One barrier to greater dialogue and understanding within Jurisprudence is the inability to appreciate the variety of forms and purposes among the different theories of or about (the nature of) law. Legal theorists are frequently to blame for the failure of clearer discussion about methodological issues, as they are often not as clear as they might be regarding the nature of claims they are making (e.g., descriptive versus prescriptive, conceptual versus empirical) or regarding the larger project of which their theories are a part. This entry attempts to offer a rough overview of the types and purposes of legal theory. The focus is primarily on theories about the nature of law. Other types of theories – e.g., regarding rights or the best approach to legal or judicial reasoning – may warrant a different analysis.
In general, one might divide theories about social practices and institutions roughly into three broad categories: (1) descriptive theories – theories that purport to state what is the case, offering an overview of current practices or understandings; (2) analytical or conceptual theories – theories that make claims about the intrinsic or necessary nature of some practice or institution; (3) theories which contain elements both of description and prescription; and (4) purely prescriptive, normative or critical theories – theories that argue for how practices or understandings should be reformed. (As will be discussed, the second category, analytical/conceptual theories, can also be seen as a subset of the first category, descriptive theories.) The categories will be discussed, in turn, in the coming sections (1.1, 1.2, 1.3, 1.4). Later sections will offer brief discussions of the related questions of whether methodological questions are specific to legal theory (2.0), the ontology of law (3.0), and the purposes of legal theory (4.0).
1.1 purely descriptive theories
Legal theorists often refer to their theories as “descriptive” – but “descriptive” comes in many variations, some of which, like analytical and conceptual theories, are sufficiently distinctive that they will be discussed separately in the next section, 1.2. There are also interesting types of theories that seem to be neither descriptive nor prescriptive, but to be, in some ways, in-between. Those types of theories are discussed in section 1.3.
In general terms, a theory is “descriptive” if it purports to describe what is the case, rather than to make judgments about the (moral or other) value of the current situation, or to offer arguments for how things should be done differently. Many types of theories that are about law, but are not “jurisprudential theories” (narrowly understood), are clearly descriptive: e.g., sociological, anthropological, and psychological theories about the way people behave in legal roles or in response to legal regulation; and historical accounts of why particular legal systems developed the way they did.
When one is offering a theory meant to range over a large number of instances of some institution or practice (across jurisdictions, or over time), there is always the problem of how to combine the data. For example, how does one have a “theory of law,” when legal systems (however understood) clearly differ from country to country and in any given country over time? To try to offer a purely descriptive theory of a vast social practice like law seems at risk of becoming little “more than a conjunction of lexicography with local history”. (Finnis 1980, 4).
One needs some means of organizing the data that is the subject of one’s descriptive theory, and there are debates within the literature regarding how such selection can or should be done. For example, John Finnis accused Hans Kelsen of having erred in his construction of a theory of law, in that Kelsen purportedly tried to find a “lowest common denominator” – that which was common to all legal systems – rather than doing as Finnis argued should be done in such cases: finding what was characteristic of law in its fullest or most mature instantiation, even if some, or even many, legal systems did not have all of these characteristics. (Finnis 1980, 9-11)
Also, though a theory may not be “prescriptive” in the sense of suggesting reforms of current practices, to make some point about nature or purpose, the theory will likely emphasize some aspects of the practice and/or downplay others. For this purpose, many writers (e.g., Waluchow 1994, 15-29; Raz 1994, 219-221) differentiate forms of non-moral evaluation needed to construct descriptive, conceptual or analytical theories from any sort of moral prescription.
H. L. A. Hart famously argued for organizing a theory of law by viewing the practice at least in part from the perspective of a participant who “accepts” the legal system as giving him or her reasons for action. (Hart 1994, 79-91) Though this “hermeneutic” approach to theory construction is not universally accepted, even Ronald Dworkin and John Finnis, who criticize Hartian legal positivism in particular and the project of purely descriptive legal theory in general, accept the idea of building a theory around an insider’s perspective (though their ideas about how to build on an insider’s perspective differ in important ways from Hart’s views). (Dworkin 1986; Finnis 1980, 3-18; see generally Bix 1999).
1.2 ANALYTICAL OR CONCEPTUAL THEORIES
Many jurisprudential theories purport to offer true claims about law generally. Such theories are usually making analytical or conceptual claims about law, as opposed to making a claim that applies only to a particular legal system — at a particular moment in time. (Also, there are some theorists for whom it is not clear whether their theories are best understood as conceptual or as “merely” descriptive — e.g., this seems to be a matter of ongoing debate regarding the best characterization of John Austin’s work (Cotterrell 2003, 81-83).)
“Analytical” or “conceptual” theories usually purport to be “descriptive”, in the sense that they purport to describe the way things are rather than to criticize or to prescribe. However, such theories are usually not “merely descriptive”, in the sense that the theorists is doing more than merely reporting data or observations.
Conceptual analysis usually involves a philosophically ambitious claim that the theory has captured what is “essential” to some concept of practice, characteristics “necessary” for a practice or institution to warrant the label in question. While such claims about “nature” and “essence” were traditionally associated with Platonic metaphysics, there are less metaphysically ambitious modern versions of such claims. (e.g., Bix 2003b). For example, Raz (1996) defends an understanding of legal theory as conceptual analysis, and in doing so argues that it such theories try to explain “our concept of law”, not some universal or timeless (Platonist) concept of law. For Raz, legal theory is an attempt to get a clearer insight into an idea that is central to a community’s self-understanding.
Hans Kelsen’s “pure theory of law” (reine Rechtslehre) (e.g., Kelsen 1967; 1992) can also be seen as a special form of analysis, one grounded on a neo-Kantian methodology. At its essence, Kelsen’s theory is an effort to determine what follows from the fact that people sometimes treat the actions and words of other people (legal officials) as valid norms.
There have been a number of challenges to the value or tenability of purely descriptive or descriptive-conceptual theories in jurisprudence. Some of the challenges have come from various versions of natural law theory (e.g., Finnis (2000)), an approach that asserts that moral objectives and moral ideals are inherent to the nature of law, and therefore central to its understanding.
John Finnis (2003) has also offered a separate challenge: that if the descriptive/conceptual theory of legal positivism is understood as determining the nature of law, where this “law” is understood as separate from the normative question of how judges should decide cases (see, e.g., Raz 1998, 4-6) or how citizens should act in the face of government decrees, then this is an uninteresting and unworthy inquiry.
A different sort of challenge has come from those who doubt conceptual analysis either generally, or at least in areas where the concepts have normative overtones. The general challenge comes from naturalism (not to be confused with natural law theory, naturalism argues for a more empirical or scientific approach to topics, like epistemology, formerly approached in an a priori or conceptual way), and has been advocated in legal philosophy primarily by Brian Leiter (2003). Leiter, building on the well-known philosophical work of W. V. O. Quine (1951), claims that there are no “conceptual” truths to discover, and that theorists about law (or judicial reasoning or other legal phenomena) should confine themselves to empirical investigations of actual practices.
The “normative” challenge to descriptive conceptual analysis asserts that conceptual analysis of a concept like “law” (or “democracy” or “justice”) is inevitably contestable and evaluative. (e.g., Dworkin 2004; Perry 1998). Stephen Perry argues that one must inevitably choose among alternative tenable theories about law, and that this selection must be made on political or moral grounds. Ronald Dworkin’s interpretive theory of law portrays theories about (the nature of) law as theories that attempt to show the value of legality as part of a larger web of political and moral values.
1.3 BETWEEN DESCRIPTION AND PRESCRIPTION
Despite the commonly accepted distinction between description and prescription, there are a number of approaches and types of theory that seem to lie uneasily somewhere on the spectrum between “purely descriptive” and “purely prescriptive”: (1) “Semantic theories.” Ronald Dworkin famously characterized H. L. A. Hart’s theory of legal positivism as being best understood as a semantic theory – a definition about the meaning of the word “law.” (Dworkin 1986; 2004) However, this characterization was rejected by Hart himself, as well as most commentators. (Hart 1994, 244-248; Endicott 2001; for a defence of Dworkin’s claim, see Stavropoulos 2001). At a minimum it is worth noting that no theorist proferring a theory of law has characterized his or her own theory as being (“merely”) a definition of the word “law.”
(2) Variations on descriptive theory. Even descriptive theories themselves often seem to deviate from pure description. As was summarized earlier, the construction of a theory about some social institution or practice requires some amount of selection or simplification: (a) to prevent the theory becoming simply a messy restatement of complex reality; (b) to allow the theory to extract some basic insight about the institution or practice; and/or (c) to allow the theory to focus on the “fullest” or “highest” instantiation of the institution or practice, rather than what appears to be common to all instantiations. These forms of selection within descriptive theory come under various titles: “principles of theory construction” (e.g. Waluchow 1994, 19-21), emphasizing important features, and Weber’s “ideal types” (cf. Finnis 1980, 9-11).
(3) Rational Reconstruction. Within doctrinal legal scholarship, it is conventional (in many countries) for commentators to try to restate court decisions, or whole areas of law, in a way largely consistent with the outcomes of the cases, but restating the justifications offered to make them more persuasive. (In common law countries, such restatements of areas of law that had been primarily developed by judges – like Contract Law, Tort Law (e.g., Owen 1995), and Property Law – sometimes go under the title, “Philosophical Foundations of the Common Law.”)
(4) Ronald Dworkin’s Interpretive Approach. Dworkin’s influential legal theory (e.g., Dworkin 1986) is grounded on the “constructive interpretation” of official actions. (“Constructive interpretation” is the principle Dworkin would have applied not only for resolving legal disputes, but also for constructing theories about the nature of law.) For law, Dworkin would have judges decide cases by first finding the theory of past official actions (legislation, constitutional provisions, and judicial decisions) related to the dispute that would simultaneously adequately fit those past actions while making that area of law the best it can be (morally or politically). Dworkin’s interpretive approach, if followed in practice, would likely approximate “rational reconstruction”.
1.4 Prescriptive theories
Along with descriptive theories, and the variations of descriptive theories, described above, theories can of course be prescriptive: not focusing on describing current practices, but focusing instead on urging a new or reformed practice. The role of such theories in modern legal theory will be discussed more fully within the overview of purposes of legal theory (4.0).
2.0 ARE THE QUESTIONS SPECIFIC TO LEGAL THEORY?
Many legal theorists seem insufficiently attentive to the work already done elsewhere in social theory regarding the problems of theory-construction. For the most part, theories about law will raise the same questions as theories in other social sciences. For example, one way to distinguish theories is to take into account different schools of social theorizing: e.g., whether the focus is on individuals or on structures, and whether the basic account of social action is behavioristic or hermeneutic (also sometimes called “interpretive” or “Verstehen”). (Bix 2003a, 7-8; Lucy 1999, 17-32). This distinction has its greatest force in theories of social action within legal systems – e.g., theories of judicial behaviour – but it also has implications for more abstract theories of law. For example, H. L. A. Hart’s criticism of John Austin’s command theory of law, and Hart’s subsequent development of his own version of legal positivism, is centrally understood in terms of a hermeneutic rejection of a more empirical or more “scientific” approach. (Hart 1994, 18-123; Bix 1999) Already mentioned is the critique by some commentators of conceptual theories of law on the basis that such theories are generally unsupportable in law.
This is not to claim that there is nothing distinctive about law or legal theory. For example, theories of law may be distinctive in that law seems to function both as a kind of social institution and as central to our practical reasoning (e.g., Finnis 2000, 1602-03); and this “double life” may be a key to understanding the difficulty in constructing a theory about the nature of law.
3.0 ontology (Basic building blocks)
One type (or topic) of legal theory involves a metaphysical (ontological) explanation of law and legal concepts. The Scandinavian legal realists, in particular, focused on this question, though aspects of such questions can be found in a wide range of other theorists. The Scandinavian realists, building on views that paralleled (though did not equate with) logical positivism, were sceptical of entities that could not be understood in terms of observable, empirical data. (e.g., Hägerström 1953, Ross 1957, Lundstedt 1956, Olivecrona 1971)
If one rejects the metaphysical/ontological challenge of the Scandinavian legal realists, one is still faced with questions regarding the metaphysical status of concepts and claims used in legal practice. This inquiry is sometimes presented from a different perspective: in terms of the nature of legal truth, or even general questions of legal reasoning. For legal truth: when one says that a certain law-related claim is correct (“X has a right to possession of A” or “there is a valid contract between R and S”), in light of what is that claim true or false? None of the obvious alternatives seems attractive. On one extreme, already discussed, is the metaphysically sceptical position that demands that legal concept be reduced to observable, empirical terms. A different sort of sceptical view would argue for reducing legal concepts to descriptions and predictions of official actions. This perspective, sometimes described as a “predictive theory of law”, has some initial attractions, but also well-known weaknesses. (e.g., Hart 1994, 88-91)
At the other extreme would be a kind of Platonism in which legal concepts are thought to correspond with metaphysical entities (one modern theorists whose work sometimes comes close to this sort of Platonism is Michael Moore (e.g., 2000)). Related efforts try to ground the objectivity underlying legal concepts through a kind of “natural kinds” analysis, derived from theories of meaning and reference. (Stavropoulos 1996)
Various attempts have been offered to create a middle position – one purportedly more in line with common understandings of terms, without requiring commitment to ambitious or unusual metaphysical claims. Among the more prominent such theories are institutional fact theories (e.g., MacCormick & Weinberger 1986) and Dworkin’s interpretive theory of law (Dworkin 1986). Alternatively, some theorists have simply urged that one can avoid possible ontological issues by equating the meaning of legal terms and concepts with the rules for their use – a response to metaphysical questions similar to that urged by Ludwig Wittgenstein. (Hart 1954; Bix 1995)
4.0 Purposes of legal theory
As earlier mentioned, the topic of the purposes of legal theory is intimately tied up with the topic of the nature(s) of legal theory. A wide variety of purposes are served by theories about law. Some theories, especially those purporting to be analytical or descriptive, can be justified on the narrow basis of seeking truth and knowledge. Here such knowledge can be seen as the straightforward collection of facts that would result from a simple description, or the deeper sort of insight or understanding that might result from a quasi-descriptive model. (However, such knowledge, once gained, might also then play a role in an evaluative or prescriptive theory — a point important to the writings of a number of theorists, including both Jeremy Bentham (1996 ) and H. L. A. Hart (1958).)
In discussing types of descriptive theories, including analytical and conceptual theories (1.1, 1.2, 1.3), it was noted that many commentators argue that it is necessary, or at least valuable, for the construction of a theory of law to involve some amount of evaluation and selection. This view entails certain ideas about theories of law: that a certain lack of fit with the data is an acceptable cost for the insight a good theory might offer. (A comparable point is often made generally about modelling of behaviour, both in the physical sciences and the social sciences.)
A related point: jurisprudential theories are often offered as “explaining” a social practice or institution. “Explanation” is a central, but frequently poorly articulated, notion in discussing the point of descriptive theories. The reason one might put up with some simplifications, or even distortions, of the empirical reality in a (descriptive) theory is the benefit a good theory can offer by way of “insight” – showing something central to the nature of a social practice, or at least something interesting all instances of some category of practices or institutions seem to share.
Relevant to the earlier discussion of “rational reconstruction” (1.3), it should be noted that it often plays a significant role in the teaching of law and the training of legal advocates. However, the “rational reconstruction” used in teaching legal advocacy might differ in small but significant ways from the one offered for purely scholarly purposes. An advocate must have an eye not only to the best reconstruction of a muddled doctrinal area, but also the reconstruction that would seem best to the judges this advocate would face. Thus, if the best reconstruction of the prior cases would justify some right, but the current members of the country’s highest court are unlikely to recognize the right, the advocate might be better served by a reconstruction that excludes that right (at least until the membership on the highest court changes).
Critical theories of law aim more towards reform of current laws and practices rather than (mere) increased knowledge of or insight into those laws and practices. Many of the influential American legal theories of the 20th and 21st century – from American legal realism to law and economics, and including critical legal studies, critical race theory, and feminist legal theory – are best understood as being essentially criticisms of the current approach to legal regulation and/or judicial decision-making, combined with suggestions for how the system could be improved. (The reform- or justice-centred nature of critical legal studies, critical race theory and feminist legal theory are relatively self-evident. American legal realism and, especially, law and economics, may be harder cases, as both include claims that seem to be descriptive or analytical – that seem to be claims about the basic nature of rules, decision-making or law generally. However, both schools of thought are grounded on a view that law is instrumental, and the question quickly becomes, for theorists in both schools, either which ends law should pursue or how best to achieve the ends already chosen.)
When the first edition of Neil MacCormick’s H.L.A. HART appeared in 1981, it was notable for being the first book-length treatment of Hart’s legal philosophy, which was comprised largely of a sympathetic analysis of Hart’s most famous and enduring work, THE CONCEPT OF LAW (1961). As such, MacCormick’s book filled an important gap in the then existing literature for an accessible, introductory text on the main themes in Hart’s jurisprudence.
Nearly thirty years later, and sixteen years after Hart’s death, the scholarly production of critical literature on his thought has grown to enormous proportions, including several monographs, countless journal and law review articles and, most recently, a comprehensive biography, Nicola Lacey’s widely acclaimed A LIFE OF H.L.A. HART: THE NIGHTMARE AND THE NOBLE DREAM (2004). While the intellectual landscape has thus changed considerably, the need for a reliable introductory text on Hart’s contributions to jurisprudence is no less necessary. In this regard, the second edition of MacCormick’s book, updated to account for the major developments since its original publication, such as the posthumously published Postscript to THE CONCEPT OF LAW, continues to fill an important niche in Hartian scholarship.
MacCormick, who attended Hart’s lectures at Oxford as a graduate student in the early 1960s, does not undertake a complete survey of Hart’s scholarship, but instead aims more modestly at a “friendly” introductory account that provides a “sympathetic reconstruction of Hart’s main ideas” (p.13). At the same time, the analysis is not without critical bite, since MacCormick repeatedly claims throughout the book that “Hartian doctrine . . . points in the right direction but does not take us far enough” (p.159). He thus purports at various points (about which more below) to extend Hartian insights in a more rigorous and thoroughgoing manner than Hart himself did. The end result, as MacCormick admits, is a substantially amended conception of law. Indeed, Hart’s reaction to the first edition was to insist that “he considered himself a more hardened positivist than MacCormick had depicted,” which “made him out to be more of a natural lawyer than he wanted to be” (p.15). The present edition would no doubt elicit the same reaction.
As in the first edition, the heart of the book remains MacCormick’s reconstruction of Hart’s general theory of the structure of modern municipal legal systems as “a union of primary and secondary rules,” which is set forth primarily in Chapters 9 through 11. In an “inversion of Hart’s order of [*42] proceeding” in THE CONCEPT OF LAW, this discussion is preceded by a series of chapters examining “the building blocks of Hart’s theory of legal order” (pp.61, 117), namely his conception of social rules, his views on positive and critical moral theory, and the notions of obligation, duty and wrongdoing, power-conferring rules, and rights. It is followed by chapters devoted to an exposition of Hart’s theory of crime and punishment, the relation of law and morality, and an epilogue addressing the methodological concerns discussed in the Postscript.
Before turning to these issues, MacCormick begins, usefully I think, with a brief biographical sketch that sets the stage by reviewing Hart’s early career as a practicing lawyer, his credentials as a proponent of egalitarian social democracy, which crucially informed his reformist impulses, and the philosophical milieu in postwar Oxford, which was “the intellectual context to which his analytical jurisprudence belongs” (p.23). Perhaps surprisingly, in discussing the distinction between “law” and “politics” that figures so prominently in American constitutional discourse, MacCormick candidly observes that Hart’s views on the relation of law and political morality did not escape the parochial concerns of his time and place. “Though he claimed it applied to legal systems quite generally,” MacCormick writes, “Hart’s theory of law bears some of the marks of the . . . unspoken assumptions of the English lawyer” and is therefore “clearly recognizable as the work of an English lawyer of the twentieth century” (pp.8, 10).
While this might have been nothing more than a modest acknowledgment that no scholar can really lay claim to an Archimedean vantage point from which to evaluate social phenomena, it raises some thorny issues for Hartian jurisprudence. Stephen Perry (1996) has usefully distinguished between methodological and substantive versions of positivism, which are logically distinct. On the methodological side, as MacCormick suggests, Hart’s theory makes a claim to universality, in the sense that all genuine legal phenomena are assumed to possess a certain set of shared features or characteristics, regardless of time and place. In this view, the task of the legal theorist is to provide a morally neutral “descriptive account of what societies living under law all have in common” (p.210).
To be sure, such a description should take into account the participant’s perspective, the so-called “internal point of view,” which may or may not include normative considerations, but any such connections between law and morality are a strictly contingent matter. Hart is not much concerned with the reasons why officials accept properly pedigreed social rules. In particular, while moral considerations might be a criterion of validity in some legal systems (e.g., the Bill of Rights to the U.S. Constitution), this is by no means a conceptual necessity. Hartian reportage thus aspires to be a scientific, explanatory-descriptive enterprise that may be conducted independently of any context-dependent features of a particular legal system, such as the moral and political values which it subserves.
But serious reservations have been raised about the plausibility of this entire [*43] approach to jurisprudence. In the first place, as Perry points out, existing legal systems are artifacts of human cultural construction, rather than a “natural kind,” such as the elements of the periodic table, each instantiation of which may be said to share some common essence or function. For this reason, a general theory of law as such may not be a scientifically fruitful explanatory category, perhaps no more enlightening, as Brian Bix (1999) puts it, than “a theory of all objects that begin with the letter ‘N’.” Whatever jurisprudes are up to, it does not seem to be science in any ordinary sense of the term. (But see Brian Leiter (2003) for a vigorous dissenting view).
Moreover, whatever might plausibly be said about all legal systems, past and present and across cultures, is likely to be so abstract as to be without much practical significance. As noted above, Hart readily concedes that legal phenomena must be understood and described from a hermeneutic point of view, with the important proviso that the theorist need not personally embrace this perspective. Be that as it may, this is always the perspective of those persons who are actual participants in a particular legal culture. In understanding legal phenomena, we always start from what Perry calls a “local methodological stance,” since “it is by no means evident how we would go about formulating pre-theoretical propositions about ‘all’ legal systems.” With such an understanding firmly in hand, we might well be in a position to formulate a more general concept of law in a comparative fashion, but since legal institutions and practices are intelligible only in view of some inherently contestable function, value, or purpose, such as the promotion of justice and the common good, such an inquiry seems to be an unavoidably normative enterprise.
It is not clear to me precisely where MacCormick stands on this issue. On the one hand, he seems to insist, with Hart, that the validity and content of law are strictly matters of social fact. This, in turn, leaves open the possibility that “faithful reports can be given about the character and content of some body of law even by somebody who has no commitment to the particular values in which these laws are grounded” (p.204). From this it also follows that “the bare existence of a legal system as a system of rules carries no guarantee concerning the substantive justice or moral satisfactoriness of the content of these rules” (p.208).
On the other hand, Hart defended the separability thesis, at least in part, in frankly normative terms, namely as a warning against “the risk of moral complacency about the uses of the concept of law.” Hart’s positivism thus remains “grounded in practical, indeed moral, concerns, not purely epistemic ones,” and MacCormick concedes that it is not clearly “tenable to base one’s methodology on a claim about the moral basis for insisting on detached juristic inquiry” (p.209).
MacCormick then takes this line of argument a step further. Following John Finnis, he says that “it can be argued most persuasively that law must in principle be oriented toward the common good of the community whose law it is and seek to realize justice among its members” (pp.208-209) (emphasis added). Understood in this way, the law as it is actually [*44] implemented in any given jurisdiction can and often does fall short of the distinctive values it ought to realize, but if an aspiration toward justice and the common good is an inherent feature of the concept of law, then the status of purely descriptive conceptual analysis seems uncertain. Indeed, MacCormick (2007) claims elsewhere that where a given law or set of laws “cannot be accounted for under any possible conception of justice that could reasonably be adopted or advocated by a reasonable person willing to subject his or her beliefs to discursive scrutiny, then what is thus done by way of rules and practices of governance would not properly count as law.” In such “extreme cases,” he says, the ostensible laws “carry no element of genuine obligation with them, though they may be backed by coercion enough” (p.209).
This is a congenial conclusion, perhaps, but it is certainly not Hart’s view of the matter. As is widely known, Hart’s project in THE CONCEPT OF LAW was to rescue positivism from the reductive Austinian formulation according to which law was essentially an order backed by an effective threat of coercion. The problem with this picture, Hart pointed out, is that one’s being obliged to obey a command under the threat of sanctions in the event of noncompliance in no sense gives rise to a genuine obligation. In order to distinguish habitual obedience, motivated perhaps by fear of punishment, from actual rule-following behavior, Hart insists that those to whom the rules apply must adopt a certain “reflective critical attitude,” namely they must willingly accept the rules as a shared or common “standard of behavior.”
In a pre-legal society, the “primary rules” of behavior are obligatory essentially because they are accepted as legitimate by the members of a group as forming part of its conventional morality. By contrast, in a modern legal system, Hart makes no appeal to the content of the primary rules. Instead, such rules are obligatory by virtue of their origin, because they are properly enacted according to a valid secondary rule, which is itself accepted as a social rule of the group.
Moreover, it is sufficient if the public officials who are responsible for the formulation and implementation of the primary rules – legislators, judges, and lawyers – willingly accept the secondary rules in a normative sense. For the system to function, the mass of the citizenry need only generally obey, although Hart allows that in a healthy society, most citizens will also adopt the internal point of view. In the last analysis, however, being under a legal obligation is, after all, to be subject to sanctions for disobedience, even for the members of a systematically oppressed minority. As Hart soberly observes, “coercive power, thus established on its basis of authority . . . may be used to subdue and maintain, in a position of permanent inferiority, a subject group . . . For those thus oppressed there may be nothing in the system to command their loyalty but only things to fear. They are its victims, not its beneficiaries.” Although MacCormick bristles at the suggestion (pp.194-195), it is difficult to escape the conclusion that Hart’s theory of law collapses into the very Austinian model he intended to reject.
These are difficult questions and much more would have to be said to make [*45] these brief comments convincing. But my suggestion is that, if Hart never really escaped the perspective of the typical socially liberal, upper middle class English lawyer of his day, as MacCormick says, then perhaps the substantive values he ascribes to the concept of law were not and could not have been as descriptively neutral as he supposed.
Stephen Perry’s commentary raised a number of quite interesting issues about the right way to understand Raz’s account of the authority of law and its compatibility with the kind of “descriptive” jurisprudence to which Raz and Hart (and me, in a different way) are committed. When Professor Perry’s paper is publicly available, I will try to address some of those issues, but for now I wanted to at least record one point on which I clearly misconstrued Perry in Chapter 6 of my book (“Beyond the Hart/Dworkin Debate: The Methodology Problem in Jurisprudence”). In an earlier paper, Perry had written: “If [the service conception of authority] is right, then the anarchist thesis that the state could never have the moral authority it claims is wrong.” I critiqued this as follows (p. 172 of my book):
Raz’s account of authority is perfectly compatible with ‘the anarchist thesis that the state [more precisely, the laws of the state] could never have the moral authority it claims,’ because Raz’s thesis is only that all laws (sincerely) claim moral authority, not that they actually have it. The anarchist thesis, in Razian terms, is simply the claim that law always fails to satisfy the Normal Justification Thesis. Nothing in Raz’s theory of authority or of law precludes it.
What Perry meant, however, is that the anarchist thesis that authority is impossible is wrong if Raz’s service conception of authority is plausible, and that’s right: the service conception of authority explains how someone can have a justified claim of authority over another (rational, autonomous etc.) person, which the (Wolffian) anarchist denies is possible. It is true that the service conception of authority is compatible with the anarchist claim that no state ever has authority, but that was not, in fact, the anarchist thesis at issue for Perry.
. L. A. Hart is heir and torch-bearer of a great tradition in the philosophy’ of law which is realist and unromantic in outlook. It regards the existence and content of the law as a matter of social fact whose connection with moral or any other values is contingent and precarious. His analysis of the concept of law is part of the enterprise of demythologizing the law, of instilling rational critical attitudes to it. Right from his inaugural lecture in Oxford he was anxious to dispel the philosophical mist which he found in both legal culture and legal theory . In recent years he has shown time and again how much the rejection of the moralizing myths which accumulated around the law is central to his whole outlook. His essays on ‘Bentham and the Demystification of the Law’ and on ‘The Nightmare and the Noble Dream’ showed him to be consciously sharing the Benthamite sense of the excessive veneration in which the law is held in common-law countries, and its deleterious moral consequences. His fear that in recent years legal theory has lurched back in that direction, and his view that a major part of its role is to lay the conceptual foundation for a cool and potentially critical assessment of the law are evident.
This attitude strikes at the age-old question of the relation between morality and law. In particular it concerns the question whether it is ever the case that a rule is a rule of law because it is morally binding, and whether a rule can ever fail to be legally binding on the ground that it is morally unacceptable. As so often in philosophy, a large part of the answer to this question consists in rejecting it as simplistic and misleading, and substituting more complex questions concerning the relation between moral worth and legal validity. Let us, however, keep the simplistic question in mind; it helps to launch us on our inquiry.
Three theses with clear implications concerning the relation between law and morality have been defended in recent years. They can be briefly, if somewhat roughly, stated as follows:
The sources thesis: All law is source-based.
The incorporation thesis: All law is either source-based or entailed by’ source-based law.
The coherence thesis: The law consists of source-based law together with the morally soundest justification of source-based law.
A law is source-based if its existence and content can be identified by’ reference to social facts alone, without resort to any’ evaluative argument. All three theses give source-based law a special role in the identification of law’. But whereas the parsimonious sources thesis holds that there is nothing more to law than source-based law, the other two allow that the law can be enriched by non-source-based laws in different ways. Indeed, the coherence thesis insists that every legal system necessarily includes such laws.
The main purpose of this essay is to defend the sources thesis against some common misunderstandings4 and to provide one reason for preferring it to the other two. The argument turns on the nature of authority, which is the subject of the first section. In the second section some of the implications of this analysis are shown to be relevant to our understanding of the law. Their relation with the three theses is then examined. The connection between law and authority is used to criticize Dworkin’s support of the coherence thesis, as well as the incorporation thesis advocated by Hart and others. The rejection of these views leads to the endorsement of the sources thesis. The essay concludes with some observations concerning the relations between legal theory, law, and morality. Throughout, the argument is exploratory rather than conclusive.
I. AUTHORITY AND JUSTIFICATION
Authority in general can be divided into legitimate and de facto authority. The latter either claims to be legitimate or is believed to be so, and is effective in imposing its will on many over whom it claims authority , perhaps because its claim to legitimacy is recognized by many of its subjects. But it does not necessarily possess legitimacy. Legitimate authority is either practical or theoretical (or both). The directives of a person or institution with practical authority are reasons for action for their subjects, whereas the advice of a theoretical authority is a reason for belief for those regarding whom that person or institution has authority. Though the views here expressed apply to theoretical authorities as well, unless otherwise indicated I shall use ‘authority’ to refer to legitimate practical authority . Since our interest is in the law we will be primarily concerned with political authorities. But I shall make no attempt to characterize the special features of those, as opposed to practical authorities in general or legal features of those, as opposed to practical authorities in general or legal authorities in particular.
The distinction between reasons for action and reasons for belief may’ be sufficient to distinguish between practical and theoretical authorities, but it is inadequate to distinguish between authorities and other people. Anyone’s sincere assertion can be a reason for belief, and anyone’s request can be a reason for action. What distinguishes authoritative directives is
their special peremptory status. One is tempted to say that they are marked by their authoritativeness. This peremptory character has other led people to say that in accepting the authority of another one is surrendering one’s judgment to him, that the acceptance of authority is the denial of one’s moral autonomy, and so on. Some have seen in these alleged features of authority a good deal of what often justifies submitting to authority. Many more derived from such reflections prove that acceptance of authority is wrong, or even inconsistent with one’s status as a moral agent. Elsewhere I have developed a conception of authority which accounts for its peremptory force while explaining the conditions under which it may be right to accept authority. Let me briefly repeat the main tenets of this conception of authority. Its details and the arguments in its support cannot be explored here.
Consider the case of two people who refer a dispute to an arbitrator. He has authority to settle the dispute, for they agreed to abide by his decision. Two features stand out. First, the arbitrator’s decision is for the disputants a reason for action. They ought to do as he says because he says so. But this reason is related to the other reasons which apply to the case. It is not just another reason to be added to the others, a reason to stand alongside the others when one reckons which way is better supported by reason. The arbitrator’s decision is meant to be based on the other reasons, to sum them up and to reflect their outcome. He has reason to act so that his decision will reflect the reasons which apply to the litigants. I shall call reasons of the kind which apply to the arbitrator dependent reasons. I shall also refer to his decision as a dependent reason for the litigants. Notice that in this second sense a dependent reason is not one which does in fact reflect the balance of reasons on which it is based. It is one which is meant to, i.e . which should, do so.
This leads directly to the second distinguishing feature of the example. The arbitrator’s decision is also meant to replace the reasons on which it depends. In agreeing to obey his decision, the disputants agreed to follow his judgment of the balance of reasons rather than their own. Henceforth his decision will settle for them what to do. Lawyers say that the original reasons merge into the decision of the arbitrator or the judgment of a court, which, if binding, becomes res judicata. This means that the original cause of action can no longer be relied upon for any purpose. I shall call a reason which displaces others a preemptive reason.
It is not that the arbitrator’s word is an absolute reason which has to be obeyed come what may. It can be challenged and justifiably disobeyed in certain circumstances. If, for example, the arbitrator was bri, was drunk wconsidering the case, or if new evidence of great importance unexpectedly
turns up, each party may ignore the decision. The point is that reasons that could have been relied upon to justify action before his decision cannot be relied upon once the decision is given. Note that there is no reason for anyone to restrain their thoughts or their reflections on the reasons which apply to the case, nor are they necessarily debarred from criticizing the arbitrator for having ignored certain reasons or for having been mistaken about their significance. It is merely action for some of these reasons which is excluded.
The two features, dependence and preemptiveness, are intimately connected. Because the arbitrator is meant to decide on the basis of certain reasons, the disputants are excluded from later relying on them. They handed over to him the task of evaluating those reasons. If they do not then reject those reasons as possible bases for their own action, they defeat the very point and purpose of the arbitration. The only proper way to acknowledge the arbitrator’s authority is to take it to be a reason for action which replaces the reasons on the basis of which he was meant to decide.
The crucial question is whether the arbitrator’s is a typical authority, or whether the two features picked out above are peculiar to it, and perhaps a few others, but are not characteristic of authorities in general. It might be thought, for example, that the arbitrator is typical of adjudicative authorities, and that what might be called legislative authorities differ from them in precisely these respects. Adjudicative authorities, one might say, are precisely those in which the role of the authority is to judge what are the reasons which apply to its subjects and decide accordingly, i.e. their decisions are merely meant to declare what ought to be done in any case. A legislative authority, on the other hand, is one whose job is to create new reasons for its subjects, i.e. reasons which are new not merely in the sense of replacing other reasons on which they depend but in not purporting to replace any reasons at all. If we understand ‘legislative’ and ‘adjudicative’ broadly’, so the objection continues, all practical authorities belong to at least one of these kinds. It will be conceded, of course, that legislative authorities act for reasons. But theirs are reasons which apply to them and which do not depend on, i.e. are not meant to reflect, reasons which apply to their subjects.
The apparent attractiveness of the above distinction is, however, misguided. Consider an Act of Parliament imposing on parents a duty to maintain their young children. Parents have such a duty independently of this Act, and only because they have it is the Act justified. Further argument is required to show that the same features are present in all practical authorities.
Instead, let me summarize my’ conception of authority in three theses:
The dependence thesis:
All authoritative directives should be based, among other factors, on reasons which apply to the subjects of those directives and which bear on the circumstances covered by the directives. Such reasons I shall call dependent reasons.
The normal justification thesis:
The normal and primary way to establish that a person should be acknowledged to have authority over another person involves showing that the alleged subject is likely better to comply with reasons which apply to him (other than the alleged authoritative directives) if he accepts the directives of the alleged authority as authoritatively binding, and tries to follow them, than if he tries to follow the reasons which apply to him directly.7
The preemption thesis:
The fact that an authority requires performance of an action is a reason for its performance which is not to be added to all other relevant reasons when assessing what to do, but should replace some of them.
The first and the last theses generalize the features we noted in the arbitration example. The normal justification thesis replaces the agreement between the litigants which was the basis of the arbitrator’s authority . Agreement or consent to accept authority is binding, for the most part, only if conditions rather like those of the normal justification thesis obtain.
The first two theses articulate what I shall call the service conception of authority. They regard authorities as mediating between people and the right reasons which apply to them, so that the authority judges and pronounces what they ought to do according to right reason. The people on their part take their cue from the authority whose pronouncements replace for them the force of the dependent reasons. This last implication of the service conception is made explicit in the preemption thesis. The mediating role of authority cannot be carried out if its subjects do not guide their actions by its instructions instead of by the reasons on which they are supposed to depend. No blind obedience to authority’ is here implied. Acceptance of authority’ has to be justified, and this normally’ means meeting the conditions set in the justification thesis. This brings into play’ the dependent reasons, for only’ if the authority’s compliance with them is likely to be better than that of its subjects is its claim to legitimacy justified. At the level of general justification the preempted reasons have an important role to play. But once that level has been passed and we are concerned with particular action, dependent reasons are replaced by’ authoritative directives. To count both as independent reasons is to be guilty of double counting.
This is the insight which the surrender of judgment metaphor seeks to capture. It does not express the immense power of authorities. Rather it reflects their limited role. They are not there to introduce new and independent considerations (though when they make a mistake and issue the wrong decrees they do precisely that). They are meant to reflect dependent reasons in situations where they are better placed to do so. They mediate between ultimate reasons and the people to whom they apply.
II. AUTHORITY AND THE LAW
I will assume that necessarily law, every legal system which is in force anywhere, has de facto authority. That entails that the law either claims that it possesses legitimate authority or is held to possess it, or both. I shall argue that, though a legal system may not have legitimate authority, or though its legitimate authority may not be as extensive as it claims, every legal system claims that it possesses legitimate authority. If the claim to authority is part of the nature of law, then whatever else the law is it must be capable of possessing authority. A legal system may lack legitimate authority. If it lacks the moral attributes required to endow it with legitimate authority then it has none. But it must possess all the other features of authority, or else it would be odd to say that it claims authority. To claim authority it must be capable of having it, it must be a system of a kind which is capable in principle of possessing the requisite moral properties of authority. These considerations, I shall argue, create a weighty argument in favour of the sources thesis. Let us review them step by step.
The claims the law makes for itself are evident from the language it adopts and from the opinions expressed by its spokesmen, i.e. by the institutions of the law. The law’s claim to authority is manifested by the fact that legal institutions are officially designated as ‘authorities’, by the fact that they regard themselves as having the right to impose obligations on their subjects, by their claims that their subjects owe them allegiance, and that their subjects ought to obey the law as it requires to be obeyed (i.e. in all cases except those in which some legal doctrine justifies breach of duty).
Even a bad law, is the inevitable official doctrine, should be obeyed for as long as it is in force, while lawful action is taken to try and bring about its amendment or repeal. One caveat needs be entered here. In various legal systems certain modes of conare technically unlawful without bso in substance. It is left to the prosecutorial authorities to refrain from prosecuting for such conduct, or to the courts to give absolute discharge. Where legally recognized policies direct such authorities to avoid prosecution or conviction, the conduct should not be regarded as unlawful except in a technical sense, which is immaterial to our considerations.
Does the fact that the law claims authority help us understand its nature in any way, beyond the sheer fact that the law’ makes this claim? If of necessity all legal systems have legitimate authority, then we can conclude that they have the features which constitute the service conception of authority. But it is all too plain that in many cases the law’s claim to legitimate authority cannot be supported. There are legal systems whose authority cannot be justified by the normal justification thesis or in any other way. Can it not be argued that, since the law may lack authority, a conception of authority cannot contribute to our understanding of what it is, except by showing what it claims to be? This conclusion is at the very least premature. It could be that, in order to be able to claim authority, the law must at the very least come close to the target, i.e. that it must have some of the characteristics of authority. It can fail to have authority. But it can fail in certain ways only. If this is so, there are features of authority that it must have. If so, we can learn from the doctrine of authority something about the nature of law.
Note that nothing in this suggestion assumes that all the necessary features of the law are necessary features of every practical authority. The law may well have others. Indeed, I am already assuming that the law does have others, since it is not necessary that every person who has legitimate authority’ claims to have it, as the law necessarily does. All that we are trying to establish is whether some necessary’ characteristics of law are necessary characteristics of authority, which the law must have if it is to be capable of claiming authority.
I suggested above that only those who can have authority can sincerely claim to have it, and that therefore the I a w’ must be capable of having authority’. This claim is so vague that, even if correct, it cannot be more than a gesture towards an argument. What might that be? Consider the fact that the law is a normative system. If it were not, it would be incapable of having practical authority. If the law were a set of propositions about the behaviour of volcanoes, for example, then it would not only lack authority over action, it would be incapable of having such authority. The statement that a normative system is authoritatively binding on us may be false, but at least it makes sense, whereas the claim that a set of propositions about volcanoes authoritatively determines what we ought to do does not even make sense.
But cannot one claim that a person X has authority which it would make no sense to attribute to X? The claim makes sense because we understand what is claimed, even while we know that it is not merely false but is necessarily, or conceptually, false. For example, what cannot communicate with people cannot have authority over them. Trees cannot have authority over people. But someone whose awareness of what trees are is incomplete,
a young child, for example, can claim that they do have authority. He is simply wrong. Similarly, even if he is aware of the nature of trees, he may make an insincere claim to that effect. Perhaps he is trying to deceive a newly arrived Martian sociologist. Notice, however, that one cannot sincerely claim that someone who is conceptually incapable of having authority has authority if one understands the nature of one’s claim and of the person of whom it is made. If I say that trees have authority over people , you will know that either my grasp of the concepts of authority or of trees is deficient or that I am trying to deceive (or, of course, that I am not really stating that trees have authority but merely pretending to do so, or that I am play-acting, etc.).
That is enough to show that since the law claims to have authority it is capable of having it. Since the claim is made by legal officials wherever a legal system is in force, the possibility that it is normally insincere or based on a conceptual mistake is ruled out. It may, of course, be sometimes insincere or based on conceptual mistakes. But at the very least in the normal case the fact that the law claims authority for itself shows that it is capable of having authority.
Why cannot legal officials and institutions be conceptually confused? One answer is that while they can be occasionally they cannot be systematically confused. For given the centrality of legal institutions in our structures of authority, their claims and conceptions are formed by and contribute to our concept of authority. It is what it is in part as a result of the claims and conceptions of legal institutions. This answer applies where the legal institutions themselves employ’ the concept of authority. But there may’ be law in societies which do not have our concept of authority. We say of their legal institutions that they’ claim authority because they claim to impose duties, confer rights, etc. Not having the concept they cannot be confused about it, though we can be confused in attributing the claim of authority to them.
The argument of the last four paragraphs has established, first, that one can fail to have authority because one is incapable of possessing authority (though even those capable of having authority may fail to have it), second, that since the law claims authority it is capable of having authority. There are two kinds of reason for not having authority. One is that the moral or normative conditions for one’s directives being authoritative are absent.
Typically, this will be either because the normal justification, explained above, is unavailable or because, though available, it is insufficient to outweigh the conflicting reasons which obtain in this particular case. The second kind of reason for not having authority is that one lacks some of the other, non-moral or non-normative, prerequisites of authority, for example, that one cannot communicate with others.
It is natural to hold that the non-moral, non-normative conditions for having authority are also the conditions of the ability to have authority. A person’s authority may be denied on the ground that he is morally incompetent or wicked. But such facts do not show that he is incapable of having authority in the way that trees are incapable of having authority. Nazi rules may not be authoritatively binding, but they are the sort of thing that can be authoritatively binding, whereas statements about volcanoes cannot. Most arguments about the authority of governments and other institutions revolve around their moral claim to the obedience of their subjects. The existence of the non-moral qualifications is taken for granted. The argument does not start except regarding persons and institutions who meet those other conditions. That is why they are thought of as the conditions which establish capacity to possess authority.
If this view is correct then, since the law necessarily claims authority , and therefore typically has the capacity to be authoritative, it follows that it typically has all the non-moral, or non-normative, attributes of authority. The remainder of my argument, however, does not depend on this strong conclusion. We will concentrate on two features which must be possessed by anything capable of being authoritatively binding. These two features will then be used to support the sources thesis.
It is convenient to concentrate attention on instructions or directives. The terms are used in a wide sense which can cover propositions, norms, rules, standards, principles, doctrines, and the like. In that sense the law is a system of directives, and it is authoritative if and only if its directives are authoritatively binding. Likewise, whoever issues the directives has authority if and only his directives are authoritatively binding becaushe makes them, that is (1) they are authoritative, and (2) part of the reason is that he made them.
The two features are as follows. First, a directive can be authoritatively binding only if it is, or is at least presented as, someone’s view of how its subjects ought to behave. Second, it must be possible to identify the directive as being issued by the alleged authority without relying on reasons or considerations on which directive purports to adjudicate.
The first feature reflect.’ the mediating role of authority. It is there to act on reasons which apply to us anyway, because we will more closely conform to those reasons if we do our best to follow the directives of the authority than if we try to act on those reasons directly. Hence, though the alleged authoritative instruction may be wrongly conceived and misguided, it must represent the judgment of the alleged authority on the reasons which apply to its subjects, or at least it must be presented as the authority’s judgment. Otherwise it cannot be an authoritative instruction. If fails not because it is a bad instruction, but because it is not an instruction of the
right kind. It may be an instruction given for some other occasion, or in jest, or an order or threat of a gangster who cares for and considers only his own good. Strictly speaking, to be capable of being authoritative a directive or a rule has actually to express its author’s view on what its subjects should do. But given that this element is one where pretence and deceit are so easy, there is little surprise that appearances are all one can go by here, and the concept of de facto authority ‘ as well as all others which presuppose capacity to have authority, are based on them. If the rule is presented as expressing a judgment on what its subjects should do, it is capable of being authoritative.
The second feature too is closely tied to the mediating role of authority. Suppose that an arbitrator, asked to decide what is fair in a situation, has given a correct decision. That is, suppose there is only one fair outcome, and it was picked out by the arbitrator. Suppose that the parties to the dispute are told only that about his decision, i.e. that he gave the only correct decision. They will feel that they know little more of what the decision is than they did before. They were given a uniquely identifying description of the decision and yet it is an entirely unhelpful description. If they could agree on what was fair they would not have needed the arbitrator in the first place. A decision is serviceable only if it can be identified by means other than the considerations the weight and outcome of which it was meant to settle.
This applies to all decisions, as much to those that a person takes for himself as to those taken for him by others. If I decide what would be the best life insurance to buy, it is no good trying to remind me of my decision by saying that I decided to buy the policy which it is best to buy. It means that I have to decide again in order to know what I decided before, so the earlier decision might just as well never have happened. The same applies to the subjects of any authority. They can benefit by its decisions only if they can establish their existence and content in ways which do not depend on raising the very’ same issues which the authority is there to settle.
Can it not be objected that my argument presupposes that people know the normal justification thesis, and the others which go with it? To be sure such an assumption would not be justified. Nor is it made, All I am assuming is that the service conception of authority is sound, i.e. that it correctly represents our concept of authority. It is not assumed that people believe that it does.
It is worth noting that a set of conditions rather like the pair I have argued for can be derived from a much weaker assumption than that of the service conception of authority’ explained above. I will call this the alternative argument. Its premise is nothing more than the claim that it is part of
our notion of legitimate authority that authorities should act for reasons, and that their legitimacy depends on a degree of success in doing so. Even those who reject the service conception of authority will accept conditions similar to the two I have argued for if they accept that legitimacy depends on (a degree of) success in acting for reasons. It is obvious that this weak assumption is enough to hold that only what is presented as someone’s view can be an authoritative directive.
Instead of the second condition, that directives be capable of independent identification (i.e. independent of the reasons they should be based upon), two weaker conditions can be established. I will assume that authorities make a difference, i.e. the fact that an authority issued a directive changes the subjects’ reasons. It follows that the existence of reasons for an authority to issue a directive does not by itself, without the directive having actually been issued, lead to this change in the reasons which face the subjects. Therefore, the existence of reasons which establish that a certain directive, if issued, would be the right one to have issued cannot show that such a directive exists and is binding. Its existence and content, in other words, cannot depend exclusively on the reasons for it. The existence and content of every directive depend on the existence of some condition which is itself independent of the reasons for that directive. Moreover, that further condition cannot simply be that that or some other authority issued another directive. Often the existence of one law is a reason for passing another.
But we have just established that the existence of a law cannot depend simply on the existence of reasons for it, on reasons showing that it would be good if people behaved in the way it prescribes, or that it would be good if the law’ required them to do so. Therefore, the existence of one directive, though it may show’ that another is desirable or right, cannot by itself establish its existence.
III. THE COHERENCE THESIS
The previous section argued that, even though the law may lack legitimate authority, one can learn quite a lot about it from the fact that it claims legitimate authority. It must be capable of being authoritative. In particular it must be, or be presented as, someone’s view on what the subjects ought to do, and it must be identifiable by means which are independent of the considerations the authority should decide upon.
It is interesting to note that legal sources meet both conditions. To anticipate and simplify, the three common sources of law, legislation, judicial decisions, and custom, are capable of being sources of authoritative directives. They meet the non-moral conditions implied in the service conception of authority. Legislation can be arbitrary, and it can fail to comply with the
dependence thesis in many ways. But it expresses, or is at least presented as expressing, the legislator’s judgment of what the subjects are to do in the situations to which the legislation applies. Therefore, it can be the product of the legislator’s judgment on the reasons which apply to his subjects. The same is true of judicial decisions. Judges may be bribed. They may act arbitrarily. But a judicial decision expresses a judgment on the legal consequences of the behaviour of the litigants. It is presented as a judgment on the way the parties, and others in the same circumstances, ought to behave.
Similarly with custom. It is not normally generated by people intending to make law. But it can hardly avoid reflecting the judgment of the bulk of the population on how people in the relevant circumstances should act. Source-based law can conform to the dependence thesis. It therefore conforms to the first of our conditions which are entailed by the fact that the law claims authority.
Legal sources also conform with the second of our two conditions, since they are capable of being identified in ways which do not rely on the considerations they are meant to decide upon. An income-tax stis meant to decide what is the fair contribution of public funds to borne out of income. To establish the content of the statute, all one need do is to establish that the enactment took place, and what it says. To do this one needs little more than knowledge of English (including technical legal English), and of the events which took place in Parliament on a few occasions. One need not come to any view on the fair contribution to public funds.
As was noted above, all three rivals, the coherence, the incorporation, and the sources theses, are united in attributing a special significance to source-based law. The preceding simplified account illustrates the way central features of the law’ can mesh in with and acquire a special significance from the service conception of authority and the two necessary features of law which it entails. It does not follow that these are the reasons normally given for the centrality of source-based law. The coherence thesis represents an account which is at the very least indifferent to the considerations outlined above, I have identified it as the view that the law consists of source-based law together with the morally best justification of the source-based law. This may look an unholy mixture of disparate elements. But it need not be. In the hands of its best advocate, R. M. Dworkin, it embodies a powerful and intriguing conception of the law.
Dworkin’s conception of the law, expressed in various articles over many years, is not easy to ascertain. Some points of detail which are nevertheless essential to its interpretation remain elusive. Many readers of his celebrated ‘Hard Cases’ (1975) took it to express a view of law which can be summarized in the following way:
“To establish the content of the law of a certain country one first finds out what are the legal sources valid in that country and then one considers one master question: Assuming that all the laws ever made by these sources which are still in force, were made by one person, on one occasion, in conformity with a complete and consistent political morality (i.e. that part of a moral theory which deals with the actions of political institutions), what is that morality?”
The answer to the master question and all that it entails, in combination with other true premises is, according to this reading of Dworkin, the law. The master question may fail to produce an answer for two opposite reasons, and Dworkin complicates his account to deal with both. First, there may be conflicts within a legal system which stop it from conforming with any consistent political morality. To meet this point Dworkin allows the answer to be a political morality with which all but a small number of conflicting laws conform. Second, there may be more than one political morality meeting the condition of the master question (especially once the allowance made by the first complication is taken into account). In that case Dworkin instructs that the law is that political morality which is, morally, the better theory. That is, the one which approximates more closely to ideal, correct or true morality.
In his ‘Reply to Seven Critics’ (1977) Dworkin returns to the question of the nature of law. He gives what he calls too crude an answer, which can be encapsulated in a different master question:
“To establish the content of the law of a certain country one first finds out what are the legal sources valid in that country and then one considers one master question: What is the least change one has to allow in the correct, sound political morality in order to generate a possibly less than perfect moral theory’ which explains much of the legal history’ of that country’ on the assumption that it is the product of one political morality?”
That (possibly less than perfect) political morality is the law. Both master questions depend on an interaction of two dimensions. One is conformity with ideal morality, the other ability to explain the legal history of the country. The new master question differs from its predecessor in two important respects. First, its fit condition concerns all the legal history of the country. Acts of Parliament enacted in the thirteenth century and repealed fifty years later are also in the picture. They also count when measuring the degree to which a political morality fits the facts. The earlier test refered only to law still in force. Only fitting in with them counted. Second, the new master question gives less weight to the condition of fit. It is no longer the case that the law consists of the political morality which fits the facts best, with ideal morality coming in just as a tie-breaker. Fit (a certain unspecified level of it) now provides only a sort of flexible threshold test. Among the (presumably
numerous) political moralities which pass it, the one which is closest to correct morality is the law.
I hesitate to attribute either view to Dworkin. The articles are not clear enough on some of the pertinent points, and his thought may have developed in a somewhat new direction since these articles were written. Luckily, the precise formulation of the master question does not matter to our purpose.
Enough of Dworkin’s thought is clear to show that its moving ideas are two. First, that judges’ decisions, all their decisions, are based on considerations of political morality. This is readily admitted regarding cases in which source based laws are indeterminate or where they conflict. Dworkin insists that the same is true of ordinary cases involving, say, simple statutory interpretation or indeed the decision to apply statute at all. This does not mean that every time judges apply statutes they consider and re-endorse their faith in representative democracy, or in some other doctrine of political morality from which it follows that they ought to apply these statutes. It merely means that they present themselves as believing that there is such a doctrine. Their decisions are moral decisions in expressing a moral position. A conscientious judge actually believes in the existence of a valid doctrine, a political morality, which supports his action.
If I interpret Dworkin’s first leading idea correctly and it is as stated above, then I fully share it. I am not so confident about his second leading idea. It is that judges owe a duty, which he sometimes calls a duty of professional responsibility, which requires them to respect and extend the political morality of their country. Roughly speaking, Dworkin thinks that morality (i.e. correct or ideal morality) requires judges to apply the source based legal rules of their country, and, where these conflict or are indeterminate, to decide cases by those standards of political morality which inform the source-based law, those which make sense if it is an expression of a coherent moral outlook.
Notice how far-reaching this second idea is. Many believe that the law of their country, though not perfect, ought to be respected. It provides reasonable constitutional means for its own development. Where reform is called for, it should be accomplished by legal means. While the law is in force it should be respected. For most, this belief depends to a large degree on the content of the law. They will deny that the laws of Nazi Germany deserved to be respected. Dworkin’s obligation of professional responsibility is different. It applies to every legal system simply because it is a legal system, regardless of its content. Furthermore, it is an obligation to obey not merely the letter of the law but its spirit as well. Judges are called upon to decide cases where source-based law is indeterminate, or includes unresolved conflicts, in accordance with the prevailing spirit behind the bulk
of the law. That would require a South African judge to use his power to extend apartheid.
Problems such as these led to the weakening of the element of fit in the second formulation of the master question. But then they also weaken the duty of professional responsibility. There is an attractive simplicity in holding that morality requires any person who joins an institution to respect both its letter and its spirit. If this simple doctrine does not apply to judges in this form, if their respect for theinstitution, the law, is weakened from its pure form in the first master question to that of the second, then one loses the theoretical motivation for such a duty , at least if it means more than saying that one ought to respect the legal institutions of a particular country because their structure and actions merit such respect, or to the extent that they do.
These are some of the doubts that Dworkin’s second leading idea raises. My formulations of the two leading ideas (and of the doubts concerning the second) are mere sketches. They are meant to outline an approach to law which gives source-based law a special role in the account of law on grounds other than those explained in the previous section. It is easy to see that Dworkin’s conception of law contradicts the two necessary features of law argued for above. First, according to him there can be laws which do not express anyone’s judgment on what their subjects ought to do, nor are they presented as expressing such a judgment. The law includes the best justification of source-based law, to use again the brief description given in the coherence thesis of which Dworkin’s master questions are different interpretations. The best justification, or some aspects of it, may never have been thought of, let alone endorsed by anyone. Dworkin draws our attention to this fact by saying that it requires a Hercules to work out what the law is. Nor does Dworkin’s best justification of the law consist of the implied consequences of the political morality which actually motivated the activities of legal institutions. He is aware of the fact that many different and incompatible moral conceptions influenced different governments and their officials over the centuries. His best justification may well be one which was never endorsed, not even in its fundamental precepts, by anyone in government. Much of the law of any country may, according to Dworkin, be unknown. Yet it is already legally binding, waiting there to be discovered. Hence it neither is nor is presented as being anyone’s judgment on what the law’s subjects ought to do.
Second, the identification of much of the law depends, according to Dworkin’s analysis, on considerations which are the very same considerations which the law is there to settle. This aspect of his theory is enhanced by his second master question, but it makes a modest appearance in the first as
well. Establishing what the law is involves judgment on what it ought to be. Imagine a tax problem on which source-based law is indeterminate. Some people say that in such a case there is no law’ on the issue. The court ought to ask what the law ought to be and to decide accordingly. If it is a higher court whose decision is a binding precedent, it will have thereby made a new law. Dworkin, on the other hand, says that there is already law on the matter. It consists in the best justification of the source based law. So in order to decide what the tax liability is in law, the court has to go into the issue of what a fair tax law would be and what is the least change in it which will make source-based law conform to it. This violates the second feature of the I a w’ argued for above.
It is important to realize that the disagreement I am pursuing is not about how judges should decide cases. In commenting on Dworkin’s second leading idea I expressed doubts regarding his view on that. But they are entirely irrelevant here. So let me assume that Dworkin’s duty of professional responsibility is valid and his advice to judges on how to decide cases is sound. We still have a disagreement regarding what judges do when they follow his advice. We assume that they follow right morality, but do they also follow the law or do they make law? My disagreement with Dworkin here is that, in saying that they follow pre-existing law, he makes the identification of a tax law, for example, depend on settling what a morally just tax law would be, i.e. on the very considerations which a tax law is supposed to have authoritatively settled.
For similar reasons Dworkin’s theory violates the conditions of the alternative argument, the argument based on nothing more than the very weak assumption that authorities ought to act for reasons and that the validity of authoritative directives depends on some degree of success in doing so. This assumption leads to the same first condition, i.e. that the law must be presented as the law-maker’s view on right reasons. As we have just seen, Dworkin’s argument violates this condition. He also violates the other condition established by the alternative argument, that the validity of a law cannot derive entirely from its desirability in light of the existence of other laws. Dworkin’s theory claims that at least some of the rules which are desirable or right in view of the existence of source-based law are already legally binding.
Dworkin’s theory, one must conclude, is inconsistent with the authoritative nature of law. That is, it does not allow for the fact that the law necessarily claims authority and that it therefore must be capable of possessing legitimate authority. To do so it must occupy, as all authority’ does, a mediating role between the precepts of morality and their application by people in their behaviour. It is this mediating role of authority which is denied to the law by Dworkin’s conception of it.
IV. THE INCORPORATION THESIS
The problem we detected with the coherence thesis was that, though it assigns source-based law a special role in its account of law, it fails to see the special connection between source-based law and the law’s claim to authority, and is ultimately inconsistent with the latter. It severs the essential link between law and the views on right action presented to their subjects by those who claim the right to rule them. In these respects, the incorporation thesis seems to have the advantage. It regards as law source-based law and those standards recognized as binding by source-based law. The approval of those who claim a right to rule is a prerequisite for a rule being a rule of law. Thus the law’s claim to authority appears to be consistent with the incorporation thesis.
I should hasten to add that many of the supporters of the incorporation thesis do not resort to the above argument in its defence. Nor do they interpret the centrality of source-based law to their conception of law in that way. They regard it as supported by and necessary for some version of a thesis about the separability of law and morals. Jules Coleman, for example, is anxious to deny that there is a necessary connection between law and morality. He mistakenly identifies this thesis with another: “The separability thesis is the claim that there exists at least one conceivable rule of recognition and therefore one possible legal system that does not specify truth as a moral principle as a truth condition for any proposition of law”. If this were a correct rendering of the separability thesis stated by Coleman in the first quotation above, the incorporation thesis entails separability. But Coleman’s rendering of his own separability thesis is mistaken. A necessary connection between law and morality does not require that truth as a moral principle be a condition of legal validity. All it requires is that the social features which identify something as a legal system entail that it possess moral value. For example, assume that the maintenance of orderly social relations is itself morally’ valuable. Assume further that a legal system can be the law in force in a society’ only if it succeeds in maintaining orderly social relations. A necessary connection between law and morality would then have been established, without the legal validity of any rule being made, by’ the rule of recognition, to depend on the truth of any’ moral proposition.
Supporters of the incorporation thesis may’ admit that, while it is not sufficient to establish the separability thesis, at least it is necessary’ for it, and is therefore supported by it. The separability thesis is, however, imp. Of course the remarks about orderly social relations do not disprove it. They are much too vague and to do that. But it is very likely that there is some necessary connection between law and morality,
that every legal system in force has some moral merit or does some moral good even if it is also the cause of a great deal of moral evil. It is relevant to remember that all major traditions in western political thought, including both the Aristotelian and the Hobbesian traditions, believed in such a connection.12 If the incorporation thesis seems much more secure than the separability thesis, it is because if seems to be required by the fact that all law comes under the guise of authority, together with the considerations on the nature of authority advanced in the previous sections. The law is the product of human activity because if it were not it could not be an outcome of a judgment based on dependent reasons, that is, it could not provide reasons set by authority.
There may, of course, be other cogent reasons for favouring the incorporation thesis. They will not be explored here. Instead I will argue that the thesis ought to be rejected, and that the support it seems to derive from the argument about the nature of authority is illusory. In fact the incorporation thesis is incompatible with the authoritative nature of law. To explain the point let us turn for a moment to look at theoretical authority.
Suppose that a brilliant mathematician, Andrew, proves that the Goldbach hypothesis, that every integer greater than two is the sum of two prime numbers, is true if and only if the solution to a certain equation is positive. Neither he nor anyone else knows the solution of the equation. Fifty years later that equation is solved by another mathematician and the truth of the Goldbach hypothesis is established. Clearly we would not say that Andrew proved the hypothesis, even though he made the first major breakthrough and even though the truth of the hypothesis is a logical consequence of his discovery. Or suppose that Betty is an astrophysicist who demonstrates that the big bang theory’ of the origin of the universe is true if and only if certain equations have a certain resolution. Again, their resolution is not known at the time, and is discovered only’ later. It seems as clear of Betty’ as it was of Andrew that she cannot be credited with proving (or disproving) the big bang theory even though the truth (or falsity ) of the theory’ is entailed by her discovery. Now imagine that Alice tells you of Andrew’s discovery, or that Bernard tells you of Betty’s. Alice and Bernard are experts in their respective fields. They give you authoritative advice. But Alice does not advise you to accept the Goldbach hypothesis. She merely advises belief in it if the relevant equation has a positive solution. The fact that the truth of the hypothesis is entailed by her advice is neither here nor there. The same applies to Bernard’s advice based on Betty’s work.
All this is commonplace. Nor is it difficult to understand why one cannot be said to have advised acceptance of a particular proposition simply on the ground that it is entailed by another proposition acceptance of which
one did advise. People do not believe in all that is entailed by their beliefs. Beliefs play a certain role in our lives in supporting other beliefs, in providing premises for our practical deliberations. They colour our emotional and imaginative life. More generally, they are fixed points determining our sense of orientation in the world. Many of the propositions entailed by our beliefs do not play this role in our lives. Therefore they do not count amongst our beliefs. One mark of this is the fact that had people been aware of some of the consequences of their beliefs, rather than embrace them they might have preferred to abandon the beliefs which entail them (or even provisionally to stick by them and refuse their consequences, i.e. embrace inconsistencies until they found a satisfactory way out), This consideration explains why we cannot attribute to people belief in all the logical consequences of their beliefs. It also explains why a person cannot be said to have advised belief in a proposition he does not himself believe in. (Though it is possible to advise others to take the risk and act as if certain propositions are true even if one does not believe in them and equally’ possible to advise believing in a proposition if it is true.)
Advice shares the mediating role of authoritative directives. It too is an expression of a judgment on the reasons which apply to the addressee of the advice. Because the advice has this mediating role it can include only’ matters on which the adviser has a view, or presents himself as having one (to cover cases of insincere advice). Since a person does not believe in all the consequences of his beliefs he does not, barring special circumstances, advise others to believe in them either.
The analogy with authority is clear and hardly needs further elaboration. The mediating role of authority implies that the content of an authoritative directive is confined to what the authority which lends the directive its binding force can be said to have held or to have presented itself as holding. It does not extend to what it would have directed, given a chance to do so, nor to all that is entailed by what it has directed. It will by now be clear why’ the incorporation thesis must be rejected if the law does necessarily claim authority. The main thrust of the incorporation thesis is that all that is derivable from the law (with the help of other true premises) is law. It makes the law include standards which are inconsistent with its mediating role, for they were never endorsed by the law-making institutions on whose authority they are supposed to rest. The mistake of the incorporation thesis is to identify being entailed by the source-based law with being endorsed by the sources of law.
Law is a complex social institution, and some of its complexities help mask the incorporation thesis’s mistake. When thinking of a piece of advice or of an authoritative directive we tend to think of them as having one
author. In the law, as in other hierarchical institutions, matters are complicated in two respects. First, authoritative directives are typically issued by institutions following an elaborate process of drafting and evaluation. Second, they are often amended, modified, and their content amplified and changed by a succession of subsequent legislative, administrative, and judicial actions. A convention of reference sometimes exists which allows one to refer to a statute, or to the original judicial decision, when citing a legal rule, even though they are no more than the starting-point in the development of the rule, which is in a very real sense the product of the activities of several bodies over a period of time.
These complications mean, of course, that the rule as it is now may include aspects which cannot be attributed to its original creator. They are part of the rule because they are attributable to the author of a later intervention. For example, typically successive judicial interpretations change or add to the meaning of statutes. Likewise, though we attribute beliefs and intentions to institutions and corporations on the basis of the beliefs and intentions of their officials, the attributing functions may sometimes sanction holding a corporate body’ to have had a belief or an intention which none of its officials had. This is not the place to inquire into the rules of attributions invoked when we talk of the intentions or beliefs of states, governments, corporations, trade unions, universities, etc. All that is required for our present purposes is that attribution is made in a restrictive way which does not allow one to attribute to such a body’ all the logical consequences of its beliefs and intentions. Restrictions to all the foreseen or foreseeable consequences are the ones most common in the law. This is enough to show’ that the incorporation thesis receives no sustenance from the institutional complexitiof the law, since it insists that the law includes all the logical consequences of source-based law.
In disputing the incorthesis I am not denying two other points which are asserted by D. Lyons in the most thorough going defence of this position. First, I agree with him that judges who work out what is required by, for example, the due process provision of the American constitution are engaged in interpreting the constitution. Lyons is mistaken, however, in thinking that it follows from that that they are merely applying the law as it is (at least if they succeed in discovering the right answer). Judicial interpretation can be as creative as a Glen Gould interpretation of a Beethoven piano sonata. It is a mistake to confuse interpretation with paraphrase or with any other mere rendering of what the interpreted object is in any case. Second, Lyons is quite right to think that there is more to the law than is explicitly stated in the authoritative texts. Authorities can and do direct and guide by implication. It does not follow, however, that they imply all that is entailed
by what they say, let alone all that is entailed by it with the addition of true premises. The limits of the justifiable imputation of directives are no wider, I have argued above, than the limits of the imputation of belief.
V. THE SOURCES THESIS
The last section established that not all the moral consequences of a legal rule are part of the law. But it leaves open the possibility that some are: that some moral consequences of a legal rule can be attributed to the author of that legal rule as representing its intention or meaning and thus being part of the law. I will not present a refutation of this possibility. The purpose of the present section is more modest. It argues that the authoritative nature of law gives a reason to prefer the sources thesis. It leaves open the possibility that additional considerations lead to a complex view of the law lying between the incorporation and the sources theses.
Let us distinguish between what source-based law states explicitly and what it establishes by implication. If a statute in country A says that income earned abroad by a citizen of A is liable to income tax in A, then it only implicitly establishes that I am liable to such tax. For my liability is not stated by the statute but is inferred from it (and some other premisses).
Similarly, if earnings abroad are taxed at a different rate from earnings at home, the fact that the proceeds of export sales are subject to the home rate is implied rather than stated. It is inferred from this statute and other legal rules on the location of various transactions. By the same reasoning it also established that not all the factual consequences of a rule of law are part of the law.
The two examples differ in that the statement that I am liable to tax at a certain rate is an applied legal statement depending for its truth on both law and fact. The statement that export earnings are taxed at a certain rate is a pure legal statement, depending for its truth on law only (i.e. on acts of legislation and other law-making facts). The Sources thesis as stated at the beginning of this chapter can bear a narrow or a wide interpretation. The narrow thesis concerns the truth conditions of pure legal statements only’.
Pure legal statements are those which state the content of the law, i.e. of legal rules, principles, doctrines, etc. The wide thesis concerns the truth conditions of all legal statements, including applied ones. It claims that the truth or falsity of legal statements depends on social facts which can be established without resort to moral argument.
The fact that the law claims authority supports the narrow sources thesis because it leads to a conception of law as playing a mediating role between ultimate reasons and people’s decisions and actions. To play this role the law must be, or at least be presented as being, an expression of the judgment of some people or of some institutions on the merits of the actions it requires. Hence, the identification of a rule as a rule of law consists in attributing
it to the relevant person or institution as representing their decisions and expressing their judgments. Such attribution need not be on the ground that this is what the person or institution explicitly said. It may be based on an implication. But the attribution must establish that the view expressed in the alleged statement is the view of the relevant legal institution. Such attributions can only be based on factual considerations. Moral argument can establish what legal institutions should have said or should have held but not what they did say or hold.
We have already traced one source of resistance to this conclusion to the assumption that if attribution is on factual rather than moral grounds then it must be a non-controversial, easily established matter which requires at most the application of a procedure of reasoning having the character of an algorithm to some non-controversial simple facts. The assumption that only moral questions can resist easy agreement or solution by algorithmic procedures has nothing to recommend it, and I in no way share it. The case for saying that attribution of belief and intention to their author is based on factual criteria only does not rest on the false claim that such attributions are straightforward and non-controversial. A second source of resistance, also noted above, derives from overlooking the greater complexity involved in attributing views or intentions to complex institutions whose activities spread over long stretches of time, and the tendency to think that nothing more is involved in these cases than is involved in attributing beliefs or intentions to individuals.
But there is a third difficulty with the view I am advocating which must be addressed now. One may ask: if an authority explicitly prohibited e.g. unfair discrimination, is not the fact that certain cases display unfair discrimination evidence enough for attributing their prohibition to the authority? Two considerations are usually brought to support the view that these reasons are sufficient to determine the content of the law on such matters. I shall try to rebut this view by showing that these supporting considerations are mistaken. First is the claim that the only alternative view holds that the law is determined only regarding cases which the law-maker actually contemplated and had in mind when making the law. This, let it be conceded right away, is not merely false but very likely an incoherent view. Second (and it does not matter that this point may be incompatible with the first), it is sometimes said that the only alternative view assumes that the law’-makers intend their particular view of what is unfair discrimination to become law even if they are wrong.
Suppose that the fathers of the constitution outlawed cruel punishment. Suppose further that it is beyond doubt that they thought that flogging is not cruel, and finally, that in fact (or in morals) it is cruel. Are we to
assume that the law-maker’s intention was to exclude flogging from the scope of the constitutional prohibition of cruel punishments? Would not the correct view be that in making cruelty a bench-mark of legality the law-makers intended their own judgment to be subject to that criterion, so that, though believing flogging not to be cruel, they expressed the view that if it is cruel it is unlawful?
Both points have a deja vu aspect. They depend on the unimaginative assumption that either the law is determined by the thoughts actually entertained by the law-maker when making the law or it must include all the implications of those thoughts. Since it must be granted, and I do grant, that it is not the first, the second is supposed to be the case. This was the structure of Lyons’s argument regarding the explicit content thesis. As he saw it, either the law is confined to its explicit content or it contains all its implications. Since Hart rejects the second alternative he was saddled by Lyons with the first. Since Lyons sees, as everyone must, that the first is wr, he embraces the second. The two considerations explained above are the psychological variants of Lyons’s linguistic dichotomy. They contrastnot actual language with its implications but actual thoughts with their implications.
The answer to both arguments is the same: the dichotomy is a false one. There are other possibilities. Sometimes we know of a person that, for example, if only’ he realized that certain forms of psychological abuse are cruel, would not be so indifferent to them. At others we know that if he were convinced that they’ are cruel he would find some other way to justify them. He would come to believe that cruelty is sometimes justifiable. In attributing such views to people, one does not endorse either of the two unacceptable views mentioned above. Naturally it is often impossible to impute any such view to a person. The question whether he would have maintained his intention to prohibit cruel punishment had he known that capital punishment is cruel (assuming for a moment that it is) may admit of no answer.
Furthermore, and this is often overlooked, the sources thesis by itself does not dictate anyone rule of interpretation. It is compatible with several. It is compatible, for example, with saying that, if it is known that the lawmakers prohibited cruel punishment only because they regarded flogging as not cruel, then that law does not prohibit flogging. It is also compatible with the rule that the law is confined in such cases to the intention expressed by the law-maker. This is to prohibit cruel punishment. Since, by this rule of interpretation, no more specific intention is attributable to the law-maker, the law gives discretion to the courts to forbid punishments they consider cruel (this reflects the lack of specificity in the law) and instructs them to forbid those which are cruel. Which of these, or of a number of alternative interpretations,
is the right one varies from one legal system to another. It is a matter of their own rules of interpretation. One possibility is that they have none on this issue, that the question is unsettled in some legal systems. The only point which is essential to the sources thesis is that the character of the rules of interpretation prevailing in any legal system, i.e. the character of the rules for imputing intentions and directives to the legal authorities is a matter of fact and not a moral issue. It is a matter of fact because it has to sustain conclusions of the kind: “That is in fact the view held by these institutions on the moral issues in question.”
Two further points have to be made to avoid misunderstanding. First, none of the above bears on what judges should do, how they should decide cases. The issue addressed is that of the nature and limits of law. If the argument here advanced is sound, it follows that the function of courts to apply and enforce the law coexists with others. One is authoritatively to settle disputes, whether or not their solution is determined by law. Another additional function the courts have is to supervise the working of the law and revise it interstitially when the need arises. In some legal systems they are assigned additional roles which may be of great importance. For example, the courts may be made custodians of freedom of expression, a supervisory body in charge both of laying down standards for the protection of free expression and adjudicating in disputes arising out of their application.
Second, it may be objected that relying on the mediating role of authority’ becomes an empty’ phrase when it comes to legal rules which have evolved through the activities of many hands over a long time. The fact that we implicitly or explicitly endorse rules of attribution which sanction talk of the intention of the law where that intention was never had by anyone person does not support the argument from the mediating role of the law. It merely shows it to be a formalistic, hollow shell. This objection, like some of the earlier ones, seems to betray impatience with the complexities, and shortcomings, of the world. Every attribution of an intention to the law is based on an attribution of a real intention to a real person in authority or exerting influence over authority. That intention may well relate to a small aspect or modification of the rule. If the intention of the law regarding the rule as a whole differs from that of any single individual, this is because it is a function of the intentions of many. Sometimes, but by no means always, this leads to reprehensible results. Be that as it may, the view propounded here will in such circumstances highlight the indirect and complex way in which the law has played its mediating role.
All the arguments so far concern the narrow sources thesis only. Nothing was said about its application to applied legal statements. I tend to feel that it applies to them as well, since they are legal statements whose truth value depends on contingent facts as well as on law. If one assumes that
contingent facts cannot be moral facts, then the sources thesis applies here as well. That is, what is required is the assumption that what makes it contingently true that a person acted fairly on a particular occasion is not the standard of fairness, which is not contingent, but the ‘brute fact’ that he performed a certain action describable in value-neutral ways. If such an assumption is sustainable in all cases, then the sources thesis holds regarding applied legal statements as well.
The considerations adumbrated above dispel some of the misunderstandings which surround the sources thesis. First, it does not commit one to the view that all law is explicit law. Much that is not explicitly stated in legal sources is nevertheless legally binding. Second, the sources thesis does not rest on an assumption that law cannot be controversial. Nor does it entail that conclusion. Its claim that the existence and content of the law is a matter of social fact which can be established without resort to moral argument does not presuppose nor does it entail the false proposition that all factual matters are non-controversial, nor the equally false view that all moral propositions are controversial. The sources thesis is based on the mediating role of the law. It is true that the law fails in that role if it is not, in general, easier to establish and less controversial than the underlying considerations it reflects. But this generalization is exaggerated and distorted when it turns into the universal, conceptual dogmas of the explicit content or the non-controversiality theses.
The sources thesis leads to the conclusion that courts often exercise discretion and participate in the law-making process. They do so when their decisions are binding on future courts (even where the decisions can be modified or reversed under restrictive conditions) and where their decisions do not merely reflect previous authoritative rulings. Saying this does not mean, however, that courts in exercising their discretion either do or should act on the basis of their personal views on how the world should be ideally run. That would be sheer folly. Naturally judges act on their personal views, otherwise they would be insincere. (Though the fact that these are their views is not their reason for relying on them. Their reasons are that those propositions are true or sound, for whatever reason they find them to be so.) But judges are not allowed to forget that they are not dictators who can fashion the world to their own blueprint of the ideal society. They must bear in mind that their decisions will take effect in society as it is, and the moral and economic reasons they resort to should establish which is the best or the just decision given things as they are rather than as they would be in an ideal world.
Finally, the sources thesis does not presuppose a non-naturalist ethical position. Even if a certain social fact entails certain moral consequences it can still be a source of law. It is a source of law as the social fact it is, and not
as a source of moral rights and obligations. It is a source of law under its naturalistic rather tunder its moral description.
VI. THE ROLE OF VALUES IN LEGAL THEORY
According to R. M. Dworkin, legal positivists endorse the model of rules because of a poltheory about the function of law which they think is to ‘provide a settled public and dependable set of standards for private and official conduct, standards whose force cannot be called into question by some individual official’s conception of policy or morality. The argument of this article shows that something like Dworkin’s description applies to my argument. But notice that Dworkin’s remark suggests that legal positivists endorse the non-controversiality and the explicit contents theses, which I do not share. Besides, it is misleading to regard the thesis and argument explained here as moral ones. The argument is indeed evaluative, but in the sense that any good theory of society is based on evaluative considerations in that its success is in highlighting important social structures and processes, and every judgment of importance is evaluative.
Given the centrality of that feature, it is justified to interpret the action of law-makers who are in a hurry to get back home, who vote without paying attention to what they are voting for, in the way described. Two features stand out. First, while this is an evaluative judgment, it is not a judgment of the moral merit of any1hing. Second, its application depends on the fact that the perception of importance of the feature focused upon is shared in our society’, that it is shared, among others, by’ the law-makers themselves.
The concept of law is part of our culture and of our cultural traditions. It plays a role in the way in which ordinary people as well as the legal profession understand their own and other people’s actions. It is part of the way they ‘conceptualize’ social reality. But the culture and tradition of which the concept is a part provide it with neither sharply defined contours nor a clearly identifiable focus. Various, sometimes conflicting, ideas are displayed in them. It falls to legal theory to pick on those which are central and significant to the way the concept plays its role in people’s understanding of society, to elaborate and explain them.
Legal theory contributes in this respect to an improved understanding of society. But it would be wrong to conclude, as D. Lyons has done, that one judges the success of an analysis of the concept of law by its theoretical sociological fruitfulness. To do so is to miss the point that, unlike concepts like ‘mass’ or ‘election’, ‘the law’ is a concept used by people to understand themselves. We are not free to pick on any fruitful concepts. It is a major task of legal theory to advance our understanding of society by helping us understand how people understand themselves.
To do so it does engage in evaluative judgment, for such judgment is inescapable in trying to sort out what is central and significant in the common understanding of the concept of law. It was my claim in this chapter that one such feature is the law’s claim to authority and the mediating role it carries with it. The significance of this feature is both in its distinctive character as a method of social organization and in its distinctive moral aspect, which brings special considerations to bear on the determination of a correct moral attitude to authoritative institutions. This is a point missed both by’ those who regard the law as a gunman situation writ large and by those who, in pointing to the close connection between law and morality, assume a linkage inconsistent with it.
Let me exemplify the difference between my conception of the role of evaluation in explaining the nature of law and that of Dworkin by considering one central objection to the sources thesis. Some people object not to the attribution of intention to legislators or interpreters of the law in itself, but to the presupposition of the sources thesis that whenever one is faced with valid legislation one can also find an intention behind it. Is it always the case? Do we not know that sometimes members of Parliament vote knowing nothing and intending only to get home as early as possible? An adequate answer to this and related questions has to await a comprehensive treatment of interpretation and the role of intention within its context. A brief indication of the direction in which an answer is to be sought will have to do.
Let us start by considering the view which denies the important of the law-maker’s intention to our understanding of’ the law. To the question: ‘why should one assign any importance to a particular text as legally binding?’ that view will reply: ‘because is was endorsed by the proper constitutional procedure.’ To the question ‘how’ should the text be interpreted other than by’ reference to the intentions of its author or of those whose action maintains its force as law’? the answer would refer to existing conventions of interpretation which need not refer to anyone’s intention. There is nothing wrong with these replies. They merely raise further questions. Why does the endorsement of a certain text in accord with those procedures endow it with a special status? Is it some form of magic or fetishism?
That procedure is a way of endowing a text with legal force because it is a procedure designed to allow those in authority to express a view on how people should behave, in a way which will make it binding. That it is such a procedure, and not just any arbitrarily’ chosen ritual, is part of what makes it into a legal procedure. The law-making procedure includes conventions of interpretation. A change in the conventions of interpretation of a legal system changes its law. Consider the simple example: of a change from an understanding of “person” to include only people to a reading of it which covers foetuses as well. Law-makers need not intend anything other than that the bill should become law with the meaning given it by the conventions of interpretation of their country. To deny them that intention is to deny that they know what they are doing when they make law.
How is this sketchy reply to the objection to he defended? It turns on evaluative conceptions about what is significant and important about central social institutions, i.e., legal institutions. But in claiming that these features are important one is not commending them as good. Their importance can be agreed upon by anarchists who reject any possibility of legitimacy for such institutions. All that is claimed is the centrality to our social experience of institutions which express what they claim to be the collective and binding judgement of their society as to how people should behave.