‘. . .the workings of the judicial process [are] conducted upon the tacit assumption that the common law (we are not concerned here with statute) always provides an answer to the matter in issue, and one which is independent of the will of the court. Put differently, the conventions of legal argument embody a belief in the theoretical possibility of a comprehensive gapless rule of law. It is as if lawyers had all been convinced by Dworkin, though none of them have.’
(Simpson, Reference Simpson and Twining1986, p. 9)‘There is no doubt that the familiar rhetoric of the judicial process encourages the idea that there are in a developed legal system no legally unregulated cases. But how seriously is this to be taken?. . .it is important to distinguish the ritual language used by judges and lawyers in deciding cases in their courts from their more reflective general statements about the judicial process.’
(Hart, Reference Hart1994, p. 274)‘When we recognise that respected judges, like Cardozo, whose opinions resemble those of other judges, write off the bench about judges’ legislative functions, and we note that their views are often not challenged by other judges, we should pause in concluding that judges who write typical opinions either do not believe in discretion or are hypocrites.’
(Greenawalt, Reference Greenawalt1975, p. 384)I. Introduction
These three quotations are directed on their surface towards the issue of whether or not judges exercise discretion. But they raise a wider issue too. Why should there be a divergence between what judges say, and what they believe? The extent and nature of this divergence is central to much jurisprudential debate. Hart admitted to the existence of conditions that might lead to rhetoric, namely the inevitability of penumbral or hard cases, without discussing the reasons why judges might continue to speak in such cases ‘as if’ they were finding law, and not making it. Dworkin took the manner of judicial speech in such circumstances as evidence that, even in hard cases, judges were indeed finding and not making law (or at least only exercising weak discretion). Various realist and critical theories identify disingenuousness in claims to apply rather than make law not only in hard cases, but in all cases. Alongside different views as to the extent to which judges fail to acknowledge what they are in fact doing, when deciding law, there is also a range of views as to the reasons for these potential failures. Judges lack the political authority of legislatures and, as such, could expect to be criticised for undertaking a legislative function. Parties who lose in such cases might expect to be particularly aggrieved. As well as avoiding criticism, judges gain authority, and prestige, by claims that they primarily exercise a technical function in identifying the law. And, on a more positive note, a judicial reluctance to admit to making law can be seen as part of a judicial reluctance actually to make law, which can in turn be seen as a commitment to various constitutional principles, such as democracy and the doctrine of the separation of powers.
Investigating the scale and nature of this divergence between what judges say, and what they know to be the case when making decisions, has taken many forms. Much of the debate between Dworkin and his critics has not been empirical so much as analytical. The practical and linguistic difficulties of presenting any claim of there always being ‘right’ answers to making legal decisions led Dworkin to back down from his earlier more absolutist position. And of course there is empirical evidence, in that judges, off the bench, have been willing to admit that they make or have made law, though they have tended to stress the exceptional nature of this. Deconstructing or ‘trashing’ judicial decisions has seemed both to confirm the fact of judicial law making, and challenge judicial claims that this is a rare occurrence by demonstrating the alternative possibilities within any court decision. Those who accept that legal sources fail to explain the decisions of judges have looked to external explanations for the construction of judicial choices, with Feminist, Marxist and Critical Race theories making significant contributions at this point.
So, the issue of judicial failure to acknowledge the ‘real’ possibilities of choice is not a minor one in so much jurisprudential discussion. That said, is there anything new that can be said about it? And would such a contribution allow us to see the prior debates in a new light? We believe that the writing of Niklas Luhmann, most particularly his book Law as a Social System (Reference Luhmann and Ziegert2004), offers such a possibility. In this article, we apply his systems theory, particularly as set out in that book, to ask the question of the title of this article: why do judges talk the way they do? Why is there a divorce, on at least some occasions, between what they believe, and what they say, and why does the repeated exposure of this divorce leave their communications unaltered, in that their legal judgments continue to fail to recognise this divorce?
To ask such a question as this, using the writings of Niklas Luhmann, will surprise many. For example Roger Cotterrell, in his recent book Law, Culture and Society: Legal Ideas in the Mirror of Social Theory (Reference Cotterrell2006), argues that an adequate understanding of legal ideas is impossible without adopting a perspective informed by social theory, but proceeds to reject Luhmann’s theory as a possible grounding for such a perspective on the grounds that the theory exhibits an almost ‘impenetrable abstraction’: that it neglects to examine ‘the changing character of the social in concrete terms in relation to law’, and fails to explore ‘the details of the discursive character that it attributes to developed law’(p. 23). In a similar vein Dennis Galligan, in his recent book Law in Modern Society (Reference Galligan2006), urges the need for sociolegal studies to ‘take law seriously’. For Galligan, ‘Law is taken seriously for the purposes of understanding its role in society by examining the concept itself, and then seeing how it interacts with other aspects of society’ (p. 6). But he rejects the use of Luhmann’s systems theory for this task (preferring to utilise Hart) for reasons which include the claim that since, for Luhmann, ‘human actors are not regarded as part of law’s operations [an] enquiry [into the actions of judges and] the reasons they give, and the way that cases are framed in legal terms. . .is out of bounds’ (p. 41).
Our task here is therefore an ambitious one. In order to overcome ‘impenetrable abstraction’ and to demonstrate the potential of Luhmann’s theoretical insights to explain the discursive nature of modern law, we wish to apply systems theory to a concrete example: judicial communications. As we hope our introduction has shown, this is an important example, one that provides a central focus of many legal theories, and the starting point for others. In seeking to show that Luhmann’s systems theory can be ‘useful’,Footnote 2 we face a further challenge in explaining the nature of Luhmann’s distinction between psychic systems and social systems. For it is true, as Galligan has stated, that humans are different from the social systems in which they participate as addressors, and addressees. But a clear analysis of what communications consist of, within Luhmann’s theory, will, contrary to Galligan’s claim, not only increase our understanding of exactly those aspects of a legal system that he feels are excluded by this type of explanation, but also explain the limits of an individual’s ability to exercise choice when participating in social systems (including law). The question, ‘why do judges talk the way they do?’ is an opportunity to explain the nature of this participation.
In the next section of this article we provide a brief introduction to modern systems theory, and the role played by its concept of redundancy in forming a link between the social and the individual. In the following two sections we apply these insights to judicial communications and the commitments that they represent, and the character of judicial discretion.
2. Social systems, psychic systems and redundancy
The social, within Luhmann’s theory, is communication.Footnote 3 We cannot per se experience another human’s thoughts. Until thought is communicated, which may be by speech, gesture, writing, and even on occasions by inaction and silence, we cannot claim to ‘know’ what is in another person’s mind. And, of course, even communicated thought is, if we consider it carefully, only a communication. We never actually experience the other person’s thought. Similarly, we can communicate about, or think about, another person’s body. But that person’s body remains separate from ours, and indeed even separate from its own thoughts.Footnote 4 In this sense, humans lie outside communication. This may strike the reader as specious, or unnecessarily complicated, but legal theory, and legal doctrine, is full of communications about these unreachable aspects of a human being that seem to belie the fact that they are unreachable. So whenever you read or hear words like ‘inclinations’, ‘motives’, ‘purposes’ and ‘feelings’ you are engaging with communications about something that itself lies outside communication. The consciousness of an individual, the thoughts and feelings that they experience are, within Luhmann’s theory, a system, but he terms this system to be a psychic systemFootnote 5 and carefully distinguishes it from social systems, which consist of communications.Footnote 6 Thus, when Galligan says that for systems theory the human is outside systems of communication he is correct (as regards human minds and bodies), but when he goes on to conclude that systems theory has nothing to do with the reasons that judges give for their decisions or the manner in which legal cases are framed he is wrong, as these reasons are communications (the central ingredient of this social theory).Footnote 7
Individuals communicate (as do corporations, and computers). Luhmann has drawn on communication and information theory to delineate carefully what this involves, and what in turn constructs the possibilities of communication. In particular he makes use of two distinctions: redundancyFootnote 8 and information, and redundancy and variety.Footnote 9 These two distinctions are interrelated. Information (meaning that is new) can only be conveyed against the background of redundancy (meaning that is unchanged). What within any communication is redundant, and what is information, is not static and given. But the meaning generated by a communication is always the difference between the two. The first of these distinctions is best understood by considering a single communication. The information provided by a communication is what is new. But in order for a communication to provide information other communications must remain the same. Redundancy in communication is a necessary background to information. We experience this distinction on an everyday basis without thinking about it. Road signs take a standard form. The symbols used, by their very homogeneity, make it possible to extract information from them easily: ‘Expect – deer/bends/junctions, etc.’ Time is another classic example. When we are told that the time is 9.25, this makes sense against a background of communicating time in terms of 60 minutes to an hour. Timetables represent a further refinement. Not only do these rely on the limited number of ways in which we record time, but by setting out information in standard arrangements they make it easy to extract what is different (information) from a background that remains unchanged (redundancy).
Whilst the distinction redundancy/information goes to the actual information generated by a particular communication, the distinction redundancy/variety goes to the potential information that could be communicated. Against any given background of unchanged communication, what is the range of new information that can be generated at any given moment? In the case of our timetable, there are any number of times that can be given for the services being timetabled. So we have a lot of variety. But, on the other hand, there are not a lot of ways in which time itself can be communicated within that timetable if the communication is going to be successful.
The example of communications about time and timetables may suggest that redundancy and variety have a linear relationship, whereby things that begin as variety become redundancies, creating new variety (new possibilities) – standard time creates the possibilities of timetables – developing timetabling creates increased possibilities for communicating about the organisation of events, etc.Footnote 10 And quite a lot of society’s systems of communication (including law) evolve in this manner. But this is not always the case. The information that can be generated against the background of any given redundancy includes the possibility of creating new redundancies that can challenge, and displace, previous ones. However, it is impossible to make a communication that immediately deconstructs the communications that constitute the background redundancy that generates the information contained in that communication. So the process by which variety can lead to new redundancies that displace earlier redundancies is constrained. Instead of such instantaneous deconstruction of prior redundancies, variety has to lead to the establishment of new redundancies from existing ones. Only on this basis can rival redundancy, and the displacement of earlier redundancy, occur. What we have here, as Luhmann rightly observes, is an evolutionary process:Footnote 11 one can only evolve new communications from the communications that are already available, even ones that will replace those which currently exist.
Redundancy represents a real constraint on communicative possibilities. We have to use redundancy to convey information. And the costs of abandoning redundancies are often considerable. The example of changing the basis for telling time makes the point very well. Should we decimalise time? Well, not for those communications where this would require new clocks, the re-programming of all computers, the translation of large amounts of existing technical and non-technical literature, and the alteration of vast amounts of general social communication. Dispensing with what is the redundant background to information always requires that information (if we still wish it to be communicated or transmitted without entropy, or random errors) be generated through a new redundancy. Not only will this require a cost in terms of creating information in new ways, but one cannot assume that attempts to generate new ways of conveying information will be successful. They are only successful if they succeed.
Now consider the position of human actors. To communicate information, humans have to make use of the store of available redundant communications. As individuals, they cannot successfully communicate (transfer or receive information) without utilising redundancy. Of course one can imagine setting up new ways of communicating between pairs of individuals – along the lines of a private language – but think of the constraints. Not only would those individuals have to utilise existing social redundancies to construct their new communications, but they would not be able to use them with others without a process of initiation. And what is going to motivate all these other individuals to adopt this new system, given the presence of an alternative and in the short run lower-cost alternative. Of course if, like the changing of a currency or driving on a particular side of the road, these new communications have to be simultaneously introduced to a large number of persons all at once then the costs of change are enormous.
How much of a restraint does redundancy place upon what can be communicated? The examples used above, most particularly that of time, suggest that redundancy can be a major constraint. But the use of the particular example of time also points by contrast to the fact that little of our communication seems quite as rigid as this. It may be convenient to communicate in the same manner on lots of occasions, but does not the vast store of communications available to us as background give us a substantial ability to vary what can be said? If we regard our store of communications as an undifferentiated mass, then this seems pretty much the case, even allowing for such central co-ordinating communications as a common basis for identifying time. But if instead we accept (or at least engage with) Luhmann’s description of modern society as consisting of functionally differentiated and constantly evolving subsystems of communication,Footnote 12 then we can see that redundancy may provide significant and systematic constraints on what can be communicated in particular contexts. The claim that society consists of separate systems of communication (the legal, political, economic, media, scientific, etc.),Footnote 13 necessarily involves the claim that redundancy is system-specific. Luhmann’s theory of society as made up of autopoietic systems of communication requires us to consider not only that redundancy is necessary for information, but that redundancies are not identical within the differently functioning systems or subsystems of society.Footnote 14 What constitutes redundancy at any moment within legal communication is different from what constitutes redundancy within other systems of communication. Thus what can be communicated within the medical system as a subsystem of science is different from what can be communicated within the political system or the media about medicine.Footnote 15 Science has generated communications that provide background redundancy to all sorts of scientific information, as has politics for political information, media and law, with respect to the information that they generate in turn. And Luhmann’s combination of an understanding of society as separate systems of communication together with the concept of redundancy not only leads us to conclude that redundancy is system-specific, in the sense that different systems can employ different redundancies on a routine basis, but also requires us to accept that the redundancy that can be utilised within a system, at any particular moment, is established by that system.
Returning to the human actor, communication requires participation in a system of communication. Successful communication requires the utilisation of redundancy in order to communicate information. And that redundancy depends on the system that the individual is seeking to participate within, and the operation within that system which the individual is seeking to affect. This does not reduce the individual to an automaton, whose communications within a system are entirely determined by the system. But it does create a situation in which the actor is severely constrained in how and what they can communicate. And it also provides a link between the system of communication and the actor. If the actor is motivated to communicate, the terms of that communication are dictated by the system in which they participate. The redundancies of that system and its variety (what can be successfully transmitted at any point in the system’s communications) are given by the system. Human actors, through participation, learn to recognise redundancy, and to utilise it in the construction of their communications. And, on the negative side, there is a cost to any human actor who fails to identify redundancies, which is to fail to make the communication successfully, and to achieve the operation that they would have wished to achieve. This has implications for many of the claims made about the nature of individual actions and beliefs.
These observations help and indeed require one to understand that individuals can and do participate in more than one subsystem, and are able to communicate different things within these different systems. So, for example, we should not presume that everything that judges communicate is a legal communication just because they are judges. Judges who make press statements are attempting to have their communications selected by the mass media. As such, there is no reason to expect that these communications will generate further legal communications. Whilst such communications are likely to use the words ‘legal’/‘illegal’ (the code of the legal system),Footnote 16 the aim of such communications, if selected by the media, is for them to be coded by the media using its code (information/non-information, or more colloquially, news/not news).Footnote 17 Similarly, when judges give public lectures, or publish law review articles, their communications will be selected in different ways (applying different codes and programmes) and generate different further communications, in different systems: education, politics, media, law, etc.
Utilising the distinction redundancy/variety not only refers to what can be communicated at a particular moment, but allows us to describe the process whereby new possibilities of information are generated. We can attempt to track communications that generate new redundancies, and new variety. Like Luhmann, we regard this distinction as an important medium to allow one to describe and analyse the evolution of systems. And, if we take on board the notion of separate systems of communication, and the ability of systems to determine what redundancy is available for the next communication, then we also have to recognise that the ability of individuals to communicate about their ‘internal’ feelings will vary from system to system, and within systems. This leads inevitably to a considerable inconsistency in the correlations between individuals’ consciousnesses and their communications. We can surmise, from the fact of an individual’s communications, that they were motivated to communicate. We can further surmise that they were motivated to identify and engage with the redundancy present at the moment of communication. But we cannot read off from their communications exactly what those internal feelings might have been. In particular, as in the case of law, where those communications take a normative form, and implicitly communicate an internal commitment to the legal system, or particular legal norms, we cannot assume, from the fact of communication, anything more than a desire to execute a particular operation.
Applying our discussion of redundancy, we need to examine communications, including judicial communications, in terms of the operations which they affect, and the variety of communications that are open to judicial actors given the presence, at the point of their participation in the legal system, of redundancy. In the next two sections we take two issues, central to much legal theory, and explore them using this analysis. The first is the claim that public officials, most particularly judges, must have a greater commitment or more committed internal attitude to the legal system than other actors. The second is that referred to in the first of the initial quotations: the issue of judicial discretion. We will investigate what operations are enabled via a redundancy represented by ‘conventions of legal argument [that] embody a belief in the theoretical possibility of a comprehensive gapless rule of law’, that would not be affected otherwise.
3. Judicial communications and ‘commitment’ to the legal system
Let us begin with the position facing judges involved in adjudication on substantive legal issues. The context for these issues is a mass of legal communications. For a start judges are judges, appointed as such to courts, with jurisdiction to decide the issues before them (if their capacity or jurisdiction is not accepted, then these background questions will themselves become legal issues). Cases will not arrive in courts by themselves, but only as a result of originating instruments, discovery, etc. The substantive issues must themselves be identified by reference to statutes, precedents, etc. These communications will typically have a normative form, and the legal information which they generate will have normative meanings. Judges are called upon to adjudicate, which, in our system, requires them to identify who is entitled to decisions in their favour. This in turn requires judges to identify which legal norms are available to them, and how they direct them to decide the legal issues before them. Our legal system requires judges, in nearly all instances, to give reasons for their decisions. Whilst this can be justified by reference to various values, an important consequence of these kinds of communications from a systems theory perspective is that they facilitate the evolution of the legal system. Without reasons being given in the past, judges would not have the enormity of resources (redundancy) available to convey information now. And with these reasons in turn, further resources are created for further legal communications. Reasons are also important for appeals, which in turn generate resources for further legal communications.
At the moment of decisions, judges can make a variety of determinations, but they do so against a background of considerable redundancy.Footnote 18 And they are not alone in this. For the parties before them to have reached courts with legal issues there will be an enormous number of further potential legal issues which either have not been raised, or have been resolved. In this sense, the issues which judges have to decide have been framed by the communications which lead to those points, and the store of redundancy available to judges, at those points, to make successful legal communications. As human actors participating within society, many different kinds of communication are open to judges. They can declare their beliefs that the legal issues in the cases before them ought to be decided in particular ways not suggested by the law, or even declare their agnosticism. They can similarly communicate their opinions as to the legitimacy or otherwise of the legal system as a whole, and any reforms they may wish to recommend. But for judges undertaking the legal operations of adjudication, the task facing them is to make legal communications. They have to reach decisions that are recognisable as decisions, within their legal systems. And if they want to make decisions that are likely actually to dispose of the issues before them, they have to make decisions that are either unlikely to be appealed, or likely to be upheld on appeal. As we have just described, the situation in which they find themselves to be judges is created through a vast number of legal communications. It is a context created by the system of communication in which they operate, through conducting trials/appeals and making decisions to participate in it or not. As human actors they can always fail to communicate, either wilfully or not. But if they are going to communicate, then the legal information that they are able to communicate is going to be a function of the redundancy that provides a background to their communications.
Assuming that our judges continue to participate in their legal systems, what choices are open to them? Can they decide that the normative form of legal communications, expressed, for example, in terms of legal rights and duties, is inappropriate, and adopt a different ‘detached’ form of communication? How can they express their authority to decide the matters before them as judges, without accepting the normative communications that have put them into the position of being judges, establish the previous proceedings as trials or appeals, and frame the matters they have been asked to decide as legal issues? This is not to deny that these legal communications could not be rephrased in a different form, or that these judges, within different systems of communication (media, education, politics) could not do this themselves.Footnote 19 But, at the particular moments which require decisions that can be recognised as legal decisions, none of these options are open to them. It is not just that judges might have sworn oaths, or that they might feel that it is prudent or right to do what others expect of them. The expectations of others, expressed in terms of redundancy, lie in the fact that the redundancy which judges can utilise to generate the legal information which they wish to put into communications is the same redundancy open to their legal audience in identifying what, within their communications, represents legal information.Footnote 20 Indeed, it will include the same redundancy that enables judges to understand why they are judges and not ushers, why their courts are courts and not theatres, and what the parties have argued before them and what it is that they are called upon to decide.
How should one describe judges’ positions: the fact that the redundancy available for them to make communications that are likely to dispose of their cases is the same body of rules, principles, etc. that will be used by their intended audience to decode their communications and identify the information within them? One can, if one likes, call this a ‘practice theory of rules’,Footnote 21 but the practice in question is communication within a system of communication. There is no practice outside these communications. And the fact that the ‘practice’ would have no normative force if it were not followed by others has no separate existence or manifestation apart from its existence within these communications. If others do not utilise normative communications in a particular way, then you cannot utilise those communications in order successfully to communicate normative meanings. If they do, you can. Of course, one can also talk about legal standards being adopted as guides to appropriate behaviour, but with judges the behaviour in question is the making of communications.Footnote 22 Therefore, there is actually no need for judges to make statements confirming that the rules which they are proposing to adopt are being followed by other judges. If the parties which these judges’ communications are directed towards were not constructing their communications using the same redundancy, then these judges would not have the ability to make successful communications. To phrase this in the terms of one of Dworkin’s arguments, judges operating within the legal system cannot refer to the standards which they apply as having their validity because of their common application by other judges. They must communicate as committed vegetarians would talk about their own decisions not to eat meat. On these communication menus, there are no non-vegetarian options. Judges cannot sensibly communicate that legal standards are not appropriate guides to their behaviour. Since the behaviour in question is legal communication, the only way in which that behaviour can occur is through the application of these standards to their behaviour. If judges wish to make legal communications (i.e. to continue as judges making their decisions) they will have to apply existing legal standards to their own behaviour. They have no other resources open to them.
This is not a claim that judges applying the law always approve of what they are communicating. Nor is it a claim that they will not, outside actual adjudication, criticise the decisions that they and other judges might have taken in the course of litigation, or criticise the character of the law in some of its particular branches, or even question the legitimacy of the whole legal system. Still less is this a claim that judges cannot make legal communications in terms of legal rights and duties whilst believing, i.e. thinking, that the standards that they are applying are actually completely inappropriate. The beliefs of judges, except to the extent that these are manifested by communications, remain unknown to anybody but themselves. But what we can say is that such critical options are not open to them as legal communications at the moment of their legal decisions. Then what, you might say, of judges who make legal decisions and insert a note of reluctance regarding what the law requires them to do?Footnote 23 Are they not expressing a different, less committed attitude towards the law? And is that not also a legal communication? A bold answer here would be ‘no’. What legal meaning is communicated by such statements? They do not in any way alter these judges’ communications of what they find the law to be. Indeed, a divergence between what they find the law to be and what they would like it to be are exactly what they are communicating by their expressions of reluctance. Nor can such statements represent a refutation that the law as found represents an appropriate guide to these judges’ behaviour. Again, as the behaviours in question are the decisions which have been communicated, it would be operationally self-refuting for judges to claim that the considerations that led to their expressions of reluctance actually outweighed their reasons for their decisions.
If the judiciary, when giving judgments, are unable to escape from utilising communications that have a normative form, and an implicit meaning that the authors are committed to the law as a standard for action, what can we conclude, from the mere fact of those communications, about judicial beliefs as to the legitimacy of the legal system? The correct answer is nothing. And that answer requires us to reconsider much of the legal theory that has been devoted to claims about the internal commitments that are required from legal actors in order for a legal system to exist. Starting with Hart, there has been a long-standing debate on the ‘internal attitude’ that must be evidenced by public officials, and in particular the judiciary, for a legal system to exist. If the premise of this debate has been that judges, when acting as such, giving judgments, could exhibit (which means communicate) any other attitude, then that premise is simply false. And in replying to various predictive theories of law, from Austin to the American Realists, by pointing to their failure to acknowledge the forms of communication which he gives as examples of an internal attitude, Hart was doing no more than taking particular legal operations (such as the making of legal arguments in courts, or the giving of legal judgments) and pointing to the impossibility of such operations being achieved by different forms of communications than those actually employed. Judges communicate through communications which have a normative form: rules and other standards.Footnote 24
If Hart’s major contribution to legal theory was an important recognition of something that is inescapable (judicial actors have to use rules, and the meaning of judicial communications that take the form of the application and identification of rules have an implicit meaning that the actor regards those rules as appropriate standards for the resolution of the disputes before them), what follows from this? Well, for a start, it makes it more difficult to talk sensibly about a judicial commitment to the legal system, if this is taken to assume the possibilities of judicial non-commitment to the legal system. By acting as judges, which means by making legal communications within a role and situation constructed by other legal communications, judges daily make statements whose implicit meaning is that they accept the legal system as a normative order. Indeed, they cannot carry out their jobs, making judicial decisions, without expressing such commitments as they go about these tasks. We can reason that if judges found that their daily legal communications caused them intolerable psychic discomfort, they could be expected to resign, and find other occupations, so that every judge can be presumed not to be in such a state. But above this threshold, it is difficult to draw conclusions from legal communications about the inner motivations and feelings of judges (i.e. such motivations and feelings cannot simply be imputed to them as part of the process of decoding their legal communications). Indeed, it is even difficult to agree how to describe these internal feelings. At a common-sense level, it seems to make sense to claim that judges who mentally regret the operations carried out through their communications are less committed than those who do not. And if they take the opportunities, within other systems, such as the media or politics, to express that regret, we might be inclined to describe them as less committed to the legal system than judicial actors who do not make such extra-legal communications. But is it more accurate to describe such judicial actors as actually more ‘committed’ or less committed to the legal system than their more silent colleagues? Are actors who can demonstrate that they have ‘overcome’ their political or moral reservations and still given effect to the law less committed to the legal system than colleagues who, as far as we know, have experienced no such reservations about the moral or political implications of parts of their judicial role? And whilst such conflicts of commitment, or their expression outside the legal system, may have their casual effects, they will not prevent the legal system from continuing; for what level of commitment does the legal system need, in order to continue, except the willingness of its actors to continue communicating?
Our analysis here is not completely different from what has gone before, particularly the work of Neil MacCormick. He acknowledged, in Legal Reasoning and Legal Theory (Reference Maccormick1978), that judicial actors can find themselves, at least on some occasions, in situations where the legal communications available to them are severely constrained, appearing to require one solution as a conclusion from deductive reasoning. And that once such a situation has occurred, whilst the next operation (an order in favour of the ‘winning’ party) does not follow as a matter of logic, but only if the actor chooses to make the requisite communication, ‘Given the institutional pressures within the legal system – the opinion of the profession, the possibilities of appeal, etc., and given the external pressures of adverse press publicity and Parliamentary comment and the like, it would be so strange as to be barely imaginable that a judge having established the justifiability of one decision by logical argument from sound legal premises and findings of fact should then issue some diametrically different order’.Footnote 25 But systems theory invites us to take the implications of such examples much further. For a start, if the ability of a judge to fail to give appropriate effect to an easy case is ‘so strange as to be barely imaginable’, how much more strange is a failure to adopt normative forms of communication when undertaking a judicial role, at any point. And what makes a communication ‘so strange as to be barely imaginable’? MacCormick includes the actor’s awareness of the communications of systems in which s/he is not participating, and may never participate: the media and politics. But the starting point for identifying what is ‘strange’ in a judicial communication, and one that operates regardless of the judge’s awareness of the communications of other systems, or the likelihood that other systems will produce any critical communications, is the same redundancy that generates all legal communications at those points in the legal system. If judges make communications that the legal system identifies as ‘so strange as to be barely imaginable’, they will be identified, within the legal system, as errors.Footnote 26 And an awareness of what constitutes error, in this sense, will have guided actors’ selections of communications at every stage in the processes that have led to these particular decisions.Footnote 27
MacCormick has also anticipated some of this analysis in his attempts to refine Hart’s distinction between internal and external attitudes. He has argued that the internal attitude comprises a cognitive element (understanding the presence and nature of rules) and a volitional one (wishing those rules to be applied). With this further distinction, MacCormick has attempted in turn to identify three positions open to those engaging with social rules: the insider who accepts rules by wanting them to be applicable to their behaviour and those of others (whether enthusiastically or simply to avoid the disadvantages of disobedience by others); the moderate outsider (the ‘hermeneutic approach’) who understands the meaning of the rules to insiders but has no desire for her/himself or others to comply with those rules; and the extreme external observer who can identify regularities of behaviour but has no concept that these regularities are related to the presence of a rule.Footnote 28 We believe that our analysis, using Luhmann, can take this further. First, the rigour of Luhmann’s separation between systems of communication and psychic systems avoids the slippage from analysing statements which have an implicit point of view to ascribing that point of view as a necessary (internal) fact for the making of those statements.Footnote 29 And second, MacCormick fails to analyse the conditions that establish when these different kinds of communication can be used. So, for example, he asserts that the legal scientist and social theorist can make statements that do not purport to reveal anything about their own commitments, critical morality or group membership without considering the difficulties facing judges, who can’t make statements that do not purport to reveal commitment. MacCormick, following in the Hartian tradition, seeks to analyse language, and categorise the kinds of statements that can be made, and the actions that they can perform. Luhmann, through his theory of systems, requires us to consider how the context of communications is established through communications, and how that context (redundancy) structures the possibilities of what can be communicated at that point. Using Luhmann to go beyond MacCormick, we have to examine how redundancies within systems restrict the choice of forms of communication available to the actors who participate in those systems.
A further implication of this analysis is that one needs to question the use within legal theory (and social theory) of the concept of conventions.Footnote 30 The judicial contribution to the existence of law as a system has been described in terms of judicial conventions about the use of appropriate sources of law, the correct manner of speech, etc. These ‘conventions’ point to something outside the system itself, as the mechanism for the legal system’s ability to maintain its autonomy and evolution. If conventions mean widespread acceptance, common positive commitments internal to the relevant actors (or common perceptions of criticism from other systems), then one can sensibly talk about the legal system changing when its conventions change. A shift in conventions allows for a change in what can be communicated. For example, changing commitment to parliamentary sovereignty, and an increasing acceptance of European integration, might explain developments in the treatment of sources of UK law. By contrast, redundancy would invite one to consider, at each moment in the system, what communications were available to the actors in question. These communications are the conventions. There is nothing outside them, which limits the possibilities of communication. And the possibilities of communication (the conventions) alter as those communications occur, in that those communications can generate new redundancies and new variety. To ask about the strength of a convention, using systems theory, is to identify what operations cannot occur at present, without utilising the particular communications in question. The cost of abandoning conventions is the cost of abandoning particular kinds of redundancy, which is the need to find other ways to carry out the same or similar operations using different communications, and the risk, in any attempt to do so, of failure.
Following on from this, there is also a tautology in Hart’s claim that the judiciary must accept a rule of recognition as a common basis for identifying what counts as a rule of the legal system. Without here entering into the extensive controversy of exactly what counts as the rule of recognition (rules, principles, standards, etc.), the need for a judiciary to utilise a common basis for what is capable of identifying a legal communication is simply a re-statement of the basis for communication: redundancy. Judges cannot communicate with each other without utilising the redundancy that is available to them.
In keeping with Luhmann’s theory, we must also advance our arguments on the basis that systems of communication need to be identified separately from institutions. Thus judges, giving legal judgments, make legal communications. But senior judges, for example, giving public lecturesFootnote 31 or writing law review articles are not making legal communications.Footnote 32 And it is their different ability to make communications as they participate in different systems of communications that supports the claim that the beliefs of judges differ from what, within their judgments, they say.Footnote 33 In addition, although the focus of this article is on judicial communications, it might assist our understanding of the limited options open to adjudicating judges to consider the communication options open to other participants within the legal system – lawyers and laypeople.
Whilst the ability of laypeople to make legal communications, generating legal information, is more limited than that of judges deciding cases, moving through our society without making legal communications is virtually impossible. What layperson is going to fail to claim under contracts, or assert property claims, let alone fail to express the legal position when their car is obstructed by illegal parking, or they are assaulted? Whilst these events may be acknowledged as harms or wrongs, they will also be recognised as involving legal rights and crimes. Which individual is not going to communicate that they have rights as a tenant, or owner-occupier, or have the ability to leave their property to people on their death via their wills? Imagining an individual who never utilised legal communications, even in order to construct a position from which they could make other kinds of claims within other systems (‘leaving all your property to my sister is wrong’),Footnote 34 is very hard. Even tramps are likely to communicate thefts and assaults, let alone recognise the signs that indicate when they are likely to commit a trespass. And again, this leaves those individuals in a position that can be described as operationally self-refuting. What does it signify when someone makes an appeal to the law? By making this form of communication they are not communicating that an event is uniquely illegal. They are drawing on a redundancy that overlaps with much of what is utilised by judges and lawyers. Laypeople do not expect that everything that they would like to be the case is the law. They know that legal rights derive from gifts, contracts, transfers, and that law is made in parliaments, courts and town halls. They also understand such constructs as rights, i.e. not just predictions of the likelihood that harms or benefits will happen in particular circumstances. By participating in the legal system through making legal communications, these laypeople will exhibit Hart’s internal attitude. But what they will not exhibit when making such communications is the state of their beliefs, whether these represent approval or disapproval of the standard to which they are referring, or approval or disapproval of the legal system as a whole.
Moving on to lawyers, when individuals seek legal advice, the communications that result draw on a redundancy that overlaps with that used by judges, and generates legal information of a similar complexity. And, as with laypeople, some of what is then communicated will take the form of a prediction of what might occur, rather than a statement which simply has a normative meaning. The so-called ‘bad man’ approach to legal advice tells clients not what the law requires, but only what is likely to befall them from the activities of legal authorities if they proceed with particular courses of action. Legal advice, unlike judicial decisions, can take this form. Judges cannot express their own decision as a prediction of what others will do. They express their decisions as a conclusion on the law, even if they concede that another court might take a different view. By contrast, lawyers can express vicarious opinions, in terms of what the courts are likely to decide. And they can further distance themselves from judicial communications by expressing views on the likelihood of judicial outcomes in terms of predictions of both decisions and the likely consequences that will follow from them. But even here one needs to recognise the limited ability of lawyers, when giving advice, to abandon the normative form of legal communications. Indeed, if lawyers deconstructed every legal communication that they utilised in order to reconstruct legal advice into predictions, it would be impossible for that advice to be given. What lawyers in fact do is engage with legal communications in their classic form, in order to construct narrow issues, and only at that point reconstruct communications that have a normative form into predictions of consequences that do not. Having given advice in that form, they then have to re-engage with normative communications in order, like judges, to generate the legal arguments that will need to be presented to the court. These arguments will not duplicate the predictive elements found in the advice, not simply because lawyers can be expected to be partisan, but also because such communications will not be recognised by judges as legal arguments. They will not offer them reasons that can form part of their legal judgments. Thus it is not that detached, vicarious normative communications cannot form any part of the legal system. It is rather that the legal system establishes where and when such communications can form part of itself by reference to the information that needs to be generated in order to carry out particular legal operations.
What we can conclude from the above discussions is that whilst there are detached and committed participants (whether these are judges, lawyers or laypeople), if that means that some participants in the legal system experience internal approval when making legal communications, and others wish that what was available to be communicated was different, this does not result in detached and committed communications.Footnote 35 The legal system, especially in connection with adjudication, is not made up of committed and detached communications. Adjudication can only be carried out through committed communications. Detached communications would either fail to carry out the adjudication, or would be identified by the system as reasons why the decision should be regarded as an improper one. Successfully carrying out adjudication involves only committed communications, which have normative meanings, because these provide the background redundancy which enables one to generate legal information – a legal decision justified by legal reasons.Footnote 36
A lot of jurisprudential discussion of the internal attitude, including that by Hart himself, addresses the question of what motivates individuals to make legal communications, i.e. what internal commitment or desire is necessarily coupled with the kinds of expressions (calls to comply, criticisms for deviation, etc.) that constitute the manifestation of the internal attitude. Arguments commonly take the form of identifying what is either necessary or sufficient for those communications to generate the meanings (information) that they do. Hart himself was quite an agnostic in regard to this question: ‘both this general obedience and the further use of and attitudes to the law may be motivated by fear, inertia, admiration of tradition, or long-sighted calculation of selfish interests as well as by recognition of moral obligation. As long as the general complex practice is there, this is enough to answer affirmatively the inquiry whether a legal system exists. The question of what motivates the practice, though important, is an independent inquiry.’Footnote 37 In our opinion this agnosticism is appropriate, as Hart is correct to assert that the question is not what motivates individuals to make legal communications, but what they communicate when they are so motivated. In making legal communications individuals communicate legal information, and the information that they communicate does not alter its form (i.e. cease to be normative) because of the motivation that has led them to make the communication in question. Of course, if they communicate their beliefs at the moment of making such communications, then this may deprive the subsequent communications of their normative form and meaning. But, first, this will not be a result of the existence of motivations as beliefs or commitments, but because of their manifestation and existence as communications. And second, such manifestations raise the issue which we have been discussing: are the communications which represent manifestations of belief compatible with, or do they add anything to, the legal communications that the individuals in question are motivated to make? In some circumstances, communicating a commitment or motivation at the same time as making a legal communication will undermine the ability to make the communication one is motivated to make.Footnote 38
If detached forms of legal communication are not available to judicial actors participating in the legal system, what other restraints might be described usefully in terms of a system’s redundancy? The other example we wish to explore here is the judicial failure to acknowledge that they are making law, at the moment of its making, in their judgments.
4. Judicial discretion
Systems theory does not regard human beings as simply robots, participating in systems of communication without exercising elements of choice. One choice is not to participate, although, as we have indicated, never participating in the legal system is extremely difficult. Other choices go to the making of communications. This is not a free choice, as we have tried to show in our previous discussion. There are lots of communications which, whilst they may bear a closer resemblance to the beliefs and commitments which motivate individuals to make legal communications, cannot form part of what is communicated without preventing the desired communications from being made as effective communications. But choices remain. Lawyers choose what arguments to make in order to produce the decision which favours their clients. And the choices here will not only go to the substantive content of the arguments, but the manner of their presentation. And, likewise, judges will have to exercise choice in the construction of their judgments. All this is fairly uncontroversial. But what remains a major issue in jurisprudence is the extent to which judges exercise choice in establishing what the law is. And a major part of this controversy, as illustrated by the three quotations which introduced this article, is the manner in which we should understand the judicial failure to admit whilst adjudicating that their decisions are, on at least some occasions, the result of individual choices rather than the application of pre-existing standards.
Before we proceed to explore the judiciary’s failure to admit changing or making law in their judgments, we need to make a few remarks on the nature of this phenomenon. There is plenty of evidence that judges change law. Not only were judicial decisions the primary source of English law from the twelfth century to the second half of the nineteenth century, but they were also the means by which that law altered from its original customary content. And, whilst legislation has taken over as the major mechanism by which law is both established and changed, the amendment of law through judicial decision has continued. Judges, in articles, books and public speeches, have felt no need to deny the alterations in the law which their brethren, or even themselves, have affected to the law. One can even find examples of decisions in which the subject of discussion, and the issue to be decided, is how to respond to a change in the law. Thus, for example, in R v. Cottrell Footnote 39 we are told: ‘It is artificial to pretend that the law was not changed, or to dress its impact in the jurisprudential disguise that the law had, in Blackstone’s word, been “discovered”.’ Cottrell was a Court of Appeal decision on the right of defendants to appeal against their convictions on the basis of a change in the law, which had to deal with the consequences of the House of Lords decision in R v. J Footnote 40 to alter fifty years of settled law on the right to prosecute for indecent assault where the time limit for unlawful sexual intercourse with a child under sixteen years of age had passed. Whilst the Court of Appeal in Cottrell dismisses the artificiality of a pretence that law is ‘discovered’, the House of Lords judgments in R v. J contain no awareness, either in the majority judgments, or Baroness Hale’s strongly critical dissent, that any of their Lordships are involved in an exercise in changing the law.Footnote 41
Whilst re Cottrell contains an admission that it is ‘pretence’ for judges to deny changing the law, R v. J is an example of the communications that generate such criticisms.Footnote 42 Their Lordships presented judgments that explored what the law was, rather than admitting that their decision would establish or change the law. They also failed to acknowledge that a hard case, which could generate divergent views of the law, must necessarily involve the exercise of some choice or discretion. And lastly, in the exercise of that choice, there was no recognition that factors such as their gender or biography (and in other hard cases their race and class) could operate as factors influencing their judgments. How do we explain such, to quote Hart, ‘rhetoric[al]’ omissions? Hart also talks about ‘ritual’ language, and the quote from Simpson at the start of this article, speaks of ‘conventions of legal argument [which] embody a belief in the theoretical possibility of a comprehensive gapless rule of law’. What is the nature of this ritual or convention?
One can offer various normative or functional explanations of the judicial failure to acknowledge what they do (establish or change law) when they do it.Footnote 43 As with MacCormick’s explanation of the judicial duty to give effect to clear law, one can attribute this to various sociological or political factors: a desire to increase their own, and their profession’s, social standing; or the need for unelected officials to avoid political criticism. And one might also hope, perhaps through academic critique, to expose this element of law-making, and all the unacknowledged elements that contribute to judicial decisions, and thereby help to produce different outcomes in the future. In keeping with Luhmann’s theory, and the forms of autonomy which his theory recognises, we prefer to begin with an exploration of the operations which are carried out by legal decisions, and to consider how these would be made more difficult, or impossible, for judges, if they did not utilise the redundancies currently available within the legal system. In so doing, we hope to identify the costs to such actors of ceasing to utilise that redundancy, and to consider how such costs, which operate at the moment of each communication, might structure the participation of judicial actors within the legal system to a greater and more predictable extent than theories that rely on those actors’ awareness of, and response to, the anticipated consequences of their actions in terms of gender, class, race or professional interests.Footnote 44
Let us begin with those explanations of judicial communications which seek to demonstrate their self-serving qualities. A convention amongst judges to hide their law-making function at the moment of legal decisions is difficult to reconcile with a willingness of judges, in law review articles and public speeches, and even as in R v. Cottrell in some decisions, to admit that their decisions make, and don’t simply identify, the law. And it is their willingness to admit that they make law which provides evidence that at least some of their communications have a disingenuous quality that can be described as ‘rhetoric’. Why hide in one moment what can be revealed, in public forums, at others? Of course, these public admissions are usually accompanied by claims that law-making by the judiciary is of a different quality from that of the legislature, that it is more incremental, or interstitial. Some will see rhetoric not only in the failure to admit to making law when doing so, but also in these further claims that judicial law-making has a limited character. For whether one focuses on acknowledged ‘landmark’ decisions, or simply seeks to expose the alternatives that are available in more ordinary cases, it is easy to show that the changes in law affected by the judiciary are, on at least some occasions, as politically significant as much of what is carried out through legislation. So the convention, interpreted from this perspective, is to admit to the fact of law-making (but not at the time when it occurs), and to disguise its extent and the political nature of the choices which accompany it. By what mechanism is such a fine balanced self-serving convention maintained in place against the vagaries and judgments of individual judges, who might be expected to hold different views as to the likely consequences, whether in the media or politics, of its infringement? What would be a systems theory explanation?
We cannot overcome the separation between psychic and social communication and know what is in the minds of judges. But we can begin this enquiry into the restrictions on the possibilities of communication open to judges by asking what restraints hypothetical honest judges might encounter (our version of Dworkin’s Hercules!).Footnote 45 Honesty is subjective, which creates a situation in which the more learned our imagined judges, and the more self-aware they are, the more difficulties they may have in articulating all of the factors within their consciousnesses (or even their unconscious motives) which have influenced their decisions. Think of the enormous difficulties. Within moral philosophy there is an extensive debate between cognitivism and non-cognitivism; this debate raises the possibility that it is not values but desires that motivate human choices, with reason operating only as a mechanism to increase our ability to achieve desires. Closely allied to this are debates about ethical relativism, which raise the question of whether values, or desires, are the result of culture, class, race or gender. And for those who believe that the acquisition of values or desires is more individualistic, one might look to psychology or biography. So honesty about the nature of choice, if this means identifying the reasons why judge X chose to make decision Y, could take us through moral philosophy into anthropology, sociology and psychology, with little or no certainty of a definite conclusion. This leaves our honest judges in a difficult position, and the more self-conscious and widely read they are, the more difficult that position could be. The ‘nightmare’ level of honesty would be judges who adopted the style of some post-modern writers, attempting to deconstruct their decisions at the moment of deciding, by attempting to identify, with total honesty, what has brought them to believe that X rather than Y is the appropriate outcome for the case at hand.
We can dispose of this ‘nightmare’ scenario by reflecting on what is involved in ‘making law’ within our system of precedent. Higher courts do not simply announce their decisions. They give reasons. And the reasons provided will be intended to persuade others (judges, lawyers and sometimes even politicians and laypeople) that their decisions are appropriate. This is a significant constraint. It will take us well away from our ‘nightmare’ scenario. Rarely will judges’ self-conscious observations on the basis for their moral values persuade other judges of the correctness of their decisions, whether those decisions are presented as a conclusion of law or a deliberate attempt to make new law. If judgments are intended to persuade, then the communications chosen to construct them must be appropriate to that task. But what is it that makes particular kinds of reasoning appropriate to that task? We can call the standards of legal rhetoric a convention, and just accept that ‘they are because they are’ amongst sufficient members of the judiciary. But systems theory invites us to go further, and examine what operations these reasons facilitate that would not occur, or not occur in the same way, if different reasons were given.
First, we need to remember that the giving of reasons by judges operates within a system of precedent. As such, the reasons provided have to guide the making of further decisions. This provides a filter against all kinds of ‘honesty’. Confessions of an idiosyncratic kind that reveal why particular judges reached particular decisions make no contribution to this process. In effect, as precedents, they leave matters undecided, as there is no reason for persons with different histories/gender/class to reach the same decisions. In order for decisions to operate as precedents, judges have to remove themselves from their judgments, whatever their beliefs as to the manner in which their own personal experiences might have shaped their decisions. This is the beginning of a process that can, from outside the legal system, be described negatively as ‘dishonesty’, ‘hypocrisy’ or ‘rhetoric’, or positively as ‘impartiality’ and ‘objectivity’.
Second, as pointed out by Hart, when judges are called upon to choose between two legal solutions to a legal issue, part of their task is to integrate their chosen solution into the existing law.Footnote 46 Again, this is part of the system of precedent. Law evolves by each new law fitting into the existing law. We might call this an examination of the ‘legal consequences’. It involves a consideration of the implications for existing law if the chosen solution is the law. One solution may lead to one configuration of legal relationships, whilst the other leads to a different one. In making this comparison, judges are involved in a contemplation of redundancy: ‘what possible differences to future legal communications follow if these matters are decided one way rather than the other?’ And from the perspectives created by these alternative configurations of hypothetical future legal communications, judges have a limited ability to view the implications for the rest of society. To some extent this description of alternatives involves an element of persuasion, showing how judges have chosen the better outcome. But restricting this persuasive element is a communication that is a necessary part of the evolution of law: what might it mean to establish this precedent?
A third factor in this assessment of judicial communications is the nature of the procedures through which judges establish the existing law. When identifying settled law, judges make communications that do not differ in form from those used to establish new law. (Indeed, that is what leads to accusations of ‘rhetoric’.) In other words, the processes which are necessary in order to choose between alternatives are not different from the procedure involved in identifying what alternatives are available to be chosen, including those cases where a judge might conclude that there are no alternatives, i.e. that the matter is disposed of by settled law. In a case of settled law we might expect literal meanings and purposeful interpretation to coincide, principles to support this, minimal disturbance to other areas of law from adopting this interpretation, and whilst the legal result might not be able to be presented as reasonable (in the sense that it is beyond criticism) at least it would not appear wholly unreasonable (in the sense that it is indefensible). But if this is what constitutes settled law, then the difference between finding law and making law is not a difference of kind, it is a movement along a continuum. Elements of what constitute settled law may find themselves located on each side of a choice between offered alternatives. This means in turn that there is no clear demarcation between areas of law that are settled and those which are not. What then would it mean for judges to exhibit honesty when they move along this continuum? One envisages a situation in which judges who feel that they are involved in the making of law should offer one of two kinds of admission: either that there is no settled law, and so the decision represents an addition to law, or that there is settled law, but the judges in question regard the existing law as so unsatisfactory that they are going to have to change it.
What are the problems with such additional elements of honesty? To put this in systems theory terms, how would the adoption of such communications change the nature of legal communications and the operations that they carry out? Would adjudication continue as adjudication with this small change? Or would the introduction of such ‘honest’ communications block or impede the making of communications and the operations they carry out (which participants are currently motivated to achieve). And if these impediments to future communications represent the costs of such admissions, what exactly are the gains (remembering that the participants in the process have to be motivated to make these new forms of communication)?
There are penalties for this kind of honesty. Judges who are not self-conscious in their judgments as to what represents settled law will have a distinct advantage over those who are. These judges will simply not have to make such honest admissions, or at least not make them at the same point on the continuum as their less arrogant colleagues. Of course if the admission that one was making law had no adverse implications for our judges (as is the case for example when statutes expressly give discretion to individual judges), then there might be no immediate disadvantages from making such confessions. But our judges have to decide on their communications by reference to the legal system as it currently is, and not as they might wish it to be. And given where we presently are, judges who admit to making new law other than as authorised by statute or precedent are at a disadvantage in deciding the cases before them, and judges who go so far as to admit to overruling settled law because they do not regard it as appropriate (‘the law is currently X, but we think it ought to be Y’) are at an even greater disadvantage.Footnote 47 The most obvious disadvantage is not the external one of political, media or academic criticism, but the internal one of failing to achieve a successful legal operation: to actually dispose of the cases before them. Such statements are, within the legal system as it currently exists, going to communicate information that is decoded as an ‘error’.Footnote 48 In the case of lower court judges, it is going to produce a version of the easy case: the easy appeal. In the case of a supreme court, such communications are not open to correction by appeal. But one should not assume that this makes a significant difference. The techniques of legal argumentation that identify errors in a lower court judgment (a contemplation of legal consequences that includes a sense of what can no longer be decided if this communication were accepted by one’s audience) apply to lower and higher court judges alike. Whilst the error would be less easy to remedy, contemplation of what it represents to the legal system (in terms of a loss of redundancy) is still present. One example of this is the House of Lords’ unwillingness to abandon its practice of regarding its own decisions as binding precedents, evidenced by the delay in introducing a practice direction allowing overruling, and the limited number of times it has been used since established in 1966. There are huge costs in terms of a lack of redundancy if a supreme court makes communications which acknowledge its willingness to change law solely on the basis that it is no longer satisfactory. And the practice direction only deals with the ability of the House of Lords to change its mind about what the law has always required, i.e. to find the law to be different, not to openly make different law. This further move, especially if it involves an abandonment of the techniques of statutory interpretation, would amount to communications that were truly, to again quote MacCormick, ‘so strange as to be barely imaginable’.Footnote 49
If honesty is not to be its own reward, then it will in practice be discouraged. Judges will not be prompted to make admissions that will almost certainly undermine their attempts to make the law into what they might like it to be, since honesty will tend to work to prevent their desired outcomes from occurring. However, could there be an institutional change to achieve this result – to remove the penalty for honesty and the reward for ‘hypocrisy’? Well, this would require a change in the legal system. What communications within the legal system could alter it to allow this to happen? We cannot say that such a change could never occur – systems theory can never predict how a system might change itself over the long term. But we can see why, at present, such a change would be quite difficult. The most flexible instrument for changing the legal system that we currently possess is legislation. But this gives us a problem. If judges were told in legislation that they could make new law whenever they felt that the law was not settled, or even that they could make new law whenever it was settled but they disagreed with it, then ‘rhetoric’ or ‘hypocrisy’ would cease. But this would be because the range of discretion that has already been expressly handed to judges in areas such as sentencing and matrimonial settlements would have been extended to the whole of the law. And not only is such a mandate unlikely to be enacted, it would not actually do what those who want an end to hypocrisy are asking for. They want judges to be honest, i.e. to admit that they make new law despite the absence of a statutory authority to do so. They are not asking for judges to be given such monumental powers that the need to be ‘hypocritical’ will never arise.
Assuming that the legislature is not going to grant judges an express general power to alter the law, we find ourselves in one of the situations of inconsistency, if not paradox, that systems theory leads us to expect to find within all systems of communication.Footnote 50 Whereas the legislature cannot or would not give judges a blanket power to make new law, nevertheless it needs judges to carry on with the task of applying the legislature’s law. And it also needs judges to interpret and alter their interpretations of law to take account, to quote Hart, of ‘our relative ignorance of fact. . .[and]. . .our relative indeterminacy of aim’.Footnote 51 And with this last ‘need’ we can say that, in effect, the legislature needs ‘hypocrisy’. To be effective, the legislature needs the judges to carry out operations that it cannot expressly authorise. Judges must make law, in the sense of dealing with the inevitable contingency within a system that cannot provide in advance for all situations, by using communications that do not admit that law, in this sense, is being made. These particular examples of ‘hypocritical’ communication facilitate a workable version of what is commonly referred to, within the legal and political systems, as the doctrine of ‘separation of powers’. Whilst the need for judges to make law is inescapable, their inability to proclaim that this is what they are doing, as they do it, leads to different forms of law-making than would occur if judges could utilise communications that confessed to their law-making role as they were doing so.Footnote 52 Whilst there is undoubtedly a creative element in making new law through the development of precedent, it is still a continuation of the techniques involved in ‘finding’ settled law; most of the redundancy remains the same. And as such, it does not provide the same opportunities as would be present if a judge could simply say: ‘at this point I have run out of law, so the best law I can create in this situation is. . .’ (let alone say ‘here I find the law so inappropriate that I am going to change it to. . .’).Footnote 53 And this form of communication (their inability to proclaim that they are making law as they make it in the course of their judgments), rather than any personal political commitment held by one judge, or by judges in general, results in law-making that can be described as ‘incremental’. By contrast, if judges could communicate ‘honestly’, and admit to making law within their judgments whenever they felt that their decisions were not dictated to them by prior standards (or felt that those prior standards were inappropriate), then the content of the doctrine of separation of powers would depend on different communications, namely those expressing each judge’s assessment of the deference appropriate to Parliament.
So, the conventions which allow discretion whilst failing to acknowledge it are not something which judicial actors have to accept in the sense of a consensus within their consciousnesses about the practices that they must commonly perform, or the political or professional reasons why such practices might be a good thing. They are first and foremost the experience of carrying out legal operations (generating information) against a background of redundancy, which enables routine decision-making to occur. The individual actor who attempted to break with the ‘convention’ would be unlikely to find that their ‘novel’ communication was absorbed within the legal system rather than being recorded as an error, still less that it could generate a new redundancy for the making of routine legal decisions. There is no obvious way (at least in the present) for the judiciary to decide both easy and hard cases on a routine basis except by using communications that fail to distinguish clearly (from the perspective of at least some academics) between the finding of law and the making of law. However much their power may resemble that of Parliament (in terms of the significance of the legal changes produced by their operations), the communications which they use have to take a different form, one that fails to recognise that law is ‘made’ by judges who make decisions, and takes the form of a discovery of the existing law.
5. Conclusion
This article has sought to indicate the reasons why judges, and other actors within the legal system, or indeed any system, might need to utilise communications that do not reflect their own (or perhaps even any human’s) understanding of what is occurring.Footnote 54 Attempting to explore why this might be a necessary part of any system’s operations, and then identifying the operations which this makes possible, provides us with a radically different insight into the nature of judicial communications. Debates which are informed by ordinary language philosophy, speech act theory and other forms of linguistic philosophy, and focus on language and not systems, fail to identify the constraints placed by systems on actors’ use of language. Arguments about the construction of meaning which limit themselves to language and its use, but ignore the restraints which arise from law’s existence as a separate system of communication (with its own redundancies), have produced a series of irresolvable disputes about what truly occurs within legal systems. Systems theory offers legal theorists a new direction, with new insights. This is a journey which has implications for more schools of jurisprudence than analytical positivism. To call communications ‘rhetorical’ or ‘hypocritical’ draws attention to something important. But without analysis of why such communications occur, one is left uncertain as to the possibilities for things to be other than they are. Those who see hypocrisy as a personal failing might hope for more ‘honesty’. Those who see ‘hypocrisy’ as a political strategy, intended by its users to maintain the legitimacy of the legal system in the face of possible criticism, will expect changes in the political or legal spheres that diminish the need for deference to lead to immediate changes to the frequency and nature of ‘hypocritical’ communications. Systems theory, by focusing on the operations that communications make possible, identifying the need for new communications to replace former redundancies if the possibilities of communication (variety) are to alter, and acknowledging the evolution of systems into ever more complex forms, offers a powerful alternative explanation for many of legal theory’s central concerns.
Understanding how redundancy constrains the possibilities of what can be said, including the ability of an individual to make statements about their internal states, and how these constraints are system-specific, so that they vary between systems and between different points in the same system, offers a hermeneutic approach that can be applied not only to law, and the particularities of its ‘rituals’ and ‘rhetoric’, but to other subsystems as well: science, politics, media, the economy, etc. For, in the resistances which social systems of communication offer to psychic systems, with which they can only loosely ‘structurally couple’,Footnote 55 individual actors are constrained. Understanding the constraints of judicial communication mirrors the participation of other actors within their systems of communication, and helps with our understanding of the autonomy not only of law, but of these other functioning systems in contemporary societies as well.
