Article body


In his remarkable article, “Human Interaction and the Law”, published in 1969, Lon Fuller argued that the primary form of law was customary.[1] Law was grounded in particular practices, emerged from those practices, and served to facilitate human interaction within them. Even statutes, he argued, were best conceived as punctual interventions by the legislator, comprehensible only against a background of customary norms.[2] This article shares that understanding of the importance of customary law. Law is grounded, fundamentally, in the practices of particular societies. All law, even legislation, finds its meaning in interpretive relationship to those practices. To understand law is to understand norms’ relationship to the web of human interaction in a given society.

Fuller explained this relationship in essentially pragmatist terms, treating law as though its primary function was to facilitate the interaction of individuals and groups. This pragmatist account has been developed in illuminating ways by Gerald Postema, who has gone substantially beyond Fuller in explaining the form of reasoning appropriate to a body of law continually articulated in relation to a set of social practices.[3] There is much value in these pragmatist conceptions and this article draws upon them, but they are ultimately unsatisfying. Especially troubling is the way in which their explanation of customary law departs significantly from how participants in legal orders conceive of their orders. That lack of resonance points to substantial limitations in the account, limitations that I seek to remedy in this article. Nevertheless, this article owes an important debt to the work of Fuller and Postema. It does not reject the pragmatist account; rather it supplements this account, building into it a significant dimension of substantive normative deliberation.

It will be clear by now that by “customary law” I do not mean exclusively the law of traditional or indigenous societies. The customary dimension with which I am concerned exists in all law, except the most despotic—and even then it is difficult to imagine a working system that lacks all trace of customary elements for any length of time. Thus, this article does not follow the lead of some scholars by presuming that there is a stark contrast between “custom” and “law”, with the latter conceived in positivist (or at least entirely state-centred) terms.[4] Rather, this paper asserts that customary law is the primary form of all law and then develops a conceptual framework for exploring how a legal order relates to the complex array of practices within a given society. It therefore applies as much to non-indigenous as to indigenous legal orders. Indeed, Fuller and Postema frame their projects primarily in relation to non-indigenous law—in Fuller’s case, American common law; in Postema’s case, both the English and American traditions of common law.[5] Legal practitioners and theorists within these traditions may not conceive of their orders in customary terms, but that fact points to the inadequacy of their theories. This article, building upon the work of Fuller and Postema, seeks to provide a more satisfying account.

Still, this article does grapple, in comparative fashion, with indigenous legal orders, especially those of northern North America. It does so in part as a way of identifying elements of customary law that have largely been lost from view in our understanding of non-indigenous legal orders. Encountering what is, for a non-indigenous scholar, the unfamiliarity of indigenous legal orders can be a way of throwing into relief dimensions of law that exist in non-indigenous orders but that are so taken for granted that we no longer see their significance, no longer see how they might have been different. Using comparisons across the indigenous/non-indigenous divide to explore the nature of customary law also helps ensure that the theory will be appropriate to law as it is defined and deployed in different societies. In turn, this comparison can help limit the extent to which our definitions smuggle in highly particular and contestable content, content that can distort the legal structure of indigenous/non-indigenous relations, and that indeed can be downright oppressive when applied across legal orders framed in very different terms.

Finally, in exploring the potential for normative engagement across the indigenous/non-indigenous divide, this article forms part of a larger project—in which my colleagues John Borrows, Val Napoleon, and Andrée Boisselle (among others) participate—of attempting to articulate the intellectual project behind the introduction of a new Bachelor of Indigenous Law at the University of Victoria. In this program, students would study specific indigenous legal traditions (including, for indigenous students, their own), in conjunction with the common law, obtaining degrees in both.

My focus in this article is on the normative content of customary law, not primarily on the processes by which customary law is defined and elaborated in particular societies. The processes are very important: all normative orders are marked by disagreement, all need to have ways of resolving disagreement, and much of the normative character of the community is shaped by the specific nature of those mechanisms and the power relations that characterize them.[6] But although I refer to such processes at several points, my primary concern is with the normative content of customary law and specifically with the relationship between this content and practices within society more broadly. Robert Ellickson has noted that most law-and-society scholars tend to treat observed norms as exogenous, declining to explain the commitments that drive those norms and, not incidentally, providing very little explanation for how participants reason with their norms.[7] This article responds to that challenge.

This article attempts to develop a general conception of how legal reasoning is related to human societies—how law, conceived in customary terms, exists in intimate relation with the array of practices in any human society. But I should be clear about this theoretical aspiration. My purpose is not to stipulate the form that legal orders should take, developing a common structure of normative obligation to govern all societies. Quite the contrary. Examining the nature of customary law across legal cultures demonstrates the way in which any legal order embodies particular, culturally shaped ways of conceiving of the very structure of normative obligation and normative relationship. Any order develops its own language of normative analysis, its own grammar, that materially shapes how the fundamental elements of legal relations are conceived. I shall explore the nature and effect of that grammar over the course of this article. The realization of that diversity reveals what is at stake in the encounter between indigenous and non-indigenous legal orders (or indeed between any substantially different legal cultures). It helps us see the possibilities and pitfalls of different ways of accommodating indigenous concerns, and it points toward more respectful modes of interaction.

I. The Pragmatist Account of Customary Law

In his 1969 article, Fuller described customary law as follows: “Customary law is not the product of official enactment, but owes its force to the fact that it has found direct expression in the conduct of men toward one another.”[8] Fuller sometimes treated all law that was embedded in social practices as customary law, although at other times he limited the concept to legal principles that had not been officially declared.[9] By this latter definition, judicial decisions would not be considered customary law. I will follow the first usage. All law has to have some mechanism of social determination—some way of determining which norm, or which interpretation of a norm, is to serve as society’s law when there are rival interpretations. This mechanism may involve determination by an “official” (such as a judge) or by some other social process (deliberation and consensus, the determination of a council of elders, assertion and acquiescence, and so on). It is an error to think that norms emerge in customary legal orders without any intervention of human agency.[10] The simple existence of mechanisms for specifying norms does not undermine those norms’ customary nature. The crucial fact is that customary law exists in intimate interdependence with social practices, not that it has been exempt from some form of official determination.

Fuller goes to considerable lengths to emphasize that customary law is not the result of mere habit, mere unreasoning repetition. He takes vigorous issue with Thomas Holland’s assertion that customary law develops like a path across a field: people happen to walk on the same line, gradually beating a path into the ground until a right of way forms. Rather, in Fuller’s view, customary law is always marked by the need to organize and facilitate interaction.[11] It has a reasoned dimension. Participants perceive the value of norms to their interaction and therefore follow those norms in their conduct. Fuller refers to customary law as a “language of interaction”, which generates complementary and stable expectations of conduct among participants in a social order.[12] Customary law enables participants to coordinate their actions through effective communication, predictably anticipating each other’s actions. The perception that an action is obligatory arises—customary law is created—when the participants “have come to guide their conduct toward one another by these expectancies.”[13]

The focus of the pragmatist account on coordination and facilitation has a number of very great strengths. To begin with, it emphasizes the eminently social nature of law. Law is not an abstract theory of justice. It is a method of social organization that is grounded in a particular society, governing relations within that society. It is, in Fuller’s words, a “program for living together”.[14]

Coordination is therefore the most common of common denominators in law. The members of a society may not share much, but if they are to live in any kind of order they at least need some method, some principles, even if rudimentary, for coordinating their actions. The participants may not have a strong subjective commitment to these understandings—they may, if given the choice, prefer other norms to the ones that are now in place—but if they are to have any order at all, some rules of conduct need to be acknowledged as the society’s rules, even if individuals disagree. Coordination therefore captures the most basic, indeed definitional, requirement of a legal order: that participants’ conduct be ordered by some social process, by some socially determined principles.

The establishment of this process, these norms, is a considerable accomplishment precisely because people do disagree on what they should be—at least if one looks beyond the norms’ abstract formulation and considers their practical implications in specific situations. In all societies, mechanisms need to exist for determining which norms, among the range of possibilities, should be treated as the society’s norms. Again, these outcomes do not have to be a matter of full substantive agreement; there simply has to be a method for determining what the society’s norms are. And note that the essential element here is the specification of the norms, not necessarily their coercive enforcement. The pragmatist theorists are quite right to emphasize that coercion is not essential to law. The specification of norms may be sufficient to prompt voluntary compliance, as when a trade association rules upon a contractual dispute among its members or an ombudsman makes a finding on citizens’ entitlements.[15]

The pragmatist account also focuses our attention usefully on the relationship that exists between customary law and the practical conduct of interaction within a given society. The two are closely intertwined, indeed mutually constitutive. The practices can only work efficiently, frequently can only exist at all, if the norms are observed—if, for example, agreements are generally taken to be binding when concluded in the customary manner. The norms, in turn, are developed and then elaborated by participants who reflect upon the practices, consider their demands, try to articulate appropriate rules of conduct, and think about what those rules should mean in particular cases. It makes sense for participants to take the practices as the starting point, because, after all, practices furnish both the need for coordination and the concrete examples of living together from which one can fashion norms. Even among people who share little else—who have different ideas of justice, different interests, perhaps even different metaphysical commitments—the practices provide a nexus of interaction, from which norms to govern those practices can be identified.[16]

Moreover, we see here how that process is always a reasoned one, involving at least an implicit deliberation to identify standards by which to align one’s actions with those of others. It is not the result of mere force of habit.[17] Indeed, often deliberation is both an explicit and dominating characteristic of customary legal orders, in indigenous and non-indigenous contexts alike. Postema says of the English common lawyers of the seventeenth and eighteenth centuries:

To the common lawyer’s mind, these three activities—articulating standards, showing them to be reasonable and sound, and applying them to particular cases—were not three separate processes, but rather interrelated moments of a single process of discursive reasoning. Through disciplined movements in this process, common law rules emerged and common law doctrine evolved. In their view, the life of the law was disciplined reason.[18]

Note that this also means that customary law is not locked in the past. On the contrary, it requires critical reflection upon conduct in a particular society and continual elaboration of the concepts used to understand and structure that conduct.

Participants may economize in that reasoning, accepting rules of thumb developed in past deliberations so that they do not continually have to deliberate from the ground up.[19] That economization may even be fundamental to the notion of law because (1) if people thought everything through for themselves each time they acted, they would lose the capacity for a common rule; and (2) the common rule always includes elements that go beyond any individual’s personal rationalizations for the rule: it at least incorporates the social warrant that the rule has obtained, and, in achieving that warrant, the rule has generally drawn upon a breadth of experience that goes well beyond that of any one individual. It represents, in some sense, the distilled lessons of all who have participated in its creation and elaboration (though the process of distillation is less than systematic).[20]

But economize as one does, the rule never completely escapes its reasoning, as one sees when rules are applied. To apply them well, one has to have some sense of their fit with broader practices, some judgment of the rationales that underlie the rule. Otherwise, one’s application is likely to be tone-deaf, unable to differentiate between distinctions that speak to the basis of the rule and those that are spurious. Indeed, the skillful application of rules is dependent on a sense of propriety, of the fit between rules and their social roles, not simply on a parsing of their linguistic content. It is dependent on a normative sensibility, grounded in familiarity with the practices of a particular society.[21]

The pragmatist emphasis on the relationship between norms and practice also avoids the dual errors of treating legal reasoning as though it were either (1) self-contained, autopoetic, generated from its own internal resources;[22] or (2) a purely intellectual process (even if that process is not confined to legal materials), where the only things that are considered are prior articulations within a given normative tradition. The raw material of customary law is not limited to books and statements; its own distinctive practices are not simply those of exegesis. This fact distinguishes the method of customary law from some approaches to hermeneutics, where normative deliberation is conceived as an entirely intellectual process, working purely and simply off of others’ past expressions.[23] As Postema explains, implicit rules

arise from conduct, not conception. Verbal formulations may more or less accurately capture the rules implicit in the conduct, but the formulations are always post hoc and strictly answerable to the conduct. No formulation is authoritative in virtue of its public articulation alone. ... Of course, agents, or even observers, may give verbal expression to such rules, but their ability to guide action is dependent on the substratum of ongoing practice. ... [T]his substratum will rise from time to time to the surface and, through the parties’ relatively explicit awareness of it, play a vital role in determining solutions to their problems.[24]

Postema’s emphasis on conduct over conception should, of course, still be read against Fuller’s insistence on the reasoned quality of customary law. Postema agrees that what he calls “implicit law” is brought into being by purposive effort, continuing: “Typically, implicit rules are side-products of intentional, rational interaction. They arise from and are sustained in this interaction.”[25]

One way of explaining the relationship between norms and practices is that statements of norms never exhaust the capacity for learning from reflection upon experience. They are always approximations, distillations, interpretations, that are perennially subject to further evaluation and refinement as a result of experience.[26] And experience is always refractory: although norms may serve to shape our actions, the success of those actions is dependent on factors that go beyond our knowledge and intentions. An unexpected response, an initiative that causes suffering when it was expected to improve lives (such as the many attempts to assimilate indigenous peoples to a thoroughly non-indigenous way of life), may cause us to re-evaluate our norms—in the case of assimilation, it may cause us to reformulate our very conception of equality.

The expressions of customary law therefore exist in an interpretive relationship with practice, mediated by other participants’ interpretations and actions. That reasoning is necessarily evaluative. Participants seek to weigh the impact of past norms, judge the appropriateness or acceptability of that impact—what has proven important in previous decisions, what has proven ill-conceived—and seek to revise the norms and their application accordingly.[27]

These evaluations are an important focus of deliberation in customary orders. The members’ ultimate opinions may differ but, as Postema notes, their participation both in the practices and in past interpretations allow them to anticipate, to some degree, the solutions that their fellows are likely to propose with respect to the application of norms and to the norms’ future development.[28] They can anticipate the trajectory, or range of possible trajectories, of their colleagues’ future interpretations. This process of reflection and anticipation also reveals how participants develop a level of expertise in their society’s law: the deeper their engagement with the practices, interpretive resources, and interpretative methods of their society, examined over a span of time, the better their capacity to judge the acceptability of future interpretations of those norms—and indeed the better the arguments they can muster for the interpretations that they personally favour.[29] The intersubjective dimension of legal deliberation suggests another reason why participants’ grasp of their law is perennially limited—indeed subject to what is almost an axiom of limitation: the ultimate development of the law never depends on the views of one person alone, but always requires a convergence among participants’ interpretations and conduct.

The pragmatists’ emphasis on coordination also furnishes criteria that participants use to evaluate a particular set of norms, at least ostensibly. One can judge a body of norms by the norms’ success in coordinating interactions. I say “ostensibly”, however, because coordination alone is less useful as a source for evaluative criteria than one might think. It is virtually tautological that law coordinates action in society precisely because law exists to allow members of a group to live together in reasonable order. One cannot conceive of a working legal order that does not “coordinate action”. The critical questions, of course, are in what manner, to what ends, and how well it coordinates, and these point toward criteria of evaluation that go well beyond the fact of coordination. As we will see, this insufficient specification of the criteria for evaluation constitutes the principal weakness in the pragmatist account.

The pragmatists’ emphasis on the socially grounded character of customary law has led some to champion the common law against legislation, as though legislatures did little more than meddle with the more embedded and responsive common law, which they took to be internalized in the minds and conduct of its subjects.[30] That is not a view I share. All law is shaped by human agency, from the judges’ choice of one interpretation among many in their application of the common law, to scholars’ interpretation of the substance of the civil law in academic treatises, to the acts of the legislature. Legislatures generally deploy that agency in a more transparent manner than do judges, drawing upon the views of a wider swath of society through processes that allow for more direct participation of the citizenry in rough equality. They can accomplish objectives that would be impossible to achieve through the courts, including the elimination or substantial restructuring of social practices that may well be marked by profound asymmetries of power. The capacity for rapid and democratically authorized governmental action, possible only through legislatures, is worth preserving (even if in exercising that power, legislatures should be alert to and deferential toward the areas of special competence of courts).

But it is important to realize that legislation too works, to some degree, in the manner of customary law. First, the legislatures’ own procedures, their place within the legal order as a whole, even the deference owed to their enactments, are in large measure defined and regulated by customary norms.[31] Second, their enactments generally take the form of interventions into a body of existing law, deflecting the evolution of that law rather than creating law out of whole cloth. They adjust customary law, shape it, as much as they displace it. Third, much legislation is framed on the basis of the legislators’ interpretations of social mores and is meant to have the same close connection to social practices that judicial decisions do. There is still something to be said for the now-archaic idea that Parliament is the highest court, articulating the custom of the country.[32]

Fourth, and most importantly, even if legislatures do make changes to the law without much concern for existing practices, and even if courts do respect the right of the legislature to make those changes (as they ought to do), laws once passed are nevertheless quickly “customized”—overtaken by the process of interpretation and application, elaborated and extended.[33] The statutes, if effective, come to structure relations on the ground such that a statute’s interpretation, and the practices performed under it, converge into (in a phrase Postema uses in a different context) “a substantial congruence (but not identity).”[34] It may well be that over time these further interpretations become the primary reference points for future decisions and the terms of the statute itself fade into the background. Indeed, significant portions of what we now think of as the English common law (contributory negligence, for example) were founded upon statute, but the terms of the statutes have long since disappeared in the judicial elaborations encrusted upon them.[35] Moreover, this growth of interpretation and elaboration is common to all legal orders. It is not restricted to English law and its descendants. Even within the French civilian tradition, for example, a layperson, faced with a legal issue, would have great difficulty reasoning to a dependable conclusion from the terms of the Civil Code alone. In that tradition, the process of customization may be accomplished more by scholars than by judges. But scholars too develop the meaning of the Code through continual engagement with an always expanding range of applications.

There are, then, very great strengths to the pragmatist account of customary law. But what about the weakness identified above: the comparative underdevelopment of the criteria by which participants judge the quality of their law?

The pragmatist description of the role of law focuses overwhelmingly on coordination, paying very little attention to other roles that the law might serve. It especially overlooks the use of the law to affirm judgments of substantive value and bring conduct into accord with those judgments. Yet these judgments lie at the very core of what most participants think their law is all about. If, for example, one asked a Cree hunter, “Why are human beings obliged to treat the remains of harvested animals with great respect?” he or she would be unlikely to respond anything like, “Because it facilitates human interaction.” [36] The same would be true if you asked a non-indigenous citizen about the prohibition on murder, or about virtually any other provision of the criminal law, or even about the obligation to compensate for damage caused by one’s negligence. In fact, for the vast majority of the principles of any legal order, “the coordination of human interaction” would at least underspecify the considerations that account for the content of the law. For a great many norms, the disjuncture would be so dramatic that the coordination of human interaction would seem like a complete non sequitur.

Fuller and Postema are, of course, alive to this objection. It is all the more striking, then, that they still find it difficult to move beyond coordination. Fuller tends to construct his account of customary law around the rules of the road (especially the rule that one should pass oncoming traffic on the right), for which the rationale underlying the rules is, unusually, coordination alone.[37] He discusses the question of murder, but he primarily emphasizes that even here coordination plays more of a role than one might think. He notes in particular that the law of murder was designed to provide a substitute for the blood feud. This response tends to confirm rather than answer the objection, however, because of its abrupt change of focus from the rationale for the prohibition to the social mechanism used to address it (Fuller does acknowledge that other considerations operate in the prohibition on murder, although he does not develop them).[38]

In his writings on customary law, Postema too tends to circle back to coordination. This approach is especially puzzling given his close attention to material in which much more than coordination is at play. In his most recent work, for example, he develops a wonderful analogy between reasoning in law and the interaction of a jazz ensemble. In discussing jazz, he expressly invokes criteria for evaluation beyond mere coordination, noting that jazz follows a “musical-aesthetic, rather than instrumental or discursive, logic.”[39] And at a later point, “The parts do not merely avoid musical collisions, they enhance the musical significance of each other and of the whole.”[40] But when he returns to law, he again speaks in terms of “solutions to cooperation problems” and to best solutions that are defined in terms of Pareto optimality or correlated equilibria.[41] There is great potential in all this discussion for the development of a dimension that goes beyond coordination. Indeed, Postema acknowledges that coordination alone does not exhaust individuals’ normative interactions.[42] But that dimension is largely undeveloped.

I suspect that some theorists (though not, I believe, Postema) are strongly attracted to the language of coordination in part because, by avoiding substantive normative judgments, that language apparently allows one to finesse the problem of disagreement. The language of coordination instead offers a standard of assessment that is ostensibly neutral and uncontroversial. The theorist simply presents his or her interpretation as the parties’ own law—law that has emerged from practice—and that is, for that very reason, supposedly exempt from contention. The emphasis on coordination serves the rhetorical purpose of minimizing the role of agency in determining the content of the law. As should already be clear, I do not accept this approach: except in rare cases of pure coordination (such as the rule that one should drive on the right), law always involves interpretive judgments that embody significant normative claims. It is thus inherently open to dispute.

However, there is another, more limited sense in which Fuller and Postema are justified in drawing attention to the role of coordination in all law, including such things as the prohibition on murder: the fact that all law fashions a societal response to prohibited conduct and in fashioning that response a range of competing possibilities have to be narrowed down to one. On this view, the language of coordination really comes down to the need to have a common rule; it says very little about the content of that rule.[43] This, I believe, is the role that coordination does play in legal orders: it underlies the need for a common rule and explains why members of the society may feel that they should follow that rule, even if they disagree with its particular content. But an infinite range of potential rules can satisfy that requirement. In the choice among various possibilities—in arguments over what the legal order should prohibit, or in attempts to interpret or extend an existing legal order—substantive normative considerations and not the mere fact of coordination take centre stage. The failure to include those considerations is the major failing in the pragmatist account of customary law. Can it be remedied?

II. Neglected Dimensions of Customary Law

I begin by exploring in more depth what the pragmatist account neglects. This exploration helps to identify the kinds of things that the pragmatist account leaves out, establishing their nature and significance to law. It also makes clear how the neglected dimensions interact with the concerns of coordination and facilitation. The two sets of concerns do not compete. They are integrally intertwined, both essential, both simultaneously operative. My purpose is not to set aside the pragmatists’ emphasis on coordination, then, but to see how it works in tandem with the more substantive elements of the law. Grappling with the dimensions neglected by the pragmatist account serves to locate those elements with accuracy, charting their interrelationship with the concerns of coordination.

Many of the examples I use are drawn from indigenous legal orders. Those examples further this paper’s project of engaging in comparative inquiry across a broader range of legal cultures than generally occurs in comparative legal scholarship. Identifying elements of connection and of difference across indigenous and non-indigenous legal orders equips us more adequately to understand how those orders might productively relate to one another. It certainly reveals the variety of law and the challenges that are involved in normative dialogue across substantial differences of legal culture, and, precisely because of those challenges, what one can learn from engagement.

I explore the elements neglected by the pragmatist account under three headings: Part II.A, persons and property (by far the largest section of the three); Part II.B, requirements of legal community; and Part II.C, the sacred or mythic dimensions of law.

A. Persons and Property

I start with a theoretical framework that, like the pragmatist account, focuses overwhelmingly on the coordination of diverse individuals’ actions, is reticent in judging questions of substantive value, and tends to evaluate law by using criteria of practical efficacy rather than substantive justice: law and economics. This examination will be useful for several reasons.

First, law and economics aims to provide rigorous means for coordinating expectations and assessing efficacy in legal regulation. It articulates more precise standards for “the coordination of human interaction” than do the pragmatists, who generally appeal to coordination only in abstract terms. Even given that high level of articulation, law and economics remains incomplete in instructive ways. That incompleteness reveals similar limitations in the pragmatist account and suggests precisely how and where it needs to be supplemented.

Second, the similarity in the two approaches is not haphazard: not only are both committed to functionalist explanations that remain relatively agnostic as to questions of substantive value, but there is a significant intersection in their theoretical resources. In particular, both Fuller and Postema appeal to game theory and related heuristic tools.[44]

Third, law and economics scholars have specifically addressed the question of indigenous landholding. An examination of how they have done so not only reveals the normative assumptions built into an approach that ostensibly focuses on coordination alone—and thereby creates space for alternative accounts based on very different assumptions—but it also helps establish that functionalist explanations of all kinds, including the pragmatist account, tend to provide only the most approximate explanations for why law takes the forms it does. A purely functionalist analysis may stipulate outer limits to the possibilities of legal form. It may state conditioning factors. But it leaves unexamined much of what determines the content of legal regulation. Many functionalist accounts smuggle in that additional content, encasing it in the account’s assumptions about the structure of human interaction. We will observe this smuggling in law and economics. One of the primary purposes of this paper is to reveal that additional content, incorporating it expressly into a theory of customary law.

To begin, in law and economics, the market is ostensibly used to secure the coordination of individuals’ diverse aims in a manner that does not prejudge questions of substantive value. These questions are left, ideally, to the participants’ own choices. But that agnosticism as to value can only go so far. For one thing, the very structure of law and economics manifests greater commitment to substantive ends than is initially apparent. This valuation plays a crucial role in the choice of the market itself as a mechanism for reconciling disagreements. Moreover, it becomes indispensable when true markets do not exist and the theorist has to predict what an equivalent outcome would be, as is often the case in the economic analysis of law. In the hands of Richard Posner and many other law and economics scholars, the foundational value is wealth maximization.[45] For Michael Trebilcock it is individual autonomy.[46] What is more, all the best practitioners of law and economics acknowledge that whatever the basis for the commitment to markets, law and economics can only provide an incomplete theory of law. They freely acknowledge that competing principles, founded on different conceptions of substantive value, have to be part of a full theory.[47]

This incompleteness is often conceived in terms that suggest that the norms shaping the market have to be balanced against other values. An example would be the debate over the commodification of human body parts, which is generally discussed as though it were about constraining the pursuit of wealth in order to protect the sanctity of the human body.[48] But law and economics is incomplete in a more foundational sense. It is not just that market mechanisms need to be balanced against other principles—as though a system of law could work on market principles alone, but in a rudimentary and impoverished way. Rather, those very mechanisms presuppose a definition of agents and entitlements that the principles sustaining the market take for granted.

Law and economics focuses, overwhelmingly, on how agents deploy their property. Its natural focus is transactional, as is evident in the prominence of supply and demand curves in its textbooks and the very use of markets as the theory’s principal analytical tool. It has much less to say about who counts as an agent or the particular structure of agents’ property rights.[49] But any market-based system of law has to be founded on a reasonably determinate conception of those matters. It has to identify who participates in markets, what resources they control, and what precisely control consists of. Those conceptions make up the bulk of what are called (in the civil law tradition) the law of persons and the law of property. They provide the necessary elements out of which markets are created. Built into those elements are normative assumptions that go well beyond the simple coordination of individuals’ divergent desires.

Take the example of slavery. Law and economics scholars naturally assume that all adult human beings have legal personality and that they cannot be owned—that the descendants of slaves can participate in market transactions on a basis of formal equality with all others. But of course, this is a recent development in the legal traditions of Europe and is based on principles that go well beyond the maximization of wealth. It is true that slavery can be analyzed in terms of market efficiency. There is, for example, a large literature on whether slavery was an efficient mode of production in the antebellum United States, with the consensus now being that there were circumstances in which slavery was efficient.[50] But even those debates necessarily beg the question, “Efficient for whom?” and implicitly answer, “For the non-slave population.” Law and economics scholars rightly foreclose such answers in their prescriptions for today’s law, and they do so because they take all people as agents, not objects, for reasons that have nothing to do with market efficiency.[51]

The same might be said for other principles that define the parties to legal orders. Slavery involves an extreme denial of civil equality, but one can apply similar reasoning to more limited forms of inequality—for example, the use of zoning requirements to impose racial limitations on who can purchase property in particular areas. Those limitations also determine the capacities of the persons that then participate in legal transactions. Some definition of legal personality, equal or unequal, must underlie the markets that law and economics scholars deploy. Other dimensions of legal orders also serve to define the actors who then engage in legal transactions. Migration and naturalization, and the regulation of the age of majority, specify who holds various sets of rights. Family law imposes conditions on individuals’ use of property. Until not very long ago it severely limited women’s legal personality (of course, what qualifies as a family is itself a matter of current dispute). All of these structures embody fundamental normative choices, choices that do not depend on considerations of wealth maximization and that must be resolved, one way or the other, before markets can be constructed. If that seems surprising, it may be because we have come to take the outcome of so many of these choices for granted.

The continued significance of these choices is clear if one considers one salient contrast between indigenous and non-indigenous orders: the status of animals as bearers of rights and obligations. The James Bay Cree norm invoked above—that the remains of animals must be treated with respect—is premised on the assumption that animals are persons in society with humans, that they give themselves to humans in the hunt, and that, in return, humans incur obligations to them.[52] Law’s markets would appear very different if animals formed members of the society within which law’s efficiency was calculated.

Similarly, when one turns to what one owns, one finds normative judgments encoded in the very structure of property rights. What, for example, does one own when one owns a tract of land?[53] Although left unexamined in the great bulk of law and economics literature, this is a non-trivial and value-laden question. Even among market-based legal orders, the ownership of land is structured in profoundly different ways. The common law conceives of such interests as bundles of particular rights that can be easily disassembled. The civil law in the French tradition works with a notion of full ownership: it presumes that the most natural relationship to the land is one of complete dominion and treats all limited interests as temporary departures from that state. Different legal systems conceive of the physical and temporal boundaries to rights in land differently, some excluding mineral rights, some treating rights over water or airspace separately from rights over land, some allowing people to own highly artificial bundles of rights with respect to apartments (condominiums), some allowing ownership to apply to limited periods of time, recurring year over year (time-sharing). The precise way in which one resolves these issues reflects the conception of ownership’s social role and understandings about human beings’ relationship to the resources in their environment. But what is true of all of the solutions is that the structure of the interests is not merely left up to individuals; it is to some extent written into the structure of the law itself.

Now, law and economics scholars do subject these interests to judgments of efficiency. Like most, Posner presumes that absolute rights of ownership are the most economically efficient, at least in situations of low transaction costs where markets can operate effectively.[54] This supposed efficiency was indeed a crucial reason for the abolition of feudal tenures in early nineteenth-century Europe and North America (or, in the common law, the evisceration of such tenures). Absolute rights of ownership were created precisely to permit the development of land free from the constraints of competing interests.[55] But how does the presumption of absolute rights fit with the division of ownership into increasingly complex regimes such as condominium or time-sharing, or the creation of finely divided property rights to permit the commercialization of highly specific aspects of things? Multiple rights have been recreated in these regimes, so that once again the ability to reconfigure or alienate things is constrained, often with the support of law and economics scholars (and of course, the very notion of what constitutes a thing is far from obvious[56]).

The fact is that when applied to the structure of rights, the interests of efficiency are frequently ambiguous. This problem is general, especially because the boundaries of rights often have to serve for an undefined but large number of potential uses, and therefore form at best a compromise among various efficiency claims.[57] The ambiguity is more pervasive than much of law and economics scholarship suggests.[58] As a result, the definition of things tends to precede, not follow, the analysis, and considerations other than efficiency shape to an important degree scholars’ conception of the way in which rights over things ought to be structured.

This implicit structure is clear when one looks at indigenous rights to land, which are often based on very different principles concerning relationships between human beings, the land, and resources—principles that shape how life on the land is regulated. There is, in fact, considerable discussion of indigenous rights to land in the law and economics literature. Many of those discussions focus on the emergence of property rights against a backdrop of common property. They elaborate a theme of the “tragedy of the commons”, in which the absence of property rights in indigenous societies leads to the overexploitation of resources and individual rights of ownership are then developed to institute an economically efficient pattern of use.[59] These discussions are not, however, all they seem.

First, their high abstraction becomes clear when one realizes that they discuss the emergence of property rights not against a backdrop of the complex rights held by families, clans, or kinship groups in actual indigenous societies, but against the assumption of an unregulated free-for-all. “Common property”, in the canonical law and economics accounts, consists of utterly unregulated resources of which anyone can avail themselves at will. But this is not an accurate description of any North American indigenous society, at least none for which we have any evidence. Those societies all had and generally still have sophisticated rules for the harvesting of resources. Thus, while new land-use regimes may have displaced prior ones, often complex and well-elaborated ones, they did not simply crystallize out of a propertyless state of nature.

Consider, for example, the detailed regulation of resource exploitation among the James Bay Cree. The James Bay Cree are a subarctic people, living in a region that can support only a modest population across a vast area. Hunting groups would move extensive distances seasonally in order to harvest particular resources. The Cree would, and to a large extent still do, divide into smaller groups—often of about four families—to hunt game for much of the year. They would coalesce into larger congregations at other times, especially for the goose hunts.

Yet the lands and animals of the Cree lands are subject to a complex regime of regulation. In recent times, this regime was (and indeed is) constructed around the hunting territories used by particular hunting groups. Each territory (of vast extent—an average of 1302 square miles in the northern reaches of the Mistassini nation’s territories, an average of 300 to 550 square miles further south[60]) is under the stewardship of a group leader who manages the harvesting of game. The consensus of recent anthropological opinion is that these territories build upon the regulation of hunting in the pre-contact period. They are primarily concerned with managing wildlife and maintaining relations among the Cree, not directly with controlling land. The description of the territories is therefore geared to the demands of harvesting game; their definition tends to be vague around the edges. The composition of hunting groups is varied from time to time in response to local scarcity of game, the need of a hunting group to rest its land, or more personal factors. On the territories, different wildlife resources are subject to different controls: hunting for furs is confined to the hunting group that holds the territory, subsistence resources are available to groups passing through. There are highly detailed ethical and ritual injunctions to be followed in hunting and in dealing with animals’ remains.[61] One sees then, among the James Bay Cree, complex principles for the regulation of land and wildlife. The situation is a far cry from the unregulated free-for-all of the “tragedy of the commons”, sufficiently far that it undermines the complacent assumptions underlying that story. Indeed, there are instructive analogies between Cree interests in land and traditional common law interests, although the interests are conceived in substantially different terms.[62]

Martin J. Bailey, in one of the most careful and original studies of indigenous property rights in the law and economics literature, recognizes the existence of strong mechanisms for regulating even collectively held resources. As a result, he treats family-held property, even property held by an extended family, as “private property” (and he suggests that property owned “by a large lineage group” might also be considered private).[63] Even with that expansive definition of private property, he argues that what remains as “common property” is often optimally allocated when held in common.[64] The realization that property held by collectivities, even very large ones, might still operate efficiently allows exceptionally wide latitude for different ways of structuring both who owns the right and what can be owned and to what extent—all within the constraints of an economically viable social order, as Bailey acknowledges.[65] This latitude suggests that, far from being an important determinant of the structure of economic interests, economic efficiency plays merely a conditioning role, with the structure being determined by other considerations.

The story generally told by law and economics scholars about the emergence of private property is best considered a myth of the emergence of economic man—akin to the political myth of the establishment of organized societies through social contract—rather than a plausible ethnohistorical account. The accounts of the rights’ emergence are not grounded in observation. They are projected back upon an imagined evolution of human resource use from free-for-all to private property. This evolution is structured by a predetermined story of progression from “primitive” to more advanced societies where the measure of advancement is precisely the emergence of private property rights.[66] My point is not that indigenous societies were infallible stewards of their resources, always regulating resources so as to avoid overexploitation. Like any other society, indigenous peoples could exhaust a resource as a result of miscalculation, the discovery of a destabilizing new technology, ruinous competition with other groups, or pressures caused by resource commercialization or population displacement following European settlement. My point is rather that a wide range of structures for regulating resources appears to be consistent with sustained economic activity over time.

In fact, in their accounts of the potential efficiency of various structures of property rights, law and economics scholars often appear to proceed post hoc. The economic analysis does not so much explain the configuration of the rights as take that configuration as its starting point, and then seek to describe how rights of that kind might be compatible with efficiency.[67] The efficiency analysis is read back across rights that are, in effect, taken as given. This approach can be valuable. It can, for example, reveal the effectiveness of unfamiliar legal forms—forms that might otherwise be discounted simply because they are unfamiliar. But one should not reverse the direction of inference so that one presumes that efficiency requires a single model for all human societies.

In law and economics, economic efficiency is often treated, at least implicitly, as though it were a sufficient determinant of law. No one doubts that efficiency is a consideration and that any legal order that profoundly impaired a society from sustaining itself would have great difficulty surviving. But it may well be that economic efficiency serves more as a limiting set of parameters—or better, a set of conditioning factors—than anything approaching a complete determinant. If that is the better description, then a wide range of possible legal forms may be compatible with economic success, and, within that range, the particular form of a society’s law may be determined by considerations that have nothing to do with economic efficiency.[68]

Now to return to the pragmatist account of customary law. Here too, the primary role of law is taken to be coordination. Law is shaped by the practices within a particular society and exists to facilitate those practices. The pragmatist account does not require that coordination occur through markets, nor that coordination be shaped by a commitment either to individual autonomy or to wealth maximization. It is open, then, to a much wider range of ways of coordinating human action than is law and economics. But this account continues to emphasize a pragmatic test for law: does law serve to foster communication and coordination among members of society, such that their mutual cooperation is facilitated? This question underspecifies law in a manner very similar to what we have just seen in law and economics. It runs the risk of collapsing into a tautology—laws exist, therefore they coordinate human action—when in fact coordination does virtually no work in explaining how a legal order relates to the practices of a particular society.

In any society, coordination has to occur in some determinate way. The full specification of the mode of coordination in that society will necessarily involve a richer set of normative principles than is captured by a bare commitment to coordination. Coordination captures the aspiration to order but says virtually nothing about the terms on which order is to be achieved—to take the examples discussed in this section: who counts as a person for the purposes of legal relations, or what are the forms through which things can be held and deployed. In those terms, an additional array of normative considerations will be operative. These considerations are an equally fundamental dimension of any legal order and are generally of more relevance than coordination in any particular application of the law. To adopt a linguistic analogy that will be developed further in this paper (and that builds upon the analogy to language that was so important to Fuller): these distinctive terms constitute the grammar through which the particular legal order achieves social coordination; they constitute the grammar which any participant in the order has to employ.[69]

It is essential that every legal order have such a determinate language through which coordination is achieved, not that this language be the same for all legal orders. The foundational categories of the grammar can differ profoundly in different legal orders. This difference was the essential message of Clifford Geertz in his immensely influential paper, “Local Knowledge: Fact and Law in Comparative Perspective”.[70] The understanding of what constitutes a fact and what constitutes a law, who are legally relevant agents, what are their intrinsic relations, how they are located within the natural world, what types of relations are written into that world, perhaps even what counts as “order”—are all structured, at least in some considerable measure, by presuppositions that are not derivable from the aspiration to order itself. These presuppositions are built into the terms that people bring to the very effort to coordinate their actions.

As an example, consider the array of kinship regimes that exist in all societies. These regimes determine people to whom one is related and whom one can marry, sometimes introducing distinctions that are entirely foreign to non-members. Among the Anishinabek of northern Ontario and Manitoba, for example, one refers to one’s “parallel cousins” (the children of the male siblings of one’s father, or the children of the female siblings of one’s mother—although in each case, one needs to use the expanded notion of “sibling” that results from this system of kinship) as one’s “brothers” and “sisters”, who then fall within the incest taboo. In contrast, one’s “cross cousins” (the children of the female siblings of one’s father, or the children of the male siblings of one’s mother) are not siblings and thus are potential marriage partners.[71]

Kinship rules control marriage and norms of social interaction, familial obligation, patterns of inheritance, often the structure of political authority within a people, and often also shape diplomatic relations between one people and neighbouring ones. Undeniably, they play a coordinating role. One set of explanations in anthropology has emphasized this role, attempting to account for the practices in functionalist terms by noting that kinship rules build alliances and promote genetic diversity. But these explanations alone are unable to produce the great variety of kinship systems, a variety engrained in characteristics of the systems that matter deeply to their adherents and that concretely shape their relations. It may be that some general, structural distinctions (the existence of some sort of incest taboo, for example) respond to features of human reproduction and interaction that are challenges for all human societies. But the ways in which these relations are defined in practice are remarkable in their variety. Those differences encode different visions of human relationships. The best discussions of kinship in the anthropological literature take that substantive content seriously, exploring the webs of meaning through which familial relations are defined and organized.

In his explication of “salience reasoning” (the form of intersubjective, analogical reasoning that underlies the emergence of conventions), Postema argues that “a conceptually sophisticated, but often transparent, deliberative framework” gives distinctive form to agents’ common normative reasoning. This framework

is defined by a conceptual template and associated attitudes that together shape deliberation by relating deliberators and their decisions to the deliberations and decisions of others and to the outcomes of their individual and concerted actions. These deliberative perspectives or frameworks do not figure in the content of desires, goals, or principles of choice, but rather structure and delimit the deliberative domain in which such factors are used by agents to arrive at rationally grounded decisions. They provide the basic materials agents need (1) to conceptualize their practical environment, (2) to analyze the particular problem of choice and action they face, colored by the desires, aims, goals, and values to which they are committed, and (3) to survey feasible solutions.[72]

Postema describes this framework by identifying six features that relate to collective agency, the form of participants’ intentions, and their reliance upon a common stock of knowledge. His description is illuminating in its exploration of structural features of collective moral reasoning but it remains abstract and formal. I am suggesting that the very language that participants use to conceptualize and analyze—the very concepts they employ to conceive of normative challenges—have particular normative dispositions inscribed into them. These dispositions shape participants’ deliberations and, alongside participants’ reflections on the stock of common experience, account for the salience of potential solutions.[73]

In a backhanded way, Fuller demonstrates that rich normative content is built into the very legal categories through which social coordination is achieved. In his writings on the rule of law, he sets out a number of principles that he argues are inherent, at least as dispositions, in all legal systems; they constitute an “inner morality” of law.[74]

The content has a distinctly liberal individualist cast. It is premised on the expectation that law consists largely of a set of explicit rules intended to allow self-directing individuals to organize their activities in a manner that coordinates with the activities of others[75] (as opposed, for example, to norms communicated in narrative form, or that are focused primarily on activities conducted fundamentally in concert). Fuller’s account is also, like most accounts, idealized: it does not discuss differences of resources or of social position that substantially affect individuals’ ability to be self-directing, or how different systems of property rights may foster or discourage these inequalities. Fuller treats the content he identifies as though it were inherent in the very idea of law. But surely the existence of quite different ways of ordering societies suggests that this is not so—suggests, in other words, that Fuller was unpacking the implications of a particular legal tradition and generalizing them. Paying attention to other traditions may reveal that richer substantive choices are built into Western legal orders than we generally perceive—choices that are rendered invisible if we treat law simply as a means of coordination for individual actors. It may also reveal alternative forms of social ordering from which non-indigenous institutions might learn.

Now, by emphasizing that such differences exist, I do not mean to suggest that they are usefully analyzed in a simple dichotomy between individualist and collectivist orientations. Every legal order is addressed to people who are in very significant ways self-directing. Every society wrestles with the dual nature of persons both as members of a community and as actors in their own right and resolves that tension in a manner that gives some rein to both. One conclusion that can be drawn from the foregoing discussion of law and economics, for example, is that in liberal societies such questions as who counts as an agent, or what are the structures through which persons can have rights over things, are in effect determined at a collective level; they are not subject to the unencumbered choices of individuals.[76] Indigenous societies may answer these questions very differently. They may conceive of persons’ relations to the land differently (for example, building a notion of stewardship, of responsibility, into indissoluble unity with notions of ownership). They may conceive of those who possess rights differently (perhaps privileging units based on kinship, although note that family property laws in many Western countries also do this to some extent). It would be a serious error to presume that the result is necessarily less respectful of individuality simply because it differs from Western legal orders. The differences in structure may form such a complex geography that it is not possible to arrange them on a spectrum from individualist to collectivist.[77]

B. The Requirements of Legal Community

The error of such a presumption becomes clear when we look at the next neglected dimension of customary law. Those of us who live under Western legal orders tend to assume that, to have a legal order at all, there have to be authoritative ways of resolving disputes, with the solution then being imposed on the parties. Even theories that de-emphasize enforcement tend still to emphasize that norms have to be authoritatively determined.[78] Thus, although Western legal orders may permit a wide sphere of individual autonomy, that autonomy occurs within a normative framework that is subject to centralized determination so that, from the point of view of any individual participant, the legal order is fixed by a mechanism independent of his or her will. Indeed, the existence of such an externally determined order—clear, certain, and enforced—is often taken to be the precondition of individual freedom.

The situation is very different in many indigenous orders. When colonists encountered indigenous societies in North America, they were often struck by the extent of individual autonomy in those societies. It was amazing to them that such polities could survive at a time when European societies tended to be hierarchical, authoritarian, and draconian.[79] Among many North American indigenous peoples the respect for autonomy extends to the very interpretation of society’s norms. There is great reluctance to impose a particular interpretation of the law either on any member (in some societies) or on someone of high rank (when the society is hierarchically ordered). Such an imposition is considered deeply incompatible with the person’s dignity. Indeed, this respect for a person’s moral autonomy may contribute to the prevalence of banishment as a punishment in many indigenous societies: rather than forcing compliance, the community treats offenders as having, by their conduct, placed themselves outside society.

But how can such orders function? For one thing, they tend to rely heavily on deliberation, diplomacy, and mediation. Norms are determined through discussion designed to promote the emergence of consensus without loss of face on the part of any participant. In a wonderful paper, Val Napoleon has described the functioning of a feast in resolving a controversy among the Gitxsan of northern British Columbia.[80] That feast had been called by one chief as part of the raising of a totem pole carrying a disputed crest. A chief from another village claimed that the crest was his and denied the right of the first chief to use it. In the dispute, the chiefs of clans related to those of the disputing chiefs (in a specific manner prescribed by Gitxsan law) played a mediating role, and the aggrieved chief himself stayed away from the feast and the pole-raising both to indicate his displeasure and to avoid open conflict (although his allies did attend). At the feast itself, those in attendance (importantly including chiefs not directly allied to the disputants), through a series of speeches, recounted histories, juxtaposed those histories, supplemented, corrected, and stated their approval for them, until gradually a salient position began to emerge. The aggrieved chief was treated with honour in other ceremonies at the time of the pole-raising, ensuring that the dispute remained limited to the issue of the crest alone. The aggrieved chief’s allies discussed the proceedings of the feast with him, assured him of their support, noted the damaging consequences of continued conflict, but then left the decision of whether to acquiesce in the emergent solution to him. Ultimately he did.

This mode of decision making depends directly on the cost of withdrawal from the order—or, to put it positively, on the parties’ willingness to continue to live together—as the principal inducement to secure compliance with the society’s law. Within broad but significant limits, in virtually any order, participants have a substantial interest in acquiescing to the order’s norms in order to obtain the benefits of social cooperation, even if those norms are not exactly what they would wish. In non-indigenous societies, this interest has been overlain by the structure of the state, which uses coercive authority to maintain the society’s norms and boundaries. In contrast, indigenous societies have tended to be much more willing to tolerate continued disagreement among leaders with respect to the society’s laws, they have often relied upon the withholding of cooperation rather than compulsion as a principal sanction, and they have also generally been more willing to accept the withdrawal of a dissenting section from the community rather than require compliance.[81] As a result, reconfigurations of indigenous societies do occur, although they are rare because of the significant costs incurred when one goes it alone. Moreover, when reconfigurations do happen, generally they are partial, with many ties maintained. The result is a system that comes much closer to being self-enforcing, founded, at least apparently, on voluntary adherence.

Now, this does mean that indigenous societies also tend to be small-scale, structured around complex relations among members—relations often conceived largely in terms of kinship. The exercise of authority within those societies does not have the impersonal, formal character that one associates with the liberal state. Even here, however, the situation is not quite as clear as it may appear. Kinship relations are more than just familial relations. They establish structures of membership and obligation that extend well beyond one’s close family. They can extend over great distances to people one has never met and can even bind one to members of neighbouring peoples.[82] The kinship relations at the foundation of many indigenous societies often seem, then, to have something of the formal, ascribed character one associates with citizenship. Moreover, much of the law of those societies is also conceived in terms that tend to separate it from the personal wills of the society’s current leaders. Among the Gitxsan, for example, this conception includes the inscription of the law in ancient histories and lineages, and the sense that today’s leaders carry on responsibilities that extend well beyond their lifetimes, an understanding reinforced by the inheritance of chiefly names and the Gitxsan belief in reincarnation.[83]

I do not mean to minimize differences that exist between indigenous and non-indigenous social organization. Of course, kinship creates a complex structure of internal differentiation, stipulating differential rights and obligations that control such fundamental relations as marriage. Societies based on kinship are also relatively impervious to outsiders (although this is moderated by the possibility of adoption). There would appear to be limits, then, to the scale of government, openness to new members, and ability to choose one’s life partner within forms of social organization founded on kinship. Most contemporary indigenous societies have tended to loosen or move away from strict rules of kinship, seeking to develop modes of government that combine elements of their traditions in new ways, often joined with statelike structures.[84] My point is simply to note the possibility of customary legal orders founded on very different premises of social organization, and to draw attention to the particular combination of consensual adherence and embedded structure common to many indigenous societies. The societies are not founded simply on consent, although consensual processes do play a prominent role in the interpretation and application of their law. Rather, they presuppose a web of relations that is largely taken as given.

This web draws attention once again to the ways in which social ordering always takes place using a particular language of social interaction.[85] And it also emphasizes the necessary copriority of a sense of community, which any order of consensual decision-making, indeed any order of justice, has to presuppose.[86] As Charles Taylor has argued regarding the autonomous individual on which Western political theory is based,

The modern subject is ... far from unmediated in fact. He may be, relative to the local community; but he cannot be, relative to the whole society. On the contrary, he is sustained on the one hand by the culture, which elaborates and maintains the vocabulary of his self-understanding; and on the other by the society in which he has a status commensurate with free subjectivity: ... the equal bearer of rights, who is producer and citizen. All this underpins my identity as free individual, which could not long survive a state of nature.[87]

Indigenous peoples do not generally draw upon the same frameworks of understanding as non-indigenous societies, although they do cherish conceptions of human freedom. Those conceptions take form against a backdrop of social commitment engrained in an ordered structure of society.

Comparisons across different orders of customary law allow us to see these contrasting frames and to realize more clearly what is at stake when one frame is forcibly imposed upon another, as settler societies have long attempted to do with indigenous societies. They also allow us to see roles and possibilities for change that are characteristic of non-indigenous institutions, but that are obscured by the fact that we take those institutions for granted.

For example, in recent years considerable attention has been paid to the existence of socially generated normative orders—to legal pluralism—in non-indigenous societies. Many approaches to such orders, it seems to me, implicitly see that pluralism as existing within structures whose essential stability is provided by the state. They conceive of non-state legal orders operating, in other words, within an arena of social relations that state institutions play a large part in maintaining.[88] Studying indigenous legal orders may help us understand what is necessary to sustain truly self-sufficient, non-state legal orders. Those requirements may be something like the tight, stabilizing web of relations provided by the structure of kinship. If kinship relations are eroded, some other structure, perhaps adapted statelike institutions, may have to be developed in their place.

Studying indigenous societies may help us understand the capacity to sustain viable polities without insisting on a highly centralized legal order—polities that allow for divergent understandings, permit significant normative autonomy among subunits, and foster cohesion through practices of negotiation and consensus-building, rather than through authoritative interpretation and imposition of a centralized order. Recent developments in Canadian constitutional law suggest a willingness to contemplate such features as integral to the Canadian constitutional order, a willingness arguably shaped in part by Canada’s fraught experience of treaty relations with indigenous peoples.[89]

C. Sacred or Mythic Dimensions of Law

Thus far, I have tended to discuss aspects of customary law that fit reasonably well with the view that law’s purpose is to structure and control human practices. But even a cursory acquaintance with North American indigenous orders reveals a pervasive emphasis on the sacred, on the mythic. These elements do not fit at all well with a purely pragmatist conception of legality.

My colleague John Borrows, for example, has begun a lecture by holding a rock from near his home community of Neyaashiinigming, describing how the law is inscribed in that rock.[90] Val Napoleon speaks of the importance of reincarnation to the Gitxsan legal order.[91] Kiera Ladner describes the ways in which Blackfoot governance is shaped by lessons learned by considering animal behaviour.[92] As we have seen, James Bay Cree conceive of their relations with animals in terms of a sociality similar to that existing among humans.[93] Many indigenous peoples ground their law in narratives peopled with non-human beings. How then should those of us who are not members of those societies understand this dimension of customary law, given that we do not share the metaphysics?[94]

An important starting point is to realize that non-indigenous law too draws on metaphor, myth, and narrative.[95] This is certainly true of constitutional law, the area in which I tend to work. Canada’s constitutional discourse has, for example, been shaped in important ways by the assertion that Canada has “two founding peoples” (French and English) and that it was created by a compact variously defined as between French and English, between Quebec and the rest of Canada, or between all of the original colonies that ultimately formed Canada.[96] These stories have been used to claim that Canada has two linguistic cultures of roughly equal stature. They have been cited to support constitutional interpretations in which provincial autonomy is paramount.

There are, of course, other constitutional narratives that cut across these stories. One such narrative was captured in the phrase “from colony to nation,” which emphasized the progressive development of Canada toward a more perfect nationhood. That phrase now appears archaic, but its underlying story has real staying power.

Along with others, I was present at a remarkable after-dinner conversation between former Prime Minister Pierre Trudeau and the Canadian political philosopher Charles Taylor. The conversation took place in October 1994, at the Faculty Club of McGill University. It turned on whether the Canadian constitution should have been patriated in 1982 over the objections of the government of Quebec.

Prior to patriation, the Canadian constitution had been a statute of the British Parliament, amendable only in Britain (although, by convention, amendments were passed only at the request of Canada). As prime minister, Trudeau had vigorously pushed for patriation, and in 1981 he succeeded in obtaining the agreement of the federal government, and all provinces except Quebec, to a package that included a charter of rights and a domestic amending formula. He then put that request to the Government of the United Kingdom and the proposals were adopted. The making of that request, over the objections of the government of Quebec, was (and is) a matter of deep controversy in Quebec, not so much because of the content of the proposals but because of the failure to secure Quebec’s agreement. That dissatisfaction contributed to a resurgence of the nationalist movement in Quebec.[97]

The conversation at the McGill Faculty Club between Trudeau and Taylor occurred in the lead-up to the October 1995 referendum on Quebec’s independence, which came very close to succeeding. Taylor questioned Trudeau on the wisdom of proceeding with patriation without Quebec’s support. In the end he posed this question: “If patriation ultimately leads to the breakup of the country, will it have been worth it?” Trudeau thought for a moment and then replied, “A country that cannot amend its own constitution does not deserve to exist.”

Trudeau’s response reflects a metanarrative of what it takes to have a country. It is not a product simply of functionalist or rationalist analysis. Such myths and narratives do considerable work in both constitutional politics and constitutional adjudication. The compact theory was frequently asserted in connection with the interpretation of the Canadian constitution, and it was recently cited again to support the position that Quebec could unilaterally secede from Canada. In the postwar period, the idea that Canada has become a true nation, represented by the government of Canada, has underpinned interpretations of the Canadian constitution that have allowed an expansion of federal authority. And one could cite comparable myths in other countries: the long influence of the idea of “the ancient constitution” in British constitutionalism;[98] the contest over the fundamental characterization of the United States in the period up to and through the Civil War; and appeals to the “social contract” in the constitutional discourse of many countries.

In the private law, consider, for example, the myths that underpin specific conceptions of property: the idea that property is based on the transformation of land through labour, or the often religiously framed conception that humanity has an obligation to use the earth most productively. Both these myths contributed to the refusal to recognize indigenous rights in land and have long influenced statutory and non-statutory rules for the pre-emption of land. Myths also underlie the conception that rights of ownership are absolute as a matter of natural law because they are founded on the capacity to consume things, an idea that has shaped the structure of property rights in the civil law.[99] And, as we have seen, the tragedy of the commons is best considered to be a fable about the emergence of economic man. Myths, which continue to shape legal interpretation, are not confined to indigenous peoples.[100] And indeed, indigenous peoples have often been on the receiving end of European myths of origin, not least those concerning the conditions of the “state of nature.”[101]

In non-indigenous societies, these myths now tend to be secular rather than sacred. But one nonetheless still finds elements in the law that verge on the sacred, or at least on a strongly charged aesthetic. This is true, for example, in the invocation of the sanctity of the body, which shapes criminal law, the law of civil trespass, obligations with respect to the disposal of human remains, the capacity to hold property in human remains, and the regulation of the commercialization of body parts. Similarly, many of the concerns of environmental law are informed by conceptions of natural purity, beauty, wilderness, and stewardship that go beyond purely functionalist concerns. Such fundamental aspects of the private law as legal personality, human equality, the family, and the rights that one can exercise over land have historically been shaped by religious and metaphysical concerns. That shaping is less obvious today, but that may be because we share the commitments on which today’s conceptions of legal personality and property are based. Moreover, this sacred/aesthetic dimension is not separable from the normative. It is not, for example, merely a lens through which one might examine the law, as one might read a legal text as though it were a literary text. The sacred/aesthetic dimension is integral to the norms; it provides much of their content and furnishes the frame through which they continue to be interpreted.[102]

One of the ways in which such myths operate is by helping to shape the normative sensibility of those involved in applying and obeying the law. The language of legal texts is never self-sufficient. Judgment is required to interpret and apply them, and judgment depends upon a close understanding of the aims underlying the rules, the considerations at play in practical situations, the ways in which the rules’ aims relate to those considerations (including when situations are significantly similar and significantly different), and the manner in which rules should be shaped by competing rules (including the relative importance of rules, and the relationship between the normative principles underlying the rules). These elements of judgment are all the more important in the case of customary law where there is no canonical expression of a rule. Instead, principles of conduct are worked out incrementally, by reference to existing understandings of social norms and practices.

Metaphors, myths, and narratives assist this judgment by providing an overarching framework to guide interpretation and by suggesting the relative importance of norms. They can supply a range of examples of the norms in action, furnishing models for how the norms should be applied. They often record a wide swath of the experience of social interaction, weighed, assigned value, and ordered, against which possible formulations of norms can be tested and refined. And they give salience and a memorable quality to certain norms, so that those norms are retained, internalized, and sometimes made a focus of identification and allegiance.

All legal orders, including non-indigenous ones, use metaphor, narrative, and myth in these ways. However, note that this discussion has tended to translate the sacred into other terms. It has domesticated the sacred by treating it simply as the bearer of normative content or as a way of reinforcing elements of normative judgment. The supernatural dimension—the spirit world, the dimension that tends to be all-important to a believer—falls out of the analysis. Surely we need to take those metaphysical assertions seriously if we are to deal with claims of the sacred in a manner adequate to their force and meaning for believers.

The problem is a real one, and differences of metaphysics can certainly cause problems in interaction across normative orders. To see how seriously, consider an argument floated by historian Keith Carlson concerning the 1906 delegation of Salish leaders from British Columbia to King Edward VII. That delegation sought to secure redress for a long list of grievances, and specifically sought the fulfillment of promises said to have been made by the first governor of the colony and later by the Governor General of Canada. The representatives toured London, met with the King, and returned to argue that the King had given them assurances. But the problem is that there is no independent evidence that King Edward made those promises in the short audience he granted the delegation. Carlson wonders whether in fact the assurances had been made, in the delegation’s eyes, by a much earlier King Edward—Edward the Confessor or Edward I—whose tombs and relics (in one case a shrine, in the other the coronation chair then containing the Stone of Scone) the delegation had visited in Westminster Abbey and which had greatly impressed them.[103]

Now, this particular interpretation is presented as a conjecture by Carlson, but it is undeniable that many indigenous peoples believe in the ability to communicate with ancestral spirits, accept the existence of connections between persons alive today and those in the past (especially those bearing the same chiefly names), believe in the existence of spirits associated with objects non-indigenous people consider to be inanimate, or believe that human beings are distinctively connected to particular animals or tracts of land.[104] Each of these beliefs has strong normative consequences, and it has those consequences precisely because of the metaphysical significance of the beliefs. Moreover, in cases of cross-cultural interaction, there will be situations where one needs to adjudicate among these beliefs, such as in Keith Carlson’s hypothesis of a promise made by an ancestral spirit, or a claim that indigenous land is inalienable because of a people’s sacred ties to that land. How can we come to terms with different visions of the spirit world?

I am too much a child of my own society simply to say that we should accept the truth of these spiritual beliefs. One does no honour to deeply held beliefs if one treats them as though their truth or falsity was of so little consequence that one can accept any belief as true. But I do know that we have to worry over the implications of these understandings of the world, not dismiss them out of hand.

Although there is no complete solution, precisely because there are deep differences that divide us, we can do the following: First, we can keep in mind the influence of mythical or metaphorical elements—even conceptions of the sacred—in non-indigenous legal orders, so that we realize that at least part of the problem is one of divergent myths, divergent metaphysical beliefs, supporting different norms. At least at that level, the problem continues to be one of pervasive normative diversity. Second, we should also keep in mind that there are broader and deeply held spiritual claims that go beyond mere metaphor, claims of great significance to believers.

Third, we should ask, then, in any particular situation, whether there is reason to forbear, not because we accept the truth of the beliefs, but because of their demonstrable importance to the people concerned. This is the classical foundation for religious toleration.[105] It is particularly appropriate in the indigenous/non-indigenous context because of the impositions visited upon indigenous people as part of colonization. The coercive nature of that encounter can lead us to misperceive and misunderstand assertions—we can fail to attend, for example, to indigenous understandings of the obligations of the treaties or to indigenous conceptions of humans’ relationship to the land—or simply to dismiss beliefs, failing to see that they are analogous to commitments still influential within non-indigenous society. Our actions have impeded indigenous peoples’ ability to develop and express their distinctive understandings, not least by placing their languages and lands under heavy pressure. There is reason to make space.

Fourth, it is also worth considering whether we might learn something from indigenous understandings. This is not to denigrate our epistemological assumptions but rather to emphasize the continual value of humility. We often treat religious differences as though they are impervious to discussion, but Jeffrey Stout has argued convincingly that there are ways in which genuine deliberation can occur across metaphysical differences. He urges us not simply to assume that we cannot communicate, but to try and see.[106] When we do, we may find that we have access to valuable insights that we would otherwise have missed. There is, for example, now a substantial body of work on the insights to be drawn from indigenous knowledge in environmental management, insights that are themselves grounded in indigenous metaphysics, such as indigenous peoples’ assumption that animals are intentional beings in society with humans.[107]

Fifth, we should always remember that not every difference needs to be adjudicated: we can agree to disagree, persisting in our different conceptions of the ways in which humans relate to animals, for example, and mediating between those conceptions in our approach to the regulation of harvesting. That option is not always available. It would not be available, for example, if one sought to rely squarely on the promises of the Edwards. But that option (or partial variants of it) is available more frequently than we tend to think. We should be alert to those possibilities.

III. The Grammar of Customary Law

I began this paper by accepting several key features of the pragmatist account: the primacy and ubiquity of customary law; the grounding of customary law in a particular society’s practices; the process of reasoning and deliberation surrounding the derivation of norms from those practices; and the intersubjective dimension of that process of deliberation and interpretation, in which a social interpretation is determined by means of social institutions.

I have not accepted, however, the pragmatists’ suggestion that the content of customary law is best understood in terms of the facilitation of human interaction. That suggestion, it seems to me, provides altogether too limited an account of the normative content of law—so partial that it is misleading, detrimentally affecting the other dimensions of the pragmatist account, including the way in which law’s relationship to social practices is conceived. In this final section of the paper, I wish to offer a more adequate account.

First, it is important to give facilitation its due. By emphasizing law’s role in coordinating the actions and intentions of diverse individuals, the pragmatist account does focus attention on law’s enabling role: by settling the terms of interaction, law establishes the preconditions for effective cooperation. It also makes clear that for that settlement to be effective, the terms have to be socially determined and socially validated: they have to become engrained in action, dependable standards that people can expect to be followed by diverse parties (parties who may have different normative views from their own), not decrees dependent on one person’s whim. Fuller captures this social warrant of the law in his parable of the king who, in order to secure his subjects’ obedience to the law, must himself act consistently with the rules he promulgates. He thereby sets in motion a process by which the law is separated from his will and achieves a social grounding in the interaction of members of the society.[108] The pragmatists’ emphasis on law as facilitating social action directs us to this process of social determination, stresses the need for a common rule, and insists that this common rule—precisely because it has to coordinate the actions of diverse parties—must have some independence from the wills of each of those parties. However, Fuller also emphasizes, rightly, that the common rule is continually open to elaboration and revision as participants reflect upon their moral experience.[109]

All this is very important. It emphasizes law’s social operation, social efficacy, and social determination. It counters theories of law that emphasize merely the downward operation of sovereignty, or that try to derive the content of law from nothing more than a philosophical conception of justice.

But the emphasis on facilitation does not specify, to any significant degree, what determines the content of law. This is because an enormous variety of potential rules can conceivably furnish the means of social coordination. It is not the case that in order to secure human cooperation a particular structure of norms is required, so that one can explain the content of the law by pointing to the demands of coordination. Many different forms of law—different definitions of legal agency, different conceptions of property, different understandings of obligation and responsibility—can all meet that desideratum. For anyone working within a particular legal order, coordination will therefore play a very minor role in explaining the particular content of the rules, and will not provide sufficient criteria for criticizing and revising those rules. Instead, coordination occurs through distinct legal languages. Those languages set the terms through which coordination is achieved; indeed they set the foundational concepts by which norms and practices are conceived.

This language is the “grammar of customary law” signalled in the title of this paper. I use “grammar” in the broad sense in which Wittgenstein used the term: to include not just the express rules that purport to regulate the use of a language, but rather the way in which a language’s structure and terms enable and constrain what a competent speaker can say intelligibly.[110] Through the distinctive grammar of each customary legal order, much of the normative content of that law is encoded, organized, and expressed.

Sometimes the linguistic analogy is very close indeed. Consider the example given above of the Anishinabek use of kinship terms, where there is no separate word for immediate siblings (the offspring of one’s own mother and father) and the term for siblings is applied to a class of both siblings and cousins. Engrained in that linguistic practice is a clear disposition to treat those cousins and one’s siblings similarly, extending to them one’s close concern and treating them all as being within the range of the incest taboo. The very linguistic usage encodes that practice.[111]

On other occasions, normative dispositions are not so much encoded in the words themselves as enshrined in complexes of concepts. Thus, broader categories of kinship—membership in clans, relationship within a particular lineage—serve to shape relationships across whole societies and at times across national boundaries. These kinship categories regulate such things as entitlement to chiefly names, obligations of mutual support, and privileged roles as witnesses in ceremonies, which in turn serve to affirm laws and settle disputes. Similarly, foundational conceptions of land determine what it is permissible to do with land—indeed, they define the very object to which rights and responsibilities attach.

The linguistic analogy is apt because often these elements are built into our conceptual structure so deeply that it is only through considerable effort that we can conceive of them being different. Richard Overstall captures this well in relation to Gitxsan law: “Kinship ... is not just about how a Gitxsan person assumes legal responsibility for people and things—it is also about how he or she knows them. It is a knowledge system.”[112] The normative effect of these terms, the normative dispositions they express, are most readily apparent when we study legal cultures that are structured quite differently or encounter situations in which the possibility of a different characterization begins to be evident. The section of this paper dealing with the requirements of legal community provides an example. The encounter with indigenous legal orders suggests how it might be possible to conceive of and sustain a legal order that relies directly on participants’ own realization of the demands of continued social cooperation, without the centralized determination and enforcement of norms.

Of course, much of the normative content of any legal order is not engrained in the terms of the order. Any legal order contains room for debate in the use of those terms to deal with disputes. The terms themselves provide foundational concepts, structure the inquiry, and embody normative dispositions, but consistent with those linguistic resources, a range of arguments can be made. Here again the linguistic analogy holds true: languages shape what can be said—they make some things straightforward, other things very difficult to say—but they do not determine everything that is said. Languages can be deployed in making a variety of arguments.

The same holds true for the grammar of customary law. Val Napoleon, for example, has described the ways in which the Gitxsan have responded to the increasing prevalence of intraclan marriage by using adoption to restore the proper clan relationships. That solution is not perfect—members know that an unsatisfactory situation is being regularized and some would rather that the situation had not arisen in the first place—but the response continues to draw upon distinctively Gitxsan forms.[113] She notes the need to develop new norms to deal with Gitxsan who lack a clan (because they have married outside the community, or because of the accumulated impact of the exclusionary band membership rules imposed by the Indian Act).[114] And she recounts a rare instance in which the Gitxsan determined that a chiefly name be shared because one of the claimants was better situated to care for the resources of the territories attached to that name, while a second claimant was better suited to fulfill the chiefly duties in public deliberation.[115] Any society generates disputes about the content and application of norms. Normative orders are complex, subject to varying interpretations. Norms are always expressed in general terms; they require assessment and adjustment in their application to the vagaries of life in society. Deliberation is inherent in law.

It is useful, then, to recognize three levels of normative determination in customary legal orders. First, there is coordination, which emphasizes the need for a common order of norms and thus for a social process by which norms can be determined. This process in turn means that norms will have some autonomy from the will of individuals. Second, there is the language through which norms are expressed—the terminology, the conceptual structure, the basic architecture that determines who are subjects of legal rights, what counts as an object of legal rights, and so on—in other words, the grammar of the law, in which rules and principles are framed and through which coordination is achieved. Third, there are the particular debates that occur using these concepts—debates that weigh those concepts, deploy them, regulate disputes, and through which the concepts themselves are continually adjusted and refined.[116]

The second and the third levels are not strictly separate from one another. In part, this lack of separation exists because the conceptual structure of the law is never univocal. The structure does contain normative dispositions—it makes some things much more easily affirmed than others, and the very way in which it states issues tends to define a privileged class of solutions (such as treating marriage to parallel cousins as a form of incest, as in the Anishinabek example given above[117]). But like any language, it also has a measure of flexibility, so that alternative arguments can be presented. The concepts themselves can be subjected to criticism by, for example, probing and revising conceptual distinctions. And over time, the grammar of the law will evolve as a result of core concepts being applied to new situations, new concepts being grafted onto the original concepts to address new challenges, adjustments to the original understanding of concepts as their implications are explored in varied circumstances, the reformulation of concepts as they are juxtaposed with others in the normative order or are extended by analogy to speak to additional situations (such as the use of adoption to regularize marriages that violate kinship rules, as Val Napoleon has described[118]). Thus, the distinction between the second and third levels is more one of degree than of kind, various concepts being embedded to different degrees in the legal culture.

Nevertheless, it is useful to distinguish between the levels in order to capture the fact that some concepts become so engrained in the way we think about issues that they become part of the landscape, taken for granted in day-to-day applications of the law. Even if it is possible to work around these concepts, that work-around requires very substantial effort to redefine terms, justify redefinitions, and then deploy the redefined terms in a manner that runs against the established grain.[119] Such reconceptualizations are much more difficult to achieve than arguments that take the existing terms as given. To take one example, it is easier to argue that the concept of marriage is able to embrace same-sex unions than it is to argue that the very notion of marriage should be reconceived to recognize an array of very different unions, including relationships that have no heterosexual equivalent. What is more, the underlying conceptual structure is more likely to be shared than are specific normative judgments—although by “shared” I do not mean agreed, but rather that it furnishes the terms through which participants have come to define their normative positions.[120]

In all of this, social practices serve as an essential touchstone for the evolution of the law. Arguments that people make about norms inevitably refer to how those norms are likely to operate against the complexity of human interaction. But those practices, that interaction, are not the purely functional relations that the pragmatist accounts portray. The practices are complex amalgams, already infused with linguistic categories and normative dispositions (as, for example, in the case of kinship or marriage).[121] The practices are best conceived in Wittgensteinian terms, with normative language entirely interwoven in them.[122] And the relationship between norms and practices is not one of simple facilitation. Rather, it is better expressed as one of normative consonance, in which one’s arguments address whether a particular norm is likely to fit well with the judgments of parties engaged in practices, to foster the achievement of ends to which the practices are directed, to instill in that interaction qualities taken to be important across the society as a whole, and so on.[123]

As this list suggests, the possible criteria for normative consonance are many and varied. There is certainly room for disagreement in their application. There is, for example, significant and, I suspect, ineradicable disagreement over the ends to which marriage is directed, precisely because marriage is a complex institution serving many ends and to which participants have attached a wide range of normative expectations. Judgments as to consonance can very well conflict, just as the judgments of people involved in practices conflict. Normative consonance does not produce, then, a single right answer, but rather a range of plausible arguments: different possible characterizations of marriage, different readings of the comparison between these characterizations and same-sex relationships, and so on. Participants advance these arguments by interpreting the practices and explaining their import. It then becomes necessary to decide among these arguments by some process of social determination, through which the norm to govern the society is determined.

The need to coordinate compels such a common solution, at least to some degree. But note that the process of advancing possible solutions and of deciding between them does not appeal purely and simply to the needs of coordination. Instead, it takes the form of a complex normative deliberation, starting from the established grammar of the society’s norms, reading that grammar against the various practices to which it relates, and criticizing and refining one’s understanding of the norms. Sometimes it takes the form of substantially revising society’s norms if their premises no longer seem consistent with the decision maker’s understanding of the normative order as a whole, if they appear to conflict with other cherished norms, or if the consequences of their past application now appear to have been deleterious. [124]

The ultimate result will likely not accord with any one individual’s view of what the normative order should be. It is likely to be a distillation of views, having sufficient support from members of society to secure their willing compliance but nevertheless abstracting from any individual member’s actual opinions and beliefs. The final step of coming to a societal solution inevitably involves a measure of final determination, of stipulation, from among a variety of alternatives held by members of the society. As we see in the case of religious belief, that distillation may tolerate areas where members can continue to pursue divergent practices, with no socially determined result imposed. In other areas, a single result will be established in the face of continued disagreement among members of the society.

Cass Sunstein has argued that in adjudication we should aim for “incompletely theorized agreements”—results that are framed only at the level of generality that is necessary for social cooperation, without demanding that people agree on the far-reaching normative theories that might support those outcomes.[125] Sunstein’s argument points in the right direction in its modest ambition and its focus on the need to determine a working result, not a grand theory, for our social order. But I would cast the process somewhat differently. First, Sunstein is wrong to suggest that public deliberation can be limited to the level of practical compromise without reaching back to underlying arguments of justification. Those underlying arguments must be engaged; they remain the active impetus for people’s normative commitments, even if wholehearted agreement is too high a standard for their resolution.

And this brings us to the second point: “agreement” is an impossibly and unnecessarily demanding standard. In any society, all members do not and cannot agree on social norms, even when those norms are defined with theoretical modesty, and even when a robustly deliberative process exists to narrow disagreement and ensure a large measure of popular support. Our disagreements are simply too deeply rooted. But members realize that, within broad limits, they have to put up with normative positions, normative interpretations, that they do not fully agree with if they are to live in society with people who have their own minds. The benefit of living in a society governed by some conception of justice is better than living in no society. The range of potentially acceptable norms is broad, given the interest members have in securing the benefits of social cooperation. That said, even within that range, the degree of consonance matters: the greater the degree of substantive support from members of the society, the more effective and stable the norms are likely to be, because members will both know and accept them more readily. Legitimacy is a relative concept, in which there can be greater or lesser degrees of acceptance. It is not on-off, either accorded or not.

In emphasizing the complexity of social practices and arguing that they are already infused with linguistic and normative content, content that inevitably shapes the process of fitting norms to practices, I do not mean to suggest that the linguistic forms are entirely self-standing, without any connection to material relations. It is of course true that human beings operate in a world that transcends their concepts, and that the world is, to a degree, refractory to concepts. Concepts are frameworks of understanding that order and shape how we conceive of the world. They can be destabilized by changes in that world; they can, over time, be considered less effective than alternative conceptions in explaining the world. And indeed, one sees in indigenous normative orders continual processes of change and adaptation of those cultures’ distinctive concepts in a manner not unlike those one sees in non-indigenous orders.[126]

It is precisely because our conceptions are always provisional, revisable, and even defeasible—always existing in complex relation to a world that extends beyond them—that it is possible to translate across linguistic and conceptual frameworks and to learn from others’ conceptual schemas. There are areas of overlap in experience, analogies among our attempts to create meaning, which can serve as starting points for mutual understanding. And of course, the extent of overlap can be increased by entering into the practices of another society, either through living within that society or through attending to its narratives and explanations. The relation of concepts to the world therefore provides potential reference points for communication and comparative evaluation.

The process of reasoning across legal languages is not easily accomplished. It may, for example, require immersion in the entire complex of environment, language, world view, deliberation, and practices of a society, in order to truly understand how concepts are deployed. The process of engaging in comparisons is certainly not reducible to evaluation on the basis of a neutral, norm-free, functionalist common denominator, as many functionalist approaches implicitly assume. Both the world and our schemas are sufficiently complex that there are likely to be strengths and weaknesses to any conceptual framework. Our very capacity to evaluate strengths and weaknesses is inevitably constrained by our vastly greater comfort with and mastery of the scheme in which we have come to form our own opinions.

But most importantly, we can only discuss these comparisons in language. Our theories never get outside language. We can do our best to translate across languages, but whatever we say is inevitably afflicted by the limitations—and the strengths, and the normative overtones—of the tools we use to say it. This fact again provides us with reasons not to dismiss an alien metaphysics out of hand but instead to remain open to insights that may, in the end, prove translatable. And even if they are never translatable, we may simply have insufficient justification, insufficient understanding, to impose our own view.


The conception of customary law set out here has important implications for the way in which one engages in comparative legal analysis and develops and assesses institutions that operate at the interface between legal cultures. A full exploration of those consequences will have to await future papers,[127] but their main lines should now be clear.

They include, for example, the need for humility and respect in the encounter, precisely because differences of law are tangled up with differences of normative language, which are in turn the product of extended experience and long reflection. Their meanings can be profoundly difficult to access, requiring immersion in the practices and concepts of the society. They can also be vulnerable to dislocation by practices and concepts imposed on a society in ignorance of the society’s own normative resources (even if that imposition is well-meaning). The risk is all the greater in situations of substantial asymmetries of power, when conceptual incomprehension is compounded by differences of material interest, as indigenous peoples have found to their severe detriment. Differences of legal culture therefore furnish important arguments for institutional autonomy and suggest an ethic that should temper interactions across the normative divide. Arguments for recognizing different traditions of customary law meld with arguments for self-government.