Skip to navigation – Site map

HomeNuméros26Revusov kanonConceptual Jurisprudence

Revusov kanon

Conceptual Jurisprudence

An Introduction to Conceptual Analysis and Methodology in Legal Theory
Kenneth Einar Himma
p. 65–92
Konceptualna jurisprudencija [bs]


This essay attempts to provide an accessible introduction to the topic area of conceptual analysis of legal concepts (or “conceptual jurisprudence”) and its methodology. I attempt to explain, at a fairly foundational level, what conceptual analysis is, how it is done (i.e. its appropriate methodology) and why it is important in theorizing about the law. I also attempt to explain how conceptual analysis is related to other areas in philosophy, such as metaphysics and epistemology. Next, I explain the enterprise of conceptual jurisprudence, as concerned to provide an account of those properties that (1) distinguish things that are law from things that are not law which (2) constitute the former things as law, illustrating this explanation with what I hope are intuitive examples. Three different methodological approaches are also explained and evaluated. Finally, the practical importance of conceptual jurisprudence is discussed.

Top of page

Full text

1 Introduction

1The defining project of general or conceptual jurisprudence—i.e., the conceptual analysis of law—is to provide philosophically rigorous explications of various concepts that figure prominently in discourse about law. That is, conceptual jurisprudence is concerned with giving an explication of the nature of law and other important legal practices. While many words, such as “chess,” do not pick out concepts that seem important enough to merit a deep philosophical analysis, this is not true of the concept of law. The practices characterized as “legal” or involving “law” are of great moral and prudential significance; persons are incarcerated, fined, required to pay compensation for injuries caused by legally culpable behavior, and even executed for violating the law. It is a matter of clear practical importance that we ensure our practices satisfy rigorous norms of political morality, and we cannot understand what norms apply without an adequate understanding of the concept of law.

2 This essay seeks to provide an accessible, but fairly comprehensive, introduction to conceptual theorizing about law. To this end, the essay will discuss the nature and methodology of conceptual analysis, the relationships of conceptual analysis to metaphysics and to epistemology, and the practical significance of conceptual analysis.

2 Types of Theorizing about Law

3Theorizing about law covers a range of topics. Most theorizing in law probably seeks to ascertain the content of relevant law on issues where the reasons fall fairly evenly on both sides; such articles strive for results that can be used by lawyers and judges to guide their deliberations. Some law review articles are concerned with defending a claim about what the law should, for non-moral practical or moral reasons, be on a particular issue. Others are concerned with justifying certain areas of law; with identifying the foundational principles that purport to determine the content of more specific norms; or with clarifying the meaning of certain terms. Yet others are concerned with explaining the nature of certain legal practices.

4Despite this wide variety of interests on matters related to law, theorizing about law can usefully be divided into three rough categories. Empirical legal theory is usually concerned with identifying or explaining certain features or properties of existing legal systems; such theory is, at least, descriptive in character and focuses on contingent properties (i.e. properties that something does, in fact, possess but might not have possessed it) of the legal systems under study. An empirical legal theorist, for example, might be concerned with identifying or explaining the content of legal norms that purport to govern information privacy in the U.S. Similarly, she might be concerned with explaining the function that some set of legal practices in Colombia purports to serve.

5In contrast, normative legal theory is largely concerned with determining the properties that legal norms or institutions, as a matter of political morality, ought to have or, otherwise put, must have to be morally legitimate. A normative legal theorist, for example, might argue that law, as a matter of substantive moral theory, should protect information privacy in a number of specified ways. Normative legal theorists are typically concerned with issues of moral legitimacy—what restrictions on citizen behavior may, as a matter of political morality, justifiably be enforced by the police power of the state.

6General, or conceptual, jurisprudence is concerned with giving what is called a “conceptual analysis” of core legal concepts; that is, conceptual jurisprudence is concerned with explicating the core concepts of our legal practices, including the interrelations among them. In particular, conceptual jurisprudence seeks to explicate the concepts of law, validity, and legal system and thereby seeks to clarify the logical relationships between these concepts and other concepts potentially related to them, such as the concepts of morality, authority, legal and social obligation, etc. A theory in conceptual jurisprudence will explicate the content of each concept and locate them among a general conceptual framework that guides both our linguistic practices regarding the relevant concept-words and our legal practices themselves.

7This essay will be concerned with explaining conceptual jurisprudence/analysis and the various methodologies that have been proposed as identifying the principles that are most likely to result in a successful piece of conceptual analysis. In addition, this essay will provide a cursory evaluation of the various methodologies proposed for conceptual analysis.

3 Understanding Conceptual Analysis

3.1 What is a Concept, Anyway?

8Conceptual analysis is the analysis of concepts; and this raises a difficult question: what, exactly, is a concept? The notion of a concept is somewhat mysterious and not well understood, but it seems clear, at the very least, that concepts are, or correspond with, mental elements needed to think about things of the relevant sort. For example, it is not possible to think, much less talk, about electrons without having or grasping the concept of an electron.

  • 1 Some caution is required here. It is not exactly clear what is the thing of which the concept is a (...)

9Although this claim much might seem uncontroversial, it tells us less about what a concept is than might initially appear. It doesn’t tell us, for example, anything about the nature of a concept; the way in which concepts are acquired; or the extent to which they are shared among members of a community. But it is clear that the possession or apprehension of a concept is a necessary condition for being able to think or talk about the thing of which the concept is the concept of that thing.1

10There are different views about what concepts are: (1) concepts are psychological states representing ideas or things; (2) concepts are abilities of a special kind—namely, the ability to discriminate one kind of thing of which a concept is a concept from another kind of thing of which another concept is a concept (having the concepts of tree and bush simply is a matter of being able to distinguish trees from bushes); and (3) concepts are meanings or “senses” of words.

11 Obviously, the methodology of conceptual analysis must be responsive to what a concept is. For example, if concepts are representational mental states, a proper analysis of any concept would require saying something about the content of the representation as well as about the nature of the relevant state (which might be a compound state involving beliefs and dispositions of various kinds).

12 Regardless of what concepts turn out to be, they are intimately associated with language. People use language to express concepts. We may not have a word associated with every concept we have. But we have a large number of words to express, refer to, convey, or pick out concepts: we use “love” to think or talk about love; “law” to think or talk about law, and so on. We use a concept-term that picks out a particular concept to talk about things that fall under the concept.

  • 2 I am agnostic with respect to how these practices arise or should be characterized beyond claiming (...)

13 Our ability to talk to each other about something a concept is a concept of (say, law) says something about the relationship between the content of a concept and our use of the associated concept-term. It is reasonable to think that we have some pre-linguistic capacity to develop concepts and develop some concepts long before we learn the associated word; without such a capacity, we could not learn a language. But the contents of our concepts are highly responsive to the linguistic practices involving the associated words because we cannot share an understanding of how to use a word without sharing a grasp of the core content of a concept.2 If you and I have radically different concepts of law, we will not be able to understand what each other is saying when using the concept-term “law.”

14The claim that concepts are associated with linguistic practices does not imply that concepts are linguistic entities; in particular, it does not imply that concepts are meanings. There is nothing in the claim that the content of our concepts is shaped by the content of linguistic practices that would entail that concepts are meanings, rather than mental representations or abilities. My concept of water might simply be an ability to distinguish things that are water from things that are not, even if that ability is responsive to how people around me use the term “water.”

3.2 Traditional Conceptual Analysis as Explicating Fregean Senses

15The traditional methodology for doing conceptual analysis presupposes that concepts are grounded in meanings or Fregean senses. Conceptual analysis attempts to provide a story, grounded in ordinary intuitions about the Fregean sense of the concept-term, about something that falls under the relevant concept. This story, grounded as it is in views about the meanings of the terms, purports to describe the very nature of the thing the concept is a concept of.

16It would be helpful to consider a comparatively unproblematic concept to illustrate how conceptual analysis has traditionally sought to identify the nature of a kind of thing through a philosophical explication of the meaning of the relevant concept-term. Consider the concept of a bachelor. Putting aside some minor concerns about the ordinary meaning of bachelor, it is reasonable to think that, for any X, X is a bachelor if and only if X is an unmarried adult male. If correct, this analysis of the concept provides a list of properties that not only distinguishes bachelors from non-bachelors but also explains why something that is a bachelor falls under that category. That is to say, the analysis of the concept identifies those properties that explain why something is a bachelor in the following sense: the instantiation of the properties of being unmarried, adult, and male constitute anything that instantiates them as a bachelor. Insofar as being unmarried is a conceptual feature of being a bachelor, given the meaning of the concept-term “bachelor,” being unmarried is part of the very nature of being a bachelor.

17Here it should be noted that the idea here is not that instantiation of these properties causes something that instantiates them to be a bachelor; whatever it explains why something is a bachelor will also explain why something is unmarried, adult, and male – and the latter explanation will differ from one person to another as the causal explanation will make reference to contingent properties and factors. “What makes Jim a bachelor?” is a very different question from “Why is Jim a bachelor?”; at a minimum, answering the latter might require knowledge of personal information about Jim, while answering the former surely does not.

18Constitution is a different notion than causation. Being a floating mass of water vapor constitutes something as a cloud; it does not cause something to be a cloud. It is the possession of these properties that determine a floating mass of water vapor’s status as a cloud; it is a cloud in virtue of being a floating mass of water vapor. Constitution is not an event that occurs in time, and this is, in part, what distinguishes constitution from causation, which is an event that occurs in time. Why a particular collection of water molecules is a cloud or a floating mass of water vapor requires a different kind of explanation and analysis. It requires an explanation that is largely empirical in nature, requiring reference to laws of nature and events occurring in time that express the causal antecedents and causal results that go into an explanation of why this particular set of water molecules was transformed into that particular floating mass of water vapor that is the cloud of interest. Conceptual analysis is concerned not with causal explanations, but with the properties that constitute something as falling within the reference of a concept-term, such as “bachelor,” “cloud,” or “law.”

19It is sometimes thought that traditional conceptual analysis (TCA), inasmuch as it seeks to identify meanings through ordinary intuitions about the application of the relevant concept-term, involves little more than providing a dictionary definition. Brian Leiter (2003: 45), for example, argues that TCA is nothing more than “glorified lexicography”:

Conceptual analysis, as Jackson conceives it, becomes hard to distinguish from banal descriptive sociology of the Gallup Poll variety. Indeed, Jackson says explicitly that he advocates, when necessary, “doing serious opinion polls on people’s responses to various cases”! But this now seems to blur the line between conceptual analysis and lexicography: for does not lexicography aim to track statistically normal usage of words or concepts, precisely the pattern of usage a well-designed opinion poll would detect?

20This is a mistake. TCA might start from something that resembles lexicography, as it is grounded in meaning, but it is considerably more than this. TCA goes deeper than just identifying shared views; that, of course, is the job of a lexicographer who records empirical patterns of word-usage. Conceptual analysis attempts to theorize these views by identifying deeper philosophical commitments they imply or presuppose, as well as more general principles that explain them. While this might or might not be a distinctively philosophical enterprise, it goes well beyond the empirical task of identifying shared intuitions or core features of our linguistic practices.

21There is an easy way to see this point: simply compare what lexicographers have to say about the word “law” with what Hart has to say by way of explication of the concept of law. Here is how the lexicographers for The Oxford American Dictionary have defined law:

law | noun 1 (often the law) the system of rules that a particular country or community recognizes as regulating the actions of its members and may enforce by the imposition of penalties: they were taken to court for breaking the law | a license is required by law | [as adj.] law enforcement.
an individual rule as part of such a system: an initiative to tighten up the laws on pornography.
such systems as a subject of study or as the basis of the legal profession: he was still practicing law | [as adj.] a law firm. Compare with jurisprudence.
a thing regarded as having the binding force or effect of a formal system of rules: what he said was law.

22Notice how much that Hart’s theory addresses that is overlooked by the lexical definition. First, there is no mention here of many pieces central to Hart’s analysis: social practices; the rule of recognition; secondary and primary rules; legal validity; etc. Second, the lexicographer’s job is accomplished in a few lines, while Hart took more than 200 pages to give an analysis of the concept of law in The Concept of Law. If Hart starts from the shared views about the meaning of “law,” it should be clear that he is also doing something radically different from what lexicographers are doing – and going much deeper into what law really is, as such.

23 In this connection, it would be helpful to return to the concept of bachelor. According to the lexical definition, a bachelor is an unmarried adult male; the properties of being unmarried, adult, and male exhaust the nature of a bachelor, on this definition. Unfortunately, this analysis of a bachelor leaves important questions open – questions that cannot be answered by a dictionary, because it is the dictionary definition that is indeterminate with respect to these questions. For example, it is not clear whether the Pope is a bachelor. Many people think, when confronted with the issue, that the Pope is not a bachelor, despite being unmarried, adult, and male. Likewise, many people are reluctant to use the term “bachelor” to describe a gay man who lives in jurisdictions not recognizing marriage equality. This suggests that the dictionary definition must be supplemented by, at least, one additional condition: it is a necessary condition for being a bachelor that one be “eligible” in the appropriate way to get married; the problem with both cases described above is that neither person is eligible in the right way for marriage to be considered a bachelor. Of course, the appropriate analysis of “eligible” is needed, among other things, to complete the analysis, but this example should make it clear that TCA, at the very least, purports to go beyond the surface meanings provided by lexicography.

4 The Relationship between Conceptual Analysis and Epistemology: The Traditional View

4.1 The Traditional View Described

24The traditional view of the epistemology of conceptual analysis is that the methodology for justifying conceptual claims is a priori. While it is true that one cannot identify the core content of the patterns of usage that link words and concepts without empirical experience, no further experience, on this view, is needed to justify true conceptual claims. Once it is observed that people use the word “bachelor” only to refer to unmarried men, no further experience is needed to justify the claim that no woman is a bachelor. This claim is justified as a valid logical deduction from the conceptual claims that only men are bachelors and that no woman is a man.

25Not every purely deductive argument, of course, will be as easy as those described in the last paragraph. This will be obvious to anyone who has read Hart’s The Concept of Law or Raz’s The Authority of Law or to anyone who has done any serious mathematics. It took hundreds of years to find a proof for Fermat’s Last Theorem; and the paper proving it was more than one hundred pages long – and something that no layperson could hope to understand.

26Of course, there is more to TCA than simply making inferences from core patterns of linguistic usage. Frank Jackson (1998: 31) describes another methodological element in discussing the concept of free will:

What we are seeking to address is whether free action according to our ordinary conception, or something suitably close to our ordinary conception, exists and is compatible with determinism, and whether intentional states according to our ordinary conception, or something suitably close to it, will survive what cognitive science reveals about the operations of the brain /…/ But how should we identify our ordinary conception? The only possible answer, I think, is by appeal to what seems to us most obvious and central about free action, determinism, belief, or whatever, as revealed by our intuitions about possible cases.

  • 3 There is little disagreement among legal theorists, legal practitioners, or even laypersons on this (...)

27TCA attempts to identify conceptual content that goes beyond the most obvious, and hence “paradigmatic,” features of our shared practices by considering intuitions about possible cases. While the relevant intuitions are ordinary in the sense of being widely shared, they frequently make explicit features of our practices of which ordinary speakers might not be cognizant until they consider such cases. For example, most speakers will probably not realize that our concept of law does not entail that law must conform to morality3 until asked whether the Nazis had a legal system or whether the Jim Crow enactments were laws. Such cases help to flesh out the content of our shared practices in more specificity, identifying features of the concept that will have to be explained theoretically.

  • 4 There are two kinds of norm that might figure into analyzing a concept: epistemic norms, like that (...)
  • 5 Many theorists, including Ronald Dworkin, believe it is not possible to give a purely descriptive a (...)

28 On the traditional conception, the relevant intuitions are purely descriptive and do not include moral intuitions about right or wrong, good or bad. Identifying the content of a concept, on this view, is not governed by moral principles,4 even when the concepts are used to evaluate behavior. For example, it is irrelevant in evaluating two proposed analyses of the concept of law whether one analysis is morally better than another—though moral standards obviously should play a role in deciding what norms should be enacted into law. TCA is purely descriptive and purely general in the sense that these intuitions range over all possible cases.5

  • 6 As Raz (1994: 216–217) has put the point, conceptual analysis is concerned with our concepts, as th (...)

29 What if people disagree on the relevant intuitions? There are a couple of different ways to reconcile such disagreement with a “modest” approach to traditional conceptual analysis that understands conceptual analysis as concerned with identifying the meaning of a concept-term as used by a particular community of speakers.6 First, one could take the position, discussed in more detail below in the next subsection, that such disagreement signals a deeper disagreement that indicates membership in two communities that do not share the same concept. A modest approach seems to presuppose that conceptual analysis is appropriate only within a community of speakers that share the relevant intuitions and social practices. Second, one could take the position that intuitive disagreement can take place within a community of speakers sharing the same concept but that such disagreement signals that our understanding of the concept is indeterminate with respect to the pertinent issues.

30 Jackson distinguishes an “immodest” conception of conceptual analysis from the modest conception, claiming that TCA should be thought of as modest. According to the immodest conception, conceptual analysis gives us insight into what the world is like; that is, an analysis of the content of our concept of law, for example, would give us insight into the essential nature of law as it really is independent of our linguistic practices and conceptual frameworks. According to the modest conception, conceptual analysis merely “tells us what to say in less fundamental terms given an account of the world stated in more fundamental terms” (Jackson 1998: 44)—the fundamental terms being terms defined by the practices of the relevant community of speakers.

  • 7 Beaney 1996; see also Blackburn 1996.

31The talk of moving from more to less fundamental terms reflects the commonsense view of analysis as attempting to resolve and explain what is complex in more simple terms. The idea is that when we analyze a concept, we break it down into simpler constituents so as to display its logical structure and give an explanation of its content that transcends, but incorporates, its ordinary lexical meaning—an idea that is quite common historically. Although theorists have frequently supplemented and modified the analysis of “analysis,” the basic elements of that concept have nonetheless remained largely stable.7

32But the important point here is that modest conceptual analysis presupposes an account of the world “stated in more fundamental terms” as its starting point. This account, however, must be someone’s account; and, as we have seen, it is a shared account grounded in shared intuitions and is, thus, our account of the relevant thing of which the concept is a concept. TCA thus presupposes a certain story about things—one that is our story—and does not purport to describe the world as it is independently of our stories about things.

33Jackson understates the role of linguistic practices in determining the content of our concepts. While it might be true that all possible concepts exist in logical space along with every other abstract object, the content of our concepts are at least partly fixed by our linguistic practices. Indeed, as critics of TCA concede, if we used the symbol “water” to refer to something other than the clear liquid to which it is used to refer, it would not be a conceptual truth that water is H2O—though there would undoubtedly be some conceptual truth corresponding to the relationship between some term and H2O. Concepts might be abstract objects independent of our social activities about which there are utterly mind-independent objective truths, but language is a social construct grounded in a social practice; and which concepts our words pick out or express is partly defined by the content of those practices.

  • 8 As Hart (1994: v) put it: “Many important distinctions, which are not immediately obvious, between (...)

34This suggests that explanations of the traditional methodology for descriptive conceptual analysis are not entirely accurate. Although philosophers frequently justify conceptual claims by an appeal to “ordinary intuitions,” the relevant intuitions are shared views reflecting the core practices for using the terms. If, as Jackson (1998: 33) maintains, “[t]he business of consulting intuitions about possible cases is simply part of the overall business of elucidating concepts by determining how subjects classify possibilities,” then shared views about language will condition the relevant intuitions because how subjects classify possibilities depends on the core practices for using the words. We classify things by using words, which in turn, on the traditional view, express, mean, or convey concepts. Analyzing the content of a concept, on this view, requires identifying shared practices concerning the word.8

35Some of the relevant views might be non-linguistic in character and express some widely accepted belief about a mind-independent feature of reality to which we want to refer. But, in such cases, these non-linguistic “intuitions” about the world will likely play an important role in defining core features of our practices regarding a concept-term. For example, part of what will explain why we use the term “water” to refer to only clear liquids with the chemical structure of H2O is the scientific discovery that water is H2O; this non-linguistic consideration helps to explain why we have converged on using “water” to apply only to H2O. The content of natural-kind concepts frequently changes in response to scientific discoveries because scientific discoveries can frequently result in changes in the way we use the relevant concept-terms. Conventions for using a word, or the definition of a word, can, and frequently does, change over time.

36But the relevant views are usually informed by our core linguistic practices regarding the use of the associated term. I have, for example, the intuition that all bachelors are adult males because, as an empirical matter, no one uses the term “bachelor” to refer to boys or women, and not because I have some non-linguistic intuition about the nature of bachelors. Our shared practices converge on calling only men “bachelors”—and my intuition that only men are bachelors is informed by and expresses those shared linguistic practices.

  • 9 Schauer (2006: 860–861) makes exactly this point: “Like his philosophical colleagues at the time, [ (...)

37TCA, then, has an undeniably empirical element.9 What views are shared among a community is a contingent matter that cannot be determined without empirical observation. While it is true that one reliable way to do the empirical work, as Jackson suggests, is to take an opinion poll, it is not usually necessary. Armchair sociology of the sort at which philosophers excel (usually without realizing that they are doing sociology) is sufficiently reliable as long as one belongs to the linguistic community under consideration.

4.2 Two Methodological Challenges to the Traditional View: Normative and Naturalized Jurisprudence

38The views described in the last subsection are no longer undisputed among philosophers. First, the very notion that a priori knowledge is possible has become controversial among epistemologists embracing a reductive empiricism. Second, and more to the point for our purposes, the underlying assumptions of the traditional view of conceptual analysis have been challenged by certain methodological disputes among theorists in conceptual jurisprudence, regarding the proper methodology for evaluating conceptual claims—in particular, conceptual claims regarding law.

39There are three different methodologies for engaging in conceptual analysis. TCA, as we have seen, adopts a descriptive methodology in the sense that conceptual claims are justified by only factually descriptive claims about the content of the relevant social practices (which typically include linguistic practices); TCA is also guided by epistemic norms of good theory construction and reasoning, such as norms requiring consistency, coherence, consilience, etc., but moral norms are irrelevant in assessing or building an analysis of the content of a concept, under TCA. TCA begins from our ordinary social practices and attempts to make inferences about the deeper implications of those practices forming the starting point for TCA in a way that conforms to the epistemic norms mentioned above.

40Although TCA is “descriptive” in the sense that conceptual claims must be justified by recourse to certain social practices and might, thus, seem to embrace an empirical epistemology, the content of the relevant social practices are grounded in the presumably shared core meanings of the relevant concept-term by the appropriate community in the appropriate practices – which form the ordinary intuitions described in the preceding subsection’s discussion of the traditional views above. From that foundation, further claims fleshing out the content of the concept beyond its core meanings in the appropriate community of speakers are thought justified a priori, if justified at all, according to proponents of TCA. TCA, thus, assumes explicitly that conceptual claims are a priori in character.

41TCA also assumes that conceptual claims are analytic in character in the sense that their truth-value depends only on the content of the meanings of the relevant words. The distinguishing feature analytic claims has to do with the truth-makers for such claims—i.e. what factors determine whether the claim is true or false: intuitively expressed, the idea is that a claim is analytic if and only if its truth-value is determined entirely by the meanings of the relevant terms. TCA attempts to dig out the deeper conceptual commitments that are implied by certain paradigmatically analytic statements, such as, for example, a law is a norm. Whether all the logical implications of an analytic statement are also analytic is interesting issue, as it depends on whether one sees the notion of analyticity as being only semantic in character or whether one sees analyticity as incorporating some vague epistemic notion of being self-evident or easily seen to be analytic. Either way, as will be discussed below, the claims that flesh out the content of a concept are, under TCA, thought to be necessarily true – regardless of whether all the relevant implications of an analytic claim are themselves analytic.

42There have been two influential challenges to TCA. First, W.V.O. Quine argues that the notion of analyticity on which TCA depends cannot be given an explication that would be sufficiently rigorous to ground the methodology associated with TCA. On Quine’s view, the problem with defining analyticity in terms of being true or false wholly in virtue of the meaning of terms is that the notion of “meaning” is too “obscure.” But any other candidates for explicating analyticity fail because they are as much in need of clarification as the notion of analyticity itself. Consider, for example, the idea that analyticity can be explained in terms of synonymy: the claim would be that a claim is analytic if and only if the subject term is synonymous with the predicate term. The problem, according to Quine, is that the notion of synonymy is as much in need of philosophical explication as the notion of analytic and cannot count as an adequate explication of the latter. Insofar as all possible explanations of analyticity share this defect, all such accounts are “circular” and hence illegitimately used as the ground for thinking philosophy has a distinctive methodology exemplified by TCA.

  • 10 See, e.g., Grice & Strawson 1956.

43While quite influential in legal philosophy, these arguments are no longer as influential in other areas – most likely, because they are vulnerable to many plausible criticisms. Indeed, in some respects, the arguments seem straightforwardly problematic. If the claim that a concept is obscure were, by itself, enough to justify disqualifying it from appearing in philosophical discourse, one would have to disqualify such concepts as number and quarks, both of which are very difficult to explicate in a clear intuitive way. Further, if the unavailability of a non-circular explication of a concept were sufficient to warrant disqualifying it from philosophical discourse, many concepts of philosophical importance would be disqualified: there are some mathematical notions (such as the notion of a set, which is defined in terms of obvious synonyms: ‘group’ and ‘collection’) and moral notions (such as the notion of good) for which no non-circular explication can be given.10

44A second challenge to TCA claims that the methodology’s reliance on intuitions is problematic because intuitions differ from culture to culture and hence are too unreliable to ground a philosophical explication of anything. But there is nothing either surprising or problematic about this. Conceptual frameworks are defined by language and language is a social artifact that differs from one culture to the next; insofar as a word does not translate perfectly in another language, intuitions about the relevant concept-terms will not line up. That, however, is not a problem once the truism is understood that TCA seeks to explicate our concepts – concepts that are shared within a particular community, an assumption that should, as we have seen, be attributed to Jackson’s modest approach to TCA.

45Despite the questions concerning the merits of these arguments, they have nonetheless been extremely influential in debates about the methodology of conceptual jurisprudence. In particular, these arguments have led to a call for a “naturalized jurisprudence.” Leiter, the most prominent proponent of naturalized jurisprudence, argues that TCA should be rejected on the grounds of these arguments in favor of a methodology that is continuous with scientific methodology and hence empirical in character. Such a methodology must reject all the notions that Quine rejects in “Two Dogmas,” including modal notions, such as necessity and possibility. This implies the rejection of the idea that true conceptual claims are necessarily true and can come to be known a priori. Insofar as philosophical methodology should be regarded as “continuous with science” and hence as empirical in nature, conceptual claims are contingently true, if true at all.

46Apart from the problems associated with the arguments thought to ground a naturalized jurisprudence, there are further worries about naturalized jurisprudence. No one has articulated a reasonably clear statement of how a naturalized jurisprudence should proceed with respect to analyzing concepts. There are many questions about law a naturalized methodology can answer—indeed, any empirical question about the law that has a determinate and determinable answer. But how a naturalized methodology could be deployed in the service of conceptual analysis is far from clear—certainly, no one has been able to explicate it with the rigor and elegance that Frank Jackson explicates the methodology of TCA. Indeed, for this reason, it is not unreasonable to think naturalized jurisprudence is nothing more than a skeptical theory of conceptual jurisprudence, asserting, in effect, that conceptual jurisprudence is impossible—just as Quine believed metaphysics was rendered impossible by his rejection of the modalities. This does not seem to be a “replacement” methodology in the sense that we simply substitute naturalized methodology for TCA and keep on doing what is the same thing. Naturalized methodology so transforms the nature of a conceptual inquiry that it is no longer clear that the relevant inquiry supported by this methodology ought to be characterized as a “conceptual” inquiry.

  • 11 Bourget & Chalmers 2014.

47Of course, one can always accept these implications, as well as the other implications of Quine’s analysis for philosophical theorizing about numbers, moral notions, and other notions that cannot be given a non-circular explication; there is nothing resembling an obviously decisive counterargument given above. But there nonetheless persists the strong intuition that there is something valuable about TCA, conceived as resting on analytic claims and expressing necessary truths about the relevant thing. Indeed, traditional conceptual analysis, metaphysics, and normative theorizing in morality and epistemology continues to be prominent in philosophy, despite the Quinean challenge. Indeed, according to a recent study, 64.9% of philosophers accept the analytic-synthetic distinction, while only 27.1% reject it.11

  • 12 For a helpful discussion of this motivation for normative jurisprudence, see Dickson 2001: Ch. 1, S (...)

48A second methodological challenge to the ascendancy of TCA is normative jurisprudence. Normative jurisprudence is typically grounded in a rejection of the view that giving an account of what law is (as such) is a different enterprise than giving an account of what law ought to be, construed to imply that conceptual methodology should be purely descriptive.12 Intriguingly, Hart (1994: 211) seems to suggest, contrary to nearly everything else he said on methodology, that a conceptual theory of law is sometimes properly evaluated on the basis of normative considerations that go beyond the norms governing sound reasoning and argument:

A concept of law which allows the invalidity of law to be distinguished from its immorality, enables us to see the variety and complexity of these separate issues; whereas a narrow concept of law which denies legal validity to iniquitous rules may blind us to them.

49One should not make too much of this argument, as Hart is clear in regarding conceptual methodology as properly descriptive in character; however, it is important to realize that this remark is in tension with Hart’s clear commitments insofar as recourse to any such practical considerations in justifying a conceptual theory is irrelevant on a purely descriptive approach to conceptual methodology. Strictly speaking, recourse to such considerations presupposes the truth of one possible thesis of a normative methodology – namely, that an adequate explication of the concept of law should have some beneficial social consequences and, thus, that the adequacy of a conceptual theory of law depends on the practical social implications of adopting it.

50More commonly, proponents of normative jurisprudence assert that the concept of law cannot be adequately explicated without recourse to moral norms that define what the content or point of law should be. Law as such is, thus, partly defined in terms of the moral value of its point, content, or practices. On this view, as Dickson (2001: 7) describes it, “the task of characterizing law ‘as it is’ is necessarily and inextricably bound up with one’s understanding of how law morally ought to be, such that the two enterprises cannot be separated”.

  • 13 Influential proponents of normative methodology include Perry (2001), and Waldron (2001).

51Each of these claims about the relevance of practical norms (beyond the norms of rationality) has some powerful implications for methodology. Normative methodology resembles TCA in locating the starting point of conceptual analysis in widely shared understandings concerning the relevant social practices and paradigms picked out by the relevant concept-term. Where normative methodology departs from TCA, however, is in rejecting the traditional view that the ordinary understandings of the practices and paradigms define the exclusive touchstone for testing the theory; instead, an accurate analysis of legal concepts must answer also to substantive moral norms that stipulate either what the content of these concepts ought to be or what social consequences a conceptual theory of law ought to have.13

52But why think that one cannot give an account of what law is as such without considering what law morally should be? One class of arguments—perhaps, the most influential—proceeds from the assumption that normative concepts (i.e. concepts with normative content, such as that picked out by “good”) must sometimes be analyzed to conform to some relevant set of moral norms. Since, for example, one might think that law’s conceptual function is to create the conditions of justice, the concept of law is partly normative in content, requiring that an analysis of law harmonize with certain moral norms.

53Ronald Dworkin offers one of the most influential defenses of the idea that the very concept of law is partly normative in character and requires a normative methodology to explicate. On Dworkin’s view, certain concepts are “interpretive” in character and, as such, must be analyzed against the backdrop of some relevant class of substantive norms. Dworkin attempts to show that the concept of law is interpretive and hence warrants a normative methodology by showing an analogy between the concepts of law and courtesy, the latter of which he takes to be paradigmatic of interpretive concepts. Dworkin (1986: 47) begins with a sustained argument for the conclusion that courtesy is an interpretive concept:

Imagine the following history of an invented community. Its members follow a set of rules, which they call “rules of courtesy,” on a certain range of social occasions… For a time this practice has the character of taboo: the rules are just there and are neither questioned nor varied. But, then, perhaps slowly, all this changes. Everyone develops a complex “interpretative” attitude toward the rules of courtesy, an attitude that has two components. The first is the assumption that the practice of courtesy does not simply exist but has value, that it serves some interest or purpose or enforces some principle—in short, that it has some point—that can be stated independently of just describing the rules that make up the practice. The second is the further assumption that the requirements of courtesy – the behavior it calls for or judgments it warrants—are not necessarily or exclusively what they have always been, so that the strict rules must be understood or applied or extended or modified or qualified by that point. Once this interpretive attitude takes hold, the institution of courtesy ceases to be mechanical; it is no longer unstudied deference to a runic order. People now try to impose meaning on the institution—to see it.

54Dworkin goes on to analogize the norms of law to the norms of courtesy to support his view that the concept of law is interpretive—i.e., that the real nature of the concept of law is interpretive and not preinterpretive as the positivist supposes. Thus, he concludes, an analysis of the concept of law must depend on certain moral norms that define its normative purpose.

55 There are a number of potential objections here to Dworkin’s line of argument. First, it is not clear that Dworkin has succeeded in showing that the concept of courtesy has interpretive content that warrants a normative methodology. While it may be true that courtesy seeks, as a conceptual matter, to realize some kind of point or value, it does not follow that the content of the concept of courtesy cannot be explicated without recourse to moral norms—even if the value courtesy seeks is a moral value. The point of courtesy is likely defined by social practices that can be observed as a means of identifying that point. What must, of course, be responsive to that conceptual point of courtesy are the substantive norms of courtesy; knowing the point of courtesy helps us to identify what the norms of courtesy should be. But it, arguably, gets the matter backwards to think that the norms of courtesy must be consulted to identify what the content of the concept is, even beyond identifying the point of the concept. It is not clear how norms of courtesy could even be identified and distinguished from other kinds of norms without having an understanding of the conceptual point of courtesy. Indeed, the norms of courtesy are what they are because, if followed, they enable a subject to conform her behavior in a way that achieves the point of courtesy; the point of courtesy seems to serve as a touchstone for the correctness of the observed norms. If this is correct, then the content of the point determines the content of the norms; the content of the norms does not determine the content of the point, as the objection assumes.

56Second, and more importantly, it is simply not clear that a concept that has normative content can be adequately explicated only through a methodology that incorporates moral norms. The concept of morality, for example, has normative content: the use of the words “moral” and “immoral,” when used in conjunction with a description of an act, express values that provide reasons for action. That X is immoral is a good reason for not doing X. Yet William Frankena provides a plausible analysis of the content of the concept of morality without recourse to a morally normative methodology, relying instead on TCA. As this theory has been described:

Among those who use “morality” normatively, all hold that “morality” refers to a code of conduct that applies to all who can understand it and can govern their behavior by it. In the normative sense, morality should never be overridden, that is, no one should ever violate a moral prohibition or requirement for non-moral considerations (Gert 2011).

57This analysis is gleaned from a purely empirical observation of how people use the relevant concept-terms, which is, of course, simply a matter of teasing out the implications of the shared patterns of ordinary usage among the relevant class of speakers. It is, thus, not clear that one can infer that a normative conceptual methodology is appropriate for concepts with normative content.

58 It is true, of course, that conceptual jurisprudence cannot be done without making certain kinds of value judgment that depend on norms other than those governing good reasoning and argument. One cannot begin an analysis of the concept of law without being able to pick out features of law that are theoretically important. It might be true, for example, that all judges wear robes, but this is not a feature of law that is sufficiently crucial to law that it must figure into an explication of the concept of law. Conceptual jurisprudence must, as Dickson (2001: 51–57) points out, therefore be “indirectly evaluative” in the sense that it requires recourse to norms that enable us to distinguish important common features of law from unimportant common features of law; however, a successful methodological approach to law need not be “directly evaluative” in the sense that it requires recourse to moral norms that state what content the law or relevant social practices should have.

59 As Hart (1987: 39) puts this important point:

  • 14 Others defending this distinction include, e.g., Coleman 2001, Marmor 2001, Waluchow 1994, Dickson (...)

[A]n analysis which allots a place to moral claims and beliefs as constituents of social phenomena must itself be guided, in focusing on those features rather than others, by some criteria of importance of which the chief will be the explanatory power of what his analysis picks out. So his analysis will be guided by judgements, often controversial, of what is important and will therefore reflect such meta-theoretic values and not be neutral between all values. But again there is nothing to show that this analysis is not descriptive but normative and justificatory.14

60Accordingly, Hart distinguishes between an indirectly evaluative approach according to epistemic norms that help pick out features of theoretical importance and a normative approach that seeks to justify some of the various practices and structures that determine the relevant concepts of interest.

5 The Relationship between Conceptual Analysis and Metaphysics

  • 15 It is important to note an important difference here between a modest and an immodest approach to c (...)

61As conceived by TCA, the analysis of a concept discloses the nature or essence of the thing referred to by the corresponding concept-word. Thus, a conceptual analysis of law is intended to tell us something about the nature or essence of law; that is to say, it is supposed to tell us something about not just all existing legal systems, but all conceptually possible legal systems. Thus conceived, a conceptual analysis of law consists in a set of conceptually (or metaphysically) necessary truths and thus constitutes a piece of metaphysical theorizing – just as an analysis of the concept of free will is a piece of metaphysics. As Hart cautioned, conceptual analysis may begin from words, but it is important to realize that an adequate explication of a concept will include truths that go far beyond the conventions that establish the core meanings of those words. The project of analyzing the content of legal concepts is the project of exploring the metaphysics of law, explicating its nature in terms of claims that are necessarily true—the hallmark of metaphysical inquiry.15

62 The metaphysical character of conceptual analysis, traditionally conceived, is logically connected to the view that the epistemology of conceptual analysis is a priori in character. Empirical observation is essential, primarily, for identifying only the contingent sensible properties of particular things. For example, empirical observation is needed to come to know that a bag of rocks contains four rocks, but does not, from the standpoint of ordinary intuition, seem to be needed to know that 2 + 2 = 4; it would be silly to go out and attempt to confirm the truth of this arithmetical proposition by a series of trials in which the sum of putting two rocks together with two more rocks is identified and recorded. We need empirical observation to know that there are two rocks in the bag because that could be false; in contrast, we come to know that 2 + 2 = 4 by grasping, somehow (and the operation of the a priori faculty, if such there be, is quite mysterious from the standpoint of epistemological theory), that it could not be false.

63Insofar as the essential properties of law are possessed by every possible legal system and not by just the existing legal systems we can observe by our senses, empirical observation is irrelevant in conceptual analysis—except as is necessary to understand the concept-term well enough to pick out paradigm instances of the things to which it refers. Accordingly, the methodology for conceptual analysis, thus conceived, would be a priori.

  • 16 An interesting question is whether, assuming the starting point of TCA consists entirely in analyti (...)

64 Similarly, the metaphysical character of conceptual analysis, traditionally conceived, is thought to be logically connected to the view that the deeper content of some concept can be derived from some set of comparatively obvious analytic truths about the thing the concept picks out. If a claim, C, is true wholly in virtue of the meanings of the words used to express it, then the denial of C is a contradiction. Since a contradiction is impossible, C is necessarily true and hence metaphysical in character, as metaphysics deals with issues concerning the nature of things. The content of the concept, then, will be fleshed out by a set of claims that are necessarily true and derived from the analytic truths that form the starting point of for TCA.16 Again, TCA presupposes conceptual analysis is metaphysical in character.

  • 17 Moral objectivism is the view according to which the truth-value of moral claims is determined by m (...)

65 As noted above, there have been two methodological challenges to TCA—normative jurisprudence and naturalized jurisprudence—and they disagree on the issue of whether conceptual analysis is metaphysical in character. Normative jurisprudence, as will be recalled, holds that an analysis of the concept of law must conform to certain relevant moral norms, just as it must conform to certain norms governing rational discourse. While normative jurisprudence does not entail a meta-ethical position about whether morality is objective (i.e., “moral objectivism”), most theorists favoring a normative methodology also hold moral objectivism.17 Insofar as moral objectivism entails, as is commonly held, that moral norms are necessarily true, if true at all, applying a morally normative methodology to conceptual analysis will likewise flesh out the content of the relevant concept in terms of necessary truths that express the nature of the relevant thing. Normative methodology, thus, entails that the character of conceptual truths is metaphysical.

66 What is considerably less clear is whether normative methodology presupposes that conceptual claims are analytic, as TCA appears to. The problem here arises in connection with the character of moral claims. Assuming, as normative jurisprudes typically do, that morality is objective, moral claims might be necessarily true, but not obviously in virtue of being analytic claims; it does not seem true by definition that killing a human being is wrong. One can take the position, of course, that the content of the concept of law is moralized in the sense that the relevant moral claims are built into the definition of law, but this is a deep and difficult issue that cannot be adequately addressed here. Even so, it seems nonetheless clear that normative jurisprudence presupposes that conceptual analysis of law is metaphysical in character.

67 In contrast, naturalized jurisprudence, as is readily evident, denies the most basic assumptions of TCA about the character of conceptual analysis. First, insofar as naturalized jurisprudence is grounded in Quine’s rejection of the analytic-synthetic distinction, it, too, rejects the idea that conceptual analysis is even partly grounded in analytic claims – since, after all, there are no such things that can be legitimately defined. Further, insofar as a naturalized methodology is grounded in Quine’s argument, naturalized jurisprudence follows Quine in rejecting the philosophical legitimacy of the modalities of necessity and possibility. Finally, insofar as naturalized jurisprudence takes the methodology of conceptual analysis to be grounded in empirical claims and continuous with scientific methodology, conceptual claims will be contingent in character; after all, empirical observation is not needed, according to the traditional views, to observe features of all possible worlds. Naturalized jurisprudence, thus, denies that conceptual analysis is a metaphysical enterprise that results in a theory comprised of necessary truths, expressing the nature of the relevant thing under consideration.

6 Conceptual Analysis of Law: Conceptual Jurisprudence

68Fundamental to the conceptual analysis of law, as traditionally conceived, is the metaphysical thesis that, in any conceptually possible legal system, there are certain properties that constitute something as law. Thus, for example, a conceptual theory of law would include a list of properties that constitute anything that has them as a legal system (“law” as institutional system of norms), as well as a list of properties that constitute anything that has them as a legal norm of some legal system (“law” as individual norms). The instantiation of the relevant properties constitutes a system/norm as one of law in exactly the same sense that being unmarried constitutes a man as a bachelor. Any institutional system of norms instantiating the appropriate properties is, for that reason, a legal system; any system that does not is, for that reason, not a legal system. Similarly, any norm instantiating the appropriate properties is, for that reason, a law in some legal system; any norm not instantiating the appropriate properties is, for that reason, not a law in some legal system.

  • 18 P, of course, might express either a very complicated list of properties that include disjunctions (...)
  • 19 As Hart (1994: 103) puts the point, “To say that a given rule is valid is to recognize it as passin (...)

69Conceptual theories of law have some intriguing implications that enable us to make more specific observations about the content of particular laws and legal systems. Each of the major conceptual theories of law – positivism, natural law theory, and Dworkin’s interpretivism – begin from the assumption that law is, in part, manufactured by contingent social practices, including those involved in legislation and adjudication. This means that the content of legal norms, including those that define the contours of the legal system to which those norms belong, can vary from one legal system to the next. One consequence of this idea is that in every conceptually possible legal system there is a set of norms that establish the existence and structure of a legal system and provide criteria that determine when a norm counts as a valid law of the system. If S is a legal system and P is a statement that describes the properties that constitute a norm as law,18 then P states necessary and sufficient criteria of “legal validity” in S in the following sense: for any norm n, n is a law in S at time t if and only if n instantiates P at t.19 Another way of putting this is as follows:

The Differentiation Thesis. In every conceptually possible legal system S, there is a set CoV (for “criteria of validity”) such that, for every norm n, n is a law in S at time t if and only if n satisfies the criteria in CoV at t.

70The Differentiation Thesis, then, asserts that every legal system contain membership conditions defining the criteria of validity for that system – and these criteria are understood to differ from system to system insofar as the relevant social practices differ.

  • 20 The claim that the Differentiation Thesis is not an epistemic thesis is also true on a naturalized (...)

71It is important to realize that, as traditionally conceived, the Differentiation Thesis is a metaphysical thesis – and not an epistemological thesis.20 The Differentiation Thesis neither presupposes nor implies any claims about the extent to which the criteria of validity can be identified or applied in any possible legal system. Indeed, the Differentiation Thesis, by itself, is logically consistent with a variety of claims about how much we can know about the content of the law – both the legal norms establishing validity criteria and the legal norms that are law because valid under the validity criteria. It is true, of course, that (1) the content of all legal norms ultimately depend, in large measure, on social processes that can be empirically observed and (2) the norms of a system must be understandable in order to be sufficiently efficacious to constitute a legal system. It is also true that the Differentiation Thesis, together with both (1) and (2), entail that we can come to know something about the content of the law in a particular legal system by observing enough about the practices that give rise to the criteria of validity and the activities of officials that conform to these criteria. But it is a mistake to think that the core theses of a conceptual theory of law, which includes the Differentiation Thesis, implies that the criteria of validity in any or every legal system constitutes an epistemic test that settles the question of what the law is in every dispute that might arise. That view was once incorrectly attributed to positivism as the so-called Pedigree Thesis during the early years of the debate between Dworkin and positivists, but is improperly attributed to any of the major conceptual theories of law. Whatever views a theorist takes on this will have to be grounded in other commitments.

72Still, it is reasonable to think that statements of the criteria of validity could not provide a decision procedure for identifying the law. As Ronald Dworkin argues so forcefully in Taking Rights Seriously, appellate judicial practice is concerned with adjudicating cases that are hard in the sense that the reasons supporting either of two conflicting holdings are similarly weighty. Such cases turn out to be very difficult for judges to decide. While one might take the position, as Raz does, that hard cases arise because there is a gap in the law on the issue, it is sometimes a tricky matter to determine, even on the assumption Raz is correct, what constitutes a hard case. There are, to put the point modestly, probably no decision procedures to be found in complex legal systems like those of modern municipal states.

73Further, the Differentiation Thesis implies nothing about the nature or moral character of the law in any particular legal system. Legally valid content might – or might not – be necessarily constrained by moral principles or the inherently interpretive character of law. Since the Differentiation Thesis is agnostic with respect to such claims, it is consistent with classical natural law theory, legal positivism, and Dworkin’s constructive interpretation and hence cannot distinguish positivism from other conceptual theories of law. Up to now, every conceptual theorist has assumed it.

74One can doubt that the criteria of law can be fully identified in every legal system, but general jurisprudence could not get off the ground without assuming there is some distinction between law and non-law. That latter distinction is so fundamental to general jurisprudence and so intuitively plausible that one would need a very good reason to justify rejecting it. The Differentiation Thesis simply expresses that legal norms are distinct from other norms in every society with a legal system. The job of every conceptual theory of law—i.e. of conceptual jurisprudence, in general—is to explain what properties distinguishes these norms that are law from norms that are not law and constitute the former as legal norms.

7 Conceptual Analysis: Theoretical and Practical Significance

  • 21 Some legal academicians overestimate the practical consequences of a conceptual theory of law. Some (...)

75One might legitimately wonder whether conceptual jurisprudence has any practical applications in resolving substantive problems concerning what the content of a particular law is, or should be. Many persons in the legal academy have developed something of an aversion, if not hostility, to the highly abstract concerns of conceptual legal theory, believing that an analysis of what law is by nature is as normatively impotent as an analysis of what a bachelor is by nature. At first blush, it would appear that understanding the concept of a bachelor has no implications that would answer practical questions that might arise in connection with how bachelors should act or should be treated. A theoretical definition of “bachelor” just seems, by nature, too abstractly descriptive to yield any interesting practical conclusions.21

  • 22 Posner 1996.

76As it turns out, some prominent legal scholars have argued for precisely this thesis with respect to the enterprise of conceptual jurisprudence. Richard Posner, for example, devotes a substantial portion of the first of his Clarendon Law Lectures, which were given at Oxford in 1995, to arguing that conceptual jurisprudence is “futile, distracting, and illustrative of the impoverishment of traditional legal theory.”22 The following passage is representative of the disdain with which he regards conceptual jurisprudence:

I have nothing against philosophical speculation. But one would like it to have some pay-off; something ought to turn on the answer to the question ‘What is law?’ if the question is to be worth asking by people who could use their time in other socially valuable ways. Nothing does turn on it (Posner 1996: 3).

77Indeed, Posner (1996: 3) goes so far as to argue that “the central task of analytic jurisprudence is, or at least ought to be, not to answer the question ‘What is law?’ but to show that it should not be asked, because it only confuses matters”.

78 As might be evident from the discussion of “bachelor” at the beginning of this section, Posner’s concerns are not utterly counterintuitive, but the challenge seems most plausible directed at purely descriptive approaches to conceptual analysis, such as TCA and naturalized jurisprudence. After all, there is a general problem of deriving normative practical claims from a set consisting of nothing but purely descriptive claims; if purely descriptive claims about what law is are logically independent from normative claims about what law should be, as is presupposed by TCA and naturalized jurisprudence, then the latter claims cannot be deduced, without more, from the former. Nevertheless, it is important to realize that proponents of normative methodology explicitly deny, as we have seen, the claim that what law is and what law should be are logically unrelated issues. Whether this denial is enough to insulate normative approaches from Posner’s criticism is not entirely clear; however, there is enough of a difference between normative and descriptive approaches to warrant some caution in thinking this criticism would sweep across all the methodological approaches.

79In any event, Posner’s view that conceptual jurisprudence should not be done, even if he is correct in thinking that such analysis results in theories that are normatively important, is problematic. To see the problem, it would be helpful to consider how Posner’s criticism would come off in another context. Though a great deal of work in pure mathematics has been used to create technologies that greatly improve our lives, one of the most celebrated mathematical accomplishments of recent years is not thought to have such applications. A few years ago, Andrew Wiles of Princeton University devised what is believed to be a successful proof for Fermat’s Last Theorem, which asserts that there are no positive integers x, y, z, and n > 2 such that the equation xn + yn = zn is true. Despite the fact that no one in the mathematical community necessarily believes this will ever have practical applications that result in the betterment of the human condition, tens of thousands of hours were devoted by mathematicians eager to finding a proof or disproof of this proposition.

80What surely explains this, at least in part, is that we value knowing things for other than instrumental reasons. Sometimes knowing a proposition p is instrumentally valuable in the sense that it that knowledge provides the means to an end of producing a useful technology. Much mathematical theory can obviously be justified by the value of the practical benefits to people, but that is not the only reason to value mathematical inquiry. If knowledge is not always intrinsically valuable (i.e. valuable for its own sake as an end-in-itself), it is sometimes intrinsically valuable. Knowledge of abstract truths defining basic structures of significant practical value (such as mathematics) seems valuable for its own sake; indeed, it is not unreasonable to think that all knowledge is (or should be) intrinsically valuable to a rational being. The same would be true of the concept of law; assuming an analysis of the concept of law has no practical instrumental value, it is plausible to think that understanding the nature of an institution that restricts the freedom of autonomous agent is intrinsically valuable.

81But is Posner correct in thinking that nothing of substance turns on a conceptual analysis of law? Certainly, there is something that can be said for his view. Consider, for example, the dispute between Dworkin and Hart on the issue of judicial discretion. Dworkin once argued that the issue of whether it is fair for judges to assess damages against a defendant in a hard case depends on whether the correct way to characterize the judge’s decision is as implicit in preexisting law or as creating new law; as Dworkin (1967: 31) put the point:

If [judges have discretion to decide hard cases by making new law] we must acknowledge that the murderer’s family in Riggs and the manufacturer in Henningsen were deprived of their property by an act of judicial discretion applied ex post facto. This may not shock many readers—the notion of discretion has percolated throughout the legal community—but it does illustrate one of the most nettlesome of the puzzles that drive philosophers to worry about legal obligation. If taking property away in cases like these cannot be justified by appealing to an established obligation, then another justification must be found, and nothing satisfactory has been supplied.

82Such takings of property are morally illegitimate because it is unfair to take property from a person under a law that did not exist at the time of the behavior that gave rise to the liability—or so the argument goes.

  • 23 Some cases are hard enough that it takes a judge of exceptional (as Dworkin puts it, “Herculean”) a (...)

83 But Dworkin’s theory seems to fare not much better on this count than Hart’s view that judges decide hard cases by making new law. What distinguishes hard cases from other cases, on Dworkin’s view, is that, in hard cases, there is more than one holding that coheres with the existing institutional history. What makes an issue of law hard, then, is that the law (construed to include the existing institutional history) doesn’t give reasonable notice of what, if anything, it requires.23 If, however, fairness precludes taking property from a defendant under a law that did not exist when the behavior giving rise to the liability occurred, it also precludes taking property from a defendant under a law that does not give reasonable notice to the defendant that the behavior at issue gives rise to such a liability. Just as it is unfair to punish a child for breaking a rule that was not communicated to her in terms that she could reasonably be expected to understand, it is unfair to compel a person to pay damages under a law that does not adequately inform the person that the behavior at issue gives rise to liability under the rule. It is not enough that to justify assessing damages against a person that she has a legal obligation; that obligation must be communicated in terms that a competent speaker of the language can understand.

  • 24 Cf. Murphy 2001.

84 These examples are far from being conclusive, of course. But they do provide some reason to think, as Posner does, that we cannot solve substantive problems of law and morality simply by moving concepts around.24 Concept-words are used to group acts, events, and entities and hence do no more than pick out particular classes of acts, events, and entities. It seems clear that we cannot solve any interesting moral problems merely by altering our conceptual characterization of some act, event, or entity. Whether hitting someone in the face is morally wrong cannot turn on whether it is properly characterized as an “assault” or as a “kiss”; if it is properly characterized it as a “kiss,” then we will have to rethink our moral views about the permissibility of kissing. What substantive normative qualities any particular act or event has cannot turn on how it is grouped through our linguistic practices with other acts or events.

85 But Posner fails to realize that this isn’t the only way in which an analysis of legal concepts might make a difference in our lives. Proponents of conceptual analysis who feel the need to justify their view by reference to socially useful consequences typically do so by pointing to its epistemic benefits. In discussing the value of conceptual analysis in the area of metaphysics, Frank Jackson (1998: 30) writes:

Although metaphysics is about what the world is like, the questions we ask when we do metaphysics are framed in a language, and thus we need to attend to what the users of the language mean by the words they employ to ask their questions. When bounty hunters go searching they are searching for a person and not a handbill. But they will not get very far if they fail to attend to the representational properties of the handbill on the wanted person. These properties give them their target, or, if you like, define the subject of their search. Likewise, metaphysicians will not get very far with questions like: Are there Ks? Are Ks nothing over and above Js? and, Is the K way the world is fully determined by the J way the world is? in the absence of some conception of what counts as a K, and what counts as a J.

86The point of conceptual analysis, on Jackson’s influential view, is not that it entails solutions to important substantive problems of law, ethics, or politics. Rather it is that conceptual analysis helps us to solve those problems by helping us to see and formulate them more clearly. As Coleman (2001: 13) puts this absolutely essential point, “Conceptual analysis is essentially a philosophical enterprise: its aim is to help us think more carefully”.

87 Even so, the issue of determining whether and how theories of conceptual jurisprudence are valuable is an important one. It is always helpful to know whether a particular area of theorizing results in any practical benefits to humanity beyond increasing the store of human knowledge. At this point, too little work in this area has been done to offer any general conclusions.

Top of page


Michael BEANEY, 1996: Analysis. Stanford Encyclopedia of Philosophy. Ed. Edward N. Zalta. URL:
Simon BLACKBURN, 1996: Oxford Dictionary of Philosophy. Oxford: Oxford University Press.
David BOURGET & David CHALMERS, 2014: What Do Philosophers Believe? Philosophical Studies 170 (3): 465–500.

Jules COLEMAN, 2001: The Practice of Principle: In Defence of a Pragmatist Approach to Legal Theory. Oxford: Oxford University Press.

Julie DICKSON, 2001: Evaluation and Legal Theory. Oxford: Hard Publishing.
Ronald DWORKIN, 1967: The Model of Rules. The University of Chicago Law Review (1967) 35: 14–46.
——, 1977: Taking Rights Seriously. Cambridge (Ma.): Harvard University Press.
——, 1986: Law’s Empire. Cambridge (Ma.): Harvard University Press.
——, 1996: Objectivity and Truth: You’d Better Believe It. Philosophy & Public Affairs 25 (1996) 2.

John FINNIS, 1980: Natural Law and Natural Rights. Oxford: Oxford University Press.

Bernard GERT, 2011: The Definition of Morality. Stanford Encyclopedia of Philosophy. Ed. Edward N. Zalta. URL:
Herbert Paul GRICE & Peter F. STRAWSON, 1956: In Defense of a Dogma. Philosophical Review 65 (1955) 2.

Herbert L.A. HART, 1987: Comment. Issues in Contemporary Legal Philosophy: The Influence of H.L.A. Hart. Ed. Ruth Gavison. Oxford: Oxford University Press.
——, 1994: The Concept of Law. Revised Ed. Oxford: Oxford University Press.

Frank JACKSON, 1998: From Metaphysics to Ethics: A Defence of Conceptual Analysis. Oxford: Oxford University Press.

Brian LEITER, 2003: Beyond the Hart/Dworkin Debate: The Methodology Problem in Jurisprudence. American Journal of Jurisprudence (2003) 48: 17–51.

Andrei MARMOR, 2001: Positive Law and Objective Values. Oxford: Oxford University Press.
Liam MURPHY, 2001: The Political Question and the Concept of Law. Hart’s Postscript: Essays on the Postscript to The Concept of Law. Ed. Jules Coleman. Oxford: Oxford University Press.

Stephen PERRY, 2001: Hart's Methodological Positivism. Hart’s Postscript: Essays on the Postscript to The Concept of Law. Ed. Jules Coleman. Oxford: Oxford University Press.
Richard POSNER, 1996: Law and Legal Theory. Oxford: Clarendon Press.

Joseph RAZ, 1979: The Authority of Law. Essays on Law and Morality. Oxford: Clarendon Press.
——, 1994: Ethics in the Public Domain. Oxford: Oxford University Press.

Frederick SCHAUER, 2006: (Re)taking Hart. Harvard Law Review 119 (2006).

Jeremy WALDRON, 2001: Normative (or Ethical) Positivism. Hart’s Postscript: Essays on the Postscript to The Concept of Law. Ed. Jules Coleman. Oxford: Oxford University Press.
Wil WALUCHOW, 1994: Inclusive Legal Positivism. Oxford: Oxford University Press.

Top of page


1 Some caution is required here. It is not exactly clear what is the thing of which the concept is a concept. For example, the extensional meaning of the term “legal system” is the class of all things that are properly characterized as being legal systems; accordingly, the extensional meaning, or referent, of the term refers to a set—and not to some general unitary legal system. The intensional meaning, of course, is the idea or content expressed by the term “legal system.” It should be understood that we are speaking somewhat metaphorically when we talk about the thing of which something is a concept. If there is a thing that is law as such or a bachelor as such, it is not clear (at least, not to me) what kind of thing it is, beyond its not being identical with either the extensional or intensional meanings of the relevant concept-term. But while it would be helpful to have a more developed account of those special “things,” we can speak somewhat metaphorically without undesirable consequences as long as we keep in mind that we are speaking a bit loosely. Such talk, however, is quite useful in discussing the various topics with which this essay is concerned. But if one is still a bit uncomfortable, one can think of these things as being any member of the relevant extension provided that we remember that we are interested in that particular member insofar as it falls under the concept.

2 I am agnostic with respect to how these practices arise or should be characterized beyond claiming that (1) they are social practices that arise and are maintained in part because they are shared across a community and that (2) core elements of this practice yield norms that have prescriptive force among members of the relevant community. For example, one might think nothing more specific regarding our practices for using “bachelor” than that they comprise various patterns of usage shared among a community. Alternatively, one might think these practices constitute conventions. Either way, the practices are social in character and express a shared norm that “bachelor” is properly used only of men: it is uncontroversial that someone who uses “bachelor” to talk about a dog has made a mistake.

3 There is little disagreement among legal theorists, legal practitioners, or even laypersons on this issue. Indeed, most natural law theorists in the Thomist tradition deny that Aquinas claimed that unjust norms could not be law.

4 There are two kinds of norm that might figure into analyzing a concept: epistemic norms, like that requiring consistency, and moral norms, including those that govern states. There is no controversy about whether epistemic norms should play a role in theorizing. See Coleman 2001.

5 Many theorists, including Ronald Dworkin, believe it is not possible to give a purely descriptive analysis of evaluative concepts like law. See, e.g., Dworkin 1986.

6 As Raz (1994: 216–217) has put the point, conceptual analysis is concerned with our concepts, as those concepts are constructed by our social practices—“our” referring to members of a particular community of speakers, namely us.

7 Beaney 1996; see also Blackburn 1996.

8 As Hart (1994: v) put it: “Many important distinctions, which are not immediately obvious, between types of social situation or relationships may best be brought to light by an examination of the standard uses of the relevant expressions and of the way in which these depend on a social context, itself often left unstated. In this field of study it is particularly true that we may use, as Professor J.L. Austin said, ‘a sharpened awareness of words to sharpen our perception of the phenomenon’.”

9 Schauer (2006: 860–861) makes exactly this point: “Like his philosophical colleagues at the time, [Hart] examined ordinary usage closely for the distinctions it embodied and rigorously analyzed and defined the terms used to mark legal concepts /…/ Conceptual analysis, as Hart surely must have recognized, inevitably rests at least in part on empirical observation, and this empirical foundation of conceptual analysis could also implicitly undergird his claim to be doing descriptive sociology.”

10 See, e.g., Grice & Strawson 1956.

11 Bourget & Chalmers 2014.

12 For a helpful discussion of this motivation for normative jurisprudence, see Dickson 2001: Ch. 1, Sec. A.

13 Influential proponents of normative methodology include Perry (2001), and Waldron (2001).

14 Others defending this distinction include, e.g., Coleman 2001, Marmor 2001, Waluchow 1994, Dickson 2001, and Finnis 1980.

15 It is important to note an important difference here between a modest and an immodest approach to conceptual analysis. Insofar as a modest approach is explicitly grounded in contingent social practices like those that define the language of the community, conceptual truths are necessary relative to that set of particular commitments; we impose a certain metaphysical structure on the world through conceptual commitments that arise out of social practices we construct. In contrast, insofar as an immodest approach seeks to identify truths that are independent of human practices, conceptual truths would simply be necessarily true – and without regard to any particular conceptual framework that people impose on the world through linguistic and other social practices.

16 An interesting question is whether, assuming the starting point of TCA consists entirely in analytic truths, the implications of a set containing only analytic truths are necessary also analytic in character. The claim that law consists of norms is surely analytic in character, if any statement is. However, the claim that the existence conditions for a legal system include a social rule of recognition that is practiced by officials is not obviously analytic – even if it is, in fact, a necessary truth about legal systems, by nature – and is derived somehow from the claim that law consists of norms.

Indeed, one can legitimately question whether the starting point—even for TCA—consists wholly in analytic truths. The observations that purport to pick out “paradigmatic” features of law and legal systems do not seem to be entailed by definitions of law. Rather, once we have a pre-theoretic understanding of law, which is partly informed by definitions, the identification of paradigms seems, at least, partly informed by value judgments that distinguish important necessary features of law from unimportant, and are, arguably, synthetic necessary truths.

17 Moral objectivism is the view according to which the truth-value of moral claims is determined by mind-independent considerations, like preferences, beliefs, practices, or desires. Every natural law theorist in legal theory is a moral objectivist. Likewise, Ronald Dworkin is a moral objectivist, and one who calls for a normative methodology. See, e.g., Dworkin 1996.

18 P, of course, might express either a very complicated list of properties that include disjunctions or conjunctions. So insofar as law is a cluster concept or family resemblance concept, the statement of the conditions will include a host of disjunctions joining conjunctions, which express the various degrees of resemblance sufficient to confer the status of law on a norm.

19 As Hart (1994: 103) puts the point, “To say that a given rule is valid is to recognize it as passing all the tests provided by the rule of recognition and so as a rule of the system.”

20 The claim that the Differentiation Thesis is not an epistemic thesis is also true on a naturalized methodology.

21 Some legal academicians overestimate the practical consequences of a conceptual theory of law. Some academic lawyers believe, incorrectly, that positivism entails legal formalism, a theory of adjudication that has long been refuted. Since the character of adjudicative practices in any legal system depend, according to positivism, on the content of its rule of recognition and the content of the rule of recognition varies from legal system to legal system, positivism does not entail any particular theory of adjudication. While positivism allows for the possibility of a legal system in which formalist norms govern judicial decision-making, it also allows for the possibility that Dworkinian norms govern judicial decision-making.

22 Posner 1996.

23 Some cases are hard enough that it takes a judge of exceptional (as Dworkin puts it, “Herculean”) ability to discern the correct answer to the question.

24 Cf. Murphy 2001.

Top of page


Bibliographical reference

Kenneth Einar Himma, Conceptual JurisprudenceRevus, 26 | 2015, 65–92.

Electronic reference

Kenneth Einar Himma, Conceptual JurisprudenceRevus [Online], 26 | 2015, Online since 03 November 2015, connection on 24 May 2022. URL:; DOI:

Top of page

About the author

Kenneth Einar Himma

Visiting Professor, Department of History of Philosophy, Tomsk State University. 
Part-time Lecturer, University of Washington School of Law

 Kenneth Einar Himma
University of Washington School of Law
William Gates Hall
Box 353020
Seattle, WA 98195-3020
United States


By this author

Top of page


All rights reserved

Top of page
  • OpenEdition Journals
Search OpenEdition Search

You will be redirected to OpenEdition Search