Jump to ContentJump to Main Navigation
Philosophical Foundations of the Nature of Law$

Wil Waluchow and Stefan Sciaraffa

Print publication date: 2013

Print ISBN-13: 9780199675517

Published to Oxford Scholarship Online: May 2013

DOI: 10.1093/acprof:oso/9780199675517.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 22 June 2018

Law as Power

Law as Power

Two Rule of Law Requirements*

(p.129) 6 Law as Power
Philosophical Foundations of the Nature of Law

Bruno Celano

Oxford University Press

Abstract and Keywords

This chapter considers the shape taken by social power when the rule of the law is envisaged as an ethico-political ideal. It discusses the Rule of Law as a specific mode of the exercise of social power, and what is valuable in it. It focuses on two Rule of Law requirements, consistency and compliability. It argues that they contribute to defining a distinctive mode of social power, one that shows respect for its subjects. Power can be effectively exercised by systematically flouting these two desiderata. Consistency and compliability are required, however, if the law is to treat its subjects as autonomous, responsible agents.

Keywords:   Rule of Law, power, normative consistency, ought, legislation, prescriptions, rational lawmaking, paternalism, publicity

1. Introduction

“Law as power” is a somewhat neglected topic in contemporary analytical jurisprudence. To be sure, attention has been paid, from Hart (and Kelsen) onwards, to normative legal powers (i.e. norm-creating and norm-applying powers, instituted by legal power-conferring norms). “Brute” social power, however, and law’s relation to it, are, in post-Hartian jurisprudence, largely ignored.1

This will be my topic in this chapter. I shall discuss, that is, the way the law operates as a mode of the exercise of social power. I approach the issue from a specific angle (section 2). The subject of my inquiry is the shape social power takes when the rule of the law is envisaged as an ethico-political ideal. I discuss, that is, the Rule of Law as a specific mode of the exercise of social power, and what is valuable in it (a preliminary characterization will be offered in section 3).

I concentrate on two Rule of Law requirements, consistency (roughly, the avoidance of conflicts) and compliability (roughly, conformity to the principle “ought” implies “can”) (section 4). These two desiderata contribute, I argue, to defining a distinctive mode of social power, one that is characterized by its showing at least a minimum of respect for the dignity of its subjects. As we shall see, power can be effectively exercised by systematically flouting these two desiderata. Entire power structures may flourish around vast areas of inconsistent or non-compliable directives. Consistency and compliability are required, however, if the laws are to treat their subjects as autonomous, responsible agents (sections 4, 5). Under both (p.130) respects, it turns out (section 6), power exercised through the Rule of Law qualifies as public, rational, and non-paternalistic.

2. Social Power through the Lens of the Rule of Law

Law’s standing vis-à-vis social power is a central issue in legal theory—witness the works of Weber, Kelsen, Ross, and, in Italy, N. Bobbio. Not to mention, of course, Hobbes, Bentham, Austin, or Marx.

One obvious way of approaching the subject is by investigating the relations between law and coercion, and the different ways in which law may be coercive. One does not have to endorse the (implausible) view that coercion is the differentia of law as contrasted with other kinds of social order,2 nor the claim that coercion is an essential feature of law, in order to grasp the significance of this connection—law operating as a means of social power through its being coercive. Even if we grant that coercion is not a necessary feature of law, it would be wrong, I think, to deny jurisprudence the task of clarifying the issue. For this to be a legitimate task for (general) jurisprudence it is sufficient, I think, that typically (epi to polu) law is coercive.3

Be that as it may, my focus is not, in this chapter, on the coerciveness of the law. What I am interested in, is law understood as a distinctive mode of the exercise of social power. This evokes grand sociological generalizations—e.g. the law as one of the manifestations of the state’s monopoly of legitimate force; law as an instrument of ideological hegemony, or social oppression; the law as the organization of force in society, involving the institutionalization of the use and the threat of physical force. I am going to approach the issue, however, from a different angle.

The object of this chapter is the shape social power takes when the rule of the law is envisaged as an ethico-political ideal. By “the rule of the law as an ethico-political ideal” I mean what is usually called “the Rule of Law” (RoL, for short). I understand the RoL in the formal, “thin,” sense,4 which has now become usual among jurisprudents. So understood, the RoL comprises a loose cluster of (1) formal features of the laws (prospectivity, publicity, relative generality, relative stability, intelligibility and relative clarity, compliability, consistency), plus (2) institutional and procedural desiderata (such as, for instance, that the making of individual norms, applying to individual cases, be guided by general rules).5 The RoL will be the lens through which I will look at my subject. I shall consider the RoL, in the specified sense, as a distinctive mode of the exercise of social power.

(p.131) Before we start, two clarifications are in order. (1) My concern is, first, with what the RoL, regarded as a distinctive mode of social power, is—what its main features are. And, second, with what is valuable in it: what the values are that this kind of power serves, and how. (Under this respect, mine will be a substantive ethico-political inquiry.) I sidestep the issue of whether the concept of law entails the idea of the RoL, or perhaps some minimal version of it. The answer to this question is, to my present purposes, immaterial. (2) The subject of this chapter is the law as a vehicle of “social power,” or of “brute” power (as contrasted with legal normative powers; see section 1). What should be meant by these terms?

Power is, of course, a puzzling, highly problematic notion. The word is polysemic. According to some,6 the concept of power—power in society—is an essentially contested concept. In what follows, I shall rely on an intuitive understanding of what social power is, what kinds of social relationships do qualify as power relationships, and so on. There are, in fact, accounts of social power which impose very stringent conditions on the applicability of the notion. According to some such austere accounts, some of the “varieties of power” that I shall mention may not qualify as forms of power at all. A fully-fledged deployment of my argument would require a fully developed account of this notion. As I said, I shall not attempt this here. I rely on an intuitive idea, no doubt vague, imprecise, generic, of what power relations in society are, and of their varieties. Thus, I stipulate, very roughly, that “social power” consists, basically, in this: someone exercises a measure of control over someone else’s behavior, thought, or feeling, in the pursuit of his ends or interests (which may include controlling someone else’s behavior, thought, or feeling). My hope is that the legitimacy of this intuitive, unrefined notion, and of the characterization of the phenomena we shall encounter as varieties of “power”, will be apparent.

Thus, the concept of social power will remain unanalyzed. It will be used as a primitive. A word, however, is needed about the characterization of the family of power relations I shall discuss as forms of “brute” social power. In this context, “bruteness” is to be understood as relative:7 the relevant forms of power may be said to be “brute” relative to the layer of normative powers (powers to create and apply norms) instituted by legal power-conferring rules. This does not mean that they may not be complex or institutionalized phenomena in their own right, relative to further, more basic kinds of facts.

3. Rule of Law Power: A Preliminary Characterization

Laws meeting RoL requirements may have almost any content. What is peculiar, as regards the RoL, is, however, the form that the exercise of power takes. The RoL is, in the first instance, a specific mode of the exercise of power.

(p.132) It is not at all unusual to understand the RoL as a specific mode of the exercise of power, in particular, political power. In order to grasp its specificity we have to consider what other modes of the exercise of power, political and generally social, it should be contrasted with.

As a specific mode of power the RoL is usually contrasted with “managerial governance or rule by decree.”8 Or it is contrasted with “arbitrary” power, meaning by this either power not exercised in conformity with the rules it should conform to;9 or public power exercised in the pursuit of private interests.10 It is also often contrasted with “manipulation,” or forms of “manipulative” power.11

I take my cue from this last suggestion. RoL power will be contrasted, in what follows, with varieties of “manipulative” power.

“Manipulation” is a generic, vague, and ambiguous word, carrying evaluative and strong emotive connotations. In order to render the contrast between RoL power and “manipulative” varieties of power meaningful, and illuminating, we have to make its meaning determinate. I propose to do this by stipulating two conditions, which our account of RoL power—and, by contrast, of the “manipulative” varieties of power it is, ex hypothesi, opposed to—will have to satisfy. The first condition concerns the descriptive import of the word. The second concerns its evaluative implications.

(1) “Manipulation” evokes the idea of treating human beings as things, to be modified, altered, etc. as one sees fit, irrespective of their beliefs, attitudes, or will. Conversely, a non-manipulative form of power (e.g. the RoL) will have to be understood as characterized by the treatment of its subjects as sources of deliberation, capable of making choices in the light of their beliefs and attitudes, and of acting on the basis of these choices.

Admittedly, this does not say much, and can easily be shown to be insufficient. (So, for instance, when X makes Y do A by relying on Y’s belief that in circumstances C one ought to do A, and by deceiving Y so that he mistakenly believes that circumstances C do in fact obtain, isn’t X “manipulating” Y’s behavior?12) But it will be enough as a preliminary characterization. We shall make it more determinate as we go on.

(2) “Manipulation” is, presumably, power which does not respect the dignity of its subjects.

It is a widely held opinion that the RoL is a mode of power characterized by (a minimum of) respect for the dignity of human beings, or at least that “observance of the rule of law is necessary if the law is to respect human dignity.13 Our account will have to explain in what sense this is so, and why.

(p.133) The first condition gives us the beginning of the required explanation. Presumably, “manipulative” power does not respect the dignity of its subjects because it treats them as things. This, however, is only the beginning of an explanation. We have to show in what ways subjects, under the RoL, are not treated as things, and how—and to what extent—these contribute to an attitude of respect for the dignity of human beings.

So, RoL power will have to be understood, in contrast with “manipulative” forms of power, as power that, first, treats subjects as centers of deliberation and that, second, in so doing (pro tanto) respects their dignity as human beings. I try to show that this is so, and why, by focusing on two RoL requirements, consistency and compliability.

4. Two Rule of Law Requirements

How does law’s conformity to these two requirements contribute to defining a distinctive mode of social power—one that is not “manipulative,” and thus (pro tanto) respects the dignity of those over whom it is exerted?

The conception of the RoL adopted here (section 2) is mainly instrumental. (Fuller’s tale concerning the different ways in which Rex may fail in his enterprise aptly illustrates this kind of approach.14) The RoL is understood, basically, as comprising what is instrumentally necessary in order to achieve a certain goal.

Which goal? A first, approximate answer is: guiding human behavior. So, for instance, directives typically have to be prospective, and intelligible (prospectivity and intelligibility are, as we have seen, two among the various RoL desiderata), if they are to be capable of guiding human behavior.

This is, however, generic and imprecise. There are many ways of guiding (taking the word “guide” at its face value) human behavior. Most of the requirements of the RoL (in particular, those in the first group; section 2) specify, more or less directly, what is instrumentally required in order to achieve the end of guiding human behavior through rules.15

These two ends—“guiding human behavior,” generally, and “subjecting human behavior to the guidance of rules”—are different. The latter is a specification of the former: subjecting human behavior to the guidance of rules is a specific way of guiding it. As we shall see, the difference between these two goals is crucial in explaining how RoL power satisfies the two conditions set out in the preceding section (not treating subjects as things and, to this extent, respecting human dignity).

What is involved, then, in “subjecting human behavior to the guidance of rules,” as the proper end which accounts for RoL desiderata? One obvious ingredient is generality: the word “rule” designates, we may assume, a general directive. But this (p.134) is only part of the answer. By focusing on consistency and compliability, rather than generality, I mean to explain further facets of what is involved in the idea of subjecting human behavior to the guidance of rules, in its difference from the idea of guidance, as such.

In other words. “Subjecting human behavior to the guidance of rules” is the basic purpose grounding, in its main outlines, the RoL. In order to understand what the RoL is, and what it requires, we have to get this purpose right. What is involved in it, in addition to the idea of guidance through general directives, and what follows from this? Focusing on the two requirements, consistency and compliability, is a way of answering this question.

In understanding what the end which explains RoL desiderata is, and how it differs from that of guiding (without specification) behavior, the crucial point is what may be termed “normal,” or standard, guidance: that is, guidance through an understanding, by the subject, of what the law, according to its tenor, requires of him, and that it is required of him. The RoL comprises a set of conditions which have to be satisfied for the normal, standard guidance (so defined) of human behavior.

The simplest case is given by the idea of “telling people what they should do” (and, maybe, threatening them of unpleasant consequences in case of non-compliance). When X tells Y to do A, typically, X tries to guide Y’s behavior in the direction of doing A through the understanding, by Y, of what is required of him (namely, doing A), and that it is required of him. This is the standard case. All sorts of non-standard cases—non-standard forms of guidance—can be devised,16 as will be borne out by our examination of the requirements of consistency and compliability (below, in this section, and in section 5). In accounting for RoL desiderata, and for their point, the general form of the explanation will have to be, in outline: it is necessary, in order for the laws to achieve their standard purpose—guiding human behavior, according to their tenor, through their being understood by their subjects—that they satisfy condition C, for such-and-such reasons.

So, the task is to show why consistency and compliability are instrumentally necessary for standard guidance, and what follows from this. We have to ask ourselves, first, what kinds of power may be served by the systematic violation of these two requirements, and, second, in what ways these may turn out to be “manipulative” in character, and thus prove disrespectful of the dignity of those who are subject to them.

In answering these two questions, I shall work with a simplified model of the RoL. I regard the RoL as an ethico-political ideal concerning the relations between a lawgiver (in the widest sense) who issues prescriptions and those who are subject to these prescriptions17—an ideal, then, concerning the shape prescriptive (p.135) relationships (i.e. the kind of relationship which comes into being, by virtue of the happy issuing of a prescription, between a lawgiver, on the one hand, and those to whom her prescription is addressed, on the other hand) should take. How to transpose this model to a more complex reality (some legal norms, we may plausibly assume, are not prescriptions) is, admittedly, a problem.18

This is a “legislative” conception of the RoL. At least prima facie (but see in this section, sub (2)) legislating plainly is the issuing of prescriptions. So understood, then, the RoL is the rule of legislation. What follows concerns only this version of the RoL.

My justification for adopting this model (apart from the gains of simplicity) is twofold.

  1. (1) In the jurisprudential tradition from which the conception of the RoL adopted here (“thin” RoL; section 2) stems, the RoL is, in fact usually understood along these lines (witness, for example, Lon Fuller’s tale of Rex, which is the matrix of these theorizations),19 as a matter of the conditions enacted directives have to meet if the purpose intended by those who have enacted them is to be achieved.

  2. (2) In contemporary legal systems, the role of legislation is paramount. “Legislative” RoL captures at least a very important layer of the idea of the RoL. True, the apparently obvious claim that legislation is the issuing of prescriptions—that, then, statutes are prescriptions—is by no means aproblematic. Legislation proper, as it occurs in developed legal systems, has many complex, articulated procedural and institutional features, which have no obvious equivalent in the case of simple acts of prescribing.20 Most important, there is no obvious way in which a multi-membered legislature, composed by individuals and groups who sharply disagree with each other on the relevant issues, and making decisions on the basis of majority rule, may be assimilated to an individual, enacting his own will by expressing it in the form of a prescription.21 These difficulties notwithstanding, it seems to me, the model—a “prescriptive” understanding of the RoL—remains (p.136) illuminating, especially insofar as the RoL is envisaged—as it, in fact, usually is—as an ethico-political ideal concerning the relations between those who govern and the governed, between rulers and ruled. Pondering the formal features of prescriptions remains, it seems to me, the first step in understanding the nature of legislation (I shall not defend this view here; the reader may take it as an axiom).22

Thus, RoL requirements (in particular, those in the first group, section 2) will be understood as features prescriptions have to exhibit if they are to be capable of achieving, and of achieving well, the purpose of standard guidance of the behavior of their addressees. Prescribing is a form—a paradigmatic form—of “telling people what they should do.”23 Thus, standard guidance is the typical, normal purpose of the activity of prescribing—it may be termed its constitutive purpose. The practice (or the “institution”) of prescribing would not exist, standard and non-standard cases alike, were it not commonly and (usually) rightly assumed, when prescriptions are issued, that they aim at standard guidance.

RoL desiderata (those in the first group), then, will be features that prescriptions typically exhibit, and conditions prescriptions normally meet, in varying degrees. The possibility of non-standard prescriptions is not ruled out. In particular, as we shall see in a minute, abuses are possible: people can exploit the opportunity of issuing prescriptions—they can make a strategic use of the institution of prescribing—in order to pursue in deviant, indirect ways, non-standard, undisclosed purposes.

So, what kinds of power may be served by the issuing of prescriptions that systematically violate either of our two requirements, and in what ways may such legislation be “manipulative” of those who are subject to it, and not respect their dignity?

Imagine a power relation of the following kind: the lawgiver does not intend the addressee to do what he prescribes him to do. On the contrary, he often issues inconsistent or non-compliable prescriptions (e.g. by commanding actions, or omissions, which run counter to the most deeply ingrained inclinations of the vast majority of human beings). What the lawgiver intends, is precisely that the addressee should fail to comply, and, as a consequence, develop a sense of puzzlement, fear, anxiety, or a strong feeling of guilt. Thanks to that, it will be easy, then, to make the addressee do, by further means, what the lawgiver actually wants him to do.

This is, I submit, a paradigm of a power structure that works through the manipulation of subjects. I shall now try to unpack what is implied in this minimal Gedankenexperiment. In so doing, by contrast, the main features of RoL power will (p.137) emerge—and it will be apparent in what sense, and why, the RoL meets the two conditions laid out in section 3.24

4.1. Consistency

I begin with some assumptions concerning the notion of “inconsistent” prescriptions, and how purported logical relations between prescriptions should be understood. These will be the conceptual tools to be employed in the inquiries that follow. These assumptions will be merely stated, here. Their defense is deferred to another occasion.

By “inconsistency” (between prescriptions) I mean three kinds of cases: (1) conflicts proper (one and the same action A is both obligatory and forbidden: OA & FA); (2) contradictions (it is both obligatory that A and permitted that not A; or, A is both forbidden and permitted);25 (3) cases where two conditional directives, referring, respectively, to the condition that p and the condition that q, reconnect to these conditions either conflicting or contradictory deontic consequences (as defined in cases (1) and (2)), and, further, p and q jointly occur (it is left open the possibility that p and q are one and the same condition; when this is so, trivially, when the one occurs the other also necessarily occurs). In the first case, and the corresponding case in the third group, whatever one does is wrong (these are dilemmatic situations). In all cases, the relevant prescriptions cannot both be obeyed or satisfied.26 This marks a difference between statements of fact (or, propositions), on the one hand, and prescriptions, on the other hand. Statements of fact (or propositions) may be true or false; that two statements of fact (or propositions) are inconsistent means that there is no possible world in which they are both true. Prescriptions are neither true nor false. That two prescriptions are inconsistent means that there is no possible world in which they are both followed or satisfied (with one qualification, to be spelt out soon).

The first of our two RoL desiderata requires the avoidance of inconsistencies. Why? Inconsistent prescriptions (on any plausible account of the existence conditions of prescriptions), or the issuing of such prescriptions, are not a logical impossibility. A world in which prescriptions which cannot both be obeyed or (p.138) satisfied exist, or are issued, is a possible world (as opposed to a world in which some of the propositions which are true in it are not true). Purported “logical” relations between prescriptions, I assume,27 mirror the logical relations proper holding between their respective obedience- or satisfaction-statements (meaning, by this, statements of fact, or propositions, describing the actions which satisfy the relevant prescriptions). Specifically, two prescriptions are inconsistent if and only if their respective obedience- or satisfaction-statements are inconsistent.28

But why should we single out, as a separate class, those relations between prescriptions that track the logical relations (for example, inconsistency) holding between their respective obedience- or satisfaction-statements? Granted, “logical” relations between prescriptions mirror logical relations between their respective obedience- or satisfaction-statements. But why do they matter? Why is it important to single out, in the form of principles of a purported “logic” of prescriptions, the conditions under which sets of prescriptions are, by logical necessity, jointly satisfiable? Why is it important to single out the class of those prescriptions—inconsistent prescriptions—which cannot (logically) jointly be satisfied?

The answer is, I think, this:29 purported logical principles governing prescriptive discourse have to be understood as criteria of rational law-giving (i.e. as principles a rational law-giving will should conform to), under one crucial assumption: whoever prescribes X to do A wants X to do A. It is under this assumption that the conditions under which sets of prescriptions can (logically can) be jointly satisfied provide criteria of rational law-giving. A rational lawgiver cannot want what is (what she believes to be) logically impossible to be the case. This is why the issuing of inconsistent prescriptions may normally be deemed irrational.

The relevant assumption is, then, that the lawgiver wants the addressee to do what she tells him to do;30 and, further, she intends, by uttering the prescription, to make him perform the action. This assumption is crucial, for this reason: it is by virtue of this assumption that, in general, the issuing of prescriptions becomes possible. Understanding an utterance as a prescription involves the (defeasible) ascription, to the utterer, of the assumed intention: the lawgiver wants that the subject perform the prescribed action; and he intends, by prescribing it, to make her perform the action.

This is what I shall call “normal,” or “standard” legislative intention.31 Normal legislative intention is the intention to pursue a prescription’s constitutive purpose, (p.139) namely, standard guidance (section 4). Non-standard legislative intentions are possible. Prescriptions can be insincere.32 Ascription to a lawgiver of a normal legislative intention is a defeasible presumption. It is, however, standardly true; and an explicit denial of this condition (“I hereby order you to do A, but I don’t care whether you do it or not”) would prevent a prescription from coming into existence. The institution of prescribing would not exist, standard and non-standard cases alike, were it not commonly and (usually) rightly assumed, when prescriptions are issued, that the lawgiver wants that the subject perform the prescribed action, and he wants, by prescribing her to perform it, to make her perform the action.

Prescriptive inconsistency, then, is no logical impossibility. Its avoidance is, rather, a matter of rational law-giving. “Rational” in what sense? Instrumentally rational, relating to a certain goal: standard guidance, i.e. guiding the behavior of subjects in the prescribed direction—making the subject perform the prescribed action—through their understanding of this.

So, the existence of jointly non-satisfiable prescriptions is a logical possibility; the issuing of inconsistent prescriptions is possible. And—what is crucial to our present purposes—it may well be instrumentally rational, in the pursuit of non-standard legislative goals. Possible reasons—i.e. non-standard legislative intentions—are the following. (I take as paradigmatic those cases where whatever one does is wrong; the items in this list may be weakened or qualified in order to account for the other possibilities.)

(1) The lawgiver may wish to induce in the subject a state of puzzlement (“What on earth is required of me?”), of frustration, anxiety, a sense of being “entrapped,”33 a nervous breakdown,34 or cognitive dissonance.35 (2) He may want to have a (purported) reason for punishing her anyway. (3) He may want to induce fear, or terrorize her. (4) He may have the purpose of humiliating her, or (5) of inducing feelings of guilt.

These are all non-standard legislative intentions (the list has no pretension of being exhaustive) which may render instrumentally rational the issuing of inconsistent prescriptions. They all identify cases of abuse of the institution of prescribing:36 specifically, the lawgiver exploits—i.e. makes a strategic use of—the form of (p.140) prescriptive discourse in order to bring about a result different from, and irreducible to, standard guidance.37

Now, it is important to realize that robust power structures, having a firm hold on their subjects, may in fact come into being and prosper—in families, workplaces, churches, associations of various sorts—around the issuing of inconsistent prescriptions, and through it.38 Flouting the consistency requirement may be a perfectly rational way of engendering puzzlement, frustration, fear, guilt, etc., thus rendering the subjects malleable to the lawgiver’s will.

Power structures that operate in this way. I submit, may well be characterized, in a straightforwardly intelligible sense, as of a “manipulative” kind (our first condition, section 3). The non-standard modes of control on our list are ways of forcing the subject in an artificially created quandary, of inducing in her unpleasant feelings or emotions, or a contrived experience of uneasiness, and to capitalize on this. These are ways of, metaphorically, treating subjects as things to be modified as one sees fit: not treating them as centers of deliberation—i.e. as at least potentially autonomous agents, who can be (part) authors of their life.39

Considering these various non-standard legislative intentions is, then, a good way of making determinate, or at least less generic, the idea of (not) treating human beings as things, and, thus, of seeing under what respects RoL power is opposed to the manipulation of human beings. Correspondingly, power structures that rely on the non-standard modes of control on our list prove, in various ways, disrespectful of the dignity of their subjects (our second condition, section 3).

In acting in these ways, first, the lawgiver does not respect the subjects’ status as potentially autonomous agents (we have just seen this). By acting in these ways, second, he may positively intend to dismantle, and succeed in dismantling, the self-esteem of subjects—provided that the latter are disposed, for whatever reason (section 4.2), to take seriously the lawgiver’s dictates. This is a further way of not respecting them. And, third, in all these ways, when apparent—to both lawgiver and subjects or to a third party (cf. section 4.2)—the lawgiver does not respect the dignity of subjects because, and insofar as, he shows disrespect for them. (One way of not respecting someone is to show disrespect for him.) The attitude evinced through the issuing of inconsistent prescriptions may be rendered as “It doesn’t (p.141) matter what you may think you should do. Calculations are useless.40 I can do whatever I want of you;”41 and “I don’t care about what you may think I think, believe, or will. I am free from any discoursive obligation toward you” (“You may think you are somebody worth talking to, but you are not”). Such messages show disrespect for those to whom they are addressed,42 and may work as tools in demolishing their self-esteem.

These are different, cumulable, and mutually reinforcing ways of not respecting the status of human beings as at least potentially autonomous centers of deliberation and agency. It is in this way that the non-standard modes of guidance on our list prove disrespectful of the dignity of subjects.

4.2. Compliability (The Technology of Guilt)

The requirement of compliability may be understood in various ways, according to how we understand the idea that it should be “possible” to perform the required act (or, to forbear from acting in the relevant way). So, for instance, the principle may be understood as imposing the condition that the required act should be logically, or conceptually possible (thereby ruling out, for example, the injunction to draw a square circle, and the like). Here, as usual in the literature, I shall understand the principle as requiring the prescribed acts (or omissions) to be “humanly possible”—i.e. as requiring them to be, in addition to being logically or conceptually possible, and physically possible, acts whose performance is generally within the scope of the abilities and capacities of normal (whatever this may mean) human beings.43

The principle “ought” implies “can,” I assume, is not a logical principle: it is not logically inconsistent to claim that Xs ought to do A, although they cannot do A (the former claim does not entail the negation of the latter). Relating to the field of prescriptive discourse, it expresses, however, a conceptual constraint: prescriptions flouting this constraint are abuses of the institution, parasitical on the normal or standard case. This, once again, should be understood in the sense that conforming to the principle is a matter of rational law-giving.44 “Rational”—instrumentally rational—with respect to a certain purpose: the purpose of standard guidance. Typically, understanding an utterance as a prescription involves the ascription, to the utterer, of a normal legislative intention (a defeasible presumption; section 4.1): whoever prescribes to X that he should do A wants X to do A, and he wants to make X do A. This is why prescribing to X the performance of A, and at once openly admitting that A is, for human beings, impossible, defeats the (p.142) prescription. It would be irrational to want somebody do something—and to try to make her do it—if one believed it to be impossible.

So, let us ask ourselves, as we did in the preceding section, what kind of power may be served by flouting the compliability condition, and why such forms of power may turn out to be of a “manipulative” kind, and not respect the dignity of their subjects.

The answers to these questions are, in fact, the same as in the case of the consistency condition. This is no surprise, since inconsistency, as defined here (section 4.1), is a special case of non-compliability: inconsistent directives cannot be jointly followed (first case) or anyway satisfied (second case). The features indicated in the preceding section as making power manipulative, and disrespectful, in the case of inconsistent prescriptions, are, in fact, features that apply to non-compliable prescriptions generally. The list of the reasons making instrumentally rational the violation of the requirement—non-standard legislative intentions—may, however, be enriched. To those listed in section 4.1 we may add the following two items.

  1. (1) The lawgiver may want to put on trial the subject’s disposition to obedience, or deference to his authority.

  2. (2) He may wish to make her try to perform the (apparently) required act. This may happen either because he wants the subject to succeed in the assigned task to the (by hypothesis partial) extent that she can,45 or because he wants her to fail. And this, once again, may engender frustration, anxiety, etc.46

One especially important possibility is that of pursuing the strategy of prescribing the impossible in order to make the addressees feel guilty. Systematically flouting the compliability requirement may be a way of inducing in the subject the sense of his inadequacy, and weakness, and of making him feel guilty about that (we may call this the “technology of guilt”). Logical impossibility (due to inconsistency, or to other forms of conceptual impossibility) may be a way of achieving this, but, if we assume the relevant logical relations to be evident to the subject’s eye, and to be common knowledge among lawgiver and subject, this way of pursuing the strategy may require too much by way of irrationality on the part of subjects. Generic non-compliability—prescribing what is (though logically possible) not humanly possible—is a much more efficacious method (not least because, as remarked by Fuller,47 the line between what is and what is not humanly possible is often uncertain). One way of acquiring and exercising power over human beings is by inducing in them guilt for their (purported) constitutive inadequacy, or (p.143) weakness, and setting ourselves as their healers, because we are uniquely authorized to guarantee them forgiveness for their faults, or because we know how, and are able to, supplement them in their inadequacy or weakness (we have to manage to make them believe this, of course; believing it ourselves may help). And one way of doing this is by issuing prescriptions we know they will not be able to comply with—setting a standard we know they will not be capable of living up to. That is, we can do it by flouting the requirement that whoever prescribes wants the addressee do what he prescribes her to do, and tries, by issuing the prescription, to make her perform the desired action. In such cases, we do not actually want the addressees to do what we (seem to) require of them; rather, we want them to fail. It is thanks to their expected failure that we (mean to) acquire and exercise power (by prescriptions, or any other means) over them. In this way, guidance—non-standard guidance—may be enabled by non-compliable directives.48

This is the place to introduce, and discuss briefly, a complication implicit in what has been said already (and which may have perplexed the reader in the preceding paragraph). It concerns the import of both requirements, compliability and consistency. The issue is what may be termed the “epistemic dimension” of the requirements—and of the RoL generally. When we say that the RoL requires compliability, do we mean that the relevant directives should simply be compliable (no matter what the parties believe thereabout), or that they should be (perhaps mistakenly) believed (by both lawgiver and addressee?), or even known to be fulfillable?49

There are various possibilities; and, correspondingly, different versions of the compliability condition. I consider only some of them, those that seem to me significant re our main issue, “law as power.”

The action prescribed, A, may, unknown to the parties, in fact be impossible, or it may be believed, either truly or falsely, to be impossible. The former hypothesis, it seems to me, is, for our purposes, uninteresting.

The hypothesis that A is falsely believed to be impossible does not, it seems to me, bear peculiar implications, apart from the fact that it may happen that A is in fact performed, and thus discovered, ex post (contrary to what was previously believed), to be possible. Let us consider, then, the hypothesis that A is truly believed to be impossible. We can further distinguish three sub-hypotheses, according to who believes this, either the lawgiver, or the addressee, or both. In the first of these, issuing a prescription may be a way of making the addressee try to perform A, fail, and (provided that, for whatever reason, the addressee is disposed to take the lawgiver’s dictates seriously) of engendering frustration, or guilt.

Suppose now that both the lawgiver and the addressee believe A to be impossible. We can further distinguish two possible scenarios. The fact that A is impossible (p.144) may be common knowledge among lawgiver and addressee (it is out in the open between them that A is impossible), or not (each one of them believes A to be impossible, but does not believe the other to believe this).50 In the former case, prescribing the impossible may be a way of overtly showing disrespect, and (once again, provided that, for whatever reason, the addressee is disposed to take the lawgiver’s dictates seriously) of humiliating the addressee, or of demolishing his self-esteem (“You are not even worthy of a compliable prescription”).51

Often, as remarked, the distinction between what is and what is not humanly possible will be uncertain (to the lawgiver, to the addressee, to both…). This may give rise to a vast array of epistemic nuances in the technology of guilt. Generally speaking, it seems to me, the relevant epistemic and, generally, intentional framework may be largely indeterminate, inchoate, untidy, muddled, even incoherent—consisting of more or less rational epistemic attitudes and dispositions of the parties involved (specifically, dispositions of deference toward the lawgiver, on the part of the subjects).52 Within a pathological parental relation, for example, a child may feel guilt, or lose his self-esteem, for not complying with prescriptions commonly known, by her and her parent, to be unfulfillable (and, maybe, issued with that purpose). On the other hand, irrational lawgivers may well want subjects do what they believe to be impossible—and, thus, sincerely set for them standards they know the subjects will not be capable of living up to.


In principle, then, power can be effectively exercised by systematically flouting our two requirements. Power structures, in various social settings, may flourish around areas of inconsistent or non-compliable directives, by relying on their psychological effects (ranging, as we have seen, from puzzlement to guilt).53

A Foucaultian remark may further clarify this. According to Foucault, writes John Scott, “any form of power, other than mere force or physical repression, depends on the formation of individuals into subjects with appropriate motives and desires.”54 Power both requires, and brings about, “subjectivation.” Specifically, both inconsistency and uncompliability may be means of constituting subjects apt for subjection to (non-RoL) power. Systematically flouting our two requirements (p.145) does not bring about specific motives, beliefs, or attitudes. It may, however, be useful in inducing or reinforcing a generic habit of submission and docility.

It is not necessary that the strategy be intentional. As we have seen (section 4.2) the epistemic and, generally, intentional dimension of non-standard guidance may be indeterminate, inchoate, untidy, incoherent, both on the part of the subjects, and on the part of lawgivers. It may be enough that the relevant directives are, and are more or less confusedly perceived as being, unfulfillable or inconsistent, that the effects are those described (ranging from puzzlement to guilt or fear), and that these effects feed back on the guidance structure, providing for its maintenance. The whole process may go on without anybody intentionally practicing a strategic use of non-compliable or inconsistent legislation—or, specifically, of the technology of guilt.

So, the issuing of non-compliable or inconsistent directives can both be strategies for gaining a secure hold on subjects: ways of laying down the foundations of robust power structures. Not, however, respectful ways. Rather, they can be ways of forcing subjects in a no-way-out situation (and, thus, of suspending their agency), of dismantling their self-esteem, of showing disrespect for them, in order to make it easier to guide them (by whatever means may be conducive to the end) in the desired direction. When effective in making the subjects docile to guidance, power structures of this sort may well be characterized as manipulative. Subjects are not treated as agents, but as things to be pushed or modified according to the power-holder’s will. And, in so manipulating subjects, non-standard guidance, as I have characterized it here (through nonconformity with our two requirements), falls below the threshold of a minimum of respect for human dignity.

6. Rule of Law power: Some Central Features

The two RoL requirements we have considered illustrate a threefold conclusion: RoL power is public, rational, non-paternalistic.55

(1) We have considered the possibility of power structures built around the systematic flouting of the requirements of consistency or compliability. In both cases, typically, power is non-public, in this sense: the power-holder keeps hidden something which is essential for the intended workings of the power structure—namely, the reason why he is issuing inconsistent or non-compliable directives (in order to punish the subject no matter how she will act; or, in order to induce feelings of guilt, etc.). If that were to become public, the power structure would cease to work. Power exercised by flouting the two requirements is liable to be “unmasked.”

(p.146) The RoL is, in this sense, publicly exercised power: the power structure works, if and to the extent that the subject understands that, and how, it is working. This understanding is part of the power structure itself.

In other words, RoL power may be characterized as “public” in the following sense: the strategies and forms of power which systematically resort to inconsistent or non-compliable directives are, typically, secret: their explicit avowal would refute them, both in the sense of making them inefficacious, and in the sense of showing them to be disreputable. In such “secret” power structures, inconsistent, or non-compliable, directives are useful, only so long as it is not publicly known why they are issued.56

Thus (or at least this is the pro tanto conclusion the examination of our two desiderata leads us to), RoL power works through the understanding, by the subject, of its workings: the subject understands that something is required, and what is required, of her, and this understanding is a necessary condition for the kind of guidance intended—for her doing what she is intended to do, in the way she is intended to do it.

I conclude that, to the extent that it satisfies our two desiderata, the law is, pro tanto, a method of social control that works openly and publicly in the production of the desired behavior. Its mode of operation is such that it has to stand out in the open for it to work in the intended way.

(2) As we have seen (section 4), power exercised by flouting the two requirements can be instrumentally rational. It all depends on what the (non-standard) purpose of the lawgiver is. There is, however, a specific respect under which RoL power is rational power: it is power for rational subjects. Specifically, I shall now argue, power exercised by complying with our two requirements is power for rational people.

Think of the attitude which may be expressed by the issuing of inconsistent prescriptions: “It doesn’t matter what you may think you should do, calculations—or the weighing of reasons for and against alternative courses of action—are useless. What will happen to you does not depend on your deliberation.” Guidance through the issuing of prescriptions, when it is standard guidance, is a rational mode of guidance, in this sense: it is a method of control which relies on—and, in the first instance, leaves room for—the subject’s calculations and deliberation. The issuing of inconsistent prescriptions, for the purposes of some form of non-standard guidance, suspends this. The addressee is not being treated as a rational (calculating, reasons-weighing, deliberating) agent anymore.

The same holds in the case of non-compliability, where this is an instrument for non-standard guidance. When non-compliability is common knowledge among (p.147) the parties, the attitude expressed by the lawgiver is very similar: “You are not worthy of being told what you are supposed to do. Your fate does not depend on your understanding or deliberation.” When the issuing of non-compliable directives is an—undisclosed—way of making the subject try to perform the (apparently) required action, or of inducing feelings of inadequacy and guilt, the subject is, once again, not being treated as a rational agent.

This leads us to an important point, which has surfaced already at times (sections 4.2, 5), but is worth emphasizing. I have summarily described some non-standard forms of guidance through the issuing of prescriptions, built upon the violation of our two RoL requirements. It is important to note that power structures of this kind may rely, for their efficacy and operation, on irrational attitudes and on fallacious inferences on the part of the subjects. To take two extreme cases, it is (barring, perhaps, far-fetched scenarios) a wholly irrational reaction to lose one’s self-esteem in the face of inconsistent directives, or to feel guilt for not complying with patently non-compliable prescriptions. Both attitudes may be the upshot of unconscious drives, misplaced affections, unwarranted fears, or any sort of irrational attitude. Power structures relying on these sorts of psychic mechanisms are no less real for that.

Thus—and this is an important methodological point—when we consider the RoL as a specific mode of power, and when we try to grasp what its features are by contrasting it with other varieties of power, it is not only power as effective in controlling rational subjects that we have to contrast it with. There is no reason to exclude from the comparison forms of (non-standard) guidance that rely on the irrationality of subjects.

When seen in the light of these possibilities, RoL power—specifically, power conforming to our two requirements—looks, in a sense, quite naive: X tells Y to do A, and announces to Y what the consequences of non-compliance will be, in the expectation that Y will ponder whether to comply or not. It is in this, minimal, sense that RoL guidance—as illustrated by our two requirements—is guidance for rational subjects.57

(3) From what has just been said about RoL power as power for rational subjects it follows that RoL guidance, as illustrated by a power structure that meets our two requirements, is non-paternalistic. Not in the sense (the focal sense of “paternalistic”) that it rules out interference—or, specifically, coercive interference—in the activities of subjects in order to promote their own good. Laws conforming to the RoL desiderata may well be paternalistic in this sense. It is, rather, a matter of the kind of attitude expressed in the exercise of RoL power.

When the government treats its subjects in accordance with the RoL, it treats them as adults, capable of making their own decisions on the basis of their own preferences and their own understanding of the relevant facts. It tells them (p.148) explicitly “I want you to behave in such and such a way; these will be the consequences—I shall inflict you such and such a harm—in case you don’t; now it’s up to you.”

This mode of exercising power over a human being should be contrasted with the way in which children are often treated. In order to make children do what we want them to do we sometimes tell them lies (“Candy shops are closed now”); we fake non-existing unpleasant consequences (“The wolf will come and get you”); in various ways, we distort reality. Or we try, unknown to them, directly to manipulate the environment, or their preferences, by working behind their back, so to speak. Or, again, we rely, in trying to make them do what we want them to do, on an aura of parental authority, or on symbols. In acting in these ways, we do not recognize to children the dignity of responsible agents, capable of autonomous choice; we do not treat them as autonomous agents capable of—and entitled to—making their own choices on the basis of preferences and beliefs which are in fact their own. It is characteristic of standard guidance through prescriptions—thus, of RoL power—that it gives the subject a choice—albeit, often, a forced one.58

Furthermore, in a prescriptive relationship the subject to whom a prescription is addressed is kept at a distance, so to speak. She is not regarded by the lawgiver as an appendix to, or an extension of, his own body, as merely a tool, or as one commodity among others at his disposal, or again as something in the environment to be manipulated (be it in view of its own good). Causal efficacy on her conduct is mediated by her own understanding of its being exerted, and how (cf. in this section, sub (1))—and this is common knowledge between the two. In this way, too, RoL power is non-manipulative.

These two features—giving the subject a choice (albeit a forced one) and keeping her at a distance (in the specified sense)—render RoL power very different from the kind of power parents often exercise over their young children, shepherds over the flock, propaganda experts over their target, etc. It may be said that, if the supreme achievement of power consists in securing the willing compliance of its subjects,59 it is characteristic of RoL power that it aims at securing unwilling compliance.

Under both respects, one distinctive feature of RoL power is that rulers regard their subjects, literally, as addressees—i.e. as subjects capable, and worthy, of being addressed. To borrow a phrase from Strawson, their dealings with them, as addressees, are not premised on “objectivity of attitude:” a “purely objective view of the agent as one posing problems simply of intellectual understanding, management, treatment and control.”60

(p.149) This is what happens in cases of standard guidance, where (inter alia) our two requirements are satisfied. Things stand differently within power structures where flouting consistency and compliability may turn out to be instrumentally rational. Where, for example, pastoral power is being exerted, inconsistent or non-compliable directives may be good means in the enterprise of leading the subject in the desired direction, be it his own good. Or, again, in issuing patently non-compliable directives in order to induce feelings of guilt power-holders are not treating subjects as addressees (in the specified sense).

Thus, under all three respects—publicity, rationality, non-paternalism—RoL power, as illustrated by laws meeting our two requirements, qualifies as a mode of power which (minimally) fits the dignity of adult, rational, autonomous agents, to be contrasted with manipulative forms of power.61 When the government treats its subjects in this way, it recognizes to them the dignity of beings worthy of being publicly, openly addressed, and of being guided through their understanding of the way in which power is being exerted over them, and their deliberations on the merits of the case. In short, it treats them with, and shows them, respect.

This concerns only the form of the relationship, not the laws’ content. It is compatible with all sorts of disrespect and unjust discrimination. But it positively is, it seems to me, a valuable feature of laws meeting RoL desiderata.


Bibliography references:

Alchourrón, C. E. and Bulygin, E. (1984). “Pragmatic Foundations for a Logic of Norms,” Rechtstheorie, 15: 453–64.

Anscombe, G. E. M. (1958). “On Brute Facts,” in G. E. M. Anscombe, Ethics, Religion and Politics. Oxford: Blackwell, 1981, 22–5.

Austin, J. L. (1962). How to Do Things with Words: The William James Lectures. Delivered at Harvard University in 1955. Second edition edited by J. O. Urmson and M. Sbisà. Oxford: Oxford University Press, 1975.

Bobbio, N. (1999). Teoria generale della politica. Edited by M. Bovero. Turin: Einaudi.

Celano, B. (1990). Dover essere e intenzionalità: Una critica all’ultimo Kelsen. Turin: Giappichelli.

Celano, B. (2010). “Normative Legal Positivism, Neutrality, and the Rule of Law,” paper presented at the conference “Neutrality and Legal Theory” (Girona, Spain, May 21–2, 2010), 〈http://ssrn.com/author=1605026〉 and 〈http://www.filosofiayderecho.es/pt/tienda/?page_id=3&lang=es http://www.filosofiayderecho.es/pt/tienda/?page_id=3&lang=es〉 (accessed March 20, 2011). (p.150) Forthcoming in J. Ferrer Beltrán, J. J. Moreso, and D. M. Papayannis (eds.), Neutrality and Theory of Law. Dordrecht: Springer.

Celano, B. (2011). “Publicity and the Rule of Law,” forthcoming in L. Green and B. Leiter (eds.), Oxford Studies in Philosophy of Law, vol. 2. Oxford: Oxford University Press.

Celano, B. (2012). “Hart’s Blind Spot: Il tassello mancante in The Concept of Law,” Rivista di filosofia del diritto, 1: 405–26.

Elster, J. (1983). Sour Grapes: Studies in the Subversion of Rationality. Cambridge: Cambridge University Press.

Elster, J. (1993). Political Psychology. Cambridge: Cambridge University Press.

Festinger, L. (1962). A Theory of Cognitive Dissonance. Second edition (first edition 1957). Stanford, CA: Stanford University Press.

Finnis, J. M. (1980). Natural Law and Natural Rights. Oxford: Clarendon Press.

Foucault, M. (1981). “Omnes et Singulatim: Towards a Criticism of ‘Political Reason’,” in S. McMurrin (ed.), The Tanner Lctures on Human Values. Salt Lake City: University of Utah Press, 225–54.

Fuller, L. L. (1969). The Morality of Law. Revised edition. New Haven: Yale University Press.

Grice, H. P. (1957). “Meaning,” in H. P. Grice, Studies in the Way of Words. Cambridge, MA: Harvard University Press, 1989, 213–23.

Hart, H. L. A. (1961). The Concept of Law. Oxford: Clarendon Press.

Hart, H. L. A. (1983). Essays in Jurisprudence and Philosophy. Oxford: Clarendon Press.

Hart, H. L. A. (1982). Essays on Bentham: Studies in Jurisprudence and Political Theory. Oxford: Clarendon Press.

Hurd, H. M. (1990). “Sovereignty in Silence,” Yale Law Journal, 99: 945–1028.

Kelsen, H. (1945). General Theory of Law and State. Cambridge, MA: Harvard University Press.

Kramer, M. H. (1999). In Defense of Legal Positivism: Law without Trimmings. Oxford: Oxford University Press.

Kramer, M. H. (2007). Objectivity and the Rule of Law. Cambridge: Cambridge University Press.

Lukes, S. (2005). Power: A Radical View. Second edition (first edition 1974). Basingstoke: Palgrave.

MacCormick, N. (1973). “Legal Obligation and the Imperative Fallacy,” in A. W. B. Simpson (ed.), Oxford Essays in Jurisprudence. Second series. Oxford: Clarendon Press, 171–201.

MacCormick, N. (1985). “A Moralistic Case for Amoralistic Law?,” Valparaiso Law Review, 20: 1–41.

Marmor, A. (2004). “The Rule of Law and its Limits,” Law and Philosophy, 23: 1–43.

Raz, J. (1979). The Authority of Law: Essays on Law and Morality. Oxford: Clarendon Press.

Raz, J. (1986). The Morality of Freedom. Oxford: Clarendon Press.

Raz, J. (1996). “Intention in Interpretation,” in J. Raz, Between Authority and Interpretation: On the Theory of Law and Practical Reason. Oxford: Oxford University Press, 2009, 265–98.

Schauer, F. (2010). “Necessity, Importance, and the Nature of Law,” paper presented at the conference “Neutrality and Legal Theory” (Girona, Spain, May 21–2, 2010), 〈http://www.filosofiayderecho.es/pt/tienda/?page_id=3&lang=es http://www.filosofiayderecho. (p.151) es/pt/tienda/?page_id=3&lang=es〉 (accessed March 20, 2011). Forthcoming in J. Ferrer Beltrán, J. J. Moreso, and D. M. Papayannis (eds.), Neutrality and Theory of Law. Dordrecht: Springer.

Schiffer, S. (1988). Meaning. Second edition (first edition 1972). Oxford: Clarendon Press.

Scott, J. (2001). Power. Cambridge: Polity Press.

Searle, J. R. (1969). Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press.

Strawson, P. F. (1962). “Freedom and Resentment,” in P. F. Strawson (ed.), Studies in the Philosophy of Thought and Action. Oxford: Oxford University Press, 1968.

Strawson, P. F. (1964). “Intention and Convention in Speech Acts,” in P. F. Strawson, Logico-linguistic Papers. London: Methuen, 1971.

Tamanaha, B. Z. (2004). On the Rule of Law: History. Politics, Theory. Cambridge: Cambridge University Press.

Von Wright, G. H. (1951). An Essay in Modal Logic. Amsterdam: North Holland.

Von Wright, G. H. (1963). Norm and Action: A Logical Enquiry. London: Routledge & Kegan Paul.

Von Wright, G. H. (1968). “The Logic of Practical Discourse,” in R. Klibansky (ed.), Contemporary Philosophy: A Survey, vol. 1: Logic and the Foundations of Mathematics. Florence: La Nuova Italia, 141–67.

Von Wright, G. H. (1983). “Norms, Truth and Logic”, in G. H. von Wright, Practical Reason. Oxford: Blackwell, 130–209.

Von Wright, G. H. (1984). “Bedingungsnormen: Ein Prüfstein für die Normenlogik,” in W. Krawietz et al. (eds.), Theorie der Normen: Festgabe für O. Weinberger zum 65. Geburtstag. Berlin: Duncker & Humblot, 447–56.

Von Wright, G. H. (1985). “Is and Ought,” in E. Bulygin et al. (eds.), Man, Law and Modern Forms of Life. Dordrecht: Reidel, 263–81.

Waldron, J. (1999a). Law and Disagreement. Oxford: Oxford University Press.

Waldron, J. (1999b). The Dignity of Legislation. Cambridge: Cambridge University Press.

Waldron, J. (2007). “Legislation and the Rule of Law,” Legisprudence, 1: 91–123.

Waldron, J. (2008a). “Hart and the Principles of Legality,” in M. H. Kramer et al. (eds.), The Legacy of H. L. A. Hart: Legal, Political, and Moral Philosophy. Oxford: Oxford University Press, 67–84.

Waldron, J. (2008b). “The Concept and the Rule of Law,” Georgia Law Review, 43: 1–61.


* I wish to thank Francesca Poggi and Stefan Sciaraffa for their comments on an earlier version of this chapter.

(1) Hart (1961: ch. 4) rightly criticized the view that the law should be understood as the gunman situation writ large—thus, he rejected Austin’s concept of the sovereign as the source of all legal rules—as naive and simplistic. Nothing comparable, however, was put in its place. Law’s foundations are seen in the practices of the club of officials. Brute social power—the role of law as a cog in the workings of social powers—is, in a way, Hart’s blind spot. (I have argued for this claim in Celano (2012). Nothing in the argument of this chapter depends on the correctness of this assessment of Hart’s jurisprudence.)

(2) Kelsen (1945: part I, ch. 1).

(3) Cf. Schauer (2010).

(4) Tamanaha (2004: 91).

(5) For a list of these institutional and procedural requirements see e.g. Raz (1979: 215–18): “the making of particular laws (particular legal orders) should be guided by open, stable, clear, and general rules;” “the independence of the judiciary must be guaranteed;” “the courts should have review powers over the implementation of the other principles;” “the courts should be easily accessible;” “the discretion of the crime-preventing agencies should not be allowed to pervert the law.” For similar lists of the RoL requirements, see Fuller (1969: ch. 2); Finnis (1980: 270–1); Marmor (2004: 5ff.); Kramer (2007: ch. 2).

(6) Lukes (2005).

(7) Anscombe (1958).

(8) Waldron (2008a: 78).

(9) Bobbio (1999: 182–3).

(10) Raz (1979: 219–20).

(11) Fuller (1969: 163–7), and, on Fuller’s views thereabout, Waldron (2008b: 17); Raz (1979: 221); Finnis (1980: 273); Kramer (2007: 184). The RoL is famously contrasted with “the rule of men.” An understanding of this Protean contrast, however, is—barring rhetorical formulae—the end, rather than the beginning, of theorizing.

(12) Cf. Raz (1979: 221).

(13) Raz (1979: 221). That the point of (many of) the RoL requirements is to make possible, inter alia, (certain forms of) respect for human dignity, is common ground. Cf. Fuller (1969: 162–3); Finnis (1980: 272–3); MacCormick (1985: 26); Marmor (2004: 21, 32); Waldron (2008a: 76).

(14) Fuller (1969: 33–8).

(15) In Fuller’s phrase, “the enterprise of subjecting human conduct to the governance of rules” (1969: 106).

(16) A simple case of non-standard guidance: parents may sometimes, in order to make their young children do A, tell them not to do A, relying on their children’s standing disposition to do the opposite of what they tell them to do.

(17) “Prescriptions are given or issued by someone. They ‘flow’ from or have their ‘source’ in the will of a norm-giver…They are, moreover, addressed at some agent or agents, whom we shall call norm-subject(s).” Von Wright (1963: 7). Further features which Von Wright sees as “characteristic of norms which are prescriptions” are of no interest for us here.

(18) In order to make room for power-conferring rules (and, especially, for rules conferring to private individuals the power to achieve some ends of theirs: “If you wish to do this, this is the way to do it” (Hart (1961: 28)), we should understand “prescribing” as including cases of telling people how to pursue the goals they want to achieve (or telling people how to do what they want to do). Cf. Raz (1979: 215): “power-conferring rules are designed to guide behavior.” Prescribing, so understood, covers both the issuing of mandatory directives and the issuing of power-conferring rules. In allowing in the text for the possibility that some legal norms are not prescriptions, I have in mind (not, power-conferring rules, but) norms that, at least apparently, are not “issued” by any “lawgiver.” This topic will not be further discussed here. “Prescriptive” RoL at least captures, I shall assume, one facet of the RoL.

(19) Waldron (2007: 109–10) rightly observes that Fuller’s treatment of the subject in Fuller (1969: ch. 2) “illustrates a strong…tendency to associate the rule of law with formal features of legislation, as opposed to other modes of law and law-making.”

(20) MacCormick (1973: 114–15).

(21) Waldron (1999a: part I) and (1999b: 26–8).

(22) As is well known, arguments against a “prescriptive model” of governance through law are legion. Cf. Hart (1961: 20–2); Hurd (1990).

(23) “Telling” people what they should do, as I mean it here, refers to cases of issuing prescriptions, not to “detached” statements of what the subject should do according to a given set of prescriptions. Raz (1979: 153–7).

(24) RoL requirements interact, and they work together in enabling the achievement of the purpose and values the RoL serves. Consistency and compliability are not—neither separately, nor jointly—sufficient conditions. So, the argument is not that when the laws are consistent, and it is possible to comply with them, then they satisfy the two conditions laid out in section 3. Rather, an understanding of the point, and import, of the two desiderata illustrates what the nature of RoL power is. Analogous arguments could be developed, mutatis mutandis, with regard to other RoL desiderata.

(25) I assume, following one usual approach in deontic logic, that OA entails PA (i.e. A is permitted), and, correspondingly, that FA entails that ¬A is permitted. So understood, the conceptual relations between the operators “obligatory,” “forbidden,” and “permitted” mirror those holding between the alethic modalities “necessary,” “impossible,” and “possible.” Cf. von Wright (1951: 1–3) and (1968: 142–3).

(26) I stipulate—following von Wright (1983: 139); but cf. also Hart (1982: 116)—that a permission is satisfied when its subject(s) at least sometimes avail themselves of the permission—i.e. they, at least sometimes, perform the permitted action. This seems fair to what may be taken to be the point of permissive norms.

(27) Following von Wright (1983), (1984), (1985); but cf. also Hart (1982: 116) and (1983: 325–7).

(28) This does not apply to pairs of permissions: that the satisfaction-statements of two permissions are logically inconsistent does not entail that they conflict (PA and P¬A are not inconsistent). Hart (1983: 327). The need for this—logically untidy—qualification is a consequence of the stipulation in n. 26.

(29) Cf. von Wright (1983: esp. at 132, 141–2, 149); (1984: 452–3, 456); (1985: 271). Cf. also Fuller (1969: 66); Hart (1982: 116); Alchourrón and Bulygin (1984: 458); Celano (1990: 268–82).

(30) Von Wright (1963: 7, 119); (1983: I, 8). Cf. also MacCormick (1973: 104–5); Hart (1982: 247, 249); Celano (1990: 127).

(31) The intentional structure constitutive of prescriptions is, in fact, more complex than this. The workings of a prescription rest, typically, on a set of nested, self-reflexive intentions. Grice (1957); Strawson (1964: 256–7); Schiffer (1988: 19); Celano (1990: 127–51, 205–13, and 2010: section 6). Cf. also Raz (1996: 283). Typically, the lawgiver has (1) the intention to make the addressee perform a certain action; (2) he intends to make the addressee perform the action as a consequence of his uttering a sentence; and (3) he intends to make the addressee perform the action (as a consequence of his uttering a sentence) by virtue of the recognition, by the addressee, of these very same intentions, (1) to (3). There is, however, no need for us to follow these complications here.

(32) Austin (1962: 16); Searle (1969: 60, 64ff.); MacCormick (1973: 105); Hart (1982: 249).

(33) Cf. Raz (1979: 222).

(34) Fuller (1969: 66).

(35) A lawgiver issuing inconsistent prescriptions may want to put the subject in a conflict situation; and, through the latter, to engender cognitive dissonance. Festinger (1962: 39–40) distinguishes between conflict (in a choice situation) and dissonance (in a post-choice situation, following the choice of one of the conflicting options). In conflict, one is pushed (by the conflicting reasons for and against conflicting options) in opposite directions. When dissonance occurs, after the choice, one is pushed in one single direction, toward the reduction of dissonance.

(36) Cf. Austin (1962: 16); Hart (1982: 247).

(37) A further hypothesis is that of a “pluralist” lawgiver, who issues conflicting prescriptions because he wants different, conflicting values to be pursued by his subjects. (Thanks to M. C. Redondo for pressing this point on me.) This hypothesis may perhaps be of some interest in accounting for the conflicts engendered by the various rights, principles, and values which make up a substantial part of the content of contemporary (especially European) constitutions, and of many supranational instruments. I shall not discuss this issue here, and concentrate on the kinds of “strategic” uses of inconsistencies, and the power structures building upon them, discussed in the text.

(38) A special case, by no means a marginal one, are “double binds,” generated, within deep or standing personal relationships, by the repeated occurrence of conflicting directives of different levels of abstraction. J. Elster (in (1993: 81); see also (1983: 60ff.)) discusses “contradictory injunctions,” such as “Don’t be so obedient!” or “Be spontaneous!:” their “overt content” “contradicts their pragmatic presuppositions;” they thus place subjects “in an impossible situation,” requiring of them “an impossible undertaking.”

(39) Cf. for this characterization of autonomy Raz (1986: 369).

(40) On this, see section 6, sub (2).

(41) Not caring about consistency is a sign, and an intimation, of unlimited power. Cf. for a parallel point about compliability Fuller (1969: 70–1).

(42) Cf. Marmor (2004: 32) making a parallel point about compliability.

(43) Thus, what is claimed is not that laws should be tailored to suit individual capacities and abilities. The inference from “cannot” to “therefore, he is under no obligation of doing it” does not hold, when the “cannot” is due to individual peculiarities. Cf. Hart (1961: 174); Kramer (2007: 130, 166). Rules are addressed to “normal” (whatever this may mean) subjects.

(44) Cf. the discussion in von Wright (1963: ch. 7, especially 119, 122), to which I am strongly indebted here. Cf. also von Wright (1985: 269).

(45) Fuller (1969: 71).

(46) Non-compliable directives may also be rational in the light of further ends, other than the guidance, be it standard or non-standard, of human conduct. So, for instance, if one takes the function of law to be that of providing the resolution of disputes, unfulfillable demands may be perfectly in order. Kramer (1999: 46–7); (2007: 131). Further, the imposition of strict liability involves a violation of the principle “ought” implies “can. Thus, whatever reasons may, in some contexts, make it rational to impose strict liability, make it rational to issue unfulfillable directives.

(47) Fuller (1969: 79).

(48) J. Elster (1993: 81) mentions, as an example of orders that require from subjects “an impossible undertaking,” orders that require “willing what cannot be willed” (“Forget it!,” “Don’t even think about that!”). Elster remarks that such orders, “if taken seriously,” may engender “pervasive guilt.” The technology of guilt works at its best when the relevant directives concern such internal states as believing, desiring, and feeling (“Love Big Brother!”). Cf. also Elster (1983: 60ff.).

(49) I shall only discuss compliability, but the discussion also applies, mutatis mutandis, to consistency.

(50) Common knowledge involves an indefinite series of epistemic iterations. It is common knowledge that p among X and Y when X knows that p, Y knows that p, X knows that Y knows that p, Y knows that X knows that p, X knows that Y knows that X knows that p, and so on. For refinements and details cf. Celano (2011: section 5). But, it seems to me, if we stop short of common knowledge there is no reason to go beyond the first iteration in the present context.

(51) Cf. Fuller (1969: 71, 162).

(52) On this cf. also section 6, sub (3).

(53) A form of power that may usefully resort to these strategies is what M. Foucault (1981) calls “pastoral power:” power exercised over human beings on the model of the shepherd’s leading of his flock.

(54) Scott (2001: 92).

(55) These features are grounded, in different ways, in all the RoL desiderata, and how they work together. Here, we consider them only to the extent that their instantiation is illustrated by the two desiderata (satisfying the two desiderata is not, however, a sufficient condition; see n. 24).

(56) Publicity, as I have said (section 2), is usually listed as a further RoL requirement. Are we conflating different phenomena, blurring conceptual distinctions? No, for two reasons. (1) What the RoL requires is publicity of the laws. Here, we are considering publicity as a property of the way in which laws conforming to the RoL influence the subjects’ behavior. (2) Different RoL requirements are interconnected, in various ways (so, for instance, prospectivity, intelligibility, and publicity are, in the RoL, mutually related features). So, it is no surprise that consistency and compliability should point to a facet of the ideal of publicity. Further aspects of this ideal are sorted out in Celano (2011).

(57) Cf. Raz (1979: 222): “a legal system which does in general observe the rule of law…attempts to guide [people’s] behaviour through affecting the circumstances of their action. It thus presupposes that they are rational…creatures and attempts to affect their actions and habits by affecting their deliberations.”

(58) This should not be taken as holding across the board. Think, for instance, of threats having a “Your money or your life” structure. These do not in fact offer the subject the choice they apparently offer her. In case the subject complies, the gunman will get her money. In case she doesn’t, the gunmen will get both her life and her money. This is, in fact, no (well-formed) alternative. The latter hypothesis includes the former—they are not logically independent. The real alternative (and the choice) is between losing or not losing one’s life: one’s money is lost anyway.

(59) Lukes (2005: 10–11).

(60) Strawson (1962: 87). “To adopt the objective attitude to another human being is to see him, perhaps, as an object of social policy; as a subject for what, in a wide range of sense, might be called treatment; as something certainly to be taken account, perhaps precautionary account, of; to be managed or handled or cured or trained” Strawson (1962; 79). Strawson writes that “if your attitude towards someone is wholly objective, then though you may fight him, you cannot quarrel with him, and though you may talk to him, even negotiate with him, you cannot reason with him.” But this, it seems to me, downplays what is involved, by way of reasoning with someone, in talking to him, or in negotiating with him.

(61) I say “minimally” because conformity to the RoL, as remarked in section 3 (following Raz), should be understood as being only a necessary condition for the law to respect human dignity.