|Naturalism and Agency in the Living Law
There is a common tendency in some branches of the sociology of law – especially in the literature of legal pluralism – to see social norms as being deeply rooted in patterns of human interaction. There is some variation in how this relationship is described, but many approaches speak of norms emerging from social interaction in a manner that suggests 1) that any prolonged human interaction tends to generate norms; and 2) that this occurs spontaneously rather than through deliberate human action. Indeed, the temporal ordering I have suggested here – interaction occurring first, with norms then being generated from that relationship – may understate the connection posed by some theorists. Some treat norms as being virtually coeval with any peaceable human relationship.1 On this view, any ordered interaction implies the simultaneous existence of norms in the same way that the physical constitution of an organism presupposes its skeleton. There is a strong implication of necessity, then, in such visions of normative orders. A normative order is considered to be a natural dimension of any human interaction, generated through the day-to-day business of human life, perhaps even definitional of the existence of society.
The tendency to present norms as emerging naturally out of human interaction also typically presents a society’s normative order as non-contentious. A normative order consists of those rules that the members of society recognise and obey. The norms are not the result of argument and imposition; they are inscribed in members’ conduct. Implicitly, then, these approaches align with theories of social organisation based on consent. The parties may not have deliberately chosen the norms by which they are governed, but they have internalised those norms, living them. The socially-grounded law is portrayed as a unified and harmonious body of norms, highly adapted to a particular social milieu and exempt from disagreement and contention. It is not so much that disagreement is denied but that it has no point of entry into the theory. The law is given directly by social interaction, not by processes of human debate and decision-making. Processes that maintain the normative order are conceived as mechanisms for the enforcement of pre-existing rules, not as devices by which the content of those rules is itself determined. Not only does this distort the nature and role of those processes (as I discuss below), it also opens up a deep gulf between social law and the state’s law. The former is natural, consensual, communal, and conflict-free; the latter is artificial, coercive, peremptory, and conflict-ridden.
In this paper, I want to defend a view that all law – non-state law as well as state law – is inherently non-consensual, that it is always to some extent peremptory and imposed, establishing a collective position against a backdrop of deep-seated normative disagreement. All legal orders, of whatever kind, have to have mechanisms for fashioning these collective positions out of the welter of disagreement. This does not undermine the key insights of the theorists of the social law, but it does warn us against the tendency to treat the social law as natural, as emerging harmoniously from practice. Instead, it insists on the existence of contestation and dissent and focuses attention on the means by which contestation is settled. In so doing, it emphasises the sources of dynamism and change in the law and provides a foundation for the more nuanced comparison of, and examination of the interaction between, state and non-state legal orders.
I will explore these questions specifically in relation to Eugen Ehrlich’s Fundamental Principles of the Sociology of Law,2 examining the extent to which he falls prey to the naturalising fallacy. As will become clear, I believe that he falls into this danger in crucial parts of his argument – although in drawing attention to those failings I don’t mean to deny the great force of his accomplishment in emphasising that law is intimately entangled in social interaction more generally. On the contrary, I accept his foundational insistence that we focus our attention on how law is lived in day-to-day interaction and that we examine the close structural relationship between state law and other forms of social normativity. My argument is that we understand those dimensions best when we take normative disagreement seriously. Indeed, the very essence of law – and of normative orders generally – involves the fashioning of an emphatically social outcome in the face of disagreement.
Precisely because I share Ehrlich’s concern with the relationship between state law and other forms of normativity I will generally avoid the term ‘law’ when presenting my own views, referring to ‘norms’ or ‘normative order’ instead. For many legal theorists, especially those influenced by legal positivism, ‘law’ is necessarily and indissolubly tied to the state. The application of the term to norms produced by other mechanisms therefore prompts an immediate and dogmatic objection. I want to sidestep that stumbling-block. I do not deny that the involvement of state institutions makes a difference to our analysis of the legitimacy, salience and means of operation of norms. (Although it is often too simple to say that the difference flows merely from the fact that the norms are ‘state norms’: it is generally necessary to examine what kind of institution has produced them by what means – legislature? civil service? courts? specialised administrative agency? Their scope of application, means of enforcement and even legitimacy will vary depending on the precise way in which the state is implicated. Moreover, if there is considerable mobility between non-state and state norms – if, for example, state institutions draw their standards in part from the activities of trade associations or from largely self-regulating professions, or depend upon non-state means for their enforcement – the very task of establishing the line between state and non-state norms can be challenging). But even granted that it is important to distinguish between state-sanctioned and non-state-sanctioned norms for some purposes we still need to preserve a space for comparing different modes of social regulation and for examining the interconnections between them, as Ehrlich does to such illuminating effect. That objective is sometimes best achieved by treating state and non-state norms as species of a single genus. That is what I do in this paper. Indeed, my primary purpose is to establish that non-state norms share a fundamental characteristic with state norms: they all confront the fundamental problem of how to establish a common standard in the face of pervasive normative disagreement.
But is this just a straw man? Surely any scholar of law – jurist, sociologist or anthropologist – will acknowledge that in any real society opinions about norms vary and that any system of social regulation must possess ways to settle these disputes. The anthropological literature – on which legal pluralists rely – is replete with descriptions of authority, disputation and dispute resolution in non-state orders.
This is true, but nevertheless the temptation to naturalise non-state orders, portraying them as actuated by a unified set of beliefs, is a real one, well worth recognising and guarding against. Such a tendency is common, for example, in anthropological accounts that seek to describe the internal rationality of normative orders, especially accounts that try to establish the distinctness and integrity of those orders in comparison to other ways of organising social life. Often the author wants to defend a subaltern order against a dominant one, and this increases the temptation to minimise internal disagreement and emphasise cohesion. This can be true of hermeneutic approaches, which seek to explain social norms in terms of a comprehensive and internally consistent world-view. Clifford Geertz’s work, stimulating though it is, is vulnerable to this criticism. He acknowledges complexity especially in the existence of legal mixité (in one striking passage he notes the many legal traditions present within Ethiopia and argues for an ‘abnormal’ or ‘nonstandard’ discourse to deal with it).3 But it is difficult to see how this fits with his primary emphasis on the internal rationality of normative world-views. One suspects that, for Geertz, legal mixité primarily serves as a source of disorder and destabilisation in relation to world-views conceived in much more consistent and unified form. For Geertz, disputing mechanisms are primarily means to re-affirm, elaborate and defend the internal rationality of the order. They tend to take the latter as given, not as something that has to be made and remade through the agency of social actors, as this paper will insist.
Indeed, we can fall prey to the temptation to minimise conflict and to naturalise social orders whenever we try to describe those orders primarily in terms of their substantive content. This is, of course, a common approach in ethnography, where one sets out to describe the beliefs and concepts of a particular culture. It is also very common – indeed standard – in the general run of doctrinal literature in Law, which seeks to present a rationalised, systematised, coherent and comprehensive understanding of an area of the law.4 There is good reason to attempt accounts of content: broad differences certainly are apparent when one compares different normative orders; differential content determines much of the texture of day-to-day life in those societies; and in its broad lines, that content may have stability and resilience through time. But we should never lose sight of the fact that those portraits are simplified, tidied-up versions of the contexts they purport to describe. To adopt a phrase commonly used by lawyers, they represent the ‘best view’ of the law, in which the writer seeks to establish more coherence and consistency than exists in the phenomena themselves. In contrast, any real legal order is characterised by the co-existence of variant interpretations, variant conceptions of justice, variant theories of the whole. Any order must have mechanisms to address those contending positions and make authoritative determinations, at least provisionally. When those mechanisms are taken seriously, the specific content becomes less significant. Descriptions of content appear more like snapshots in time, instructive in themselves, but immediately subject to mutation and change. To fully understand a normative order, one has to understand the processes of change: what drives it; how it works with the past; how it is regulated; how provisional outcomes are determined and applied.5
There is, in short, a temptation in the practical or empirical literature to treat normative orders as more consistent than in fact they are, and to minimise the role of human agency in fashioning and refashioning that content. The temptation is also manifested in the theoretical literature. It is not uncommon to see consciously made orders contrasted to orders that emerge out of practice, with the latter treated as though they are the product of something other than conscious decisional processes. To take one recent example, Simon Roberts adopts the adjective ‘acephalous’ to describe non-state orders. He does so to distinguish orders based on centralized political control from orders that do not involve imposition by a governing class; he wants to confine ‘law’ to the former (although he does acknowledge that the latter are normative orders). But surely ‘acephalous’ connotes more than the mere dispersion of decision-making. It obscures human agency altogether in the establishment of order, suggesting that there is no directing intelligence or combination of intelligences, or if there is, such an intelligence works purely by consensual negotiation.6 In doing so, the gulf between highly centralised state-structured orders and non-state orders is exaggerated.
I suspect that one reason for this tendency is a desire to affirm the integrity of non-state orders against the state, because the state is seen as distant, unresponsive, repressive and (with respect to subaltern orders) colonialist. There are strong affinities between this literature and classical anarchism, as writers seek to affirm means of social organisation that are taken to be consensual, untainted by authority and imposition. I share the need to take non-state orders seriously. I acknowledge that state orders can do great violence to non-state mechanisms of social ordering. But we cannot achieve an adequate understanding by portraying non-state orders as more unified and conflict-free than they are, and by obscuring the mechanisms of decision within those orders.
Ehrlich’s theory strays into this naturalising fallacy, especially (as I will argue) in his fundamental distinction between ‘rules of decision’ and ‘rules of conduct’. And although a full assessment lies beyond the scope of this paper, one wonders whether Ehrlich was tempted by the attractions of anarchism. The period is right, and Ehrlich shares Proudhon’s and Kropotkin’s hostility to the state running roughshod over more local orders, an idealisation of local orders as more consensual and responsive than the imposed order of the state, and Proudhon’s particular concern with the role of localised associations and voluntary acts (especially various contractual forms) in creating normative communities.7 One would not want to conclude too much. Characteristically, Ehrlich hedges his bets; there are passages, for example, where Ehrlich appears to accept the inevitability of state institutions. But the parallels are thought-provoking.
Ehrlich’s work is subtle and rich, carrying great insight. It is not my purpose to pigeon-hole much less to dismiss him. This paper is premised on the value of engaging with Ehrlich’s approach, an engagement from which I have learned much. Indeed often Ehrlich appears to have been on the track of the insights for which I will be arguing, but those insights were inadequately integrated into his theory. My fundamental argument is that, when engaging with Ehrlich’s work, we get a better sense of the nature and dynamics of law, and of the presence of agency within its formulation, if we take the true extent of social disagreement seriously.
INTERPRETATION AND THE LIVING LAW
Let me begin the primary argument of this paper with a story. In the 1990s I was a member of a research group concerned with developing a conceptual framework for the sociology of law, working especially with the notion of ‘the emergence of norms.’8 Most of the members of the group approached the issues in the spirit of legal ethnographers, locating norms within a thick description of the circumstances in which the norms were formed. But one of our colleagues took a distinctly different approach. He too was concerned with the emergence of norms but he conceived of that emergence entirely in terms derived from the work of Friedrich Hayek, as the spontaneous ordering of the market.9 His interpretations did not fit at all well with those of the rest of us. We were attempting to provide rich cultural accounts of the emergence of the norms, highly attentive to the peculiarities of each context; he provided what seemed to us to be stripped-down and functionalist accounts, accounts which filtered all human action through the self-regarding rationalism of economic man. There were times when we wondered whether our colleague and we were engaged in the same type of intellectual endeavour.
There was indeed – and still is – a deep gulf between Hayek’s vision of human interaction and those I find persuasive. But I now think that there was more in common in the approaches taken by our Hayekian colleague and by the rest of us than I had initially realised. Every attempt to derive legal rules from the lived interaction of the members of a particular society is necessarily interpretive, necessarily simplifying, seising upon particular dimensions as especially significant. We make decisions as to which are the best among rival interpretations. We impose greater order upon the material than the practices alone generate. The Hayekian approach was particularly far-reaching in its abstraction, universalising in its assumptions about human behaviour, and thorough-going in its individualism, as it proceeded to reduce human interaction to the self-interested deployment of one’s property. But as an attempt at theorisation it was not profoundly different in spirit from our more culturalist approaches. Our approaches too took a complex and ambivalent experience and simplified it, producing a unified portrait of a normative order said to have emerged from interaction.
In doing so, we tended to occlude variant interpretations, disagreement, and the mechanisms by which that disagreement was settled in the various contexts in which we worked. We tended to reduce contending interpretations to a single vision of the normative order. We might have avoided such a highly simplified account. Instead of attempting to state the emergent law of a particular context as though it were a set of determinate rules engrained in human action, we might have set out to describe the legal culture of that context. We might have explored the diversity of normative arguments that had become entwined with the practices, identified the salient fault-lines within the society, described the relative strength of the various interpretations of the normative order, determined the relationship to practice that gave those interpretations their strength, and suggested how the social milieu’s normative debates tended to be settled.10 We would then have portrayed a more contentious milieu, albeit one that took distinctive shape and structure from its practices, traditions and historical interaction.
Now, I think there is a reason why we were induced to push beyond this picture of a diverse and somewhat untidy normative field to an account of what we took to be the milieu’s unified and coherent system of right. Those responsible for maintaining, interpreting and applying a normative order do aspire to such unity and coherence; they do attempt to fashion a consistent normative order for the social field with which they are concerned. Our mistake lay in thinking that this was accomplished fact, and in purporting to describe the intended normative order as though it were already ingrained in interaction. We failed to realise that any normative order always involves the continual creation and re-creation of that order in the face of disagreement. We short-circuited that process, treating the order as though it were already achieved when it was very much under construction, indeed perennially so. Moreover, when describing the nature of that order, we would often pick what appeared to us to be the best interpretation of the order’s norms, in a manner that seriously confused our role as observers with the role of those participants charged with providing authoritative interpretations of that normative order.11
Normative orders are, in a sense, always open, always subject to further interpretation, refinement, elaboration and extension. This is true even of the social law – in Ehrlich’s terms the ‘living law’. The normative meaning of practices is always subject to ambiguity and disagreement. The practices may well furnish grounds for evaluating rival interpretations, providing a basis for arguing why one provides a better understanding of the practices’ normative significance than another. In that sense, norms can be fashioned and then evaluated in relation to practice. That is the great truth in theories of social law, providing the foundation for our intuition that practices and norms evolve together and allowing us to argue intelligibly about the principles that might best order our societies. But given the complexity of practices in any actual social field, and given the variety of considerations that are used to shape and interpret them, the practices themselves are never so univocal that they simply generate norms. At most they support a range of normative considerations, which then have to be narrowed and adjudicated (in the broad sense of adjudication described below) to determine a single rule for society.
Every normative order aspires to have such single rules, at least to govern some matters. If people want to live in an ordered society, in which at least some matters are resolved by means other than brute force, there have to be ways of coming to a resolution despite the persistence of normative disagreement. Those mechanisms may well rely upon – indeed they should rely upon – knowledge of social practices so that the resolution fits well with the pattern of normative traditions present within the society, and so that the ultimate resolution complements those practices (or, if the intention is to change the practices, so that one knows accurately what one is changing). But that knowledge will only rarely result in a single unambiguous conclusion. Generally there will be a range of conflicting interpretations which then have to be adjudicated. That adjudication can occur through a wide variety of means: legislation; adjudication by a judge or panel of judges; agreement; or normative assertion, compromise and acquiescence. One can even, in some circumstances, agree to disagree. One can recognise continuing spheres of normative autonomy. But in all these cases, mechanisms exist for determining what the societal resolution to the divergence of interpretation will be. This resolution will always be artificial, in the sense that it will always establish a single or at least greatly simplified outcome in the face of disagreement. It will always be, in this sense, peremptory. Moreover, the underlying disagreement will generally continue. It may resurface and prevail in some future controversy.
In fact, it is not too much to say that this last, decisional stage of settling norms is fundamental to our idea of law as an emphatically social order, which governs us regardless of our individual agreement or disagreement. Law necessarily imposes a socially determined rule against a more complex and ambiguous background of normative assertion and counter-assertion.
An example may help clarify what I mean. To establish that the argument applies in the case of non-state norms, I have chosen a clear case of a socially-generated rather than governmentally-decreed norm. In Vancouver, British Columbia, it is generally accepted that when a traffic light turns amber, not one but two cars that have been waiting to turn left across a line of on-coming traffic are permitted to make their turns, once the line of on-coming traffic has come to a stop. This seems such an invariant practice as to amount to a rule of customary law. Drivers accustomed to Vancouver almost universally understand this rule: those turning left feel justifiably aggrieved when on-coming drivers come through the intersection on a late amber, encroaching on the left-turners’ time to turn; drivers waiting to turn left behind the first two cars are annoyed if one of the first two cars doesn’t turn, thereby delaying the rest of the line; and drivers on the cross street will patiently wait for the left-turners to complete their turns before they enter the intersection.
Now, one can see how such a rule would come to exist. It is clearly structured by particular aspects of the situation. All have experienced the frustration of waiting to turn left when faced with a continuous line of cars coming in the other direction. All drivers can understand that left-turners must be given an opportunity to turn or a long line of cars will develop at busy intersections, all trying to turn left. The changing of the lights appears to be the best moment to allow them this opportunity. Moreover, if the on-coming traffic stops promptly when the light turns amber, there is often enough time for two cars to turn before cars on the cross street can get started. Finally, at many large and busy intersections, if on a green light the left-turners advance to make their turns there will be two cars in the intersection.
But it is also clear that the solution adopted in Vancouver is not a foregone conclusion. (In fact, in my city, Victoria, there appears to be no such custom, even though we are just across the Strait of Georgia from Vancouver, in the same province, with many drivers who would have driven extensively in Vancouver.) A strict rule against a car entering an intersection to turn left before the route was clear might have been enforced. The practical challenge of turning left could have been solved by greater reliance, by the city, on advance-arrows and left-turn lanes. The Vancouver solution would appear to depend upon the wider existence of norms of courteous behaviour; it may be breaking down as traffic volumes increase and drivers become more aggressive, running through intersections on late amber lights or surging ahead when their light turns green.
How, then, did the customary rule come to be established? As with many social rules it is impossible to know, but I suspect it happened something like this: Faced with heavy traffic and several cars waiting to turn left, drivers of the second left-turning car began to take the opportunity to follow through with their turn after the light turned amber. Other drivers, recognising the value of more than one car clearing the intersection, acquiesced, and did not press into the intersection when their light turned green. Moreover, when they too were in a left-turning situation, they began to adopt the practice. The police, realising the benefit to the flow of traffic, tolerated the practice. The custom became generalised. Note two things about this: First, there is a time before the establishment of the norm when its existence has to be settled. Here it is settled by assertion and acquiescence, followed by its reproduction in other contexts. Second, note that it is not mechanically produced by the practice. A process of reasoning and justification enters in. The acquiescence flows from the parties’ recognition that the practice carries a number of benefits. There is an element of implicit deliberation. This reinforces a point made by Lon Fuller: customary law is not merely the sedimentation of habit; it is always a reasoned solution.12 In the case of this simple norm, where all drivers from time to time find themselves in the same predicament, all understand the considerations in issue, and the range of potential solutions is limited, assertion and acquiescence is sufficient to settle the norm. In more complex situations a more elaborate mechanism would generally be required: in the case of non-state norms, determination by an elder or other authority figure, compromise among the members of a family, recommendation by a trade association, debate leading to a majority opinion backed by the threat of ostracism and denunciation, and so on.
There are, then, two dimensions to all law that is grounded in social practices: 1) an interpretive dimension, in which participants propose and deliberate about rules, justifying their solutions on the basis of the exigencies of the situation, the lessons of experience, and broader attitudes already established within society; 2) a decisional dimension, in which a collective resolution is established from among the proposed interpretations. The former involves arguments of justification on the basis of practice and tradition, and forms the foundation for the legal pragmatist’s or legal anthropologist’s assumption that norms are derived from practice and culture. But practice and culture do not speak with a single voice, and there therefore have to be mechanisms for settling upon a specific outcome. That gives rise to the second dimension, which forms the foundation for legal theories that emphasise jurisdiction, authority and command. It furnishes the kernel of truth in positivistic accounts of law.
Virtually all normative orders involve both dimensions, blended in different proportions sometimes in different institutions. Even structures that seem to depend overwhelmingly upon jurisdiction and command, such as legislatures, undoubtedly rely on interpretive methods in their deliberations. In addition, the success of legislative norms depends on their appropriateness to social life and not simply on their coercive enforcement (as Ehrlich argued at many places in Fundamental Principles of the Sociology of Law) – although different institutions may perform such interpretive roles better or worse depending on their particular make-up, and a complex society may well adjust institutional role to institutional form. Institutions may also be subject to criticism on the basis of the scope of experience to which they refer when crafting their rules. They may look exclusively to the practices of a dominant cultural group, for example, neglecting those of a powerless minority. This may be one reason for adjusting the jurisdiction of institutions to conform to deep and pervasive cultural divisions (as in the case of indigenous self-government). But all law, both state and non-state, is articulated against a backdrop of experience, with mechanisms for bringing normative debate to a close. All law is both interpretive and, to a degree, peremptory.
NATURALISM AND AGENCY IN EHRLICH’S SOCIOLOGY OF LAW
How are these dual dimensions of law reflected in Ehrlich’s sociology of law?
Although Ehrlich acknowledges their presence in what he calls juristic law (statements of law that result from the synthesising and systematising efforts of scholars), he tends not to treat other forms of law as embodying both dimensions. He treats the ‘living law’ (the rules of conduct which he sees as being embodied in social life) as though it were simply latent in social practices. The decisional dimension is virtually absent in his account of living law and even the interpretive dimension gets short shrift: the only interpretive role that he directly acknowledges is that of the legal scientist, who observes the legal order and attempts to describe its character as a matter of fact. He pays very little attention to