Varieties of Vagueness
Keith C. Culver *
Department of Philosophy and Centre for Social Innovation
University of New Brunswick, Fredericton, NB.
A critical notice of Timothy A.O. Endicott, Vagueness in Law (Oxford: Oxford University Press, 2000). Subsequent references appear parenthetically in the text.
It will surprise no one to hear that laws are often vague, and, moreover, that laws are vague for a variety of reasons. Some vagueness results from specific intentions of users of legal language. Laws may be poorly drafted, or drafted using general terms intended to capture a wide yet unspecified range of affairs. Vagueness of these kinds is, for the most part, noticeable yet bearable. Poor drafting can be fixed, and open-textured laws can sometimes be made more specific if their application proves troublesome. Yet in the trade-off between specificity and openness, a more peculiar kind of vagueness seems intrinsic to the nature of some legal norms. Their open texture cannot be sharpened to a univocal statement of what they require: they are indeterminate, leaving borderline cases where judges lack legal resources to resolve a dispute one way or another. At least part of the difficulty in handling norms of this sort comes from the mixture of reasons for their vagueness. When we observe, for example, that the Canadian Charter of Rights and Freedoms contains a vague provision permitting 'reasonable limits' 1 on fundamental rights and freedoms, we recognize also that mere understanding of the meaning of 'reasonable' goes only so far in dissolving its vagueness. 'Reasonable limits' are not just any artefacts of language use. They are normative standards, notoriously elastic and insusceptible to univocal and final statement in application to concrete cases. Here the real trouble begins: there is at least the air of paradox in the contrast between the fact of vagueness in laws and the opposed determinacy demand intrinsic to the rule of law - the requirement that laws be framed in a way that makes them capable of being obeyed. When has meaning been stretched too far? What justifies judicial setting of borders on the application of indeterminate propositions of law? After all, if 'reasonable limits' have no determinate edges, our most fundamental rights and freedoms have no inviolable borders, and the rule of law has failed. [End Page 109]
The danger of vagueness is most readily displayed by contrasting vague laws against our daily practice of saying that it is either true or false that some particular law applies in a given situation. This is called the principle of bivalence, meaning that there are only two possibilities when considering whether or not a particular law governs a particular situation. This feature of legal norms is captured in informal logicians' talk of the 'truth functionality' of meaningful and usable propositions such as laws. A proposition is truth functional if it is, in principle, capable of being shown to be true or false upon provision of the appropriate kind of evidence. Our ordinary practice is prominent in the way judicial decisions are criticized: a decision is often derided for failing to apply the law correctly, or for applying the wrong law, but it is ordinarily not an option to claim that the decision was a bad one because the case at hand sits in a border area in which the law neither applies nor fails to apply. And yet, as I have suggested above, there are propositions of law, such as the constitutional provision for 'reasonable limits' on rights and freedoms, that seem destined to generate situations in which there is no obvious way to determine that that the vague law applies or does not. The stakes are deceptively high here. If it is true that at least some propositions of law are vague, and that we cannot know that they are vague until we try unsuccessfully to apply them in some situation, we are stuck with a 'mole' in our midst, an unknown saboteur of the rule of law. At times and in situations we cannot predict, vague legal norms will breach the promise of the rule of law that legal norms protect our current expectations about the future legal consequences of our conduct.
There has been no shortage of attempts to understand the tension between the requirements of the ideal of the rule of law, on the one hand, and the existence of vague propositions of law, on the other. The history of these attempts owes a great deal to the tenacity of the last century's loosely associated 'Legal Realist' writers. 2 They disputed what Timothy Endicott identifies as the 'standard view' of adjudication, that judges neutrally apply propositions of law in situations where it is always true or false that a given proposition of law applies. On the standard view, propositions of law are 'juridically bivalent' in the sense that judges always have the means to understand those propositions such that they apply or fail to apply to a given situation, resulting in clear limits on the reach of laws into citizens' lives. The Legal Realists often argued against this standard view on the grounds that judges themselves are too variable in their inclinations to produce consistent interpretation and application of propositions of law. Later theorists such as H.L.A. Hart absorbed [End Page 110] developments in philosophy of language and identified another problem in the standard view of adjudication: some legal propositions appear to be vague, or, on Hart's considered view, neither true nor false. These propositions are indeterminate and so neither apply nor fail to apply to borderline cases. What judges do with these propositions, Hart suggests, is to use their discretion in various ways to extend or withhold a particular law purported to govern a particular situation. 3 This solution has the virtue of facing up to the problem that vague legal propositions leave borderline cases. Yet the rule of law is clearly disrupted by the suggestion that judges reach to something other than the particular law at hand to decide a borderline case. Perhaps more importantly, this solution seems to be excessively bound up with the meaning and application of words, without sufficient regard for the judges and officials whose use of law shapes it in various ways. Ronald Dworkin has dubbed this problem the 'semantic sting,' affecting those whose otherwise useful attention to legal language has left them unable to reconcile the rule of law with the existence of vague propositions. Dworkin has offered an alternative that saves the rule of law while famously insisting that there can be a single right answer to questions of law even where law is vague. 4 By interpreting a law in its morally best light, in the context of a legal system aiming to uphold the values of the best available political morality, 5 judges can settle borderline cases in a way that respects the rule of law, simply elaborating and enforcing the lessons of the legal system as a whole. Dworkin's fictional judge, Hercules, represents the ideal performance of such a system, where superhuman judges with superb historical knowledge and vast contemporary knowledge use their extraordinary abilities to interpret and apply the best-justified understanding of propositions of law. This imaginative solution is not, however, without limitations. Actual legal systems lack superheroes and may further lack determinate resources to allow individual judges to reach unanimity as to what a reading of a law in its morally best light reveals about how it should be applied in a borderline case.
These two opponents loom larger than life in the debate over law and adjudication, at least in the twentieth century, but they are by no means the only paths to resolution of the problems posed by vague propositions of law. The 'deflationary' approach proposed by Andrei Marmor (and surveyed by Endicott) is still another contender. Marmor builds on Dworkin's idea that a vague proposition of law should be interpreted in [End Page 111] light of its purpose, but suggests that a far simpler method than Dworkin's can be found. Rather than searching for the meaning of a proposition of law in the plain meaning of its terms, or by understanding it in its morally best light, Marmor proposes that we understand it as if it were the product of some definite speaker. 6 This solution requires far less of individual judges and citizens attempting to understand and apply vague propositions of law. It will come as no surprise that this apparently common-sense approach has difficulties and critics, and the debate rages on. Or perhaps it is more accurate to say that the debate sputters on, since there is a cogent line of argument that questions the significance of the contrast between vagueness and the rule of law. Joseph Raz, for example, accepts that some propositions of law are neither true nor false. Yet, because he characterizes the rule of law as but one variably realized virtue of law, he can deny any outright clash between vagueness and the rule of law. The ideal of the rule of law is perhaps in tension with vagueness, but this tension only amounts to an expected, limited, and desirable flexibility in the way vague laws are applied by judicial officials. On Raz's view, 'Conformity to the rule of law is a matter of degree. Complete conformity is impossible (some vagueness is inescapable) and maximal conformity is on the whole undesirable (some controlled administrative discretion is better than none).' 7 There is a keen resemblance between the spirit of Raz's view and Endicott's final conclusion, and here, as in so much else, the substance of interest lies not in the sameness of spirit but in the route Endicott takes to his conclusion. With this overview, we have at least a glimpse of some of the stakes and issues driving the debate over the nature and effect of vague propositions of law and can now turn to Endicott's efforts to sort it all out.
In the midst of this winding debate, Endicott's thesis in Vagueness in Law is a comforting one. Just as we ordinarily suppose, vagueness in laws is bearable, and even problems of higher-order vagueness can be resolved in a manner that leaves the ideal of the rule of law coherent and sustainable. Beneath this modest thesis lies the substantial interest of the book, as it deepens our understanding of precisely why vagueness can be bearable. The core of the book consists of three carefully explored themes: the nature of vagueness, its implications for bivalence, and the subsequent implications of ineliminable indeterminacy for the rule of law. None of these themes is entirely novel. There has been substantial recent work on indeterminacy, truth functionality of propositions of law, and the implications of these ideas for thinking about adjudication and [End Page 112] the nature of legal theory. Brian Bix's Law, Language, and Legal Indeterminacy takes up a similar set of questions and shares both Endicott's sympathy for Wittgensteinian approaches to language and his emphasis on non-linguistic dimensions of vagueness. 8 Endicott's concerns also overlap, at least thematically, with Dennis Patterson's Law and Truth 9 and its evaluation of theories of what makes a proposition of law true. Do we really need another book treating these questions? The answer must be that there is always space for another good book that advances a familiar debate in new directions. Set against that standard, Endicott's arguments deserve serious attention. He takes up questions at the most abstract end of legal theory and brings developments in philosophy of language to legal theory in ways often usefully independent of immediate practical concerns about what judges in fact do with laws. Endicott deepens our understanding of the bearable tension between vagueness and rule of law with great clarity and precision, never succumbing to the risk of being vague about vagueness. Whether his argument goes deep enough into the problem of vagueness in law is a separate question, and I shall add to my exploratory remarks some reservations about the reach of Vagueness in Law.
I Theme 1: The nature of vagueness
Much of the force of Vagueness in Law relies on the interaction of a claim and two related devices. The 'indeterminacy claim' suggests that laws are often indeterminate in their requirements in particular cases. Two devices enable us to explore the implications of this claim for the possibility of the rule of law. The first device is the 'Tolerance Principle,' the idea that terms such as 'heap' can tolerate extension of their use in indiscernibly small steps, just as the removal or addition of a single grain of sand from or to a heap does not change the fact that it is a heap. The paradox of the heap, the 'Sorites Paradox,' illustrates the application of this principle. If we construct a 'sorites series' of individually tolerably small steps of extension of a term, we eventually arrive at the paradoxical conclusion that a heap still exists even when all of the individual grains of sand have been removed. Endicott's most compelling illustration of the paradox is drawn from the Criminal Justice and Public Order Act, which gives British police authority to end rave events whose music is played at a volume 'likely to cause [End Page 113] serious distress' to local citizens: 'The power applies to "a gathering ... at which amplified music is played during the night (with or without intermissions), and is such as, by reason of its loudness and duration and the time at which it is played, is likely to cause serious distress to the inhabitants of the locality."' 10 Even once we overcome the question of what counts as 'music,' the phrase 'serious distress' is vague. As Endicott puts it, 'Somewhere between the silent and the seismic, there is music to which the police power is not clearly applicable, and not clearly inapplicable' (57). Yet the tolerance principle paradoxically allows us to move from end to end of this continuum without any principled way of setting a beginning point for 'serious distress.'
Endicott contends that the puzzle posed by the tolerance principle and our capacity to construct sorites series for vague terms shows important weaknesses in both theories of law and theories of adjudication. The first theme, then, is found in exploring the impact of the tolerance principle and sorites series on attempts to give fuzzy terms (and, perhaps, concepts) sharp edges. Throughout the second and third chapters of the book, Endicott surveys largely philosophical arguments about the nature of and importance of vagueness, introducing early on the metaphor of core and penumbra present first in Benjamin Cardozo's writing 11 and later given wider expression by Hart. Endicott distinguishes usefully between the possibility of radical indeterminacy in application of language and the separate possibility that this kind of indeterminacy might lead to radical indeterminacy in findings of law, such that legal questions might lack a single right answer. There may be special vagueness-reducing mechanisms in law that reduce the dangers posed by linguistic imprecision, so our worries about linguistic vagueness may not correspond directly to vagueness in application of particular laws in particular legal systems. The distinction enables him to take up philosophical concerns without worrying simultaneously about immediate application of philosophical and linguistic insight to questions of law. The second chapter, for example, takes up Derridean deconstruction, demands for context-sensitive interpretation, and the influence of Wittgenstein before rejecting radical indeterminacy of language on at least the grounds that none of the famous defenders of something like the claim of radical indeterminacy in fact turns out to defend anything so extreme in daily life. As Endicott finds David Gray Carlson admitting, deconstruction does not 'deny the phenomenon of successful communication. Without question, every day provides examples of spoken sentences understood by listeners in the way the speaker would have them understood. But there is no guaranty of the successful arrival of the message, because language has a life in context that is beyond the control of the speaker.' 12 [End Page 114] Endicott's investigation works in this overlap between language and the life of language, engaging the ontological and epistemological limits on successful arrival of the largely linguistic normative messages found in laws.
Perhaps the more interesting part of these explorations arrives with Endicott's distinction of no less than ten varieties of vagueness. Each attempts to handle what H.P. Grice identified as those expressions for which 'there are cases (actual or possible) in which one just does not know whether to apply the expression or withhold it, and one's not knowing is not due to ignorance of the facts' or of the meaning of the expression. 13 Accounts and dimensions of vagueness range from the view that it is a matter of (1) imprecision, to Hart's familiar talk of (2) open texture, on to (3) incompleteness and (4) incommensurability and Endicott's coinage (5) 'immensurability,' referring to 'the property that something has if and only if it can be assessed in some respect in which it cannot be measured' (46). Reasonable and substantive disagreement regarding the application of expressions gives rise to the possibility of (6) contestability as a form of vagueness, not entirely dissimilar to the idea of (7) family resemblance as a way to explain the unity amongst prima facie different applications of an expression. (8) Dummy standards require a decision maker to posit a standard within a context established by (9) pragmatic considerations, which, nonetheless, may leave residual pragmatic vagueness where evaluative considerations surrounding the application of an expression do not decisively incline competent application one way or another. Vagueness of this sort remains in a case where we try to say whether someone is, for example, a baseball player and recognize that application of the term is sensible only in light of certain purposes, such as identification of participants in amateur sport or distinction between persons wearing similar uniforms. Finally, Endicott identifies (10) ambiguity as a candidate for generation of vagueness beyond the double meaning of an expression.
II Theme 2: Vagueness and bivalence
These varieties are the fodder for consideration of the implications of vagueness for bivalence about propositions of law: Can we get around vagueness to get to bivalence and certainty in statement of the requirements of laws, as promised under the ideal of the rule of law? The most obvious place to look is to theories of adjudication and law, divided by Endicott into those that suppose that vagueness is always eliminable and those that construe vague propositions of law as neither true nor false and accept that the internal resources of law sometimes run out, with [End Page 115] various consequences. Endicott argues briefly that a vagueness-denying doctrine can be drawn out of Hans Kelsen, then focuses the bulk of his attention on Dworkin's famous contention that there is one right answer to a legal dispute. Dworkin has a standard answer to the objection that there cannot be a right answer where legal rules are expressed in vague terms: additional mechanisms (rules of adjudication or construction) can resolve apparent vagueness by 'requiring that the rule be applied only to cases "in the indisputable core of the language."' 14 All other instances of doubt regarding the application of the rule must amount to treating it as false. In this way Dworkin supposes that he gets around the possibility that some expressions may be indeterminate in the sense of being not true and not false. Endicott adroitly takes up this suggestion that what those such as Hart and Raz (and, in turn and in his own fashion, Endicott) accept as ineliminable indeterminacy amounts to a conjunction of an assertion that a proposition X is not true, and a second assertion that proposition X is not false. But is this really what indeterminacy means? And can Dworkin survive the challenge of indeterminacy so easily while showing that Hart and Raz cannot? According to Endicott, Dworkin's solution is still susceptible to higher-order vagueness even once we accept that certain legal concepts may reasonably be applied as different conceptions of the same concept. Those terms
such as 'cruel' and 'fair' and 'courtesy,' which are Dworkin's examples of concepts which admit of different conceptions, are vague ... They share with words like 'middle-aged' and 'red' the susceptibility to the sorites paradox that leads to the claim that there are indeterminacies in the application of vague words. That problem is not an artefact of a theory of law that suffers from a semantic sting: it is a problem for an interpretive theory too. (72)
There is, however, a germ of good sense in Dworkin's argument, gathered up later in Endicott's consideration of the challenge posed by vague propositions of law to the judicial duty to resolve disputes. Dworkin's argument about the variability of conceptions of the same concept is part of his argument about the duties of judges in the constructive enterprise of adjudication, so, to some extent, Dworkin's tangling with critics on the vagueness of expressions understood as not true and not false leads us away from the real tenor of Dworkin's view on vagueness. On his view, which merges accounts of law and adjudication in a comprehensive theory of law, the object of the theory is not a set of words but the practices of judges. This insight underwrites much of the force of Endicott's escape from the challenge of indeterminacy to the rule of law. [End Page 116]
III Theme 3: Implications of the collapse of bivalence for the rule of law
A final, non-Herculean attempt to eliminate vagueness can be found in the 'epistemic theory' Endicott draws from the work of Timothy Williamson. That theory holds that there are sharp boundaries giving distinct edges to propositions of law, but there are epistemic problems regarding how we are to know just where those sharp boundaries lie. There are two parts to the argument: on logical grounds, the epistemic theory 'does not require us to assert bivalence for all vague utterances, it only requires us not to deny it for any vague utterance.' 15 And where some vague utterance appears to resist attempts to reduce it to bivalence, we must employ the resources driving the semantic argument offered by the epistemic theory, which gives 'meaning as use' fresh expression. According to the semantic argument, 'correct application of words supervenes on the dispositions of speakers' (101), so the task is to identify the relevant community of language users and to discern their dispositions. Endicott provides a series of objections to the possibility of finding a community and set of dispositions that univocally reveal the vagueness-dissolving supervenience relation and proposes a further, more general objection to the epistemic theory, arguing that general evaluative terms carry a kind of necessary vagueness: '"Right" and "wrong" are vague because they are abstract terms used to evaluate practical decisions, and the considerations relevant to practical decisions are very commonly susceptible to sorites paradoxes. There is no precise amount of time that you should spend with your children. There is no precise mean between being stingy and being profligate' (126). The epistemic approach lacks immunity to repeated application of the tolerance principle to generate sorites series, and so must be rejected as unable to overcome higher-order vagueness.
At this point, 150 pages into a 203-page book, we arrive at last at Endicott's positive program: the 'Similarity Model.' This model proposes to accept the tolerance principle, accept the existence of sorites series, and accept the existence of vagueness as necessary imprecision, all while avoiding disastrous conflict between indeterminacy of propositions of law and the determinacy demand of the rule of law. The similarity model builds on the slogan that meaning is use. Meaning is not just any use, Endicott argues: meaning is use of paradigms in a kind of pragmatically warranted fashion. Endicott proposes to deny gaps and sharp boundaries between truth and falsity of vague propositions, asserting instead that '[t]o understand a vague word is to be able to use paradigms, and vagueness is flexibility in their use' (155). The difficult part of the argument comes with the idea of acceptable flexibility, a notion that seems [End Page 117] prima facie liable to the same sort of objection Endicott has mustered against linguistic-ontological accounts of vagueness through repeated application of the tolerance principle to generate sorites series. But the prima facie difficulty of the idea is quickly dispelled. Despite the resemblance of this view to Dworkin's interpretive theory and Marmor's deflationary alternative, Endicott claims to provide something different. The key to his argument is a kind of re-socialization of the activity of interpreting vague propositions of law, leaning heavily on the duty of judges to resolve disputes. Endicott's 'simple account' of interpretation argues that in light of the ineluctability of the problem of higher-order vagueness even when using the best account of vagueness, 'we cannot say that interpretation eliminates indeterminacy in the application of legal rules, or even that it tends to reduce indeterminacy' (158). What we can say in the face of vagueness is that 'the need for a decision is a fundamentally important feature of adjudication and does justify choosing a quietist rave ... But it does nothing to support the notion that the law (or morality) requires a particular choice' (167). This is not meant to open the way to the claim that courts are arbitrary: in ordinary circumstances, determinate rules can be applied, and are applied, without recourse to the special activity of interpretation. But where rules are indeterminate, the courts are in a bind, and some further reason beyond the legal norm must settle its application.
Endicott's rather brief eighteen-page final chapter tackles the tension between the ideal of the rule of law and the fact of indeterminacy. The central argument is quite simple: attention to the ideal of the rule of law soon reveals that it is incoherent if understood bivalently, as a state of affairs that either occurs in some social situation or does not. To take up just one of the aspects of the rule of law explored by Endicott, the rule of law claims to preclude arbitrariness by requiring stability in the body of norms making up a given system. Does the requirement of stability generate a paradox when run against the social need for mechanisms of legal change? Is the rule of law impossible because change is inconsistent with stability? Plainly not. Only an exaggerated conception of the social need for stability supposes that a legal system ceases to exist or lacks legitimacy if some portion of its norms are subject to change. What is important in the rule of law's guarding against arbitrariness is that the change of legal norms is regular - rule-governed and predictable, such that certain arbitrary attempts to change norms can be ruled out. For a legal system to be in 'legally good shape,' as Endicott says (borrowing from Finnis), 16 it is sufficient that these aspects of the rule of law be realized for the most part, most of the time. The ideal of the rule of law is [End Page 118] constituted by vague standards, so there is no sense in which indeterminacy of some norms is necessarily inconsistent with the rule of law (recalling that Endicott accepts the existence of non-interpretive, uncontroversially correct application of legal norms). In fact, a sensible understanding of the rule of law and its variable instantiation in actual legal systems suggests that we must sometimes be prepared to accept that it is a rough standard, realized to a greater or lesser degree. Endicott's closing words are a final shoring up of regularity in life under law, arguing that where indeterminacy enters, it is never justifiable for judges to simply flip a coin: to do so would flout 'the need for judicial discipline' (201). Judges must exercise their duty to resolve disputes with regard to the merits of each dispute and its surrounding factors. Sometimes this may mean taking a counterfactual legislative point of view to the dispute; sometimes this may be impossible, and some other route must be explored. Vagueness may be ineliminable, and even the rule of law may contain vagueness; but vagueness is never a licence to abandon the attempt to justify decisions in light of the variably realized virtues of the rule of law.
IV The reach of vagueness: Does Endicott have vagueness firmly in his grasp?
Endicott's explorations occur in something of a boundaryless lost quarter between philosophy of language, epistemology, ontology, and practical problems of application of vague laws in concrete circumstances. To his credit, Endicott tries to occupy an unambiguous, nuanced point of view in this murky area. His deliberate, systematic approach quickly sets his views apart from naïve alternatives. He explains: 'The indeterminacy claim that I defend is not the notion that there is no right answer to a controversial question,' and, further, 'I do not claim that vagueness is a purely linguistic feature of law' (5). The focus of the book is further settled by Endicott's explicitly staying out of some familiar debates: 'I do not argue in favour of any general position on the old, convoluted disputes among legal positivism, natural law theory, and other anti-positivist theories (or among forms of legal positivism)' (6). These are significant caveats, so it is reasonable to ask what the book really does accomplish once it has made clear what it doesn't set out to do. Is it a piece of practical philosophy? A special kind of advice? Or should it instead be evaluated as a piece of conceptual analysis needing no particular practical applicability as an indicator of its worth?
There is little use of statute or case law in Vagueness in Law. A few references and some thought experiments provide the instances of vagueness up for discussion. The book is plainly not a manual for practitioners, so there is little merit in objecting to its focus on the grounds that real cases of troublesome vagueness do not arise sufficiently frequently [End Page 119] to warrant sustained attention. Perhaps we should instead understand the book as a piece of general jurisprudence or of theoretical philosophy of law. Here a number of concerns should be raised. My first concern arrives with an unavoidable choice on Endicott's part: to take up just some of the questions concerned with vagueness and its relation to laws. Has he taken up the right questions, or, more importantly, enough of the right questions? One way to begin to express the worry that he has not is to point to the distance between the general title of the book, Vagueness in Law, and the quite narrow focus of the bulk of the book on particular laws understood as relatively discrete expressions of norms. Endicott writes as though he accepts as roughly synonymous the ideas of an individuated legal norm, an expression of law, and a proposition of law. Sometimes, of course, the vagueness in individual laws is properly attributed to a subset of the terms of a particular law, as in the phrase 'serious distress' in the example of the quietist rave. It is entirely appropriate for at least this reason to take up arguments regarding vagueness as a largely linguistic phenomenon with accompanying issues in epistemology and ontology. This kind of focus on the legal language carrying an individuated legal norm certainly leaves room for a broader understanding of vagueness in law. Endicott develops just this kind of view when he understands such phrases primarily as objects of adjudication. Once we have socialized language by pointing to constraints on its use such as judges face in the institutional context of courts, we are no longer talking solely about words when we talk about legal vagueness. Endicott's discussion of the 'boundary model' as an alternative to his 'similarity model' is another clear step beyond understanding vagueness as a matter of philosophy of language. Yet even this quite broad focus far beyond the semantic sting falls substantially short of capturing the full habitat of vagueness. I suspect that this omission of scope amounts to a significant deficiency in Endicott's argument. I shall argue below that he cannot explain adequately the compatibility of indeterminacy and the rule of law without extending his discussion to vagueness in the structuring concepts of talk of law and the effect of that vagueness on the rule of law and on conceptions of the scope of the judicial function. I shall make my way to that argument via further consideration of the point of view of Endicott's arguments.
It is a commonplace in legal theory that the longstanding debate over the nature and purpose of legal theory has served in part to alert us to the plurality of useful ways to talk about law. If we reject the proposition that philosophy of law or jurisprudence must at least ultimately, even if not immediately, be a kind of practical philosophy, we can accept that philosophy of law can be a theoretical enterprise without immediate responsibility to practical demands. As the preceding discussion shows, Endicott appears to accept that his arguments can be understood largely [End Page 120] as a theoretical enterprise. Yet once we are this far in our understanding of Endicott's argument and its point of view to law, it is unclear how Endicott's dominant concern is to engage theoretically the application conditions of vague legal propositions. Vagueness of particular propositions of law seems to be a practical matter closely aligned with prevailing judicial practices affecting positions of subjects of legal norms in particular social situations. Often we know with certainty that we have a problem of indeterminacy, rather than mere open texture, only at the point of attempted application of a proposition. Endicott appears to acknowledge the practical nature of the problem in his eventual rescue of the rule of law by appeal to judicial duty to resolve those particular disputes where ineliminable vagueness arises for linguistic, epistemological, or other reasons. (It is a judicial duty of resolution, rather than, say, a legislative duty of determinacy, to which Endicott appeals.) In what sense, then, is his approach to vagueness a theoretical rather than practical approach?
We can borrow here from the useful scheme developed by Jules Coleman in The Practice of Principle 17 to distinguish between three theoretical approaches to law. Conceptual-explanatory accounts aim at a richer explanatory account of characteristically legal concepts and their use in life under law. Causal-functional approaches accept that law performs certain social functions, such as maintenance of social order, and inquire into causal relations in the social operations that produce these social functions. Constructive interpretations merge analysis and normative argument in a way that explains law while simultaneously seeking to understand and evaluate its justification according to the best available understanding of political morality. Read through this scheme, Endicott's arguments seem for the most part to be conceptual-explanatory, taking particular vague propositions as exemplars of general features of language and norms. There may also be a causal-functional account of judicial handling of vague propositions of law in light of the requirements of the ideal of the rule of law, to the extent that Endicott seems to portray dispute resolution as a judicial function that proceeds in certain ways when ineliminable vagueness in propositions of law triggers certain judicial operations. There is also an argument for the vagueness of evaluative standards that verges on a substantive position in moral epistemology. Yet beyond that tendril, there is no morally committed argument regarding vagueness, so it would be unreasonable to read Endicott as offering a constructive interpretation. On the whole, he offers conceptual-explanatory arguments, and in his deployment of those arguments he purports to be able to avoid the debates over the nature of law. In this last delimitation of his task I think Endicott's point of view involves an error that leads his overall argument to miss the mark [End Page 121] of vagueness. The vagueness in law amenable to the conceptual-explanatory approach he adopts is inextricably bound up with those debates in the theory of law that he supposes he can skirt around by limiting his discussion to implications of indeterminacy in exemplar propositions for the ideal of the rule of law. The trouble is that talk of the rule of law as an ideal appeals, at least implicitly, to a surrounding 'web of concepts,' to borrow Raz's phrase. 18 That surrounding web of concepts includes those concepts inhabiting debates in the theory of law regarding conceptual vagueness in the structuring concepts of law and legal system and the resources they provide for making sense of indeterminacy of particular propositions of law. That kind of vagueness is precisely what Endicott has refrained from discussing on the grounds that he need not. But he cannot have it both ways. He cannot both cast his investigation as a theoretical rather than practical one that justifiably selects a very few instances of vagueness in propositions of law as exemplars in order to provide a conceptual-explanatory account of vagueness and then withhold comment on the possible indeterminacy of the structuring concepts of law and legal systems that provide the theoretical backdrop against which talk of 'simple' theories of non-interpretive application of determinate propositions of law makes sense. The appeal to judicial duty simply complicates matters. It inserts a logically contingent, though typically socially evident, practice of interpretation and adjudication into the uneasy truce between exemplar instances of vagueness taken from daily legal life and the overriding concern with vagueness as a general phenomenon in language and norms.
Here the going becomes far more difficult than Endicott allows in his neat carving out of a space for his approach to vagueness, particularly in his use of the idea of what it is for a legal system to be in 'legally good shape.' This phrase is borrowed from a sentence of Finnis's that reads, in full, 'The name commonly given to the state of affairs in which a legal system is legally in good shape is "the Rule of Law" (capitalized simply to avoid confusion with a particular norm within a legal system).' 19 Certainly Finnis, and likely Endicott also (judging by his exemplar instances of vagueness drawn from common law systems), sees the rule of law as the first virtue of a legal system - that unitary object of legal theory most commonly found in the form of the nation-state. Endicott's focus on the effect of indeterminacy of particular propositions of law on the rule of law operates in the aftermath of an assumed answer to the question 'Is this legal system in legally good shape?' This object-level question can only be sensibly asked against the backdrop of a meta-level conceptual- explanatory [End Page 122] account of the concept of law and the web of allied concepts. 20 So Endicott's argument about indeterminacy and its threat to the rule of law seems inextricable from the possible indeterminacy of the concept of law that sits in the conceptual web together with the concept of legal system. And that concept of law - and the idea of a unitary, criterial concept of law setting a test for legal good shape - is precisely what is under challenge today. Raz, for example, argues that
We explain concepts in part by locating them in a conceptual web. These aspects of conceptual explanations can be said to be statements of conditions for the application of the concept only by stretching the idea of a condition for application ... Explanations are of puzzling or troubling aspects of concepts, and they are therefore almost always incomplete. 21
Raz writes further,
One important point reinforces the previous one. There is no uniquely correct explanation of a concept, nothing which could qualify as the explanation of the concept of law. There can be a large number of correct alternative explanations of a concept. Not all of them will be equally appropriate for all occasions. Appropriateness is a matter of relevance to the interests of the expected or intended public, appropriateness to the questions which trouble it, to the puzzles which confuse it. These vary, and with them the appropriateness of various explanations. 22
The varying purposes of conceptual-explanatory accounts of the concept of law (and legal system) lead us only partway to a further kind of meta-level conceptual vagueness handled by theories of law expected to handle, in turn, the problem of vague propositions of law. It may be, as I have argued recently, that some structuring concepts of legal theory are essentially vague because of the method of their construction (and not simply vague as a matter of incompleteness, as Raz suggests). On the relational account of the concept of law I attribute to Hart, 23 the concept of law produced by his descriptive-explanatory approach is essentially vague because it represents the dynamic, shifting relations between law [End Page 123] and social rules, law and morality, and law and coercion. This account of the concept of law proceeds from an ordinary understanding of instances of law to build a much richer account of the concept out of the relations between law and closely associated social phenomena. And here, perhaps, is the root of a more interesting observation about the predilection with paradoxical interaction between legal bivalence and the rule of law: if at least some thick legal concepts underlying the rule of law are essentially vague, so much the worse for the supposed paradox. The steady, implacable opposition is a mirage, and Finnis is correct in his contention, rejected by Endicott, that juridical bivalence is 'a technical device for use within the framework of legal process.' 24 Recognition of the mirage may also make space for causal-functional, naturalized explanations of legal life. Whenever we are stuck with more than a problem of understanding what our commitment to the rule of law amounts to, we need further resources to tackle remaining questions about the nature and reach of the ideas of law and legal system, and a naturalized explanation may be the resource required to extend our understanding beyond the conceptual.
Essential vagueness about the nature of the concept is a matter of concern for deeper reasons too, connected to the exercise of recrafting of the ideal of the rule of law 25 in ways beyond the provision of carefully nuanced accounts (such as Endicott's) of its nature as a variably achieved ideal that collapses only as certain deficits drain its strength. A variety of metaphors have been used to express the increasingly sophisticated normative environment of global social life. Talk of the collapse of 'black box' theories of legal systems carries some intuitive heft, but perhaps not so much as the idea that legal systems are surrounded by a 'semi-permeable membrane,' a kind of film surrounding legal life and separating out those other dimensions of normative life, such as etiquette, that only rarely supply norms worth giving the special force of legal norms. This metaphor seems to be expressed empirically in ways that show just how deeply troublesome it is to reconcile the tension between indeterminate propositions of law and the rule of law by appeal to judicial duty, without tackling at the same time the question of vagueness of the structuring concepts of legal life. Where black box legal systems have gradually become semi-permeable membranes that accept and build into themselves new normative resources while rejecting others, the idea of a judicial duty to resolve disputes suddenly becomes so complex that I suspect it of being rather less useful in its middleman role than Endicott [End Page 124] indicates. Endicott purchases a particular view of the sustainability of a particular expression of the rule of law in discrete legal systems by appeal to judicial duty to resolve disputes. But is this duty in the new, post-black box legal world what it always was? 26
Consider how the concept of law, the instantiation of the ideal of the rule of law, and the duties of judges change with the rise of the restorative justice movement within the paradigmatically discrete legal system of Canada. 27 If we take prospectivity as a characteristic of the ideal of the rule of law sustained by judicial practices, there is prima facie a deficit in the rule of law in Canada, where an objective understanding of criminal wrongdoing 28 is expressed by use of 'Sentencing Circles,' particularly in Aboriginal communities. An initial finding of wrongdoing is given its final characterization as a wrong demanding redress by a collective assessment of the wrong done an actual, particular community. Here there is a tension between the general ideal that legal prohibitions against criminal wrongdoing ought to specify prospectively, and with reasonable precision, the legal consequences of wrongdoing and the dimension of restorative justice, which renders justice not as a prospectively available measurement but as an activity that expresses the virtue of a particular community. It is an attempt to restore that community's integrity while returning the wrongdoer into the community with fresh understanding and acceptance of the virtues of the community. 29 It is not much of an objection to this claimed tension to say that even an offender facing restorative justice has a prospectively guaranteed expectation of facing up to a sentencing circle. The content of 'facing up to a sentencing circle' is so complex and variable, depending on particular factors in particular communities, that prospective knowledge of the activity of the [End Page 125] sentencing council is at least epistemically unavailable to the prospective wrongdoer and may be altogether unforeseeable. The hallmark of restorative justice in the form of sentencing circles is its capacity to capture the particularity of wrongdoing, in both the character of the wrongdoer and the harm done a particular community, in ways beyond the capacity of abstract prospective measurement of wrongdoing in conceptions of justice lacking a restorative element. 30
The addition of restorative justice activities to the Canadian legal system raises questions about the identity of the Canadian legal system and about the content of the web of concepts that underwrite the system. But more immediately, restorative justice has a limiting effect on the power of the idea of a judicial duty of resolution and the centrality of the adjudicative function in an account of the concept of law. Talk of a judicial duty to resolve disputes makes sense on the assumption of a capacity to resolve disputes. Where that capacity is thinned, so is the sense in which judges can reasonably be held accountable for performing a duty of resolution as a matter of judicial discipline. Restorative justice brings the potential for a judicial handing-off of application of vague propositions of law to devices, such as sentencing circles, whose particularist application of those propositions produces a resolution whose character may owe very little to judicial input. The judicial duty to resolve might be exercised, then, in a very thin fashion as a kind of referral, followed by certification of the results of the activity of restorative justice carried out as a result of the referral. When a vague proposition of law (e.g., 'serious distress') threatens the rule of law in situations of this sort where restorative justice is the activity of applying the vague proposition of law, the judicial duty of resolution is unlikely to guarantee consistent results in repeated encounters of that proposition. This is possible precisely because the adjudicative function is sidelined in favour of the expressive activity of a community restoring justice in itself. Whether this shows a further threat to the ideal of the rule of law is of some interest, but it seems to me that the more troublesome dimension of legal life brought out by this example is connected with pluralism about the concept of law and allied concepts. What occurs in Canada when sentencing circles are used is a complex, continuing reconstruction of the concept of law and the conception of the Canadian legal system and the place of the ideal of the rule of law within it. Where multiculturalism and claims to recognition are blended into law, there is a challenge to the ideal of the rule of law altogether, and not merely controversy [End Page 126] regarding application of the ideal. In the terms of autopoietic analysis, 31 normative talk that was mere 'noise' outside the system of meanings comprising the Canadian legal system is brought into the system, expanding and changing individual meanings within the legal system and changing the identity of the system itself. If all of this is plausible, it seems to me to weigh in favour of positivist descriptive-explanatory approaches to law that have no difficulty accepting widely varying purposes and ideals occupying life under law, including a rule of law in which determinacy is not the pre-eminent feature. I suspect that Dworkinian constructive interpretation is, for its part, weakened by shifting ideals of the kind identified above, which impair its ability to provide a plausible story of law unified by a commitment to the best justified political morality. But these issues are for another day. Perhaps most importantly for a final evaluation of Endicott's arguments, it seems that these reflections on the reach of vagueness lead us quite naturally back to foundational questions about how to characterize law and legal system. No matter whether we side with positivism or anti-positivism, we find ourselves facing again the questions raised in the debate Endicott supposes he has evaded.
I have gone quite far beyond Endicott's avowed concern with the challenge
posed to the rule of law by vague laws. Yet that is just my intention. I
have not disputed the internal logic of Endicott's arguments, and again
I should say that Vagueness in Law is very much worth reading,
carefully and more than once. There is a great deal to be learned from
Endicott's patient dissection of vagueness in propositions of law. Where
I diverge, I do so to point to inadequately acknowledged complications
in Endicott's too-brief reconstruction of the ideal of the rule of
law. Vagueness as an ontological and epistemological puzzle extends
far into the structuring ideas of legal life, and an adequate account
of vagueness in law must at least acknowledge, if not investigate,
the reach of vagueness.
* I owe thanks to Michael Giudice for his remarks on a draft of this essay.
1. Canadian Charter of Rights and Freedoms, s. 1, Part I of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (U.K.), 1982, c. 11.
2. For an account of their concerns see N. Duxbury, Patterns of American Jurisprudence (Oxford: Clarendon Press, 1995). Chapter 2, 'The Evolution of a Mood,' is especially useful as a nuanced account of the Legal Realist ideals.
3. H.L.A. Hart, The Concept of Law, 2d ed. (Oxford: Clarendon Press, 1994).
4. R. Dworkin, Law's Empire (Cambridge, MA: Belknap / Harvard University Press, 1986).
5. As Dworkin writes in more general terms, 'in order to make of it [the object of interpretation] the best possible example of the form or genre to which it is taken to belong.' Ibid. at 52.
6. A. Marmor, Interpretation and Legal Theory (Oxford: Clarendon Press, 1992).
7. J. Raz, 'The Rule of Law and Its Virtue' in J. Raz, The Authority of Law (Oxford: Clarendon Press, 1979) reprinted in K. Culver, ed., Readings in the Philosophy of Law (Peterborough, ON: Broadview Press, 1999) 13 at 22.
8. B. Bix, Law, Language and Legal Determinacy (Oxford: Clarendon Press, 1993).
9. D. Patterson, Law and Truth (Oxford: Oxford University Press, 1996).
10. Vagueness in Law at 57, citing Criminal Justice and Public Order Act 1994 (U.K.), 1994, c. 33, s. 63(1).
11. Ibid. at 8, citing B.N. Cardozo, The Nature of the Judicial Process (New Haven, CT: Yale University Press, 1921) at 130.
12. Vagueness in Law at 16, citing D.G. Carlson, 'Liberal Philosophy's Troubled Relation to the Rule of Law' (1993) 43 U.T.L.J. 257 at 282.
13. Vagueness in Law at 31, citing H.P. Grice, Studies in the Ways of Words (Cambridge, MA: Harvard University Press, 1989) at 177.
14. Vagueness in Law at 63, citing R. Dworkin, 'No Right Answer?' in P.M.S. Hacker & J. Raz, eds., Law, Morality and Society (Oxford: Clarendon Press, 1977) 58 at 68.
15. Vagueness in Law at 100, citing T. Williamson, Vagueness (London: Routledge, 1994) at 193.
16. Vagueness in Law at 184, citing J. Finnis, Natural Law and Natural Rights (Oxford: Clarendon Press, 1980) at 270.
17. J. Coleman, The Practice of Principle (Oxford: Clarendon Press, 2001).
18. J. Raz, 'Two Views of the Nature of the Theory of Law' in J. Coleman, ed., Hart's Postscript (Oxford: Clarendon Press, 2001) 1 at 10.
19. Finnis, Natural Law and Natural Rights, supra note 16 at 270.
20. An alternative question that might asked using the same evaluative standard of 'good shape,' but with respect to the concept of law rather than to any particular legal system, is this: 'Is law in legally good shape?' But that question, which Endicott does not ask, generates complex questions about properties, predication, and identity, among other things, and is in any case simply unhelpful as a question likely to produce practically or theoretically interesting results. The only viable question Endicott's argument allows him to ask, then, is the question about particular discrete legal systems.
21. J. Raz, 'Two Views of the Nature of the Theory of Law: A Partial Comparison' (1998) 4 Legal Theory 249 at 257.
23. K.C. Culver, 'Leaving the Hart-Dworkin Debate' (2001) 51 U.T.L.J. 367.
24. Vagueness in Law at 72, citing Finnis, Natural Law and Natural Rights, supra note 16 at 280.
25. E.g., D. Dyzenhaus, ed. Recrafting the Rule of Law: The Limits of Legal Order (Oxford: Hart Publishing, 1999).
26. Let me mark here, without examining it in depth, the question of whether Endicott can really sensibly talk of judicial duties, given his abstention from practical talk of particular duties in particular systems. It seems that Endicott's point of view limits him, rather, to talk of institutional forces bearing on judges, some of which may be claimed to be judicial duties while others might include the weight of certain interpretive traditions. The point of this criticism is that Endicott's wavering point of view on vagueness in law leaves unclear the nature of the tool of judicial duty used to extricate the rule of law from the threat posed by indeterminacy.
27. We might call this an example of internal legal pluralism within the nation-state, setting aside for the moment the equally interesting questions of the effects of external legal pluralism in international and transnational legal contexts.
28. The emphasis on objective dimensions of wrongdoing typically exists even with respect to wrongdoing traditionally requiring a subjective fault element.
29. See, e.g., Canada, Department of Justice, The Aboriginal Justice Strategy, online: Department of Justice <http://canada.justice.gc.ca/en/ps/ajln/strat.html> (last modified: 24 April 2003); and 'Restorative Justice: A Program for Nova Scotia' at http://www.gov.ns.ca/just/rj/rj-contents.htm, instantiated in, e.g., the Cape Breton Island Community Justice Society <http://www3.ns.sympatico.ca/icjs/>.
30. On these questions generally see, e.g, A. Duff, Punishment, Communication and Community (Oxford: Oxford University Press, 2000).
31. See, e.g., Z. Bankowski, 'How Does It Feel to Be on Your Own: The Person in the Sight of Autopoiesis' (1994) 7 Ratio Juris 254; N. Luhmann, 'Law as a Social System' (1989) 83 N.W.U.L.Rev. 135.