Issues in Contemporary Legal Philosophy: The Influence of H.L.A Hart, Oxford

In the context of law they have not received, until recently, the systematic attention that they deserve. Partial answers to these questions could be glimpsed from the arguments brought by theorists of law against other theories. Hart's analysis of a variety of legal theories and their inadequacies is justly seen as a classic in this kind of work.² The time has come to conduct discussions about the nature of legal theory in a more systematic way, and Dworkin, by going beyond merely proposing a new theory of law and looking into the nature of the enterprise of theorizing about law, has taken an important step in this direction.

It is almost a truism to say that a good theory of law should give us insight into the nature of law and its working and an adequate description of this complex social institution. By now we all know that the data to be adequately illuminated are themselves complex and full of apparent contradictions: as Hart pointed out, we know how to identify law, or at least the relevant sources of law, yet there is a continuing controversy about the nature of law and its meaning, and about the content of the law in particular cases. We use the word 'law' in a variety of meanings and contexts. We may say, for example, that in 1944 a certain law was passed in the German Parliament. We may also say, in a different context, that there was no law in Germany in those years. Another standard problem of jurisprudence is an adequate account of adjudication: people go to court in order to have an authoritative decision on their dispute, and are willing to accept the verdict of the law, assuming this verdict is clear. Yet they often find that lawyers and judges disagree about what the law is. A good theory of law, we hope, should explain and give a coherent account of these facts and apparent contradictions.

For a long time Dworkin has been criticizing what he calls the 'ruling theory of law', which is in part Hart's version of legal positivism. The novelty of his present approach is that he seems to couch his critique of contemporary theorizing about law in more general terms. Dworkin now argues not simply that there are in the field misguided and misleading theories of law, and that he can propose a better one; he claims that there is something wrong and misguided about our perception of the whole purpose of seeking a theory about law. Among the misguided theories, which he dubs 'semantic theories of law', are the positivism of both Austin and Hart, legal realism, and natural law theories. In the paper before us Dworkin argues in favour of a new perception of the enterprise: articulating theories about law should be seen as an interpretive process, law should be seen as an interpretive concept. All conceptions (theories) of law should be seen as the products of such a process, and the process is a kind of general constraint on all theories of law. By implication, semantic theories of law do not meet these initial constraints. Dworkin then proceeds to sketch his own proposed conception of law.

There is a great deal one can say about Dworkin's own conception of law, but in this comment I shall limit myself to Dworkin's claims about the nature of theorizing about the law.

It is difficult to assess Dworkin's claims on the basis of a partial elaboration, as the one before us inevitably is. None the less I shall take a risk and claim that although Dworkin's analysis contains, as usual, many valuable insights, its implications are less radical than he would like us to believe. All theories of law might indeed be the products of a process of interpretation, broadly speaking. But there are many types of possible interpretations, and Dworkin is mistaken in thinking that only one sort is applicable to concepts such as law; 'semantic theories' cannot be dismissed in such a cavalier fashion (and Dworkin knows it); the virtues of interpretation are exaggerated: the process, in itself, cannot overcome the difficulties Dworkin has pointed out. And the elegance of this broad claim has lured Dworkin into disregarding some important distinctions.

Anyone familiar with Dworkin's writing knows how difficult it is to comment on his work, since it is always full of interesting details, which may not be crucial to the argument, but call for some acknowledgement none the less. I shall try to strike a balance between picking out these details and getting at the substance of his claims.

The sense of breakthrough in Dworkin's piece is generated by his success in identifying a basic flaw in all legal theorizing to date. I believe this perception is misleading. Let me start with two relatively small points, designed to support Dworkin's claim. It is important to do so before we come to the substance of Dworkin's claims, so that we may gain a clearer picture of the state of the art of theorizing about law which is the subject of Dworkin's paper.

I. THE CRITIQUE OF SEMANTIC THEORIES

Dworkin argues that all previous attempts to provide a theory of law share one common feature: they are attempts to provide and articulate criteria for the use of the word 'law', criteria which will help us in making sense of our attempts to identify the law, especially in hard cases. He calls such theories 'semantic'. Dworkin argues forcefully that semantic theories cannot account for the nature of disagreements about the law, and he thus concludes that all legal theorizing has so far been misguided.³ I believe Dworkin is constructing a straw man here. Classical legal theories have always been attempts at understanding the phenomenon of law not at defining the word 'law'. Usually, they were not directly concerned with the resolution of particular cases, either hard or easy. Legal theory was classically seen as an attempt to identify the features which make law a unique social institution, differentiated from morality and religion. The linguistic formulations of the question concealed a belief that linguistic usages reflect social reality in some way, and that there was something to be gained from attending to them. But the search has always been the same: an attempt to understand the nature of law. Thus it is not surprising that the discarded semantic theories return as conceptions of law under Dworkin's analysis. And it is similarly not surprising that Dworkin has to continue his debate with positivism although he seemed to prove its basic inadequacy by simply showing it to be a semantic theory.

In short, I do not think positivism, realism, and natural law theories are semantic theories of law. Thus Dworkin has not shown that the enterprise has been misguided. (It is curious to note, however, that even if they had been semantic theories, they would not necessarily have some of the flaws Dworkin ascribes to them: Dworkin holds against semantic theories that they advocate a unified use of 'law' across countries and periods, and are thus inadequate. The literature about translations and time-sensitive theories of meaning is wide enough to suggest that such theories might not be so 'blind' after all. I shall return later to Dworkin's claim that an adequate legal theory must be time- and community-sensitive.)

II. CONCEPT AND CONCEPTION

Another general complaint of Dworkin's is that theorists seek to resolve, on the level of a concept of law, questions which should be dealt with only on the level of conceptions of law. According to Dworkin we mistakenly believed that seeking a concept of law was something illuminating, which should provide some tentative answers to some basic questions of jurisprudence (for example that of the relationship between law and morals). It now seems as if we were chasing a wild dream. Such questions can only be resolved by conceptions of law, and regarding them as 'conceptual' is thus a category mistake.⁴

If we accept, with Dworkin, that the concept of law is merely a non-controversial statement of a consensus, then Dworkin is tautologically right; according to this analysis there cannot be a conceptual contro­versy by definition. Yet if one wishes to draw the concept/conception distinction as one between a root principle and instantiations of it (the way Perelman did with justice), there might be controversy on the concept, and it might be fruitful to present some of the central controversies in the field as conceptual ones.

Whatever we may think of the right way of drawing the distinction, Dworkin has not shown that a category mistake was indeed committed. In fact, he himself lapses, at times, into the convenient way of describing these central controversies as 'conceptual'.

These two points are not crucial. Yet it is important to make them before we turn to Dworkin's suggestions as to the 'true' nature of theorizing about the law. They show that the field is not empty, after all.

To argue convincingly that our conception of the enterprise of theorizing about the law should change, Dworkin must show that the old kind of theory does not enable us to do what we can do under his kind; and that his kind of theory helps us avoid some of the immanent difficulties of the kind of theory he wishes to replace. I am afraid Dworkin fails on both grounds. Dworkin does succeed in a less ambitious claim: his analysis of interpretation deepens our understand­ing of adjudication (although not of the attempt of the social theorist to articulate a concept or conceptions of law).

Ill. LEGAL THEORIES AND ANSWERING LEGAL QUESTIONS

Usually, when people think of a theory of law, they think of an attempt to identify the features which are unique to law and distinguish it from other social phenomena. This was clearly the conception shared by Austin and Hart. At times it seems that Dworkin accepts this concep­tion. Yet at others it becomes clear that one of his theses is that a theory of law in this sense is just a part of legal practice, and that legal theory should make a difference to the way judges decide hard cases. For Dworkin, the social scientist's attempt to understand what law is and the judge's attempt to identify the law applicable to a case are questions of the same nature. The second naturally requires a much more detailed elaboration. But, basically, both are questions of an identical process of interpretation. (It has become quite fashionable lately to discuss social phenomena in terms of interpretation, but for Dworkin there is a special point in this choice of terms: hard cases are usually described by lawyers as raising questions of interpretation.)

Dworkin's approach is based on the fact that judges in hard cases argue about what the law is, that is, identification of the law of their system is an indispensable part of their work. There is no doubt that the identification of one's law on a question is central to adjudication, and that answering it involves many theoretical considerations. However, these are not considerations of legal theory, in the limited sense described above. The essence of law is not affected by the question of whether a particular law in a particular system severs the relationship between a murderer and his inheritance. Whatever the law's position on this question, a theory of law in this limited sense can 'live' with it. It follows that the theory, in itself, cannot dictate any answer to the question of what the law says or should say on such an issue. It is a question of law within this system, not a question of the general theory of law.

A possible source of confusion in this respect is that at least some of the theoretical considerations involved in solving hard cases are questions of legislative intent. Some of these questions are conceptual, and thus are not a part of any particular legal system. The logical meaning of application of rules and some of the difficulties with the idea of identifying legislative intent are indeed general, and they have to be used in interpreting particular laws in particular cases. But they are not, in the sense just mentioned, a part of an attempt to identify the unique features of law. Such theoretical analyses will form part of the general theory of law, and will apply with equal force to all legal systems. In this they are different from the traditions in a given legal system concerning interpretation or the force of precedent. These latter doctrines are clearly a part of the law, not of the general theory of law, although they may be fruitfully analysed in terms of those general theoretical observations.⁵

In other words, Dworkin was weakening considerably the distinction between law and theory of law, a central point in the debate between Hart and Bodenheimer.⁶ The distinction is not always easy to make, but I believe that it is worth making because the law is what is binding on us, and a theory of law is our way of understanding what is binding us. The law will be binding even if we do not understand it. The rule of literal interpretation may be binding on us by precedent even if we find that this rule is not logically coherent. It is satisfying if our doctrines always fit our logic, but there is no guarantee that they must. It is useful, therefore, to maintain the distinction between the law and the theory.

I believe this confusion is intensified by the fact that both theories of law and its interpretation are, in effect, attempts at the identification of law. But here again the difference is clear: the identification of the law on a given question within a legal system is a process different from the social scientist's attempt to differentiate between law and other social phenomena.

This confusion of the identification of law in the process of adjudica­tion and its identification in theoretical analyses of law has a number of important implications, some of which are problematical. The most central, to which I shall return in the next section, is the question of whether theories of law are normative or descriptive. Law itself is clearly normative, and if a theory of law is only the general part of law, it follows that it, too, is normative. In fact, Dworkin explicitly sees this conclusion as one of the virtues of his conception of a legal theory. I shall argue that it is a major disadvantage.

A second implication is that theories of law are time- and community-dependent. The more abstract parts may be applicable more widely, but basically a theory of law is a theory of the law of a community at a certain time. This implication follows not only from Dworkin's continuity between law and theory of law, but also from his conception of the interpretive process itself: it is one of the essences of this process that it is done periodically, and its product changes with the content of the practice interpreted. Dworkin specifically says that there is no reason to expect that a theory of law, no matter how abstract, could be an adequate interpretation of systems called 'law' in all societies at all times. Again, I believe that by following Dworkin on this point we are unnecessarily relinquishing a valuable tool.

IV. LEGAL THEORY: DESCRIPTION OR JUSTIFICATION?

One important implication of Dworkin's analysis is that the process of reaching a conception of law becomes necessarily 'reforming', since the last, or 'post-interpretive stage' in the exercise, which is none the less a part of interpretation as a whole, is to reduce the discrepancies between the pre-interpretive data, and the interpretive stage in which the practice is analysed in terms of its point. In other words, the conceptions of law yielded by Dworkin's interpretation are necessarily evaluative. The same conclusion follows from Dworkin's suggested concept of law, seeing law as a justification of the use of the coercive force of the community.

There is no need here to reopen the old debate within jurisprudence on the value and possibility of non-evaluative conceptions of law. I still adhere to the view that such conceptions are both possible and useful, even indispensable if we are to provide an adequate analysis of law. I have argued that Finnis in effect accepts this,⁷ and I now argue that Dworkin too has to do so. He agrees that there are contexts in which the most useful way of expressing a stand on the law is by using a non-evaluative (pre-interpretive, by him) conception. We must concede that one should not advocate a system of articulating conceptions of law which inevitably excludes such conceptions. As Dworkin himself rightly insists, we should not pre-empt important questions by delimiting the rules of the game.

V. THE UNIVERSALITY OF CONCEPTIONS OF LAW

Dworkin's analysis of law as an interpretive concept, under his own model of interpretation, makes classical legal theory impossible. In other words, Dworkin is challenging here the possibility and the value of 'general jurisprudence', the attempt to analyse societies at their most general (like the attempt to analyse human nature) and identify those features common to all social organizations which might lead to the need for similar institutions and practices. Law is one such social institution found in all societies and exhibiting a core of similar features. The advantages of having a universal conception of law, if it is possible, are many. The constructs of scholars such as Austin, Hart, Kelsen, and Weber, to mention just a few, were explicitly meant to be universal in this sense. I do not think it has been shown that this undertaking is impossible. On the contrary: the intellectual history of legal philosophy is full of insights gained within this kind of enterprise. So for Dworkin to propose a concept of law that will be necessarily local seems a disadvantage. What is the sense of talking about international law, for example? Why give up such an important tool of comparative sociology? It may well be that such a universal concept will have to be very general and abstract, and will not be too useful for deciding questions of law in any given community. This will again strengthen the distinction between law and legal theory which Dworkin finds wrong and misleading. But in the spirit of conceptual pluralism favoured by Dworkin it seems wrong to exclude, at the outset, through the definition of the process of interpretation, the possibility of an 'ideal type' of law of a kind that has yielded theoretical benefits in the past.

It should be added that this debate about the utility and possibility of 'ideal types' is not special to law, and indeed has been raging in sociology and psychology for quite some time. The universality of the conception reflects much more than the historical fact that different conceptions were called by the same name, and are thus strands in the same rope, as

Dworkin would have it. One does not have to be an essentialist to argue that many valuable insights may be lost if we do not, from time to time, compare our ways of dealing with human and social problems with the ways of others. I shall risk being called a 'semantic scholar' and say that for these comparative lessons to be drawn we need some universal conceptual scheme, in terms of such basic legal functions such as dispute resolution, which may help us both to understand the similari­ties and highlight the differences between societies.

Even if we are not sure that such concepts are useful for all practical purposes, why adopt constraints on theorizing which will prevent us from using them when the need arises? Dworkin does not provide an answer to this question.

These two points show that we lose something by accepting Dwor-kin's invitation. What are, according to Dworkin, our main gains? First and foremost is a better understanding of adjudication, and through it, of the role law plays in society.

VI. THE NATURE OF DISAGREEMENTS BETWEEN JUDGES

Anyone familiar with Dworkin's work will not be surprised to learn that one of his starting points, one of the most important pre-interpretive data he identifies and seeks to explain, is the fact that there are often disagreements between judges. Furthermore, judges present these disagreements as ones about the law. If we take them seriously, we must give an account of law under which law can be controversial. Dworkin's main complaint against semantic theories of law, primarily positivism, is precisely that it refuses to allow that there may be non-empirical controversy about the identification of law. According to positivism, law consists of the pre-existing standards identified by origin or sources. Most disagreements arise in cases where the law, thus defined, does not settle the question. The judge, therefore, must go 'beyond the law' to decide the issue before him. According to Dworkin, positivism must deny that these disagreements are genuinely about the law and it must label as simple minded or fraudulent those in the legal community who claim that this is their nature.

One of the main advantages of Dworkin's picture of conceptions of law as resulting from processes of interpretation is that conceptions of law are, according to him, controversial. Thus disagreements about these conceptions, which affect judicial decisions in hard cases, are disagreements about the law. Dworkin's theory thus does justice to the way lawyers and judges talk about disagreements without seeing them as either cheaters or simpletons.

I believe Dworkin is basically right in contending that what judges disagree about is, in one sense, the law. Judges do indeed interpret the law, and I find Dworkin's model of interpretation quite illuminating: I accept that in this context of interpretation there is an inherent and inevitable reforming stage, and that this is part of the interpretive process. That is why judges claim that they are arguing about the law, and rightly so. What they are doing is al! within their legal activity, and one should not forget this. Dworkin succeeds in drawing our attention to the structure of argument about the law, in contexts where the question is the identification of the content of the law relevant to a particular issue. (Dworkin's analysis is confined to 'static' interpreta­tions and developments of the law. We should remember that most laws, unlike new rules of courtesy, are created by legislatures, in a 'dynamic' way. This is another argument for maintaining the distinction between law and legal theory.)

However, a basic rule with theory formation is that it is wise to see first if the available theories, which have made a contribution to our understanding of a phenomenon, may be adapted to the change. I believe they can, and that this adaptation is more fruitful than the total substitution Dworkin suggests. One of the virtues of the adaptation is that it does not conceal the insights gained by the 'old' theory, discarded by Dworkin's new emphases.

Adjudication is no doubt a complex social practice. In hard cases it involves interpretation. Let us accept Dworkin's model of this process. He sees that, analytically speaking, three distinct stages are involved. The question is really what we should call 'law', in a particular context: the post-interpretive conclusion of the majority, or the pre-interpretive data shared by all judges? Once we see that we have a problem of naming, much of the intensity of the debate disappears. Surely it is important to emphasize that the move from the first stage to the third involves judicial responsibilities? Surely it is important to point out that the shared pre-interpretive sources are partly what gives the whole enterprise its legitimacy? Surely it is important to be aware that the parties may be required to know the law of the first stage, and only speculate about the law of the third stage?

It follows, then, that both pre-interpretive law (the law identified by sources, or 'positivists'' law, and post-interpretive law (Dworkin's law) are important entities. But I want to go further than argue that Legal Theory and the Problem of Sense emphasizing 'first stage law' is legitimate. I take it to be an important insight of Hart, elaborated by Raz, that what the latter calls the 'sources thesis' is not merely a contingent feature of law.8 A more comprehensive discussion of this point will clearly go beyond a comment, but I trust that the thrust of the argument is familiar. Thus the loss of this emphasis is not merely the loss of some feature of law. It may lead us to underrate one of the important ways in which law as we know it fulfils its social functions.

One last comment is called for: Dworkin seeks to enlist our sympathies for his views by arguing that his opponents must ascribe to judges and lawyers either simple-mindedness or fraud. I agree with Dworkin that a theory which takes what judges say as reflecting what they think, and assumes that judges know what they are doing, should be more attractive than one resting on such ascriptions. However, it is not clear which of the two approaches takes more seriously the way judges feel about what they are doing. When you compare the words of candid judges out of court to their exercises in interpretation, you see that many of them fight as scholars against the rhetorical myth they must perpetuate as judges.9 Dworkin has not linked his position on the nature of legal theory to his famous one-right-answer thesis of adjudication (a task which may prove difficult), but Dworkin himself provided the framework within which judges may make sense of what they are doing, so that they acknowledge both that their disagreements are ones about the third-stage law, and that in deciding hard cases they must go, in a sense, beyond the first-stage law.

I have enumerated above some classical problems which have been seen as crucial indicators of the adequacy of theories of law. Does Dworkin's approach improve our ability to understand these questions and deal with them? It should be remembered that the test here is a strong one, since courtesy requires that existing theories should not be stigmatized as totally inadequate unless our ability to deal with issues under the newly proposed theory is substantially improved. A mere showing that a new way of looking at things is not less helpful will not do for this purpose. I believe that Dworkin's approach is acceptable on these grounds, in the sense that it allows us to discuss these questions, but it does not seem to be superior to existing accounts. Moreover, there is a price to be paid in terms of our ability to deal with some of the issues.

VII. GROUNDS OF LAW AND FORCE OF LAW

Clearly one of the most important questions of legal and political philosophy is where to draw the line between the identification of what the law says and the moral decision of whether the law should be obeyed. One of Dworkin's initial points of criticism was that existing legal theories present many questions of law as if they were questions of fidelity to law. Yet, although Dworkin's account of the obligation to obey the law and of the division of labour between lawyers and political scientists on this matter is very illuminating, it does not differ substantially from that suggested by other scholars, no matter what their position on the nature of legal theory and on what is the best legal theory. Thus, Dworkin's specific legal theory does not offer us a better discussion of this question than is available to us under existing accounts.

VIII. WICKED LEGAL SYSTEMS

Another issue which has tested legal theories in recent history is the treatment of legal systems so wicked that they do not meet minimal standards of morality. Hart's defence of positivism was, in part, its better ability to deal with such problems: unlike some versions of natural law theories, it granted such systems the name of 'law' but denied that there was a duty of fidelity to that law, on moral grounds.'0 Fuller, on the other hand, thought a lot could be gained from a legal theory that would deny such systems the name of 'law', and would thus aid in severing the link between the system and our inclinations and sense of duty to obey."

Dworkin's approach suggests that an important distinction should be made between a conception of one's own legal system and one's conception of another legal system. I believe the distinction is indeed important, and is similar to the difference, pointed out by Hart himself, between the relationship of acceptance of one's system and the recording of rules of conduct by an anthropologist. Yet I fail to see how this distinction improves our ability to deal with the theoretical or the practical questions raised by wicked legal systems in ways not already put forward by others.

IX. CRITERIA FOR ADEQUACY:

JUSTIFICATION OF DECISIONS IN HARD CASES

I shall conclude with another claim made by Dworkin, which has been referred to above. Dworkin asserts that legal theories are committed theories, that their cutting edge is the guidance of judges in hard cases. It is implied that one of the criteria of adequacy of a legal theory should be its conduciveness to the making of 'good' decisions. As mentioned above, this implication is related to Dworkin's method of deriving conceptions of law and to his refusal to draw a line between law and legal theory.

Legal theory at its most abstract does not help the judge in hard cases simply because its statements are not sufficiently specific to do so. I accept that articulations of theories of precedent and legislation are, as Dworkin says, both a part of legal theory and legal system-dependent, and that such theories may guide answers to hard cases, or at least provide rationalizations to decisions in such cases. Since previous decisions in hard cases form a part of the pre-interpretive data of this particular legal system, these doctrines are affected by such decisions in hard cases. I doubt, however, that even a theory of precedent or a theory of legislation, in itself, could dictate any answer to hard cases. In all the sample cases Dworkin cites we can easily present the disagreement in ways other than as a controversy about the force of precedent. As Dworkin himself indicates, the judgement of the majority in Riggs v. Palmer does not have to be based on the rejection of a literal interpretation of the statute. The best account may well be the idea that statutes should be read against a background of principles. Here, too, I do not see where Dworkin's new conceptual apparatus or approach aids us in understanding or talking about legal phenomena, including adjudication, in ways which are not compatible with existing theories of law. More important still, Dworkin's discussion does not support his view that theories of precedent or adjudication should be conclusively evaluated by their direct contribution to good decisions in particular hard cases.

X. CONCLUSION

To sum up, then, I agree that both the articulation of a concept of law and more detailed conceptions of law and of legal doctrines may profitably be seen as exercises in interpretation, taken in a broad sense. Articulating ideal types may be seen as interpreting reality, but the considerations of this kind of interpretation are cognitive and theoreti­cal, not practical as they must be in the context of adjudication. I believe it is extremely important to remember that there may be different kinds of interpretations. In some contexts we are interested in understanding, not in reform; even when seeking understanding, there are many different points of view which may dictate different emphases and approaches. A general methodology of articulating conceptions of law should heed these distinctions.

Generally speaking, criteria of adequacy for theories of law cannot be uniform. Adequacy is a relative idea: we must always know the tasks we want the theory to fulfill in order to judge its adequacy. For a long time legal theorists sought legal theories without being sensitive to this plurality of tasks. Many sterile and barren debates resulted. Today we may be ready to give up hope of finding a uniformly valid legal theory and proceed to the less ambitious but more promising job of articulating conceptions of law for particular purposes, and joining them when we can to a complex picture of law. 'Classical' legal theories, Hart's included, have not been successfully dismissed by Dworkin. They still offer us many of the enduring, and possibly eternal, perspectives of analysing law.

Notes

² See especially his discussion of Austin and the Realists in Concept and essays 12—16 in Essays.

³ This point is put somewhat differently in Dworkin's contribution to the book, but his original thrust is still reflected in his reference to 'semantic' rules of identification.

⁴ This point has also suffered in the abridgement. Both concepts and conceptions of law are, according to him, products of interpretation. Yet the 'concept' reflects the agreed features of the practice of law, and the conceptions reflect different, more complete and detailed theories of law.

⁵ For a good illustration of a logical discussion of the notion of rule-following and interpretation see e.g. M. Moore, 'The Semantics ofjudging', (1981) Southern California Law Review, vol. 4 (1981), 151. Although such discussion is not system-dependent, and it might be presupposed by all law andlegal theory, it is not a part of what is classically called 'general jurisprudence'.

⁶ Bodenheimer, 'Modern Analytical Jurisprudence and the Limits of its Usefulness', University of Pennsylvania Law Review, vol. 104 (1956), 1080—6; and H. L. A. Hart, 'Analytical Jurisprudence in Mid-Twentieth Century', U. Pa. L. Rev., vol. 105 (1957), 953-75-

⁷ See my review of Raz and Finnis 'Positivism, Natural Law and the Limits of Jurisprudence: A Modern Round', Yale Law Journal vol. 91 (1981), 1250.

⁸ See Raz, The Authority of Law, Oxford (1979), chaps. 3, 4; and The Concept of a Legal System, 2nd edn., Oxford (1980), postscript, 210—16.

⁹ Many judges have found the tension difficult. Cardozo's classic The Nature of the Judicial Process, New Haven (1923) is said to have passed without calls for his resignation only because of Cardozo's undoubted integrity and stature as a judge.

¹⁰ H. L. A. Hart, 'Positivism and the Separation of Law and Morals', Harvard Law Reviemvol. 71 1958, 593, reprinted inEssays, 49, 72-8.

¹¹ See Fuller's response to Hart in Harvard Law Review, vol. 71 (1958), 630 and especially his treatment of the subject in The Morality of Lam, revised edn. New Haven (1969).