1Paolo Sandro’s book The Making of Constitutional Democracy: From Creation to Application of Law (Hart Publishing, 2022) embarks on a twofold task: to “defen[d]”1 the distinction between creation and application of law, and to show how fundamental tenets of modern constitutional theory depend on that distinction. The first task is probably unnecessary; the second task yields exciting insights, which would be enriched by a better account of normative power.
- 2 Ibid. He goes as far as to say that scepticism about the possibility of application of law “is nowa (...)
- 3 As Sandro knows and to some nice extent captures, particularly in chapter 2.
- 4 But Sandro’s is hardly “the first of its kind” (Sandro 2022: 3; similarly at 227): consider for ins (...)
2I say that the first task is probably unnecessary because I don’t think sceptical views about the distinction are nearly as extended or convincing as Sandro claims they are when he sets out to “resist” them — indeed to “resist” the claim that “there is no such thing as application of law”.2 For centuries, if not millennia,3 legal thought has been premised on the assumption that law created in the past provides reason to engage with it now, by way of giving effect to what was then settled and must (so far forth) not be reopened for discussion. In this sense it is a fundamental presupposition of legal reasoning that there is such thing as law application as distinct from law creation, even if few theorists have a dedicated account of the idea of law “application”,4 and even if much theoretical concern with this idea has (perhaps understandably) exhibited a critical hue. None of this is to suggest that an explanation of the idea is worthless. Quite the contrary. While it is well understood that there is a rich spectrum between sheer creation and sheer application of law, few have ventured to map that spectrum in ways relevant to both legal theory and practice. Sandro makes a significant contribution in identifying different variables in play as one moves closer to either pole. The value of this effort is not to trump what would otherwise be default scepticism (there is no such default), but to highlight what is at stake, both legally and institutionally, in engaging in different kinds of discretion. This feeds into the book’s second task, to which the notion of power is central. This second task will be the focus of my discussion: in particular, its account of normative power.
3In what follows I shall highlight three respects in which the book’s account of normative power is incomplete at best. They weaken the book’s whole enterprise, including its account of law application. They concern: the limits on normative power (section 3), the distinctive way normative power is exercised (section 4), and whether normative power is other-regarding (section 5).
4The root of these shortcomings, and the key to overcoming them, lies in the less than fully articulated account of power developed in chapter 1 of the book.5 I therefore start, in the next section, by looking at this account of power and the three loose ends it leaves.
- 6 Sandro 2022: 28.
- 7 Sandro 2022: 28.
- 8 Sandro 2022: 28.
- 9 Sandro 2022: 29.
5Sandro’s discussion of power gets off to a sound start as he disambiguates different senses of the (English) term “power”. They include the contrasting notions of “power” understood as an ability to do something, and “power” understood as “potentiality, or also the strength (or its measurement) of an entity”6 — as in the power of an engine or a CPU. After further remarks on alternative distinctions, however, Sandro settles for a master dichotomy between “power to” and “power over”. Both these senses appear to be (Sandro does not say so) instances of the initial first sense of power, that is, they capture some kind of ability to do something. By “power over” Sandro understands “the immanent or original ability of most living beings to affect reality, that is, to shape and change the structure of the sensorial world”:7 for example, by transforming a piece of wood or leaving traces on soil. By “power over” he understands “the capacity to influence and cause behaviour… on the part of others”:8 for example, through acts of persuasion, coercion or command. It follows that both “power to” and “power over” are abilities wielded by animate beings as opposed to inanimate things. We can further surmise that the exercise of “power over” involves human agency, for Sandro says it can be possessed “not just by individuals and groups” but also by human artifacts such as institutions.9
6The master dichotomy between “power to” and “power over” does not aid Sandro’s argument. The remit of chapter 1’s discussion of power is to set the basis for the understanding of normative power that will be invoked in later chapters. Perhaps most immediately, in the context of chapter 1, the remit of the discussion is to identify political power as a species of normative power. The dichotomy is not fit for either purpose. For one thing, the contrast it draws is unstable. The very term “power to” does not denote something opposed to “power over” but rather a general category of which “power over” is a species: after all, “power over” is also an ability to do something (to influence and cause behaviour on the part of others). To discern how Sandro envisages the contrast between “power to” and “power over” one must attend to the above definitions. As per the definitions, “power to” is an ability to affect the sensorial world, whereas “power over” is an ability to affect the behaviour of others. Both are powers to affect something; they differ in what is affected by their use.
7Or so it seems. Nevertheless, upon closer scrutiny, the criterion I have just outlined also proves to be a shaky basis on which to distinguish “power to” from “power over”. For example, I can affect the sensorial world by opening the door, or by persuading you to open the door. In the latter case I have affected the sensorial world by affecting behaviour. Conversely, one can affect behaviour by affecting the sensorial world — say, by building a border wall. The criterion needs revising, both in order to be stable and in order to fit Sandro’s examples. The key distinction he seems to have in mind concerns not what is ultimately affected but how it is: that is, the primary means to which the agent resorts in doing what he does. When someone goes about persuading, coercing, or commanding another, their act has a nature that is altogether different from the act of building a wall or closing a door. The former involve engaging with another’s rational agency in a distinct way, which is absent in the latter kinds of act. But Sandro does not spell out what that distinct way is, and that omission carries over to the most important move in chapter 1’s discussion of power.
- 10 Sandro 2022: 29.
- 11 Sandro 2022: 30.
- 12 Sandro 2022: 29.
8That move is the contrast, within the category “power over”, between “normative power” and “non-normative power over”.10 According to Sandro, that contrast “sits prominently at the core of political and legal theory”.11 One would have therefore expected a more thorough explication of the contrast than the account delivers. “Non-normative power over”, we are told, “amounts to influence, persuasion, or (brute) force or strength: the natural fact that one animal is stronger than others and can influence or control their behaviour”.12 We may pause for a moment to question the inclusion of brute force in this list. It is one thing to persuade you to open the door and quite another to use you as a human weapon to break the door open. In the latter case, when I use brute force, I engage with your body as an element of the sensorial world rather than with your rational agency. It would be more coherent with Sandro’s discussion to regard the deployment of brute force on another person as an exercise of “power to” rather than “power over”. Be that as it may, it is in the idea of normative power that both Sandro’s and our interest lies.
- 13 Sandro 2022: 29.
- 14 Sandro 2022: 29.
- 15 Sandro 2022: 27, 30, in approving reference to Raz 1990: 2-3.
9Sandro only explains the idea of normative power by appealing to shorthand definitions found in the literature, which themselves have received fluctuating treatment by different writers. He describes normative power indistinctly as “the meaning of practical authority”,13 “the normative power to change another’s normative relations”,14 and the “right to rule”.15 He does not endeavour to articulate what exactly distinguishes normative power from “non-normative power over”, or to pronounce further on the former’s features. There are at least three respects in which this imprecision compromises later stages of the book’s argument.
- 16 Sandro 2022: 28 (the general definition of “power over”).
- 17 Sandro 2022: 29, my emphasis.
10One respect is the other-regarding focus of Sandro’s account of normative power. As we have seen, he speaks of power to affect behaviour “on the part of others”,16 and to change “another’s normative relations”.17 That focus is perhaps consistent with chapter 1’s immediate concern with political power, and thus with power as a relation between ruler and ruled — between the authority changing relations and the subjects liable to have their relations changed. But it sits uneasily with the exercise of other kinds of normative power, which become prominent later in the book, including powers to bind oneself (as well as others) through the making of contracts, marriages, wills, or other forms of relation contemplated in the law. Note how this skewed focus is a direct consequence of the master dichotomy between “power to” and “power over”, where the latter is conceived, in a rather top-down way, in terms of affecting behaviour (naturally someone else’s), rather than (as it should be) exclusively in terms of shaping normative relations, i.e., the reasons for such behaviour.
- 18 See e.g., Raz 1975: 102ff.
11Another loose end left by the account of power concerns the distinct way that “power to” and “power over” are exercised, and in particular, the distinct way, if any, that normative power is exercised as compared to “non-normative power over”. This tallies with a well-known conundrum in the literature on normative power.18 Normative power certainly “change[s] another’s normative relations”, as Sandro says, but so can an act of coercion or assault. In being assaulted, someone acquires rights and powers they would otherwise not have. A mere residence change can alter the duties, powers, and rights of tax authorities in relation to the relevant citizen, and may be sought by the latter for this very reason. We (including Sandro) would not wish to describe these as exercises of normative power, but Sandro does not equip us with considerations to justify why.
- 19 Sandro 2022: 30ff.
- 20 Sandro 2022: 37.
- 21 Sandro 2022: 37.
- 22 Hart 1961: 91ff.
12The third loose end I am concerned with is left towards the end of chapter 1, in the context of a discussion of the emergence of political authority (“From Powers to Power…”).19 Here, Sandro points toward an additional criterion for distinguishing between normative power and “non-normative power over”. He argues that the former, unlike the latter, “must be necessarily generated by something”.20 By what? Normative power cannot exist, he says, in the absence of “social rules” that regulate how to exercise it.21 “Non-normative power over”, by contrast, can be wielded in a social vacuum. Coercion does not presuppose social rules. Sandro’s exposition of the social rules presupposed by any exercise of normative power runs parallel to Hart’s account of the emergence of secondary rules to supplement the defects of a regime of primary rules, thereby introducing institutionalisation into what would otherwise be a purely customary order.22 Unfortunately Sandro does not work these considerations back into his original distinction between normative power and “non-normative power over”. Crucially, as we will see, he does not spell out the implications they have for the question of limits on normative power that will be the subject of his second chapter.
13I now turn to the problems generated by each of these loose ends. I tackle them in reverse order.
- 23 Sandro 2022: 59.
- 24 Sandro 2022: 50, 53.
- 25 Sandro 2022: 59.
- 26 Sandro 2022: 64.
14Take the question of limits first. A central thesis of the book’s second chapter is that “[c]onstitutionalism is realized whenever the exercise of political power through law is limited juridically”.23 Sandro eloquently explains such limitation of law by law in terms of a “duality” between two bodies of law within one and the same legal order.24 He dubs the two bodies “lex” and “ius”. “Lex” is the immediate product of political power and hence of the sovereign’s will. “Ius” is positive law that does not likewise have its source in the sovereign’s will. It may consist in a tradition of judge-made law, or law developed through judicial construction and enforcement of a constitutional document. Crucially, both the source and the administration of “ius” “are institutionally beyond the reach of the political authority”.25 Where law thus unfolds into “lex” and “ius”, where law itself is legally limited, where such “legal otherness”26 is found, constitutionalism takes hold — so the book argues.
- 27 Sandro 2022: 60.
- 28 Sandro 2022: 44.
- 29 Sandro 2022: 51-52.
- 30 Sandro 2022: 24.
- 31 Hart 1961, especially chapters 4 and 6.
- 32 Hart 1961: 111-112.
- 33 Section 2 above.
15But Sandro wavers on whether such legal limitation of law-making power is inherent to law, or rather is a feature of certain political cultures that legal systems may more or less approximate to. On the one hand, some of Sandro’s remarks in chapter 2 suggest that legal limits on law-making power are contingent, something to be fought for, which accordingly may be lacking. He speaks of “the potentially unlimited nature of the normative power of the ruler and the need for its limitation”.27 He circumscribes his interest in the book, and the focus of the ensuing discussion of political power, to “modern constitutional democracies”.28 He even refers to “ius” as an “invention” of the late Roman Republic.29 Yet, in other ways his discussion points to “legal otherness” as an inherent feature of law. The ample historical evidence drawn on to discern varieties of “legal otherness” suggests that its roots lie well beyond “late modernity”.30 Moreover, the book’s discussion of “ius” as a body of customary rules regulating the making of law is close again to Hart. Hart insists, in his critique of Austin, that at the foundation of any legal system lies not a person or group, who is habitually obeyed, but a customary rule of recognition.31 Like Hart’s rule of recognition, which is manifested in a practice among legal officials, Sandro’s “ius” chiefly develops through an official practice whose source and administration are beyond the legislator’s reach. Hart’s rule of recognition can be regarded as a body of positive law,32 like Sandro’s “ius”. Indeed, as we have just noted,33 Sandro says at the end of chapter 1 that any normative power must be “generated” by social rules. It is to be regretted that these various threads of the book are never openly connected.
- 34 Hart (1961: 67-68), for one, insinuates this distinction but does not dwell on it.
- 35 Sandro (2022: 26, 44) broadly claims political power has an “internal relation” to law.
16The question left open here is an important one. Sandro is not alone among legal and political theorists in failing to address it head-on, and to disambiguate important sub-questions that make it up. For instance, it may be one thing to say that political power presupposes a body of social rules and another to say that such rules are limitations on that power. Rules merely establishing the manner and form of law-making are far from placing the kinds of limits on the content of legislation we find in modern constitutional practice.34 Nor is it perhaps adequate to conflate legal with political power, or either with normative power more generally, when then the issue of limits is at stake.35 The book’s theoretical set-up and Sandro’s expertise would have furnished a good opportunity to shed light on these issues and their implications for, inter alia, an understanding of sovereignty, statehood, and even core doctrines in public or private law. That the book at least provokes reflection on point is to its credit.
17Chapter 6 is a crucial step in the book’s defence that there is such a thing as law application. As far as I can see, Sandro’s core argument is this: there is such a thing as law application because legal power-conferring rules require application, and every legal order has power-conferring rules. It is worth noting from the outset that, had Sandro developed a firmer stance on the question of limits earlier in the book, had he less ambiguously concluded in chapters 1 and 2 that even a law-maker’s decisions are necessarily governed by and subject to law, his defence of the inevitability of law application might well have taken a shortcut. For it would have followed from that conclusion that even law-making —the central case of law creation— involves an application of law by way of giving effect to pre-existing legal provisions. Sandro takes a longer route, which peruses legal powers of all kinds within a legal system, including private powers. That broader focus is welcome, as it enables him to stress the pervasiveness of law application. Unfortunately, however, the book’s initial imprecision about normative power takes its toll on the argument here as well.
- 36 Sandro 2022: 233-234.
- 37 Sandro 2022: 234.
- 38 Hart 1982: note 8 at 196; see further Köpcke 2019: 23ff.
- 39 Sandro 2022: 29 (see section 2 above).
18The chapter does not spend much time arguing that there are such things as power-conferring rules in law. The discussion quietly slips in the distinction between legal duty-imposing and power-conferring rules,36 and in a short paragraph it refers to Hohfeld’s and Hart’s accounts of the latter.37 Hart, by contrast, devotes substantial portions of his book to discrediting various views, including Austin’s, that reduce legal powers to legal duties. Hart certainly leaves much “unfinished business” (Hart’s term) in accounting for legal power-conferring rules.38 As Hart is aware, any defence of the distinctness of power-conferring rules must justify the need, the rationale, of individuating legal material into rules of this kind. This, Sandro does not do. The account of normative power he has provided up to that point, in terms of changes in normative relations and the pre-existence of positive rules, is plainly insufficient for this purpose. For one thing, as I noted earlier, conduct governed by legal duty-imposing norms also “change[s] another’s normative relations” in law.39 It may even be performed for just this purpose: people have been known to commit minor offences to secure a warm shelter in prison during Winter. For another thing, the ability to thus alter legal relations presupposes the existence of a set of positive rules: namely those that impose the legal duties. These considerations are insufficient to get the duty/power distinction off the ground.
- 40 Sandro 2022: 232.
- 41 Sandro 2022: 232.
- 42 Sandro 2022: 232.
- 43 Sandro 2022: 232-233.
19Failure to get the duty/power distinction straight spills over into Sandro’s account of law application. The account is correspondingly problematic. The account’s starting point is that law application is a species of law compliance. By complying with a norm, Sandro understands satisfying its antecedent — doing “what a norm requires”.40 Sandro goes on to distinguish “merely complying” with a norm from doing so “intentionally”.41 Complying intentionally means, to him, complying “in order to satisfy whatever the norm requires of us”.42 Applying the law, he concludes, is complying with the law with just this kind of intention.43
20It should immediately dawn on the reader that this account of law application is overinclusive. It is coextensive with what is normally called “obedience”. One obeys a norm by doing what it says because it says so. Stopping at a red light is an instance of law application on this view, provided the agent acts because the law says so. Sandro is of course right that legal duty-imposing norms do not require application, in the sense that mere compliance, for whatever reason, will fend off sanctioning for legal breach.44 But the sole fact that stopping at a red light because the law says so counts as applying the law on Sandro’s account, demonstrates that there is something deeply amiss with the account, at least as it stands.
21Sandro seems to be after something else. He wants to say that there is some tight connection between law application and legal powers, a connection that does not exist, or is not quite as tight, as between law application and legal duties. But defining application in terms of intentional compliance will not do. It is not even clear what it means to comply with a power-conferring rule because the law says so. Power-conferring rules do not require that one does what they say in the manner that duty-imposing rules do. A power-conferring rule sets out the steps to follow if one wishes to attain a certain result, but does not pronounce on whether one should attain the result. (There may of course be a legal duty to attain that result, and thus to use the power, as in the case of many official powers.) To be sure, a power-conferring rule imposes a requirement in an instrumental sense: it is those steps, and not some others, that are to be followed if the relevant result is to be attained. But is this a requirement of intentional compliance?
- 45 Sandro 2022: 232.
- 46 See further Köpcke 2019: 109, 120-121.
22Sometimes Sandro suggests that the distinctive mark of the exercise of a legal power is awareness of the rule, rather than the intention to comply with it.45 There is some good sense to this. Clearly the exercise of certain legal powers involves, almost necessarily, some awareness on the agent’s part that he is following a legal rule, perhaps even of the particular rule he is following. This is especially true of public powers. Given the fiduciary nature of public power, core tasks of legal officials can only be discharged when and to the extent that there is a legal warrant for them.46 Certain formal instances of the exercise of private powers may also fall into this category; think of the process of conveyancing a house (Sandro’s example). But it does not apply to countless ordinary exercises of legal power, such as shopping at the corner store. Conversely, compliance with certain tax duties is sometimes inseparable from awareness of the existence and content of the relevant rules.
- 47 In this and the next two paragraphs, I draw on my own account of legal power in Köpcke 2019, chapte (...)
23Perhaps the criterion can be refined by adding two other elements scattered throughout Sandro’s discussion.47 According to chapter 1, normative power is power “to change… legal relations”, and according to chapter 6, power-conferring rules require “intentional” compliance. Associating legal power with an intention to change legal relations is therefore compatible with Sandro’s account. Does this not run us into difficulties already flagged above? Not, if two crucial qualifications are made. The first one concerns the unreflective corner store customer. He acts with the intention of changing legal relations under an appropriately general description, such as “getting the carrots”, “buying dinner”, etc. Even a house buyer may lack “awareness of the rule”, despite intentionally going through the steps in order to bring about the relevant normative result, however non-technically understood by him.
24The second qualification is more important. Sandro himself touches on it in passing. Sandro anticipates the objection that power-conferring rules may be complied with unintentionally, as where one utters the wrong words, or accidentally presses a digital button, but is nevertheless held to be bound in law. He rightly replies that, in law, an agent’s intention is determined by reasonable inference from his or her actions.48 What remains to be said is that exercises of legal power always thus appear to be intentional because the intention involved in exercising a legal power is of a particular kind. It is an intention to perform the act by means of the very manifestation of that intention. Legal powers are exercised, as it were, by saying so. This is made explicit in the legal expression “hereby”, characteristic of formal exercises of legal power, but it is likewise the way one makes a binding offer to buy the carrots by muttering “these please” and placing them on the counter. Any private and public powers are exercised by saying so, whatever else it takes to exercise them, and even though requirements on the expression of one’s intention vary greatly between more formal exercises of power and more casual or informal ones. What matters for our purposes is that the act of saying so is of a different kind altogether from that of stopping at a red light, moving house, or waving a knife at another person. This is a criterion to stably distinguish between exercises of normative power and other kinds of act, including deployment of “non-normative power over”. One may persuade or coerce by saying something, but not by saying so. “I hereby persuade” won’t win an argument any more than “I hereby trespass” amounts to a tort.
25Note that the distinctive mark of power-conferring rules I am highlighting focally concerns the way they are complied with rather than the intention they are complied with. As a matter of fact, the intention may be absent in accidental exercises but the modus operandi or technique remains. It is a technique that amounts to saying something and thereby making it true — in the framework of a set of pre-existing rules that so provide. This, in my view, is the relevance of Sandro’s insight that the existence of a normative power presupposes a regime of (positive) rules.
26Where does all this leave Sandro’s argument about the inherence of application to law? Recall that Sandro brings in legal power-conferring rules to show that their exercise requires law application, and thus that application is inherent to law. But we ruled out his account of application as intentional compliance and found that the latter is not a mark of power-conferring rules anyway. Does our revised account of power-conferring rules closely relate to the idea of law application?
- 49 For this understanding of law application, see Köpcke 2019: 105ff. Sandro (2022: 228ff) dismisses a (...)
27Answering this question would require a revised account of law application. I will not provide it here except for noting that the place to start, it seems to me, is not compliance. In assuming that application is a species of compliance, Sandro takes an unduly narrow focus. In this regard his account of application is underinclusive. From the outset, it rules out the important sense in which a legal rule can be and is routinely applied simply by being used in reasoning to reach a normative conclusion — as, for example, where a lawyer advises that when “applying the new tax regulation” one’s duties are such and such.49 In this kind of scenario, one applies a law without thereby complying with a power-conferring rule, and thus without changing normative relations. I am inclined to agree with Sandro that there is a significant connection between law application and the exercise of legal power, but I deny that it is as close as he portrays it. He wants to argue that only in exercising legal power can one apply a legal rule, and that all legal power-conferring rules — also those in the private domain — require application (in the sense that application is what it takes to comply with them). For my part, I see little gain in defining law application in terms of a notion we already have a name for: namely, the exercise of legal power. It is more illuminating to track the ordinary understanding of law application as something only some exercises of legal power involve, and which can likewise be undertaken independently of an exercise of power. Focus on the intersection between law application and legal powers, particularly on the idea of official power, invites reflection on the need for officials not only to have but also to provide a legal warrant for their actions, and the reasons for doing so. It is a fertile ground to probe important doctrines in legal and political theory, including Sandro’s defence of the inherence of application to law.
28I come, by way of conclusion, to the other-regarding focus of Sandro’s account of normative power: the fact that he characterises normative power as power over another.50 We have seen that the idea of normative power runs through a number of different junctures of the book’s argument. In particular, it underlies Sandro’s discussion of the limits on political power, and thus of “lex” and “ius”, in chapter 2. It also underlies the discussion of law application, and thus of powers in the law, in chapter 6. But the other-regarding focus is ill-suited to this second discussion. Here power understood as a “right to rule” does not fit as comfortably as it does in the context of political power. Many powers in private law cannot, without distortion, be conceived of as powers over someone else. The power to make a contractual offer is focally a power over oneself. It is an ability to bind oneself to another, by bestowing rights and further powers on them. On Sandro’s scheme, however, the power to make a contractual offer is a power over the offeree — potentially a power over every member of the public — who wields a corresponding liability to be bestowed a faculty. Self-binding powers, so central to private arrangements, are poorly accounted for on the model of ruler and ruled.
- 51 Except in a cryptic passage in the conclusion of chapter 6 (Sandro 2022: 245), which is concerned o (...)
29This incongruence is but a symptom of a deeper malaise. The book does not connect its discussion of political power (chapters 1 and 2) with its discussion of powers in the law (chapter 6).51 But the connection is central to the remit of Sandro’s enterprise and its concern with democracy. The model of ruler and ruled envisages a law-maker issuing commands that bind the ruled. If the law-maker is democratic, there is a sense in which the law-maker binds itself, a sense that already escapes the top-down logic of “power over”. But if, moreover, the laws thereby made are not only duty-imposing but also power-conferring, as is inevitably the case in any mature legal order, this gives rise to an allocation and sharing of power that is of utmost relevance to the aims of constitutional democracy. The exercise of legal power enables an agent — you or I — to shape normative relations as he or she sees fit, within the framework that regulates the exercise. Though the task of legally moulding the normative landscape begins, in a sense, in the hands of the law-maker, it is pursued and sustained on an ongoing basis through the choices of public institutions and private citizens within the scope afforded to their self-direction, both collective and individual. Sandro fails to show how legal powers are privileged tools to realize democratic ideals – to make constitutional democracy, through the exercise of a unique power to make something true by manifesting one’s intention to that effect. Having narrated in chapter 2 the transition “from powers to power”, that is from natural powers to political power, Sandro surprisingly neglects the ulterior step from power to powers: from political power to legal powers. He neglects the step from application back to creation of law.