Using Systems Theory to Study Legal Pluralism: What Could Be Gained?

Abstract

This article examines the ability of modern systems theory to provide a foundation for understanding the problematic notion of legal pluralism, and to the ability of scholars to apply that understanding to engage in the study of pluralistic legal orders. In particular, it develops the observations of systems theory of the relationship between state law and violence by adopting one of its linked ideas, that of structural coupling. It also considers the role played by translation when law is identified by reference to the application of the legal code: legal/illegal. The whole analysis is underpinned by systems theory's account of the differences between studying premodern and modern societies.

This article examines the ability of modern systems theory (autopoietic or neo-systems theory) to provide a foundation for understanding the problematic notion of legal pluralism. But first we need to indicate why legal pluralism is problematic for sociolegal studies in particular and jurisprudence in general.

Legal pluralism seeks to extend the study of law beyond state and interstate legal orders to include non-state-sourced forms of law.¹ In so doing, it raises a specter such that law ceases to be identifiable as a separate social formation, as the border between the legal and the social is dissolved. Phrases such as “law from below” (Merry et al. 2010) or “an oppositional postmodern understanding of law” (Santos 2002) or “law between the global and the local” (Goodale 2007) or “legal hybridization” (Santos 2006) tend to reverse the hierarchical assumptions implicit in much legal practice and scholarship, and phrases such as “the more the merrier” (Melissaris 2004), or those applying the distinction between “law as one” and “law as many” (Davies 2005), imply that law can only be captured in a combination of conceptions rather than in a single conception, and in a range of linguistic forms rather than in any single form (Chiba 1998). A concept of law unencumbered by its associations with the nation-state and the activities of legal officials can easily make it impossible to understand—and thereby study—law as a separate formation (as occurs when law is identified with multiple social norms, general accounts of social control, strongly held commitments, cultural conceptions, and so on). Indeed, once we move beyond the study of law as state law, what prevents us from identifying different criteria for what constitutes law, according to our own observational standpoints and research objectives? The problem that then arises, and that appears to be a feature of legal pluralism–motivated studies of legal orders, is that these studies tend not to accumulate into anything like a coherent and integrated body of knowledge.²

If, as we hope we have briefly demonstrated, legal pluralism raises difficult issues for sociolegal studies, there seems to be good reason to attempt to find alternative approaches that facilitate better understanding and offer a better foundation for research. And it is in this light that we propose systems theory. However, we undertake this task with the knowledge that the general features of the theory have generated considerable skepticism toward its potential to inform something as focused on the local and specific as legal pluralism. As a theory of society, systems theory has a level of abstraction that seems far removed from the study of concrete legal orders, let alone the empirical study of particular aspects of those legal orders. Because the theory identifies society with its communications, the global nature of modern communications requires it to be a theory of world society, rather than one limited to a spatially bound community.³ And further, by identifying society with communications, the theory uncomfortably places the biological and thinking human being outside of society. Its hermeneutics are rooted not in the intentions of human actors, but in the meanings generated by those actors through their participation as communicators within subsystems of communication such as law, the economy, science, politics, and education.

There is also the troubling issue of which version of the theory should be utilized. The works of its creator,⁴ Niklas Luhmann, span 40 years and include some 70 books and over 400 scholarly articles. Over his lifetime he reacted to criticism, engaged in considerable self-reflection, and made numerous adjustments to and restatements of the theory.⁵ Luhmann's writing style is dense and often enigmatic, with frequent resort to metaphors. In developing the theory, Luhmann drew upon general systems theory, evolutionary theory,⁶ cybernetics, and semiotics. And if all this were not enough, the theory shares some of the features of Marxism, in that alongside the works of Luhmann, one also has the work of various “Luhmannians” who take a variety of approaches. These encompass “strict literalists,” who insist that a correct interpretation of the theory requires fidelity to Luhmann's own works; “liberals,” who reinterpret and elaborate on his basic concepts; and “pragmatists,” who take some of his concepts and seek to apply them to their own subject areas.⁷ In light of all this, it is hardly surprising that there has been plenty of resistance, particularly in the Anglo-American academy, to the use of systems theory in sociolegal studies. To those coming to the theory for the first time (and not a few trying for the second or third), the considerable effort required to become familiar with the theory, with its network of linked understandings, seems unlikely to justify the benefits, in terms of the new insights that might result.

Taking the particular example of legal pluralism, we hope to persuade readers that the effort is indeed worthwhile despite these difficulties. We would position ourselves with the “pragmatists,” among whom we also would place Gunther Teubner, though he would also be considered one of the most important “liberals” who have both developed and criticized Luhmann's writings.⁸ Teubner is the author of two articles (Teubner 1992a, 1997) arguing that systems theory could be used to study legal pluralism. He claims that systems theory can provide a basis for studying what is legal within society without either conflating all law with the official law of the state or losing the ability to separate what is legal from the rest of society. We wish to explore his claim, but to do so in a manner that can accommodate a readership that is not already familiar with the theory. To do this we need an accessible starting point for our discussion. We have therefore chosen to begin by looking at a short section from Brian Tamanaha's 2001 book A General Jurisprudence of Law and Society. In this section, Tamanaha, a leading legal pluralism scholar, addresses the same task we undertake here: what are the possible benefits of using systems theory to study plural legal orders? Tamanaha found systems theory unsuitable and inappropriate. By responding to his arguments, we hope to offer this readership an accessible introduction to the parts of systems theory that are particularly relevant to legal pluralism. We stress that our purpose in utilizing Tamanaha's work on legal pluralism is not to engage in a criticism of his own preferred approach. It is rather to use his rejection of systems theory as a starting point from which to introduce the reader to the theory, and to explore what it could mean to study legal pluralism through the lens of systems theory.⁹

The first section of this article consists of Tamanaha's own presentation of systems theory,¹⁰ the reasons why he found it unsuitable, and our responses to his critique. Following on from this we have selected a number of issues that illustrate how systems theory might be deployed to engage in legal pluralism study, but recognize that in the time and space available we can address only a few of the many issues that could be considered. Thus, in the second section we focus on how systems theory treats the relationship between state law and violence (the link between law and the state being, as we have illustrated, a key concern for legal pluralist study); in the third section we consider how systems theory approaches the issue of translation (particularly since legal pluralism confronts very different legal orders operating with their own terminology); and in the final section we examine how systems theory constructs the differences between modern and premodern societies. Each of these issues is of significance to legal pluralism, and each, we suggest, can be illuminated by systems theory observations.

Tamanaha's Criticisms of Systems Theory

In his book A General Jurisprudence of Law and Society, Tamanaha addresses the problems that underlie all attempts to study legal pluralism (as we have similarly attempted in our introduction). How can one develop a sociolegal positivist concept of law unencumbered by its associations with the nation-state and the activities of legal officials, without losing the ability to study law as a separate formation? Tamanaha's solution to these problems is a conventionalist/hermeneutic approach: to identify as law whatever a significant number of participants within a particular community refer to when they apply the term law.¹¹ Like systems theory, this is a nonessentialist approach to the subject, and it assumes that the legal system involves phenomena that establish their existence independently of the observer who studies them. In the course of presenting this solution, Tamanaha engages with Teubner's suggestion that systems theory's identification of the legal system as one of society's subsystems of communication, which reproduces itself through the application of the legal code (legal/illegal), might also offer a useful nonessentialist basis for the study of legal phenomena. This engagement takes the form of setting out a hypothetical conversation (2001: 189) and considering how one would use systems theory to analyze it. The conversation addresses a situation that might arise in relation to state law. We reproduce that conversation both to begin to explain what is pluralistic about systems theory and to assess the strength of Tamanaha's objections to it:

• Smith. “The value of NEWCORP's stock will increase by at least 50 percent, and perhaps by 100 percent, when this takeover bid is made public tomorrow. If we buy now we will make millions.”

• Jones. “You're right, we could easily double our assets. But it's illegal, and we might get caught. We could go to prison.”

• Smith. “Sure it's illegal, but the risk of being prosecuted is small. We'll be rich if we do it, so it's worth taking the chance.”

• Jones. “Okay, we probably won't go to prison, but it's still illegal, and furthermore it's immoral. It's wrong to break the law, and even if it weren't illegal it would be wrong and unfair to everyone else to use this information. Crime and immorality never pay.”

According to systems theory, modern society contains separate subsystems of communication: the economic system, the political system, mass media, science, the education system, the legal system, and so on. Each system is autopoietic: it forms its elements (communications) from itself (those of its communications that are available). Each subsystem's communications apply a code unique to that subsystem. In the case of law, the code applied is legal/illegal; that of the economy is payment/nonpayment; that of the mass media is information/noninformation; and so on. Applying this kind of analysis, Tamanaha assumes that systems theory would treat this conversation as part of the legal system, or rather, that it would treat parts of this conversation as part of the legal system: those communications that are creating legal meanings through the application of the code legal/illegal. He feels able to diagnose the other systems that the speakers above are drawing upon (or participating in). The opening sentence forms part of economic communication. Smith anticipates that payment for shares now will lead to considerable profit. Jones identifies the action as illegal, with possible consequences. At the end of the sequence, Jones, drawing on communications that are neither legal nor economic but recognizably moral, makes claims about the morality of these actions.

Having used his understanding of systems theory to analyze this exchange, Tamanaha questions the benefits of such analysis. First, he believes that systems theory complicates a simple situation: “What appears to be a rather simple conversation is extraordinarily complex when analyzed from an autopoietic standpoint” (2001: 190). Second, one may have difficulty identifying exactly which code is being applied (is Smith's second statement legal, economic, or both?), which leads one to “wonder what legal pluralists would gain by travelling down this path” (2001: 191). Third, Tamanaha believes that the theory is overinclusive. By identifying the legal system with the application of the code legal/illegal, systems theory ends up including communications “which most people would not consider ‘law'” (2001: 191). And lastly, he feels that systems theory is underinclusive, at least in relation to state law, as its focus on communication “eliminates raw physical violence from within the law—thereafter it may at most be considered an effect or consequence of law as communication, or a part of law's environment” (2001: 191). Tamanaha argues that, at least as regards state law, systems theory's approach is deficient if it fails to include “an analytical apparatus which would include the material power of law as central to its existence while excluding …” (in a reference back to his previous objection) “such marginal phenomena as the private conversation between two individuals contemplating a criminal course of action” (2001: 191).

Starting with Tamanaha's first objection, is this really a simple conversation? The only simple factor is the number of people involved: two.¹² But the complexity of meanings involved is significant. How can someone make a meaningful claim that shares bought today will double in value by tomorrow, except by drawing upon a network of communications concerning the stock market and share prices (the economic system)? How does he assert that using his knowledge of those economic communications (including communications about facts relevant to those communications) will expose them to a potential prison sentence for insider dealing without drawing upon a network of legal communications? And how is the claim of unfairness made without reference to moral ideas of freeloading? This hypothetical conversation is anything but simple, and its complexity reflects the complexity of modern society, which requires individuals to participate in many systems of communication.

Within systems theory, a conversation of this kind is itself a system: an interaction. Whether it also has an existence within a functional social subsystem such as law or the economy is not something that can be determined with any finality by an outside observer who observes no more than this conversation. The role of functional social subsystems in this situation is complex, but we can make it simpler by asking ourselves a question: how can this conversation have an existence beyond itself—i.e., after the conversation ends, and including more persons than Jones and Smith? One possible and, indeed, likely answer is that it may not. It may exist only as an interaction. But an alternative possibility is that it does exist separately within functional social subsystems. But we have to wait and see whether this is the case—whether the legal system recognizes this conversation as a conspiracy to commit a crime, whether the economic system identifies it as an agreement to make a purchase, and so on. The fact that this conversation between Jones and Smith may not in fact belong to any social system does not mean that systems are irrelevant to the meanings generated within it. Familiarity with social subsystems, and the manner in which they ascribe meanings to communications, allows the participants in such interactions to include far more complex and structured meaningful communications than would otherwise be possible.

This is what Tamanaha is actually discussing when he considers the difficulties (and successes) of allocating particular sentences within this conversation to particular subsystems. Consider his claim that this conversation might involve economic, legal, and moral systems. Drawing on those systems to construct his hypothetical, he knew that we too, as his readers, would recognize them as communications raising legal, moral, and economic issues. The distinction between an interaction and a social subsystem gives us our response to his claim that some of the sentences are ambiguous as to whether they belong to the legal system, the economic system, or both. None of these communications will belong to a social subsystem unless that system recognizes them as belonging to itself, in which case they will have a dual existence: within the interaction, and again, within the social system. The best that can be achieved by an observer of this conversation who is seeking to understand the meanings within it is to observe which social systems the participants are alluding to, or seeking to replicate, within their conversation. This is systems theory's version of what is often referred to as a hermeneutic approach. One can do this with some success, and some ambiguities, exactly as Tamanaha does in his discussion. The point here is not that systems theory provides us with a means to break down every “private conversation” (an interaction that has no further existence) into its constituent parts. Rather, it is that systems theory allows us to observe the nature and existence of the systems that allow conversations within a modern world to consist of such complex meanings as does this hypothetical one.

An observer of this interaction can observe how social systems are replicated within conversations, including aspects of their functional differentiation, and, we would suggest, how such conversations utilize the distinction between conditional program and code that, at the level of social systems, maintains their autonomy. The code of a social subsystem is its binary distinction, which has no inherent meaning. In the case of the legal code, what is legal is not illegal, and vice versa. What generates meaning are communications that take the form of observations on the application of the code.¹³ These are communications such as “Situation X is legal because of A,” “Situation Y is illegal because of B,” and so on. Through such communications about the application of their codes, systems build up ever more complicated communications about when each side of the code has been applied in the past and when each side will be applied in the future. These communications are a system's conditional programs (establishing conditions for the application of one or the other side of the code).

Applying this distinction between conditional programs and codes, we might ask ourselves whether the reference to legality in Smith's second line (“Sure it's illegal, but the risk of being prosecuted is small. We'll be rich if we do it, so it's worth taking the chance”) operates as code or program. Law is not the only system that refers to legality or illegality, but it is the only system that codes in terms of legal/illegal. If someone regards the risk of his conduct being found illegal as a cost against which to weigh the potential financial benefits, is this the replication of an economic or legal communication? We can answer this question by considering how the reference to illegality operates in the sentence. Does the acknowledged value of the illegal behavior have any relevance to its illegality? By contrast, does the risk of going to prison for illegal behavior have any relevance to the financial benefits? It is clearly one way—illegality is relevant as a cost to the financial benefit of insider dealing when deciding whether or not to buy the shares. The size of the financial benefit (Smith and Jones can double the value of the shares) does not alter the legality of what is being suggested. Having accepted the illegality of the proposed actions, Smith is replicating economic communications in which illegality is a factor relevant to the making of payments. Thus, the fact that the legal system may not include this conversation does not mean that the legal system is irrelevant here. It is the presence of the legal system, and the economy, that structures the possibilities for expectations within this conversation. To make this point another way, if there were not preexisting differentiated systems that transcended particular interactions, we could not recognize that this conversation could have separate economic, legal, and moral elements.

This is not a claim that communications within any interaction could always be clearly identified by an observer as replications of particular social system communications. Indeed, the converse will more likely be the case, especially in interactions that function as part of what Habermas termed “the lifeworld,” or general social communication. To understand both the benefits and the limitations of using systems theory here we need to consider how social systems might operate to restrict and thereby stabilize¹⁴ (rather than determine) the meanings generated within interactions. What Tamanaha introduces here as the “legal” elements of a “simple private conversation” is the application of state legal norms by individuals to their own contemplated behavior.¹⁵ What stabilizes these meanings are the actual communications of the legal system, which occur when the system recognizes communications as belonging to itself. In this conversation Jones and Smith are involved in a hypothetical application of the state norms of criminal law to their own contemplated behavior. What stabilizes this part of the conversation is the fact that the state legal system can recognize conversations such as this as evidence of an intention to commit a criminal conspiracy, and on rare occasions, it actually does so.

To summarize at this point, Tamanaha's position is in truth much closer to systems theory than he might suppose. Both approaches might exclude this conversation from the ambit of the legal system. Tamanaha would not include this conversation as “law” because the actors would not apply this title to their conversation. Systems theory would not include interactions as such within the legal system, though it would recognize that some interactions’ communications could also exist as social system communications. Tamanaha claims that the ability of such “private” communications to become part of the legal system lies with the individuals themselves. If they, or at least a significant number of them, describe these conversations as “law,” then this is their status. Systems theory proceeds on the basis that the process of inclusion within a functional social subsystem is not established through consensus (the number of individuals who express a similar view) but through the operations of that system. The enormous numbers of systematically connected communications that circulate within the legal system of a modern society and create ever more complex meanings¹⁶ do not represent a consensus of individual opinions. And these connected communications severely limit the ability of any individuals to declare for themselves what should be described as “law.” Within modern society, such attempts by individuals to claim to define something as “law” have to operate alongside a system that describes itself as “law,” in contradistinction to morals, or the economy, or politics, and maintains the separation of itself from those other systems by the application of a different code: legal/illegal. The communications of these individuals also have to operate alongside other functional systems, such as the economy, politics, and the mass media, which generate their own internal versions of the legal system. The communications about law within the economy, politics, and the mass media will not correspond in a number of ways to the legal system's own communications. Nevertheless, the object of these other systems’ communications will be the legal system, rather than whatever individuals, even in significant numbers, might apply the label of “law” to.

If both Tamanaha's approach and that of systems theory decline to recognize private conversations as part of the legal system, does this mean neither approach can inform the study of lay communications, and what does this mean for legal pluralism? Do these approaches limit law to the communications of formal institutions: courts and “officials”?¹⁷ And, if so, does this mean that state law is the only thing that can be observed as law? Tamanaha's approach would exclude lay views about what the law requires but include lay views about what constitutes law. So if laypeople were to consider law as something that came from the legislature, courts, and officials, what issues from those sources would become “law” for these people, even if they did not have any clear understanding of what such “law” required. Similar results follow with lay understandings of religious law, or customary law. When people interpret human conduct by reference to what they understand to be the law—whether state, religious or customary law—they are unlikely to regard their own expressed opinions as themselves “law.” The Torah, the Koran, or the Bible, along with the words of people who have authority to interpret such religious texts, are likely to be understood as the “real law.” This is a consequence of Tamanaha's conventionalist approach to the identification of law. He believes that practical advantages follow from this approach, and has more recently claimed that extending legal pluralism to include “day to day human encounters” leads to the conclusion that “every form of norm governed social interaction is law,” so that we end up “swimming, or drowning, in legal pluralism” (Tamanaha 2008: 393).

However, we would argue that the approach of systems theory, which concentrates on coding, has more potential to extend the study of what is legal beyond a focus on formal sources than does an approach that identifies as “law” only what a significant number of participants, if questioned, would describe as “law.”¹⁸ It takes the study of law to what, using a biological metaphor, might be called its capillary level. Luhmann used a different metaphor, a distinction between the centers and peripheries of systems. In the case of law, the center is the courts, with their peculiar responsibility of deciding what the law is, even when there are no adequate reasons for reaching a particular answer. This responsibility, a “prohibition on the denial of justice” (Luhmann 2004: 284–296), has led courts to develop doctrines with a level of systematic complexity that is beyond the capacity of most laypeople to utilize. Luhmann locates lay legal communications, whether with other laypeople or with legal professionals, at the periphery. This spatial metaphor does not represent a hierarchy. All legal communications are just that—legal communications. Those within courts are no more legal than those without.

Legal meaning occurs whenever the system is applying the code legal/illegal, implicitly or explicitly, to construct itself and its environment. It constructs both itself and its environment through its operations, and lay communications form part of this process. While a discussion of hypothetical legal operations may remain at the level of an interaction, lay communications, no less than official ones, can perform legal operations. For example, laypeople who change their legal status through marriage, or alter the distribution of their wealth through gifts and contracts, utilize communications that transcend any accompanying interactions (conversations). Legal operations are not limited to what courts or officials regard as a valid application of rules, though it is difficult to envisage a legal operation in modern society that has no relation to the rules and norms recognized by courts, tribunals, or some other such institution. So, for example, a marriage may be conducted according to norms recognized by courts (the registrar qualified to carry out the ceremony). However, the parties may be too closely related to be eligible for a valid marriage. While the marriage may later be declared invalid, one cannot say that the ceremony failed to perform a legal operation. From the date of the ceremony until the flaws are recognized, the legal status of the parties changed for all sorts of purposes both inside and outside of the legal system. For Tamanaha, the inclusion or exclusion of the above examples within the legal system depends on whether the laypeople recognize themselves as legal actors. Laypeople are likely to be sure that the personnel who staff institutions are legal actors, but less sure of their own status, even when they participate in such institutions (for example, as defendants or contracting parties). Applying systems theory, the inclusion of these communications within the legal system (as communications that constitute legal operations) depends neither on the participants’ themselves recognizing the nature of their own communications, nor on their own status as system agents.

Legal communications at the periphery are likely to make reference to those at the center. Even without the involvement of legal professionals, laypeople within modern societies understand that law emanates from legislatures, courts, and town halls, and they recognize certain formal documents, notices, and street signs as communications that draw their authority from these sources to declare what is legal or illegal. But likewise, the center does not ignore legal communications at the periphery. The work of legal professionals to constantly reinterpret lay demands as contracts, trusts, debts, and so on involves communications that repeatedly generate a flow of communications back to the courts. And this flow is not limited to what is processed by legal professionals. The legal system does not exist solely as communications about laypeople by professionals and the courts. For example, while a discussion about the illegality of a particular action may remain at the level of an interaction, it may also lead to a person's making a claim for compensation, reporting a crime, or campaigning for a change in the law. Without these lay communications, which attribute legal significance to events and provide the basis for further communications about those events, we would have a much reduced legal system. And the ability of such communications to form part of the legal system does not depend on laypeople's accurately anticipating how a legal professional would articulate the matter in question. Or, to put this in the terms of systems theory, while the conditional programs that operate at the periphery are likely to be influenced by those at the center (as with lay attribution of legal norms to courts and statutes), they do not duplicate them.

Thus systems theory provides a method to observe lay communications in two ways. There are those lay communications that form part of legal operations and are part of the legal system. And there are those that are not part of that system, but are stabilized and structured through participants’ familiarity with ones that are, as well as their experience of communications about the legal system within other systems (most notably the mass media). Inside the legal system, the possibilities for legal meanings are further stabilized by the manner in which communications at the periphery link to communications at the center. In law, the center is the courts. Other social systems also demonstrate this center-periphery relationship: in the economy the looser meanings as to what might constitute a payment within the periphery are stabilized when they become linked to communications within institutions such as banks (as when money is deposited in accounts).

Systems theory addresses not only interactions and social systems, but also communications that occur inside institutions. For example, it is a mistake to conclude that because a court is a legal institution, and a judge has a legal role, all communications that issue from courts and judges belong to the legal system. A recent development in the UK court system has been the increased use of press releases.¹⁹ The fact that these are issued by Her Majesty's Courts Service, or that they repeat statements made by judges in open court, does not make these communications legal. They are offered as media communications, with the intention that the media will take them up and use them in the process of constructing stories about the legal system and other topics. Thus, not only “private conversations,” but also the communications of institutions, are often complex. Both are sites in which different systems are drawn upon in order to stabilize meanings. Similar analysis applies to judges who give public lectures or even (though this is subtler) issue homilies when sentencing criminals. While some part of a judge's remarks may enter social systems, via media reporting or appeals, much of what a judge says on such occasions remains at the interaction level.

Violence, State Law, and Systems Theory

With regard to Tamanaha's objection that a systems-theory approach would not sufficiently address the role played by violence within state law, this claim is not about the use of the code to identify what is legal, but about the ability of systems theory to build from this nonessentialist starting point. Such building is particularly relevant to legal pluralism study that needs to locate state law within the social orders and other “laws” that it examines. Tamanaha seems to believe that, even if the use of the code legal/illegal could provide a suitable nonessentialist starting point for the study of law, the rest of systems theory would block attempts to develop adequate accounts of particular forms of law. This is especially so with reference to state law, as systems theory will be unable to give recognition to the material power and effects of law. He believes that the understanding of law as a system of communication eliminates raw physical violence from consideration.

We can begin to come to grips with this claim if we remind ourselves that systems theory not only studies law as communication, but also regards society as communication. Does the theory therefore have nothing to say about the nature and forms of physical violence within society? It is true that raw physical violence—the kind that damages human beings in the same manner whether it comes from hurricanes, car crashes, or executions—lies outside of communication. But communication is what gives social meaning to violence and divides it into what is considered natural rather than intentional, as well as what is legal, political, or even athletic. As such, just about everything that we mean when we refer to violence lies within the ambit of systems theory. With particular regard to state legal systems, it is clearly important to understand the links between the kinds of communications issued by legislators and judges, and the pain that wardens inflict on prisoners. But again, this can be understood in terms of communications. As Rob Cover points out in his seminal essay on the need to take account of the violence that accompanies legal communications, it is only the fact that prison warders are willing to react to two rather similar forms of judicial communication (an order to execute and a stay of execution) in quite different ways that makes it fruitful for opposing counsel on death penalty appeal cases to conduct their extensive arguments about the legality of the imposed sentence (Cover 1986: 1623). Applying systems theory, reference to the termination of the life of a prisoner, or not, in response to one of these two communications is an “effect” of communication. The chemical, physical, and psychic processes involved in one person's taking the life of another are not themselves communications. But the institutional practices that have resulted in the predictability of such responses are constructed through communications and, as such, are open to observation through the application of systems theory analysis. And what this involves is an empirical question. If the link between law and violence is built in to our definition of state law—as occurs, for example, if we define state law as a monopoly of violence, or the infliction of evil in response to disobedience to a sovereign—we impede our observation of the manner in which power that leads to penal violence is distributed through communications (particularly political and legal ones) and the occasions when communications are absent.

The ability of communication to result in the infliction of physical pain is important to our understanding of not only the operation of state law, but also the relationship between state and nonstate legal orders—that is, pluralism in general. Cover has observed that it is not the relative integrity of their respective interpretations that leads to state laws’ domination over rival legal orders, but the greater likelihood that resistance to state legal norms will result in the infliction of physical suffering. The relative ability to avoid sanctions being applied to oneself, or to have sanctions applied to those with whom one has a dispute or grievance, provide motivations for communication.²⁰ As such, the manner in which state law distributes and utilizes the ability to inflict physical and mental suffering affects the possibilities open to rival systems of law. Legal theory has typically struggled with this issue in one of two ways. Theories like those of Bentham and Austin define law in terms of the ability to inflict sanctions, which not only neglect nonstate forms of law, but also seem to ignore both the kinds of law that operate in the absence of sanctions and the occasions when law (which typically applies sanctions to a breach of duty) will not be able to do so. Hart, building on the observation that legal meanings continue to be generated in the absence of sanctions, offers a replacement theory.²¹ For him, obedience, at some unquantifiable level of significance, is a precondition to any claim by a group administering a system of rules to be legal officials, and to any possibility for the system that they apply to be called “law.” But once this level of obedience operates, the only acknowledgments of the role played by sanctions is that they will be understood as “deserved” by those who breach primary rules, and that those who administer punishments can explain their behavior, to themselves and to others, as acts taken in accordance with legal rules. Neither of these approaches—state law as sanctions or state law understood without reference to sanctions—seems satisfactory for those who wish to compare state law with nonstate forms of law, especially in those situations where the two coexist.²²

While it would be wrong to say that systems theory excludes violence from society, how does it include it within law? What is its explanation of the relationship between legal norms and the actions of wardens, police and bailiffs, which provide so much of the motivation to make legal communications? The theory initially excludes violence from law, in the sense that what law distributes through its communications (its medium) is legality, not power. Power, which includes physical coercion, is distributed through the political system. These are two subsystems with their own codes and conditional programs, and each can only communicate about the other through combinations of its own communications. This means that each system can only construct an internal description of the other. The idea that law and politics are closed to each other in this way, with the distribution of political force left to the political system while the legal system can only distribute legality, seems an unpromising beginning for identifying state law and its ability to distribute violence. To proceed, we need to introduce another concept from the theory: structural coupling.

Structural coupling is systems theory's answer to the question of how social subsystems can be both autonomous and coordinated. Because each system uses and accounts for a different code, and creates its own internally interconnected configuration of communications, there are no common or metameanings through which this coordination can be achieved. Thus, even when communications with an identical semantic form exist simultaneously within two different systems, the meanings they generate are not the same.²³ Structural coupling occurs when a system “presupposes certain features of its environment on an ongoing basis and relies on them structurally,” such that “the forms of a structural coupling reduce and so facilitate influences of the [system's] environment on the system [italics in original].”²⁴ In the case of state law we have two systems, each of which has coevolved in response to the mutual predictability of some of the other's operations. A political system can rule through law via legislation, on the basis of the regularity of the legal system's reaction to these particular political communications. A legal system can provide a stable reaction to the political system by ignoring most of its communications (lobbying, political speeches, leafleting, lapsed bills, and so on) and focusing on the meanings of bills passed by Congress (to which it attributes a unitary will despite all the evidence to the contrary). The relative predictability of the legal system's reaction provides an incentive for much political communication to take the form of attempts to introduce new laws. Law has offered politics a form in which to express power with technical precision (in comparison to an attempt to direct power solely by reference to class interests, personal loyalties, ideology, and so on). And in turn, politics has provided law with an enforcement mechanism that provides incentives for its communications. If we return to our example of the warden's attention to the legal niceties of different orders surrounding the death penalty, the legal system is able to distribute a near monopoly on violence only where the political system has coevolved in response to structural coupling to create the relationship generally known as the rule of law.

The extent to which the political system has been stimulated to seek the benefits of constructing the basis of collectively binding decisions by reference to legality has varied over time and space. Defining law, or even state law, as a system that distributes some minimum level of political violence necessary for effective enforcement deflects attention from empirical questions that go to the heart of legal pluralism. We are likely to ignore not only the extent and manner in which structural coupling between law and politics has evolved over time, and in different areas of the globe, but also the possibility that structural coupling between law and systems other than politics raises other forms of coevolution than that of nonstate law. By contrast, if we view state law as a coevolution of two separate systems, each with its own code,²⁵ then we can explore the possibility that law can structurally couple with other social systems, such as the economy and religion. Much writing on legal pluralism is directed against the assumption that law must be understood to originate from the state, either explicitly or through acquiescence. Recognizing the mutual closure of the two systems, the manner in which each has reacted to changes in the other, and the way in which each constructs its own version of the other to react to is a methodology that is not limited to state law. So, for example, the economy and law have a close structural coupling oriented around contract. While the economy constructs a contract in terms of a likely flow of future costs and benefits, law communicates in terms of a reconfiguration of legal relationships—new rights and duties. The ability of the economy to reuse legal communications in order to provide stability to economic transactions²⁶ is not coextensive with the political system's use of legality to distribute political power. A focus on state law is likely to result in the absence of state enforcement of contract law's being viewed as a failure of law, or a rejection of law. But as the example of lex mercatoria demonstrates, some transnational contracts do not originate in state law (explicit or implicit) and do not rely on the enforcement mechanisms of particular states, or sometimes of any states. The ability to establish a transnational contract law that is not dependent on any particular state's political system is a coevolution of law and the economy. It represents the creation of both new law and new economic transactions, with each responding to the other but drawing upon its respective existing store of communications to produce the new development. The economy does not give meaning to legal transactions in terms of the complex relationships between duties and rights that constitute their existence within the legal system. Instead, legal contracts are reinterpreted in terms of the likely profits and losses that may arise from different courses of action. But the role of legal agreements, and their ability to be translated into economic costs and benefits, is not limited to those that are created according to state legal rules or enforced through state or regional government sanctions. In the study of legal orders, we should move away from notions of a threshold level of effectiveness closely linked to the availability of state sanctions. Instead, we need to consider the effects of legal communications—not, we stress, effects in terms of any linear theories of legal cause and social effects, but effects in terms of the nature and extent to which the communications of a particular legal order are or are not reconstituted as communications within the economy, or within religion, or even within culture, and vice versa.²⁷

Law in Translation

The question of whether “law” exists raises questions of translation. Tamanaha, with his conventionalist approach, relies on the practices of translators:

Whatever translations there are for “law,” including whatever indigenous terms are used to designate the existing state legal apparatus, will satisfy the requirements of the conventionalist approach. Thus, existing translations of the term “law” will have already done most [of] the necessary work for the identification of “law” in non-English contexts (keeping in mind that disputes and borderline cases will remain, with the default rule being inclusion) (Tamanaha 2001: 203).²⁸

The nature of translation is important to, and has been taken seriously for some time by, others engaged in legal pluralism studies (Allott 1980). It also has significant implications for systems theory's reliance on codes. Could such reliance be a more useful starting point for engaging in legal pluralism study than some other approaches to translation? Indeed, could it reduce the manifest problem of translation? To start considering this question we need briefly to consider how translation operates.

What allows us to say that a term used in one language has equivalence in another? At least since Saussure, we have known that signs generate meanings within language through their relationships to each other, and at least since Wittgenstein, we have known that these relationships are established by use, not rules. Thus there is never an exact translation of the terms of one language into another, as the respective terms will always have possibilities of connection within one language that are different from the possibilities of connection within another. The problem is “doubled” within systems theory by the claim that meanings created by communications are a consequence of the connections between communications within subsystems, so that even the meanings of identical terms within identical sentences used by different subsystems cannot be the same.

How do we know that individuals who do not speak English are referring to “law”? There is no such thing as a one-word “literal” translation; even words that represent physical objects do not translate unless those objects are used in manners that the translator perceives to be similar. Complex words like right, duty, and corporation are not, as Hart stresses, reducible to objects. Rather, they take on meaning within complex arrangements of words²⁹ so that the translation of one such word requires a translation of accompanying words, to see if a similar word game is being played in another language. A nonessentialist approach removes the tools whereby the translation of the word law might be disciplined.³⁰ One cannot look to function or essential forms (or translations of function and essential forms) to justify what is considered to be the use of a word equivalent to law in another language. One may be content to rely on the work of translators at the level of accepting that droit/loi and recth/gesetz are references to law, but is this due to the number of times this translation occurs, or to the fact that legal translators, like most comparative lawyers, are content to accept that whatever else might not be law, state-enforced rules certainly are? Can we have the same confidence where translators consider a practice to be “customary law” or “religious law”? On what basis would translators be expected to distinguish what was “merely” custom, or religion, from practices that had the additional, or alternative, status of law? How does systems theory deal with these issues?

The identification of legal communications with the application of the code legal/illegal results from an observation on the manner in which the legal system constructs itself. It is a distinction that is different from the code applied by other systems. In the case of the economy the distinction is between payment and nonpayment. The application of the code provides the basis for the next application of the code. A conventionalist approach does not capture this use of codes. There is no claim that each application of the code has the same semantic form. For example, the fact that the convictions of defendants make their continued detention legal is not a statement that any person needs to make either at the time of the convictions or thereafter. Similarly, within the economy, the fact that someone redeems a mortgage creates the conditions for further transactions; no one has to utter the word payment. A binary code is a distinction with a positive and a negative side.³¹ Functional differentiation could not occur if every social system relied on the same mathematical symbols (+/−) to indicate this process of coding. Different codes are applied. But the semantics that represent the application of codes are not immutable. As such, claims that the code of the science system is truth/nontruth, or that the code of the media system is information/not information, or that the code of the legal system is legal/illegal do not amount to claims that these words are used on every occasion when coding occurs. The claim that the code of law is legal/illegal is persuasive because, at least in the English-speaking world, these words are frequently used in the operations of the legal system.³² But as a theory of society, one needs to understand how codes operate in each social system, not just the legal system.³³ And this may in turn increase our understanding of the nature of the legal code.

If we consider the media, the code identified by Luhmann is information/not information, a code that unifies the mass media as a system that includes advertising, news, and entertainment (Luhmann 2000b). But if we were to contemplate which words represent this distinction within daily newspapers, we might come up with “news/not news” or “story/no story,” since we can readily imagine the rejection of a submitted article in terms of the statement “This is not news,” or a journalist arriving at a location asking, “What is the story here?” We can apply a similar analysis to the manner in which scientific communications link. The claim that science codes in terms of true/not true (false) makes sense as a description of a code that links scientific communications and makes them a system rather than a random set of communications that have no relationship to each other, without the participants having to use the words true or false within their communications. Similarly, we can understand that a medical system is not identical to the science system, and that the positive/negative coding that makes possible the connection of medical communications, and thus makes them meaningful (since connection is what generates meaning), is itself understandable as therapeutic/nontherapeutic without these words being applied on each occasion when the medical system's communications apply negative or positive values. Thus attempts to articulate these codes can only partially be explained by conventionalism, if conventionalism is based on the semantics of the participants. Teubner refers to the implicit or explicit invocation of the legal code (Teubner 1997: 14–15). Codes can be “implicit” because they are abstractions. If one had regard only to the “news” industry, and sought the code that linked all “news” items, one might describe it as “news/not news.” But “the news” is only part of the mass media. If one sought to describe and identify the “implicit” code that allows communications to connect from advertising, news, and entertainment, it would be “information/not information” rather than “news/not news.”

The first issue is whether one can identify the codes through which communications can interconnect and generate system-specific meanings. One is looking for a way of describing how a particular system applies a negative or positive value throughout all of its operations. One might do this in the English-speaking world by using the codes payment/nonpayment, legal/illegal, true/not true, information/not information, government/opposition, and so on. One does this accepting that a modern society has economic, legal, scientific, mass media, political, and other systems, and that these are in some way different from each other. Looking for codes and applying the distinction program/code are systems theory's answer and methodology for seeking to identify and describe this process of differentiation, as well as the nature of the relationships that are possible among these different parts of society.

With this understanding of codes, translation is not the problem that faces us. One does not simply go to translators and ask them what word they might substitute for legal, information, true, and so on. With systems theory, one starts seeking to identify how religion, law, and economy are differentiated and how they interrelate within another national territory, within an entire region, or around the world. The fact that the observed local participants do not speak English would not prevent a person who undertook this exercise and sought to describe it to an English-speaking audience from using English versions of the codes. A person who did this and sought to describe her results to a non-English-speaking audience might choose words from that audience's language to describe the different codes that make it possible for an economy to be separate from a political system, for example. The actual language used by the participants observed when coding these systems is not critical to the presentation of differentiation. One might say that Tamanaha would start with translation (of the word law) and then treat as law whatever his translators told him was referred to by this translation; whereas systems theory would start with the observation that the differentiation of law, economy, politics, mass media, religion, and other systems is a feature of modern society, continue by identifying how that differentiation operates, and only then translate the results.³⁴ Michael King offers a useful example of how this process occurs in practice:

A family in a remote village in Botswana watching the O.J. Simpson trial on a television set understand that what they are seeing is law and that this makes it different in nature to health, religion, politics or any other kind of communicative event. … [O]ne does not need to go back very far to arrive at a time when the notion of legal as being quite distinct from other meanings would have been incomprehensible (King 1997: 123).

King is pointing to the fact that these villagers have experienced and would understand differentiation. They would know that this was not a religious ceremony or a political event or even a resolution of a dispute between members of a community, but a legal proceeding in contradistinction to all these other things. Exactly which words the villagers in King's example use to describe law, economics, or religion is not what is important. Rather, what is important is their participation in these separate social subsystems. While this will not involve conditional programs identical to those of the United States, this participation can be observed and mapped with reference to the codes of the respective systems.

This issue of translation is related to, and informed by, another topic of concern to what we have described as a second wave of legal pluralists: globalization. The reductionist nature of codes—the fact that they exist only as negative/positive distinctions, but nevertheless manage to be different positive/negative distinctions within different social systems—allows for connections within a system despite changes in language. This is easiest to see in the case of the economic system, which is the system that is most widely accepted as a global one. While each side of a global economic transaction may speak a different language, the positive/negative coding whereby economic communications help carry out operations and establish the basis for further communications occurs nevertheless. According to Luhmann, this coding, which has been translated into English as payment/nonpayment, might be translated into any number of languages. But speakers of any language will recognize that something positive has occurred in terms of expectations of an increase or a decrease in resources. To use a metaphor here, codes can travel where conditional programs cannot. The ability to trade with people who speak another language allows economic communication to be global, but it does not mean that the conditional programs created with each language are interchangeable. This, unlike coding, would require translation.

Exploring Legal Pluralism through Systems Theory

One of the stimulants for legal pluralism has been the discovery that defining law in terms of state law results in the conclusion that premodern societies that exist without states are without law (Griffiths 2002: 293–294). Similar conclusions result from attempts to define law without reference to the state yet retain the features associated with state law, such as institutions, universality, formality, and so on. Does the same thing occur when the legal system is identified as a system coding in terms of legal/illegal? The answer is yes. Does this matter? The first point to note is that this conclusion is not reached through the application of an arbitrary classificatory scheme applied to normative phenomena; instead, it arises out of a theory of society. As such, if one objects to this approach to the study of law within society, one needs to take account of the explanatory potential of systems theory as a means to understand society and the interactions within it, rather than focusing on the pejorative consequences of finding that particular societies may not have a legal system.

Following on from this, systems theory analysis of law is the application to law of a theory that understands modern society as consisting of functionally differentiated social systems (see Luhmann 1977, 1997: chapter 4, “Differenzierung”). If we accept Luhmann's claim that modern societies have evolved from relatively undifferentiated societies, in which communications that we would today identify as religion, politics, and law did not exist in separate systems, to ones in which they do, then there is nothing descriptively inaccurate in concluding that what anthropologists and historians identify and study as premodern societies lack autonomous legal systems. Law as a separate social formation is, within this theory, law as a social system whose separation from other systems is achieved through a binary code, and whose relationship with other systems requires one to apply the distinction between code and conditional program. Other sociological approaches would seem to reinforce the descriptive accuracy of this understanding of modern society—for example, both Durkheim's focus on the division of labor and Weber's observation that modern society consists of competing rationalities acknowledge differences between the modern and the premodern in terms of an intensification of differentiation.

This process of differentiation also has altered the nature of societies’ hierarchies. Premodern societies were organized in terms of vertical segmentation (tribes) and horizontal classifications (classes and estates), and these hierarchies were understood as something natural, or given, within a society that is bound, forming an organic hole. Functional differentiation fractures society, so that the possibilities of communication (opportunities for meaningful action) are structured separately within each system. This does not mean that opportunities are distributed equally among all people, so that the inequalities associated with tribe, kinship, or class disappear. But because tribe, kinship, and class do not form the basis of the communications that organize operations within these separate systems, their evolution and their colonization of society make it increasingly difficult for these vertical and horizontal hierarchies to present themselves as “natural” and, as part of this, to maintain an organic and holistic account (self-description) of society.³⁵

On the basis of this theory of the evolution of modern society, there is good reason to resist equating what anthropologists and historians identify as premodern forms of law with the law of modern societies. The societies in which this common term is supposed to be applied lack the separation of social formation that allows one to identify something as “law” in contradistinction to morals, economy, politics, and other systems. This is not the case at the level of premodern societies. One can call the normative order of a society that does not distinguish between law and other communicative systems “law” if one wishes, and thus avoid the unfortunate legacy of legal theories that associate the presence of law with progress and civilization. But one faces considerable difficulties with this approach if one attempts to reapply it when undertaking ethnographical or anthropological studies of modern societies.³⁶ The acceptability of calling everything “law” within a society that does not have differentiated social formations (social subsystems) provides an impoverished scheme of analysis when attempting to describe a society that does have these formations. It does not aid analysis to deny that the evolution of modern societies has resulted in the development of separate social formations that society has constructed as “law” in contradistinction to economy, morals, or politics.

Systems theory provides a sociological understanding of the difficulties of having a common referent, “law,” for both state law in modern societies and nonstate law in premodern ones. But can it do more? Can it inform our understanding of the nature of each of these phenomena, as well as the relationship between them? When assessing the ability of systems theory to meet this challenge, one should not conclude that systems theory identifies the legal system with state law, in the sense that the characteristics of state law limit the possibilities of further legal evolution. Because systems are differentiated by their codes, and not by their current structures, those structures can only stabilize the current applications of the code; they cannot determine the future evolution of the system. In the case of state law, this represents a stage in the evolution of the legal system within modern societies, but not its end state. The prevalence and central role played by national legislation within modern legal systems generates a self-description of law, or jurisprudence, which is an abstract and more general representation of these kinds of legal communications: sovereignty, national constitutions, and so on.³⁷ But systems theory does not require us to accept that the communications generated within the legal system at a moment in its evolution—including those whereby the legal system describes itself to itself as a totality—are essential, or that they determine the future possibilities of what can be legal within that system. Any attempt to insist that all law is state law denies the presence of law within premodern society, but as Fischer-Lescano and Teubner (2004) have aptly demonstrated, it also dis-applies this label to the ever-expanding transnational legal communications that apply the code legal/illegal.³⁸ The assertion that premodern societies have no legal system (a system separate from politics, the economy, and religion) is a valid claim, if the modern is understood in terms of a transition to functionally differentiated social systems. But this does not compel us to accept that legal systems are unable to evolve beyond the forms associated with the nation-state.

What, then, of nonstate law within a nation-state territory? Here there has been an important shift in one of the central assumptions that informed many of the classic case studies in legal pluralism: the tendency to treat the studies’ communities, typically villages, as closed societies. It is now accepted that local communities, even in some of the most remote places in the world, are interpenetrated by elements of the wider societies that surround them.³⁹ More recent studies acknowledge and explore the manner in which these outside elements, which typically include state laws, impact the norms of the local communities. This change of assumption informs, for example, Moore's (1978) concept of semiautonomous fields and Santos's (2002: 437) idea of “interlegality.” But this recognition also diminishes the strength of any argument that a social theory developed in order to explain the nature of modern society has no application to the kinds of societies and case studies that form the subject of legal pluralism. Participation within a wider modern society requires the members of a local group that may have identified itself by reference to a common religion or ethnic heritage to engage with social systems. The members of this group become patients within a health system, pupils within the education system, voters within the political system, consumers and entrepreneurs within the economy, and litigants or defendants within the legal system. In particular, what pluralists might call religious or customary law becomes only one part of a life that is also experienced through these other systems (see King 1995 and Luhmann 1984).

This process has implications for the practice of comparing the strongly held commitments of local communities with the rules and practices of a state legal system. The threat to such local commitments is not limited to how state law might threaten the ability of a particular religious or ethnic group to continue to live according to its own norms. The ability of these particular communities to exist within a modern society, while retaining a strong and widely distributed sense of a particular way of life, is premised upon their isolation from a wider society that exhibits functional differentiation (King 1995: 112–113). The economy, the political system, the science system, the medical system, and the educational system all pose problems for this form of social life. It is true that the threat of violence, organized and administered via a state legal system, represents a particularly harsh risk for any community that operates on the basis of a self-conscious insularity within modern society. Nonparticipation in systems is an option whose costs to the individual rise when the state imposes penalties for this. But there is a definite sense of legal fetishism in seeing the threat to such communities as something that comes solely from a state legal system, rather than from the fracturing of community commitments and understandings that results from participation in all of these social subsystems. As part of this, one must also consider the extent to which different legal orders generate relationships of structural coupling with the economy, political system, mass media, science, and so on. If, for example, it is local custom, not state law, that provides the norms that operate within the economy at a local level, then people are motivated to continue to utilize custom. But where economic transactions take new and dynamic forms, as where there is participation in global markets, a positive legal order is more likely to generate norms that can incorporate the semantics of those transactions. In order to study legal pluralism within modern society, we have to consider the ability of different legal orders to generate their respective legal meanings and, as part of this, to examine their respective abilities to incorporate the semantics of other social subsystems.

Conclusion

In this article, we have attempted to demonstrate the potential for using systems theory to study forms of law that are not restricted to the features associated with state law, without moving to an acceptance that all forms of normativity or social control constitute “law.” We began by looking at the presentation of this theory by Tamanaha, who claimed that the weakness of a systems theory approach lay within Luhmann's wider theory of modern society as characterized by differentiated social systems. We have argued by contrast that studying law using this theory of society is a strength, not a weakness. Modern society has evolved in a way that allows the legal to be constructed as something different from the rest of society—the moral, the economic, the political, and so on. Systems theory, with its identification of modernity with functional differentiation within separate social subsystems, each self-producing through the interlinking of their own communications and applying its own code, offers a basis for understanding why, in the modern world, it is possible to distinguish law from other subsystems of society, and why that distinction is not dependent on any essential set of institutions or structures. Systems theory neither limits its recognition of the legal to state law nor ignores state law and its coercive potential. Rather, systems theory's explanation of state law in terms of the structural coupling and coevolution of law and politics provides a basis for exploring the relationships between law and other subsystems that generate and sustain nonstate law. In other words, through its understanding of state law it stimulates the study of forms of coevolution in which patterns of legal pluralism arise.

It may well be that the pluralistic elements of this approach will not satisfy all who consider themselves legal pluralists. Nevertheless, the recognition that all communications that apply the code legal/illegal form part of the legal system—not simply the communications of particular members of state institutions—moves us away from the assumptions of theories that identify all law with the state. And this does not lead to the conclusion that all norms, rules, or values, no matter how strongly actors may be committed to them, constitute law, whether or not they apply this term themselves, a conclusion that has bedeviled the legal pluralist project. Systems theory potentially enables the insights and motivations associated with the first wave of legal pluralism studies to be imported into the second wave—in other words, to enable legal globalization to be studied as a branch of legal pluralism, without the restrictions of state-centered perspectives and in the light of the understandings built up in earlier legal pluralism studies.