Do Philosophy and Sociology Mix? A Non-Essentialist Socio-Legal Positivist Analysis of the Concept of Law

Transcription

1 Oxford Journal of Legal Studies, Vol. 24, No. 4 (2004), pp Do Philosophy and Sociology Mix? A Non-Essentialist Socio-Legal Positivist Analysis of the Concept of Law KENNETH EINAR HIMMA In A General Jurisprudence of Law and Society, Brian Z. Tamanaha articulates and defends a socio-legal positivism that recognizes the myriad ways in which the concept of law is conventionally used. As he describes it, socio-legal positivism holds that a legal system exists whenever there are legal actors (conventionally identi- Wed as such) engaged in producing and reproducing a legal system through shared secondary rules, regardless of their eycacy in generating widespread conformity to the primary rules, regardless of their evectiveness, or lack thereof, in maintaining social order. 1 Ordinary linguistic usage, then, plays a central role in determining the content of the concept of law. Tamanaha concludes that [l]aw is whatever people identify and treat through their social practices as law (or droit, recht, etc.) (GJ 166; emphasis omitted). Socio-legal positivism is usefully characterized in terms of its relationship to Hartian positivism. Like Hartian positivism, socio-legal positivism accepts the Separability Thesis and the Social Fact Thesis, which, taken together, explain the content of law as manufactured through coordinated social practices of some sort. 2 Unlike Hartian positivism, socio-legal positivism denies the following claims: (1) law has institutional features that are essential to its being characterized as law ; (2) law has something that is plausibly characterized as the function of law; (3) it is a conceptually necessary condition for the existence of a legal system that oycials accept the norms dewning the legality criteria as standards governing their behaviour; and (4) it is a conceptually necessary condition for the existence of law that legal norms are eycacious in guiding citizen behaviour. A Review of Brian Z. Tamanaha, A General Jurisprudence of Law and Society (Oxford University Press, 2001). I am indebted to Brian Tamanaha and Julie Dickson for very helpful comments. 1 Tamanaha, A General Jurisprudence of Law and Society, 148. Hereinafter GJ. 2 The Separability Thesis asserts that there are no necessary moral constraints on the content of law, while the Social Fact Thesis asserts that the content of law is manufactured according to social processes. Tamanaha seems to accept both: the existence and content of law is determined by some range of facts about human beings in a social setting facts about their behaviour, history, institutions, beliefs, and attitudes (GJ ; quoting David Lyons, Moral Aspects of Legal Theory (Cambridge: Cambridge University Press, 1993)), 77. Oxford Journal of Legal Studies, Vol. 24, No. 4, Oxford University Press 2004; all rights reserved

2 718 Oxford Journal of Legal Studies VOL. 24 In addition, socio-legal positivism denies the claim, highly inxuential among legal sociologists, that the content of law mirrors the content of prevailing social norms. As is readily evident, socio-legal positivism can fairly be characterized, to use a term coined by Frederick Schauer, as an extremely thin version of Hartian positivism. 3 Tamanaha s project is an ambitious one that criticizes and defends a number of diverent kinds of claim. Some of these claims are, as one would expect from a legal sociologist, contingent, descriptive, and empirical in character; a claim about what law characteristically does is an empirical claim about what most legal systems do. Others are necessary and conceptual in character; claims about essential characteristics or functions purport to be true of every conceptually possible legal system. Indeed, the breadth of Tamanaha s project can be seen in the range of theorists whose views he takes on. Tamanaha s targets include not only conceptual theorists like Joseph Raz, Ronald Dworkin and H.L.A. Hart, but also sociologists like Emile Durkheim. Tamanaha s book overs important conceptual, philosophical, and sociological insights into law; as such, it is an important contribution to both the philosophical and sociological literature a very impressive feat. The writing is clear; and the arguments for both the sociological and philosophical claims are rigorous, detailed, and nuanced. For these reasons, I recommend it, without reservations, to both legal philosophers and sociologists. Nevertheless, I argue below that the book s sociological arguments are more successful than the conceptual arguments. 4 Although Tamanaha s sociological analysis succeeds in refuting the traditional sociological view of the relationship between law and society, his conceptual analysis is somewhat less successful as a challenge to Hartian positivism. In particular, I argue that his view that [l]aw is whatever people identify and treat through their social practices as law is far too thin to be theoretically adequate as an account of the concept of law. 1. The Mirror and Function Theses As the title announces, Tamanaha s aim is to construct what he calls a general jurisprudence of law and society. At least part of this complex idea will be quite familiar to theorists who spend their time attempting to explain the concepts that Wgure centrally in legal practice which, of course, includes the concept of law itself. As Austin and then Hart explained it, a general jurisprudence purports to identify and explain elements central to law in the following sense: any normative system that lacks one of these elements is simply not accurately 3 Frederick Schauer, Positivism through Thick and Thin in Brian Bix (ed), Analyzing Law (Oxford: Clarendon Press, 1998). 4 Even so, Tamanaha hits the bull s-eye on at least one critically important positivist doctrine: the Hartian claim that it is a necessary condition for the existence of a legal system that it be minimally eycacious in regulating the behaviour of citizens. See below at 730.

3 WINTER 2004 Do Philosophy and Sociology Mix? 719 described as law. Thus, it is not just that these elements are present in every conceivable legal system; it is rather that the very presence of such elements in a system, taken together, warrant characterizing it as a system of law. Nevertheless, it is important not to be misled by the reference to conceptual jurisprudence. Tamanaha s goal is not simply, or even principally, to produce an analysis of legal concepts as positivists, constructive interpretivists, and natural law theorists strive to do. Rather, Tamanaha s primary concern is to explain the relations between law and society at a suyciently high level of generality to be applicable to all possible societies and legal systems. The stated goal is to construct a general jurisprudence of law and society and merely explaining the concepts that are central to law and legal practice does not have any obvious connection to the relationship between law and society. Tamanaha s discussion is organized around an evaluation of two theses that, taken together, make up what he takes to be the traditional view of law and society. The Wrst is the Mirror Thesis. As Tamanaha states this thesis, law is a rexection a mirror of society (GJ 1). On this conception of the relationship between law and society, the content of the law rexects the content of various social norms and practices that govern life in the society. The second thesis is the Function Thesis. As Tamanaha describes this thesis, law maintains social order by establishing and enforcing the rules of social intercourse, and by resolving disputes (GJ 2). Thus stated, the idea is that law has some central or essential purpose that is concerned with maintaining social order by enforcing rules that mirror other socially accepted norms. Tamanaha argues that both claims are taken for granted by many legal theorists especially those in working in empirical areas of legal theory like legal sociology. Endorsing the Mirror Thesis, legal sociologist Lawrence Friedman asserts that necessarily, [legal systems] rexect what is happening in their own societies. 5 Endorsing one strong version of the Function Thesis, sociologist David Dudley Field argues Where there is no law there can be no order, since order is but another name for regularity, or conformity to rule. 6 Although we will see that the Mirror and Function Theses can be interpreted in a number of ways, the theorists quoted above interpret them as making very strong claims about the content and purpose of law. Tamanaha believes that each of these theses is problematic and attempts to show both should be rejected. Though most of the discussion in the book is directed at evaluating the Mirror and Function Theses, Tamanaha s principal concern is to articulate his own socio-legal positivist jurisprudence of law and society. He treats the traditional view that consists of these two theses as a foil 5 Lawrence Friedman, Borders: On the Emerging Sociology of Transnational Law, Stanford Journal of International Law, vol 32, no. 1 (1996) 72; quoted at GJ 2. 6 David Dudley Field, Magnitude and Importance of Legal Science in S.B. Presser and J.S. Zainaldin (eds), Law and Jurisprudence (St. Paul, MN: West Publishing, 1995) 713; quoted at GJ 2 3.

4 720 Oxford Journal of Legal Studies VOL. 24 for the development of his own positive views on law, society, and the relationship between them. 2. The Plan of the Book In Chapter 1, Tamanaha sets out the general framework for the discussion. The Wrst component requires an elaboration of the Mirror and Function Theses. The second outlines traditional views on the relationship between law and society in terms of the relationships between three basic elements: Custom/Consent; Morality/Reason; and Positive Law, which Tamanaha describes as the rules articulated and enforced by an institutionalized authority (GJ 4). As Tamanaha describes the traditional views: [A]s characterized by Western legal and social theory, the relations among these three elements are complex and far-ranging. They are thought to have a shared origin; they are thought to rexect one another, to feed or inxuence one another, and to determine the eycacy of one another. The relations between custom/consent and positive law, and between morality/reason and positive law, are parallel in many respects, especially with regard to the issue of legitimation. Coinciding with this general intimacy and with their mutually supporting relations, these three elements also exist in tension as competing sources of power and authority, overlapping one another in their reach (GJ 9). Chapter 2 shows the inxuence of the traditional view in Western legal theory. Tamanaha begins by showing that the principal elements of the view, including the Mirror and Function Theses as well as the analytic framework described above, are all present in Plato s and Aristotle s work. Further, he argues that the principal elements of this view are present in both neo-classical natural law and legal positivism. 7 Natural law theory is sometimes interpreted as claiming that it is conceptually impossible for there to be unjust law, suggesting that the content of the law, though universal and based in reason, is not necessarily responsive to the contingent customs of any particular society. But a more modest interpretation holds only that unjust laws are inconsistent with certain moral ideals of legitimacy; the more modest interpretation, then, recognizes that diverent legal systems contain diverent legal norms that mirror the accepted customs. Positivist theories, on Tamanaha s view, are more explicit in emphasizing the connection between the content of law and the content of the accepted social morality. As Hart put it in The Concept of Law, The law of every modern state shows at a thousand points the inxuence of both the accepted morality and wider moral ideas. 8 Tamanaha Wnishes the chapter by arguing that the traditional view is 7 Tamanaha does not address the issue of whether these elements are present in Dworkin s third theory of law. It is worth noting that Dworkin argues that the law of a community includes those principles of law that, as an objective moral matter, show the pre-existing legal materials and institutional history in the best moral light (i.e. come nearest to, as an objective moral matter, justifying the coercive enforcement of the law). See Ronald Dworkin, Law s Empire (Cambridge, MA: Harvard University Press, 1986). 8 H.L.A. Hart, The Concept of Law, Rev. edn (Oxford: Oxford University Press, 1994), 199. Hereinafter CL.

5 WINTER 2004 Do Philosophy and Sociology Mix? 721 presupposed, to a greater or lesser extent, by comparative law theories, legalsociological theories, and even Marxian theories. In Chapter 3, Tamanaha argues against two myths that explain the origin of legal systems in ways that suggest the content of the law will rexect the prevailing social norms and customs. According to the evolutionary myth, systems of law evolve naturally out of simpler forms of social organization grounded in the acceptance of customary norms; when these simpler forms of social organization can no longer evectively regulate behaviour, formal centralized structures arise to ensure the continuing obedience of citizens to the customary norms. According to the social contract myth, systems of law originate as the outcome of an agreement that takes place at some particular moment (or moments) in time. Tamanaha rejects these theories on historical grounds in favour of a theory that explains the origin of law as the outcome of conxicts in which one group seeks to impose its will on another. While Tamanaha concedes that some legal systems might have originated in the ways described by the two myths, he argues that [t]he majority of state legal systems in existence today originated through imposition from outside or were created by imitation by local authorities to meet the threat posed by conquest from outside powers (GJ 69). Chapter 4 describes a number of important changes in the way legal theorists and practitioners conceive the proper relationships between law and social norms. According to Tamanaha, scepticism about the objectivity of morality and the ability of reason to discern ultimate goods has contributed to the decreasing relevance of morality in the decisions of legal practitioners, such as judges and legislators. In addition, he argues that the growth of bodies of law that, at best, connect only tenuously with community standards and customs has diminished the role of these norms in shaping the content of the law; business law, administrative law, and law constraining legislative activity lack any substantial foundation in social customs, which are contrived to deal with radically diverent problems. Reliance on the content of custom has increasingly been supplanted by democratic procedures; subject consent is increasingly inferred from democratic participation instead of from the incorporation of widely accepted social norms into the law. Indeed, according to Tamanaha, the formal ideal of the rule of law which is consistent with considerable immorality in the content of law has replaced substantive moral constraints on lawmaking activities as law s chief legitimizing rationale. Chapter 5 consists largely of empirical arguments against the Mirror Thesis. First, Tamanaha points out that many existing legal systems have imported bodies of law from other legal systems. In particular, many colonized nations, such as India, have retained the legal systems and bodies of law that were imposed upon them by colonizing nations, despite the lack of social continuity between them. Second, non-state systems of law tend in many nations to be more evective in regulating behaviour than the oycially sanctioned legal system precisely because the content of state law does not suyciently rexect the content

6 722 Oxford Journal of Legal Studies VOL. 24 of prevailing social norms. Finally, the trend towards increasing social, political, and economic globalization has resulted in a trend towards increasing homogeneity in the content of the law of participating nations. In many instances, the desire to reap the benewts of a global economy has led many nations to enact laws rexecting economic and social realities of radically diverent cultures. Local customs and mores, on Tamanaha s view, do not play even a minimal role in shaping these bodies of law. In Chapter 6, Tamanaha develops his socio-legal positivist approach to law that he believes avoids problems that plague Hartian positivism. The discussion in this chapter is organized around a sustained critique of Hartian positivism that focuses on tensions among four central elements: (1) the meta-conventionalist 9 claim that ordinary linguistic usage determines the social phenomena to which the term law is correctly applied; (2) the conventionalist claim that the content of the legality criteria is determined by a conventional rule of recognition; 10 (3) the essentialist claim that certain features of law (such as the existence of a conventional rule of recognition) are essential to law; and (4) the functionalist claim that the function of law, as Hart put it, is to provid[e] guides to human conduct and standards of criticism of such conduct (CL 249). The most important elements of Tamanaha s critique are as follows. First, on his view, the functionalist claim clashes with the essentialist claim because institutional systems lacking the essential features of law can perform law s function and because institutional systems with the essential features of law can fail to perform law s function. Second, the conventionalist claim is in tension with the functionalist claim because [a] given function can be satiswed by any number of conceivable practices (GJ 149). Third, the meta-conventionalist claim is in tension with the essentialist claim because the claim that ordinary usage determines the content of the concept law seems inconsistent with there being any essential features of law (i.e., features that are necessary to law and legal systems). Tamanaha concludes that [law] has no essence (193). Tamanaha s view that law is whatever people call law is an attempt to incorporate the insights of Hart s conventionalist claims while rejecting his functionalist and essentialist claims. As Tamanaha points out, people may conventionally designate as law an institutional system of norms that fails to perform the function of guiding behaviour because people look to other systems of norms for supreme guidance; people in Yap, Micronesia, have very little idea of what the law requires or prohibits. Likewise, people may conventionally identify systems as systems of law despite the fact that they lack certain features thought to be 9 The terminology here is mine. 10 The terms legality criteria and rule of recognition refer to diverent abstract propositional entities though they are obviously related in important ways. For a discussion of the distinction between the two, see Kenneth Einar Himma, Final Authority to Bind with Moral Mistakes: On the Explanatory Potential of Inclusive Legal Positivism forthcoming in Law and Philosophy; and Kenneth Einar Himma, Making Sense of Constitutional Disagreement: Legal Positivism, the Bill of Rights, and the Conventional Rule of Recognition in the United States, Journal of Law in Society, vol. 4, no. 2 (Winter 2003),

7 WINTER 2004 Do Philosophy and Sociology Mix? 723 essential to law; people might, for example, identify a system as law despite the fact that it lacks a centralized mechanism for coercive enforcement or despite the fact that it lacks a monopoly on such enforcement. Socio-legal positivism is consistent with characterizing such practices, norms, and systems as legal in character precisely because it rejects Hart s functionalist and essentialist claims. 11 Chapter 7 attempts to reconcile socio-legal positivism with the legal pluralist view, nearly unquestioned among legal sociologists, that there can be, and frequently are, many legal systems operating within one society. Such systems include the legal system that gives rise to the state, of course, but they can also include systems of indigenous customary norms accepted as, and designated by, theoretically signiwcant portions of the population as law. Tamanaha believes that essentialist concepts of law are diycult to reconcile with legal pluralism because they limit the application of law to only systems that satisfy a laundry list of essential features, which may exclude systems that are recognized, accepted, and treated within a society as legal systems. In contrast, Tamanaha s non-essentialist, conventionalist analysis of law is clearly compatible with legal pluralism because it recognizes as law whatever people attach the label law to. To the extent that law is applied to a plurality of systems within a society, Tamanaha s analysis entails that there is a plurality of legal systems within that society. 12 The last chapter sets out the elements of a general jurisprudence that is broad enough to accommodate the diversity of both legal institutions and social living. To this end, Tamanaha rejects the concept of society in favor of a more general concept of social arena; the boundaries of the concept can be drawn in any way desired [and] as determined by the purposes of the study (GJ 207). Additionally, Tamanaha describes a number of diverent sources of social order: (1) an unarticulated and unrexective substrate of shared habits, modes of action, and practices; (2) shared norms; (3) self-interested instrumental behaviour that leads to mutually benewcial voluntary exchanges; (4) agreements; (5) feelings of social identiwcation and attachment (including love, altruism, and group identiwcation); and (6) coercion. Finally, Tamanaha lists a typology of various kinds of law, which include state law, customary law, religious law, international law, and transnational law. Tamanaha concludes the book by articulating what he takes to be the core questions for sociological investigation about law. As Tamanaha puts it, [f]or any given social arena, the core questions posed will be: (1) to what extent is 11 It is worth reiterating that socio-legal positivism shares with legal positivism its commitment to the ideas (1) that there are no necessary moral constraints on the content of law (the Separability Thesis) and (2) that the content of law is ultimately manufactured according to social processes (the Social Fact Thesis). 12 For an interesting attempt to reconcile legal pluralism with Hartian positivism, see Detlef von Daniels, Is Positivism a State-Centered Theory? in Kenneth Einar Himma (ed), Law, Morality, and Legal Positivism (Proceedings of the 21st World Congress of the International Association for Philosophy of Law and Social Philosophy), Franz Steiner Verlag, forthcoming 2004.

8 724 Oxford Journal of Legal Studies VOL. 24 (state, customary, international, religious, natural, indigenous, etc.) law a mirror of prevailing customs and morals? and (2) to what extent does... law contribute to the maintenance of the social order (GJ 231)? The underlying idea is that the extent to which law mirrors society to maintain social order will diver from one society to the next. In some social arenas, the norms of law will closely rexect other norms; in others, there will be little resemblance between legal and other social norms. 3. The Critique of the Mirror Thesis To get a sense for whether Tamanaha s critique of the Mirror Thesis succeeds, we need a clear sense of what this claim amounts to. The claim is stated in a number of diverent ways throughout the book: (1) law is a rexection a mirror of society (GJ 1); (2) a society s law will rexect its patterns of life and morality (GJ 2) 13 ; (3) the law of every modern state shows at a thousand points the inxuence of both the accepted social morality and wider moral ideas (GJ 2; quoting Hart, CL 199); and (4) law rexects/mirrors society (GJ 51). As it turns out, there are a number of very diverent interpretations of this claim. For his part, Tamanaha observes that the Mirror Thesis is ambiguous between a number of interpretations: William Ewald observed that there are strong versions of the mirror thesis Law is nothing but X, Law is wholly explicable in terms of X and there are weaker versions Law and X are closely related. And he noted that there are diverent dimensions of mirroring, whereby X can be Wlled in with economics, culture, politics, climate, and geography, or some combination thereof (GJ 51). As it turns out, there are considerably more interpretations than are suggested in this passage. In addition to the determining which set of norms are being mirrored, there are at least three more important areas of ambiguity in the statements of the Mirror Thesis quoted in the Wrst paragraph of this section, leading to a very large number of interpretations. First, it is not clear exactly what kind of modality is being expressed. On one construction, the Mirror Thesis expresses a necessary truth of some kind and there are three possibilities here: logical, conceptual, or causal. 14 On this construction, it is a logically/conceptually/ nomologically necessary truth that the content of the law in any given society rexects the content of its social norms and customs. On another construction, the Mirror Thesis expresses a contingent truth that is grounded in empirical 13 Tamanaha is quoting Kent Greenawalt, Law and Objectivity (Oxford: Oxford University Press, 1992) A proposition is a logically necessary truth if it is not possible for it to be false no matter what the world might otherwise have been like; e.g. the proposition expressed by either Sam is in the room or Sam is not in the room is logically necessary. A proposition is a conceptually necessary truth if it is logically implied by propositions that express the content of some concept; e.g. the proposition expressed by laws are norms is conceptually necessary. A proposition is a nomologically necessary truth if it is logically implied by the causal laws that are true of this world; e.g. the proposition that expresses some general truth about gravitational force is nomologically necessary.

9 WINTER 2004 Do Philosophy and Sociology Mix? 725 observations: in most/all observed legal systems, the content of the law rexects the content of the prevailing social norms and customs. Since, strictly speaking, the empirical construction applies only to existing legal systems, a third construction can be distinguished that states a probabilistic thesis about future legal systems: it is likely that the content of the law in future legal systems will also rexect the content of the prevailing social norms and customs. Second, it is not clear exactly how much of the law must rexect or mirror prevailing social norms to satisfy the Mirror Thesis. On one construction, the content of every single law rexects/mirrors the content of prevailing social norms. Though this interpretation is obviously quite strong, the use of the term mirror suggests such a construction. My image in a looking glass mirrors my appearance in the following sense: the appearance of every feature of my rexection in the looking glass closely resembles the appearance of some feature of me, and the appearance of every feature of me closely resembles the appearance of some feature of my rexection in the looking glass. On another construction, the content of most laws rexects/mirrors the content of the prevailing social norms. On the weakest construction, the content of a theoretically signiwcant number (large enough to warrant the supposition, at the very least, that it is not an accident or chance occurrence) of laws rexects/mirrors the content of prevailing social norms. Third, it is not clear what is meant by the terms rexect and mirror. On one construction, a law rexects/mirrors prevailing social norms if and only if the content of the law reproduces the content of some prevailing social norm or norms. On another construction, a law rexects/mirrors prevailing social norms if and only if the content of the law is logically consistent with the content of the set of prevailing social norms. A law that reproduces the content of some social norm is consistent with prevailing social norms (assuming the social norms themselves are consistent), but a law need not reproduce the content of some norm to be consistent with the set of social norms (assuming the set of social norms is not complete in the sense that there is a norm that governs every pair of alternative behaviours). Much, of course, turns on how we interpret the Mirror Thesis. On the one hand, the very strongest interpretation is wildly implausible. On this interpretation, it is a logically necessary truth that every law in every legal system reproduces the content of some prevailing social norm or norms; that is, the claim that some law does not reproduce the content of some prevailing social norm or norms is selfcontradictory. Such an interpretation, of course, cannot be reconciled with ordinary intuitions and practices regarding law and use of the associated concepts. Whether the US federal statute prohibiting unauthorized computer intrusions is a law has absolutely nothing to do with whether its content reproduces the content of a social norm in force among some segment of the population; it counts as law regardless of whether there is a corresponding social norm being practiced in the US.

10 726 Oxford Journal of Legal Studies VOL. 24 It is true that Hart seems to endorse a modal version of the Mirror Thesis, but this version is much weaker than the strongest version and is not really part of his core views about law. H.L.A. Hart suggested more as a working hypothesis than as some sort of Wrm conviction about an immutable truth that there are certain rules that had, as a matter of natural necessity, to be part of the law in order to do what law normally does. He called such norms (which include, e.g. norms prohibiting violence and theft) the minimum content of the natural law. But this version is much weaker than the strongest version described above in two respects. First, the character of the modality, as Tamanaha is careful to observe, is causal rather than conceptual or logical; Hart hypothesized that the law could not be causally eycacious in regulating human behaviour if it did not contain certain rules protecting human survival. Second, there was nothing in Hart s view that implied even that the content of many, much less most, laws had to incorporate the content of some pre-existing social norms; as far as Hart s view was concerned, the minimum content of the natural law could represent a very small portion say 1% of the content of the law in some municipal legal system. On the other end of the interpretive spectrum, the weakest claim is simply too modest to be of theoretical interest. The claim that many laws in many legal systems are consistent with some set of prevailing social norms is too weak at least on ordinary understandings of what many amounts to to be worth theoretical investigation. The claim that, as an empirical matter, there are many such societies seems too obvious to be worth an expenditure of theoretical resources. Such thoroughgoing ambiguities make it diycult for the reader to determine Tamanaha s targets and fully evaluate his arguments; but it seems to me that his arguments are strong enough to refute a more modest version that asserts that the content of most laws in most existing legal systems reproduce (or directly incorporate) the content of some social norms. The bulk of his case is empirical. As noted in the last section, Tamanaha argues that the standards of business law, administrative law, and law constraining legislative activity do not have counterparts in social norms because social norms generally govern the interaction of individuals in their personal capacities. Further, he argues that many nations have simply carried over bodies of law that were imposed upon them by colonizing nations and lacked any sort of substantive foundation in prevailing social norms. In addition, the trend toward social and economic globalization assures the increasing homogeneity of law around the world. It is hard to Wnd fault with Tamanaha s critique of even this more modest version of the Mirror Thesis. If anything, the case against this interpretation of the thesis is so strong that it is tempting to wonder why so many sociologists (and this interpretation states a sociological thesis) would have accepted this to begin with much less the stronger modal versions. After reading Tamanaha s analysis, it is hard to fathom how legal sociologists would have missed, for example, the fact that very important bodies of law governing governmental

11 WINTER 2004 Do Philosophy and Sociology Mix? 727 functions do not have mirror images in social norms. It is likewise hard to fathom how they would have overlooked the signiwcance of the fact that social norms have failed to keep pace with the rapid changes in information and medical technologies or the signiwcance of the increasing inxuence around the world of business models developed in the US and Western Europe. Tamanaha s critique of this version of the Mirror Thesis seems devastating. Nevertheless, a somewhat more plausible interpretation of the thesis holds that the content of most laws in most systems is consistent with the content of moral norms that are generally accepted and practiced by the bulk of the population. While it might be true, as Hart believed, that citizens need not take the internal point of view towards the primary or secondary norms for a stable legal system to exist and function, it is reasonable to think that a stable legal system cannot exist in a society where citizens vehemently reject the content of signiwcant areas of law as would likely be the case where the law conxicts with important moral norms accepted as legitimate standards by citizens. Citizens would likely have a very strong reaction to a body of administrative rules that allowed oycials to behave dishonestly with citizens that would likely impair the smooth functioning of the legal system. Still, Tamanaha s larger point remains correct: as he would surely point out, this kind of claim is simply not something that can be taken for granted as some sort of bedrock sociological principle. Rather, it is an empirical claim that rests on assumptions about how people are likely to react to certain kinds of behaviour by oycials and hence must be conwrmed according to exactly the same methodological principles that govern any other sociological study. Indeed, on Tamanaha s view, sociological study of any particular legal system should begin with the core question of whether the content of the relevant kind of law mirrors (in the appropriate sense) prevailing customs and morality. 4. The Critique of the Function Thesis While the Function Thesis is less ambiguous than the Mirror Thesis, there are several ways to interpret it. 15 To begin, there are at least a couple of diverent empirical interpretations. According to one, the Function Thesis asserts that, in most societies with a legal system, the law plays a primary role in maintaining social order among citizens. According to another, the Function Thesis asserts that, in most societies with a legal system, the law is intended (by some appropriate class of persons) to maintain social order. Further, the Function Thesis can be construed as asserting a conceptual claim about the nature of law. On this construction, the conceptual function of law the point of law as such is to maintain social order by providing rules that guide the behaviour of citizens. 15 As will be recalled, the Function Thesis asserts that law maintains social order by establishing and enforcing the rules of social intercourse, and by resolving disputes (2).

12 728 Oxford Journal of Legal Studies VOL. 24 Though Tamanaha is concerned to address a number of these claims, his principal target is the conceptual interpretation of the Function Thesis. The claim that law has a conceptual function seems to be a core thesis of most analyses of the concept of law. H.L.A. Hart, for example, somewhat reluctantly took the position that the point of law is to guide behaviour; as he put it, I think it quite vain to seek any more speciwc purpose which law as such serves beyond providing guides to human conduct and standards of criticism of such conduct (CL ). 16 Ronald Dworkin argues that the conceptual function of law is to justify the state in using coercive measures to maintain social order. 17 Lon L. Fuller argues that the point of a legal system to achiev[e]... [social] order... through subjecting people s conduct to the guidance of general rules by which they may themselves orient their behaviour. 18 Though these theorists disagree on the precise formulation of the Function Thesis, all seem to take the position that the conceptual function of law is to maintain social order by guiding citizen behaviour. Tamanaha rejects the claim that law has a conceptual function. For starters, Tamanaha points out that there is no function performed by only law or legal systems; for any proposed function f, it is not a suycient condition for a system of norms S to be a system of law that S performs f. Indeed, all systems of norms purport at least to maintain social order by guiding behaviour which is the function that is most commonly attributed by positivists to the law. Morality, for example, purports to do exactly that; and the scope of its rules are at least as wide as the scope of the legal norms of any developed legal systems with which we are familiar. There is, of course, nothing in the claim that the conceptual function of law is to maintain social order by guiding behaviour, as traditionally understood by those who endorse it, that implies that it is a suycient condition for S to be a legal system that S maintains social order by guiding behaviour. Indeed, being able to guide behaviour was just one of a number of properties that a normative system must satisfy in order to count as a legal system on Hart s view. In addition, the oycials must accept and follow a rule of recognition that provides standards for creating, changing, and adjudicating laws, while citizens must generally obey the rules valid under the legality criteria dewned by the rule of recognition. Tamanaha realizes this, but argues that the additional features that Hart posits are insuycient to distinguish state legal systems from sets and systems of norms that are not, on Hart s view, legal systems: 16 Although the view that law has a conceptual function is commonly attributed to Hart, he nonetheless seems to reject the view in the same passage that contains the remark that is typically thought to justify the attribution: Like other forms of positivism my theory makes no claim to identify the point or purpose of law and legal practices as such; so there is nothing in my theory to support Dworkin s view, which I certainly do not share that the purpose of law is to justify the use of coercion. In fact, I think it quite vain to seek any more speciwc purpose which law as such serves beyond providing guides to human conduct and standards of criticism of such conduct (CL ; emphasis added). 17 See Dworkin, Law s Empire, n Lon L. Fuller A Reply to Professors Cohen and Dworkin, 10 Villanova Law Review 655 (1965), 657.

13 WINTER 2004 Do Philosophy and Sociology Mix? 729 Hart s resort to secondary rules was not as evective in distinguishing legal from nonlegal systems of social control as he apparently believed. Hart s concept of law has been adopted by many legal pluralists, who assert that society consists of a multitude of legal institutions with rule-making and rule-enforcing power, from corporations to universities, to community associations, to sports leagues... All these institutions exert social control through administering a system of secondary and primary rules (GJ 138). Tamanaha concludes that Hart s analysis fails because social life is thick with phenomena that he would not have wanted to call law (GJ 138). 19 Tamanaha makes a convincing case that the existence conditions Hart posited fail to distinguish systems of law that give rise to a state from other institutional normative systems, but this does not, strictly speaking, refute the claim that law has a conceptual function. Assuming that legal pluralism is false, Tamanaha s observation, at best, refutes the claim that the existence conditions that Hart identiwes exhaust the content of the relevant concept of law; and this criticism can be answered by specifying additional existence conditions that only the state systems satisfy. 20 But assuming that legal pluralism is true as Tamanaha believes, his observations do not even constitute a criticism. If pluralism is true, then it is a virtue, rather than a vice, that Hart s theory does not pick out only state systems of law. Tamanaha, however, has a second objection to the conceptual interpretation of the Function Thesis. He argues that there is no function that is performed by every law or legal system; thus, for any proposed function f, it is not a necessary condition for a system of norms S to be a system of law that S performs f. In particular, there are institutional normative systems that are clearly legal systems despite the fact that the law in those systems does not succeed in guiding behaviour: Yap [Micronesia] had a legal system, with a legislature, a handful of judges and attorneys, a small police department, and a complete legal code based in its entirety on laws transplanted from the United States. But vast portions of the Code had never been applied, few lay people had any knowledge of the content of the laws or of the operation of the legal system, a large proportion of social problems were dealt with through traditional means without participation of the state legal system, and indeed on most islands there was no legal presence at all. The day-to-day behaviour of the people was not governed by state law, but by their own cultural norms... They did not identify with the legal system in any way. For most Yapese, when confronted with the law, it was like being confronted with the command of an alien sovereign, despite the fact that they were an independent country and this was their own legal system. While they did not routinely act in conxict with the law... it could not be said that they were obeying or complying with the primary rules in Hart s terms, since they were ignorant of these rules and paid them almost no heed (GJ ). 19 Tamanaha, however, criticizes liberal versions of legal pluralism that characterize entities like corporations and sports leagues as legal systems (GJ ). 20 Tamanaha acknowledges this and suggests that a conventional identiwcation of legal oycials might solve the problem (GJ 142).

14 730 Oxford Journal of Legal Studies VOL. 24 Tamanaha believes that this refutes the claim that the conceptual function of law is to guide behaviour because it shows that it is not a conceptually necessary condition for law to exist that it succeed in guiding citizen behaviour. Yap has a legal system, despite the fact that the citizens are largely unaware of its existence and operations and hence cannot be said to be guided by it. Tamanaha s claims about the institutional normative system in Yap and the role it plays in the lives of citizens seem correct. On the one hand, Yap is accurately characterized as having a legal system since all of the distinguishing furniture is there: courts, legislatures, attorneys, and coercive mechanisms for enforcing the law. On the other, the citizens of Yap are not accurately characterized as being guided by the law since they have little knowledge of its content. Although their behaviour happens to conform to the content of the law, this is because the relevant content of the law largely conforms to the content of social norms generally obeyed by citizens something that is unbeknownst to the citizens. Since the citizens are unaware of the law s content, it would badly misdescribe the situation to characterize the citizens as being guided by the law. Tamanaha s insightful analysis of Yap s system has at least two very important theoretical implications. First, it suggests a very plausible hypothesis that, if true, would refute certain empirical interpretations of the Function Thesis. The behaviour of citizens of Yap conforms to the law, not because they are motivated to obey the law, but rather because they are motivated to obey other norms that play a more central role in their lives. This, of course, is hardly unusual. Most people are morally decent and refrain from committing theft, murder, and physical assaults, honour their contracts, etc, because such behaviours are required by moral norms and not because such behaviours are required by law. If this is correct (as seems reasonable to hypothesize), it refutes the empirical claim that the law in most legal systems actually functions to guide citizen behaviour. Second, as Tamanaha points out (GJ 148), it also refutes Hart s eycacy condition, which is largely unquestioned among positivists. Hart argues that it is a conceptually necessary condition for a system S to constitute a legal system that S be generally eycacious in regulating citizen behaviour; as Hart puts the point, those rules of behaviour which are valid according to the system s ultimate criteria of validity must be generally obeyed (CL 116; emphasis added). But Yap has a legal system despite the fact that the citizens of Yap are not fairly characterized as obeying the laws. Again, while their behaviour conforms to the law, it is because they are obeying other social norms and not because they are obeying the law. The importance of this point should not be overlooked: the eycacy condition is almost universally regarded as a core tenet of any plausible analysis of the concept of law. That Tamanaha s analysis provides such a strong challenge to the eycacy condition is a striking achievement. Nevertheless, Tamanaha s argument falls short as a criticism of the conceptual interpretation of the Function Thesis. As it turns out, there is nothing in the

15 WINTER 2004 Do Philosophy and Sociology Mix? 731 view that the conceptual function of law is to maintain social order by guiding behaviour that implies it is a necessary condition for something to count as law that it maintains social order by guiding behaviour. To say that some kind of entity K has a conceptual function f is to say only that something must be capable of performing f to count as a K; it is not to say that something must succeed in performing f to count as a K. Objects that have conceptual functions are capable of malfunctioning; a television in need of repair is still a television despite the fact that it does not succeed in performing its function. But a CD player is not a television, in part, because it is not capable of doing what televisions do. It is nomologically impossible for something that is built the way a CD player is built to do the job that televisions are conceived to do. Such considerations apply to the functional concepts associated with law. Joseph Raz, for example, analyzes the concept of authority as having something like a conceptual function. 21 Roughly put, he argues that the function of an authoritative directive is to substitute the judgments of the authority about what people ought to do according to the reasons that apply to them for the judgments of its subjects; the point is to pre-empt the subjects own judgments about what those reasons require them to do. But, Raz argues, a directive that says Do what the balance of reasons requires you to do is not authoritative because it is incapable of pre-empting the subject s judgment about what to do; the subject must weigh the balance of reasons to determine what the directive requires. Since such a directive cannot as opposed to merely does not perform the conceptual function of authority, it is conceptually disqualiwed from being authoritative. One might think, I suppose, that the conceptual function of a social artefact like law should be unique to that artefact. On this view, conceptual functions operate to distinguish one social artefact from another such that it is a logical or metaphysical truth that no two social artefacts can share the same conceptual function. Thus, the function of maintaining social order by guiding behaviour can be the conceptual function of only one kind of system of norms; it simply cannot be that, say, the rules of law and the rules of social morality share the conceptual function of guiding behaviour. Conceptual functions are distinguishing functions. But even if this were true, it would not follow from the claim that social morality and law are both capable of guiding behaviour that law has no conceptual function; what follows is that it is not the case that the conceptual function of law is to guide behaviour. This latter claim is consistent, of course, with the claim that law has no conceptual function, but it is also consistent with the claim that law has some other presumably more narrowly speciwed conceptual function. For example, one might think that the conceptual function of law is to 21 Joseph Raz, Authority, Morality, and Law in Raz, Ethics in the Public Domain (Oxford: Oxford University Press, 1994).

16 732 Oxford Journal of Legal Studies VOL. 24 guide behaviour by providing norms that will be enforced by the state. While it might be true that states sometimes enforce norms of social morality, those norms are enforced because they are incorporated into the law and thereby made legally binding. Such norms, then, are enforced qua law and not qua social morality. Though the claim that law has a conceptual function is not accepted by everyone, there is a fairly compelling motivation for the idea. Consider, for example, the concept of scissors. Someone who understands all of the structural features and properties of scissors but does not understand that scissors are capable of cutting certain kinds of objects has not fully grasped the concept. It is true, of course, that scissors can be used for other purposes, such as for wounding people. But one can understand the concept of scissors without realizing that they can be used to wound people. Knowing that scissors are designed to do the job of cutting thin objects like papers is essential to understanding the concept. Legal systems are like scissors in this respect. One can understand the concept of a legal system without understanding that a legal system can be used to oppress people; knowing this requires knowing psychological facts about people that are not relevant with respect to the existence conditions for law. But someone who does not understand that legal systems are capable of guiding behaviour (though they might fail to do so) because they manufacture norms has failed to grasp something that is essential to understanding law. One who does not understand this has simply not fully grasped the concept of a legal system. If this is correct, then being capable of performing this function (or something very like it) is one of the existence conditions for law. 5. The Critique of Legal Essentialism Central to Tamanaha s view that law is whatever people attach law to is the rejection of a view that is at the foundation of each of the most inxuential theories of law. According to this view, there are some features that are essential to law in the following sense: an entity that lacks these features is conceptually disqualiwed from counting as an instance of law regardless of what otherwise might be true of the world. As is readily evident, positivism, classical natural law theory, and Dworkinian constructivism all describe properties of law that are fairly characterized as essential and thus presuppose a legal essentialist view. To his credit, Tamanaha grounds his argument in a methodological constraint on conceptual analysis that is not always appreciated by philosophers who analyze legal concepts. As will be recalled, Tamanaha distinguishes two threads of conventionalism: a conventionalism having to do with the content of concepts and a conventionalism with respect to the content of the rule of recognition. According to the Wrst version of conventionalism, the content of our concepts is largely determined by conventional linguistic practices regarding the use of the