Special Issue on Lon Fuller || Implicit Law

G E R A L D J. P O S T E M A

IMPLICIT

LAW

In his Anatomy of Law, Lon Fuller warns his readers that theorists of m o d e r n law are especially susceptible to a certain kind of theoretical blindness. Partial to "the intellectual flavour of made ,,1 aw, they attend exclusively to laws enacted by authorized lawmaking institutions and identified, interpreted, and enforced by authorized law-applying institutions. Consequently, they overlook the vast body of law lying beneath the surface of the phenomena they seek to understand. Yet the existence and content of explicit laws depend on a network of tacit understandings and unwritten conventions, rooted in the soil of social interaction. We cannot fully understand law, we cannot appreciate its dynamic character, the role it plays in human affairs, or the kind of public good it offers, if we fail to attend to this implicit dimension of law. O f course, other contemporary legal philosophers have also recognized that explanations of the nature and functions of law cannot focus on law's explicit dimension alone. Hart, for example, argued that the lawmaking power of legal institutions, and the laws they create, presuppose power-conferring rules that exist not in virtue of some further lawmaking act, but in virtue of a fundamental rule o f recognition implicit in the practice of lawapplying officials. 2 Dworkin's jurisprudence also recognizes a vast reservoir of principles "implicit" in the practice of law, principles that follow from the best interpretive theory of explicit law. 3

i Lon Fuller, The Anatomy of Law (Westport, Conn.: Greenwood Press, 1968), p. 57. (Hereafter referred to as AL.) 2 H.L.A. Hart, The Concept of Law (Oxford: Clarendon Press, 1961), chapters 4, 6. 3 Ronald Dworkin, Law's Empire (Cambridge, Mass.: Harvard University Press, 1986), chapters 3, 7. Law and Philosophy 13: 361-387, 1994. © 1994 Kluwer Academic Publishers. Printed in the Netherlands.

362

Gerald J. Postema

However, Fuller's understanding o f the nature and scope o f law's implicit dimension sets him apart from these legal theories. His theory is a form o f jurisprudential conventionalism, but not o f the sort that often goes by that name in recent jurisprudential discussions. We have reason to call it common law conventionalism in virtue o f its affinities to the c o m m o n law tradition o f jurisprudence. 4 Unlike versions o f conventionalism that take social conventions concerning (or arising from) formal institutions o f lawapplying and lawmaking to hold the key to jurisprudence, Fuller argues that m o d e r n law is continuous with, and fundamentally dependent upon, informal social practices, s This may be Fuller's most important jurisprudential legacy. Its value lies in its rich insights into connections between fundamental social and moral phenomena and basic features o f law. In this essay I will concentrate on these insights, rather than on his critique o f rival jurisprudential theories. M y aim is expository and constructive. I shall try to articulate Fuller's thesis concerning the implicit dimension o f law, and to reconstruct his most important arguments for it. In the process I hope to highlight some important but underappreciated features of Fuller's conception o f law. I. A.

I M P L I C I T RULES, I N T E R A C T I O N ,

A N D LAW

Implicit Rules and Social Interaction

To understand what Fuller has in m i n d w h e n he speaks o f the implicit dimension o f law consider, first, his distinction between

4 I have tried to identify important features of that tradition in Bentham and the Common Law Tradition (Oxford: Clarendon Press, 1986, 1989), chapters 1 and 2. 5 However, Fuller is as critical of the romantic's reduction of law to folk custom as he is of some positivists' blindness to implicit law (AL, p. 70if). Even though I will not mention it any further, readers must keep this side of Fuller's argument in mind throughout the remainder of my discussion.

Implicit Law

363

made rules and implicit rules. 6 Made rules are general normative propositions conceived prior to and projected onto conduct for some purpose to which, in the intention o f the rulemaker, the rule directs conduct. Made rules are given canonical verbal formulations by a determinate author at a reasonably precise date. T h e practical force o f made rules depends on the authority o f their makers or the offices they o c c u p y ] Thus, made rules presuppose both authors and relations o f authority and subordination. In contrast, implicit rules arise from conduct, not conception. Verbal formulations may more or less accurately capture the rules implicit in the conduct, but the formulations are always post hoc and strictly answerable to the conduct. N o formulation is authoritative in virtue o f its public articulation alone. Although implicit rules arise from the conduct o f determinate agents, typically they have no precise date o f birth and no determinate authors. They presuppose no relations o f authority and subordination; thus, their practical force depends neither on authority nor on enactment, but on the fact that they find "direct expression in the conduct o f people toward one another. ''8 W h i l e they arise from conduct, implicit rules are not mere regularities of the behavior in view; they are rules for that behavior. To use Hart's familiar terms, these rules have an "internal" as well

6 AL, p. 44. This distinction is meant to apply to all sorts of rules, including but not limited to legal rules. 7 As we shall see in the final section of this essay, Fuller argues that the practical force of legal rules, even enacted rules, cannot depend entirely on relations of authority. 8 Lon Fuller, "Human Interaction and the Law," in Fuller, The Principles of Social Order, Kenneth Winston, ed. (Durham, N.C.: Duke University Press, 1981), p. 212. (Hereafter referred to as HIL.) Note that Fuller's point here is that implicit rules do not depend for their content or practical force upon, and can exist in the absence of, relations of authority and subordination. Nevertheless, as we shall see presently, Fuller holds that rules are also implicit in and structure relations among legal officials and between officials and ordinary citizens.

364

Gerald.]. Postema

as "external" aspect. T h e y are "brought into being and kept alive by purposive effort and by the way each o f the parties interprets the purposes o f the other.''9 Typically, implicit rules are side-products of intentional, rational interaction. T h e y arise from and are sustained in this interaction. Purposive action in a social context is inevitably interactive; thus, agents choose actions in part based on their interpretations o f the aims and expected behavior o f those with w h o m they interact. In interactive social contexts, individuals can achieve their own purposes only if relatively stable mutual expectations o f behavior are available. From these "intermeshing anticipations" implicit rules emerge as focal points around which persons who must coordinate their actions form reliable expectations o f the behavior o f others knowing that their counterparts are doing the same thing. Thus, implicit rules arise from and draw their practical force from the interdependence o f expectations and aims. 10 This essentially interactive dimension o f implicit law is the hallmark o f Fuller's jurisprudence. Fuller never developed a fullscale theoretical model o f this interactive and coordinative dimension o f social life, n but he identified several o f its important features and built his account o f the implicit law on them. I will mention a few o f these features. First, implicit rules emerge over time from a process of mutual accommodation and adjustment o f expectations and actions o f interacting agents. Second, the rules represent relatively stable points in the network, and this stability

9 Lon Fuller, The Morality of Law, 2d edition (New Haven, Conn.: Yale University Press, 1969), p. 195. (Hereafter referred to as ML.) 10 "Customary law [Fuller's paradigm of implicit law] arises, then, out of situations of human interaction where each participant guides himself by an anticipation of what the other will do and will expect him to do" (AL, p. 73; see also HIL, pp. 213-15, 219, 226-27). 11 For an attempt to do so with the help of machinery from game theory see my "Coordination and Convention at the Foundations of Law," Journal of Legal Studies 11 (1982): 165-203. (But see my caveat in the next footnote.)

Implicit Law

365

makes possible successful coordination o f action. 12 Moreover, the rules function in the social lives o f the agents concerned as facilitators o f successful interaction. Fuller regards this feature as an essential, if only formal and partial, purpose o f such rules. Third, the practical import o f the rules - by which I mean both their content and their reason-giving force - is a function o f their place in this network o f intersecting expectations, purposes, and actions. O f course, agents, or even observers, may give verbal expression to such rules, but their ability to guide action is dependent on the substratum o f ongoing practice. This is w h y formulations o f the rules are always post hoc and strictly answerable to the conduct from which they arise. Fourth, this substratum will rise from time to time to the surface and, through the parties' relatively explicit awareness o f it, play a vital role in determining solutions to their problems. Because parties engaged in interaction frequently face new situations and n e w practical problems, they must adjust, reinterpret, or even replace existing rules. These revising activities are possible only because the parties develop shared abilities to negotiate the n e t w o r k o f expectations and together to arrive at new, relatively stable understandings. Participating in such practices we learn h o w to anticipate the solutions our counterparts will hit u p o n as they attempt to anticipate our decisions and actions. Interpretation and extension or contraction o f existing rules in such a practice, then, is itself an essentially interactive process. 13 Finally, the successful introduction o f a new rule or a new understanding o f an established rule always depends on a wide sense o f its reasonableness. "Reasonableness" here is a function of integration o f the rule into the network o f

12 HIL, p. 215. Fuller usually speaks of"interaction" rather than "coordination." I shall use these two terms interchangeably, with the caveat that we must not equate "coordination" as I use the term here with its narrower technical use in game theory. For example, Fuller does not assume that agents who face a problem of interaction/coordination are strictly indifferent to the alternative solutions available to them. 13 ML, pp. 227-29.

366

Gerald J. Postema

expectations and interdependent purposes and intentions o f the agents involved. TM This integration is never strictly determined by what each party accepts as right or reasonable for her own part only; rather, it depends on what each can anticipate that others regard as reasonable in this sense. B.

Law and Interactive Practices

We must now link this account o f the implicit and interactive nature o f social rules to law. Fuller's general thesis is that implicit, interactive practice is pervasive in all legal systems - not just in those heavily dependent upon "customary law," but even in those apparently dominated by enacted law and formal lawmaking and law-applying institutions. We can distinguish three broad contexts o f interaction in which practices o f the sort Fuller describes are in evidence. 1~ The first context is essentially "horizontal," where interaction occurs b e t w e e n parties related as "equals." In the simplest cases these relations are bilateral, as in contractual arrangements or business partnerships; but they can also be complex, multilateral relations, as in relations amongst citizens in a political community. The second context involves interaction among parties related "vertically," as authorities to subordinates; for example, as officials and players, judges and litigants, and law-givers and subjects. In the third, interaction occurs between or among authorities or officials themselves. Here relations may be vertical, as between constitution framers and law-givers, or higher and lower courts, or horizontal, as b e t w e e n courts at the same level, or a complex mixture o f both, as between lawmakers and courts. To demonstrate the pervasive presence o f implicit law in m o d e r n legal systems, Fuller's argument proceeds at two levels simultaneously. First, he discusses examples o f the explicit law's dependence on implicit interactive practice in the above three contexts. For example, he illustrates law implicit in first-level, horizontal inter-

14 HIL, p. 227. is Here I follow "Coordination and Convention": 183ff.

Implicit Law

367

actions w i t h an e x t e n d e d discussion o f c u s t o m a r y law 16 and the implicit elements in contract law. iv H e also describes a wide variety o f third-level interactions, for example, b e t w e e n c o n s t i t u t i o n drafters and legislators, TM and b e t w e e n legislators and courts. 19 Similarly, he argues that courts depend on implicit practice insofar as a doctrine o f precedent governs their deliberations. 2° Fuller also describes at length the d i m e n s i o n s and implications o f s e c o n d level interdependence b e t w e e n legal officials and citizens. 21 At the same time his a r g u m e n t goes further. H e claims n o t just that dependence on implicit interactive practice is a c o m m o n feature o f m o d e r n law, but also that law is not possible w i t h o u t it, that it is essential to our idea o f legal order. His most explicit a r g u m e n t for this claim concerns second-level interaction. H e argues that

HIL, pp. 212-24; AL, pp. 44-47, 71-84. 17 HIL, pp. 224-30. Fuller has in mind not the law of or about c o n 16

tracts, but, as he puts it, "the 'law' a contract itself brings into existence" (HIL, pp. 224-25). Contracting parties "make" this law through their agreeing to the terms of written contract. However, Fuller maintains, the terms of the contract have practical import only in the context of the parties' practice, including the wider commercial practice in which they take part. Interactional regularities of standard commercial practice, for example, often set or influence the meaning of terms of the contract, especially when standardized clauses are used (HIL, p. 225). Even more, the specific interactions of the parties can "evidence . . . what courts call a practical construction of their agreement" which may force an interpretation of the contract contrary to the ordinary meaning of the words of the written document (HIL, p. 226). Indeed, "the written contract often furnishes a kind of framework for an ongoing relationship, rather than a Precise definition of that relationship." The written contract, Fuller argues, rests for its definition on "a kind of two-party customary law implicit in the parties' actions" (HIL, p. 226). is AL, pp. 63-68. 19 AL, pp. 57-59; ML, pp. 231-32. See also Lon Fuller, "The Role of Contract in the Ordering Processes of Society Generally," in his Principles of Social Order, Winston, ed., p. 177. 20 AL, pp. 46-47. 21 HIL, pp. 234-35; ML, pp. 206-20.

368

Gerald J. Postema

the existence o f a legal order depends on effective interaction and cooperation between citizens and lawmaking and law-applying officials. 22 1 will call this Fuller's vertical interaction thesis. From this thesis he draws the important corollary that law is "the product o f an interplay o f purposive orientations between the citizen and his g o v e r n m e n t . . . [not] a one-way projection o f authority originating with government and imposing itself upon the citizen". 23 He also offers a second, and in some ways more profound, argument for the dependence o f law on implicit interactive practice. He argues that a substantial degree o f congruence between enacted laws and background informal social practices and conventions governing horizontal relations among citizens is necessary for the existence o f law. I will call this Fuller's congruence thesis. In the next section o f this essay I will elaborate these two theses and reconstruct Fuller's argument for them. II.

A.

LEGAL O R D E R A N D S E L F - D I R E C T E D SOCIAL INTERACTION

Vertical Interaction

According to Fuller, vertical interaction takes two forms. First, he argues that the meaning and content o f laws depends on interaction between citizens and officials. The argument begins with the observation that we expect that citizens' understanding o f what the law requires o f them will determine at least in part their decisions and actions. But that understanding depends on their expectation o f how lawmaking and especially law-applying officials are likely to understand it. Similarly, officials authorized to enact, interpret, and apply the laws must anticipate how citizens will take up the laws they make or interpret, that is, h o w the rules are likely to figure in the practical reasoning o f citizens.

22 HIL, pp. 234-35; ML, pp. 39-40, 192-93, 209, 219. 23 ML, p. 204.

Implicit L a w

369

Otherwise, they will not be able to direct or guide actions in such a way as to achieve the substantive aims o f the law. Second, Fuller argues that compliance with the laws thus understood also depends on a kind o f reciprocity: On the one hand, the law-giver must be able to anticipate that the citizenry as a whole will accept as law and generally observe the body of rules he has promulgated. On the other hand, the legal subject must be able to anticipate that government will itself abide by its own declared rules when it comes to judge his actions. 24 Thus, effective functioning o f legal order, Fuller maintains, depends on a "cooperative effort - an effective and responsible interaction - between law-giver and subject". 25 Fuller insists that this cooperative effort and the "relatively stable reciprocity of expectations" that result are "part o f the very idea o f a functioning legal order". 26 If this reciprocity breaks down, or is greatly compromised, legal order will break down. This is not to say that radical social disorder would inevitably follow, but it is to say that if some kind o f social order is maintained it would lack a feature essential to law. For it could no longer adequately provide baselines for self-directed h u m a n interaction which, according to Fuller, is the central task o f law. 27 Analysis o f this notion o f self-directed social interaction will make clear the outlines o f Fuller's argument for this vertical interaction thesis. To begin, consider what is involved in law providing a basis for self-directed action. First, law regulates or guides actions o f citizens by addressing reasons or norms to them. R a t h e r than

24 HIL, p. 235; see ML, pp. 2•7, 219. 2s ML, p. 219; see ML, p. 192. Implicit in this interaction are what Fuller

calls "the internal morality of law" or the "implicit rules of lawmaking," about which there has been much debate over the years. It is rarely noticed, however, that Fuller seeks to ground these principles in the essential interactive dimension of law. 26 ML, p. 209. 27 ML, p. 210; HIL, p. 234.

370

Gerald J. Postema

altering the social or natural e n v i r o n m e n t o f action, or m a n i p u lating (non-rational) psychological determinants o f action, law seeks to influence behavior by influencing deliberation. It addresses norms to agents and expects t h e m to guide their actions by those norms. Moreover, it expects those norms to figure in deliberation, not as contextual features setting the environment or parameters o f choice, but as reasons for deliberate choice. Thus, rules are intended to be " i n t e r n a l " in two respects: (a) they figure in the deliberation o f agents, and (b) t h e y figure as reasons for, and n o t merely parameters of, deliberation and choice. 2s Second, law seeks to influence deliberation in a wholesale fashion, not t h r o u g h detailed step-by-step instructions, but through general norms that agents must interpret and apply to their specific practical situatio1:s. 29 For an agent to act in a self-directed way, on this view, is for that agent to find in a rule a reason for acting in a certain way. This requires that the agent be able to grasp the practical i m p o r t o f the n o r m , and this in t u r n involves a capacity to w o r k o u t correct applications o f the n o r m and a capacity to appreciate the reason for acting in the way indicated and to relate it in an appropriate way to other reasons the agent m i g h t have. Fuller's thesis is that it is law's fundamental task to provide base-

28 As I note below, Fuller holds that legal norms regulate only the manner in which citizens pursue their goals; they do not prescribe their goals. This view is consistent with the claim that legal norms are internal in the two respects, in particular with the second, mentioned in the text. For a consideration can figure in deliberation as a reason for one's action rather than as part of the environment of deliberation or action ("parameter") without being a complete reason for one's action. Viewed as reasons for acting in a certain manner, legal norms are partial reasons, needing to be filled out with some specification of the end of the action. They provide reasons for choosing to pursue one's end in one way rather than another, but, of course, one would not have reason to act in this way at all, if one did not think that it was a way of promoting one's end. Nevertheless, partial reasons still function as reasons rather than merely parameters of deliberation. 29 ML, p. 210; HIL, p. 234.

Implicit Law

371

lines for self-direction. Law can do so only if law-subjects are able to d e t e r m i n e for themselves the practical import o f its n o r m s in most instances. 3° It is this feature o f law that has led some philosophers to think that law by its nature secures a k i n d o f freedom. If one is governed by law, one is in one sense free from subjection to the will o f a n o t h e r person; for law leaves to subjects to determine its application to particular circumstances w i t h o u t having to be told on each occasion by s o m e o n e else (a superior) h o w to a c t . 31

Since law largely addresses self-directed action a m o n g and often in concert w i t h other h u m a n beings, the law's central task is that o f providing citizens with a sound and stable framework for selfdirected interaction. 32 This adds a further i m p o r t a n t d i m e n s i o n to the law. If self-directed action is to be possible, each agent must be reasonably confident that the practical i m p o r t o f the norms he or she finds will correspond w i t h that f o u n d by o t h e r agents. 33

30 Fuller also associates a third element with self-direction. He maintains that law's directives do not seek to define, even in general terms, goals or objectives for citizens to achieve, but rather merely furnishes baselines against which to organize their lives (HIL, p. 234, ML, pp. 207-9). Law regulates only the manner in which citizens pursue their goals but it does not define those goals for them. This seems to me to be a dubious claim, if it is put forward as universally or necessarily true of law. For the idea that law has a tutelary function has a distinguished pedigree and, even if it is objectionable, it does not seem incoherent. However, I believe nothing in the argument to follow critically hangs on this part of Fuller's conception of self-directed action, and so I do not consider it any further here. 31 While the idea is familiar to readers of Rousseau and Kant, its roots can be traced to a medieval idea of freedom. See J.R. Lucas, "The Phenomenon of Law," in Law, Morality, and Society, P.M.S. Hacker and Joseph Raz, eds. (Oxford: Clarendon Press, 1977), p. 94, citing R.W. Southern, The Making of the Middle Ages (London, 1953), p. 108. 32 ML, pp. 210, 222; HIL, p. 234. 33 This does not apply, of course, to entirely solitary rational action, if such is ever available to us. Fuller clearly assumes throughout his discussion that individual action takes place in a social context.

372

Gerald J. Postema

So a critical part of understanding the practical import of a n o r m consists in understanding h o w it will be understood by others, how it will figure in their practical deliberation, recognizing that they will be doing the same regarding one's own understanding. Fuller insists that the deliberation involved in self-directed action is necessarily interdependent. But, then, if law is to facilitate selfdirected social interaction, its norms must be c o m m o n , public norms. Citizens must be able to proceed on the presumption that others with w h o m they interact are governed by the same norms. Citizens must be able to grasp the norms and appreciate their practical force, confident that most other citizens will on most occasions read them roughly the same way. That is to say, the practical import o f legal norms must be publicly accessible. In view of this understanding o f the complex purpose defining law, Fuller's argument for the vertical interaction thesis is very straightforward. Cooperative interaction between officials and citizens is necessary, he argues, if law is to succeed in providing guideposts for self-directed social interaction. 34 Coordination o f selfdirected social interaction will be effective only if most of the parties involved can reasonably expect that most o f the others will identify and understand the applicable rules in roughly the same way they do, and can be counted on for the most part to comply with them. Thus, law-givers must shape the rules they enact or interpret in anticipation o f how citizens are likely to understand, and expect their fellow citizens to understand, the language they use and the decisions they make. Likewise, citizens will understand announced rules and decisions in light o f how they expect officials and their fellow citizens to understand and apply the rules. Moreover, Fuller argues, law-givers will see no point in enacting a rule that they cannot expect citizens to follow; similarly, citizens will see no point in complying with rules that officials probably will not follow or enforce. Facilitating self-directed social interac-

34 HIL, pp. 234-35.

Implicit Law

373

tion depends, therefore, on "a relatively stable reciprocity o f expectations between law-giver and subject". 3~ O n this view o f the dynamics o f law, the interaction between officials and citizens is best seen as an extension of the deliberative interaction already at w o r k among citizens. Hart's familiar example o f the official scorer in a game illustrates the point well. 36 The decisions and actions o f the players are intelligible only in the context o f a set o f mutually accessible rules, the understanding o f which is c o m m o n and known to be so. Introducing officials with the power to make final decisions does not alter this important fact. An official's decisions are intelligible only if they coordinate with the shared understanding o f the rules among players and scorers alike. As J.P,.. Lucas has observed, the point is not merely that an observer, fan, or player is doing in an unofficial and non-authoritative way the same as what the official scorer does, but that it is an essential condition of the intelligibility of the official scorer's activity that he is doing, only in an official and authoritative way, the same as the unoffcials are doing, and what they are doing is something which, although for convenience sake made the responsibility of an official scorer, can essentially be done by unoffcials at large. 3v T h e same is true o f law. Legal officials and citizens are engaged in the same kind o f interdependent activity. For this reason, Fuller concludes, "the creation o f an effective interaction between [officials and citizens] is an essential ingredient o f the law itself". 38

B.

The Congruence Thesis

The account o f self-directed interaction we have just considered provides materials for defense o f an even more robust claim about the implicit dimension o f law, the congruencethesis. Fuller's argument, in outline, is that legal norms and authoritative directives can 35 ML, p. 209. 36 Hart, Concept of Law, p. 138ff. 37 J.R.. Lucas, "The Phenomenon of Law," p. 94. 38 ML, p. 193.

374

Gerald J. Postema

guide self-directed social interaction only if they are broadly congruent with the practices and patterns o f interaction extant in the society generally. This is not to say that particular legal norms can never seriously conflict with informal social norms or values, even less that a single norm is invalid if it conflicts with background social values. But it is to say that a legal order could not survive, and its norms would not be intelligible to their addressees as laws, if those norms were systematically at odds with or radically isolated from ordinary, informal social practices and conventions. Let us see h o w we might fill out this argument from considerations we rehearsed above. Observe, first, that if legal norms were broadly congruent with and integrated into ordinary social practices and conventions, it is likely that their practical import would be sufficiently public for purposes o f self-directed interaction. 39 For it is reasonable to expect that those practices and conventions provide c o m m o n resources on which members o f society can draw, and can expect their fellow members to draw, in negotiating their interactions. These practices and the social capacities learned through participating in them provide a sufficiently c o m m o n background against which to interpret the law's norms and to appreciate their practical force. Techniques for determining the practical import o f legal norms w o u l d then be natural extensions o f techniques and capacities already learned and publicly available to law-subjects in their informal social interactions. O f course, this observation does not yet demonstrate the truth o f the congruence thesis, for it holds that the c o n g r u e n c e o f legal norms with ordinary social practices and conventions is necessary for self-directed interaction. At best we have suggested that the congruence is likely to be sufficient. We can defend the necessity claim if we can show that self-directed interaction would 39 This is true, that is to say, if self-directed interaction can ever be facilitated by public norms. Skepticism about this possibility is, of course, fashionable these days. Fuller, however, was never afflicted by the disease. This is not the place to attempt its cure.

Implicit Law

375

be u n d e r m i n e d if legal norms were isolated from, and substantially incongruent with, ordinary social practices. Suppose, then, that legal norms were isolated from and substantially incongruent with ordinary social practices. What resources could law-givers draw upon to assure to law-subjects the practical import o f legal norms? C o m m o n linguistic resources will be available by which they can hope to fix the meaning o f the norms; and they can assure the practical (reason-giving) force o f the norms through either (a) sanctions added to the promulgated norms, or (b) appeals to the authority o f the law-giver, or (c) appeals to the reasonableness o f compliance with a publicly articulated set o f norms. These seem to be sufficiently public to facilitate self-directed interaction, but it appears that no other resources are available. Thus, the requisite publicness o f the practical import o f the legal norms could only be the product o f the ordinary linguistic meaning o f the words used to formulate the norms plus one or another "external" sources of practical reasonableness. 4° Now, to defend the congruence thesis we must show that this strategy will undermine self-directed social interaction. Consider the following argument. Fuller argues that the content o f legal norms cannot be determined (for the purpose o f providing guideposts for self-directed social interaction) strictly from the linguistic meaning o f the words in which they are formulated. 41 Lawmakers construct legal norms in the abstract, but the norms are "projected into" the developing life o f the community, and they are practically intelligible only if they are successfully so projected. Thus, understanding a n o r m requires not merely understanding the meaning o f the words in which it is formulated, but understanding the institutions, practices, and attitudes of the community to which it is addressed. Enacted norms make sense as practical guides for 40 These reasons are "external" to the norms in question because, while they may supply reasons for complying with the norms, they do not typically supply reasons for following them. 41 Fuller's defense of this claim is suggested in his discussion of an ordinance prohibiting vehicles in the park (AL, pp. 56-58, 66).

376

Gerald.]. Postema

self-directing agents (that is, are followable as norms) only w h e n they are set in the context o f concrete practices, attitudes, and forms o f social interaction. Usually, the context is presupposed: precisely because we readily recognize, and expect others with w h o m we regularly interact to recognize, the circumstances and practices into which a n o r m is projected, its abstract formulation is no obstacle to its making public practical sense to us. To use Hart's familiar metaphor, legal norms have a "core" o f settled meaning, as well as a less determinate "penumbra," not merely in virtue o f the settled meaning of their linguistic components, but in virtue o f a standard and commonly recognized background o f attitudes and practices in which those linguistic components take on practical sense. Thus, the content o f legal norms cannot practically be determined strictly from the linguistic meaning o f the words in which they are formulated. Moreover, it may not be possible in practice for an agent sharply to distinguish between her grasp o f the meaning or content o f a norm and her sense o f its reasonableness. To make practical sense o f a norm, that is, for an agent to determine what it requires o f her in particular circumstances, she must make sense o f her acting on it. Persons charged to follow or apply an abstract norm that appears arbitrary or pointless w h e n projected into concrete circumstances are likely to regard the n o r m as practically unintelligible, and will be inclined to formulate or interpret it in a way that makes it at least minimally intelligible. We can only draw standards of intelligibility for norms that can play the public role legal norms do from c o m m o n sources of social interaction, namely, the social practices and conventions into which the norms are projected. But, then, the practical force or reasonableness of such norms must be to this extent "internal" to the norms. Practical force supplied by sanctions or authority or some other external source may adequately motivate compliance with norms once it is publicly clear what complying with them consists in, but they cannot help in determining the public content o f those norms. If this argument is sound, then, it will not be possible to provide norms for self-directed social interaction relying solely on c o m m o n

Implicit Law

377

linguistic resources and external motivations. For with only those resources to draw on, law-subjects will find it very difficult if not impossible to work out for themselves the implications o f these norms for their own particular circumstances, confident that the reading they give the norms is likely to be sufficiently close to readings others with w h o m they must interact give them. Lawsubjects will be dependent on decisions and determinations by formal institutions and officials for each particular case. Some degree o f such intervention, o f course, is inevitable, but w h e n the norms are cut off from all c o m m o n resources for interpreting and appreciating the practical import o f norms, self-directed action will be radically undermined. T h e citizens' situation would be like that of players caught up in "scorer's discretion. ''42 They must shift attention from the game they had been playing to what they can only regard as the ad hoc deliverances o f officials. Players, once focused together on their interaction structured by the rules they mutually recognize, must now focus on the decisions o f officials. N o longer able to guide their actions by their own determination o f what the rules require o f them, they become dependent on the "wills" o f the authorities. Thus, law cannot hope to facilitate self-directed social interaction unless it remains substantially congruent with and integrated into ordinary social practices and conventions in the community it serves. Fuller's conclusion, then, is that enacted, pedigree validated, authoritative norms ("made law") represent only the surface phenomena o f law. Law has a social depth which we must recognize if we are to understand adequately the nature and modes o f functioning o f these salient surface phenomena. By its very nature, law is deeply implicated in the practices and conventions o f the communities it governs. This is true to a m u c h greater degree, Fuller argues, if the legal system in question recognizes "adjudicative law" as well as enacted law, as do C o m m o n Law jurisdictions. Because o f two distinctive features, adjudicative law "projects its

42 Hart, The Concept of Law, pp. 138-44.

378

Gerald J. Postema

roots more deeply and intimately into human interaction than does statutory law." (1) It "makes" law case by case, and (2) it is an essentially "collaborative" enterprise drawing into its circle not only the whole judiciary but also actual and potential litigants. 43 Let us consider briefly these two features. Adjudicative law comes into being through the determination of particular disputes. This process differs from statutory lawmaking in several important ways. 44 F o r example, three elements kept sharply distinct by statutory law are thickly intertwined in adjudicative law: deciding the particular case before the court, articulating the rule on which the decision rests, and justifying that rule in deeper and broader matters of principle. The decisions and rules of adjudicative law do not stand apart from their justifications, but are parts of a single process of reasoning. A judicial decision, says Fuller, "is always an explained thing", and in virtue of this there is a standing invitation to contemporary and future judges and citizens to participate in the same process of deliberation. Consequently, the law made by a precedent court is not regarded as a finished product, but as amenable to reformulation in light of further reflection and experience. Interpretation of the rule of a past decision always involves thinking through the past court's practical problem again, looking behind the words of the precedent to what the court was trying to say and do. The words of the opinion are a gateway to the "unwritten law" lying behind it. Thus, the process of articulation-cum-justification of the rules of adjudicative law enlists the efforts not only of the deciding court, but also of subsequent courts and the actual and potential litigants who formulate arguments for their specific situations by reference to the precedent decisions. Moreover, these precedents not only offer "a place of c o m m o n anchorage" for argument and delibera-

43 HIL, p. 236; AL, pp. 93, 101, 109. 44 AL, pp. 89-93.

Implkit Law

379

tion, but also "a shared point o f embarkment toward new law". 4s Subsequent courts reshape past decisions to meet the demands o f emerging social conditions and needs. Proceeding case by case, the collaborative enterprise o f adjudicative law "fit[s] and refit[s] its prescriptions to the configurations o f life as they reveal themselves in litigation". 46 For this process to go on with even minimal efficacy and integrity it must maintain communication with "standards implicit in the social environment in which it functions," for they provide "the raw stuff o f [adjudicative] law". 47 Moreover, without any such c o m m o n standards or experience, litigants would not be able to participate meaningfully in the collaborative project, for it would not be possible for them to reach meaningful joinder o f issue (ibid.). However, Fuller argues, it is not necessary that the environment o f social practices and conventions secures broad agreement on general values. It is sufficient if it can provide "shared standards by which differences can be isolated, discussed meaningfully, subjected ultimately to some rational resolution". 48 A society so sharply divided that it did not even share this degree o f c o m m o n experience could not hope to maintain a legal order through adjudicative law. Similarly, a legal order o f adjudicative law could not function if it is cut off from the environment o f social practices that make meaningful, collaborative articulation o f c o m m o n , public rules possible. It remains to note that this discussion o f the vertical interaction thesis, and especially the congruetlce thesis, and Fuller's arguments for them, reveal a deep "conventionalist" strain in Fuller's jurisprudence. Moreover, it is not difficult to see that this conventionalism draws its inspiration from c o m m o n law experience.

4s 46 47 48

AL, p. 96. HIL, p. 237. AL, p. 109. AL, p. 110.

Gerald J. Postema

380 III.

T H E P L A C E OF A U T H O R I T Y A N D F O R M A L I N S T I T U T I O N S IN LAW :

T h e arguments we have explored in the previous section rest o n the premise that the central, defining function o f law is to facilitate first-level self-directed social interaction - "facilitating interaction as traffic is facilitated by road signs". 49 Yet, this premise is open to the objection that, although some laws may have this as their primary aim, laws often serve m a n y different purposes besides facilitating interaction. Indeed, some laws, for example antitrust regulations, seem bent on putting roadblocks in the path o f this type o f social traffic, preventing rather than promoting interaction. Fuller acknowledges the facts pressed b y this objection, s° but he believes that they caneasily blind, us to the deeper and m o r e pervasive coordinative features o f law. For even roadblocks can enhance traffic flow by routing traffic around hazards or trouble spots that in their absence would snarl traffic) 1 Similarly, licensing o f physicians enables individuals t o locate qualified medical attention with greate r ease and far lower cost than otherwise w o u l d be possible. Perhaps the best illustration o f this point is anti-trust legislation. The narrow aim of the legislation, o f course,is to disrupt coalition building among producers - i.e., to make this kind o f cooperation more difficult - but its larger aim is to enhance that

49 HIL, p. 231. so He mentions laws prohibiting prostitution, homosexual practices, gambling, use of drugs arid other "crimes against victims" as examples of the law's attempt to prevent rather than facilitate interaction ( H I L , pp. 231-33). He suggests that these laws are notoriously ineffective because they work against the law's primary facilitative function. I find this argument unpersuasive. There are many reasons we might cite to explain the special difficulties posed by these laws; moreover, the same "counterfacilitating" aim seems at work in anti-trust legislation to vchicb Fuller's observations do not seem applicable. 5~ I am grateful to Kenneth Winston for calling this point to my attention. --

Implicit Law

381

special kind o f cooperative activity characteristic o f market exchange, s2 More generally, Fuller argues that not far beneath the surface o f laws directed to aims other than facilitating social interaction there is often a large and indispensable coordinative element, w h i c h is nearly invisible to us because o f the law's success in achieving coordination. We can gain a sense o f this dimension if we view law in its historic development, or look to yet unsettled areas o f the law. s3 These replies are plausible, I think, but, ultimately, they are insufficient to support Fuller's strong thesis. For they show only that facilitation is a more extensive aim o f law than may be generally appreciated, but that falls short o f his claim that it is the fundamental task o f law to facilitate self-directed social interaction. However, Fuller has a quite different reply to the above objection. Observe, first, that the above objection cites as counterexamples to Fuller's thesis other, sometimes incompatible, concrete purposes that laws are often put to. Fuller might reply that this misunderstands the theoretical level at which his premise is working. H e does not deny that laws may be put to a myriad o f concrete purposes; rather, he calls attention to the more general fact o f the role law plays in the lives o f citizens. His premise calls attention

52 Typically, we think of markets as forms of competition, but the competitive activity common to market exchange is possible only within a larger context of cooperation and is more plausibly regarded as a certain kind of cooperative activity. This seems to have been the view pressed by economist Frank Knight, with whose work FullEr was very familiar. On Knight's view of markets as schemes of cooperation, see Jules Coleman, "Competition and Cooperation," Ethics 98 (1987): 76-90, esp. 84-87. s3 He argues that, historically, people looked to criminal law to organize and coordinate their collective response to wrongs that threatened their society, and even today "interactional issues" remain at the center of the debate, for example, over how to interpret conditions of self-defense, or killing to prevent a felony, for here juries are forced to face the question: What can reasonably be expected of a person in these situations? (HIL, p. 232).

382

Gerald J. Postema

to the social function o f law, not the concrete purposes that lawgivers have in mind w h e n they enact legislation. Admittedly, Fuller's use of "purpose" is misleading; but clearly he does not wish us to view law at this level o f generality as an instrument designed to serve substantive aims. For that way o f viewing matters immediately invites the questions: whose purposes? and who designed it? It is precisely the appropriateness o f these questions that he challenges, s4 The "purpose" he has in mind is not the purpose o f anyone in particular, but, rather, the distinctive point o f law, the function it performs generally in the lives o f citizens. Viewing law from this perspective invites us to raise a question we tend otherwise to beg: H o w are the explicitly identified legal norms o f this system or practice related to formal institutions o f lawmaking and law-applying and to the broader social practices o f the community? A number o f contemporary legal theorists have challenged the classical positivist view that coercion is a central and defining feature o f law. ss In the same way, Fuller challenges the centrality o f the notion o f authority in contemporary legal theories. Clearly, both coercion and authority are salient features o f m o d e r n legal systems. The argument, however, is that to put them at the center o f our conception o f law distorts our understanding o f key features o f it. For example, our understanding o f the nature of legal obligation and o f legal powers is distorted if we make coercion the focal point; similarly we distort customary law a n d t h e interdependence o f officials and citizens if we make authority the focus. Just as, according to Hart, sanction theories o f legal obligation distort the

54 I think this marks an important difference between Fuller's thesis~ concerning the facilitative task of law and the otherwise very similar view embraced by Bentham. On Bentham's commitment t o this -task see Bentham and the Common Law Tradition, chapters 5.3-5.4 and "Bentham on the Public Character of Law," Utilitas 1 (1989): 47~49. ss Hart, The Concept of Law, chapter 3.2,: 5.2.; J. Raz, Practical Reason and Norms, second edition (Princeton, NJ.: Princeton University Press, 1990), chapter 5.3. " -_--

Implicit Law

383

role of law in the practical reasoning of citizens, so, according to Fuller, the equation o f law with a set of authoritative directives identified by their sources in declarations o f formal institutions o f law distorts the role law plays in the practical reasoning of citizens. T h e mistake in both cases, we might say, is to identify a peripheral albeit salient feature o f the practice with its core. In a familiar section o f The Concept of Law, Hart highlights the essential features o f law by comparing a m o d e r n legal system with a "pre-legal society" governed only by social rules. ~6 T h e important difference, he says, lies not in the primary rules governing members o f the two societies, but in the structure of secondary rules and the institutions they define, rules that have as their principal function the maintenance o f primary rules. O f special importance in Hart's story are law-applying institutions and the social rule of recognition implicit in the ordinary practice of its officials. Fuller, I suspect, regarded this story as insightful, but incomplete. Because it is incomplete, it may lead readers mistakenly to identify law with the activities or declarations o f formal legal institutions. As if to correct the course, he proposes a contrast between law and "managerial direction. ''sv We can best u n d e r stand his portrait o f this distinction not as an attack on the positivists' c o n c e p t i o n o f law, but as a warning against drawing the contrast with informal social practices too sharply or in the wrong place. As Hart's story makes us aware of the institutional dimensions of law, Fuller's distinction isolates features of law that enable us to locate those institutions in the practice of law more generally. "Managerial direction," as Fuller conceives it, has five salient features: (1) It presupposes relations of authority, of superiors and subordinates, dominated by vertical relationships among the parties, w h e r e h o r i z o n t a l relationships a m o n g authority-equals are o f secondary importance. (2) Superiors address directives to their

56 Hart, Concept of Law, chapter 5.3. s7 See, for example, ML, pp. 207-9.

384

Gerald J. Postema

subordinates. (3) These directives specify means to achieving the superior's ends. (4) Subordinates apply these directives to serve their superior's ends. (5) While the superior's directives may take the form o f general rules, the superior's decision to formulate his directive in this way is merely tactical; nothing in principle stands in the way o f managerial direction by specific orders alone) 8 In his essay, "Freedom as a Problem o f Allocating Choice, ''59 Fuller tells a just-so story o f a kind o f tyrant's progress. It is the story o f the entirely unintentional transformation o f a system o f managerial direction into a structure o f law. The tyrant, seeking effectively to employ his subjects for the realization o f his own projects, finds reasons at each stage to grant to t h e m increasing powers o f choice and self-direction. First, he learns that, in view o f the enormous costs to him o f designing instructions for his subjects that will enable them, without further reflection or interpretation, to carry out his wishes, he can more efficiently achieve these purposes by granting them powers to interpret and apply his directives. Next he learns that they work more willingly and efficiently if they are granted scope to pursue their own ends as well as work to promote the tyrant's. Finally, he learns that they can be even m o r e effective if they have opportunities to develop their powers beyond those strictly d e m a n d e d by their assigned jobs. Fuller argues that with each attempt by the tyrant m o r e effectively to serve his own ends the tyrant grants greater scope to the self-direction o f his subjects. At the end o f this progression, the tyrant has ushered himself, qua tyrant, out o f existence, and in his place is installed a legal regime. Managerial direction is replaced by law, and subordination to the will o f the tyrant by self-directed interaction.

58 HIL, p. 214. ~9 L. Fuller, "Freedom as a Problem of Allocating Choice," Proceedings of the American Philosophical Society 112 (1968): 101-6. The tyrant's progress is told at 105-6. Kenneth Winston brought this little-known essay to my attention.

Implidt Law

385

Fuller draws the contrast between managerial direction and law as follows: In the first instance, law is concerned not with relations between authorities and citizens but with relations among citizens. 6° Moreover, the "law-abiding citizen . . . does not apply legal rules to serve specific ends s& by the law-giver, but rather follows them in the conduct o f his own affairs, the interests he is presumed to serve in following legal rules being those o f society generally". 61 Generality is an intrinsically necessary feature o f law's distinctive technique o f governance, for, as We have seen, the "law does not tell a man what he should do to accomplish specific ends set by the law-giver; it furnishes him with baselines against which to organize his life with his f e l l o w s " . 62 Admittedly, Fuller overstates the contrast between ~hese two kirids o f social ordering, just as Hart's distinction between the "pre-legal society" and a society with a full-fleclged legal system ignores a spectrum o f social orderings between the two. It is also unfortunate that Fuller makes the contrast in terms o f whose good is being served, the superior's or the citizens'. This is unfortunate, I believe, because it unnecessarily opens his account to irrelevant objections. His central point can be made even if we stipulate that the superior's aims are restricted to the community's good, or even to providing each citizen as m u c h liberty as possible for pursuing his or her own ends, and making possible their pursuit o f such ends together. For the essential difference lies riot in the content o f the specific ends or purposes that may be served in complying with directives, but rather in the relations among the directives, the directive-givers, and the directive-appliers or directive-followers. While Fuller admits that there are elements o f managerial direction in m o d e r n legal systems, and law-like elements in structures o f managerial direction, the difference between these two modes o f social ordering is a fundamental difference o f kind.

60 HIL, pp. 207-8. 6~ ML, p. 207. 62 HIL, p. 234.

386

Gerald J. Postema

The component elements - general rules, issuers o f those rules, other officials charged with interpreting and enforcing the rules, and followers or appliers o f the rules - are connected in fundamentally different ways. Each m o d e o f ordering has a different "purpose" or point and so these components are related in different ways. The reason for their interaction is different, and so both its importance to law and the kind o f social good it represents very different. 63 Managerial direction puts authority at the center o f the picture and defines the other components in terms o f this fundamental vertical dimension. Generality o f the directives, for example, is regarded as useful insofar as it serves the superior's ends. Similarly, horizontal relations among subordinates will be secondary to achieving the superior's ends. However, in law, vertical relations o f authority and the activities o f officials, while always important and often prominent, are secondary to the structure o f horizontal relations among citizens regarded as self-directing rational agents aware o f and capable o f adjusting their decisions and actions in view o f the rational deliberations, decisions, and actions o f others. In this form o f social ordering, generality is not an expedient but an essential ingredient. The model for this relationship is not, o f course, that o f boss and employee, or commander and soldier, but neither is it that o f the private arbitrator of a dispute between two largely unrelated parties, 64 rather, it is that o f the referee introduced into a game. As Fuller puts it, the task o f law is that o f "providing the citizenry with a sound and stable framework for their interactions, the role of government [that is, officials and formal institutions o f authority] being that o f standing as a guardian o f the integrity o f this system". 65 This is the fundamental contrast Fuller has in mind. Thus, while Fuller would agree with Hart that what is distinc63 64 68 65

ML, pp. 209-10. See, for example, Joseph Raz, "Authority, Law, and Morality" Monist (1985): 297-99. ML, p. 210.

Implicit Law

387

tive o f societies with mature legal systems are the formal institutions o f lawmaking, law-applying and law-enforcing, and perhaps even formal criteria o f validity, he would hasten to add that we cannot understand law's distinctive approach to social ordering if we focus exclusively on the operations o f these institutions. For this focus is likely to lead us to misunderstand the place and role o f such institutions in a society's legal practice, encouraging us to see law simply as "a one-way projection o f authority". 66 Law is not the product of authority but its necessary precondition. Authority is compatible with law w h e n it, and its directives, are rooted deeply in the social interaction o f the communities it purports to serve. For only then can it enable and enhance the self-directed action o f its citizens. Department of Philosophy, The University of North Carolina at Chapel Hill, Chapel Hill, NC, U.S.A.

66 ML, p. 204.