Résumé

Il est fait, le plus souvent, un usage sélectif et non-systématique de l’analogie du jeu en droit. Cette critique s’applique autant à Hart qu’à Huizinga, l’un et l’autre sélectionnant, pour les besoins de la comparaison, des traits différents du jeu, qu’il est cependant possible d’articuler à une même échelle évolutionniste. D’autres adoptent un modèle structuraliste (Erhmann) ou naturaliste (Midgley) pour présenter les jeux. L’étude de ces différents textes révèle, mieux que ne le ferait la présentation d’un auteur unique, (a) le faisceau des diverses caractéristiques du jeu et (b) les problèmes épistémologiques que suscite leur discussion. On suggérera que l’utilisation de concepts sémiotiques peut contribuer à clarifier autant la nature de l’analogie ludique que certaines de ses manifestations concrètes.

Summary

Much use of the « games analogy » in law is selective and unsystematic. This criticism may be applied equally to its use by Hart and Huizinga, both of whom select different features for comparison with points on the same evolutionary scale. Others adopt structuralist (Ehrmann) or naturalist (Midgley) models for their presentation of games. A review of this literature indicates more fully than does any single author (a) the set of features characteristic of games, and (b) the epistemological problems involved in discussing them. It is suggested that the use of semiotic concepts may assist in clarifying both the nature of the analogy, and some of its substantive manifestations.

1. Introduction

The Anglophone student of jurisprudence is likely first to encounter the « games analogy » [1] in jurisprudence through H.L.A. Hart’s Concept of Law. For Hart, law is a system of rules (as opposed to commands) – the notion of « rules » entailing an « internal point of view », while that of « system » involves the presence not only of « primary » rules, but also of secondary rules of recognition, adjudication and change. On these, and other, features of the modern, « advanced » legal system, Hart finds the games analogy useful. However, his use of it is partial, and at times more than a little strained.

Huizinga includes a chapter on « Play and Law » in his celebrated Homo ludens, but the connections he identifies are rather different to those of Hart. For Huizinga, « the lawsuit can be regarded as a game of chance, a contest, or a verbal battle » (1949, p.78). Yet this is seen, predominantly, as a feature of law in its « archaic » phase, where right and wrong are identified with winning and losing. Survivals of this view remain with us, according to Huizinga, and those aspects of games which contribute to Huizinga’s conception of law as play are clearly regarded as belonging to a phase of culture superseded in modern times in the West.

There is no necessary incompatibility between these two accounts. For if Hart and Huizinga adopt the same evolutionary scale (neither spells it out), together they may be taken to argue that archaic law manifests some, while advanced law manifests other features of games. Yet neither author really explains the force of the analogy. Hart does not claim that law is itself a type of game, and Huizinga, while claiming that « play » (which includes games) represents a basic facet of human culture, does not provide a satisfactory account of the survival of « play » in modern legal culture.

Steps towards a more comprehensive account of games (if not of the games analogy in law) have been taken from different theoretical standpoints. Ehrmann offers a structuralist perspective, more comprehensive in its taxonomy and more sophisticated in the relationships it perceives within the system, but less sensitive to the level of consciousness within the game, which is the focus of Hart’s concern. Midgley distinguishes between a universal phenomenon of game-playing (found also in animals), and the institution of particular games, within which she seeks some definitional common denominator (strongly rejecting Wittgenstein’s use of games to illustrate his notion of « family resemblance »).

Through a review of this literature, it is possible to clarify both the characteristics of games, and the epistemological problems involved in applying the « games analogy » to law. I argue that a semiotic framework assists us to clarify the relationship between games and law, and at the same time to view the individual « characteristics » within a single, coherent theoretical framework – one which indicates also some aspects of the analogy which have been missed in the literature, but which are required in any complete account.

Before reviewing this literature, two preliminary points properly stressed by a semiotic approach should be highlighted.

First, no behaviour pattern naturally presents itself to us as either « game » or « play », any more than such a behaviour pattern naturally presents itself to us as « law » or « rule ». The behaviour pattern may, in some instances, be instinctive, and in that sense « natural » (e.g. the play of children, as we are led to assume). But the perception of that behaviour by the participant as « game », « play », « law », « rule » – a consciousness which is taken as a prerequisite throughout most of the literature here discussed – requires the attribution of a particular meaning to that behaviour. It involves the attribution of the meaning : « this is a game », « this is play », « this is law », « this is a rule ». Any such attribution is socially, indeed discursively, constructed.

Secondly, the discursive construction of the attribute « game » (etc.) is a function of the particular discourse (and of inter-discursive relationships) within which it is constructed. It follows, therefore, that there is no necessary unity in the sense of the attribute « game » as applied to different games. This immediately entails rejection of Midgley’s a priori claim to a common denominator for games, even though it begs (at this stage) the empirical question whether there exist any common elements in all the discursive constructions of the attribute « game ». It is important to stress this point at the outset, since many treatments of the games analogy have tacitly assumed that all games are interchangeable, for the purposes of the analogy. No such tacit assumption is warranted. Football, chess and tiddly winks are all called games (and the former two, at least, figure prominently in the literature on the « games analogy »). But a systematic examination of them will show that they involve quite different combinations of sense relations. The sense in which each one is a « game » is constructed within that discourse, and is a function of the various sense relations which exist within that discourse. To say that they are all « games » because we use the same signifier in relation to them is to commit oneself to an unnecessary and unwarranted reification. It is, moreover, to beg the very questions which must be asked. In the event, we may come to conclude that it is far more sensible to compare particular aspects of legal practices with particular games, rather than to compare law to games in general.

2. Hart’s Use of the Games Analogy

Hart uses the games analogy in several different contexts ; indeed, a glance at the Index to Concept of Law suggests that this is one of his most important analytical concepts. We may distinguish four different arguments in which Hart uses the games analogy. In two of them – the internal point of view, and the relation between the rule of recognition and the finality of judgment of the adjudicator – the comparison appears apt, at least at first sight ; but in the other two – the claims to the variety of types of rule, and to the continuity of rules of law irrespective of the person of the sovereign – the comparison is, to say the least, artificial.

The games analogy is used in support of (what later in the book will be revealed as) the internal point of view, in an argument directed against those who insist on reducing all types of rule to directions to officials (1961, p.40) :

We may compare the inversion of ancillary and principal, which this extreme form of the theory makes, to the following suggestion for recasting the rules of a game. A theorist, considering the rules of cricket or baseball, might claim that he had discovered a uniformity hidden by terminology of the rules and by the conventional claim that some were primarily addressed to players, some primarily to officials (umpire and scorer), some to both. « All rules », the theorist might claim, « are really rules directing officials to do certain things under certain conditions. » The rules that certain motions after hitting the ball constitute a « run », or that being caught makes a man « out’, are really just complex directions to officials ; in the one case to the scorer to write down « a run » in the scoring-book and in the other to the umpire to order the man « off the field ». The natural protest is that the uniformity imposed on the rules by this transformation of them conceals the ways in which the rules operate, and the manner in which the players use them in guiding purposive activities, and so obscures their function in the co-operative, though competitive, social enterprise which is the game.

The last sentence of the argument indicates clearly the basis on which Hart uses the analogy. As against an artificial formalisation, he prefers the « natural protest » which reflects the ordinary instinctive reaction. In line with his ordinary language philosophy, respect is to be paid to the everyday expression of experience. On these grounds, the participant of the game knows directly when a run is scored, or when someone is out ; there is no need for judgment on this matter to be mediated through a third party. Not only does the participant know the meaning attributed to actions, in terms of the rules of the game ; he or she is also committed to those meanings. They have been internalised, hence the « natural protest » which the artificial, formal representation might provoke. Rules, then, are not merely phenomena which participants know about cognitively ; they are tools which those participants use, and they use them « in guiding purposive activities » (namely, to win).

The game, like the law, thus represents for Hart a co-operative social enterprise. Legal rules are a species of social rules and the rules of a game are another such species. The very notion of a social rule entails the existence of an « internal point of view ».

The point is certainly coherent, even if the claim made by the analogy may be considered, on further examination, extravagant. For whereas commitment to the internal point of view is arguably a necessary condition of participation in a game, such a commitment is not, even on Hart’s own argument later, a necessary condition for being subject to a system of laws. Suppose a Frenchman is led blindfold into the middle of a cricket pitch, the blindfold is removed, a bat is placed in his hand, and someone throws a ball at him. The Frenchman has the savoir faire to defend himself by using the bat to hit the ball away. But he lacks the savoir faire of playing cricket. He does not know that he is standing at the crease in front of the wicket, and that if he strikes the ball to the boundary he will have scored four runs. Suppose that he does just that. Whether or not we conclude that he has, thereby, scored the runs (which assumes only that a game of cricket is being played), we would surely agree with Hart that the Frenchman can hardly be said to be playing cricket.

The rules of the game of law, on the other hand, operate whether or not participants know of their existence. In crime, generally speaking, ignorance of the law is no defence ; and in everyday life, people make contracts all the time, without being aware that they are entering into legally binding transactions. This is not to suggest, of course, that there never exists, within the game of law, an internal point of view comparable to that which Hart illustrates from the game of cricket. The argument, rather, is that the internal point of view may claim a far greater status in games than in law.

A second use of the games analogy by Hart occurs in his celebrated argument about the game of « scorer’s discretion ». This occurs in the context of his attempt to navigate what he sees as the Scylla of formalism and the Charybdis of rule- scepticism. The fact that a legal system may have adopted secondary rules, including rules of adjudication according to which the resolution of disputes is entrusted to an authorised official whose decisions are final, does not mean, Hart argues, that the law now becomes irrelevant, and all that matters is the decision of the official, or, as in one version : « the law (or the constitution) is what the courts say it is. » Here, too, Hart appears at first sight to use the games analogy to good effect (1961, p. 138-139) :

Like the changes from a régime of custom to a mature system of law, the addition to the game of secondary rules providing for the institution of a scorer whose rulings are final, brings into the system a new kind of internal statement ; for unlike the players’ statements as to the score, the scorer’s determinations are given, by secondary rules, a status which renders them unchallengeable. In this sense it is true that for the purposes of the game « the score is what the scorer says it is ». But it is important to see that the scoring rule remains what it was before and it is the scorer’s duty to apply it as best he can. « The score is what the scorer says it is » would be false if it meant that there was no rule for scoring save what the scorer in his discretion chose to apply. There might indeed be a game with such a rule, and some amusement might be found in playing it if the scorer’s discretion were exercised with some regularity ; but it would be a different game. We may call such a game the game of « scorer’s discretion ».

We may note that the games analogy is here deployed in support of a feature of a « mature » system of law, and indeed the analogy (like that in support of the internal point of view) is designed to support the rational legitimation of the system. For the very institution of secondary rules, we learn elsewhere, is designed to promote the certainty and efficiency of the system ; here, where we have the use of a scorer (a referee, analogous to a judge), we have the benefit of not just one form of rational determination (finality of judgment of the authorised – and expert – adjudicator) but two : the scorer is not authorised to decide arbitrarily, but rather has a « duty to apply it [the scoring rule] as best he can ». Hart does not tell us how he would analyse the situation where the scorer did arbitrarily depart from the scoring rule.

We do, however, have to contemplate the possibility that the scorer on the field itself might decide to play the game not according to its rules, even when claiming to do so (thus, not accepting that he is engaged in a different game, that of scorer’s discretion). There is increasing discussion of the extent to which the football [2] referee may be swayed by the noise of the crowd. Suppose that emotions are running particularly high, with the home team in danger of losing a championship if one more goal is scored against them. A defender of the home team handles the ball in the penalty area. This is observed by the referee, who is quite convinced that this was a case of deliberate handling, thereby meriting the award of a penalty to the opposing team (indeed, we may go further and say that according to the rules of the game, the opposing team has a right to the penalty). Yet the referee, fearful of the consequences of awarding a penalty against the home team (consequences, perhaps, of riot and injury to persons other than himself), refuses to award the penalty. The result of the game is affected thereby. Is this still football, or is it scorer’s discretion ? On Hart’s argument, it would appear to be neither : it is not football, since the referee is not complying with his duty to apply the scoring rule, nor is it scorer’s discretion, since we lack here a rule (accepted from the internal point of view) to the effect : « no rule for scoring save what the scorer in his discretion chooses to apply. »

What we have here, in the sphere of sport, is a clear example of one of Hans Kelsen’s classical problems, that of the « erroneous legal decision ». We may have to decide, for football as for law, whether this is a game where rules determine outcomes, or where what counts is the decision of an authorised official, who claims to be « applying » the rules, but who in reality is at best « guided by » them. Indeed, some would say that the rules are always defeasible : the real game is « the game of the game », that set of tacit conventions which tell the insider to what extent he must observe the rules of the game (so as to retain the spectacle of events which are intelligible as an instance of that game), while at the same time departing from those rules for reasons external to the game itself.

Hart, by contrast, implies that, because the scorer has a duty to apply the scoring rule, he will always do so, or at least that his good faith in seeking always so to do will be generally recognised. Hart thus seeks to carve out a position between formalism and rule-scepticism by reliance upon the good faith of the adjudicator. The games analogy may appear apt. Football referees, like judges, may occasionally make mistakes, but they are appointed because of their expertise and judgment, and this expertise and judgment is generally recognised (or at least, must publicly not be challenged – by the players, if not by the spectators).

Yet further reflection suggests that we need not rely upon such unrealistic idealism, in order to find our way between the two extreme positions of formalism and rule scepticism. Hart is right, in my view, in seeing a distinction between the rule-based game adjudicated by a scorer, and the game of scorer’s discretion. But it is not good faith – certainly, not good faith alone – which explains the distinction. Rather, it is the presence within the game of a feature which (on the view of both Hart and Huizinga) belongs to an earlier stage of evolution : something like luck, chance or fate.

In the game of football, the ball is always in motion while it is « in play », and most of the players are in motion with it. The referee has to run with the players ; he has no fixed, privileged vantage-point, but must rely upon his legs, his eyes, and his two touch-line officials in order to inform him of the events which transpire during the game. He is continually required to make decisions on these events : thus, to determine facts, and to attribute to those facts their proper significance in terms of the rules of the game. But sometimes he fails to observe something which thousands of spectators have clearly seen (and which, nowadays, will be the subject, to the referee’s embarrassment, of interminable video replays). Nevertheless, he makes decisions, and these decisions are accepted by both players and spectators (notwithstanding vociferous ritual protests to the contrary). In short, the referee is subject to at least the same hazards of fact construction as beset the courtroom. Moreover, his judgment in applying the rules to the facts – is this a foul, should a penalty be awarded, was the player off-side ? – is subject to influences beyond the deductive application of rules to facts, as indicated in the example of the (deliberately) erroneous decision. The referee applies a form of practical reason not too far distant from MacCormick’s « consequentialist argument ». But beyond this, it is often thought that he takes a « holistic » approach to adjudication which is not in accordance with the rules of the game, and may, as applied to the individual incident, be seen as partiality (this might count as the « principles » of the game, if such a concept were recognised). After giving one team the benefit of the doubt on a few occasions when a penalty might arguably have been awarded against them, on the next occasion (if and when it occurs) he does award a penalty, even though the case for it is far less strong. Again, he may be influenced by the reputation of the team, previous media comment on his performance (for the football referee is not protected, like the judge, by contempt of court), or by the noise of the spectators.

What are we to make of such phenomena ? Are we to say that this is not a game of scorer’s discretion simply because the referee sincerely attempts, by and large, to fulfil his duty to apply the rules of the game, and players and spectators alike attribute good faith to him ? No. The interstitial arbitrariness of the decision-making of the scorer is an essential part of the game itself. It is the element of luck, or chance, which combines with skill and strategy to give the game its interest. The « opponent » of each side (to use the term in its Greimasian sense) is not merely the other side, but the elements of chance which – less predictably than the actions of the other competitor – may deprive one’s team of its victory [3].

We may now offer an alternative description to that of Hart of a game like football. There is the playing of the game – what the players do – and there is also the playing of the game of the game – what judments are made about what the players do by the scorer, who, despite his good faith in attempting to apply the rules of the game, will necessarily on some occasions come to arbitrary and contestable (but nevertheless final) decisions. The tension between the good faith application of the rules, and the necessary arbitrariness of the outcome of some of the adjudicatory decisions, is an integral part of the game, and contributes to, rather than detracts from, its enjoyment. It provides many of the « talking points » which will engage the post-match commentaries, and which, in their very discursive repetition, will accord recognition both to the authority and decisions of the referee, and to the mix of skill and chance which gives that particular game its character. In short, though the referee is located on the same time and space co-ordinates as the players, he is playing the « game of the game », one might say the pragmatics rather than the semantics of football.

The games analogy, thus understood, reveals its own weaknesses. The peculiar mix of judgment arbitrariness most frequently which may be essential to the game of football, will not provide the litigants with a great deal of enjoyment. To them, it truly detracts from the game of the law ; it is an interference, not an integral part of it.

Though Hart’s uses of the games analogy, observed thus far, are contestable, they are at least intelligible and plausible up to a point. The same can hardly be said for his other examples. In introducing the distinction between duty-imposing rules and power-conferring rules, Hart tells us (1961, p.9) :

The same contrast between these two types of rule is also to be seen between those rules of the game which veto certain types of conduct under penalty (foul play or abuse of the referee) and those which specify what must be done to score or to win.

But such « scoring rules » are hardly equivalent to power-conferring rules, as Hart understands them. The latter are described in this passage as :

rules such as those prescribing the procedures, formalities, and conditions for the making of marriages, wills, or contracts (which) indicate what people should do to give effect to the wishes they have.

We may mention just two obvious difficulties in this analogy. Everyone playing the game is seeking to win, and if winning means (as it normally does) achieving the highest score, every team is attempting to score. That is the very purpose of the game. To enter into the game is thus to commit oneself to seek to score (in semiotic terms, to accept the « contract »). That is hardly the case in Hart’s examples of legal institutions which use power-conferring rules. It is of the essence of the latter that their use is optional (as contrasted with duty-imposing rules). No-one has to make a marriage, will or contract ; whether they do so depends precisely upon « the wishes they have ». In more formal terms, the scoring rules are constitutive of the game, since it has no meaning apart from such rules [4], while the rules of marriage, wills or contract are not constitutive in this sense, but merely serve to enable the attribution of legal significance to be communicated, in contexts where the institution would, in any event, have a social meaning. Secondly, as Hart stresses elsewhere, power-conferring rules enable their users not only to enter into optional legal relationships, but thereby to alter the legal positions – including the legal capacities – of others. By « scoring » in a game, one hardly alters the capacity, within the rules of the game, of the other to participate in it.

Here too, we may note, Hart locates the games analogy at the mature rather than early stage of social advance. In another formulation of this same argument (1961, p. 238-239), the comparison is put in terms of the rules of « any complex game ».

An even odder use of the games analogy is found in Hart’s argument for the continuity of rules, as distinct from the conception of law as the command of the sovereign for the time being. The statement that a new legislator has a right to legislate, Hart argues (1961, p. 58), presupposes the existence, in the social group, of a rule under which he has this right.

If it is clear that the rule which now qualifies him was accepted during the life-time of the predecessor, whom it also qualified, it is to be assumed, in the absence of evidence to the contrary, that it has not been abandoned and still exists. A similar continuity is to be observed in a game when the scorer, in the absence of evidence that the rules of the game have been changed since the last innings, credits the new batsman with the runs which he makes, assessed in the usual way.

Similarly, he argues elsewhere (1961, p. 62) :

But, given the present acceptance of the underlying rule, the persistence of laws is no more mysterious than the fact that the decision of the umpire, in the first round of a tournament between teams whose membership has changed, should have the same relevance to the final result as those of the umpire who took his place in the third round.

In both cases, the parallel is difficult to see. The new batsman is hardly analogous to the new sovereign in the command theory, nor is the changed composition of teams (or of the umpire) playing in different rounds of the same tournament.

The artificiality of these latter examples perhaps directs our attention to a question Hart might prefer us not to raise : what game is he playing ? Surely, one might respond, he is playing the game of the analytic philosopher, who (at that period, at least) gained both pleasure and recognition by playing word-games with « games ». It was, after all, the example of games which Wittgenstein used to illustrate his concept of family resemblance. And the games analogy secured recognition as a vital concept in both the philosophy of knowledge (language-games) and ethics (e.g. Ayer on the promising game, discussed by Midgley). To some, it may appear that Jackson, here, is most definitely not « playing the game », in making such cheap shots. But the shot is, in fact, coherent with the overall semiotic framework Jackson has adopted. Jackson (1988) has advanced the notion of the narrativisation of pragmatics. This suggests that we should be sensitive to the narrative typification of acts of enunciation (in their proper social and institutional contexts). Of the Analytic Philosopher, then, we may ask : what kind of act of enunciation is likely to be regarded as interesting, intelligent, plausible. The game of the philosopher is one into which intellectual fashions (the instantiation of chance, in this context ?) are liable to intrude. Hart is, indeed, playing the game by straining to introduce the games analogy. Jackson, however, is not playing the game, by saying so.

3. Huizinga : Play and Law

Whereas Hart selects features of « complex » games, and compares them with the characteristics of a « mature » legal system (these being the characteristics which contribute most to the liberal rationality of the modern legal system), Huizinga stresses the arbitrary and indeterminant aspects of play, and attributes them particularly to the « archaic » stage of the development of law. The two theories, therefore, are complementary, in correlating different aspects of games with the characteristics of law at different stages of its (assumed) evolutionary progression. Huizinga’s examples, like those of Hart’s, provide elements which we may re-use in a semiotic analysis of the games analogy.

Huizinga conceives that he has a difficult case to make (1949, p.76) :

At first sight few things would seem to be further apart than the domain of law, justice and jurisprudence, and play. High seriousness, deadly earnest and the vital interests of the individual and society reigns supreme in everything that pertains to the law...

That an affinity may exist between law and play becomes obvious to us as soon as we realize how much the actual practice of the law, in other words a lawsuit, properly resembles a contest whatever the ideal foundations of the law may be...

Contest means play... there is no sufficient reason to deny any contest whatsoever the character of play. The playful and the contending, lifted on to the plane of that sacred seriousness which every society demands for its justice, are still discernible to-day in all forms of judicial life.

From this, we see some of the vital components of Huizinga’s position. Serious action may nevertheless have a « play-quality », which in the case of law derives from the element of contest. Play is the primary concept, contest is an important form of it, law is a site where it is found – « still ». In other words, law is play insofar as it retains the feature of agonistic contest, and this feature is more pronounced in ancient sources, and societies studied by anthropologists, than it is today. Huizinga’s chapter would appropriately be included in any course on Historical Jurisprudence. His sources are ancient Greece, Rome, the Bible, pre-Islamic Arabia, the Germanic concept of Urteil and other Germanic concepts, the ancient Hindu sacred books, and the Greenland Eskimo. The most modern of his citations are Goethe’s description of a sitting of a Venetian court in the Doge’s Palace (1949, p. 78) and Abyssinia under Italian Rule, where (1949, p. 84) :

Litigation still continued to be a passion and a sport that delighted the natives. According to an English newspaper, a judge received a visit from a man who had lost his case on the previous day, but now said contentedly : « I had a very bad lawyer, you know, all the same I’m glad to have had a good run for my money ! »

Success in the contest does not depend upon abstract conceptions of (legal or even moral) right and wrong, and the separation of the contest from such rational standards seems to increase, according to Huizinga, the further we go back in legal history (1949, p. 78) :

The active association of law and play, particularly in archaic culture, can be seen from three points of view. The lawsuit can be regarded as a game of chance, a contest, or a verbal battle.

We moderns cannot conceive justice apart from abstract righteousness, however feeble our conception of it may be. For us, the lawsuit is primarily a dispute about right and wrong ; winning and losing take only a second place. Now it is precisely this preoccupation with ethical values that we must abandon if we are to understand archaic justice. Turning our eyes from the administration of justice in highly developed civilizations to that which obtains in less advanced phases of culture, we see that the idea of right and wrong, the ethical-juridical conception, comes to be overshadowed by the idea of winning and losing, that is, the purely agonistic conception. It is not so much the abstract question of right and wrong that occupies the archaic mind as the very concrete question of winning or losing. Once given this feeble ethical standard the agonistic factor will gain enormously in legal practice the further back we go ; and as the agonistic element increases so does the element of chance, with the result that we soon find ourselves in the play-sphere. We are confronted by a mental world in which the notion of decision by oracles, by the judgement of God, by ordeal, by sortilege – i.e. by play – and the notion of decision by judicial sentence, fuse in a single complex of rules of the game. We still acknowledge the incontrovertibility of such decisions when, failing to make up our minds, we resort to drawing lots or « tossing up ».

Huizinga here associates the agonistic element with chance, and ties it clearly to a developmental theory of ethics. But it is not « pure chance » which he has in mind : even the casting of lots (« sortilege ») is an instrument of divine judgement in the sources to which he refers. Indeed, he continues (1949, p. 79) :

Divine Will, destiny and chance seem more or less distinct to us, at least we try to distinguish between them as concepts. To the archaic mind, however, they are more or less equivalent.

Elsewhere, he points out, the contest may be decided by either physical or intellectual prowess, albeit not connected directly to the issue in dispute.

The third form of play which, Huizinga argues, survives in the modern trial, is that of the verbal battle (1949, p. 84) :

We have tried to distinguish three play-forms in the lawsuit by comparing it as we know it to-day with the legal proceedings in archaic society : the game of chance, the contest, the verbal battle. The lawsuit remains, in the nature of things, a verbal battle even when it has lost is play-quality wholly or in part, actually or in appearance, with the progress of civilization. For the purposes of our theme, however, we are concerned only with the archaic phase of this verbal battle, where the agonistic factor is at its strongest and the ideal foundation of justice at its weakest. Here it is not the most meticulously deliberated juristic argument that tips the balance, but the most withering and excoriating invective. The agon in this case consists almost exclusively in the endeavour of each party to exceed the other in choice vituperation and so to keep the upper hand.

The singing-match of the Greenland Eskimo is described by way of example (1949, p. 85) :

No distinction is made between well-founded accusations, satirical remarks calculated to tickle the audience, and pure slander. For instance one singer enumerated all the people who had been eaten by his opponent’s wife and mother-in-law during a famine, which caused the assembled company to burst into tears... Most of the spectators join in the refrains of the song, applauding and egging the parties on. Others just sit there and go to sleep. During the pauses the contestants converse in friendly terms. The sessions of such a contest may extend over a period of years, during which the parties think up new songs and new misdeeds to denounce. Finally the spectators decide who the winner is. In most cases friendship is immediately re-established, but it sometimes happens that a family emigrates from shame at having been defeated.

Huizinga stresses that amongst these tribes such contests take the place of judicial decisions, and concludes (1949, p. 86) :

Here, therefore, we are dealing with a cultural practice which fulfils the judicial function perfectly in agonistic form and yet is play in the most proper sense. Everything passes off amid laughter and in the greatest jollity, for the whole point is to keep the audience amused.

There are both strengths and weaknesses in Huizinga’s presentation. The link which Huizinga suggests between archaic and modern legal procedures is well observed, and may indeed be even stronger than Huizinga suggests. For the use of « chance » is, perhaps, too broad. There are, indeed, games of pure chance (e.g. board games dependant upon the throw of a dice, such as snakes and ladders) where no choices whatsoever have to be made by the participants in the course of play – a phenomenon quite different from the games which are the primary comparators in our context, such as football or chess. Rather, the contest in Huizinga’s example proceeds on criteria which, though they may be rational (in the sense of purposive), are not regarded as relevant or pertinent to the case in issue. For example, the « verbal battle » (itself clearly a form of contest) may focus upon aspects of the lives of the participants which are quite irrelevant to « the issue », and success in that verbal battle will depend upon the « chance » of the skill of the performer, and the quality of the performance that day. Again, the use of oracles or the casting of lots is not a resort to chance, but rather to what is conceived as a superior form of adjudication, one which relies upon the omniscience of divine judgement [5].

A critic might suggest that this reformulation of Huizinga’s argument does not negate his essentially evolutionary view. The arbitrary (here in the sense of the irrelevant, or dependence upon the skills of the other – the divine) is still more prominent in the « archaic » and anthropological sources. Even if this does not reflect the progress of ethical development, it does reflect the progress of cognitive development, in the sense that we can see a progression from reliance upon irrelevant skills and knowledge to reliance upon relevant criteria as judged by rational human beings.

Yet even this reformulation will not do. Relevance itself is a discursive construction, and may differ (within one and the same society, in relation to one and the same incident) according to the criteria (or closure rules) of the group concerned. Many contemporary examples could be given. The « exclusionary rules » of the English law of evidence, for instance, are designed to exclude not the irrelevant (who would want to use that anyway ?) but rather that which is relevant to others or in other contexts.

Here, as elsewhere, clarity is gained by proper deployment of the distinction between semantics and pragmatics. In his account of the Eskimo singing-match, Huizinga did not tell us what actually was in dispute between the contestants. Let us assume, however, that it was nothing to do with the alleged but rather was reserved for cases of special cannibalism of the opponent’s wife or mother-in-law. Are we really to say that the Eskimo nonetheless regarded the cannibalism as relevant – thus, that the cannibalism was part and parcel of the same semantic construction as the dispute itself (whatever that was) ? Or should we say that the use of cannibalism by way of ritual abuse of the opponent is part and parcel of the pragmatics of the resolution of the dispute ? If the dispute was a game (on which I reserve my position), the ritual accusation of cannibalism is part of the game about the game. It accords, in that society, with the narrative typification of being a good contestant. In Hart’s terminology, it is part of the secondary rules of adjudication.

This kind of analysis takes a step towards a semiotic presentation of the games analogy. For the contemporary trial also depends, in its pragmatic aspects [6], upon a whole set of performances, judged in narrative terms, which, though they exhibit skills and perform goals, are not the direct representation of the semantics of the problem, but rather an authorised form or mediation of it. The parallels are too obvious to labour : the degree of knowledge and skill of the advocate, the non-relevant rhetoric with which the claims are presented, the experience of the judge (even, perhaps, the state of his digestion). We may invoke again the « game of the game » as played by the football referee, and the players who observe him. It is part of the pragmatics of football that lack of capacity (to see everything), and the occasional exercise of judgement on irrelevant factors, should be taken into account.

The weak point of Huizinga’s analysis resides in his conception of play, and its relationship to « the contest ». Conceptually, the two are distinct. Even if all play involves a contest (not all would agree), all contests do not involve play – at least, if by « play » we invoke Huizinga’s notions of « jollity » and « entertainment ». Games theorists of various kinds have debated whether it is a necessary feature of the game and/or play that it is closed off, separated from everyday life, and whether it is necessarily « disinterested ». These are issues to which we must return. But even if we adopt expansive conceptions of play and games in this respect, Huizinga’s own examples force us to ask the question : playful to whom ? Even if there is jollity and entertainment for the spectators, does it follow that this attribute is endorsed equally by the participants ? In the context of the modern trial, it is self-evident that what may (to put the matter at its mildest) have elements of a game or play to the advocates may completely lack that dimension in respect of the litigants themselves. Yet despite this obvious distinction, those who use the games analogy commit the same vice as does much of legal theory : they insist on reducing the trial (indeed, in many cases, « the law ») to a single phenomenon, and thereby deprive themselves of any possibility of attaining a plausible description. Semiotics would here insist (for both theoretical and methodological reasons) upon isolating the different discourses which take place within « the trial » (or « the game »), and analysing each one separately. Only thereafter does it become possible to assess the means by which the unity of the different discourses, the sense in which they belong to one and the same phenomenal event, is constructed.

A crude exemplification of this approach may assist in clarifying the relationship between « play » and « the contest ». Even if « play » is not to be attributed to the behaviour of all the participants in the trial, there is a credible argument that all of them are involved in a contest of some kind. It may not be the same contest. The parties to litigation contest the outcome of the dispute which prompted the litigation. What matters to them is the subsequent effect of the process in the outside world. The advocates, on the other hand, are involved in a contest purely within the legal process. Their contest is located at the level of the pragmatics, not the semantics of the trial. For them, the particular case may be a means rather than an end – a means towards their personal contests for professional advancement. The same may be said, mutatis mutandis, for all of Huizinga’s examples, irrespective of culture and historical period. In short, it is the notion of contest, not that of play, which provides the common thread. Play is sometimes involved, for some of the participants, but not always.

By chance (!), this finding accords with Greimasian claims regarding the basic structures of meaning. The narrative syntagm which represents a basic model of intelligibility (i.e. of the construction of sense) at the « syntagmatic » level of the elementary structures of signification, involves a « polemic » element. This may be represented by a genuine contest between people, or, more abstractly, by the striving to accomplish a goal. In other words, « contest » (in this sense) is central to the intelligibility of human action itself. By contrast, « play », at least as Huizinga appears to conceive of it, is an attribute of human action, something that is regarded as jolly or entertaining. It is therefore located, semiotically, on the same level as other attributes, such as « true », « just », « beautiful », « skillful », « courageous », etc. Many such attributions are made in the course of the trial, as they are in the course of the game. For some spectators, indeed, that may be what they come to see (in whole or in part). But it is the contest which gives the activity its meaning.

In short, the contest belongs to the semantic construction of action ; « play » to its pragmatic presentation. In offering this distinction, there is no implication that « play » is unimportant, or that semiotics is uninterested in it.

4. Other Perspectives on Games

The analysis thus far has concentrated on arguments (of Hart and Huizinga) which address explicitly the relationship between law on the one hand and play/games on the other. However, many necessary and valuable additional insights may be gained from the general literature on games. Here, reference will be made to the wider claims of Huizinga, and to discussions by Caillois, Ehrmann and Midgley.

Earlier in Homo Ludens, Huizinga provides a definition of play which includes many of the elements taken up by later authors, but which are not taken up by Huizinga himself in the context of law (1949, p. 13) [7] :

Summing up the formal characteristics of play we might call it a free activity standing quite consciously outside « ordinary » life as being « not serious », but at the same time absorbing the player intensely and utterly. It is an activity connected with no material interest, and no profit can be gained by it. It proceeds within its own proper boundaries of time and space according to fixed rules and in an orderly manner. It promotes the formation of social groupings which tend to surround themselves with secrecy and to stress their difference from the common world by disguise or other means.

Ehrmann (1968, p. 35) notes parallels in the presentation by Caillois :

.... the preceding analysis allows us to define play as an activity which is essentially :

1. free : the player cannot be obliged to participate without robbing play of its nature as alluring and joyful diversion ;
2. separate : it is circumscribed within limits of space and time which are precise and fixed in advance ;
3. uncertain : its course cannot be determined nor its outcome reached in advance, a certain latitude for innovation being left necessarily to the initiative of the player ;
4. unproductive : it creates neither goods nor wealth nor new elements of any kind ; and, except for redistribution of property within the circle of players, it results in a situation identical to that with which it began ;
5. controlled : it is subject to conventions which suspend ordinary laws and introduce temporarily a new body of legislation endowed with exclusive authority ;
6. fictive : it is accompanied by a specific awareness of a second reality or of straightforward unreality in relation to everyday life.

Nevertheless, as Ehrmann points out (1968, p. 31), Huizinga stresses the competitive aspect of play, and offers (in Ehrmann’s view) a reductive account which sees this as the common denominator, while Caillois presents a typology of play which gives equal weight to competitive games, games of chance, games of simulation, and games of physical excitement [8]. For Caillois, moreover, there is a continuum running from ludus (controlled play) to paidia (spontaneous play).

Despite the differences between them, Ehrmann points out, both Huizinga and Caillois adopt the same

world-view, a fundamentally rationalist view according to which human activities relate, on the one hand, to dreams, gratuitousness, nobility, imagination, etc. and on the other hand to consciousness, utility, instinct, reality, etc.

Yet this world-view, which separates and often privileges « reality », is itself a cultural construct, as are all the oppositions here deployed by our authors. Thus, for Ehrmann (1968, p.56) :

1. Play is not played against a background of a fixed, stable, reality which would serve as its standard. All reality is caught up in the play of the concepts which designate it. Reality is thus not capable of being objectified, nor subjectified. However, it is never neutral. Nor can it be neutralized...
2. Just as culture is, in the last analysis, communication, so is play... and game.

This structuralist approach may assist us later in the argument. For the moment, we may look briefly at these various characteristics of games, and ask whether they are severally definitive of games in general, of some particular games (in general), or of the playing of particular games in particular contexts. My argument will be that, despite strong claims for the universality of some characteristics of games, it is the last level, that of the playing of a particular game (the game of that game), where a significant lack of uniformity in the characteristics of games (and thereby the weakness of any general games analogy) may be demonstrated.

Take, first, the notion of freedom or voluntariness, stressed by both Huizinga and Caillois. The latter says that « the player cannot be obliged to participate without robbing play of its nature as alluring and joyful diversion. » [9]. But this begs the question : « alluring and joyful » to whom ? The very title of Huizinga’s book is Homo Ludens. But the most famous Roman « games » were hardly voluntary for the participants, much as they may have provided joyful diversion for the spectators. It is, perhaps, a sad reflection on human nature that human beings can derive pleasure from the unfree activities of others. It is hardly surprising, then, that Huizinga can speak of the play element in law, notwithstanding the often compulsory nature of adjudication.

Moreover, we have to ask what is the meaning of a « free » activity ? Is the presence or absence of constraint (here, to participate in a game) an objective aspect of external reality, to which we have access, or is it part of the discursive construction of the activity itself, a construction informed by social knowledge and communicated (often to the self) by discursive means ?

Just as entry into the game may be constructed as more or less free, so too games differ in the degree to which they demand adherence to their own internal rules. This is the spectrum from ludus to paidia, suggested by Caillois. And if « play » is associated with spontaneity, and « game » with control, the categories frequently meet in practice. As Ost observes, in the legal context (1988, p. 85) :

It may in fact be maintained that each game combines in a variable but irreducible proportion an element of regulation, of convention, of repetition on the one hand and an element of chance, inventiveness and imagination on the other. It may by the way be noted that very many forms of game may in turn be classified on a scale, with one of the poles being extreme regulation and the other that of extreme arbitrariness.

Ost himself notes that terminology sometimes reflects this situation : French has just the one word jeu to cover the variety of game activities, while English has both « play » and « game ». More significant still, in English we speak quite naturally of « playing a game ». Our earlier comparison of football and chess may be located along such a spectrum. Games are, for Caillois, « unproductive : it creates neither goods nor wealth nor new elements of any kind. » Somewhat more carefully, Huizinga says that play « is an activity connected with no material interest, and no profit can be gained by it. » This does not prevent Huizinga from seeing play as present in law, nor does it prevent Caillois, later, from stressing the socialising function of games : « games discipline the instincts and institutionalise them. » [10] And in the legal context, in particular, the game has a very real function, of ritualising (thus temporarily neutralising, and ultimately resolving) the conflict, which otherwise might result in the shedding of blood. Of course, ritual is something which it is always easier to recognise in others. We have no difficulty in ascribing such a function to the Eskimo song-contest. I have little doubt that the unwesternised Eskimo, observing our trial, would see it just as much as a ritualisation. On this matter, Midgley has some particularly important things to say (1974, p. 242) :

All play that deserves one’s attention involves difficulties, rising to a fight. What people take seriously, that they ritualize. Religion is surrounded by ritual, not because it is ossified and cast aside from real life, but because it is so important that it demands perfection of form. And justice demands ritual, although ritual can distort justice. Play is for Huizinga still a set of enclosed systems, only they enclose within them all the most central human activities, and do so because those activities themselves require it.

Overt conflict is not the only phenomenon which calls for ritualisation. Emotions themselves have the capacity to engender conflict and danger. They call for social regulation, and play fulfils a vital role in this. Midgley (1974, p. 246) quotes Berne (1966, p. 17-18) :

The essential characteristic of human play is not that the emotions are spurious, but that they are regulated... Pastimes and games are substitutes for the real living of real intimacy.

Sports which involve physical contact and dirt (e.g. rugby) exemplify the point. Significantly, perhaps, we play these games only with those with whom we have [11] a communitarian relationship : we play games with our friends, and sports with other teams in the « same league ». Midgley points out that games are not interchangeable forms of the ritualisation of conflict. Particular forms of conflict require particular forms of ritual. For example (1974, p. 237) :

Lawn tennis will not do instead of football for some quite interesting reasons. It is not a team game : it involves no physical contact and does not make the players dirty. Moreover there are rackets, which, if used in the spirit of football, might kill people. Any attempt to substitute it would result, either in changing lawn tennis past recognition, or (more likely) in the public’s abandoning tennis and inventing instead some much more primitive ritualized contest of the kind from which football originally sprang.

Midgley also takes issue with the rule book conception of games (the one endorsed, in effect, by Hart), as providing an insufficient account of the game. Her argument suggests that the rule book is insufficient largely because it neglects the pragmatic variables which attend the playing of the game (1974, p. 237-238) :

The Rule Book is misleading ; or rather, it misleads those few unhappy people who expect to see the whole truth about anything written down in a book. Books take obvious points for granted. For instance, the book does not mention spectators, nor the reasons for playing and the kind and degree of friendliness called for between players : nor do they mention the choice of teams and opponents, but every game makes quite complicated demands here. Nor does it mention how you give up playing, but that doesn’t show there are no proper or improper ways of doing it. (Anyone giving up chess in Russia or football in Glasgow would soon find out about that.)

Of course, both football and chess can be played with or without spectators, but it is never played neither with nor without spectators ; the presence or absence of spectators is an essential aspect of the playing of the game, if not the game itself. Any full use of the games analogy must take account of such variants. Some aspects of law are played with spectators, others without. Spectator participation differs according to the game. Which aspect of law, we must always ask, is being compared to which aspect of the playing of which game ?

Is it essential to the playing of any game that one is conscious that one is so doing ? Again, such a conceptual claim is often made. Huizinga speaks of play as standing « quite consciously outside « ordinary » life », and applies this to children’s play, the playing of sports, the actor, and the violinist (Midgley 1974, p. 241). But is it really essential, or merely a characteristic often or normally found ? And what is the consciousness here involved ? Is it a consciousness that the attribute « play » is associated with the activity, or that the activity stands, somehow, outside of everyday life ? Once again, Huizinga’s argument for the presence of play in law suggests that no categorical conclusion is possible. One hears barristers speak of advocacy as a game, but they may not apprehend it as such at the moment they are engaged in it – despite the fact that they may be conscious that they are deriving pleasure from it. One may, indeed, derive pleasure from something which is serious, and fraught with significant consequences. In responding to her last House of Commons debate as Prime Minister, on a vote of no confidence in her government, Mrs. Thatcher turned round and proclaimed : « I am enjoying this ! ». That debate, even more than most in the House of Commons, may indeed have been a ritual conflict, since Mrs. Thatcher had some hours earlier already announced her intention to resign. Nevertheless, the debate (broadcast on radio, shown in part on television) was crucial for the prospects of the Conservative Party in general. Play and seriousness were not here opposed categories – a point emphasised by Ehrmann and others in the literature.

Finally, there is the vexed question of the separateness of games from everyday life. For Huizinga, play is a transposition, an « embellished, ennobled representation (mimesis) of » ordinary life (Ehrmann 1968, p. 37). Caillois adopts a more limited conception of separateness, namely circumscription « within limits of space and time which are precise and fixed in advance » (his point 2, above). Midgley observes (1974, p. 235-236) :

I do not want her to pursue the question about the basis of promising, so much as to investigate the notion of Games as a Closed System. This, I suggested, means that they are discontinuous with the life around them. That seems to be how the term is used in mathematics ; the Theory of Games deals with a certain set of closed systems. In this use, no question arises about the reasons or motives for playing ; there is no suggestion of playfulness or jollity in the ordinary sense. But when you bring the term into moral philosophy and apply it to people’s actual activities, the reasons and motives begin to matter. Any actual activity has motives, and it won’t be a closed system, optional and removable, unless the motives are of a special kind. They must not be very strong, or it will begin to matter whether we play or not ; they must not be very specific, or it will begin to matter which game we play.

She gives as an example Bobby Fischer, playing chess. We could not say to him, especially when he is playing in a tournament : « calm down, chess is only a game. » But we could say exactly that to each other, when playing chess in our inexpert way, purely for relaxation. In other words, chess neither is nor is not separate from everyday life : it is only the playing of a particular game of chess which may be said to be so. And even in such a case, it is not the mere fact of the playing in a particular context which gives the game its special character ; it is the attribution and communication of consequentiality which makes that difference. We see that we are dealing, once again, with the pragmatics of game-playing, and with the discursive construction of modalities (here, the attributes of seriousness and consequentiality) within them.

We are now in a position, without over-elaboration, to identify and respond to the basic epistemological question about games posed by Wittgenstein. There is no point in seeking to define a game in the abstract (at the purely semantic level), since such a phenomenon has no discursive existence. We can only pose the question in relation to the playing of the game, and this playing may occur in so many different contexts, and with so many variations in its discursive construction, that we may feel confident, at best, that Wittgenstein is right in proposing a family resemblance model, and in rejecting any strict definition, which requires necessary and sufficient conditions. Midgley opposes this view, on the grounds that all games presuppose an institutional basis in promising, a phenomenon she regards as universal and natural to the human race. But even if we accept this point, it hardly amounts to a definition, since the latter requires not only necessary but also sufficient conditions, and these are not disclosed by Midgley’s argument.

We may, however, sympathise with Midgley in seeking some model of games which will enable us to account for their varying characteristics in a more systematic fashion than does the theory of family resemblance. I suggest that semiotics is able to provide such a model. Many of its elements have been adumbrated in the foregoing analysis. I conclude with an attempt to put this semiotic account into a more complete and coherent form, and to show how it may assist in understanding the various manifestations of the « games analogy ».

5. The Semiotic Model and the Games Analogy

The game is a discourse rendered intelligible through its manifestation of basic structures of signification. These structures deploy (following the Saussurean model, as adapted by Greimas) both syntagmatic and paradigmatic axes. The paradigmatic axes are the units of sense which, at various points in the syntagm, are substitutable for each other while still making sense (though a different sense) ; the relations of the units so substitutable are often formalisable by means of the semiotic square. The « serious » and the « playful », the weak and the brave, appearance and reality, sincerity and lying, are relations of this kind.

However, our main concern in this context is with the syntagmatic axis. The classical Greimasian approach divides this into three elements : « contract », « performance », and « recognition ». « Contract » is the process by which the subject of the action is instituted with a goal (though not necessarily consciously), and, by receiving such a « mandate », becomes the « subject » of the « performance » of the action. Finally, there is a process of « recognition », whereby a message regarding the success or otherwise of performance of the contract is communicated. Since the people (or other semiotic objects) who perform these roles may perform several of them, it follows that « recognition » may be a purely internal, individual process, the same person – here, the participant in a (perhaps solo) game – sending and receiving the message recognising fulfilment of the task.

Such a « narrative syntagm » was formulated through the re-analysis by Greimas of Propp’s Russian folktales, and is particularly well adapted to the semiotic analysis of human action. Most human action is purposive, and the narrative syntagm treats such purposive action as the paradigm. « Purposive », however, is not here restricted to conscious purpose, much less to the conscious purpose of the actor himself. In this context, the narrative model is applicable equally to factual and fictional human action, and indeed sees the distinction between them in discursive terms : factual human action is communicated together with a truth-claim (the modality of « existence »), fictional human action is communicated not merely without such a truth claim, but with a modality of make-believe. Equally, this semiotic model sees no conceptual distinction between speech-act(ion) and other forms of action. Thus, in principle, the same model may be used to account for the intelligibility of a story (semantics) and the telling of that story (pragmatics). It thus becomes possible to argue (Jackson, 1988) that parallel but separate models may be offered of the story (told) in the trial, and the telling of that story (the story of the trial). The behaviour of each of the participants in the trial is itself intelligible in terms of contract, performance and recognition. We should therefore look systematically at the pragmatics of the trial, as an independent form of human action, in all the aspects which the semiotic model suggests to be involved.

In the argument advanced above, a distinction has been hazarded from time to time between the game (semantics) and the playing of the game (pragmatics). But such an application of the distinction between semantics and pragmatics might at first sight appear somewhat different from that normally encountered. We can clearly distinguish the content of an utterance from the act of enunciation, just as we can distinguish the story in the trial from the story of the trial. But can we distinguish, just as clearly, the game from the playing of the game ? What meaning does the game have, in isolation from its playing ?

I suggest that we can draw such a distinction. Even without playing the game, the narrative syntagm is manifest. There will be a rule-book for the game, which exists and has meaning independently of any act of playing the game. It is, quite simply, a text which hypothesises the playing of the game, creating what might be termed « virtual » subjects, performances, and recognitions. It is in just this sense of « virtual » that Greimas and Landowski (1976) speak of semiotic objects like the company, as created (mentioned) by the text of a French commercial law statute – independently of the creation of any particular company through the use of those statutory rules. Normally, the rules of a game will themselves state what the object of the game is. That institutes a virtual subject (here, player) of the game. The rules also set out the competence of the subject, here what the player may do in order to achieve the object. It states also when the game is « won », thereby manifesting « recognition » through the sending of a modal value (success) to the player who has performed the task of what the rules define as the end of the game.

So conceived, the game does indeed appear as separate from life. This appearance derives from the absence of inter-discursive claims. This absence hardly goes unnoticed. It is precisely because the discourses of everyday life continually claim to refer to each other that the absence of such inter-discursive claims appears as marked. And in the playing of the game, too, an attempt is made to preserve this sense of separation, by suppression of those inter-discursive claims which distinguish the pragmatic from the semantic level. The players, too, make the claim, sincerely believe, and grasp the sense that this is « only a game », by suppressing the manifestation of the pragmatics of playing. And in this, there is a marked parallel with the law. From the internal point of view, at least, the trial proceeds as if it were a neutral processing of the story in the trial, not an independent set of behaviours about that story.

It is this distinction between the game (semantics) and the playing of the game (pragmatics) which assists us to resolve many of the difficulties observed in the « games analogy ». For however difficult it may appear to apply the narrative syntagm to the game itself, those difficulties disappear immediately when we address the playing of the game. For now we are back in the realms of purposive human action, whether the purpose is to win in an agonistic context, or to perform the task of simulation, etc. And since we are dealing with real players, not « virtual » subjects, we cannot give an account of the meaningfulness of their behaviour which is restricted to one set of objectives, one single narrative syntagm. In practice, the players are subject to multiple contracts, just as the advocate seeks both to represent his client’s interest and advance his own career. Moreover, once we enter the pragmatic sphere, and leave that of the « virtual », the intelligibility of the action (even, here, ludic action) is mediated through narrative typifications of ludic behaviour. We know the types of move to be expected of the left back, the centre forward, and the referee. And these narrative typifications are accompanied by the transfer of modal values : we know what typical behaviour patterns are regarded as good, skillful, beautiful, effective, – even playful, entertaining, funny.

These narrative typifications of how the game is played are not identical to the « virtual » rules of the game itself. That is why Hart’s attempt to mediate between mechanical jurisprudence and scorer’s discretion fails. Hart seeks his solution purely at the level of the game itself. That game, he tells us, contains both a scoring rule and a rule of adjudication. The scorer has (presumably, under the rules) a duty to apply the scoring rule, and (though this is more difficult to locate at the level of the game itself), we may have faith that the scorer will, indeed, attempt so to act. He may occasionally make a bad decision, which will be binding, but it is only in virtue of the scoring rule itself that we can say that the decision is bad or erroneous. The alternative account here proposed permits us to develop a more realistic account. The sense of the game as played is one which admits chance elements : errors of perception, errors of judgment, even principles and policies not envisaged by the rules of the game. These do not represent the intrusion of extraneous elements. Rather, they are part of the narrative typification of the behaviour of the official, and indeed they are recognised by the transfer of modal values : team X was lucky to win (a judgment made from within the pragmatics of the playing of the game, not an external criticism of it) ; « they didn’t deserve that penalty », etc.

The modal values exchanged in the playing of the game, and the forms of recognition accorded to that playing, explain many of the features of games described above : for example, the consciousness that it is a game, and the socialising function. This last, perhaps, exemplifies particularly well the difference between recognition of success in the game itself, and recognition of the playing of the game. In the Eskimo song-contest, social re-integration occurs not because the mother-in-law is or is not a cannibal but rather because a bond is formed through their common experience of the ritual. Mutual vituperation serves the purpose of common and communal entertainment – they all laugh together.

Thus far, the analysis may appear to suggest that there is a common semiotic analysis of all games, so that we can, indeed, offer a semiotic definition of the game. But that would be to mistake the nature of the basic structures of signification. They provide necessary conditions of the generation of sense (or, to use the technical terminology, « semiotic constraints »), but they do not claim to provide sufficient conditions. Rather, the mode of manifestation of these semiotics constraints (in principle, the same semiotic constraints for games and for law) is contingent upon the particular discourse. That makes it important to define, or rather individuate, « the particular discourse ». Semiotics adopts a severely empirical test : it is not « the game » (a reification), but rather the playing of a particular game in a particular context. And within the playing of each such game, it is important to analyse separately the positions of each participant. For just as in an English criminal trial, we may have a judge, a jury, a defendant, barristers, solicitors, clerks, witnesses, spectators, each with their own contracts to perform, their own set of competences, their own modes of recognition, and their own sets of modalities expressing the form of such recognition, so too in the game of football we have the players, the referee, the managers, the substitutes, the ground staff, the directors, the spectators. What is skillful to some is not necessarily skillful for others. What is playful to some is deadly serious to others. What is open to some is concealed or compressed to others. There is not just one game going on in the football stadium ; there are many different discursive constructions of the game.

The implications of this argument for the status of the games analogy seems to be this. There is nothing special or magical in the relationship between law and play/games ; indeed, much of the relationship claimed in the literature depends upon an artificial reification of both concepts. Rather, the interest of the comparison, and the force of many of the parallels, derives simply from the fact that both make manifest the same basic structures of signification : both have semantic and pragmatic dimensions, in both of which human action is intelligible in terms of contract, performance and recognition. There is no common set of manifestations of the game, any more than there is a common set of manifestations of « the law ». Some manifestations of law will exhibit parallels to some games or forms of play, but equally parallels may be found in other spheres of human activity.

Nevertheless, comparative investigation is useful, since hypotheses developed in one context may prove plausible in another. I conclude with some semiotic observations on the question of identity – the identity of the game on the one hand, of the law on the other.

Recall, for the last time, Hart’s problem of scorer’s discretion. If « football » were played, Hart argues, on the basis that the referee could decide for himself what the scoring rules were, that would not be football, it would be a different game (« scorer’s discretion »). But Hart’s explanation of why the judge’s (or referee’s) finality of judgment does not amount to « scorer’s discretion » is not, as we have seen, satisfactory. Rather, I suggested, the solution is to be found at the level of the pragmatics of the playing of the game, where chance elements are indeed admissible. Nevertheless, there must be a distinction between such admissibility of chance elements and « scorer’s discretion » (or should it be : « scorer’s indiscretion » ?). How do we know, despite the admissibility of elements not contemplated by the rules, that we are still playing football ? The answer, as hinted already, lies in the attribution of good faith to the referee. Hart tells us that he is « under a duty » to apply the scoring rule (though, we may add, he can only do so to the limit of his competence). Being « under a duty », for Hart, entails internalisation of that duty, using the obligation imposed by the rules as a reason for action. So we have two questions to answer. First, how has this obligation been internalised ? Secondly, how is the appearance created that the referee is actually fulfilling his obligation ? The answer to the first is to be found in the training of the referee, and his internalisation of the narrative typifications of acting like a good referee. Such an internalisation will produce the consciousness of being under an obligation.

The answer to the second question is, perhaps, more controversial. The referee must not only fulfil the obligation to apply the scoring rule, to his best endeavours (though this will be the scoring rule in the context of the narrative typification of the behaviour of a good referee) ; he must also communicate the appearance of fulfilling that obligation. How can he do so ? He cannot communicate the reasons for his decisions while the game is in progress (not, at least, to the spectators) ; he is restricted to signalling the decisions themselves. Nevertheless, the spectator has narrative typifications of the good referee, communicated through dress, gesture and activity, which serve to communicate the appearance of fulfillment of the obligation. A referee who appeared in clown’s trousers, or who gestured in ways different from the accepted code, or who sat on the ground to observe the taking of a penalty, would not be taken seriously. Logically, there is no reason why such a referee might not be fulfilling his obligation to apply the scoring rule ; nevertheless, by failing to adhere to the visible conventions, he may be suspected of lack of commitment to the invisible conventions too. Conversely, the referee who does adhere to the visible conventions will be assumed, by association, to adhere also to the invisible ones. We have, here, a hypothesis for the construction of good faith by the judge. If he is seen to be playing the game, he will be assumed so to be doing. Fulfillment of the pragmatic conditions creates a presumption that the rules of the game itself are being followed.

A different aspect of the question of identity occurs when we consider the playing of the game, or the application of a particular law, in different contexts. Football may be played in a huge variety of ways, and contexts : indoor, outdoor, five-a-side, eleven-a-side, friendly, competitive, international, league, cup, football association rules, football club rules, with a referee, without a referee, with or without strict adherence to the off-side rule, etc. What is it that makes us regard all these different activities as belonging to the same game, « football » ? There is no one authoritative set of rules accepted for all these contexts, so that we cannot say that there is one game, football (semantics), but different ways of playing it (pragmatics). Nor can we say that whatever else may vary, the object of the game is to score by propelling the ball into the net, since there are other games too in which this is the scoring criterion. We may, indeed, conclude with Wittgenstein that there is a family resemblance between the different playings of football, but that does not explain the strong claim to unity that is actually made.

Exactly the same problem afflicts the law. The same law may be applied in many different ways, in different contexts, according to the pragmatics of the situation. The situation may differ in the classroom, the Magistrates’ Court, the Court of Appeal, the House of Lords, the solicitor’s office, the barrister’s chambers, the police station, and on the street (not to ask : which street ?). How do we know that it is the « same law » which is being applied ? On each occasion, it is discursively reconstructed within the context of the particular act of enunciation. Here, too, there may be adherence to a variety of statements of the rule. Nevertheless, a strong claim of unity is made. How, in both in football and the law, is the appearance of such unity constructed ?

Once again, the answer would appear to reside substantially in the construction of the good faith of the participants, along lines similar to those sketched above. But there is also, here, an additional element. The use of the common term « football », and the use of a common reference to « the law » or « the law on police powers », or even to « section 24 of the Police and Criminal Evidence Act », makes a claim to correspondence to a common, external referent. Acceptance of such a claim is incompatible with the semiotic position. Understanding of the way in which the claim is rendered plausible, on the other hand, is an essential task of semiotic analysis.

Références bibliographiques

BERNE E., 1966, Games People Play, London, Deutsch.

CAILLOIS R., 1958, Les jeux et les hommes, Paris, Le Seuil.

CONTE A., 1983, « Paradigmi d’analisi della regola in Wittgenstein », in R. Egidi (dir.), Wittgenstein. Momenti di una critica del sapere, Napoli, Guida Editore, p. 37-82.

CONTE A., 1985, « Materiali per una tipologia delle regole », Materiali per una storia della cultura giuridica, n° 15, p. 345-368.

EHRMANN J., 1968, « Homo ludens revisited », Yale French Studies, n° 41, p. 31-57.

GREIMAS A.J. and LANDOWSKI E., 1976, « Analyse sémiotique d’un discours juridique », in GREIMAS A.J., Sémiotique et sciences sociales, Paris, Le Seuil, p. 79-128.

HART H.L.A., 1961, The Concept of Law, Oxford, The Clarendon Press, 262 p. (Clarendon Law Series).

HUIZINGA J., 1949, Homo Ludens, London, Boston and Henley, Routledge and Kegan Paul, 220 p. (International Library of Sociology).

JACKSON B.S., 1972, Theft in Early Jewish Law, Oxford, Clarendon Press, 316 p.

JACKSON B.S., 1988, Law, Fact and Narrative Coherence, Merseyside, Deborah Charles Publications, 214 p. (Legal Semiotics Monographs).

MIDGLEY M., 1974, »The Game Game », Philosophy, n° 49, p. 231-253.

OST F., 1988, « Between Order and Disorder : the Game of Law », in TEUBNER G. (dir.), Autopoietic Law : a New Approach to Law and Society, Berlin, New York, Walter de Gruyter, p. 70-96.

ROBLES G., 1984, Las Reglas del Derecho y las Reglas de los Juegos, Palma, Facultad de Derecho de Palma de Mallorca, 278 p.

L’auteur

Il est né en 1944. Etudes à l’Université de Liverpool et l’Université d’Oxford (thèse de doctorat : Theft in Early Jewish Law). Il a enseigné à l’Université de Géorgie (Etats-Unis), d’Edimbourg et de Kent, ainsi qu’à la Preston Polytechnic et à la Liverpool Polytechnic, et a présenté des cours, à titre visitant, à Jérusalem, Oxford, Nanterre et Bologne. Actuellement, il est Queen Victoria Professor of Law à l’Université de Liverpool. Il a publié Theft in Early Jewish Law (1972), Essays in Jewish and Comparative Legal History (1975), Semiotics and Legal Theory (1985), et Law, Fact and Narrative Coherence (1988), et il a dirigé, depuis 1978, The Jewish Law Annual. Il est Secrétaire général de l’Association Internationale de Sémiotique Juridique et l’un des principaux animateurs de la Revue Internationale de Sémiotique Juridique, ainsi que de Law and Critique.

Les protagonistes du débats

Herbert Lionel Adolphus HART (1907-)

Born in the U.K. and educated at Oxford, he practiced at the Chancery Bar from 1932-40, and from 1939-45 served in the British War Office. Returning to Oxford in 1945, he became Professor of Jurisprudence in 1945 (-1968) and later Principal of Brasenose College. Closely associated, especially in the early years, with linguistic philosophy, his best-known works include Causation in the Law (with A.M. Honoré), Oxford, Clarendon Press, 1959, Law, Liberty and Morality, Oxford, Oxford University Press, 1960, The Concept of Law, Oxford, Clarendon Press, 1961, Punishment and Responsibility, Oxford, Clarendon Press, 1962, Essays on Bentham, Oxford, Clarendon Press, 1982, and Essays in Jurisprudence and Philosophy, Oxford, Clarendon Press, 1983.

Johan HUIZINGA (1872-1945)

Born in the Netherlands and educated at the University of Groningen, he went on to study Indo-Germanic linguistics at the University of Leipzig, before returning to Groningen to take his Ph.D. in 1897. He became Privaat-Docent of Indian Literature at the University of Amsterdam, then Professor of History at the University of Groningen (1905-1915) and ultimately Professor of General History at the University of Leyden (1915-1945). Apart from Homo Ludens, he also wrote a biography of Erasmus, various sociological works, and is perhaps best known for his The Waning of the Middle Ages, London, Arnold, 1924.

Pour en savoir plus...

Sur la socio-sémiotique parisienne en général, voir : A.J. GREIMAS, Sémiotique et sciences sociales, Paris, Seuil, 1976 ; Id, On Meaning, Minneapolis, University of Minnesota Press, 1987 ; A.J. GREIMAS ET E. LANDOWSKI (eds.), Introduction à l’analyse du discours en sciences sociales, Paris, Hachette, 1979 ; E. LANDOWSKI, La société réfléchie, Paris, Le Seuil, 1989.

Sur la sémiotique du droit, selon l’école parisienne, voir : B.S. JACKSON, Semiotics and Legal Theory, London, Routledge and Kegan Paul, 1985, esp. Pt.II ; et (on the issue of consciousness), « Rationalité consciente et inconsciente dans la théorie du droit et la science juridique », Revue Interdisciplinaire d’Etudes Juridiques n° 19, 1987, p. 1-18 ; E. LANDOWSKI, « Sémiologie juridique », « Sémiotique juridique », In Dictionnaire de théorie et de sociologie du droit, A.J. ARNAUD (ed.), Paris-Bruxelles, L.G.D.J. and E.Story-Scientia, 1988, p. 368-372 ; Id, « Towards a Semiotic and Narrative Approach to Law », International Journal for the Semiotics of Law /Revue Internationale de Sémiotique Juridique n° I/1, 1988, p. 79-105 ; Id, « Sémiotique du droit : interdisciplinarité et pertinence », Droit et Société n° 21, 1988, p. 85-90.

* Queen Victoria Professor of Law – Université de Liverpool, U.K.

1. Throughout this article, by « games analogy » I mean the analogy of games to law.

2. = soccer ; throughout this argument, I use « football » in the British rather than the American sense.

3. Football and chess – the two games most frequently compared to law by Hart – are significantly different in this respect. In chess, the decision of the judge is exceptional, residual, and though elements of chance may not perhaps be totally excluded (the state of Bobby Fischer’s digestion on the day of a particular championship ?), they are not integral to that particular game in the way in which they are to football.

4. Conte (1983, 1985), in a sharper analytical distinction of types of rules, would say : « eidetico-constitutive ». For further analysis of such questions, see Robles (1984).

5. On the Bible, at least, this was not the standard procedure, but rather was reserved for cases of special difficulty – Jackson, 1972, p. 233-244.

6. What I have termed elsewhere the « story of the trial » as against the « story in the trial » : 1988, p. 88.

7. This corresponds to the passage from Huizinga quoted by Ehrmann (1968, p. 34), which appears to be an English translation from the French edition, the latter being a translation from the Dutch.

8. I take this to be the meaning here of « ilinx, vertigo ».

9. Point 1, quoted above.

10. Quoted in Ehrmann (1968, p. 38), a view often associated with sports (but called « games ») in traditional British schools.

11. Or seek to establish : recall the ping pong diplomacy of some years ago.