary to the economic analysis, the source of burdens (whether they stem from private forces or public policies) might require different compensation practices when analyzed through ethical norms like fairness.  But he then states that previous commentary does not indicate what the distinct ethical (non-economic) values might be that would justify such differences.  Part of the aim of this review is to suggest what these values might be and what is at stake in collectively embracing or rejecting them.  Of course, I can hope at best to be merely suggestive here, since this is not the place for extended analysis of the vexing problems of transition practices.
  To crystallize our differences, note that Kaplow locates much of the convergence he sees between fairness and economic analyses in risk aversion, itself reflective of the declining marginal utility of money.  He interprets fairness concerns to require compensation for large, concentrated losses, and notes this is consistent with risk-aversion in which large monetary losses cause disproportionately larger losses in individual utility.  Id. at 577.  But existing compensation practices do not, at least on a first-order rationalization, take economic conceptions of risk aversion into account. Effects on total wealth of policy changes are not explicitly mirrored in judicial doctrine, nor do the doctrinal categories employed seem proxies for these effects. Policy changes that affect government entitlements (the "new property") do not generally implicate constitutional compensation requirements, even through marginal effects on individual welfare are likely to be great; physical boundary crossings do require compensation, no matter how trivial in absolute terms or how significant the effects on total wealth are likely to be; when looking into whether government has taken any distinct stick in the "bundle" of property rights, judicial decisions do not turn on absolute diminution of economic value (despite misleading judicial formulations), but at whether some "distinctly perceived, sharply crystallized, investment-backed expectation" has been violated. Frank I. Michelman, Property, Utility, and Fairness:  Comments on the Ethical Foundations of "Just Compensation" Law, 80 Harv L. Rev. 1165, 1233 (1967).  These results do not reflect declining marginal utilities of money; hence, to explain existing practices, it must be the case either that (a) the fairness and economic approaches are not nearly as convergent as Kaplow suggests; (b) the economic approach must be reformulated; (c) existing practices do not embody ethical intuitions such as fairness.
  In my view, to uncover the ethical intuitions that underline compensation practices, we cannot focus exclusively on "large and concentrated losses" (through much commentary uses such language). This is too quantitative a formula.  Instead, qualitative distinctions between different kinds of instrusions government makes on individuals are central to understanding existing practices; it is these qualitative distinctions to which the relevant ethical intuitions attach.  By focusing, as many others have, on "large and concentrated losses," Kaplow's fairness discussion focuses on distributional equality as the relevant ethical ideal, but framed in this way, that ideal runs into the kinds of incoherences noted above regarding existing wealth and marginal income effects from policy reforms.  Instead, I think fairness in this area is more a matter of government's obligations of respect for individual dignity than of distributional equality. The individual right not to have government inflict certain kinds of harms, without at least paying compensation in acknowledgement of its obligation to respect the dignitary interests invaded, seems central to this area.  But further discussion must await another day.

 
[FN42]. This economic perspective on legal transitions goes back to some of the first, and best, economic writing in law.  see, e.g., Robert L. Hale, Value and Vested Rights, 27 COLUM. L. REV. 523, 528 (1927) ("There is no essential economic difference in losing the value of one's property by virtue of the suppression of the manufacture of carbon black from natural gas, and losing it by direct appropriation by the government.").

 
[FN43]. Kaplow, supra note 34, at 552-56, 553 ("All of the analysis offered in connection with losses applies, in a symmetric fashion, to gains.").  Kaplow does qualify this argument in the footnotes by stating "I am not claiming that there are no arguments that may warrant a distinction between gains and losses," Id. at 553 n. 127, but he remains committed to the view that it is at least "presumptively inconsistent to argue for special treatment of one but not the other."  Id.

 
[FN44]. See Mark Kelman, Interpretive Construction in the Substantive Criminal Law, 33 STAN. L. REV. 591 (1981).

 
[FN45]. Id. at 616, 642.

 
[FN46]. Id. at 592.  In contrast to these more rhetorically bold proclamations, Kelman at other points suggests a more modest agenda, such as merely aiming to bring to the surface questions of characterization that go "virtually unexamined."  Id.

 
[FN47]. Kelman makes this view clearest in MARK KELMAN, A GUIDE TO CRITICAL LEGAL STUDIES 86-114 (1987).

 
[FN48]. See supra note 21 and accompanying text.

 
[FN49]. See Kelman, supra note 44, at 616.

 
[FN50]. Id. at 617 (quoting United States v. Calley, 46 C.M.R. 1131, 1183, affd., 22 C.M.A. 534, 48 C.M.R. 19 (1973)).

 
[FN51]. Id. at 618.

 
[FN52]. For recognition of the way assignments of criminal responsibility are often external attributions based on various social conceptions of role, see Meir Dan-Cohen, Responsibility and the Boundaries of the Self, 105 HARV. L. REV. 959, 999-1001 (1992).

 
[FN53]. With regard to moral responsibility, this point is beautifully developed as the thesis of MARION SMILEY, MORAL RESPONSIBILITY AND THE BOUNDARIES OF COMMUNITY (1992).  With regard to criminal responsibility, see, e.g., Alan Norrie, Subjectivism, Objectivism and the Limits of Criminal Recklessness, 12 OX. J. LEGAL STUD. 45, 46 (1992) ("Recklessness, it transpires, is ultimately in its very essence a matter of socio-political construction and judgement, not an abstract, apolitical, juridical concept of individual responsibility ....").
  Kelman might perhaps believe, with some justification based on the way legal scholarship and philosophy have traditionally addressed problems of responsibility, that this kind of Kantian foundation is the basis for the way criminal law concepts are widely understood; thus one could read Kelman as seeking to show that such an understanding is misconceived.  But that would not mean that criminal law, as a social practice, is misconceived or incoherent, only that certain academic justifications for it are. Moreover, from a pragmatic perspective, even if criminal law did rest on some single underlying philosophical framework, showing that framework to be infused with contradiction would provide no critical leverage over existing practices in the absence of offering some alternative, persuasive framework not subject to similar limitations.

 
[FN54]. See Mark Kelman, Consumption Theory, Production Theory, and Ideology in the Coase Theorem, 52 S. CAL. L. REV. 669, 669 n.3 (1979).

 
[FN55]. Id.; see Richard A. Epstein, A Theory of Strict Liability, 2 J. LEGAL STUD. 151, 164-65 (1973).

 
[FN56]. See Kelman, supra note 54.

 
[FN57]. Kelman does officially disclaim making any evaluative judgment as to the empirical behavior he describes, but it is hard to avoid concluding that he believes, appropriately enough, that much of this behavior is completely intelligible and rational. Hence his references to the alternative behavior implied by the Coasian model as obsessive marginalization and a "path to the psychiatrist's couch."  Id. at 689.

 
[FN58]. This is the way Don Regan appears to find the question most intersting in his critique of Joseph Raz's arguments that values are incommensurable.  See Donald H. Regan, Authority and Value: Reflections on Raz's Morality of Freedom, 62 S. CAL. L. REV. 995, 1056-75 (1989).

 
[FN59]. For one of the most recent demonstrations that understanding liberalism as requiring collective decisions to avoid any stance on alternative, substantive conceptions of the good makes liberalism unable to resolve concrete ethical problems, see EZEKIEL J. EMMANUEL, THE ENDS OF HUMAN LIFE (1991).  For the argument that liberalism, at least in some variants, need not be understood as committed to this kind of neutrality, see Stephen A. Gardbaum, Why the Liberal State Can Promote Moral Ideals After All, 104 HARV. L. REV. 1350 (1991).

 
[FN60]. "The notion of the private will have been by hook or by crook completely eliminated from life."  PLATO, THE LAWS (739 CD), cited at p. 120.

 
[FN61]. See generally JURGEN HABERMAS, KNOWLEDGE AND HUMAN INTERESTS (1968).

 
[FN62]. See supra note 6.

 
[FN63]. As Nussbaum points out, this helps explain why Plato, and perhaps other theorists who assume value commensurability, seize on such otherwise unfathomable qualities as pleasure or wealth maximization to define their single metric of value.  P. 109.  We should see these approaches as starting, not from commitment to viewing the particular value chosen as the only ultimate value, but from the prior commitment to finding a framework for choice that will make it easier to generate single right answers.  Choosing the particular metric may be secondary to the desire to discover such a framework, which then necessitates assuming commensurability and coming up with some single metric.

 
[FN64]. See supra text accompanying notes 3-6.

 
[FN65]. See generally Richard H. Pildes, The Unintended Cultural Consequences of Public Policy:  A Comment on the Symposium, 89 MICH. L. REV. 936, 937-39 (1991).

 
[FN66]. ERIC FONER, RECONSTRUCTION 74 (1988).

 
[FN67]. Id.

 
[FN68]. See, e.g., Penn Cent. Transp. Co. v. New York City, 438 U.S. 104, 124 (1978).

 
[FN69]. A similar idea appears to underlie Jules Coleman's analysis of legal rights and the purposes that legal remedies enforcing them should be understood to serve.  JULES L. COLEMAN, MARKETS, MORALS AND THE LAW 35, 39 (1988) ( "[P]roperty, liability and inalienability rules are best understood as devices for generating or specifying the content or meaning of such rights.... It is unhelpful to think of them as tools or instruments for protecting entitlements.").

 
[FN70]. Many economists do recognize that efficiency analysis cannot determine how entitlements ought initially to be defined and distributed, and Kaplow himself at times reverts to more qualitative, traditional distinctions in defending the compensation practices his approach would suggest.  Thus, Kaplow distinguishes between government decisions that a certain activity is "undesirable," in which case retroactive application with no compensation is appropriate, and government action based on a "change in circumstance," in which only prospective application, or retroactivity with compensation, might be appropriate.  Kaplow, supra note 34, at 551-52.  These are the kinds of distinctions traditional doctrine has long made, but which economic analysis presumably means to displace with more "rigorous" analysis.  But Kaplow does not tell us what criteria distinguish these situations and whether economic analyses might be of any help in providing the appropriate criteria.  Kaplow also avoids the difficulties of this kind of classification scheme by discussing "undesirable" activities as if they "had always been harmful." Id. at 551. But that offers no guidance on how we ought to treat difficult, common cases, such as ones in which a previously legitimate private activity, carried on at the same level, comes to impose (or to be viewed as imposing) "undesirable harms" on others.  See, e.g., Hadacheck v. Sebastian, 239 U.S. 394 (1915) (brickyard established far from city, but when city expands, brickyard ordered to desist, resulting in market loss of land value from $800,000 to $60,000).

 
[FN71]. See Frank I. Michelman, Reflections on Professional Education, Legal Scholarship, and the Law-and-Economics Movement, 33 J. LEGAL EDUC. 197, 208 (1983) ("[L]aw and economics cannot entertain the idea that thought, motive, and action make one another what they are.").

 
[FN72]. This tendency to attempt to understand the meaning of one particular legal norm in isolation from other norms and the historical purposes that ground the norm occurs not just in some aspects of law and economics, but in certain contemporary styles of analytic jurisprudence.  See, e.g., Peter Westen, The Empty Idea of Equality, 95 HARV. L. REV. 537 (1982).  For a critique from within analytic philosophy of law of this approach to understanding legal norms, see Jeremy Waldron, The Substance of Equality, 89 MICH. L. REV. 1350 (1991) (book review).  For an internal critique of law and economics expressing this point, see GUIDO CALABRESI, IDEALS, BELIEFS, ATTITUDES AND THE LAW 69 (1985) ("[W]hat is deemed unreasonable behavior, no less than who is the cheapest avoider of a cost, depends on the valuations put on acts, activities, and beliefs by the whole of our law and not on some objective or scientific notion.").

 
[FN73]. For an argument that recent developments in corporate law regarding management resistance to hostile takeovers reflect just this understanding, see Jeffrey N. Gordon, Corporations, Markets, and Courts, 91 COLUM. L. REV. 1931 (1991).  In at least one specific circumstance, courts have become more willing to defer to managerial business judgment concerning which among several competing merger partners to choose (or takeover bids to endorse), even when the rejected bidder offered substantially greater immediate monetary gains to the target's shareholders.  Gordon suggests that the best explanation for this development is that courts have come to believe that the single-minded pursuit of economic efficiency in this setting can undermine cultural values that are necessary preconditions to the success of the market system.  See id. at 1986 ("[T]he success of a market-oriented system ultimately depends on the flourishing of such values as loyalty and fairness.").

 
[FN74]. DAVID HUME, A TREATISE OF HUMAN NATURE 484-514 (P. H. Nidditch & L.A. Selby-Bigge eds., 2d ed. 1978).

 
[FN75]. Id. at 491.

 
[FN76]. Cf. PUTNAM, supra note 5, at 196 ("[M]oral philosophers would do better to reflect on the conditions that make it increasingly difficult for many people to feel any sense of social solidarity at all than to issue overly elaborate arguments from unconvincing premises.").

 
[FN77]. HUME, supra note 74, at 491.

 
[FN78]. A similar point is made in John Stick, Charting the Development of Critical Legal Studies, 88 COLUM. L. REV. 407, 413 (1988) (book review) ("If moral reasoning involves not just application of rules, but also judgment of the suitability of analogies or application of values in context, as even many liberal writers would claim, Kelman's argument is incomplete.").

 
[FN79]. I do not want to imply that Kelman seems unaware of these problems.  In more recent work, he suggests that contemporary legal discourse shifts between viewing values as purely subjective matters of taste and viewing them as relatively more shared or objective; for example, Kelman describes legal norms thatassign legal rights as reflecting commitments to "value objectivism." KELMAN, supra note 47, at 86.  What interests Kelman in this work is his perception that legal discourse oscillates between these positions and, in his view, suppresses recognition of the widespread commitment in some areas of law to treating values as shared or "objective."  What interests me in this review is that we seem to treat values as incommensurable with regard to many norms other than "legal rights," and that some CLS scholars frame their critique of legal categories so broadly as to reject the appropriateness of ever conceiving values as incommensurable.

 
[FN80]. Id. at 184.

 
[FN81]. BRUCE A. ACKERMAN, PRIVATE PROPERTY AND THE CONSTITUTION 4 (1977).

 
[FN82]. In the former, the aim of legal rules is understood to be vindicating the expectations and practices embedded in and generated by existing social institutions.  According to the latter, legal rules ought to be developed in ways that accord with some larger, comprehensive, normative understanding of the aim of public policies, such as utilitarianism or Kantianism - even if doing so requires overriding socially based expectations and framing legal analysis in largely inaccessible, technical terms.  As a result, in contemporary legal culture, "the Scientific Policymakers are unable to make sense of the law, while the Ordinary Observers have lost their voice and are capable only of manipulating precedents whose deeper structures are lost from view."  Id. at 168.

 
[FN83]. Margaret J. Radin, Market-Inalienability, 100 HARV. L. REV. 1849  (1987).

 
[FN84]. As Radin points out, not all arguments against permitting market exchange of certain goods necessarily reflect these reasons against commodification.  Many arguments against permitting people to sell themselves into slavery, for example, focus on "market failures" that might characterize such a market and thus justify collective proscription of such sales.  See id. at 1863-70.  But as Radin recognizes, these arguments still conceive the good as potentially commodifiable; they merely point to contingent features of markets that make them inappropriate on instrumental grounds as exchange structures for these particular goods.  The discussion in text addresses the very different kinds of reasons against market alienability of certain goods that writers like Radin develop.

 
[FN85]. In addition, it is not clear whether critics of commodification accept a single-value or incommensurable-values conception of rationality, since one can imagine arguments of either type against commodification.

 
[FN86]. For another example of previous scholarship that resonates with the issues described here, see George Fletcher's identification in tort law of an internal conflict between fairness and utility as the basis for determining the scope of duties people owe to each other.  George P. Fletcher, Fairness and Utility in Tort Theory, 85 HARV. L. REV. 537 (1972).  Much as I have suggested here, Fletcher noted in tort scholarship a "bias toward converting values which are ends in themselves into instrumentalist goals."  Id. at 538 n.4.  He also observed that, where issues of fairness were discussed, few noninstrumentalist accounts could be found.

 
[FN87]. P. 123 (citing PLATO, THE LAWS (819DE)).

END OF DOCUMENT