1In this essay, I argue for a view that inclusive positivists share with Ronald Dworkin and strong natural lawyers:
- 1 I will also refer to criteria of legality as “criteria of law.” The criteria of legality were once (...)
- 2 Dworkin and natural lawyers, of course, make a stronger claim about the relationship between the gr (...)
The Moral Incorporation Thesis (MIT): It is logically possible for a legal system to incorporate moral criteria of legality1 (or “grounds of law,” as Dworkin puts it).2
- 3 A set is countably infinite if and only if its members can be placed in one-to-one correspondence w (...)
2It is worth noting here that MIT makes a very weak claim. MIT does not even purport to say anything about actual (or existing) legal systems. First, MIT does not assert or imply that there are, have been, or ever will be legal systems in the actual world (i.e., the logically possible world that we inhabit, which is one among an uncountably infinite number of logically possible worlds3). Second, MIT does not assert or imply anything about the probability that such systems exist – beyond the claim that this probability is non-zero. It claims no more than this: among the uncountably infinite logically possible worlds, there is at least one logically possible world in which there is something that counts as a “legal system” with moral grounds of law. That is, MIT asserts what I will call a “possibility claim.”
- 4 See, below, for a discussion of why Raz’s argument is unsuccessful, p. 99.
3I begin with an explanation of the logic of establishing possibility claims, such as MIT. At the outset, it is worth noting that the logic of establishing possibility claims is very different from the logic of establishing contingent descriptive claims or necessary claims. For this reason, some explication of the relevant features of the semantics of modal logic will be necessary here. Once the structural framework is adequately developed, the argument for MIT will be grounded on the strength of a thought experiment of a surprisingly simple kind. Indeed, the argument is inspired by a Razian argument for the possibility of a legal system without coercive enforcement machinery; on his view, a society of angels could still have a system of law without any coercive machinery. My argument will possess two theoretically important qualities that are also possessed by Raz’s powerfully simple, but ultimately unsuccessful, argument.4
4There are a number of different types of legal theory. 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 of the legal systems under study. An empirical legal theory, 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, such a theory might be concerned with explaining the function that some set of legal practices in Colombia purports to serve. In contrast, normative legal theory is largely concerned with determining the properties that legal norms or institutions 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 or that only certain protections of information privacy are justifiably enforced by the police power of the state.
5Conceptual theories of law, however, attempt to address an underlying foundational issue taken for granted by normative and descriptive theories – namely, identifying the nature of law as such – and hence attempt to answer the question “What is law?” Conceptual theories attempt to identify those features and properties that constitute the nature of law as such, according to our concept of law, and hence distinguish entities properly characterized as “law” from entities not properly characterized as “law.” These theories usually express or imply conceptually necessary conditions on what counts as law and generally conform to the following schema:
Necessarily, in any society with a legal system S, there is a set of conditions c1, c2, …, cj such that, for any norm N, N is a law in a society S if and only if N satisfies the conditions c1, c2, …, cj.
6The conditions c1, …, cj (where j represents an arbitrary natural number) have variously been called by positivists and natural law theorists “the criteria of legality,” “the criteria of legal validity,” or “the criteria of validity.” Ronald Dworkin refers to them as the “grounds of law.” I shall use the terms interchangeably, the assumption being that whatever differences there are between them are of no significance with respect to the argument of this essay.
7There are generally three major problems connected with this foundational problem. First, one must explain whether the grounds of law necessarily have the status of law; positivism, for example, holds that the grounds of law are law but not legally valid. Second, if so, one must explain why the grounds of law have the status of law; a positivist states that a conventional rule of recognition that defines the grounds of law or criteria of legal validity has the status of law in virtue of being practiced by those who serve as officials in the legal system. Third, one must explain the existence conditions for a legal system; positivism holds that a legal system exists when there is a conventional rule of recognition practiced by officials and where citizens generally obey the laws validated by the rule of recognition.
8Some of the most important disputes in conceptual jurisprudence involve the issue of whether there is a conceptual relationship between the grounds of law and moral principles – a dispute that started many years ago between the positivist Jeremy Bentham and the classical natural law theorist Thomas Aquinas and continues today among neo-natural law theories, Dworkin’s constructivism, inclusive positivism, and exclusive positivism. Indeed, legal positivism arose in response to the natural law view that there could not be unjust laws because there are necessary moral criteria of legal validity – necessary in the sense that they apply to all possible legal systems, constraining the content of law in each. Beginning with John Austin and Jeremy Bentham, legal positivists denied this strong thesis, adopting the Separability Thesis as part of its foundation: according to the Separability Thesis, there are no conceptually necessary moral grounds of law (or criteria of legality). On the positivist view, law and legal systems are social artifacts manufactured by people – and the artifactual quality of the institution and norms extends all the way down, so to speak, to the grounds of law. As H.L.A. Hart puts the thesis, “it is in no sense a necessary truth that laws reproduce or satisfy certain demands of morality, though in fact they have often done so.”5
9While disagreement continues on the issue of whether the grounds of law necessarily include moral principles, the focus of much recent discussion in conceptual jurisprudence on theoretically salient relationships between law and morality has been on whether it is possible for the grounds of law to include moral principles. The roots of this debate, which became a central concern among legal positivists during the last fifteen years, can be found in Ronald Dworkin’s influential early criticisms of legal positivism. In the mid- to late-70s, Dworkin argued that legal positivism lacked the resources to explain the role that moral standards and reasoning play in judicial reasoning and decision-making. As he (somewhat misleadingly) puts the point in “The Model of Rules II”:
- 6 Dworkin (1977: 40–41). I say “misleading” here because the suggestion is that these moral principle (...)
The origin of these as legal principles lies not in a particular decision of some legislature or court, but in a sense of appropriateness developed in the profession and the public over time. Their continued power depends upon this sense of appropriateness being sustained. If it no longer seemed unfair to allow people to profit by their wrongs, or fair to place special burdens upon oligopolies that manufacture potentially dangerous machines, these principles would no longer play much of a role in new cases, even if they had never been overruled or repealed.6
10According to Dworkin, lawyers and judges routinely ground their arguments in moral principles that have the status of law not because they have been formally or officially promulgated but rather because of the moral content of these principles.
11Positivists became divided over how to respond to Dworkin’s argument. Exclusive positivists, like Joseph Raz, Scott Shapiro, Andrei Marmor, and Brian Leiter, rejected the Dworkinian claim that these moral principles had the status of law in virtue of content. Exclusive positivists thus adopted and defended the Sources Thesis, which denies there can be legal systems with moral criteria of legal validity. Inclusive positivists, like Jules Coleman, Matthew Kramer, Wilfrid Waluchow, and myself, hold what will be called the Moral Incorporation Thesis, which I reiterate in slightly different language as follows:
Moral Incorporation Thesis (MIT): There is a logically possible legal system in which the grounds/criteria of law (or criteria of legal validity) include some moral norms.
12Although this language is typically used by inclusive positivists, it is held by a number of other theorists: strong natural law theorists, such as Aquinas and Blackstone (if it is a logically necessary truth that the ground/criteria of law include moral norms, MIT logically follows), some neo-natural law theorists, and Ronald Dworkin hold this view. It is true that each of these theorists holds a somewhat stronger claim about the conceptual relation between morality and the grounds/criteria of law; but the weaker claim expressed by MIT is logically implied by these stronger views.
13Further contributing to the importance of the debate was the emergence of positions that fall between legal positivism and the strongest form of natural law theory, which is traditionally interpreted as denying the separability thesis. These important new positions include Dworkin’s mature “third theory of law” and neo-natural law positions like Mark Murphy’s view that, as a matter of logical necessity, laws with morally problematic content are defective qua law. The development of these competing views has fueled the continuing debate over whether MIT is true.
14Of course, MIT is a conceptual claim – and a very weak one at that. The claim is not about what is necessary or even actual, which would make it a partly empirical issue; it is about what is possible. The claim is just that it is logically possible for a legal system to have moral grounds of law; otherwise put, the claim is that our legal concepts are consistent with there being moral grounds of validity – a claim that is the target of, for example, Raz’s authority argument, which denies that moral grounds of law are logically compatible with the concept of legal authority. Claims about what is merely logically, metaphysically, or conceptually possible present special difficulties in defending them. In order to make an affirmative case for this kind of claim, one has to show that one can coherently conceive of a legal system with moral grounds of law. This involves showing that the existence of a legal system with moral grounds of law is neither self-contradictory nor contradicts other core doctrines of a theory of law. Possibility or coherence claims like this are difficult to support by positive argument. In consequence, the debate has taken place with the proponent of the view that moral grounds of law are impossible giving an argument to show that the existence of such grounds is logically inconsistent with some other plausible conceptual truth about law.
15In this respect, the debate resembles a debate about the logical coherence of the very notion of an all-perfect God. Usually, someone who rejects the view that this notion is coherent will argue for an alleged inconsistency implicit in the notion: it has been argued, for example, that a being cannot be perfectly merciful and perfectly just at the same time: perfect justice requires always (or so the argument goes) giving a person what she deserves while perfect mercifulness requires sometimes giving a person less harsh treatment than she deserves. To my knowledge, theists lack a positive argument for the consistency of all the relevant notions comprising the complex concept of a being that instantiates such perfection, focusing instead on rebutting arguments asserting the inconsistency of the concept of an all-perfect God. The reason for this is that it is very difficult to make a positive argument for the claim that these ideas are consistent; we lack a very direct sense of what claims are coherent because inconsistency can lurk deep beneath the surface. For this reason, the structure of this dispute between those who affirm MIT and those who deny it generally parallels the structure of the debate concerning the coherence of the notion of an all-perfect God.
16As should be evident, possibility claims are difficult to support with positive arguments. To claim that a state of affairs S is possible is to claim nothing stronger than this: a set of propositions that exhaustively describe S is logically consistent or, alternatively put, is not self-contradictory. The claim, then, that a state of affairs is possible expresses, in essence, a logical claim about the set of sentences that exhaust the description of all theoretically salient features of S – namely that this set of sentences does not contain sentences that entail a logical contradiction.
- 7 For a classic and still useful text on modal logic, see, e.g., Chellas 1980.
17There should not be any confusion about this: it is a straightforward application of principles at the foundation of the possible-world semantics presupposed by standard systems of modal logic. Indeed, a possible world is typically defined as a maximally consistent set of sentences.7 A set of sentences S is maximal in the relevant sense if and only if, for every proposition p, S contains one of either p or not-p. A set of sentences S is consistent if and only if it is not the case that S logically implies a contradiction. Thus, a possible world is a set of sentences that contains one of either p or not-p, for every proposition p, and does not logically imply a contradiction. A positive argument for a claim that it is possible that p – the schematic form of a possibility claim – requires showing that there is a maximally consistent set of sentences containing p.
18Accordingly, a successful positive argument for MIT must be constructed to show that there is a maximally consistent set of sentences that includes the proposition that, for some institutional system L, L is a legal system that has moral grounds of law. This requires showing that there is no logical contradiction between (1) the claim that L is a system of law (with all that this presupposes by way of conceptually related claims, for example, about authority, obligation, etc.) and (2) the claim that L has moral grounds of law.
- 8 Shapiro (1998: 469–507).
19Up to this point, there have been few positive arguments offered in support of MIT – and none successful. The debate has largely taken the shape, as will be recalled, of arguments to the effect that there is no consistent set of propositions containing the following propositions: (1) the set of propositions correctly describing the existence conditions for law; (2) the proposition that, for some entity L, L is a legal system; and (3) the proposition that L has moral grounds/criteria of law. Raz’s argument is grounded in the claim that the set defined by (1) includes the proposition that law claims legitimate authority, as well as the propositions expressing his service conception of authority. His argument is that an institutional system of norms including, so to speak, moral grounds for membership in the system is logically inconsistent with its satisfying the set defined by (1). The problem, on Raz’s view, is that the existence of moral grounds of law is inconsistent with law’s conceptual claim to authority. Similarly, Shapiro’s argument is that the set defined by (1) contains the so-called Practical Difference Thesis.8 His argument has exactly the same structure as Raz’s: the idea is to show that an institutional system of norms that includes moral grounds/criteria for membership is inconsistent with the set defined by (1). On Shapiro’s view, the problem is that moral grounds/criteria of law are logically inconsistent with the claim that law is capable of making a practical difference.
- 9 See, Raz 1994.
- 10 I have challenged Raz on a number of these assumptions. See, e.g., Himma 2001a; Himma 2007a; Himma (...)
- 11 See, Himma (2000: 1–43).
20Inclusive positivists have largely defended MIT by attempting to show, on a piecemeal basis, that each such “inconsistency argument” is unsuccessful by showing that the argument in question relies on a false premise, whether explicit or implicit. Inclusive positivists, including myself, have defended MIT against Raz by showing that the Razian argument relies on false claims about the nature of law or the nature of authority.9 In particular, the defense against Raz is grounded in an attempt to rebut one or more of the following three ideas: (1) law necessarily claims authority; (2) authoritative directives provide pre-emptive reasons for action; or (3) authority is justified only insofar as its directives more accurately identify what a subject should do according to right reason than the subject’s own assessment of the reasons.10 Likewise, inclusive positivists have defended MIT from Shapiro’s argument by denying either (1) that positivism includes the Practical Difference Thesis; or (2) that the Practical Difference Thesis implies that judges must be guided in their deliberations both by the rule of recognition and the relevant first-order norms regulating the acts of citizens.11
- 12 This is not to suggest that these negative arguments cannot form the basis of a positive defense of (...)
21It is of critical importance to note that, either way, these defenses of inclusive positivists are “negative” in the sense that they attempt to show that these inconsistency arguments are unsuccessful – that is to say, these negative arguments are limited to showing that certain arguments against MIT fail. This is noteworthy because, strictly speaking, such arguments provide absolutely nothing by way of positive support for MIT. The reason for this is a simple matter of basic logic: the claim that an argument is unsound does not provide any reason for thinking that the conclusion of the argument is false. What this means is that these arguments succeed in doing no more than showing that MIT has not been shown false. By itself, none of these counterarguments give any reason to think that MIT is true simply because, as a matter of elementary logic, they cannot.12 To show, for example, that it is not true that law necessarily claims legitimate authority does not, by itself, give any reason to think there could be moral grounds of law. A review of the literature suggests that the vast majority of content defending MIT published over the last fifteen years has involved making negative arguments of this sort. Very few positive arguments have been offered in support of MIT.
22At first glance, the exclusively defensive posture of the inclusivist might seem puzzling; however, it is easy to understand why the debate has taken this shape once it becomes clear what would be involved in crafting a successful positive argument. As it turns out, the real problem is, at bottom, that the logic of the relevant modalities has not been understood well enough to do what is needed to make a positive argument in support of MIT. It is difficult enough, I would surmise, for people who lack training in metaphysics to fully understand the character of various metaphysical claims as well as what the proper methodology might be, which is, of course, a topic under dispute. Once modalities like the necessity operator (i.e., it is necessary that p or, symbolically, □p) and the possibility operator (i.e., it is possible that p or, symbolically, ◊p) are introduced into the conversation, the logical standards governing deductive argument become considerably more complicated – a complexity that reflects the corresponding complexity of metaphysics as a separate discipline. And it is important to realize here, as is frequently overlooked, perhaps because thought obvious or unnecessary, that conceptual jurisprudence is nothing more than the metaphysics of law.
23Here it is helpful to consider again the issue of whether the concept of an all-perfect being is coherent – which, it should be noted, is also a metaphysical issue. The relevant atheistic claim is that there is not (indeed, could not be) a thing that satisfies the existence conditions for perfection because some of these conditions are logically inconsistent with other conditions. As noted above, for example, it has been argued that being perfectly merciful and perfectly just cannot be simultaneously instantiated by a being because the criteria for being perfectly just are logically inconsistent with the criteria for being perfectly merciful. As I also noted above, though, theists lack a positive argument that directly addresses the consistency issue by showing that, so to speak, all the perfections are logically compatible. While there are, of course, arguments for the existence of God (e.g., the so-called intelligent design argument), they do not directly address the coherence of the concept of an all-perfect being. Of course, insofar as one of these arguments is successful in showing that an all-perfect God exists, it follows as a corollary that the concept is coherent. But these other arguments do not even attempt to directly engage with the inconsistency arguments of atheism.
- 13 There are many examples that secure the point. We could not, for example, see how Euclid’s Parallel (...)
24The reason for the comparative lack of positive arguments for the coherence of the concept of an all-perfect God is that there are special difficulties associated with arguing directly for possibility claims (i.e., in this case, it is possible for a being to instantiate each of the perfections). It is hard to see exactly how one could make a positive argument for the relevant possibility claim. How, for example, would one even begin to show that it is possible for a being to be omnipotent and omniscient at the same time? While it might seem obvious that at least that much is possible, this is because we cannot think of how an inconsistency might arise between those two qualities. If that might lead us, reasonably enough, to doubt that there is any such consistency, our inability to imagine how those two qualities might conflict does not provide any significant positive support for the claim that they do not conflict.13 When one reflects on how one might go about providing a positive argument that shows that the two qualities are consistent, one will likely come up blank – which explains why the structure of the debate as to whether it is possible for a being to instantiate all the perfections (i.e., whether the concept of an all-perfect being is coherent) parallels the structure of the debate as to whether it is possible for a legal system to have moral grounds of law. The problem is that it can be, and frequently is, exceedingly difficult to see how to show that two abstract propositions are consistent.
25If this seems surprising to philosophers, it is taken for granted in other academic subjects. Establishing possibility claims in pure mathematics is well known to present special difficulties. One issue that commonly arises in the context of studying a particular formal axiomatic system in mathematics, such as set theory, is whether the set of axioms is consistent; and whether, if so, the consistency of that set can be proved. There are two possible consistency theorems – one stronger and hence more desirable than the other. The most desirable result would be a showing that a set of axioms is absolutely consistent, which is what we usually have in mind when we say that some set of propositions is consistent. What is needed to show that a set of axioms is absolutely consistent is an interpretation of all of the symbols of the formal language of the axiomatic system in which the axioms are true; such an interpretation is known as a model for the set of axioms.
26It is absolutely crucial to note the relationship between finding a model for a set of axioms and giving a positive argument for MIT. Intuitively, the idea behind the mathematical project is to interpret the symbols of the language in such a way that we can create a coherent story in which all the interpreted axioms are clearly true. This is analogous to what must be done by way of giving a successful positive argument for MIT. If one can produce a model for the axioms of the relevant mathematical system, then it follows that the axioms are consistent – which is to say that there is some possible world in which all the axioms are true. If the proponent of MIT can produce a coherent story in which something that is clearly a legal system (under the relevant theory) also clearly contains moral grounds of law, then that is sufficient to show that the existence of moral grounds of law is consistent with the relevant theory of law – which is to say that there is some possible world where there is a legal system with moral grounds of law. In other words, if the proponent of MIT can produce such a “model,” that suffices to establish MIT by what I have been calling a positive argument.
- 14 See, e.g., “Equiconsistency,” at: <http://en.wikipedia.org/wiki/Equiconsistency>. It suffices here to cite Wikipedia because none of this is even</http> (...)
27Intriguingly, absolute consistency results are very difficult to achieve in mathematics because it turns out to be prohibitively difficult to construct something that clearly serves as a model for the relevant mathematical theory.14 The reasons for this have, in part, to do with the fact that the notion of “consistency” is also deployed within every mathematical theory. It is not as if there is some Archimedean point of view in mathematics in which one can produce an analysis of whether some set is consistent that does not presuppose some element of the system under question: standards defining the notion of consistency are (pun intended) consistently applied throughout every standard mathematical theory. But a second reason surely has to do with the difficulty of trying to construct an interpretation of the symbolic language and model from scratch that would clearly establish the consistency of the system (even if we could formulate a test for consistency from outside the system being studied). The story here is much more complicated and contentious than this might suggest but it is plausible enough that it gestures in the direction of a problem faced by theorists of any discipline attempting to establish a possibility claim (by way of proving an absolute consistency result).
28In consequence, mathematicians must, much more often than not, settle for showing what is called relative consistency. The idea is to take a set of axioms, S, that seems more perspicuously consistent from an intuitive point of view and attempt to show that the axioms of the system under study, T, are theorems of S. A proof that the axioms of T are axioms of S shows that T is consistent relative to S in the following sense: if S is consistent, then T is consistent. In the vast majority of cases, relative consistency is the best that can be done; indeed, it cannot be shown even that arithmetic is absolutely consistent.
29As intimated above, there is probably no comprehensive, accessible theoretical account, and one that commands a consensus, of why absolute consistency results are so difficult to obtain in mathematics; but, as we also saw, we can get a rough sense for why this might be true. It might not be possible to produce a rigorous theoretical explanation as to why consistency results are difficult to support with positive arguments in philosophy of law. But we can get a sense for the complexity involved in producing a model that would validate MIT.
30One tempting thought is that there are many readily available models of existing legal systems and that any one of these existing legal systems can be used to show moral grounds of law are possible. After all, it is a trivial truth – at least, under any standard system of modal logic that has direct application to philosophical analysis – that, for all propositions p, p logically implies ◊p. If one can show that some existing legal system has moral grounds of law, then it follows a fortiori that it is possible for a legal system to have moral grounds of law, which is precisely what is expressed by MIT.
31This strategy turns out to be much less attractive than it appears for reasons that, as will be shown, explain my adoption of a related but more direct approach to finding a model for MIT. It is worth noting here that this approach will lead me to construct a model that might strike many readers, untrained in modal logic, as implausibly simplistic, improbable, and otherwise problematic and unpromising. But, as we will see below, this impression is incorrect, and grounded in misconceptions about the logic and methodology of model construction of the sort offered in this essay.
32In his justifiably influential book, Inclusive Legal Positivism, Wil Waluchow adopts this strategy in what is perhaps the only sustained attempt to give a positive argument in support of MIT. Waluchow saw what he took to be a clear model for MIT in the legal practices associated with the substantive guarantees of the Canadian Charter. He argued, for example, that judges routinely have recourse to moral principles in deciding issues arising under these guarantees and are legally bound to apply these moral principles. On the strength of these practices, he concluded that the Canadian legal system is a legal system with moral criteria of validity and hence that MIT is true.
33The problem with Waluchow's analysis is that it presupposes a contentious interpretation of those legal practices. Waluchow interprets – without argument – the moral principles to which judges turn in hard Charter cases as pre-existing principles that have the status of law; the judge in such Charter cases discovers the content of the law defined by the Charter. In contrast, the exclusive positivist interprets such recourse as the judicial exercise of a quasi-lawmaking discretion to reach beyond the law in inventing (or making) new law. As Stephen Perry puts the point:
After making rather havy weather of the fairly obvious point that judicial opinions in charter cases involve moral reasoning, Waluchow attempts to show that the moral standards employed in charter cases, drawn from e.g. section 7 of the Canadian Charter of Rights and Freedoms (right to life, liberty and security of the person), or the First Amendment of the Americal Constitution (freedoms of speech and assembly), function at least sometimes as tests for the existence or content of valid laws. If this could be shown, it would establish the truth of inclusive positivism and provide a counterexample to exclusive positivism. Waluchow maintains that exclusive positivism has available to it “only one obvious possibility” to explain charter challenges, namely, the idea that provisions such as section 7 of the Canadian Charter do not set out legal criteria for validity, but simply make reference to extra-legal, moral criteria “to which judges are required or at liberty to appeal” (157). But in fact an exclusive positivist could say more that this, since he could appeal to Waluchowʼs own notion of legal-adjudicative power or duty to decide charter cases in certain ways. Although Waluchow does not mention it, Raz employs his notion of a directed power to characterize judicial review cases along very much these lines.15
34Although Perry is certainly correct in thinking this is a problem, he does not fully appreciate the character of the problem with Waluchow’s argument here.
35Perry seems to believe that the problem is just that the interpretation is contentious – a problem that Waluchow attempts to address by arguing that the inclusivist interpretation of Charter practices is superior to an exclusivist approach. As Perry describes Waluchow’s next steps in the argument:
Waluchow criticizes what he takes to be the exclusionary positivist's best possible account on three main grounds: (i) it flouts the common understanding of charter provisions as entrenched fundamental rights that are legal, not just moral, in character; (ii) it runs counter to language often found in constitutions themselves, such as the provision in section 52(1) of the Canadian Constitution Act that any law inconsistent with the Constitution is of no force or effect; and (iii) it cannot explain as easily as the inclusive account the retroactive effect of charter challenges. For these reasons, Waluchow concludes that charters and bills of rights create “legal rights whose content is partly dependent on moral considerations” (162).16
36As Perry explains the problem, Waluchow’s move seems to be the only sensible move. Waluchow needs to show that the inclusivist interpretation better coheres with the relevant Charter practices than the exclusivist interpretation; and this is exactly the move described in the text quoted above.
37Perry does not question the legitimacy of this strategy; that is to say, he does not provide any further response to Waluchow’s claim that any exclusivist interpretation of Charter practices will be marred by the problems described by (i), (ii), and (iii) in the passage above.
38Instead, Perry responds to another of Waluchow’s criticisms of exclusivist interpretations that Waluchow believes show that an inclusivist interpretation better coheres with Charter practices than exclusivist interpretations. As Perry argues:
One of Waluchow's arguments in favour of the inclusive-positivist interpretation is that the exclusive-positivist account flouts the common understanding that charter rights are legal and not just moral in character (158-59). Yet the exclusive positivist can borrow Waluchow's own notion of legal-adjudicative rights, developed as part of the latter's source-based conception of the common law in chapter 3, and argue that while charter rights might not be legal rights in the strict sense, they are not just moral either; they are legal-adjudicative rights that citizens hold against the courts. (As noted earlier, Raz makes essentially just this move, employing his very similar notion of a directed power.) Another of Waluchow's arguments in support of the inclusive-positivist account of judicial review is that it explains the retroactive effect of charter cases better than the exclusive-positivist account (160-62). But cases overruling common law precedents likewise generally have retroactive effect, so whatever Waluchow would wish to say about retroactivity in that context, as part of the source-based conception of the common law he develops in chapter 3, can presumably be relied upon by the exclusive positivist in defending a source-based interpretation of constitutional judicial review.17
39This makes it seem as though the dispute between the exclusive positivist and the inclusive positivist can simply be settled by looking at a particular legal system to see which type of interpretation better conforms to the practices. The idea here seems to be that had Waluchow succeeded in showing that the inclusivist interpretation better coheres with Charter practices than an exclusivist interpretation, this would be enough to show MIT.
40This might be a common view but it is grounded in a fundamental misunderstanding of the modal character of MIT and what that character requires by way of positive support. Even if MIT is false, as exclusive positivists believe, one might still be able to produce coherent interpretations that are, in some sense, logically grounded in MIT – and plausible ones (assuming you do not know that MIT is false).
41What makes this possible? The reason here is not necessarily the obvious point that any proposition follows from a contradiction, which would be true of the claim that some legal systems have moral grounds of validity if MIT is false. Nor is the reason necessarily that MIT contains so many conflicting principles that one can select some favored set narrowly tailored to support any proposition – this, of course, being the basis for the Critical Legal Studies claim that law is so globally indeterminate that judges can reach either of two conflicting holdings in hard cases with an ostensibly rational argument. The reason is more narrow: it is that one can still produce plausible interpretations that come out of MIT that derive neither (1) from the contradiction itself if MIT is false nor (2) from the kind of indeterminacy that the Critical Legal Theorists believe results in nearly unfettered discretion to reach any decision because any decision can be justified because of the multiplicity of conflicting values that explain the global indeterminacy of law.
42This might seem implausible at first blush, but producing a plausible interpretation grounded in MIT is exactly what Waluchow did with Charter practices. Perry did not even attempt to show that Waluchow’s inclusivist interpretation of Charter practices was inconsistent or incoherent. Rather, Perry identified the relevant Charter practice and provided an exclusivist interpretation that was claimed to cohere as well with these practices as Waluchow’s inclusivist interpretation. But even if Perry produced an exclusivist interpretation that was more coherent with the relevant practices, it would not follow that inclusivists could not produce a coherent interpretation of any particular Charter practice. That one interpretation better coheres with the relevant phenomenon does not imply that another is incoherent. Coherence is a matter of degree. Nor does the fact that one interpretation is more plausible imply that another is implausible. One theory might simply be more plausible than another. Like coherence, plausibility is a matter of degree.
43But this is also a feature of possibility claims in mathematics; it is always possible that what seems to be a model showing the possibility of a world in which the relevant axioms are all true is not really a model at all for the axioms. As will be recalled, it is nearly impossible to produce absolute consistency results for axiomatic systems in mathematics, including those that form the foundation for all deductive reasoning – i.e., mathematical logic. The reason is epistemological: we have no Archimedean point from which to assess whether a putative and plausible interpretation succeeds in modeling the relevant axioms. The fact that it seems plausible and coherent is consistent with its being false. Further, it is consistent with the interpretation being inconsistent provided that the inconsistency is difficult to see.
44Indeed, we might know a theory is inconsistent but be justified in applying various elements of that theory to certain phenomena in the world. For example, physics is divided into two analytically distinct theories: (1) the theory of the very big (i.e., the theory of relativity) and (2) the theory of the very small (i.e., quantum mechanics). Both are unparalleled in the history of physics with respect to the results they predict. However, it is well known that the two theories are inconsistent with each other in the standard sense that they cannot both be true. Indeed, this is what motivates a continuing search for a “unified theory of everything,” a search that focused for a time on string theory. But despite all this, the theory of relativity and the theory of quantum mechanics are used to interpret, explain, and predict physical phenomena.
45Returning to the dispute between inclusive and exclusive positivism, the upshot is this: Stephen Perry’s assumption that the dispute between inclusive positivism and exclusive positivism can be settled by finding some legal system with practices, like Canada, that are best interpreted as inclusive – which would show the existence of an inclusive legal system and thereby prove MIT – overlooks the fact that MIT could still be the source of plausible interpretations even if it turned out to be false. Again, pure mathematicians continue to derive useful, plausible theorems from sets of axioms that might very well turn out to be inconsistent; indeed, as we have seen, there are few, if any, models for any set of axioms that would show the absolute consistency of these axioms. Further, physicists continue to derive usual results from both quantum mechanics and the theory of relativity, despite proofs that the two theories are inconsistent and cannot both be true in this world.
46The real problem with Waluchow’s argument strategy goes deeper than Perry’s claim that Waluchow has not produced an inclusivist interpretation that better coheres with existing Charter practice than the exclusivist can produce. At bottom, what ultimately goes wrong with Waluchow's argument is that his interpretation of the relevant legal practices begs the question. Waluchow's interpretation of these practices ultimately rests on an assumption that moral criteria of validity are possible; without that, he lacks even a prima facie justification for interpreting Canadian practice that way. The interpretation is a non-starter because the very coherence of inclusivist interpretations is challenged by the exclusive positivist. If we had compelling reason to think MIT were false, then it would not matter in the least how attractive some proposed inclusivist interpretation of Charter practices might be. That interpretation is immediately disqualified as grounded in a contradictory claim.
47Here's another way to see this important point: if we already know that inclusive systems are possible, there will be an issue as to whether any particular legal system is inclusive – including the Canadian legal system. The argument Waluchow offers in support of MIT would provide a plausible reason for thinking that the best interpretation of Canadian legal practice is as incorporating moral criteria of validity – if we already know that moral grounds of law are possible. That view could be countered, of course, but it would not be refuted by Perry's observation that exclusivists interpret those same practices differently. If, again, we already know inclusive legal systems are possible, then the issue becomes which is the better interpretation of Canadian legal practice. If Waluchow’s interpretation is better (in the sense that it better coheres with the relevant legal data points), then that is good reason to think Canada has an inclusive system with moral grounds of law. If the Razian interpretation is better in the relevant sense, then that is a good reason to think Canada does not have an inclusive system. But, regardless of which one is the better interpretation, the conclusion reaches no further than to how the particular practices of the Canadian legal system should be characterized. If we are not antecedently justified in thinking that moral grounds of law are possible, there is simply no non-question begging reason to think Waluchow's interpretation is better.
48Surprisingly, what goes wrong, at bottom, with Waluchow’s strategy is that there is too much complexity built into these practices of the Canadian legal system to form the basis of an uncontroversial model of MIT. MIT asserts a very simple claim – even if it is the kind of claim that seems, as I have remarked, difficult to support with a positive argument in virtue of its modal quality. The claim is that there is some possible world in which there is something that counts as a “legal system” with something that counts as “moral grounds/criteria of law.” As it turns out, the weakness of this claim provides some guidance as to how to go about producing a positive argument in the form of a model. All one has to do is cook up a coherent story that shows (1) how there could be a legal system (2) with moral grounds of law: those are the guiding features.
49This makes it clear that we must adopt a very different starting point from the one Waluchow adopts. It is not just that Waluchow chooses an existing legal system that causes the problem; it is that Charter practices are enormously complex – so complex that they include too many features that are compatible with both inclusivism and exclusivism. There is simply too much material in the Canadian legal system that obscures the issues by requiring interpretations so complex that they contain many entry points for contesting interpretations.
50What is needed is something far simpler – and this requires that we construct a model from the ground up, taking care to ensure that we build no more into the system we wish to construct than what is reasonably necessary to provide a model for MIT. Every piece of additional information creates a logical opening for disagreement and hence for opposing interpretations. As we will see, adopting this approach will result in a model that might, at least initially, (1) seem overly simplistic, and (2) contain attributions of properties to subjects that cannot plausibly be attributed to people in this world. Indeed, one will be tempted to reject the model just on the strength of its simplicity and dissimilarities with the complexities of legal system in the actual world – that is, just on the strength of instantiating (1) and (2).
- 18 The reason is that the relevant possibility claims simply, by nature, lack the kind of explanatory (...)
51But, at this point, it should be clear that rejecting a model because it instantiates (1) and (2) would be a mistake because MIT makes such a strikingly weak claim: it asserts merely that a legal system with moral grounds/criteria of law is possible. That, by itself, says nothing about what such a legal system should look like. Indeed, it does not even imply that there is an inclusive legal system in some world that governs beings with exactly the same set of capacities and qualities we have. All that is needed by way of assumptions about the subjects is two claims: (1*) the subjects are rational; and (2*) the subjects are capable of being guided in their behavior by norms. It need not be the case, for example, that the beings in that model (or the possible world it describes) share the same cognitive abilities and psychological characteristics that people have in this world. Indeed, MIT does not imply, assert, or even insinuate that the law subjects must belong to the class of human beings.18
- 19 Raz (1999: 158; emphasis added).
52A very helpful example of what many consider to be a successful model for a possibility claim, despite the fact that Raz’s model instantiates (1) and (2), can be found in the work of Joseph Raz. Raz considers the question of whether coercion is a necessary feature of law; as he puts the matter, “Is it possible for there to be a legal system in force which does not provide for sanctions or which does not authorizetheir enforcement by force?”19 It is telling that Raz frames the question in terms of its negation: instead of asking directly whether it is a matter of conceptual necessity that a legal system be backed by coercion, he asks whether it is conceptually possible for something lacking coercive enforcement mechanisms to count as a legal system.
53The reason Raz expresses the question in terms of whether it is possible to imagine a legal system without coercive enforcement is that he is setting up an argument that uses exactly the methodology explained above. Indeed, he all but makes this explicit:
The answer seems to be that it is humanly impossible but logically possible. It is humanly impossible because for human beings as they are the support of sanctions, to be enforced by force if necessary, is required to assure a reasonable degree of conformity to law and prevent its complete breakdown. And yet we can imagine other rational beings who may be subject to law, who have, and who would acknowledge that they have, more than enough reasons to obey the law regardless of sanctions.Raz (1999: 158–159).
54Raz is presupposing here that one proper way of showing coercion is not a necessary feature of law is to show that one can – coherently – imagine a legal system without coercive enforcement mechanisms. He concedes that we might not be able to do it with human beings because of certain psychological qualities we have; however, that only makes it nomologically impossible (i.e., impossible given the causal laws in our universe). Raz stakes out the weaker claim that it is logically possible to have a legal system. He claims we can imagine a logically possible legal system without enforcement mechanisms but not nomologically possible.
- 20 We will see below that it also satisfies (1).
55Two observations should be made here. First, Raz makes clear that the model he constructs will contain rational beings as law subjects and, thus, satisfies (1*); likewise, Raz makes clear that the subjects of the model are capable of being guided by norms (i.e., these beings have “more than enough reason to obey the law regardless of sanctions”), and thus that the model satisfies (2*). Second, Raz’s explanation of his methodology is explicit in indicating that the model will instantiate (2), since it will involve beings with psychological features that human beings lack – one of the two concerns that might arise about the model I construct below of an inclusive legal system.20 It is not nomologically possible, Raz maintains, for there to be a legal system without coercive enforcement mechanisms because human beings lack crucial psychological attributes that would ensure the efficacy of the system of norms in guiding behavior. Rather, it is logically possible, which means – given his view that such a legal system is not nomologically possible – that the relevant law subjects will have very different properties than human beings possess. It is, thus, not unusual for arguments for possibility claims to rely on assumptions that do not apply to human beings or other relevant phenomena in the actual world. It is clear that he acknowledges (2) is a legitimate property of models that establish possibility claims in conceptual jurisprudence.
56It is, of course, true that the intuitive foundation for the Hartian analysis can be found in these instances of what we regard as paradigms. MIT says nothing that should lead anyone to think that any legal system with moral grounds or criteria of law must possess any of the features that we regard as paradigmatic features of a legal system. Conceptual jurisprudes look at paradigmatic features of a legal system because they do tell us something essential about law: they are paradigmatic features because they are a core – as opposed to more nearly borderline or even borderline – instance of some essential property. If there is a legal system incorporating moral grounds or criteria of law, it must have all the essential features of law. However, there is simply no reason to think that any of its features must be paradigmatic features; as far as conceptual analysis is concerned, every essential feature of such a legal system, if possible, could be near-borderline or borderline. It is true we might not recognize them as systems of law because features that are borderline fall within the open texture of a concept and thus are epistemically indeterminate. But the fact we cannot identify whether a borderline feature is actually an essential feature does not imply that the borderline feature is not an essential feature. Some borderline cases of a concept fall within the concept; some do not.
57The second concern one might have about the model I construct below – namely, property (1), which is that the model is simplistic – also arises in connection with the model Raz constructs of a legal system that lacks coercive enforcement mechanisms. His argument is in the form of a model that satisfies the constraints of modal logic discussed above:
Even a society of angels may have a need for legislative authorities to ensure co-ordination. Angels may be in agreement about both their values and the best policies for implementing them. But the sort of society described above does not presuppose such a measure of agreement. Its members may pursue many different and conflicting goals and they may share our difficulties in settling disputes and resolving conflicts ofinterests by mutual agreement. They differ from us only in having universal and deep-rooted respect towards their legal institutions and in lacking all desire to disobey their rulings. They have, therefore, all the reasons that we have for having legislative authorities and an executive.21
58It can now be seen, as mentioned above, that Raz creates a model with beings that have attributes that human beings lack, and his model, thus, instantiates (2) above. More importantly, for our purposes, is the simplicity of the model. Raz’s model is grounded wholly in the assumption that a society of angels is logically possible; he simply assumes that there is a coherent set containing all the other propositions needed to fully describe a world that is consistent with there being a society of angels. The argument is, then, just that a society of angels will need certain disputes adjudicated, and certain coordination problems solved.
59The simplistic appearance is quite misleading. A possible world consists of a coherent set containing an uncountably infinite number of propositions: it is a coherent maximal set of propositions in the sense that for every possible sentence A, either A is true in the world or not-A is true. The reason for the simplicity of the model is that no one can list an infinite number of propositions, so the definition of a model will, as a matter of nomological necessity, consist in the specification of a finite number of propositions and the assumption that the rest of a coherent model exists. Raz’s model seems simplistic because he devotes just a couple of paragraphs to sketching the model, and he makes explicit only the key ideas. The assumption that there is a maximally coherent set of propositions (defining a logically possible world) containing an uncountably infinite number of propositions is a plausible one. And, again, no one could specify every proposition in an infinite set. The simplistic concern is unfounded because based on a misunderstanding of the logic of the possibility operator.
60With these preliminaries in mind, the model of a legal system incorporating moral criteria of legality can be developed as follows – subject to the following conditions and assumptions. First, I assume at the outset that none of the exclusivist arguments against inclusive legal positivism succeed in refuting the latter – an assumption that should be quite plausible. By far, the best such arguments have been made by Joseph Raz and Scott Shapiro: Raz argues that moral grounds/criteria of law are incompatible with the concept of legal authority, while Shapiro argues that moral grounds/criteria of law are incompatible with the Practical Difference Thesis. Both arguments have generated a number of responses, which – at the very least – call into question the adequacy of the arguments. While I think a number of the arguments successfully refute these two arguments, it is safe to say that the dispute is at an impasse. Thus, it is reasonable to assume that MIT has not been refuted. The argument, of course, depends on that assumption.
61Second, I will attempt to construct this model in a way that is compatible with both (1) the fundamental theses of legal positivism and (2) fundamental assumptions about law that are shared by legal positivists and anti-positivists alike. The first step in the argument is to address (1). On the positivist assumption that the grounds of law are completely determined by conventional practices in the sense that the criteria of legality are fully defined by a social rule of recognition, I argue for the possibility of a legal system with moral grounds of law by providing a model that is grounded in some highly unusual assumptions.
62Third, the argument will specify only a few propositions that are intended to define the logically possible world. In this connection, it is helpful to recall that a possible world consists of an uncountably infinite number of propositions; an omnipotent being can specify a countably infinite number of propositions, but even an omnipotent being cannot specify an uncountably infinite number of propositions. Human beings lack the ability to specify even a countably infinite number of propositions – or even a sufficiently large finite number of propositions. Models are, in both philosophical and mathematical practice, defined by a comparatively small number of propositions, as will be the case here. It is, as we have seen, not a general criticism of arguments defining models that they are too simplistic in the sense that the model is specified through a small number of propositions.
63Fourth, and related to the first point above, the argument thus depends on the assumption that the specified propositions can consistently be joined with an uncountably infinite set of propositions to form a maximally consistent set and hence a possible world. If the assumption I made in the first of these points is not true (i.e., that none of the exclusivist arguments against MIT succeed), then this fourth assumption also fails. The exclusivist arguments would have shown that there is no set of propositions, including other propositions about the legal system, that can be joined to the propositions that define the model to form a maximally consistent set.
64Having said all this, here are the propositions defining the model. To begin, there is a conceptually possible world in which everyone is infallible in discerning the requirements of morality but frequently acts in ways contravening those requirements in order to further his or her own selfish interests. Further, in one such world, an institutional system of rules is set up to regulate the behavior of citizens who act out of selfish interest when it conflicts with morality. Moreover, the grounds for membership in this institutional system of rules are exhausted by the norm all and only moral norms constitute a rule of the system. The fact that everyone is morally infallible means that everyone always knows what the law requires with respect to any particular act.
65It is worth noting the relationship of these propositions to the propositions specifying Raz’s model. Both models assume that law subjects are infallible with respect to moral requirements; the difference is that in Raz’s model, law subjects are morally impeccable and never do wrong, as they are angels, while the subjects in my model are self-interested and frequently do wrong. But, and this is especially important to note, the law subjects of both models have a property that human beings lack – namely, infallibility with respect to the requirements of morality. As I have argued, this is simply not a problem when it comes to establishing possibility claims.
66Some further propositions must be specified as part of my model. First, to keep order, the system employs coercive enforcement mechanisms that also conform to the requirements of political morality, which helps to ensure that people who know better conform their behavior to the first-order moral norms governing their behavior thereby making the institutional system of norms efficacious in guiding behavior. Second, judges always decide disputes arising under these institutional norms correctly.
67The key steps in the argument can be sketched as follows. Since the system includes the minimum content of the natural law and the first-order norms validated by the grounds of this system are generally obeyed, the system counts as a legal system. Insofar as judges infallibly decide what morality requires of citizens and successfully persuade citizens to behave in accordance with those norms through morally justified coercive mechanisms, the norms are, at the very least, capable of providing reasons for action – and, on any sensible account do; even if the content of the law provides no reasons for action, there is no question that the coercive enforcement mechanisms do. Clearly, we have a reason to avoid coercive enforcement mechanisms. Finally, insofar as morality forms a seamless web, as is commonly assumed, there is always one correct answer and that answer is always reached and enforced by the judges, the grounds of law seem clearly to include – indeed, are exhausted by – moral principles. If that is correct, then it is possible for a legal system to have moral grounds of law and hence that MIT is true.
68I should emphasize here that this particular argument shows no more than the possibility of moral grounds of law given certain assumptions about law and about positivism. These assumptions define no more than a very narrow set of circumstances (including some epistemic circumstances) in which morality can be incorporated into the grounds of law.
69Even so, the argument I have sketched is more than strong enough to do the needed work, given what we have come to understand about the implications of a viable system of modal logic with respect to possibility claims. Indeed, it is worth recalling the relationship between the model I define here and Raz’s model that purports to show sanctions are not a necessary feature of law. To begin, he assumes that a society of angels is conceptually possible in a world that bears sufficient resemblance to ours to ground an argument about our concepts of law and authority. But notice that it is not nomologically possible for any human being in this world to be an angel. Further, Raz does not build many more explicit assumptions into the model he is constructing of a legal system free of coercion. The assumption is, and this is surely plausible enough at first blush, that everything else we need to be true of this possible world to draw a conclusion about the actual world is, in fact, true! No argument is even attempted here – and that’s fair game. There is little reason to think that the other elements we need for that possible world to be properly related to this one are not satisfied.
70The problem with Razian argument is not with its methodology; rather, he makes a conceptual mistake. The problem is that whatever system of norms would be necessary to perform these coordination functions in a society of angels, it would not be a system of law because whatever authority is needed to settle disputes, those disputes will not concern the “minimum content of the natural law”; a society of angels will know those rules without promulgation, and the angels will conform their behavior without needing to treat them as pre-emptive reasons. Angels, after all, are presumed to be both morally infallible and morally impeccable (i.e., always do what is the right thing to do).
- 22 Although Hart seems to regard the incorporation of the minimum content of natural law into the law (...)
71Accordingly, the “authority” of such a system, for that reason, simply could not be a legal authority, on Raz’s own theory of law and authority. Indeed, it is hard to see in what meaningful sense it could be true that the “law” of such a society, if concerned only to solve coordination problems (which would be the only kind of problem likely to give rise to disputes among angels), contains the minimum content of the law. Now if that is a correct analysis of the Razian claim, then a society of angels as Raz describes it could not ground a model of MIT because whatever system of norms it has to solve coordination problems does not constitute a system of law. But that is not the only problem here: grounding a model of MIT in a society of angels would be problematic only because of substantive concerns; it would not be problematic because of methodological concerns. A Razian model, like the model I attempt, would not be methodologically problematic simply because the law subjects do not sufficiently resemble human beings.22
- 23 Coleman (2001: 77–81). The assumption here is that norm could fix the content of the law without a (...)
72It is to avoid such problems that I argue for MIT on the basis of more general fundamental assumptions shared across theories. In this connection, it is worth noting that the argument I offer here might strike positivists as making some implausibly strong assumptions about how official practices determine the grounds of law. But if one allows that officials can incorporate vague principles into the law by fixing the content through the statement of a norm, as Jules Coleman suggests, then it is that much easier to make the positive case for MIT precisely because it allows for the possibility of officials characteristically making mistakes about a social rule of recognition without its changing the status of that rule as stating grounds of law.23 Of particular significance here is that the argument that can be made under Coleman’s assumption would also provide support for MIT in a manner that is consistent with Dworkin’s view about the relationship between sentences stating rules and the content of the propositions expressed by those sentences.
73Claims about what is merely possible have limited explanatory power with respect to what actually exists because the possible world that validates the claim, and this is certainly true of my arguments, might look nothing like the actual world. In the actual world, people are not morally infallible – although they are prone to doing what they believe to be wrong for the purpose of satisfying their own self-regarding interests. As far as coherence arguments of the type I offer here are concerned, they simply make intelligible the idea of moral grounds of law (i.e., as not contradicting other obvious conceptual truths about law). While Coleman’s assumption makes possible a way of interpreting legal practice in the United States as incorporating moral grounds of law, along the lines of how Waluchow interpreted legal practice in Canada through the filter of MIT, it accomplishes, I argue, no more than that. An exclusive positivist can still interpret a legal system of that type as not incorporating moral grounds of law – although this would beg the question in the absence of a compelling counterargument (which I have argued above is lacking).
74To see the plausibility of the idea that MIT has little to offer by way of explanatory potential, it is helpful to consider an issue in philosophy of religion. Some philosophers have argued that the very concept of an all-perfect God is incoherent; if that is true, then it is not possible for something to exist that instantiates all of the properties entailed by the locution “all-perfect.” So a successful argument to the effect that the concept of an all-perfect God is incoherent implies that an all-perfect God does not exist. This does tell us something about the explanation of events in the world; it tells us that they cannot be explained in terms of the acts of an all-perfect God. But notice that what explanatory power this has is minimal because it tells you nothing about what explains events in the world; it simply tells you what does not explain those events.
- 24 Some philosophers argue that if God’s existence is possible, then it is necessary. But these argume (...)
75A successful argument in the other direction has even less explanatory power. If we omit the questionable idea that perfection entails necessary existence, then a showing that the concept of an all-perfect God is coherent nets nothing by way of explanatory value. The reason is that what this shows is not that God actually exists (i.e., exists in this world) but rather that there is a logically possible world in which God exists. Given that we do not know whether that possible world is the actual one, the claim that God exists in some world tells us nothing by way of explaining events in this world.24 What would be needed at the very least is a showing that God exists in this world; at least, then, we have reason to think that some of the events in this world might be explained by God’s acts. But such a showing is not enough, by itself, to justify claims that God’s acts explain any particular event in the world.
76Exactly the same reasoning applies to MIT. Because MIT does not assert that there are actual legal systems that incorporate moral grounds, much less specify which ones do, it cannot explain any existing legal practices because it provides no grounds for identifying which ones have moral grounds of law. It asserts nothing more than the idea that moral grounds of law are coherent and thus that there is a conceptually possible legal system that incorporates moral grounds of law. The argument I give for the claim, it is true, points us in the direction of characteristics that are necessary conditions for a legal system to do this – in this case, a society where people are morally infallible but not inclined to do the right thing. But, as we are clearly not morally infallible, we cannot pick out worlds in which people are because we would have to be able to infallibly identify all the correct moral principles to determine whether people in some possible legal system have that capacity. For this reason, the explanatory power of MIT is, I argue, of negligible theoretical significance. Indeed, part of what might have temporarily quieted the debate is that the issue has come to seem substantively irrelevant to many conceptual jurisprudes. Nothing of practical value seems to turn on this.
77Indeed, many legal theorists outside of conceptual jurisprudence are quite critical of the entire enterprise of conceptual analysis. Consider, for example, Richard Posner’s plausible argument that conceptual jurisprudence is of no practical value. As he puts the point:
I grant that even if the word ‘law’ cannot be defined the concept of law can be discussed; and that is after all Hart’s title, though he uses the word ‘definition’ a lot. Philosophical reflection on the concept of justice has been a fruitful enterprise since Plato; for that matter, there is a philosophical literature on time. 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. I go further: 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.25
78I cannot argue the point here in detail, but I think that Posner (1) is largely correct with respect to the practical implications of conceptual jurisprudence but that (2) fails to recognize the intellectually (and morally – since his seems to be a moral criticism) legitimate reasons for engaging in conceptual jurisprudence. As to (1), the choice of conceptual theory, as far as I can tell (and this is the subject for a different essay) contributes nothing to the solution of certain substantive problems arising with the law and paradigmatic legal practices. Of course, it is true that certain problems can seem to disappear under one conceptual theory that arise under other contending conceptual theories; however, the appearance is deceiving. To cite just one relevant example, Dworkin argues that the idea that judges have quasi-legislative discretion implies that cases in which that discretion is utilized to create new law that is applied to the case at bar entails that judges are illegitimately applying law in an ex post facto fashion. But Dworkin’s theory faces an analogous difficulty. Assuming there is one right answer to every legal question, there is no guarantee that the judge will reach the correct decision. In that case, it might not be “law” that is applied in an ex post facto fashion; but the coercive mechanisms of law are being applied to enforce rules in that fashion – and that raises the same moral problems. But even when judges reach the right answer in a hard case, the law in a hard case fails to provide reasonable notice and hence fails to guide behavior in a way that is morally legitimate. Indeed, one can argue that the reason ex post facto laws are illegitimate is precisely because citizens do not have reasonable notice of how the coercive mechanisms of the state will be applied adversely against them. To put the point metaphorically, changing one’s conceptual framework to solve a problem is analogous to trying to smooth out a carpet where an air pocket causes a bulge. Simply pushing on it accomplishes no more than to displace the bulge to another location on the carpet.
- 26 I have discussed this problem as it pertains to other more important conceptual questions in other (...)
79Indeed, this appears true of many, if not all, problems of conceptual analysis. Consider the trivial conceptual truth that bachelors are unmarried. It is hard to see what events in the world this truth could even begin to explain. The reason for this is that, on the traditional methodology with which most philosophers are familiar, the idea that conceptual truths are necessary truths entails that the truth of a conceptual claim does not depend on any particular contingent feature of a world. Insofar as that is true, it can be known without knowing anything more about a world except the conceptual content expressed by a concept-term (i.e., the meanings of the relevant terms). It is hard to see how conceptual claims true in every possible world could explain contingent events in any particular world, including this one.26
- 27 I should note here that, although I am neither a mathematician nor a philosopher of mathematics, I (...)
80But the absence of practical benefits, contra Posner, does not imply that there is no reason to study conceptual jurisprudence. Posner’s criticism is, at the end of the day, a moral criticism to the effect that pursuing theories that have no instrumental value wastes valuable resources that can be deployed to solving practical problems. This is simply not consistent with commonly held views about the value of knowledge per se; knowledge is, on the ordinary view, worth pursuing for its own sake. Indeed, it is not implausible to think that many areas of pure mathematics are pursued without regard for whether they are potentially valuable from an instrumental point of view. Further, some of the most famous achievements in mathematics do not seem likely, at this point in time, to have much by way of practical benefits. Indeed, Andrew Wiley’s proof of Fermat’s Last Theorem does not yet seem to have any practical benefits, yet hundreds of mathematicians spent thousands of hours pursuing a proof of the theory.27 Some claims are worth knowing for their own sake – that is, because the truths they express are intrinsically, and not just instrumentally, valuable.