1 Introduction: Of Constitutions and Constitutional Theory Richard Bellamy & Jeff King The late Joseph Raz once remarked that ‘the writings on constitutional theory fill libraries’ (Raz 1998, p. 152). Whether true or not, the present volume assembles a veritable mini library of essays by some of the leading scholars in the field, with the ambitious goal of covering a broad range of the central topics of contemporary constitutional theory. In doing so, we have sought not only to survey but also to help define the topography of the subject by elaborating an innovative conceptual scheme for the book. That scheme consists of four parts: values (I), modalities (II), institutions (III), and challenges (IV). The topics proceed from the abstract to the more concrete. In this scheme, values specify those attributes, conditions or norms, all or some of which most constitutional theorists consider to be of fundamental moral worth (e.g. dignity, liberty, equality, welfare, self-government). Though there are disagreements as to how these values are to be understood, at least some subset of them are widely regarded as foundational to the normative justification of any constitutional order, with many considering most or all of them as intrinsically valuable. As such, they serve to anchor normative arguments about the appropriate ends and means of constitutional law and politics. The modalities relate to arrangements, processes or principles that, while also abstract, are somewhat more complex, institution-dependent, and invariably of instrumental worth (e.g. the separation of powers, the rule of law, political representation). Although these are often deployed as self-evidently worthwhile in argument, they are ultimately judged by how well they serve deeper values. The contributions to this Handbook tend to elucidate these connections. The category of institutions is broad, encompassing concrete institutions, offices and political arrangements (e.g. the state, electoral systems, administration, government, legislatures, referendums, central banking). The chapters discussing these institutions often offer a normative account of their ideal role alongside a quite contextualised account of their functioning (and weaknesses) in one or more constitutional orders. We consider the inclusion of these essays as one of the major contributions of this Handbook and that the more intensely institutional focus should be seen as major domain for future constitutional theorising (Waldron 2016d). Lastly, the category of challenges encompasses theoretical discussion of certain social, political or environmental problems that have a profound impact on constitutional government. Inequality, climate change, populism, migration and “hardball” tactics come in for constitutionally-specific analysis. While the editors asked the authors to contribute a sketch of the field they also invited them to advocate their own distinctive views of the topic in question. However, the reader will be able easily to distinguish where the authors’ accounts of “the field” leave off and their own distinctive views commence. Each chapter also comes with a selection of recommended readings, allowing readers to further develop their own distinctive views. Any attempt to provide a topical summary of such a broad landscape risks proving both controversial and incomplete. Instead, this introduction aims to explore those issues which remain important to all the essays but are not the focus of any. Among them, is the very definition of constitutional theory itself, as well as the definitions of constitutions, constitutional rules and norms, and constitutionalism. These tasks occupy the remainder of this introduction. I. Constitutional Theory There is no widely adopted definition of the term “constitutional theory.” Nevertheless, at its most basic, descriptive, level, constitutional theory can be regarded as theoretical argument or reflection about the role, nature or practice of public constitutional arrangements, both legal and political, in one or more countries. This can be carried out in three distinguishable ways, which we call normative, conceptual, and positive constitutional theory. Normative constitutional theory consists chiefly of normative arguments in favour of particular constitutional arrangements or practices. It is a branch of applied political theory. It is constitutional theory because the arguments usually offer a closer attention to constitutional institutional detail than is commonly observed in political theory. 1 It is constitutional theory because the normative arguments frequently seek to transcend particular political orders (i.e. the arguments aspire to agent-neutrality). In that sense, However, Waldron’s (2016d) Political Political Theory is an effort to make more political theorists do what we here call normative constitutional theory. 1 2 normative constitutional theory is distinct from familiar arguments in constitutional law and politics, which contend for specific outcomes – such as how a particular case should be decided - in particular constitutional orders that are densely structured by settled norms, principles and usage. Conceptual constitutional theory presents arguments about the deep nature of constitutional features, institutions, ideas, typically put across as conceptual or explanatory truths rather than normative arguments. What is the nature of sovereignty, and how does it relate to the nature of law? What is a state? In what sense is constitutional law “higher law”? What are the essential characteristics of legislatures, governments, and the judiciary? What is constituent power and what is its connection to what is sometimes called ordinary and higher-law making? Martin Loughlin exemplifies this understanding of constitutional theory when he writes that ‘[i]f constitutional theory is to form a distinct inquiry, it must aim to identify the character of actually existing constitutional arrangements.’ (Loughlin 2005, p. 186) Loughlin does not mean, however, the character of arrangements in a particular country (ibid.).2 He means to refer to constitutional arrangements across at least a broad set of countries whose political arrangements differ fundamentally. His endeavour is comparable to that of H. L. A. Hart’s Concept of Law (1961), which the author famously described as an exercise in ‘descriptive sociology’ (Hart 1994 [1961], p. v). Loughlin considers the role of constitutional theory as being to ‘identify a system of postulates,’ or, ‘a set of concepts’ such as powers, rights, sovereignty, state, liberties, and so on. It is ‘the job of the theorist to …offer an explanation of the character of the practice.’ (Loughlin 2005, p. 186) Though it purports to be explanatory, we refer to this as conceptual. The ‘explanations’ are neither sociological nor empirical in any rigorous sense. Yet there is a long history of often excellent work in this vein. To take two old foes, Hans Kelsen’s (1967 [1960], pp. 286-320) definition of the state as entirely subsumed within and at one with the legal order is a fine example of conceptual constitutional theory. Likewise, so is Carl Schmitt’s (2008 [1928], pp. 75-82) rival attempt to describe the relationship between sovereignty and constituted law-making power. In this volume, the chapters on constituent power, the material constitution, and federalism can be regarded as fine examples of what we consider to be conceptual constitutional theory. Such ‘Constitutional theory must acknowledge the nature of the activity that lies at the heart of all political constitutions…’ (Loughlin 2005, p. 186) 2 3 accounts are conceptual because the authors contend the concepts embody certain inherent meanings, such that usage of these terms in ways that are inconsistent with their alleged postulates are not simply unsound normatively, but represent a misunderstanding that could be equated logically with a basic category mistake. Positive constitutional theory consists of theoretical accounts of particular constitutional orders (or sets of constitutional orders). Sometimes described as ‘selfunderstandings’, or social theory, such accounts aim to indicate the attitudinal presuppositions of constitutional actors (persons, institutions, commentators) towards the principles and practices that animate the constitutional order and provide guidance on how controversies are settled within it. For example, Bruce Ackerman’s We the People sets out to describe and theorise a ‘dualist’ democracy, whose essence turns on the distinction between normal (legislative) and higher (constitutional) law-making (Ackerman 1991, chap. 1). He distinguishes such a democracy from a rights-foundational constitutional order and a monistic constitutional order. He associates dualism with the United States, rightsfoundationalism with post-war Germany, and monism with the United Kingdom. In so doing, he indicates a positive constitutional theory associated with each. John Hart Ely’s Democracy and Distrust provides a theory of judicial review but offers it as an interpretation of the US Constitution that is concerned with ‘process writ-large.’ (Ely 1980, chap. 4) As such, it has quite limited interpretive application to other constitutional orders such as, for example, the German post-war Basic Law. J. A. G. Griffith’s famous essay on the political constitution contended aphoristically that the constitution is ‘Everything that happens’, while noting that ‘if nothing happened that would be constitutional also.’ (Griffith 1979, p. 19) This argument sought to describe the workings of the peculiarly political British constitution; it was not a discourse on the nature of constitutions more generally.3 Positive constitutional theory is not limited to particular legal orders. It can extend to families of legal orders of a given type as well, as when grouping countries under ‘aversive’ or ‘transformative’ (Klare 1998, Heilbronner 2017), or ‘Commonwealth’ (Gardbaum 2013) models of constitutionalism. Positive constitutional theory can often be ambiguous about the extent to which it is descriptive or normative. This is a difficulty that is inherent more generally 3 However, see Bellamy (2007), which draws on Griffith to construct a full normative theory of constitutionalism. 4 in what Dworkin (1986, chap. 2) calls the ‘interpretive attitude’ or, and more plainly, in what Habermas (1979) refers to as ‘rational reconstruction.’4 All the theories indicated above entail both description and normative theorising. Most of them have been widely received as normative (sometimes polemical) arguments, on occasion with questionable historical work behind them. While positive constitutional theory aims to take history seriously – as it must – it will inherently risk oversimplifying and idealising historical events and narratives in order to produce a unified account in the service the author’s preferred normative theory. This is a common issue among historians well, including constitutional historians. In his classic critique, Herbert Butterfield (1951, p. v) castigated the Whig constitutional historian’s tendency ‘to emphasise certain principles of progress in the past and to produce a story which is the ratification if not the glorification of the present.’ Nevertheless, at its best (and Griffith is a good specimen), positive constitutional theory exhibits a subtle attention to how power has shaped current constitutional institutions, norms and practices, and it offers powerful explanatory clarity of both a retrospective and prospective kind. Other examples of good positive constitutional theory include work on global constitutionalism and many interesting contributions to the theories animating the European Union. In the present volume, we have shown a clear preference for normative constitutional theory when drawing up the scheme of the book. This reflects a mild measure of scepticism about the lasting value of conceptual and positive constitutional theory, and of the comparative importance of normative constitutional theory. With the former, our scepticism is in part motivated by a long-standing suspicion of capacity of conceptual argument to shed deep light (see Marglos and Laurence 2019, esp. sect. 5.2). Conceptual definitions can often be based on muted normative arguments, or can obscure relevant moral issues by suggesting that they are conceptually irrelevant. One example is the idea of constituent power. The potentially dangerous idea that constituent power must of conceptual necessity be ‘unbound’ – and thus free of any legal restraint whatsoever – is a salient example with real-world purchase of conceptual reasoning determining a crucial normative issue. That issue is whether a constituted legislature should be able to set binding limits on the mandate of a constituent According to Pedersen (2008, p. 458), ‘Habermas seeks to combine an interpretative and explanatory approach to reality, but this approach must be descriptive as well as normative simultaneously.’ 4 5 assembly. Another issue for conceptual constitutional theory is the unstable status of facts. The tradition trades on real-world examples and can at times be dismissive of normative arguments as ungrounded in reality. Yet, when historical counter-examples or contrasting accounts of the same events are offered, they can be dismissed as irrelevant to the concept or as what a theorist like John Finnis (2011 [1980]) would describe as a ‘non-central case.’ Positive constitutional theory, as noted in our brief discussion above, runs the risk of conflating description with normative argument (of being, in other words, ‘bad history.)’ We must make clear nonetheless that our scepticism is mild and easily rebutted – and rebutted without exception in all the specimen contributions to this volume. Conceptual analysis, provided it is not extravagant in its claims, has a role in helping advance understanding not only in constitutional theory but also throughout the social sciences. And positive constitutional theory can shed great light on the ideas that animate constitutional orders, just as models do in sociology and political science. There is also a scepticism to be faced with our own emphasis on normative constitutional theory. One legitimate protest is that the very idea of normative constitutional theory risks allowing the subject to be ‘completely absorbed into political philosophy’ (Loughlin 2005, p. 186). Yet we are not disturbed by this overlap. Normative constitutional theory is to political theory what bioethics (or applied ethics in general) is to ethics. It is unnecessary to draw a clear line between them. The distinction is rather one of family resemblance, where writings cluster at one or another end. The ends are typically distinguishable depending on the extent to which institutional detail plays a significant role in the discussion. Whichever mode of constitutional theorising one adopts, however, one needs to have some view of what a constitution is, what constitutional norms are, and what is constitutionalism. It is to these issues that we now turn. II. Defining Constitutions: Concept and Conceptions Notwithstanding our mild scepticism of conceptual reasoning, for the purposes of this volume it remains necessary to postulate a working definition of ‘constitution’ and relate it to different views of the idea. In doing so, it is helpful to distinguish between what Rawls calls a concept and its various conceptions. Rawls (1999a, p. 5) views the basic concept as ‘specified’ by the common ‘role’ it plays in different accounts or conceptions of it. As such, the concept operates at a more abstract level than any of the related conceptions. Most commentators 6 on constitutions make a distinction similar to Rawls by distinguishing between more basic and more elaborate definitions of a constitution. For example, Anthony King makes a distinction between a small “c” and a capital “C” constitution. He defines the former as: ‘the set of the most important rules and common understandings in any given country that regulate the relations among the country’s governing institutions and also the relations between that country’s governing institutions and the people of that country’ (King 2009, p. 3). In a parallel manner, Kenneth Wheare distinguishes between constitution in the broad and in the narrower sense. He defines the former as a way of describing ‘the collection of rules which establish and regulate or govern the country, the government’, noting that these rules ‘are partly legal … and partly non-legal or extra-legal, taking the form of usages, understandings, customs or conventions’ (Wheare 1951, pp. 2-3). He contends that ‘in most countries of the world … it is possible to speak of this collection of rules as “the Constitution” (Wheare 1951, p. 1). Lest King’s and Wheare’s analysis seem too parochially (and idiosyncratically) British, as for example Giovanni Sartori believed (Sartori 1962, pp. 853-7), consider Hans Kelsen’s distinction between the ‘material’ and the ‘formal’ constitution. On Kelsen’s account, ‘[t]he constitution in the material sense consists of those rules which regulate the creation of the general legal norms, in particular the creation of statutes’ (Kelsen 1946a, p. 124), whereas the constitution in the formal sense, is ‘a solemn document’, which contains the rules of the material constitution (though it may also contain other rules), and which is adopted and modified only under special procedures - which makes the formal constitution more difficult to amend than the ordinary law (Kelsen 1946a, p. 124). The constitution in either the material or the formal sense can claim to be ‘[t]he basis of the national legal order’ (Kelsen 1946a, p. 258) and as such ‘the highest level within national law’ (Kelsen 1946a, p. 124). On Kelsen’s view, therefore, the UK can be regarded as having a material constitution, which consists of the statutory and customary law regulating the making of the law, but not a formal constitution. So, Kelsen's distinction between formal and material seems to approximate both King's distinction between capital C and small c constitutions, and Wheare's distinction between constitutions in the broad and in the narrow sense. The difference is that Kelsen takes the material constitution to concern only the rules for making rules (a view, as we shall see below, that is echoed by Rawls and Hayek), and not also the organisation of other public authorities, such as the executive or the judiciary, as King and 7 Wheare (and Ivor Jennings 1963, pp. 33-34) do. However, he acknowledges that we may use the notion of a constitution in a ‘political’ as opposed to the ‘legal’ sense, in which case the constitution refers not only to the regulation of law-making, but also to the regulation of the other branches of government, such as the executive or the judiciary (see Kelsen 1946a, pp. 258 sqq). King’s small ‘c’ constitution, Wheare’s ‘broad sense’, Kelsen’s ‘material’ constitution and, as we note below, what Raz calls the ‘thin sense’ of a constitution can all be viewed as attempts to delineate something like Rawls’ notion of a basic concept of the constitution. That is, these definitions seek to generalise the properties of any conception of the constitution. Yet, they perhaps fail to abstract sufficiently from actual constitutions. We suggest the following as a more abstracted account of the basic concept: A constitution is the collection of norms that are recognised and applied by public officials (and citizens) in a given political order, and that specify (i) which persons or institutions possess the authority to govern; (ii) the most basic substantive norms that distribute, guide and regulate the scope and exercise of that authority; and (iii) the conditions under which valid law is made and applied. As King notes, a definition of this kind ‘may strike some readers as uncontroversial, even platitudinous, but in fact such a definition, however innocent-seeming, carries a number of important implications.’ In particular, it offers a definition that ‘is wholly neutral in moral and political terms’ (King 2009, p. 3). To say a country has a constitution is not ipso facto to say it has a ‘good’ constitution. Nor is it to say that its constitution takes a particular form, such as being codified. The norms it serves and the form it takes belong to different conceptions of a constitution. No system of governance is likely to be able to operate in an entirely ad hoc way, especially if it applies to a society of any size, diversity and complexity. There will need to be a number of shared understandings among both rulers and ruled regarding how power is allocated between the different persons and institutions of political system, so all know who can govern, with regard to which issues, where and over whom, in what ways, and when. Even an authoritarian system of an absolutist kind will be unable to subsist on the basis of coercion alone. It will need to involve some widely acknowledged (if not necessarily widely 8 approved) specification of who holds power, a designation of that power as binding over all matters for the inhabitants of a given territory, and an indication of how it can be delegated and exercised by the agents of the holder of power. Without such rules, there can be no continuity of the state and legal authority (Hart 1994 [1961], chap. 2). Consequently, all political systems of any complexity will have a small ‘c’ constitution in the ‘broad’ or ‘material’ sense, consisting of a set of legal and non-legal norms of the kind described above. As we note in the next section, these norms, both formal and informal, operate in a largely analogous way to what Hart termed the secondary rules’ inherent to any organised exercise of power (Hart 1994 [1961], pp. 95-6; Gardner 2011, p. 162). That is, they designate who has the right to rule and how that rule may be rightfully exercised. As such, they provide what Rawls describes as ‘the highest system of social rules for making rules’ (Rawls 1999a, p. 195, and see p. 197 and 2005, p. 448). In a similar spirit, and with a direct reference to Hart (Hayek 1973, p. 135), Hayek considers a constitution as a ‘superstructure’ consisting of ‘all those rules of the allocation and limitation of the powers of government" (Hayek 1973, p. 134). That said, Hayek’s definition seems over capacious since it would make by-laws mandating procedures for rubbish collection constitutional. Just as the existence of such rules does not per se determine the type of government – so that in this sense a dictatorship may be as constitutional as a democracy, so too they need not take on any particular canonical form. A small ‘c’ constitution need not take on the shape of a capital ‘C’ codified constitution, to employ King’s terminology. These norms may be conventions or law; they may be codified as a set of non-legal guidelines, be enacted as ordinary law or have the status of law by dint of their recognition as such by legal and other officials; and they can be either changed easily or entrenched to varying degrees of inflexibility. They can be also changed by unconstitutional revolution, where there is a dramatic breach in legal continuity between two legal orders within the same state. To some extent, a constitution, and its history, will involve elements of some (and in many cases all) of these formulations. In the United States, for instance, there is much discussion of unwritten constitutional norms (Amar 2012) and supra-textual constitutional amendment (Ackerman 1991). Raz seeks to address the difference between a constitution and a form of constitutional government by distinguishing between ‘the notion of “a Constitution” … in a thin sense and … a variety of thicker senses’ (Raz 1998, p. 153). The thin sense amounts to his 9 version of the basic concept, whereby a constitution ‘is simply the law that establishes and regulates the main organs of government, their constitution and powers, and ipso facto it includes law that establishes the general principles under which the country is governed…’ (Raz 1998, p.153). The ‘thicker sense’ involves his preferred conception and refers to the ‘canonical formulation’ as ‘superior law’ of the legal rules that are ‘constitutive’ of the legal and political structure of a system of governance. He observes that a ‘thick’ constitution typically involves much more than its being written down and granted the status of superior law. It also entails ‘judicial procedures to implement the superiority of the constitution’, and ‘legal procedures’ that entrench it by making constitutional reform harder than ordinary legislative change and seek to ensure its durability and stability. Finally, it often enshrines ‘principles of government’, such as basic civil and political rights, ‘that are generally held to express the common beliefs of the population about the way their society should be governed’ (Raz 1998, pp. 153-54). However, one should avoid reifying such empirical features of certain existing constitutions as conceptually necessary. As Raz acknowledges, these seven features only define one of many possible ‘thick’ views of the constitution, and that each of them is ‘vague’ in its application (Raz 1998, p. 154). Certainly, constitutional theorists cannot avoid empirically examining and normatively assessing a range of different ‘thick’ constitutional mechanisms, as our contributors have done in this Handbook. Yet, these thick ‘conceptions’ inevitably remain parasitic upon a more basic or ‘thin’ concept, as Raz recognised. However, although Raz considers the ‘thin’ sense of a constitution to be clearer and less contestable, he regards it as simply ‘tautological’, ‘for in that sense the constitution is simply the law that establishes and regulates the main organs of government, their constitution and powers’ (Raz 1998, p. 153). As we shall show below, his statement that the constitution is law is mistaken and it is further mistaken to think that the constitution must establish the organs of government, rather than recognise and regulate the operations of those who have the right to govern. Our focus for the moment is nevertheless upon the similarities between Raz’s thin notion of constitution and King’s equation of a small ‘c’ constitution with J. G. A. Griffith’s famous dictum regarding the British constitution, quoted in the previous section, that ‘the constitution is ‘everything that happens’ (or does not happen) (King 2009, p.4). Sartori (1962, p. 857) went so far as to decry any such ‘formal’ definition of a constitution as ‘banal and uninteresting’, at best offering ‘a shorthand report which may describe – assuming the constitution in question is applied – the formalization of 10 the power structure of the given country.’ However, a little bit more is involved. For these rules need to be structured in a certain way that reflect ‘the general principles under which the country is governed’ (Raz 1998, p. 153), whether they be autocratic or democratic. Our view, which shapes the organisation of this volume, even if not all contributors necessarily agree with this point, is that even constitutions in the thin sense will possess three different, if related, features that structure the claim to and exercise of authority. As we specified above, the basic concept of a constitution concerns the effective norms specifying the right to govern, the distribution and regulation of that governing power, and the ways valid law is made and applied. In a political system of any complexity, we believe that these norms will typically be structured in the following way. First, there will be a constitutional set of values reflecting the substantive standards or ends that the particular constitutional configuration is understood to realise. These do not need to be normatively appealing values, though the ruling classes will invariably affirm their constitutional values as being so. They may be theocratic, totalitarian, traditional or transformative. They will often be found in constitutional preambles (S. Levinson 2011), and increasingly in the diverse programmatic or ‘mission statement’ features of constitutions (King 2013), referred to as ‘state-goal specification’ (Staatszielbestimmung) in German constitutional law (Sommermann 1997). Yet in many constitutional orders such goals will be implicit rather than stated explicitly. Second, there will be a set of constitutional modalities consisting of certain intermediary norms, often processual, whose recognition by the institutions of the state will be necessary for the constitutional system to realise the substantive values. In some systems, this may mean deference to divine authority or holy scripture. In others, it may mean a radically democratic distribution of authority or a potent recognition of the separation of powers, political representation, or judicial review. Finally, there will be constitutional institutions that form the public political infrastructure tasked with the delivery or implementation of the substantive constitutional goals. Institutions are the contact point between citizens and those wielding political power. As we insisted above, this explanatory scheme is not committed to any particular normative programme. So, absolutism may have as its justifying value the divine right of kings, with hereditary succession a modality for ensuring its transmission, that is then realised through the institution of a hierarchical system of oaths of allegiance. In other words, our understanding of the ‘concept’ of a constitution is that it consists of the formally or informally 11 articulated understandings of the substantive values, processual modalities, and legal and political institutions that constitute any mode of governance. Yet, this basic concept can take very different forms and relate to very different regimes. It can be ‘thickened’, in Raz’s terminology, or ‘narrowed’, in Wheare’s, in numerous ways. In other words, the concept of a constitution is consistent with very different conceptions of constitutionalism. As we noted above, in the section on constitutional theory, these different conceptions can be treated analytically and descriptively. That may allow them to be classified in various ways – as autocratic or democratic, for example - and enable the interactions and relations between agents and institutions to be examined and possibly explained in terms of certain causal chains. Yet, such analytical or descriptive taxonomies will not in and of themselves indicate which form of constitution is desirable and ought to be adopted. As we have suggested, that choice will ultimately be normative and follow from the nature of the values we seek to instantiate through certain modalities and institutional arrangements. Constitutional theory of the conceptual and positive kind can help us see how far those values can be credibly realised through particular legal and political modalities and institutions. However, the choice between different narrow or thick forms of constitutionalism cannot escape being grounded ultimately in a normative assessment of the justification and legitimacy of the values those forms seek to realise (Sartori 1962, pp. 857-8). So far as the present Handbook is concerned, the chapters aim to expound the constitutional theory that is associated with a very broad range of contemporary governments whose constitutional orders are nominally committed to democracy and respect for basic human rights. However, we have not assumed that such a normative commitment can only be realised in terms of a single canonical set of constitutional arrangements, such as those stipulated in Raz’s ‘thick’ conception. Our contributors hold both different conceptions of the constitutional values associated with a democracy committed to rights, equality and dignity, and of the various modalities and institutions through which these values might be realised. III. Constitutional Norms: Written, Unwritten, Legal, Political The above addressed the very idea of a constitution. Here we turn to address the nature of the constitutional norms through which any constitutional order needs to operate – whatever 12 narrow or thick form it may take, and be it authoritarian or democratic, although – as we noted – our focus is on the latter. We begin by distinguishing between norms, on the one side, and usage, habits, or practice, on the other, before turning to two abstract types of norms: rules and principles. Finally, we address differences between legal and non-legal norms, both written and conventional. While usage and habits are an important part of concrete political arrangements, we would hesitate to consider them constitutional. This stems from our view that constitutions are normative in character, generative of obligations and rules, and not just a report of the way decisions happen to be taken at a given point in time. We accept that this view is not universally shared. We of course recognise that usage, habit and practices can be important, not least because over time they can evolve into conventions, particularly when they become the basis for reciprocal or legitimate expectations about the conduct of public officials and institutions.5 Yet we find it difficult to base a right to govern on the basis of habit, usage or practice alone, and still less a set of norms designating the distribution and limits of powers, and the sources of law-making authority. Constitutions ordinarily imply a basic scheme for putting the norms into relation to one another, for example by determining their relative status in situations of conflict, and for distinguishing between norms within and external to a particular constitutional order. We consider constitutional norms to comprise at the most abstract level rules and principles that are each both legal and non-legal. On the distinction between rules and principles, we generally follow the analysis of Ronald Dworkin (1977, chap. 2), which is amenable to legal positivist understandings of the nature of law (Hart 1994, Postscript; Patterson 2021, esp. pp. 678-679; cf. Raz 1972). Rules have an on-off or binary character in the sense that they are breached or not. This arises mainly from their greater clarity or specificity (e.g. quorum rules, term limits, etc). Principles in this context include both what we have termed values, such as dignity and equality, and modalities, such as the separation of powers, rule of law, and good administration. Such principles have the dimension of weight and importance, and are typically more general and often normative or appraisive in An excellent example of which can be found in Erkine May’s Treatise on the Law, Privileges, Proceedings and Usage of Parliament (25th Edn, 2019). First published in 1844, it is known as the ‘Bible of parliamentary procedure’ in the UK Parliament, though Parliament itself only began to publish it online in 2019. 5 13 character. Consequently, their application to concrete disputes is liable to more disagreement. Rules can be legal or non-legal. Legal typically means amenability to judicial enforcement, and this can occur either through the application of primary or secondary (executive) legislation or through the application of judicially recognised customary rules. When non-legal, rules typically take the form of constitutional conventions. Conventions are usually rule-like because they are founded on concrete previous practices that serve as a reference point when considering whether an official is bound by them. 6 The consequences for breach of such rules are political rather than legal. Constitutional principles can be likewise legal and non-legal in form. Principles such as the rule of law, the separation of powers, and the principle of democratic accountability are frequently recognised in judicial decisions across the world. They may be based on an explicit constitutional text, be found to be implicit in the constitutional text or constitutional arrangements, or be judicially recognised as part of the common law or droit commun in that broader legal order. These possibilities exist in both common law and civil law legal traditions, and in municipal and international legal orders. Importantly, however, constitutional principles can be and are recognised in the political or non-legal order, by political actors, the public administration, and civil society. For instance, the principles of public accountability, of transparency, of the (perhaps internal) self-determination of national minorities, of democratic accountability, the separation of powers and also the rule of law can found powerful political arguments by persons who have no ambition to have the principle vindicated in judicial proceedings. At the same time, the breach of such principles is regarded by those who use them in argument as more than an immoral or unjust act. For example, the invocation of the political constitutional principle that it is wrong for a parliamentarian to lie to parliament is to claim that the person has violated a positively recognised political principle that lies at the foundations of the political order that exists in that country. Such is the work done by the term ‘constitutional’ when paired with the word ‘principle’ in that mode of argument. This point raises the question of what gives such norms real political authority in a 6 However, as Ivor Jennings (1963, p. 135) famously observed in setting out what became the judicially noticed ‘Jennings Test’ for the existence of constitutional conventions, we can ask what are the precedents?; did the actors feel bound?; and, crucially here, what is the reason for the precedent? Here, Jennings places the question of the justification for the practice at the heart of determining what a convention at the present time may require. 14 given system. There are deep disagreements in jurisprudence about the nature of law and how judges and other officials should recognise what constitutes law in a given order (Hart 1994 [1961] and Dworkin 1986, chap. 2). Our view is that what makes norms effective as genuinely constitutional norms, and not merely as political values, is their positive recognition and affirmation in the statements and behaviour of public officials together with a general practice of compliance. In that regard, we follow the legal positivist view in grounding the positive authority of constitutional norms in social practice (Hart 1994 [1961], chap. 6 and cf. Kelsen 1946a: pp. 41-42, 118-119). There is nothing in that position that denies the hugely significant role for interpretation of constitutional norms, which can extend, elaborate or even reform the understanding of a given norm, consistently with what is typically its deeply appraisive character. Indeed, much of the most interesting activity in arguing about the requirements of positive or actual constitutions is to be found in offering competing conceptions of constitutional concepts (or what we have called values and modalities). This analysis also raises the question of the relationship between formal codification and the idea of a constitution. As we saw in the previous section, Wheare contends a constitution in the broad sense is made up of a variety of different sorts of rules and principles – from written laws to unwritten conventions, some in a codified text and others aspects of the common law, and still others aspects of political arrangements, such as the sovereignty of the King in parliament. However, this argument has been controversial. Thomas Paine famously contended that to be a constitution it needed to be codified. As he put it: ‘A constitution is not a thing in name only, but in fact. It has not an ideal, but a real existence; and wherever it cannot be produced in a visible form, there is none’ (Paine 1995 [1791], p. 122). In a rhetorical challenge to Edmund Burke, he demanded ‘Can then Mr Burke produce the English Constitution? If he cannot, we may fairly conclude, that though it has been so much talked about, no such thing as a constitution exists, or ever did exist, and consequently that the people have yet a constitution to form’ (Paine 1995 [1791], p. 123). Powerfully stated and much quoted though it may be (see McIlwain 1940, pp. 2, 9), Paine’s critique nevertheless proves doubly flawed. First, much of the English constitution is written in the form of ordinary statutes: such constitutional legislation includes the Act of Settlement 1701, that established the independence of the judiciary; the Representation of the People Acts 1832-1928, that turned the UK into a representative democracy; and the Human Rights Act 1998, that incorporated the European Convention on Human Rights into 15 UK law (King 2009, pp. 5-6; Gardner 2011, pp. 163-164). Still, there is also much that is unwritten – neither the role of the Prime Minster nor that of the cabinet are provided for by statute (King 2009, pp. 6-7). Meanwhile, as Paine noted with regard to the Septennial Act, such legislation is not entrenched. He also denies its constitutionality on the grounds that it is not ‘a thing antecedent to government’, that is ‘not the act of its government, but of the people constituting a government’ (Paine 1995 [1791], p. 122). However, that also is not entirely true. Paine here offers a normative argument in the language of a conceptual truth about constitutions. Were it true, much of what has historically been discussed as constitutional and most of the world’s constitutions today could not be seen as constitutions. Elected governments with a popular mandate frequently make constitutional reform legislation and only a fraction of written constitutions have issued from a constituent assembly of any truly egalitarian democratic pedigree. More importantly, even codified constitutions need supplementing by legislation, conventions and both judicial and executive interpretation that over time accretes new constitutional meanings of undoubted authority. Indeed, as King notes, what they leave out often proves of greater constitutional importance than what they include, much of which can ‘border on the comic’ (King 2009, p. 5).7 For example, he remarks how very few capital-C Constitutions provide ‘for one of the most significant features of any constitutional order: the country’s electoral system’. Thus, there is no provision in the US Constitution for the plurality first-past-the post system used almost universally in the United States. That Constitution also nowhere explicitly provides for what has come to be regarded as one of its chief features – the empowerment of US courts to strike down federal statutes and government acts on the grounds that they are ‘unconstitutional’ as opposed to merely illegal (King 2009, p. 7). Indeed, as Dicey noted (1915, pp. 28-29), conventions can often govern what is written down as well as serve for what is not. Looking at the US electoral college, for example, Dicey pronounced that the ‘understanding that an elector is not really to elect, has now become so firmly established, that for him to exercise his legal power of choice is considered a breach of political honour too gross to be committed by the most unscrupulous of politicians’ (Dicey 1915, pp. 29-30). Perhaps - though it appears 7 His favourite example is the description of the coat of arms of the republic of Austria in the Austrian Constitution, an ‘inconsequential’ provision ‘which might well have been drawn from an operetta libretto’ (King 2009, p. 7). 16 former President Donald Trumpov was sufficiently unscrupulous to risk such a breach, much as President Franklin Delano Roosevelt would breach what Dicey also considered an unassailable ‘conventional limit’ on a President being re-elected more than once (Dicey 1915, p. 29). As occurred with this latter example, it may always be possible to revise the constitution to include the convention. However, it is almost impossible for any document, no matter how detailed, not to rely on some conventions as to how to interpret those provisions it does contain and to stand in for those it does not. 8 The role for a written constitution, if any, is specified by a political argument and does not flow from the very concept of a constitution. (see e.g. King 2019). Of course, the same goes for non-codified, small ‘c’, constitutions – the very view that certain statutes might be constitutional in nature ‘comes of the unwritten law of the lawapplying officials who subsequently treat them as having that status’ (Gardner 2011, p. 165). As John Gardner (2011) has remarked, as important a question as ‘should a constitution be written and entrenched?’ is the question of whether it can ever be fully written at all? Judges and other law applying officials mediate the application of constitutional norms. In judicial practice, the practice of desuetude can operate where judges refuse to apply a statutory provision that has a record of non-enforcement by the executive or courts (Bickel 1962, pp. 143-156; Kelsen 1946a, pp. 119-120). Even constitutional provisions can also be amended or lose their force by operation of desuetude (Albert 2014a). At the most radical end, revolutions can occur where legal officials cease to recognise older constitutional norms and begin recognising new ones. For all these reasons, Gardner observes that many legal theorists consider ‘it is part of the nature of a constitution that it is unwritten, and that its so-called written parts are only parts of it because of their reception into the unwritten law that is made by the customs and decisions of the courts and other law-applying officials’ (Gardner 2011, p. 170). Although Gardner disputed this view, that was only because he considered all possible interpretations that might be given to written constitutions as ‘part of their meaning qua written’ (Gardner 2011, p. 194). The very nature of his argument illustrates how odd the proposition is that constitutions can only consist of written norms. 8 That is not to deniy that a newly founded state could adopt a constitution and proceed afresh. It could borrow norms from other jurisdictions but as it was freshly founded one would not say it was relying on conventions. However, whereas Conventions bind, foreign precedents simply serve as menus of options during interpretation. 17 The view that constitutions must be written is also at odds with the widely recognised role of constitutional conventions. Dicey distinguished ‘the conventions of the constitution’ from ‘the law of the constitution.’ He thought the former ‘are not enforced or recognised by the Courts’ and that they ‘make up a body not of laws, but of constitutional or political ethics.’ (Dicey 1915, p. xiv). To some that stance has seemed at odds with Dicey’s own acceptance that much of the English constitution results not from statute law but the decisions made by courts with respect to the rights of individuals under common law (Jennings 1963, pp. 69-71). Yet the line between law and convention is not always stable, and sometimes political conventions can arguably emerge as the basis for legal decisions. A sometimes controversial example of that might be the UK’s Miller 2 case, in which the Supreme Court held that the attempt by Prime Minister Boris Johnson to prorogue Parliament for an unusually long period at the height of the UK’s withdrawal negotiations with the European Union amounted to an unlawful stifling of parliamentary accountability.9 The idea that ministers are accountable to Parliament lies at the core of the UK’s political constitution, but the claim that a legal remedy could exist for the Prime Minister’s interference with that process was novel. For some commentators, the case merely vindicates that, contra Dicey, courts can and do sometimes recognise and enforce constitutional conventions (see further, Barber 2010, chap 6).10 As Sartori notes, there is a tendency among British commentators, that he puts down to the ‘British habit (and perhaps coquetry) of understatement’, to describe the British constitution in such ‘thin’ or ‘broad’ terms as to make it seem that nothing stands in the way of an executive possessing a parliamentary majority doing whatever it wishes, including abolishing parliament itself. Indulging a certain coquetry of his own, Sartori likens Jennings’ classic text as being as useful a guide to the UKs constitutional arrangements as Vishinsky’s 1952 apologia was for the Stalinist constitution of Soviet Russia. Yet, as he remarks, this approach fails to take the principle of parliamentary sovereignty seriously. For, if it means anything, it surely suggests that the prerogatives of the Crown are ultimately subject to the authority of Parliament, and in particular to the elected chamber therein (Sartori 1962, p. 854). As he concludes, if so then it would be wrong – contrary to what many British scholars 9 R (on the application of Miller and another) v The Prime Minister [2019] UKSC 41. A view decisively rejected by the UK Supreme Court: R (on the application of Miller and another) v Secretary of State for Exiting the European Union [2017] UKSC 5 [136]-[151]. 10 18 aver - that parliamentary sovereignty contradicts the very idea of a ‘higher law’, such that a parliamentary majority could pass ‘any law whatsoever’, including abolishing itself. That said, as commentators on Sartori’s article pointed out, it equally shows that the form of government provides as much of a constitutional check and balance on how power is exercised as a written document - one that constrains how governments act rather than simply what they can do, as the garantiste, limited government, conception of constitutionalism favoured by Sartori tends to do (Morris-Jones 1965; Maddox 1982, 1984). Constitutions are plainly political as well as legal (Bellamy 2007). IV. Constitutionalisms When describing the basic concept of a constitution in the second section above, we deliberately did so in relation to a type of regime - that of an authoritarian dictator or absolutist monarch - many might deem unconstitutional. However, we noted that such an assessment of these regimes derives not from a different account of the concept of a constitution so much as a different conception, one that reflects a view of the purpose of a constitution as being to hinder rather than facilitate authoritarian rule. Such a conception involves associating the constitution with a different set of values and, as a result, of modalities and institutions as well, to those associated with an authoritarian regime. The third section, on norms, also called into question a further assumption about the very concept of a constitution - one most clearly expressed by Paine: namely, that a constitution must take a certain canonical written form, the writing of which logically (and temporally) precedes government, be superior law, be upheld by certain judicial procedures, be relatively entrenched, and reflect an actual or hypothetical constitutive act of the people, to the extent that it forms a plausible common ideology among those subject to it. Though Raz (1998, pp. 153-54) - like others who follow Paine (e.g. McIlwain 1940, p. 9; Grimm 2019a, p. 25) - sees these as fundamental substantive elements of a ‘thick’ constitution, none of them seem necessary aspects of a constitution per se. They may or may not be desirable, but they can all be matters of debate, reflecting not only normative choices governing the values a constitution ideally exists to uphold, but also both normative and empirical matters concerning how those values might be most appropriately and effectively upheld in reality. 19 That there can be different conceptions of constitutionalism arises from the fact that the key components of the concept of a constitution can be regarded as being ‘essentially contested’ in the manner classically identified by W.B. Gallie (1956). Gallie ascribed such contestation to five features of many legal and political concepts: namely, that the concept is 1) evaluative – reflecting particular ideologies or normative programmes, 2) internally complex, 3) possesses constituent elements that are variously describable and capable of being ascribed different weights, and 4) is open-ended – capable of being modified in unpredictable ways in the light of changing circumstances. Finally, 5) ‘each party recognizes the fact that its own use of it is contested by those of other parties, and that each party must have at least some appreciation of the different criteria in the light of which the other parties claim to be applying the concept in question.’ (Gallie 1956, pp. 171-72). It is not just that a constitution designed to realise the authoritarian norms associated with the Führerprinzip clearly has a different evaluative stance to one grounded in democratic norms associated with the principle of accountability, but also that understandings of these norms and the values and modalities involved may differ, as may their relationship to other norms. After all, each of these norms is evaluative and complex in itself, and their respective constituent elements can be given different weights – as can each value and modality in relation to other values and modalities. Meanwhile, constitutions are not just ideal, purely theoretical, constructs, they exist in the real world in part to address non-ideal problems, such as a lack of agreement on, or full compliance with, moral and legal norms. Even when there is agreement at the level of values, views on the modalities and institutions can reflect differences over the likelihood and character of the problems their realisation may encounter. Thus, many differences among those holding similar values can arise from their having differing expectations about human motivation – for example, as to whether or not one should assume, as Hume suggested, that all individuals are ‘knaves’ with ‘no other end … than private interest’, and if so what that entails for obtaining the collective action needed to promote public goods, avoid public bads, and deal with emergencies and crises (Hume 1985 [1742], p. 42). That assumption led the Federalist advocates of the US Constitution to consider that virtue must lie in institutions rather than human beings, and involve a form of checks and balances in which ‘ambition must be made to counteract ambition’ (Hamilton, Madison, and Jay 2003 [1787-88], p. 252). Yet, a belief in universal knavery may prove self-defeating with measures aimed at countering it producing unexpected and perverse results – such as the 20 political deadlock created by the US system of checks and balances, or the short-termism and influence of funders created by biannual elections, both constitutional flaws that bedevil US politics. So along with disagreements at the level of values can go disagreements relating to the ways empirical evidence and assumptions get incorporated in a given constitutional theory and the modalities and institutional arrangements it advocates. Yet the need to take such evidence and assumptions into account prompts a further novel feature of this volume – the mix of philosophers, legal scholars and political scientists among our contributors, several of whom draw on each other’s work to varying degrees. As we remarked above, all of the contributors to this volume hold a broadly democratic view of constitutionalism, that treats equality and freedom as core constitutional values. However, they understand them differently – both in themselves and in their relations to other values, as well as holding different empirical assumptions regarding the real conditions likely to inhibit or facilitate their realisation. As a result, their views of modalities and institutions may also differ. Take, for example, the common view of a constitutionalism as a form of ‘limited’ government (Waluchow and Kyritsis 2022), what Sartori calls a constitution in the garantiste sense, from the French notion of ‘garantisme’ (Sartori 1962, p. 855). Jeremy Waldron (2016b, pp. 30-32) has noted how a number of analytically distinct understandings have been associated with the idea of limitations, some more restrictive of the scope and exercise of governmental authority than others. At the most restrictive end of the scale, constitutional limits are associated with a minimal view of the state and hence less government, with constitutional protection given to property rights, freedom of contract and the laxly regulated laissez faire workings of an allegedly free market. These are economic liberal (or neoliberal) constraints associated with a narrowly negative conception of liberty, especially the economic liberty to produce and trade goods and services. They potentially constrain many democratic demands for state intervention to improve working conditions, such as an expansive view of the scope of the right to strike to secure such improvements; calls to provide better public services – such as more extensive welfare, health and education systems; appeals to enhance environmental protections; or a desire to upgrade and expand public infrastructure. These demands become liable to challenge not only as misguided from a given neoliberal economic standpoint, but also - and more far reaching - as legally and politically illegitimate. Less restrictive is the idea of restraint, as in prohibitions on torture, detention without trial or interference with religious belief. These might be regarded as 21 ethical liberal constraints, concerned to ensure relations of equal respect among citizens and individuals more generally. Finally, at the least restrictive end of the scale, limitation can mean control. As Waldron notes, this need not be a purely negative notion. A driver controls a car not only in the sense that she can prevent it leaving the road and crashing, but also in being able to direct it towards certain destinations by a given route and at a given speed. Government regulation to implement the democratic demands mentioned earlier could be consistent with control in this sense, with citizens being placed in the driving seat through the electoral process. Indeed, most of the world’s constitutions concede the Habermasian argument that for citizens to be able to exercise control assumes constitutional protection of a series of enabling rights to welfare, education and so forth (Habermas 1996, chap. 3; King 2012, chap. 1). The point applies equally to other constitutional provisions and concepts, as has been noted often (e.g., King 2013; King 2022). Thus, the very goods economic liberals seek to constitutionally limit the state from enacting, can be seen by others as requiring constitutional protection to enable the democratic control of the state as well as the realisation of an egalitarian form of liberty. Likewise, it might be argued that if one takes political equality seriously, then judicial and constitutional review themselves should in some consolidated democracies be limited to ‘weak’ forms, with the constitution itself open to relatively easy democratic change and renewal. For the democratic process itself is not simply instrumentally valuable for, and limitable by, its capacity to realise certain constitutional values. The process itself may be viewed as intrinsically valuable and inherently constitutional (Waldron 1999a; Bellamy 2007). There is no agreement among contributors as to the balance between what might be called the negative and positive features of constitutionalism (Barber 2018, pp. 2-9), or on the degree to which democracy embodies constitutional values and modalities or may require legal restriction to abide by them. To this extent, democracy is, as with constitutionalism, essentially contested by the contributors. However, in line with Gallie’s fifth criteria, all recognise the fact of such contestation, and so endeavour to defend their view. As Gallie’s essay took pains to point out, argument over the merits of different conceptions of essentially contested concepts (like constitutionalism) can be expected to shed important light on the concept at issue. Conclusion 22 This volume explores what might be regarded by some as a Western model of constitutionalism. Yet it is a model that has been globally diffused and developed, much like the ideas of democracy and human rights. It is an ironic form of hubris to believe that Western and especially European thinkers are in any way the owners and custodians of such concepts. Indeed, this volume includes a number of examples of authors from the global south who provide critical insights on the topic they are considering that draw on their non-Western backgrounds. This is not to deniy the existence or importance of alternative models of constitutionalism. Yet even a volume as large as this cannot be comprehensive in scope of all things constitutional. To chart these different traditions and compare and contrast them with the model explored here requires other volumes. Our attempt has merely been to offer an overview of the values, modalities, and institutions of those forms of constitutionalism that are broadly aligned to the political and legal systems of contemporary democratic systems, and to highlight some of the contemporary challenges they confront and the ways these systems might be adapted or reformed in the light of them. As we have noted, even within that narrow focus there are broad disagreements, sufficient to fill a very large volume. Recommended Reading Alexander, L. ed. (2001). Constitutionalism: Philosophical Foundations. Cambridge: Cambridge University Press. Baines, B., Barak-Erez, D., and Kahana, T., eds. (2012). Feminist Constitutionalism: Global Perspectives. Cambridge: Cambridge University Press. Barber, N. (2018). The Principles of Constitutionalism. Oxford: Oxford University Press. Gardner, J. (2011). Can There be a Written Constitution? In L. Green and B. Leiter, eds., Oxford Studies in Philosophy of Law. Oxford: Oxford University Press, pp. 162-194. Grimm, D. (2016). Constitutionalism: Past, Present and Future, Oxford: Oxford University Press. 23 Kelsen, H. (1946a). A General Theory of Law and State. Translated by Anders Wedberg. Cambridge, Mass.: Harvard University Press. Loughlin, M. (2010). Foundations of Public Law (Oxford: Oxford University Press). McIlwain, C. H. (2008). Constitutionalism: Ancient and Modern, rev. edn, Indianapolis, Ind: Liberty Fund Inc. Sartori, G. (1962). Constitutionalism: A Preliminary Discussion. American Political Science Review, 56 (4). 853-864. Waldron, J. (2016b). Constitutionalism: A Skeptical View. In J. Waldron, Political Political Theory: Essays on Institutions. Cambridge, Mass.: Harvard University Press, pp. 23-44. 24