THEORIZING AREAS OF LAW: A TAXONOMY OF SPECIAL JURISPRUDENCE

Abstract

This paper provides a taxonomy of the different kinds of theory that may be offered of an area of law. We distinguish two basic types of philosophical accounts in special jurisprudence: nonnormative accounts and normative accounts. Section II explains the two central subspecies of nonnormative accounts of areas of law: (i) conceptual and ontological theories and (ii) reason-tracking causal theories. Section III explores normative theories of areas of law. Normative accounts subdivide into detached and committed normative accounts. Detached or committed normative accounts can be subdivided further into the following cross-cutting categories: (i) pro tanto or all-things-considered, (ii) hyper-reformist or practice-dependent, (iii) taxonomical or substantive. Section IV shows that our taxonomy does not presume a prior commitment to any particular school in general jurisprudence. This paper clarifies methodological confusion that exists in theorizing about areas of law, and contributes to the subfield of thinking generally about special jurisprudence.

I. INTRODUCTION

Legal theory has seen a surge in scholarly interest in theorizing discrete “areas of law” (variously described as “special jurisprudence”¹ or “particular jurisprudence,”² to contrast the field with general jurisprudence). General jurisprudence focuses on the nature, normativity, and operation of law and legal systems generally. It concerns itself with questions such as the conditions of a norm being a legal norm, the nature of legal obligation, whether the rule of law is inherently valuable, the nature of adjudication, and the possibility and implications of legal pluralism. The subject matter of theoretical inquiry in special jurisprudence, on the other hand, is a discrete area of law, such as labor law,³ discrimination law,⁴ tort law,⁵ family law,⁶ international law,⁷ criminal law,⁸ and constitutional law.⁹

The main purpose of this paper is to provide a clarificatory account of the different kinds of theory that may be offered of an area of law.¹⁰ Such a clarificatory account is needed, in our view, for three reasons. First, existing taxonomies of the types of account that may be offered of an area of law tend to underrepresent the possibilities. For instance, it is notable that Arthur Ripstein reports being posed the following question about his account of tort law: “is the account descriptive, prescriptive, or interpretive?”¹¹ Our analysis provides a more fine-grained delineation of the kinds of account that one could offer, drawing distinctions between different kinds of nonnormative and normative account, and explaining that interpretive theories come in significantly different forms.

Second, it is quite often unclear what kinds of claim are being made in special jurisprudence. Identifying the nature of these claims allows us to determine the considerations bearing on their truth and, consequently, allows us to assess the success of the theory. A general example and a specific one. The general example is the category of theories that self-identify as “interpretive” theories. We will show that there are importantly different kinds of interpretive theory and that these differences impact on their success conditions. A specific example is Adrienne Stone's criticism of Yaniv Roznai's claim that the power to amend constitutions is impliedly limited inasmuch as it may not be used to change the identity of the constitution: the essence of Stone's criticism is that Roznai bases this claim on his conceptual distinction between constituent power and amendment power, whereas for his claim to stand, he needs a normative argument that constituent power (as actually exercised) has greater legitimacy than amendment power (again, as exercised in practice in a given context).¹²

Third, a lack of clarity afflicts the considerations that bear on the success conditions of theories of areas of law, especially in evaluating the “transparency” and the “fit” of the proposed theory. The “problem” of transparency concerns the extent to which, if any, a theoretical explanation of an area of law should be sensitive to, even track, the internally accepted accounts of that area of law by legal officials. Does the fact that the legal officials in a particular jurisdiction largely believe that the point of criminal law is to deter criminal behavior place limits on the success of a theoretical account that refutes this claim? A second problem in special jurisprudence is the problem of fit. This concerns the extent to which an account of an area of law must explain or otherwise take account of extant features of that area. A common mistake concerning fit in comparative constitutional studies entails universal claims made by extrapolating from a very small set of “canonical” constitutions. For example, the very first paragraph of Grimm's book on constitutionalism insists that a “constitution establishes the rules by which political rule should be exercised under law.”¹³ This—German/American—understanding of constitutions as a collection of legal rules ignores vast swaths of constitutional practice that deploys nonlegal norms for purposes aspects of constitutional governance.¹⁴ Ignoring relevant legal data is such a common fault in constitutional theories in special jurisprudence that an awareness of our taxonomy would make an author less likely to commit it.

Transparency and fit affect different types of theories differently: beyond a minimal threshold, it is as pointless to criticize (what we characterize as) a “hyper-reformist theory” on the ground that it does not completely fit the existing law as it is to demand transparency in all theories of areas of law. In what follows, we aim to make some progress on the precise role that considerations of “transparency” and “fit” play within different accounts. We argue, for instance, that considerations of “fit” can have an evidential bearing on normative theories under certain assumptions.

By addressing these three points, through clarifying the different kinds of theoretical claim that might be made about areas of law, and explaining their interrelationship, we hope to make theoretical claims in special jurisprudence more transparently assessable, and genuine disagreements more readily identifiable. If one theorist is making a conceptual claim and another is making a normative claim, there may not be any real disagreement between them. If parties to debates can be clearer about the precise nature of the claims made, progress can be more readily made.

In this article, we distinguish two basic types of theoretical account: nonnormative accounts and normative accounts. Section II explains the two central subspecies of nonnormative accounts of areas of law: (i) conceptual and ontological theories and (ii) reason-tracking causal theories. Section III explores normative theories of areas of law. Normative accounts subdivide into detached and committed normative accounts. Detached or committed normative accounts can be subdivided further into the following cross-cutting categories: (i) pro tanto or all-things-considered, (ii) hyper-reformist or practice-dependent, (iii) taxonomical or substantive. Section IV will deal with some objections to this taxonomy. We also delineate different types of “interpretive” theories in this section, and show how they are all accommodated within our taxonomy. Section V concludes.

Three preliminary points may usefully be dealt with up front. First, with the possible exception of reason-tracking causal theories, our taxonomy is a taxonomy of recognizably philosophical theories of an area of law. One could, for example, offer an empirical account of the effect of the gender of the litigants on tort liability. This could be a kind of theory of the effects of gender on liability. But it is unlikely to be particularly helpful to classify this kind of empirical theory alongside the kinds of philosophical theories we consider here. No one is in much doubt as to the nature of the claims being made by those seeking to investigate the empirical effects of gender on liability. They are empirical claims; the success conditions of empirical theories are reasonably well established in the extant literature on social science methods. Our taxonomy is motivated by a concern to bring greater clarity to a domain of scholarship in which the nature of the claims being made is not very clear; this aim will not be served by a discussion of empirical theories, whose nature and success conditions are less controversial. We include reason-tracking causal theories in our taxonomy—despite the essentially empirical, rather than philosophical, character of these theories—because these theories claim to uncover normative reasons as causes that motivate the creation or development of an area of law.

Second, the reasons for grouping together items in a taxonomy is obviously motivated by a concern to identify important similarities and differences between the taxonomized items. Such similarities and differences always run along some dimension. In taxonomizing theoretical accounts of areas of law, we argue that the relevant dimension is the nature of the claim being made by the theory, because this feature of the theory determines (a) the considerations bearing on the truth of the theory, (b) the kinds of claims that are licensed by the theory, and (c) the methods by which the theory will be established or assessed. Therefore, we restrict our taxonomy to philosophical theories of areas of law, without at all implying that this type of theorizing is the only useful mode of thinking theoretically about areas of law. On the contrary, we take empirical theories of law to be necessarily complementary to philosophical theories, if one is to understand legal phenomena more fully. Third, subject to the point just made about our focus on recognizably “philosophical” theories, we use the word “theory” interchangeably with the word “account.”

A final caveat: the taxonomy offered in this paper concerns theories that concern particular areas of law. It leaves out several metatheoretical concerns in special jurisprudence: concerns such as what count as “areas of law,” whether they are social or logical categories, whether they admit to revision, normative criticism, and reclassification, and so on. We address some of these metatheoretical concerns in another paper.¹⁵

II. NONNORMATIVE THEORIES IN SPECIAL JURISPRUDENCE

In this section, we outline two broad kinds of nonnormative theories in special jurisprudence.

A. Conceptual and Ontological Theories

Conceptual theories of an area of law, say competition law, aim to give an account of the concept of competition law. “Conceptual analysis” covers a range of different approaches. What we might call “traditional conceptual analysis” normally self-describes as aiming to identify the necessary and sufficient conditions for something to count as a practice of competition law.¹⁶ Conceptual analysis may go beyond this concern for “necessary and sufficient conditions,” however. It may sometimes seek only to identify the inherently likely features of a phenomenon. For instance, “coercion” might be an inherently likely feature of law, even if it is not a necessary condition.¹⁷ Although this is the traditional or standard account of conceptual analysis, it is clear that those who seek even these traditional conceptual analyses normally seek to identify more than simply the necessary and sufficient conditions of a concept.¹⁸ It is implicitly assumed in philosophical analysis that the necessary and sufficient conditions identified are ones that identify more explanatorily fundamental features in virtue of which something counts as that thing.¹⁹

Conceptual analyses differ not only in whether they seek necessary and sufficient or only inherently likely conditions, but also in whether they claim to be offering an account of the nature of a thing, or only our “concept” of that thing.²⁰ The very term “conceptual analysis” suggests a concern not with the world out there, but with our mental categorizations of it. Some theorists employing the tools of conceptual analysis, such as drawing distinctions, classifications, and family identifications and testing propositions with real or hypothetical examples, may be better considered as engaged in ontological analysis: they are seeking to identify the necessary, sufficient, or likely features of the thing itself. It seems to us there is room for both kinds of view: one could simply be seeking to unpick, say, the necessary conditions for the application of an important concept—without making an ontological claim—or one could go, in one respect, further and claim to be uncovering the nature of the thing itself. It is not always straightforward, however, to keep these things apart. There may be some “things” in the world that exist solely as concepts. For example, a conceptual account of “unicorns” or “fairies” is entirely possible, but offering an ontological theory—as far as we can tell—is not possible for either unicorns or fairies. Closer to our purposes, there are phenomena, such as “areas of law,” whose existence depends—at least in part—on certain mental categorizations (concepts) being in existence, even if (unlike unicorns) their conceptual dimension may not be all there is to them. It seems difficult, for instance, to test whether “Y” is a necessary feature of discrimination law without testing one's intuitions about the application of the concept “DISCRIMINATION LAW” to hypothetical examples. It might be objected that there is no such need: one can simply look to the real-world practice of discrimination law, and determine whether such-and-such a feature is a necessary one. If the feature does not appear in all instances of discrimination law, then it is not necessary. But before we can identify something as an instance of the practice of discrimination law, we already need to apply our concept of discrimination law (even if that concept is, no doubt, partly shaped by extant practice).

Legal practice, like all human institutions, is dynamic and messy. Theory-skeptic legal scholars often cite the ever-changing and messy character of legal practice to deny the possibility of concept-building, or theorization more generally.²¹ Their intuition is that (admittedly messy) practices cannot be explained by reference to theories that are—aspirationally at least—tidy and coherent. What this kind of skepticism ignores is that legal practitioners themselves make such proto-conceptual theoretical judgments all the time—determining whether a norm is a norm of discrimination law, for example, is often a practical necessity in order to decide what legal consequences follow. For example, the statutory limitation period for a tort and a contract claim may be different, and a lawyer would have to have some understanding of conceptual boundaries of “tort law” or “contract law” to persuade a judge that the norm in question belongs to the first area of law, or the second, or both, or neither.²² Especially when an area of law is underdeveloped or messy, practitioners are able to form their intuitions only because they have some concept of discrimination law, tort law, or contract law to work with. If practitioners can unwittingly theorize about a phenomenon, there is no reason to claim that a more systematic and scholarly approach to its theorization is impossible. The fact that the scholar, as much as the practitioner, may be wrong about their theoretical claims is beside the point.²³ Having said that, a theorist may well conclude—after analysis—that there are no necessary, sufficient, or even likely features that explain a phenomenon—asking the conceptual question does not mean that there will, necessarily, be clear conceptual answers. One may, for example, discover that there is only a family resemblance between the components that make up an area of law. Although unlikely for long-established norms (given the coherence-seeking forces typically at work in law), one may even find that nothing other than the fact that they are all found in the same piece of legislation binds together an arbitrary set of norms, norms that are nonetheless treated by the practitioners in a given jurisdiction as belonging to a particular area of law.

Conceptual analysis, in one or more of the forms explained, is a familiar form of theorization in general jurisprudence, concerned with elucidating necessary, sufficient, or likely features that contribute to the identity of the phenomenon that is law.²⁴ In The Concept of Law,²⁵ for example, Hart was, at least in part, interested in the identification of the necessary and sufficient conditions for the existence of legal systems and the validity of legal norms. The same project can, however, also be undertaken with areas of law in special jurisprudence. An excellent example is John Gardner's essay, “Torts and Other Wrongs,” which seeks to offer an ontological analysis of “tort law.”²⁶ Here is Gardner's analysis:

the law of torts is a law of

1. (a) civil recourse

2. (b) for wrongs

3. (c) in which primarily corrective justice is attempted

4. (d) in a primarily reparative mode

5. (e) in response to claims for unliquidated sums

6. (f) where the duties breached are non-contractual.

Each of (a)–(f) is defended by further conceptual or ontological analysis of terms of each condition, and by arguments that one loses the sense of the category “tort law” without that condition. So, for example, (a) is explicated by reference to the notion of a normative legal power, and the argument that, unless we include this condition, we lose the distinction between criminal law and tort law. Both areas regulate wrongdoing, but only tort law allows the victim largely undirected control over the enforcement of the wrong.²⁷ Drawing distinctions between the phenomenon under study and other, related, phenomena, in order to sharpen our understanding of the former, is a quintessentially conceptual move.

Having separated it from criminal law, Gardner's analysis then distinguishes tort law from the law of equity through criterion (d). In his view, equity is distinguished by the primacy of nonreparative, gain-based, remedies. In maintaining a distinction between the law of tort and the law of equity, Gardner's view reflects the ordinary view of legal officials and practitioners. However, suppose we thought that gain-based remedies are not a central feature of the law's response to equitable obligations and that, on reflection, there are no conceptual or normative differences between equitable obligations and tortious obligations. We might then conclude that the law draws the boundaries of these areas of law in a way that serves no useful purpose.²⁸ If two sets of norms have the same conceptual structure and normative justification, it is unlikely to promote consistent reasoning to treat them as separate areas of law. Notice, however, that this moves beyond merely offering a conceptual analysis of the socially recognized area of law in question. To generalize, an ontological analysis of an area of law might serve as a premise in an argument that the existing recognized boundaries of two ostensibly distinct areas of law are incoherent: such a claim involves a normative taxonomic claim, i.e., a claim about how areas of law should be classified.

The methods by which conclusions concerning a particular concept are reached are typically the traditional tools of analytical philosophy, such as drawing distinctions, explicating the relationships between phenomena, and testing propositions about necessity or sufficiency with real or hypothetical examples. Accordingly, a theorist of a particular area of law, for instance, identifies norms, and inferential relationships between norms, that, according to her own conceptual intuition, constitute the area of law as distinct from or related in particular ways to other areas of law, while also giving substantial weight to the application of the concept in legal practice.²⁹ Her conclusions will be tested by consulting her own intuitions and the evidence available from the practice about real and hypothetical norms. Intuitions and practice also inform which cases are treated as paradigmatic, and which are peripheral or controversial, with varying impact on theory construction. We hasten to add that we do not assume that a “case” must be one concerning an individual engaged in a particular action whose legal character is under scrutiny: such methodological individualism could distort understanding, and often lead to ignoring structures and emergent properties of systems. A “case” for a philosopher can be different from how a lawyer understands the term, and could include a type of legal liability (the “case” of indirect discrimination liability, for example) or even an entire area of law.

To the extent that conceptual or ontological theories are mainly concerned with the nature, structure, concept, etc. of an area of law, their exclusive focus on the conceptual data provided by law-making institutions (viz legislatures and appellate courts) is understandable. However, this inquiry might benefit from more robustly empirical methods in some domains. In recent years, experimental jurisprudence—“the study of jurisprudential questions using empirical methods”³⁰—has begun to emerge.³¹ Practitioners of experimental jurisprudence survey laypeople, for example, as to their application of a concept, such as causation or intent, in hypothetical examples, to test whether the “ordinary” concept of X is affected by certain factors.³² This approach seems most likely to be fruitful insofar as the law purports to apply “the ordinary concept of X” in delimiting the boundaries of an area of law, or insofar as is one is simply seeking an account of such a legal concept; experimental jurisprudence can test the veracity of the law's claim to employ the ordinary concept. There is, of course, a real probability that the law's concept has taken on its own distinctive contours and so the evidential bearing that an account of the folk concept of X has in identifying the legal concept of X is necessarily limited.³³ More generally, it might also be fruitful to expand the theoretical data set to examine not only the docket of appellate courts, but also those of lower courts and tribunals, as well as the practice of other, nonjudicial, law-applying officers such as police and bureaucrats, in constructing a conceptual or ontological theory.

B. Reason-Tracking Causal Theories of Areas of Law

A reason-tracking causal theory of an area of law is a kind of sociopsychological empirical theory in special jurisprudence that seeks to identify the more fundamental motivating normative reasons that, as a matter of fact, guide legal officials in creating, developing, and determining the contours of that area. The sense in which such a theory looks for “more fundamental” reasons is that it does not merely describe the surface-level institutional facts of legal practice, for instance, what judges said in cases X, Y, Z, or even the explicit formulation of the norms given in legal materials. Rather, it seeks to explain what normative reasons truly motivated legal officials to say such-and-such in cases X, Y, Z or the legislators to enact a particular statute. However, despite this concern for seeking more fundamental explanations of surface-level legal phenomena, such theories are descriptive in the sense that they aim to identify those more fundamental reasons that as a matter of fact influence the institutional content of the law.

Insofar as it insists on those reasons that, as a matter of fact, influence the institutional content, this type of account is not, then, a Dworkinian “best moral justification” of the institutional facts. The best moral justification of the institutional facts may or may not be the one that motivated the judges to determine the content of the law in this way. If judges were as skilled moral reasoners as Dworkin's Hercules, then there may be a convergence between the Dworkinian best moral justification of an area and the fundamental motivating reasons for the legal content of that area. But in our world, while a morally superior explanation M1 may, sometimes—when we can rationally believe that legal officials are reasonably competent moral reasoners—be the more likely explanation, than a morally inferior explanation M2, the moral superiority of M1 would at best be one indication that it in fact motivated the judges to create the legal content. What this shows is that, for this kind of theory, the moral superiority of an explanation may have an epistemic, empirical, role in the theory.

The epistemic relevance of the moral superiority of an explanation to a reason-tracking causal theory might lead to some confusion. It might give the impression that, after all, reason-tracking causal theories in special jurisprudence are inherently “normative.” We think that this impression is misleading. This is because the truth of a reason-tracking causal theory is never logically guaranteed by its normative appeal. Rather, the normative appeal of an explanation has an epistemic, empirical significance: under certain, and in many systems rarely existent, conditions concerning the quality and independence of judges, the moral superiority of an explanation may make it more likely to have been the reason that influenced the legal content. Similarly, the fact that an explanation entails that judges were motivated by mutually inconsistent normative values does not establish that the explanation is false. It only, depending on one's views about the moral reasoning abilities of officials in particular contexts, renders the explanation less likely to be true. The epistemic relevance of “morality” or “coherence” does not undermine, then, the fundamentally nonprescriptive nature of these theories. Given this fundamental concern to identify what reasons in fact motivated the development of the legal content in the area, these theories are normatively inert. Nothing directly follows from them about what ought to be done by anyone. If we offer a compelling reason-tracking causal account of tort law in terms of efficiency as the normative driver for judicial and legislative development of tort law, this is logically compatible with tort law being of no genuine value whatsoever. Having said that, we do not deny that a reason-tracking causal account may acquire normative relevance if seen by a judge as giving her institutional reasons to decide cases in a particular way.

If what one is trying to do is offer a reason-tracking causal account, one's method should be determined by the nature of such an account. In particular, whether one seeks an account that is “transparent” to legal officials or not will depend on one's views about certain empirical questions. These empirical questions concern the relative likelihood of a reason in fact guiding official decision-making when the official does not articulate that reason (as guiding their decision-making) compared to the likelihood of the decision being guided by an explanation that is articulated by the official (or that would be able to be articulated under some suitable hypothetical conditions). The important point is that there is nothing logically inherent in the nature of a reason-tracking causal theory that insists on transparency. The force of transparency is always empirically contingent. This point tends to be neglected in discussions of methodology in special jurisprudence where it seems often to be assumed that transparency either is or is not baked into the very idea of offering a theoretical account. For instance, Andrew Gold, in an insightful discussion of “interpretive” theories of private law, initially seems to be in agreement with our view that the correctness of a nontransparent (or, transparent, for that matter) explanation of an area of law is an empirical, contextually contingent matter. For instance, he writes that “judicial opinions need not be perfectly transparent, but a theorist should at least be able to explain why judges would claim to reason in the way that the language of their opinions suggests.”³⁴ Later, however, when discussing whether a pluralistic theory of tort law is problematic from the perspective of a “coherence” criterion, his discussion of the merits of such a criterion becomes untethered from the empirical issue of how likely it is that institutional actors act on the basis of incoherent reasons.³⁵ If, however, the aim is to provide a reason-tracking causal account (which may not be Gold's aim—he describes his position as “interpretivist”—a label we deconstruct later), the answer to the relevance of “transparency” and “coherence” simply cannot be answered as an abstract theoretical matter. Their relevance is empirically contingent. Thus, we claim that their actual reasons may or may not be transparent to the judges or legislators: sometimes they would know their reasons for doing what they do, at other points they may hold their normative assumptions unconsciously or subconsciously.³⁶ Furthermore, even when their real reasons are transparent to judges, whether they report these reasons accurately in their judgments is also empirically contingent.

There is no bright line between reason-tracking causal theories and what might be termed “deeply empirical” causal theories of an area of law, for both seek to make sociological or psychological (rather than primarily philosophical) claims about law. In general jurisprudence, at least one major strand of legal realism is an empirical theory, in this sense: it claims that the primary causal explanation of judicial decisions is the classification of the facts by the judge as belonging to a certain factual pattern extraneous to law, facts that give rise to reasons that are external rather than internal to legal doctrine.³⁷

Empirical theories can be useful in testing the claims made by ontological and normative theoretical scholarship. They can highlight the gap between the law in action and the law as constituted by authoritative sources. They can expose the role of power, social identity, culture, and political ideology as factors that shape the law and influence its functioning and implementation.³⁸ For example, critical race theorists have shown how one of the factors that led to the celebrated judgment of the United States Supreme Court in Brown v. Board of Education,³⁹ desegregating America's racialized schools, was a temporary convergence in the interests of black people and some powerful white groups.⁴⁰ Much of critical legal scholarship aims not at providing theoretical explanations of legal phenomena by itself, but rather to “destabilize the complacent and/or apologetic focus on the naturalness or autonomy” of such phenomena.⁴¹ At least when informed by robust social science methods of inquiry (instead of a priori and unfalsifiable ideological presumptions), empirical theories can perform an extremely valuable role of revealing aspects of the practice that need to be considered by a normative theory; a justification of a practice may be incomplete or skewed if it fails to account for its actual operation.

III. NORMATIVE THEORIES

Normative theories are concerned with assessment of areas of law in terms of genuine reasons for action or genuine values (normative theories simpliciter), or reasons for action and values that exist from a perspective that is not necessarily shared by the theorist (detached normative theories). Normative theories simpliciter aim to identify the genuine reasons and values—or more generally the rationales—that justify particular areas of law or, indeed, require the reform or abolition of an area of law. Their overarching question is: “is X area of law justified?” or “Should society have the norms that constitute area X?” The idea of a normative theory is often treated, however, in a rather monolithic way, without acknowledgment that there are different types of normative theories that could be offered for an area of law.⁴² We will draw certain cross-cutting distinctions in this subsection to map various normative approaches adopted in special jurisprudence, fully recognizing that even the possibility of some of these distinctions may well be controversial, and rejected by some theorists.

A. Detached v. Committed Theories

A first distinction is between detached and committed normative theories.⁴³ A detached normative theory proceeds by assessing whether the area of law can be explained in terms of a putative normative value, reason, rule, principle, or set thereof, without the theorist taking a stand on whether that value, reason, rule, principle, or set thereof is a genuine value, reason, rule, or principle. An area of law is “explained by” a putative normative reason when, were the rationales valid, it would rationally support the adoption of the norms constitutive of that area. For instance, we might offer an efficiency theory of discrimination law while remaining aloof from whether efficiency is a goal that should be pursued. We would proceed by considering the implications of efficiency for the problems faced by the area of law and consider whether an efficiency-based solution to those problems would rationally support the legal rules in question (or enough of them). A detached theory is conditional; the conditionality is in relation to whether the value, etc. doing the explaining is a genuine one. It is (implicitly) of the form: if N [a putative value] is valid, then it pro tanto justifies this area. By contrast, a committed normative theory includes an assertion of commitment to the value that is serving as the explanans (explanans = the thing doing the explaining). It asserts that, say, tort law is justifiable by efficiency (i.e., that (i) tort law (or, sufficiently significant components thereof) achieves efficiency, and that (ii) efficiency is desirable). Notice, however, that a committed normative theory could still be conditional in the way we explain in the next section.

What distinguishes reason-tracking causal theories from detached normative theories? The two may seem similar in that in neither case does the theorist take a stand, qua theorist, on the genuine moral value of the explanation. A further similarity is that both only license claims about what one genuinely ought to do on the basis of some (further) normative premise that the theorist does not purport to defend. It may seem, then, that there is a case for treating detached normative theories and reason-tracking causal theories within the same taxonomic category.

The main difference, however, which justifies their inclusion in this part of the taxonomy, is that detached normative theories do not necessarily purport to establish that the reasons that they identify are the ones that in fact motivated the development of the legal content. In principle, one could argue that if Kaldor-Hicks efficiency is genuinely-to-be-pursued, then one ought to have tort law, and yet still believe that it is false as a causal account of tort law. It would be a remarkable coincidence, of course, if tort law could be justified by Kaldor-Hicks efficiency even if it never guided the development of the legal content in the area. Nonetheless, it is a logically separate kind of claim. A detached normative theorist is, strictly, freed from demonstrating that their theory actually influenced legal practice.

It might be questioned whether “detached normative theory” is a worthwhile enterprise.⁴⁴ If the theorist does not purport to identify a genuine reason that justifies the legal content of the area, and nor do they purport to be able to explain what actually influences the legal content of the practice, why bother? In our view, such theories can nonetheless be valuable. This is because it is often unclear whether something is of genuine value or genuinely permissible or required. Moral uncertainty (not merely empirical uncertainty) is a pervasive feature of practical life. It can thus be valuable to investigate whether a putative value may justify an area of law, in the hope that further moral analysis may be able to establish that it is of genuine value.⁴⁵

If the primary distinction between detached and committed normative theories is the theorist's assertion of a commitment to the genuine normative value of the propositions in the theory, it may seem odd to distinguish them. Why distinguish between theories depending on the theorist's moral commitments? The main reason for doing so is that, typically, detached normative theories do not provide us with any reason for believing in the moral (or other) value of the explanans. Their posture is merely one of implicitly stating that such-and-such, a putative moral position, justifies the law. Therefore, it is likely that, without more, we will not have sufficient grounds for concluding that the theory has any genuine normative bite if the theory is offered as a detached theory. We will likely need to look elsewhere—to committed normative theories—for arguments showing why the explanans is valuable.

B. Pro Tanto v. All-Things-Considered Theories

A normative theory may be pro tanto or all-things-considered. A pro tanto theory sets out a defeasibly sufficient justification, or set of justifications, for the area. An all-things-considered theory purports to demonstrate that the area of law is what one ought to have in the circumstances to which the theory applies. For instance, a theorist could argue that a defeasibly sufficient justification of tort law is the fact that institutional recognition of reparative, interpersonal non-agreement-based moral rights increases the probability of conformity to such rights, or usefully resolves indeterminacies in their content. However, the theorist may remain neutral on whether tort law is justified all-things-considered. It may be that tort law creates such woefully counterproductive incentive structures that it will lead to more violations of moral rights than if tort law did not exist. Or it may be that, when compared with the benefits of alternative possible institutional arrangements, the opportunity cost of tort law arrangements is too great. These kinds of defeating circumstance may well entail that tort law is not justified all-things-considered. This is not the same distinction as detached versus committed. A detached theory may only seek to identify a pro tanto justification by the lights of efficiency for the area or it may seek to show that by the lights of a moral system the theorist does not take a stand on, it is all-things-considered justifiable.

In any sufficiently complex jurisdiction, offering all-things-considered theories for most areas of law is likely to be extremely difficult, if not impossible. Doing so for legal systems generally, i.e., without reference to the particularities of a given jurisdiction, even more so. Furthermore, the application of even an all-things-considered normative theory of an area of law to a specific multifaceted real-world problem arising within that area of law is unlikely—on its own—to generate specific prescriptions to act or decide in one way rather than another, although it will narrow the field of permissible possibilities.⁴⁶

C. Hyper-Reformist v. Practice-Dependent Theories

Normative theories are unconstrained by considerations of “fit” in the following sense: it does not necessarily count in favor of a normative theory that it fits the practice. Fit bears no necessary connection to normative appeal. This is a central contrast between normative theories and conceptual and reason-tracking causal theories: the latter fail if the theory substantially fails to fit the data, and lack of fit counts against the theory.

This is consistent, however, with the following roles for “fit” in normative theories. First, if the conclusion of a normative theory is that “X area of law” ought to be reformed or abolished, or “X area of law is problematic because of feature F,” the theorist must, of course, accurately identify X and F.

Second, in relation to certain kinds of normative issue, there may be a weak evidential significance to considerations of fit. As we noted above, in relation to reason-tracking causal theories, the moral superiority of an explanation can have, in certain contexts, an epistemic significance: it can make it more likely to be the explanation of a legal norm. If a particular legal rule is found in multiple legal systems, created under conditions that are conducive to moral correctness, and arrived at independently by multiple legal systems, this at least raises a chance that the rule is justified. Of course, until a justification for the rule is actually identified, the mere fact that it is widely posited in these circumstances is unlikely to be a robust defense of the rule.

Finally, normative theories might still begin with existing practice. This might simply be because they are directly concerned to contribute to the normative assessment of present institutional arrangements. There might, also, be a theoretical value in doing so in relation to certain types of theoretical question. The law is a rich repository of moral distinctions and reasoning, developed across thousands of particular examples; it may assist in identifying important moral distinctions that would otherwise be missed by even the most ingenious exponent of trolleyology.

Normative theories can be more reformist or more vindicatory of existing practice. At one extreme, a theory could be radically reformist—hyper-reformist, we might call it—and yet still be a normative theory of a particular area of law. For instance, one could offer a hyper-reformist theory that argued that the morally ideal form of tort law is one that only protects people's freedom-based rights and no more. This theory is hyper-reformist in the sense that it may not purport to justify any significant parts of the extant area of law in question. Such a theory may hold that, for instance, the torts of defamation and privacy are unjustifiable, because they go beyond protecting people's freedom-based rights. Even so, this is a normative theory of tort law since it still prescribes the existence of something recognizably like existing practice. Consequently, it is clearly an error to criticize hyper-reformist theories on the basis that they recommend wide-ranging changes to an area of law. A theory of the morally best form of tort law, which recommends radical changes to extant practice, can still be a theory of tort law insofar as it justifies some key aspect of existing systems of tort law. For example, a hyper-reformist theory of tort law that calls for limiting its scope to the protection of people's freedom-based rights must—either explicitly or implicitly—have a view about the nature of tort law, which underpins its hyper-reformist claims, else it risks recommending the existence of something justifiable, but not properly described as tort law at all. In short, even hyper-reformist theories of areas of law are accountable to the existing contours of the practice of that area to some minimal extent.

D. Taxonomical v. Substantive Theories

A normative theory may be primarily directed at the taxonomical reconceptualization of an area of law rather than proposing a justification or critique of its norms as such.⁴⁷ For instance, the argument could be made that equitable wrongs ought to be merged with tortious wrongs to constitute a new area of law. This would not necessarily involve a proposal to reform the content of the existing law. However, conceptualizing these wrongs as part of a single area of law might improve judicial decision-making insofar as inconsistencies in the future development of the rules are more likely to be avoided, and apt analogies between normatively related rules identified.

IV. OBJECTIONS

A. Nonnormative Theories Are Impossible in Special Jurisprudence

Previous iterations of this paper have been met with the objection that nonnormative theories of an area of law—conceptual theories and reason-tracking causal theories—are impossible. Here are some more fine-grained versions of this objection and our responses to them.

1. “Legal Validity Depends on Moral Validity”

Suppose one is an antipositivist. This means that one denies that legal validity (at least ultimately) depends only on social facts.⁴⁸ Suppose further one is a moral antipositivist: one believes that legal validity depends, necessarily, on moral facts.⁴⁹ If so, then identifying the legal content of an area of law will necessarily depend on moral argument, since an area of law is constituted by binding legal norms, and whether a norm is such, on this view, necessarily depends on moral facts. If we grant all this, does it follow that conceptual and reason-tracking causal theories in special jurisprudence are impossible? It does not.

First, it is difficult to deny, as a matter of social fact, that the practice of law makes categorizations of norms into certain groups. These categorizations may be morally arbitrary or otherwise indefensible. For instance, one might think that the social categorization of norms into “equitable” and “nonequitable” does not track any moral distinction. But this is clearly a different point from whether this categorization exists as a matter of social fact. If these categorizations exist as a matter of social fact in a given jurisdictional context, then their conceptual content is a subject matter of special jurisprudence. It may turn out that the concept so revealed is incoherent, arbitrary, vague and so on—but so be it.

It is notable in this connection that Dworkin—a antipositivist—excluded the conceptual boundaries of “areas of law” from his notion of an “interpretive concept.”⁵⁰ An interpretive concept, for Dworkin, is a concept whose content is determined by moral argument. For instance, the nature of democracy, according to Dworkin, depends on what is the morally best interpretation of democratic practice.⁵¹ Dworkin did not regard the concepts that marked out areas of law (or, in his terminology, “departments of law”) to be interpretive concepts.⁵² Although he insisted that “compartmentalization is a feature of legal practice no competent interpretation can ignore,” he also claimed that the boundaries of these “departments of law” themselves are not interpretive phenomena but “based on tradition.”⁵³ Sure, a judge may well seek to alter the traditional boundaries of classification for moral reasons. But these alterations will be effected, Dworkin accepts, only “[i]f he persuades the profession of his view.”⁵⁴ So, unlike his interpretive claim that what the law is cannot be determined without answering some normative questions, the existence and scope of an area of law—for Dworkin—is determined by reference to a social fact concerning its recognition as such by legal practice.

Of course, Dworkin could be wrong about this and fail to carry through appropriately his nonpositivistic commitments into the domain of special jurisprudence. We do not think this is the case, however, with respect to the existence (as opposed to the boundaries) of an area of law. First, the central debate between positivists and antipositivists is one about legal validity—about the kinds of fact that ground legal content. Whether a norm belongs to an area of law is not a question of the norm's validity or bindingness as a legal norm. It is, therefore, logically consistent to hold a social-fact view about the contours of legal areas and a non-social-fact view about legal validity. Second, even if the classificatory bases that pick out areas of law are informed by an implicit moral principle or set of principles, those principles may or may not be fully realized in the social categorization that is employed in legal practice. It should be possible to accept two kinds of inquiries here: one that seeks to understand the bases-of-classification actually employed in legal practice to delimit areas of law, and one that seeks to articulate the categorization that would be employed if the underlying principles of the socially employed categorization were consistently extrapolated. Indeed, the view that the boundaries of an area of law are solely an interpretive phenomenon seems implausible. Imagine that for some curious historical reason, a legal system sets up separate courts to deal with “Tuesday law.” All legally relevant events that take place on Tuesdays—crimes committed, contracts agreed on, injuries inflicted—are litigated before these special courts. A separate set of the bar, called Tuesday lawyers, develops specialism in Tuesday law. Law schools have special courses teaching Tuesday law. There is even a journal called Tuesday Law Review (published, oddly, on a Wednesday). For all that, no one can quite remember why Tuesday law developed in the first place, and there seems to be no contemporary reason for its separate existence. In fact, let us suppose there are very good, even overwhelming, reasons why Tuesday law should not exist as a separate area of law, but should instead be disaggregated and merged with the rest of criminal law, tort law, contract law, and so on. It is, nonetheless, not at all clear to us in what sense one could claim that Tuesday law does not exist as an area of law in that jurisdiction. Elsewhere, we have defended the claim that an “area” of law is a subset of the legal norms in a legal system, which is intersubjectively recognized by the legal complex in that system as a subset of such norms. Its existence is, thus, a matter of social fact, although (pace Dworkin) its precise scope and boundary may require normative determination. Appeals to the purported foundations (ie aims/functions) of the candidate areas of law are inevitable in such cases.⁵⁵

Normally, the areas of law that are the subject of theoretical inquiry are those in which the relevant members of legal practice consider the norms to have some important rational connection such that membership in the subset has an important normative implication. For instance, legal actors will consider classification as part of area X as implying the apt operation of other rules of law. If something is classified as a tort, it is considered apt to respond to it legally in certain ways, for instance, with damages claims before civil courts. Thus, some might say that this is part of what it is for an area of law to exist: an area only exists if something's being classified as part of that area has a legal significance, such as giving rise to pro tanto legal reasons for certain legal responses. These kinds of categorization—what Peter Cane has called “dispositive” classifications⁵⁶—seem to us to be an important subclass of areas of law. But it seems too narrow to exclude from the objects of theoretical inquiry socially recognized classifications that do not have normative significance within legal practice. Even if the classification of a norm as a norm of labor law has no dispositive significance within legal practice, there is still a clear sense in which labor law exists as an area of law within law schools, and it could still be pedagogically rewarding and theoretically interesting to investigate, for instance, whether its norms have any rational unity.

2. “Conceptual Disagreement Reduces to Moral Disagreement”

A second objection to thinking that nonnormative conceptual theorization of an area of law is possible might derive from Finnis's methodological argument in Natural Law and Natural Rights. ⁵⁷ Finnis argues that when we adopt the internal perspective on the law, from which the law creates norms, we will find that people disagree about the necessary and sufficient conditions that a social organization must satisfy in order to amount to a “legal system.” If legal theory is to avoid simply collecting a variety of different perspectives on the nature of law, it must, for Finnis, identify the “central” case of law, which is the case of law that we end up with by arguing from first moral principles about what kinds of state institutions we need in society.⁵⁸ Again, we cannot fully engage with this argument here—our primary aim is to categorize and explain different approaches to special jurisprudence with the goal of clarifying their distinctive claims. But we are skeptical of the claim that the existence of disagreement among persons who adopt the internal perspective necessarily entails what is, in substance, full-fledged normative political philosophy, and no longer conceptual theory at all.⁵⁹ For instance, suppose that someone disagreed with one of Gardner's necessary features of tort law, say (b), that the law of torts is a law of wrongs. There are various ways in which this disagreement might be explained or resolved without resort to normative argument. As Shapiro says, conceptual disagreement might happen because one person “engages in fallacious reasoning, overlooks relevant evidence, lacks imagination, indulges in wishful thinking, or brings to bear a different worldview than their interlocutor.”⁶⁰ At any rate, even in the face of pervasive disagreement conceptual theory could still proceed to identify the conceptual implications of a prominent or widely shared understanding of the area in question. It seems, then, that nonnormative conceptual theories of areas of law are possible even if one accepts Finnis's general jurisprudence account of theorizing about legal systems.

This is not to deny that conceptual and reason-tracking causal theories may require the theorist to engage in what Julie Dickson calls “indirect” evaluation.⁶¹ They do not necessarily require the theorist to make committed normative judgments, such as “x is morally required” or “x is morally valuable.” But they may require the theorist to make noncommitted normative judgments—that is, judgments that X is required, permissible, valuable (etc.) from a certain perspective, particularly the perspective of those whose actions primarily constitute the practice. For example, the theorist may, herself, regard corrective justice as moral bunk, but to explain whether corrective justice is a necessary feature of the socially recognized concept of tort law, she may need to determine the implications of the idea of corrective justice.

3. “Detached Normative Judgments Are Impossible”

An antipositivist may still object to the various subclassifications we make within normative theories. What meaning, it might be asked, can be given to “detached normative theory” if one believes that determining the content of the law necessarily involves moral argument? Any proposition about the content of the existing law is, it might be said, a committed normative proposition, since it takes a stand on the moral justification of the existing law. This objection overlooks the possibility of offering a detached theory as a possible account of the legal content of the area. Suppose the antipositivist is uncertain about whether efficiency is a genuine value. If so, she might offer an account of the following form: “if efficiency is a value, then the legal content of this area is X, Y, Z.” For this kind of antipositivist, a detached normative theory may be a kind of preliminary to determining the true legal content of the area.

Furthermore, any plausible antipositivist theory of law must still find a way of recognizing a distinction between the law as it is and the law as it ought to be.⁶² So an antipositivist might still engage in hyper-reformist theorizing. For instance, let us take Greenberg's moral impact theory, according to which the legal obligations in a jurisdiction are the moral obligations created by the actions of legal institutions. So, for Greenberg, a statute that requires us to reduce our carbon emissions by 25 percent over the next year may successfully create a moral obligation on us to do so, even if such an obligation would otherwise be morally suboptimal.⁶³ So our categorization still makes sense from an antipositivist perspective. And, of course, if we dispensed with detached normative theory, this would exclude an important brand of theorizing about areas of law.

B. “Interpretivist Theories Are a Tertium Quid”

Some theorists offer “interpretive” theories of areas of law. The nature (and value) of these theories has itself become an object of significant discussion.⁶⁴ What seems to undergird them, as a whole, is the negative claim to be neither simply descriptions of the existing law nor purely normative.⁶⁵ Indeed, their critics accuse them of incoherently adopting a nonexistent methodological tertium quid.⁶⁶ Our aim in this section is to deconstruct different possible versions of interpretivism and show that they are accommodated by our taxonomy. This deconstruction aims to assist in the clarification of the kind of claim made by these theories. Given the variety of different kinds of claim that might be made under the banner “interpretivism,” it seems preferable to us to use more fine-grained descriptions of the kinds of theory being offered. Ultimately, we show, these theories are either nonnormative or normative theories, and fit neatly within our taxonomy.

1. Interpretivism as Detached Normative Theory

One reason why interpretive theorists may be reluctant to characterize their inquiry as “normative” is that they might be engaged in detached normative theory: avoiding the “normative” label is a means of indicating that one is remaining aloof from the question of whether the explanation of the area one identifies is truly normatively valid. For example, an interpretive theorist could purport to show that the area is explained by efficiency, bracketing whether that is a goal-to-be-pursued (the detached version).⁶⁷

This is one reading of Smith's characterization of interpretive theories: “[i]nterpretive theories aim to enhance understanding of the law . . . by revealing an intelligible order in the law, so far as such an order exists.”⁶⁸ The revelation of such an order proceeds by showing how features of the law are “best explained.”⁶⁹ If one seeks an explanation in terms of reasons for the area of law, but one remains uncommitted on whether those reasons are genuine reasons, then one offers a detached normative theory.

2. Interpretivism as Reason-Tracking Causal Theory

Smith's characterization could also be read as saying that interpretive theories are reason-tracking causal theories. On this reading, these theories seek to reveal an intelligible order—a set of reasons—which as a matter of fact guide the practice in that jurisdiction. This reading is supported by what Smith says about whether a theory should be “transparent.” Smith seems to see transparency to be significant for epistemic, empirical, reasons: it should increase our confidence that the explanation is the one actually underlying legal practice.⁷⁰ Note the importance, however, of disaggregating these two possible things one could be doing as an “interpretivist.” If one is offering a detached normative theory, there is no inherent reason to care about transparency. If one is offering a reason-tracking causal theory, one is bound to provide evidence that one's theory is as a matter of empirical fact the one that is guiding the determination of legal content. This is likely to involve a greater (epistemic) significance being attached to transparency.

This conception of interpretivist theory also seems to underpin Smith's discussion of interpretivism when he prefers a “moderate” morality criterion for what constitutes a good theory of an area of law. This criterion requires that a theory must “show how legal actors could sincerely, though perhaps erroneously, claim that the law was morally justified.”⁷¹ The rationale for this criterion, at least in part, is that the theory is more likely to be the one that explains the rules and reasoning in the area because legal officials claim that the law is morally justified.⁷²

3. Dworkinian Interpretivism

Dworkinian interpretivism seeks the explanation of an area of law that (i) best fits the features of the area and (ii) has the most moral appeal. It is distinct from reason-tracking causal theory in two ways. First, having identified the explanation satisfying (i) and (ii)—if such an explanation exists—it concludes that this is already part of the law.⁷³ So, Dworkin argues, there may be a legal right to engage in civil disobedience in a given jurisdiction even if its apex court has held otherwise, for “though the courts may have the last word in any particular case about what the law is, the last word is not for that reason alone the right word.”⁷⁴ Second, this brand of interpretivism is not committed to the idea that the explanation identified is the one that actually influences the development of the area. Of course, if the explanation cannot be said to have any causal impact, this may reduce the probability that it will fit the practice, but this is a contingent matter. Dworkinian interpretivism is essentially a kind of committed normative theory, albeit one that insists that the upshot of its argument is a conclusion about what the law is.

Within special jurisprudence, Dworkinian interpretivism has been especially popular with constitutional theorists.⁷⁵ Constitutional law is particularly amenable to Dworkinian insights because of the very wide latitude judges have in deciding politically charged decisions due to only being governed by vague and broad constitutional provisions and often conflicting precedents. This is precisely what makes it easy to characterize constitutional judgments as “wrong in law,” as permitted by Dworkin's one-right-answer thesis. It has also been applied to explain US antitrust law, in part because it is an area of law where US courts have been very willing to revisit or ignore precedents.⁷⁶ McGinnis and Meerkins's antitrust theory is Dworkinian (rather than causal) because their overarching organizational principle for this area of law—consumer welfare—explains US antitrust law (on their view) despite some legislators in fact being motivated by other purposes when legislating in this area of law.⁷⁷

4. Interpretivism as Restricted Reason Normative Theory

Here are Beever and Rickett, self-described interpretivists:⁷⁸

The theorist begins with what appear to be salient features of the case law. She then attempts to produce a theory that explains these features . . .

Beever and Rickett at points seem clear that they are not causal theorists:

the legal theorist is primarily concerned, not with the reasons individual judges may have had for reaching their decisions, but with the best explanation for those decisions.⁷⁹

Nonetheless, Beever and Rickett insist that the “best explanation” must be one that is in some way “internal” to the area.⁸⁰ In this way, these theorists restrict the kinds of reason that are permitted to enter into the best explanation—the best justification—of the area. For Beever and Rickett, the fact that a normative consideration provides a reason for having a particular feature of that area is not sufficient to make it “internal” to the area. It follows that, for them, an efficiency-based reason is not “internal” to tort law. It is not entirely clear, however, what “internality” means in this case. One possibility is that this brand of interpretivist theory asserts that judges have been influenced indirectly or subconsciously by the reasons that the theory identifies, but that judges articulate these reasons badly (inadvertently or otherwise). However, some interpretivist theories make little effort to vindicate this causal claim, such that it seems unlikely that these theorists are really causal theorists.⁸¹ If “internality” does not mean this—that is, if internality is not an indication that the theorist is trying to track the implicit, if unarticulated, explanation underlying the practice as a matter of fact, then the theory is ultimately a normative theory. If so, its exclusion of certain reasons from the justification of the practice has to be defended on normative grounds.

C. Interpretivist Theories: The Importance of Clarifying Their Nature

As noted above, interpretive theorists are typically keen to distinguish their enterprise from at least certain types of normative theory. Beever writes that “though the theory is prepared to jettison some elements of the positive law, it is not prescriptive. It is not an account of how the law ought to be in some ideal, or even non-ideal, state.”⁸² If this is just a declaration of being a detached normative theory, then it makes perfect sense. Nonetheless, and somewhat puzzlingly, it seems that interpretive theory does offer committed prescriptions: “interpretive legal theory will often call for changes to details of positive law.”⁸³ Is this call for change just a detached prescription? Maybe. Perhaps Beever is saying that “the details of the law should be changed from the perspective of, or given the rational commitments of, the area of law—but whether it really ought to be changed is another question.”

Or perhaps Beever is saying both that, from perspective of the area of law, the law ought to be changed and that this alone provides a genuine and sufficient normative reason for changing the area of law. On this latter understanding, the interpretive theorist, in prescriptive mode, claims that the law ought to be changed in order to be more consistent with whatever explains the central features of the area of law, while bracketing the issue of whether that explanation is morally valid. This prescriptive aspect of Beever-interpretivism seems problematic to us. This brand of interpretivism can only validly make conditional prescriptive claims—that is, claims that are conditional on the validity of the value (etc.) serving as the explanans. To see this, suppose that the only possible explanation of employment law in a given jurisdiction is the principles of the Cenotine religion (an imaginary religion whose precepts include doing fruitful work on every day of one's life),⁸⁴ but one branch of this legal system's employment law is not consistent with these principles. Is there a reason to alter this branch so that employment law as a whole is more consistent with Cenotine principles? If you like, there is a reason from the Cenotine perspective for doing so. However, that is no true reason. In relation to the law, if we conclude that the only possible justification of a set of rules is X, we may wish to say that, if the law is to adopt a maximally consistent set of normative propositions, then it ought to alter the law to conform to X. Regardless, consistency-based reasons are only reasons worth paying attention to if a minimal threshold of genuine value is met. In our example, Cenotine principles have to meet some objective threshold of value before we can make committed claims about making employment law consistent with them. Even if we were to conclude that the only thing that could make sense of an area of law were Cenotine—all other explanations are inadequate—this would still provide us with no reason to alter that area of law to make it more Cenotine-like unless we thought that Cenotine principles were in some way good, or unless, perhaps, people have planned their life around the area of law's moving in a Cenotine-like direction—a rule-of-law concern.

Interpretivists could avoid the problem we identify here by arguing that the explanation they offer has some genuine moral merit, even if it is not morally optimal. There may then be reasons for the law to continue along a morally suboptimal path, and interpretivists could validly offer prescriptions. This is precisely the approach that Smith adopts when endorsing “the moderate version of the moral criterion,” which is “that good [interpretive] theories show that the law is, or could be thought to be, supplied by recognisably moral principles,”⁸⁵ without also insisting that these “are the best possible moral foundations.”⁸⁶ Notice, however, that even the suboptimal moral merit of the normative propositions has to be established for any valid prescriptions to follow.

V. CONCLUSION

This paper is aimed at enhancing the clarity of theoretical debates in special jurisprudence by providing a taxonomy of the different kinds of theoretical claim that may be made about areas of law. We have proposed a fundamental division of theories into nonnormative and normative and explained the variety of different possible theoretical claims within this division. At the same time, we have shown how these divisions can become blurred in practice because of the indirect, epistemic, relevance of moral considerations to nonnormative theories. Nonetheless, the in-principle distinction is clear, should be acceptable regardless of the school of general jurisprudence one subscribes to, and bears importantly on the assessment of claims being made in special jurisprudence.