POSTSCRIPT

INTRODUCTORY

This book was first published thirty-two years ago. Since then jurisprudence and philosophy have come much closer together and the subject of legal theory has been greatly developed both in this country and in the United States. I would like to think that this book helped to stimulate this development even if among academic lawyers and philosophers, critics of its main doctrines have been at least as numerous as converts to them. However that may be, it is the case that though I originally wrote the book with English under graduate readers in mind, it has achieved a much wider circulation and has generated a vast subsidiary literature of critical comment in the English-speaking world and in several countries where translations of the book have appeared. Much of this critical literature consists of articles in legal and philosophical journals, but in addition a number of important books have been published in which various doctrines of this book have been taken as targets for criticism and a starting-point for the exposition of the critics’ own legal theories.

Though I have fired a few shots across the bows of some of my critics, notably the late Professor Lon Fuller¹ and Professor R. M. Dworkin,² I have hitherto made no general comprehensive reply to any of them; I have preferred to watch and learn from a most instructive running debate in which some of the critics have differed from others as much as they have differed from me. But in this Postscript I attempt to reply to some of the wide-ranging criticisms urged by Dworkin in many of the seminal articles collected in his Taking Rights Seriously (1977) and A Matter of Principle (1985) and in his book Law’s Empire (1986).³ I focus in this Postscript mainly on Dworkin’s criticisms because he has not only argued that nearly all the distinctive theses of this book are radically mistaken, but he has called in question the whole conception of legal theory and of what it should do which is implicit in the book. Dworkin’s arguments against the main themes of the book have been broadly consistent over the years, but there have been some important changes both in the substance of some arguments and in the terminology in which they are expressed. Some of his criticisms which were prominent in his earlier essays do not appear in his later work, though they have not been explicitly withdrawn. Such earlier criticisms have, however, gained a wide currency and are very influential, and I have therefore thought fit to reply to them as well as to his later criticisms. The first and longer section of this Postscript is concerned with Dworkin’s arguments. But I consider in a second section the claims of a number of other critics that in my exposition of some of my theses there are not only obscurities and inaccuracies but at certain points actual incoherence and contradiction.⁴ Here I have to admit that in more instances than I care to contemplate my critics have been right and I take the opportunity of this Postscript to clarify what is obscure, and to revise what I originally wrote where it is incoherent or contradictory.

1. THE NATURE OF LEGAL THEORY

My aim in this book was to provide a theory of what law is which is both general and descriptive. It is general in the sense that it is not tied to any particular legal system or legal culture, but seeks to give an explanatory and clarifying account of law as a complex social and political institution with a rule-governed (and in that sense ‘normative’) aspect. This (p. 240) institution, in spite of many variations in different cultures and in different times, has taken the same general form and structure, though many misunderstandings and obscuring myths, calling for clarification, have clustered round it. The starting-point for this clarificatory task is the widespread common knowledge of the salient features of a modern municipal legal system which on page 3 of this book I attribute to any educated man. My account is descriptive in that it is morally neutral and has no justificatory aims: it does not seek to justify or commend on moral or other grounds the forms and structures which appear in my general account of law, though a clear understanding of these is, I think, an important preliminary to any useful moral criticism of law.

As a means of carrying out this descriptive enterprise my book makes repeated use of a number of concepts such as duty-imposing rules, power-conferring rules, rules of recognition, rules of change, acceptance of rules, internal and external points of view, internal and external statements, and legal validity. These concepts focus attention on elements in terms of which a variety of legal institutions and legal practices may be illuminatingly analysed and answers may be given to questions, concerning the general nature of law, which reflection on these institutions and practices has prompted. These include such questions as: What are rules? How do rules differ from mere habits or regularities of behaviour? Are there radically different types of legal rules? How may rules be related? What is it for rules to form a system? How are legal rules, and the authority they have, related on the one hand to threats, and on the other to moral requirements?⁵

Legal theory conceived in this manner as both descriptive and general is a radically different enterprise from Dworkin’s conception of legal theory (or ‘jurisprudence’ as he often terms it) as in part evaluative and justificatory and as ‘addressed to a particular legal culture’,⁶ which is usually the theorist’s own and in Dworkin’s case is that of Anglo-American law. The central task of legal theory so conceived is termed by Dworkin ‘interpretive’⁷ and is partly evaluative, since it consists in the identification of the principles which both best ‘fit’ or cohere with the settled law and legal practices of a legal system and also provide the best moral justification for them, thus showing the law ‘in its best light’.⁸ For Dworkin the principles thus identified are not only parts of a theory of the law but are also implicit parts of the law itself. So for him ‘Jurisprudence is the general part of adjudication, silent prologue to any decision at law’.⁹ In his earlier work such principles were designated simply as ‘the soundest theory of law’,¹⁰ but in his latest work, Law’s Empire, he characterizes these principles and the particular propositions of law which follow from them as law in an ‘interpretive sense’. The settled legal practices or paradigms of law which such interpretive theory is to interpret are termed by Dworkin ‘preinterpretive’,¹¹ and a theorist is taken to have no difficulty and no theoretical task to perform in identifying such preinterpretive data since they are settled as a matter of the general consensus of the lawyers of particular legal systems.¹²

It is not obvious why there should be or indeed could be any significant conflict between enterprises so different as my own and Dworkin’s conceptions of legal theory. Thus much of Dworkin’s work, including Law’s Empire, is devoted to the elaboration of the comparative merits of three different accounts of the way in which law (‘past political decisions’)¹³ justifies coercion, and so yields three different forms of legal theory which he calls ‘conventionalism’, ‘legal pragmatism’, and ‘law as integrity’.¹⁴ All that he writes about these three types of theory is of great interest and importance as contributions to an evaluative justificatory jurisprudence and I am not concerned to dispute his elaboration of these interpretive ideas¹⁵ except in so far as he claims that positivist legal theory such as that presented in this book can be illuminatingly re-stated as such an interpretive theory. This latter claim is in my view mistaken and I give below my reasons for objecting to any such interpretive version of my theory.

But in his books Dworkin appears to rule out general and descriptive legal theory as misguided or at best simply useless. ‘Useful theories of law’, he says, are ‘interpretive of a particular stage of a historically developing practice’¹⁶ and he had earlier written that ‘the flat distinction between description and evaluation’ has ‘enfeebled legal theory’.¹⁷

I find it hard to follow Dworkin’s precise reasons for rejecting descriptive legal theory or ‘jurisprudence’ as he often calls it. His central objection seems to be that legal theory must take account of an internal perspective on the law which is the viewpoint of an insider or participant in a legal system, and no adequate account of this internal perspective can be provided by a descriptive theory whose viewpoint is not that of a participant but that of an external observer.¹⁸ But there is in fact nothing in the project of a descriptive jurisprudence as exemplified in my book to preclude a non-participant external observer from describing the ways in which participants view the law from such an internal point of view. So I explained in this book at some length that participants manifest their internal point of view in accepting the law as providing guides to their conduct and standards of criticism. Of course a descriptive legal theorist does not as such himself share the participants’ acceptance of the law in these ways, but he can and should describe such acceptance, as indeed I have attempted to do in this book. It is true that for this purpose the descriptive legal theorist must understand what it is to adopt the internal point of view and in that limited sense he must be able to put himself in the place of an insider; but this is not to accept the law or share or endorse the insider’s internal point of view or in any other way to surrender his descriptive stance.

Dworkin in his criticism of descriptive jurisprudence seems to rule out this obvious possibility of an external observer taking account in this descriptive way of a participant’s internal viewpoint since, as I have said, he identifies jurisprudence as ‘the general part of adjudication’, and this is to treat jurisprudence or legal theory as itself a part of a system’s law seen from the internal viewpoint of its judicial participants. But the descriptive legal theorist may understand and describe the insider’s internal perspective on the law without adopting or sharing it. Even if (as Neil MacCormick¹⁹ and many other critics have argued) the participant’s internal perspective manifested in the acceptance of the law as providing guides to conduct and standards of criticism necessarily also included a belief that there are moral reasons for conforming to the law’s requirements and moral justification of its use of coercion, this would also be something for a morally neutral descriptive jurisprudence to record but not to endorse or share.

However, in response to my claim that the partly evaluative issues which Dworkin calls ‘interpretive’ are not the only proper issues for jurisprudence and legal theory, and that there is an important place for general and descriptive jurisprudence, he has conceded that this is so, and he has explained that his observations such as ‘jurisprudence is the general part of adjudication’ need qualification, since this, as he now says, is only ‘true of jurisprudence about the question of sense’.²⁰ This is an important and welcome correction of what appeared to be the extravagant and indeed, as Dworkin himself has termed it, ‘imperialist’, claim that the only proper form of legal theory is interpretive and evaluative.

But I find still very perplexing the implications of the following cautionary words which Dworkin has now coupled with his withdrawal of his seemingly imperialist claim: ‘But it is worth stressing how pervasive that question [of sense] is in the issues that general theories, like Hart’s, have mainly discussed.’²¹ The relevance of this caution is not clear. The issues which I have discussed (see the list on p. 240 above) include questions such as the relation of law to coercive threats on the one hand and to moral requirements on the other, and the point of Dworkin’s caution seems to be that in discussing such issues even the descriptive legal theorist will have to face questions concerning the sense or meaning of propositions of law which can only be satisfactorily answered by an interpretive and partly evaluative legal theory. If this were really the case, in order to determine the sense of any given proposition of law even the descriptive legal theorist must ask and answer the interpretive and evaluative question, ‘What meaning must be assigned to this proposition if it is to follow from principles which best fit the settled law and best justify it?’ But even if it were true that the general and descriptive legal theorist seeking an answer to the kind of questions that I have mentioned must determine the meaning of propositions of law in many different legal systems, there seems no reason to accept the view that this must be determined by his asking Dworkin’s interpretive and evaluative question. Moreover, even if the judges and lawyers of all the legal systems of which the general and descriptive legal theorist had to take account themselves did in fact settle questions of meaning in this interpretive and partly evaluative way, this would be something for the general descriptive theorist to record as a fact on which to base his general descriptive conclusions as to the meaning of such propositions of law. It would of course be a serious error to suppose that because these conclusions were so based they must themselves be interpretive and evaluative and that in offering them the theorist had shifted from the task of description to that of interpretation and evaluation. Description may still be description, even when what is described is an evaluation.

2. THE NATURE OF LEGAL POSITIVISM

(i) Positivism as a Semantic Theory

My book is taken by Dworkin as a representative work of modern legal positivism distinguished from earlier versions, such as those of Bentham and Austin, mainly by its rejection of their imperative theories of law and their conception that all law emanates from a legally unlimited sovereign legislative person or body. Dworkin finds in my version of legal positivism a large number of different though related errors. The most fundamental of these errors is the view that the truth of propositions of law such as those that describe legal rights and legal duties depends only on questions of plain historical fact including facts about individual beliefs and social attitudes.²² The facts on which the truth of propositions of law depends constitute what Dworkin calls ‘the grounds of law’,²³ and the positivist according to him wrongly takes these to be fixed by linguistic rules, shared by judges and lawyers, which govern the use and so the meaning of the word ‘law’ both when this appears in statements of what ‘the law’ of a particular system is on a particular point and in statements about what ‘law’ (i.e. law in general) is.²⁴ From this positivist view of law it would follow that the only disagreements that there can be about questions of law are those which concern the existence or non-existence of such historical facts; there can be no theoretical disagreements or controversy as to what constitutes the ‘grounds’ of law.

Dworkin devotes many illuminating pages of his criticism of legal positivism to showing that theoretical disagreement as to what constitutes the grounds of law is, contrary to the positivist’s view, a prominent feature of Anglo-American legal practices. Against the view that these are uncontroversially fixed by linguistic rules shared by lawyers and judges, Dworkin urges that they are essentially controversial, since amongst them are not only historical facts but very frequently controversial moral judgments and value judgments.

Dworkin offers two very different accounts of how it is that positivists such as myself have come to adopt this their radically mistaken view. According to the first of these accounts, positivists believe that if what the grounds of law are was not uncontroversially fixed by rules, but was a controversial matter allowing theoretical disagreements, then the word ‘law’ would mean different things to different people and in using it they would be simply talking past each other, not communicating about the same thing. This belief thus imputed to the positivist is in Dworkin’s view wholly mistaken, and he calls the argument against controversial grounds of law which the positivist is supposed to base on it the ‘semantic sting’²⁵ because it rests on a theory about the meaning of the word ‘law’. So in Law’s Empire he set out to draw this ‘semantic sting’.

Though in the first chapter of Law’s Empire I am classed with Austin as a semantic theorist and so as deriving a plain-fact positivist theory of law from the meaning of the word ‘law’, and suffering from the semantic sting, in fact nothing in my book or in anything else I have written supports such an account of my theory. Thus, my doctrine that developed municipal legal systems contain a rule of recognition specifying the criteria for the identification of the laws which courts have to apply may be mistaken, but I nowhere base this doctrine on the mistaken idea that it is part of the meaning of the word ‘law’ that there should be such a rule of recognition in all legal systems, or on the even more mistaken idea that if the criteria for the identification of the grounds of law were not uncontroversially fixed, ‘law’ would mean different things to different people.

Indeed this last argument ascribed to me confuses the meaning of a concept with the criteria for its application, and so far from accepting this I expressly drew attention (on page 160 of this book), in explaining the concept of justice, to the fact that the criteria for the application of a concept with a constant meaning may both vary and be controversial. To make this clear I drew in effect the same distinction between a concept and different conceptions of a concept which figures so prominently in Dworkin’s later work.²⁶

Lastly, Dworkin also insists that the positivist’s claim that his theory of law is not a semantic theory, but a descriptive account of the distinctive features of law in general as a complex social phenomenon, presents a contrast with semantic theory which is empty and misleading. His argument²⁷ is that since one of the distinctive features of law as a social phenomenon is that lawyers debate the truth of propositions of law and ‘explain’ this by reference to the meaning of such propositions, such a descriptive theory of law must after all be semantic.²⁸ This argument seems to me to confuse the meaning of ‘law’ with the meaning of propositions of law. A semantic theory of law is said by Dworkin to be a theory that the very meaning of the word ‘law’ makes law depend on certain specific criteria. But propositions of law are typically statements not of what ‘law’ is but of what the law is, i.e. what the law of some system permits or requires or empowers people to do. So even if the meaning of such propositions of law was determined by definitions or by their truth-conditions this does not lead to the conclusion that the very meaning of the word ‘law’ makes law depend on certain specific criteria. This would only be the case if the criteria provided by a system’s rule of recognition and the need for such a rule were derived from the meaning of the word ‘law’. But there is no trace of such a doctrine in my work.²⁹

There is one further respect in which Dworkin misrepresents my form of legal positivism. He treats my doctrine of the rule of recognition as requiring that the criteria which it provides for the identification of law must consist only of historical facts and so as an example of ‘plain-fact positivism’.³⁰ But though my main examples of the criteria provided by the rule of recognition are matters of what Dworkin has called ‘pedigree’,³¹ concerned only with the manner in which laws are adopted or created by legal institutions and not with their content, I expressly state both in this book (p. 72) and in my earlier article on ‘Positivism and the Separation of Law and Morals’³² that in some systems of law, as in the United States, the ultimate criteria of legal validity might explicitly incorporate besides pedigree, principles of justice or substantive moral values, and these may form the content of legal constitutional restraints. In ascribing ‘plain-fact’ positivism to me in Law’s Empire Dworkin ignores this aspect of my theory. So the ‘semantic’ version of plain-fact positivism which he attributes to me is plainly not mine, nor is mine any form of plain-fact positivism.

(ii) Positivism as an Interpretive Theory

Dworkin’s second account of plain-fact positivism does not treat it as a semantic theory or as based on linguistic considerations but attempts to reconstruct it as a form of Dworkinian interpretive theory called by him ‘conventionalism’. According to this theory (which Dworkin ultimately rejects as defective) the positivist, in the guise of an interpretive theorist committed to showing the law in the best light, presents the criteria of law as consisting of plain facts, uncontroversially fixed not, as in the semantic version, by the vocabulary of law but by a conviction which is shared by judges and lawyers. This casts a favourable light on law because it shows it as securing something of great value to the subjects of the law: namely that the occasions for legal coercion are made to depend on plain facts available to all, so that all will have fair warning before coercion is used. This Dworkin calls ‘the ideal of protected expectations’,³³ but its merits for him do not in the end outweigh its various defects.

But this interpretivist account of positivism as conventionalism cannot be represented as a plausible version or reconstruction of my theory of law. This is so for two reasons. First, as I have already stated, my theory is not a plain-fact theory of positivism since amongst the criteria of law it admits values, not only ‘plain’ facts. But secondly and more importantly, whereas Dworkin’s interpretive legal theory in all its forms rests on the presupposition that the point or purpose of law and legal practice is to justify coercion,³⁴ it certainly is not and never has been my view that law has this as its point or purpose. Like other forms of positivism my theory makes no claim to identify the point or purpose of law and legal practices as such; so there is nothing in my theory to support Dworkin’s view, which I certainly do not share, that the purpose of law is to justify the use of coercion. In fact I think it quite vain to seek any more specific purpose which law as such serves beyond providing guides to human conduct and standards of criticism of such conduct. This will not of course serve to distinguish laws from other rules or principles with the same general aims; the distinctive features of law are the provision it makes by secondary rules for the identification, change, and enforcement of its standards and the general claim it makes to priority over other standards. However, even if my theory were wholly committed to plain-fact positivism in the form of conventionalism which protects expectations by guaranteeing that prior notice of the occasions for legal coercion will be generally available, this would only show that I view this as a particular moral merit which law has, not that the whole purpose of law as such is to provide this. Since the occasions for legal coercion are mainly cases where the primary function of the law in guiding the conduct of its subjects has broken down, legal coercion, though of course an important matter, is a secondary function. Its justification cannot be sensibly taken to be the point or purpose of the law as such.

Dworkin’s reasons for reconstructing my legal theory as a conventionalist interpretive theory which makes the claim that legal coercion is only justified ‘when it conforms to conventional understandings’³⁵ rest on my account of the Elements of Law in Chapter V Section 3 of this book. There I exhibit the secondary rules of recognition, change, and adjudication, as remedies for the defects of an imagined simple regime consisting only of primary rules of obligation. These defects are the uncertainty as to the identity of the rules, their static quality, and the time-wasting inefficiency of the diffuse social pressure by which alone the rules are enforced. But in presenting these secondary rules as remedies for such defects I nowhere make any claim that legal coercion is only justified when it conforms to these rules, still less that the provision of such justification is the point or purpose of the law in general. Indeed the only reference which I make to coercion in my discussion of secondary rules is to the time-wasting inefficiency of leaving the enforcement of the rules to diffuse social pressure instead of (p. 250) to organized sanctions administered by courts. But plainly a remedy for inefficiency is not a justification.

It is of course true that the addition to the regime of primary rules of obligation of a secondary rule of recognition will, by frequently enabling individuals to identify in advance the occasions for coercion, help to justify its use in the sense that it will exclude one moral objection to its use. But the certainty and knowledge in advance of the requirements of the law which the rule of recognition will bring is not only of importance where coercion is in issue: it is equally crucial for the intelligent exercise of legal powers (e.g. to make wills or contracts) and generally for the intelligent planning of private and public life. The justification of coercion to which the rule of recognition contributes therefore cannot be represented as its general point or purpose, still less can it be represented as the general point or purpose of the law as a whole. Nothing in my theory suggests that it can.

(iii) Soft Positivism

Dworkin in attributing to me a doctrine of ‘plain-fact positivism’ has mistakenly treated my theory as not only requiring (as it does) that the existence and authority of the rule of recognition should depend on the fact of its acceptance by the courts, but also as requiring (as it does not) that the criteria of legal validity which the rule provides should consist exclusively of the specific kind of plain fact which he calls ‘pedigree’ matters and which concern the manner and form of law-creation or adoption. This is doubly mistaken. First, it ignores my explicit acknowledgement that the rule of recognition may incorporate as criteria of legal validity conformity with moral principles or substantive values; so my doctrine is what has been called ‘soft positivism’ and not as in Dworkin’s version of it ‘plain-fact’ positivism. Secondly, there is nothing in my book to suggest that the plain-fact criteria provided by the rule of recognition must be solely matters of pedigree; they may instead be substantive constraints on the content of legislation such as the Sixteenth or Nineteenth Amendments to the United States Constitution respecting the establishment of religion or abridgements of the right to vote.

But this reply does not meet Dworkin’s most basic criticisms, (p. 251) for in replying to other theorists who have also adopted some form of soft positivism,³⁶ he has made important criticisms of it which if valid would apply to my theory and so call for an answer here.

Dworkin’s most fundamental criticism is that there is a deep inconsistency between soft positivism, which permits the identification of the law to depend on controversial matters of conformity with moral or other value judgments, and the general positivist ‘picture’ of law as essentially concerned to provide reliable public standards of conduct which can be identified with certainty as matters of plain fact without dependence on controversial moral arguments.³⁷ To establish such inconsistency between soft positivism and the rest of my theory Dworkin would cite my account of the rule of recognition as curing, among other defects, the uncertainty of the imagined pre-legal regime of custom-type primary rules of obligation. This criticism of soft positivism seems to me to exaggerate both the degree of certainty which a consistent positivist must attribute to a body of legal standards and the uncertainty which will result if the criteria of legal validity include conformity with specific moral principles or values. It is of course true that an important function of the rule of recognition is to promote the certainty with which the law may be ascertained. This it would fail to do if the tests which it introduced for law not only raise controversial issues in some cases but raise them in all or most cases. But the exclusion of all uncertainty at whatever costs in other values is not a goal which I have ever envisaged for the rule of recognition. This is made plain, or so I had hoped, both by my explicit statement in this book that the rule of recognition itself as well as particular rules of law identified by reference to it may have a debatable ‘penumbra’ of uncertainty.³⁸ There is also my general argument that, even if laws could be framed that could settle in advance all possible questions that could arise about their meaning, to adopt such laws would often war with other aims which law should cherish.³⁹ A margin of uncertainty should (p. 252) be tolerated, and indeed welcomed in the case of many legal rules, so that an informed judicial decision can be made when the composition of an unforeseen case is known and the issues at stake in its decision can be identified and so rationally settled. Only if the certaintyproviding function of the rule of recognition is treated as paramount and overriding could the form of soft positivism that includes among the criteria of law conformity with moral principles or values which may be controversial be regarded as inconsistent. The underlying question here concerns the degree or extent of uncertainty which a legal system can tolerate if it is to make any significant advance from a decentralized regime of custom-type rules in providing generally reliable and determinate guides to conduct identifiable in advance.

Dworkin’s second criticism of the consistency of my version of soft positivism raises different and more complex issues concerning the determinacy and completeness of law. My view advanced in this book is that legal rules and principles identified in general terms by the criteria provided by the rule of recognition often have what I call frequently ‘open texture’, so that when the question is whether a given rule applies to a particular case the law fails to determine an answer either way and so proves partially indeterminate. Such cases are not merely ‘hard cases’, controversial in the sense that reasonable and informed lawyers may disagree about which answer is legally correct, but the law in such cases is fundamentally incomplete: it provides no answer to the questions at issue in such cases. They are legally unregulated and in order to reach a decision in such cases the courts must exercise the restricted law-making function which I call ‘discretion’. Dworkin rejects the idea that the law may be incomplete in this way and leave gaps to be filled by the exercise of such a law-creating discretion. This view he thinks is a mistaken inference from the fact that a proposition of law asserting the existence of a legal right or a legal duty may be controversial and so a matter about which reasonable and informed men may disagree, and when they do disagree there is often no way of demonstrating conclusively whether it is true or false. Such an inference is mistaken because when a proposition of law is thus controversial there may none the less still be ‘facts of the matter’ in virtue of which it is true or false, and though its truth or falsity cannot be demonstrated, arguments that it is true may still be assessed as better than arguments that it is false and vice versa. This distinction between law that is controversial and law that is incomplete or indeterminate is a matter of considerable importance for Dworkin’s interpretive theory, since according to that theory a proposition of law is true only if in conjunction with other premisses it follows from principles which both best fit the legal system’s institutional history and also provide the best moral justification for it. Hence for Dworkin the truth of any proposition of law ultimately depends on the truth of a moral judgment as to what best justifies and since for him moral judgments are essentially controversial, so are all propositions of law.

For Dworkin the idea of a criterion of legal validity the application of which involves a controversial moral judgment presents no theoretical difficulty; it can still be in his view a genuine test for pre-existing law because its controversial character is perfectly compatible with there being facts (in many cases moral facts) in virtue of which it is true.

But soft positivism, which allows that a criterion of legal validity may be in part a moral test is, so Dworkin claims, involved in a second inconsistency, in addition to that already discussed on pp. 251–2 above. For it is not only inconsistent with the positivist ‘picture’ of law as identifiable with certainty, but inconsistent also with the wish which he attributes to positivists to make ‘the objective standing of propositions of law’⁴⁰ independent of any commitment to any controversial philosophical theory of the status of moral judgments. For a moral test can be a test for pre-existing law only if there are objective moral facts in virtue of which moral judgments are true. But that there are such objective moral facts is a controversial philosophical theory; if there are no such facts, a judge, told to apply a moral test, can only treat this as a call for the exercise by him of a law-making discretion in accordance with his best understanding of morality and its requirements and subject to whatever constraints on this are imposed by the legal system.

I still think legal theory should avoid commitment to controversial philosophical theories of the general status of moral judgments and should leave open, as I do in this book (p. 168), the general question of whether they have what Dworkin calls ‘objective standing’. For whatever the answer is to this philosophical question, the judge’s duty will be the same: namely, to make the best moral judgment he can on any moral issues he may have to decide. It will not matter for any practical purpose whether in so deciding cases the judge is making law in accordance with morality (subject to whatever constraints are imposed by law) or alternatively is guided by his moral judgment as to what already existing law is revealed by a moral test for law. Of course, if the question of the objective standing of moral judgments is left open by legal theory, as I claim it should be, then soft positivism cannot be simply characterized as the theory that moral principles or values may be among the criteria of legal validity, since if it is an open question whether moral principles and values have objective standing, it must also be an open question whether ‘soft positivist’ provisions purporting to include conformity with them among the tests for existing law can have that effect or instead, can only constitute directions to courts to make law in accordance with morality.

It is to be observed that some theorists, notably Raz, hold that whatever the status of moral judgments may be, whenever the law requires courts to apply moral standards to determining the law it thereby grants the courts discretion and directs them to use it according to their best moral judgment in making what is a new law; it does not thereby convert morality into pre-existing law.⁴¹

3. THE NATURE OF RULES

(i) The Practice Theory of Rules

At various points in this book I draw attention to the distinction between internal and external statements of law and between internal and external aspects of law.

To explain these distinctions and their importance I started (pp. 56–7) by examining not the highly complex case of a legal system which comprises both enacted and custom-type rules, but the simpler case (to which the same distinctions between internal and external apply) of the custom-type rules of any social group large or small, and these I call ‘social rules’. The account I have given of these has become known as ‘the practice theory’ of rules because it treats the social rules of a group as constituted by a form of social practice comprising both patterns of conduct regularly followed by most members of the group and a distinctive normative attitude to such patterns of conduct which I have called ‘acceptance’. This consists in the standing disposition of individuals to take such patterns of conduct both as guides to their own future conduct and as standards of criticism which may legitimate demands and various forms of pressure for conformity. The external point of view of social rules is that of an observer of their practice, and the internal point of view is that of a participant in such practice who accepts the rules as guides to conduct and as standards of criticism.

My practice theory of social rules has been extensively criticized by Dworkin, who, as I have already mentioned, makes a similar but in fact in many ways a very different distinction between a sociologist’s external description of a community’s social rules and the internal point of view of a participant who appeals to the rules for the purpose of evaluation and criticism of his own and others’ conduct.⁴² Some of Dworkin’s criticism of my original account of social rules is certainly sound and important for the understanding of law, and in what follows here I indicate the considerable modifications in my original account which I now think necessary.

(i) My account is, as Dworkin has claimed, defective in ignoring the important difference between a consensus of convention manifested in a group’s conventional rules and a consensus of independent conviction manifested in the concurrent practices of a group. Rules are conventional social practices if the general conformity of a group to them is part of the reasons which its individual members have for acceptance; by contrast merely concurrent practices such as the shared morality of a group are constituted not by convention but by the fact that members of the group have and generally act on the same but independent reasons for behaving in certain specific ways.

(ii) My account of social rules is, as Dworkin has also rightly claimed, applicable only to rules which are conventional in the sense I have now explained. This considerably narrows the scope of my practice theory and I do not now regard it as a sound explanation of morality, either individual or social. But the theory remains as a faithful account of conventional social rules which include, besides ordinary social customs (which may or may not be recognized as having legal force), certain important legal rules including the rule of recognition, which is in effect a form of judicial customary rule existing only if it is accepted and practised in the law-identifying and law-applying operations of the courts. Enacted legal rules by contrast, though they are identifiable as valid legal rules by the criteria provided by the rule of recognition, may exist as legal rules from the moment of their enactment before any occasion for their practice has arisen and the practice theory is not applicable to them.

Dworkin’s central criticism of the practice theory of rules is that it mistakenly takes a social rule to be constituted by its social practice and so treats the statement that such a rule exists merely as a statement of the external sociological fact that the practice-conditions for the existence of the rule are satisfied.⁴³ That account cannot, so Dworkin argues, explain the normative character possessed by even the simplest conventional rule. For these rules establish duties and reasons for action to which appeal is made when such rules are cited, as they commonly are, in criticism of conduct and in support of demands for action. This reason-giving and duty-establishing feature of rules constitutes their distinctive normative character and shows that their existence cannot consist in a merely factual state of affairs as do the practices and attitudes which according to the practice theory constitute the existence of a social rule. According to Dworkin, a normative rule with these distinctive features can only exist if there is ‘a certain normative state of affairs’.⁴⁴ I find these quoted words tantalizingly obscure: from the discussion of the example of the Churchgoers’ Rule (males must bare their heads in church)⁴⁵ Dworkin, it appears, means by a normative state of affairs the existence of good moral grounds or justification for doing what the rule requires, so he argues that while the mere regular practice of churchgoers removing hats in church cannot constitute the rule it may help to justify it by creating ways of giving offence and by giving rise to expectations which are good grounds for a rule requiring the removal of hats in church. If this is what Dworkin means by a normative state of affairs required to warrant the assertion of a normative rule his account of the existence conditions of a social rule seems to me far too strong. For it seems to require not only that the participants who appeal to rules as establishing duties or providing reasons for action must believe that there are good moral grounds or justification for conforming to the rules, but that there must actually be such good grounds. Plainly a society may have rules accepted by its members which are morally iniquitous, such as rules prohibiting persons of certain colour from using public facilities such as parks or bathing beaches. Indeed, even the weaker condition that for the existence of a social rule it must only be the case that participants must believe that there are good moral grounds for conforming to it is far too strong as a general condition for the existence of social rules. For some rules may be accepted simply out of deference to tradition or the wish to identify with others or in the belief that society knows best what is to the advantage of individuals. These attitudes may coexist with a more or less vivid realization that the rules are morally objectionable. Of course a conventional rule may both be and be believed to be morally sound and justified. But when the question arises as to why those who have accepted conventional rules as a guide to their behaviour or as standards of criticism have done so I see no reason for selecting from the many answers to be given (see pp. 203, 232 of this book) a belief in the moral justification of rules as the sole possible or adequate answer.

Finally, Dworkin argues that the practice theory of rules even if restricted to conventional rules must be abandoned because it cannot accommodate the idea that the scope of a conventional rule may be controversial and so the subject of disagreement.⁴⁶ He does not deny that there are some uncontroversial rules constituted by regular practice and acceptance, but he claims that rules so constituted include only relatively unimportant cases such as the rules of some games; but in this book a rule as important and as little controversial as a legal system’s basic rule of recognition is treated as a rule constituted by the uniform practice of the courts in accepting it as a guide to their law-applying and law-enforcing operations. Against this Dworkin contends that in hard cases there are frequent theoretical disagreements between judges as to what the law on some subject is and that these show that the appearance of uncontroversiality and general acceptance is an illusion. Of course the frequency and importance of such disagreements cannot be denied but appeals to their existence used as an argument against the applicability of the practice theory to the rule of recognition rest on a misunderstanding of the function of the rule. It assumes that the rule is meant to determine completely the legal result in particular cases, so that any legal issue arising in any case could simply be solved by mere appeal to the criteria or tests provided by the rule. But this is a misconception: the function of the rule is to determine only the general conditions which correct legal decisions must satisfy in modern systems of law. The rule does this most often by supplying criteria of validity which Dworkin calls matters of pedigree and which refer not to the content of the law but to the manner and form in which the laws are created or adopted; but as I have said (p. 250) in addition to such pedigree matters the rule of recognition may supply tests relating not to the factual content of laws but to their conformity with substantive moral values or principles. Of course in particular cases judges may disagree as to whether such tests are satisfied or not and a moral test in the rule of recognition will not resolve such disagreement. Judges may be agreed on the relevance of such tests as something settled by established judicial practice even though they disagree as to what the tests require in particular cases. To the rule of recognition viewed in this way the practice theory of rules is fully applicable.

(ii) Rules and Principles

For long the best known of Dworkin’s criticisms of this book was that it mistakenly represents law as consisting solely of ‘all-or-nothing’ rules, and ignores a different kind of legal standard, namely legal principles, which play an important and distinctive part in legal reasoning and adjudication. Some critics who have found this defect in my work have conceived of it as a more or less isolated fault which I could repair simply by including legal principles along with legal rules as components of a legal system, and they have thought that I could do this without abandoning or seriously modifying any of the main themes of the book. But Dworkin, who was the first to press this line of criticism, has insisted that legal principles could only be included in my theory of law at the cost of surrender of its central doctrines. If I were to admit that law consists in part of principles I could not, according to him, consistently maintain, as I have done, that the law of a system is identified by criteria provided by a rule of recognition accepted in the practice of the courts, or that the courts exercise a genuine though interstitial law-making power or discretion in those cases where the existing explicit law fails to dictate a decision, or that there is no important necessary or conceptual connection between law and morality . These doctrines are not only central to my theory of law but are often taken to constitute the core of modern legal positivism; so their abandonment would be a matter of some moment.

In this section of my reply I consider various aspects of the criticism that I have ignored legal principles and I attempt to show that whatever is valid in this criticism can be accommodated without any serious consequences for my theory as a whole. But I certainly wish to confess now that I said far too little in my book about the topic of adjudication and legal reasoning and, in particular, about arguments from what my critics call legal principles. I now agree that it is a defect of this book that principles are touched upon only in passing.

But what precisely is it that I am charged with ignoring? What are legal principles, and how do they differ from legal rules? As used by legal writers ‘principles’ often include a vast array of theoretical and practical considerations only some of which are relevant to the issues which Dworkin meant to raise. Even if the expression ‘principle’ is taken to be limited to standards of conduct including the conduct of courts in deciding cases, there are different ways of drawing a contrast between rules and such principles. However, I think all my critics who have accused me of ignoring principles would agree that there are at least two features which distinguish them from rules. The first is a matter of degree: principles are, relatively to rules, broad, general, or unspecific, in the sense that often what would be regarded as a number of distinct rules can be exhibited as the exemplifications or instantiations of a single principle. The second feature is that principles, because they refer more or less explicitly to some purpose, goal, entitlement, or value, are regarded from some point of view as desirable to maintain, or to adhere to, and so not only as providing an explanation or rationale of the rules which exemplify them, but as at least contributing to their justification.

Besides these two relatively uncontroversial features of breadth and desirability from some point of view which account for the explanatory and justificatory role of principles in relation to rules, there is a third distinguishing feature which I myself think is a matter of degree whereas Dworkin who regards it as crucial does not. Rules, according to him, function in the reasoning of those who apply them in an ‘all-or-nothing manner’ in the sense that if a rule is valid and applicable at all to a given case then it ‘necessitates’ i.e. conclusively determines the legal result or outcome.⁴⁷ Among the examples which he gave of legal rules are those prescribing a maximum speed of 60 m.p.h. on the turnpike road or statutes regulating the making, proof, and efficacy of wills such as the statutory rule that a will is invalid unless signed by two witnesses. Legal principles, according to Dworkin, differ from such all-or-nothing rules because when they are applicable they do not ‘necessitate’ a decision but point towards or count in favour of a decision, or state a reason which may be overridden but which the courts take into account as inclining in one direction or another. I shall, for short, call this feature of principles their ‘non-conclusive’ character. Some examples given by Dworkin of such non-conclusive principles are relatively specific, such as ‘the courts must examine purchase agreements [for automobiles] closely to see if consumer and public interests are treated fairly’;⁴⁸ others have much wider scope, such as ‘no man may profit from his own wrong’;⁴⁹ and in fact many of the most important constitutional restrictions on the powers of the United States Congress and on state legislation such as the provisions of the First, Fifth, and Fourteenth Amendments to the United States Constitution function as non-conclusive principles.⁵⁰ Legal principles, according to Dworkin, differ from rules because they have a dimension of weight⁵¹ but not of validity, and hence it is that in conflict with another principle of greater weight, one principle may be overridden and fail to determine a decision, but none the less will survive intact to be used in other cases where it may win in competition with some other principle of lesser weight. Rules, on the other hand, are either valid or invalid but do not have this dimension of weight, so if as initially formulated they conflict, only one of them according to Dworkin can be valid, and a rule which loses in competition with another must be reformulated so as to make it consistent with its competitor and hence inapplicable to the given case.⁵²

I see no reason to accept either this sharp contrast between legal principles and legal rules, or the view that if a valid rule is applicable to a given case it must, unlike a principle, always determine the outcome of the case. There is no reason why a legal system should not recognize that a valid rule determines a result in cases to which it is applicable, except where another rule, judged to be more important, is also applicable to the same case. So a rule which is defeated in competition with a more important rule in a given case may, like a principle, survive to determine the outcome in other cases where it is judged to be more important than another competing rule.⁵³

So law for Dworkin comprises both all-or-nothing rules and non-conclusive principles, and he does not think that this difference between them is a matter of degree. But I do not think that Dworkin’s position can be coherent. His earliest examples imply that rules may come into conflict with principles and that a principle will sometimes win in competition with a rule and sometimes lose. The cases he cites include Riggs v. Palmer,⁵⁴ in which the principle that a man may not be permitted to profit from his own wrongdoing was held notwithstanding the clear language of the statutory rules governing the effect of a will to preclude a murderer inheriting under his victim’s will. This is an example of a principle winning in competition with a rule, but the existence of such competition surely shows that rules do not have an all-or-nothing character, since they are liable to be brought into such conflict with principles which may outweigh them. Even if we describe such cases (as Dworkin at times suggests) not as conflicts between rules and principles, but as a conflict between the principle explaining and justifying the rule under consideration and some other principle, the sharp contrast between all-or-nothing rules and non-conclusive principles disappears; for on this view a rule will fail to determine a result in a case to which it is applicable according to its terms if its justifying principle is outweighed by another. The same is true if (as Dworkin also suggests) we think of a principle as providing a reason for a new interpretation of some clearly formulated legal rule.⁵⁵

This incoherence in the claim that a legal system consists both of all-or-nothing rules and non-conclusive principles may be cured if we admit that the distinction is a matter of degree. Certainly a reasonable contrast can be made between near conclusive rules, the satisfaction of whose conditions of application suffices to determine the legal result except in a few instances (where its provisions may conflict with another rule judged of greater importance) and generally non-conclusive principles which merely point towards a decision but may very frequently fail to determine it.

I certainly think that arguments from such non-conclusive principles are an important feature of adjudication and legal reasoning, and that it should be marked by an appropriate terminology. Much credit is due to Dworkin for having shown and illustrated their importance and their role in legal reasoning, and certainly it was a serious mistake on my part not to have stressed their non-conclusive force. But I certainly did not intend in my use of the word ‘rule’ to claim that legal systems comprise only ‘all-or-nothing’ or near-conclusive rules. I not only drew attention (see pp. 130–3 of this book) to what I termed (perhaps infelicitously) ‘variable legal standards’ which specify factors to be taken into account and weighed against others, but I attempted (see pp. 133–4) to explain why some areas of conduct were suitable for regulation not by such variable standards as ‘due care’ but rather by near-conclusive rules prohibiting or requiring the same specific actions in all but rare cases. So it is that we have rules against murder and theft and not merely principles requiring due respect for human life and property.

4. PRINCIPLES AND THE RULE OF RECOGNITION

Pedigree and Interpretation

Dworkin has claimed that legal principles cannot be identified by criteria provided by a rule of recognition manifested in the practice of the courts and that, since principles are essential elements of law, the doctrine of a rule of recognition must be abandoned. According to him, legal principles can only be identified by constructive interpretation as members of the unique set of principles which both best fits and best justifies the whole institutional history of the settled law of a legal system. Of course no court, English or American, has ever explicitly adopted such a system-wide holistic criterion for identifying the law, and Dworkin concedes that no actual human judge as distinct from his mythical ideal judge ‘Hercules’ could accomplish the feat of constructing an interpretation of all his country’s law at once. None the less the courts in his view are most illuminatingly understood as trying to ‘imitate Hercules’ in a limited way and viewing their judgments in this way serves, he thinks, to bring to light ‘the hidden structure’.⁵⁶

The most famous example, familiar to English lawyers, of the identification of principles by a limited form of constructive interpretation is Lord Atkin’s formulation in the case of Donoghue v. Stevenson⁵⁷ of the previously unformulated ‘neighbour principle’ as underlying the various separate rules establishing a duty of care in different situations.

I do not find plausible the view that in such limited exercises of constructive interpretation judges are best understood as trying to imitate Hercules’ holistic system-wide approach. But my present criticism is that preoccupation with constructive interpretation has led Dworkin to ignore the fact that many legal principles owe their status not to their content serving as interpretation of settled law, but to what he calls their ‘pedigree’; that is the manner of their creation or adoption by a recognized authoritative source. This preoccupation has, I think, in fact led him into a double error: first, to the belief that legal principles cannot be identified by their pedigree, and secondly to the belief that a rule of recognition can only provide pedigree criteria. Both these beliefs are mistaken; the first is so because there is nothing in the non-conclusive character of principles nor in their other features to preclude their identification by pedigree criteria. For plainly a provision in a written constitution or a constitutional amendment or a statute may be taken as intended to operate in the non-conclusive way characteristic of principles, as providing reasons for decision which may be outweighed in cases where some other rule or principle presents stronger reasons for an alternative decision. Dworkin himself envisaged that the First Amendment of the United States Constitution, providing that Congress shall not abridge freedom of speech, is to be interpreted in just that way.⁵⁸ Also some legal principles, including some basic principles of the Common Law, such as that no man may profit from his own wrongdoing, are identified as law by the ‘pedigree’ test in that they have been consistently invoked by courts in ranges of different cases as providing reasons for decision, which must be taken into account, though liable to be overridden in some cases by reasons pointing the other way. In face of such examples of legal principles identified by pedigree criteria, no general argument that the inclusion of principles as part of the law entails the abandonment of the doctrine of a rule of recognition could succeed. In fact, as I show below, their inclusion is not only consistent with, but actually requires acceptance of that doctrine. If it is conceded, as surely it must be, that there are at least some legal principles which may be ‘captured’ or identified as law by pedigree criteria provided by a rule of recognition, then Dworkin’s criticism must be reduced to the more modest claim that there are many legal principles that cannot be so captured because they are too numerous, too fleeting, or too liable to change or modification, or have no feature which would permit their identification as principles of law by reference to any other test than that of belonging to that coherent scheme of principles which both best fits the institutional history and practices of the system and best justifies them. At first sight this interpretivist test seems not to be an alternative to a criterion provided by a rule of recognition, but, as some critics have urged,⁵⁹ only a complex ‘soft-positivist’ form of such a criterion identifying principles by their content not by their pedigree. It is true that a rule of recognition containing such an interpretive criterion could not, for the reasons discussed on pp. 251 ff. above, secure the degree of certainty in identifying the law which according to Dworkin a positivist would wish. None the less, to show that the interpretive test criterion was part of a conventional pattern of law-recognition would still be a good theoretical explanation of its legal status. So there is certainly no incompatibility such as Dworkin claims between the admission of principles as part of the law and the doctrine of a rule of recognition.

The argument of the last two paragraphs is enough to show that contrary to Dworkin’s contention the acceptance of principles as part of the law is consistent with the doctrine of a rule of recognition, even if Dworkin’s interpretive test were as he claims the sole appropriate criterion for identifying them. But in fact a stronger conclusion is warranted: namely that a rule of recognition is necessary if legal principles are to be identified by such a criterion. This is so because the starting-point for the identification of any legal principle to be brought to light by Dworkin’s interpretive test is some specific area of the settled law which the principle fits and helps to justify. The use of that criterion therefore presupposes the identification of the settled law, and for that to be possible a rule of recognition specifying the sources of law and the relationships of superiority and subordination holding between them is necessary. In the terminology of Law’s Empire, the legal rules and practices which constitute the starting-points for the interpretive task of identifying underlying or implicit legal principles constitute ‘preinterpretive law’, and much that Dworkin says about it appears to endorse the view that for its identification something very like a rule of recognition identifying the authoritative sources of law as described in this book is necessary. The main difference between my view and Dworkin’s here is that whereas I ascribe the general agreement found among judges as to the criteria for the identification of the sources of law to their shared acceptance of rules providing such criteria, Dworkin prefers to speak not of rules but of ‘consensus’⁶⁰ and ‘paradigms’⁶¹ and ‘assumptions’⁶² which members of the same interpretive community share. Of course, as Dworkin has made clear, there is an important distinction between a consensus of independent convictions where the concurrence of others is not part of the reason which each party to the consensus has for concurring, and a consensus of convention where it is such a part. Certainly the rule of recognition is treated in my book as resting on a conventional form of judicial consensus. That it does so rest seems quite clear at least in English and American law for surely an English judge’s reason for treating Parliament’s legislation (or an American judge’s reason for treating the Constitution) as a source of law having supremacy over other sources includes the fact that his judicial colleagues concur in this as their predecessors have done. Indeed Dworkin himself speaks of the doctrine of legislative supremacy as a brute fact of legal history which limits the role which a judge’s conviction can play⁶³ and he states that ‘the interpretive attitude cannot survive unless members of the same interpretive community share at least roughly the same assumptions’ about ‘what counts as part of the practice’.⁶⁴ I conclude therefore that whatever differences may remain between rules and the ‘assumptions’ and ‘consensus’ and ‘paradigms’ of which Dworkin speaks, his explanation of the judicial identification of the sources of law is substantially the same as mine.

However, large theoretical differences remain between mine and Dworkin’s view. For Dworkin would certainly reject my treatment of his interpretive test for legal principles as merely the specific form taken in some legal systems by a conventional rule of recognition whose existence and authority depend on its acceptance by the courts. This would in his view utterly misrepresent and demean the project of a ‘constructive’ interpretation designed to show the law in the best moral light, which in Dworkin’s view is involved in the identification of the law. For this style of interpretation is not conceived by him as a method of law recognition required by a mere conventional rule accepted by the judges and lawyers of particular legal systems. Instead he presents it as a central feature of much social thought and social practice besides law and as showing ‘a deep connection among all forms of interpretation’, including interpretation as it is understood in literary criticism and even in the natural sciences.⁶⁵ However, even if this interpretive criterion is not merely a pattern of law recognition required by a conventional rule and has affinities and connections with interpretation as it is understood in other disciplines, the fact remains that if there are any legal systems in which Dworkin’s holistic interpretive criterion is actually used to identify legal principles it could perfectly well be that in such systems that criterion is provided by a conventional rule of recognition. But since there are no actual legal systems where this full holistic criterion is used, but only systems like English law and American law where more modest exercises of constructive interpretation are undertaken in cases like Donoghue v. Stevenson to identify latent legal principles, the only question to be considered is whether such exercises are to be understood as the application of a criterion provided by a conventional rule of recognition or in some other way, and if so what their legal status is.

5. LAW AND MORALITY

(i) Rights and Duties

I argue in this book that though there are many different contingent connections between law and morality there are no necessary conceptual connections between the content of law and morality; and hence morally iniquitous provisions may be valid as legal rules or principles. One aspect of this form of the separation of law from morality is that there can be legal rights and duties which have no moral justification or force whatever. Dworkin has rejected this idea in favour of the view (ultimately derived from his own interpretive theory of law) that there must be at least prima-facie moral grounds for assertions of the existence of legal rights and duties. So he regards the idea that ‘legal rights must be understood as [a] species of moral rights’ as a ‘crucial’⁶⁶ element in his legal theory and says that the opposed positivist doctrine belongs to ‘the peculiar world of legal essentialism’⁶⁷ in which it is just given to us to know pre-analytically that there can be legal rights and duties without any moral ground or force. It is I think important for understanding the kind of contribution which a general descriptive jurisprudence can make to the understanding of law to see that whatever the merits of his general interpretive theory may be, Dworkin’s criticism of the doctrine that legal rights and duties may be devoid of moral force or justification is mistaken. It is so for the following reasons: legal rights and duties are the point at which the law with its coercive resources respectively protects individual freedom and restricts it or confers on individuals or denies to them the power to avail themselves of the law’s coercive machinery. So whether the laws are morally good or bad, just or unjust, rights and duties demand attention as focal points in the operations of the law which are of supreme importance to human beings and independently of moral merits of the laws. It is therefore untrue that statements of legal rights and duties can only make sense in the real world if there is some moral ground for asserting their existence.

(ii)The Identification of the Law

The most fundamental difference relating to connections between law and morality between the legal theory developed in this book and Dworkin’s theory concerns the identification of the law. According to my theory, the existence and content of the law can be identified by reference to the social sources of the law (e.g. legislation, judicial decisions, social customs) without reference to morality except where the law thus identified has itself incorporated moral criteria for the identification of the law. In Dworkin’s interpretive theory, on the other hand, every proposition of law stating what the law on some subject is necessarily involves a moral judgment, since according to his holistic interpretive theory propositions of law are true only if with other premisses they follow from that set of principles which both best fit all the settled law identified by reference to the social sources of the law and provide the best moral justification for it. This overall holistic interpretive theory has therefore a double function: it serves both to identify the law and to provide moral justification for it.

Such was Dworkin’s theory, briefly summarized, prior to his introduction in Law’s Empire of the distinction between ‘interpretive’ and ‘preinterpretive’ law. Considered as an alternative to the positivist’s theory that the existence and content of the law may be identified without reference to morality, Dworkin’s theory as it originally stood was vulnerable to the following criticism. Where the law identified by reference to its social sources is morally iniquitous, principles providing the best ‘justification’ for it could only be the least iniquitous of principles fitting that law. But such least iniquitous principles can have no justifying force and cannot constitute any moral limit or constraint on what can count as law and since they cannot fail to fit any legal system, however evil, the theory purporting to identify law by reference to them is indistinguishable from the positivist theory that the law may be identified without any reference to morality. Principles which are morally sound by the standards of what Dworkin has called ‘background morality’⁶⁸ and not merely the morally soundest of those principles which fit the law may indeed provide moral limits or constraints upon what can count as law. I do not dissent in any way from that proposition but it is fully compatible with my claim that the law may be identified without reference to morality.

In introducing his later distinction between interpretive and preinterpretive law Dworkin concedes that there may be legal systems so evil that no interpretation of their laws which we could find morally acceptable is possible. When this is so we may, as he explains, resort to what he calls ‘internal scepticism’⁶⁹ and deny that such systems are law. But since our resources for describing such situations are highly flexible we are not bound to come to that conclusion when we can instead say that legal systems however evil are law in a preinterpretive sense.⁷⁰ So we are not forced to say of even the worst of the Nazi laws that they are not law since they may differ only in their iniquitous moral content from the laws of morally acceptable regimes while sharing with them many distinctive features of law (e.g. forms of law creation, forms of adjudication and enforcement). There may be reasons enough in many contexts and for many purposes to disregard the moral difference and say with the positivist that such evil systems are law. To this Dworkin would only add as it were a rider manifesting his general adherence to his interpretive point of view that such evil systems are law only in a preinterpretive sense.

I find that this appeal to the flexibility of our language and the introduction at this point of the distinction between interpretive and preinterpretive law concedes rather than weakens the positivist’s case. For it does little more than convey the message that while he insists that in a descriptive jurisprudence the law may be identified without reference to morality, things are otherwise for a justificatory interpretive jurisprudence according to which the identification of the law always involves a moral judgment as to what best justifies the settled law. This message of course gives no reason for the positivist to abandon his descriptive enterprise, nor is it intended to do so but even this message has to be qualified, for the law may be so evil that ‘internal scepticism’ is in order, in which case the interpretation of the law involves no moral judgment and interpretation as Dworkin understands it must be given up.⁷¹

One further modification by Dworkin of his interpretive theory has an important bearing on his account of legal rights. In his holistic theory as originally expounded the identification of law and its justification are both treated as following from that unique set of principles which both best fit all of the settled law of a system and best justify it. Such principles therefore have, as I have said, a double function. But since the settled law of a system may be so evil that no overall justifying interpretation of its law is possible, Dworkin has observed that these two functions may become separated, leaving only principles of law identified without reference to any morality . But such law cannot establish any rights having the prima-facie moral force which Dworkin claims all legal rights have. Yet as Dworkin later recognized, even where the system is so wicked that no moral or justifying interpretation of law as a whole is possible there may still be situations where individuals may properly be said to have rights with at least prima-facie moral force.⁷² That would be so where the system contains laws (e.g. those relating to the formation and enforcement of contracts) which may not be affected by the general wickedness of the system and individuals may have relied on such laws in planning their lives or making dispositions of property. To cater for such situations Dworkin qualifies his original idea that legal rights and duties with (p. 272) prima-facie moral force must flow from a general interpretive theory of the law, and he recognizes such situations as constituting independently of his general theory ‘special reasons’ for ascribing legal rights with some moral force to individuals.

6. JUDICIAL DISCRETION⁷³

The sharpest direct conflict between the legal theory of this book and Dworkin’s theory arises from my contention that in any legal system there will always be certain legally unregulated cases in which on some point no decision either way is dictated by the law and the law is accordingly partly indeterminate or incomplete. If in such cases the judge is to reach a decision and is not, as Bentham once advocated, to disclaim jurisdiction or to refer the points not regulated by the existing law to the legislature to decide, he must exercise his discretion and make law for the case instead of merely applying already pre-existing settled law. So in such legally unprovided-for or unregulated cases the judge both makes new law and applies the established law which both confers and constrains his law-making powers.

This picture of the law as in part indeterminate or incomplete and of the judge as filling the gaps by exercising a limited law-creating discretion is rejected by Dworkin as a misleading account both of the law and of judicial reasoning. He claims in effect that what is incomplete is not the law but the positivist’s picture of it, and that this is so will emerge from his own ‘interpretive’ account of the law as including besides the explicit settled law identified by reference to its social sources, implicit legal principles which are those principles which both best fit or cohere with the explicit law and also provide the best moral justification for it. On this interpretive view, the law is never incomplete or indeterminate, so the judge never has occasion to step outside the law and exercise a law-creating power in order to reach a decision. It is therefore to such implicit principles, with their moral dimensions, that courts should turn in those ‘hard cases’ where the social sources of the law fail to determine a decision on some point of law.

It is important that the law-creating powers which I ascribe to the judges to regulate cases left partly unregulated by the law are different from those of a legislature: not only are the judge’s powers subject to many constraints narrowing his choice from which a legislature may be quite free, but since the judge’s powers are exercised only to dispose of particular instant cases he cannot use these to introduce large-scale reforms or new codes. So his powers are interstitial as well as subject to many substantive constraints. None the less there will be points where the existing law fails to dictate any decision as the correct one, and to decide cases where this is so the judge must exercise his law-making powers. But he must not do this arbitrarily: that is he must always have some general reasons justifying his decision and he must act as a conscientious legislator would by deciding according to his own beliefs and values. But if he satisfies these conditions he is entitled to follow standards or reasons for decision which are not dictated by the law and may differ from those followed by other judges faced with similar hard cases.

Against my account of the courts as exercising such a limited discretionary power to settle cases left incompletely regulated by the law, Dworkin directs three main criticisms. The first is that this account is a false description of the judicial process and of what courts do in ‘hard cases’.⁷⁴ To show this Dworkin appeals to the language used by judges and lawyers in describing the judge’s task, and to the phenomenology of judicial decision-making. Judges, it is said, in deciding cases and lawyers pressing them to decide in their favour, do not speak of the judge as ‘making’ the law even in novel cases. Even in the hardest of such cases the judge often betrays no awareness that there are, as the positivist suggests, two completely different stages in the process of decision: one in which the judge first finds that the existing law fails to dictate a decision either way; and the other in which he then turns away from the existing law to make law for the parties de novo and ex post facto according to his idea of what is best. Instead, lawyers address the judge as if he was always concerned to discover and enforce existing law and the judge speaks as if the law were a gapless system of entitlements in which a solution for every case awaits his discovery, not his invention.

There is no doubt that the familiar rhetoric of the judicial process encourages the idea that there are in a developed legal system no legally unregulated cases. But how seriously is this to be taken? There is of course a long European tradition and a doctrine of the division of powers which dramatizes the distinction between Legislator and Judge and insists that the Judge always is, what he is when the existing law is clear, the mere ‘mouthpiece’ of a law which he does not make or mould. But it is important to distinguish the ritual language used by judges and lawyers in deciding cases in their courts from their more reflective general statements about the judicial process. Judges of the stature of Oliver Wendell Holmes and Cardozo in the United States, or Lord Macmillan or Lord Radcliffe or Lord Reid in England, and a host of other lawyers, both academic and practising, have insisted that there are cases left incompletely regulated by the law where the judge has an inescapable though ‘interstitial’ law-making task, and that so far as the law is concerned many cases could be decided either way.

One principal consideration helps to explain resistance to the claim that judges sometimes both make and apply law and also elucidates the main features which distinguish judicial from a legislature’s law-making. This is the importance characteristically attached by courts when deciding unregulated cases to proceeding by analogy so as to ensure that the new law they make, though it is new law, is in accordance with principles or underpinning reasons recognized as already having a footing in the existing law. It is true that when particular statutes or precedents prove indeterminate, or when the explicit law is silent, judges do not just push away their law books and start to legislate without further guidance from the law. Very often, in deciding such cases, they cite some general principle or some general aim or purpose which some considerable relevant area of the existing law can be understood as exemplifying or advancing and which points towards a determinate answer for the instant hard case. This indeed is the very nucleus of the ‘constructive interpretation’ which is so prominent a feature of Dworkin’s theory of adjudication. But though this procedure certainly defers, it does not eliminate the moment for judicial law-making, since in any hard case different principles supporting competing analogies may present themselves and a judge will often have to choose between them, relying, like a conscientious legislator, on his sense of what is best and not on any already established order of priorities prescribed for him by law. Only if for all such cases there was always to be found in the existing law some unique set of higher-order principles assigning relative weights or priorities to such competing lower-order principles, would the moment for judicial law-making be not merely deferred but eliminated.

Dworkin’s other criticisms of my account of judicial discretion condemn it not as descriptively false but for endorsing a form of law-making which is undemocratic and unjust.⁷⁵ Judges are not usually elected and in a democracy, so it is claimed, only the elected representatives of the people should have law-making powers. There are many answers to this criticism. That judges should be entrusted with law-making powers to deal with disputes which the law fails to regulate may be regarded as a necessary price to pay for avoiding the inconvenience of alternative methods of regulating them such as reference to the legislature; and the price may seem small if judges are constrained in the exercise of these powers and cannot fashion codes or wide reforms but only rules to deal with the specific issues thrown up by particular cases. Secondly, the delegation of limited legislative powers to the executive is a familiar feature of modern democracies and such delegation to the judiciary seems a no greater menace to democracy. In both forms of delegation an elected legislature will normally have residual control and may repeal or amend any subordinate laws which it finds unacceptable. It is true that when, as in the USA, the legislature’s powers are limited by a written constitution and the courts have extensive powers of review a democratically elected legislature may find itself unable to reverse a piece of judicial legislation. Then ultimate democratic control can be secured only through the cumbrous machinery of constitutional amendment. That is the price which must be paid for legal constraints on government.

Dworkin makes the further accusation that judicial lawmaking is unjust and condemns it as a form of retrospective or ex post facto law-making which is, of course, commonly regarded as unjust. But the reason for regarding retrospective law-making as unjust is that it disappoints the justified expectations of those who, in acting, have relied on the assumption that the legal consequences of their acts will be determined by the known state of the law established at the time of their acts. This objection, however, even if it has force against a court’s retrospective change or overruling of clearly established law, seems quite irrelevant in hard cases since these are cases which the law has left incompletely regulated and where there is no known state of clear established law to justify expectations.

POSTSCRIPT NOTES

Page 272. [An alternative beginning to this section is included here, as it was not discarded.]

Throughout the long sequence of his writings on adjudication Dworkin has unswervingly maintained his denial that the courts have discretion in the sense of a law-creating power to decide cases left incompletely regulated by the existing law. Indeed he has argued that apart from some trivial exceptions there are no such cases, since as he has famously said, there is always a single ‘right answer’ to any meaningful question as to what the law is on any point of law arising in any case. ⁷⁶

But notwithstanding this appearance of an unchanging doctrine, Dworkin’s later introduction of interpretive ideas into his legal theory and his claim that all propositions of law are ‘interpretive’ in the special sense which he has given to this expression, has (as Raz was the first to make clear⁷⁷ brought the substance of this position very close to my own in recognizing that the courts in fact have and frequently exercise a law-creating discretion. Arguably before the introduction of interpretive ideas into his theory there seemed to be a great difference between our respective accounts of adjudication, because Dworkin’s earlier denial of judicial discretion in the strong sense and his insistence that there is always a right answer were associated with the idea that the judge’s role in deciding cases was to discern and enforce existing law. But this earlier conception which of course conflicted very sharply with my claim that the courts in deciding cases often exercise a law-creating discretion does not figure at all in

[The text of the alternative beginning to Section 6 ends at this point.]

POSTSCRIPT 3rd ed. NOTES

This is perhaps the only postscript about which a whole book has been produced:

Jules Coleman ed., Hart’s Postscript. The essays are all valuable in understanding not only the Postscript, but also Hart’s theory. See also Ronald Dworkin, ‘Hart’s Postscript and the Character of Political Philosophy’ (2004) 24 Oxford Journal of Legal Studies 1.

FOOTNOTES POSTSCRIPT

1. See my review of his The Morality of Law (1964), 78 Harvard Law Review 1281 (1965), reprinted in my Essays in Jurisprudence and Philosophy (1983), p. 343. [Note: Footnotes to the Postscript that are enclosed in square brackets were added by the editors.] ↩

2. See my ‘Law in the Perspective of Philosophy: 1776–1976’, 51 New York University Law Review 538 (1976); ‘American Jurisprudence through English Eyes: The Nightmare and the Noble Dream’, 11 Georgia Law Review 969 (1977); ‘Between Utility and Rights’, 79 Columbia Law Review 828 (1979). All the foregoing are reprinted in Essays in Jurisprudence and Philosophy. See also ‘Legal Duty and Obligation’, chap. VI in my Essays on Bentham (1982), and ‘Comment’ in R. Gavison (ed.), Issues in Contemporary Legal Philosophy (1987), p. 35. ↩

3. Cited hereinafter as TRS, AMP, and LE respectively. ↩

4. [Hart did not complete the second of the two sections mentioned here. See Editors’ Note.] ↩

5. See H. L. A. Hart, ‘Comment’, in Gavison, above, n. 2, p. 35. ↩

6. LE 102. ↩

7. LE chap. 3. ↩

8. LE 90. ↩

9. LE 90. ↩

10. TRS 66. ↩

11. LE 65–66. ↩

12. But Dworkin warns that the identification of such preinterpretive law may itself involve interpretation. See LE 66. ↩

13. LE 93. ↩

14. LE 94. ↩

15. But note that some critics, e.g. Michael Moore in his ‘The Interpretive Turn in Modern Theory: A Turn for the Worse?’, 41 Stanford Law Review 871 (1989), at 947–8, while accepting that legal practice is interpretive in Dworkin’s sense, deny that legal theory can be interpretive. ↩

16. LE 102; cf. ‘General theories of law, for us, are general interpretations of our own judicial practice.’LE 410. ↩

17. AMP 148; cf. ‘theories of law cannot sensibly be understood as . . . neutral accounts of social practice’, in ‘A Reply by Ronald Dworkin’, Marshall Cohen (ed.), Ronald Dworkin and Contemporary Jurisprudence (1983) [cited hereinafter as RDCJ], p. 247 at 254. ↩

18. [See LE 13–14.] ↩

19. [See Legal Reasoning and Legal Theory (1978), 63–4, 139–40.] ↩

20. R. M. Dworkin, ‘Legal Theory and the Problem of Sense’, in R. Gavison (ed.), Issues in Contemporary Legal Philosophy: The Influence of H. L. A. Hart (1987), at 19. ↩

21. Ibid. ↩

22. LE 6 ff. ↩

23. LE 4. ↩

24. LE 31 ff. ↩

25. LE 45. ↩

26. On this distinction see John Rawls, A Theory of Justice (1971), pp. 5–6, 10. [In distinguishing the concept of justice from conceptions of justice, Rawls states, ‘Here I follow H. L. A. Hart, The Concept of Law . . . pp. 155–159.’ (First edition.) See A Theory of Justice , p. 5 n. 1.] ↩

27. LE 418–19, n. 29. ↩

28. See LE 31–3. ↩

29. See p. 209 of this book, where I repudiate any such doctrine. ↩

30. [This phrase is Hart’s, and does not appear in LE.] ↩

31. TRS 17. ↩

32. 71 Harvard Law Review 598 (1958), reprinted in my Essays on Jurisprudence and Philosophy (see esp. pp. 54–5). ↩

33. LE 117. ↩

34. [LE 93.] ↩

35. LE 429 n. 3. ↩

36. See his replies to E. P. Soper and J. L. Coleman in RDCJ 247 ff. and 252 ff. ↩

37. RDCJ 248. ↩

38. See this book, pp. 123, 147–54.] ↩

39. [See this book, p. 128.] ↩

40. RDCJ 250. ↩

41. See J. Raz, ‘Dworkin: A New Link in the Chain’, 74 California Law Review 1103 (1986), at 1110, 1115–16. ↩

42. [See LE 13–14.] ↩

43. [See TRS 48–58.] ↩

44. TRS 51. ↩

45. [TRS 50–8; see this book, pp. 124–5.] ↩

46. [TRS 58.] ↩

47. [TRS 24.] ↩

48. TRS 24, quoting from Henningsen v. Bloomfield Motors, Inc., 32 NJ 358, 161 A.2d 69 (1960) at 387, 161 A.2d at 85. ↩

49. TRS 25–6. ↩

50. [Dworkin discusses whether the First Amendment is a rule or a principle at TRS 27.] ↩

51. [TRS 26.] ↩

52. TRS 24–7. ↩

53. Raz and Waluchow have emphasized this important point to which I had failed to draw attention. See J. Raz, ‘Legal Principles and the Limits of the Law’, 81 Yale LJ 823 (1972) at 832–4 and W. J. Waluchow, ‘Herculean Positivism’, 5 Oxford Journal of Legal Studies 187 (1985) at 189–92. ↩

54. 115 N.Y. 506, 22 N.E. 188 (1889); TRS 23; see also LE 15 ff. ↩

55. [For Dworkin’s discussion see TRS 22–8 and LE 15–20.] ↩

56. LE 265. ↩

57. [1932] A.C. 562. ↩

58. [See TRS 27.] ↩

59. See, e.g., E. P. Soper, ‘Legal Theory and the Obligation of a Judge’, RDCJ p. 3 at 16; J. Coleman, ‘Negative and Positive Positivism’, RDCJ p. 28; D. Lyons, ‘Principles, Positivism and Legal Theory’, 87 Yale Law Journal 415 (1977). ↩

60. [LE 65–6, 91–2.] ↩

61. [LE 72–3.] ↩

62. [LE 47, 67.] ↩

63. [LE 401.] ↩

64. LE 67. ↩

65. LE 53. ↩

66. RDCJ 260. ↩

67. RDCJ 259. ↩

68. [TRS 112, 128, and see TRS 93.] ↩

69. LE 78–9. ↩

70. [LE 103.] ↩

71. [LE 105.] ↩

72. [LE 105–6.] ↩

73. [An alternative version of the opening paragraph of this section appears in an endnote.] ↩

74. [TRS 81; cf. LE 37–9.] ↩

75. [TRS 84–5.] ↩

76. [See his ‘No Right Answer?’ in P. M. S. Hacker and J. Raz (eds.), Law, Morality and Society (1977), pp. 58–84; reprinted with revisions as ‘Is There Really No Right Answer in Hard Cases?’ AMP, chap. 5.] ↩

77. [See J. Raz. ‘Dworkin: A New Link in the Chain’, 74 California Law Review, 1103 (1986) at 1110, 1115–16.] ↩