The `thin' definition of a constitution, as the body of law which constitutes and differentiates the main organs of government and their powers, and which specifies the main rights and obligations connecting the citizenry to these organs of government tell us little more than that, in this minimal sense, the European Union, although clearly lacking a documentary Constitution, already has `constitutional law'.28 To gain a closer understanding of the significance of constitutionalism and constitutional law in the European Union, and of its relationship to the various dimensions of legitimacy discussed above, we must look instead at the variety of different (but overlapping) social and political functions that may be performed by constitutional law and by constitutional discourse more generally. Constitutional law and discourse may be said to perform community-generative, substantive, technical and polity-affirming functions, and we will briefly discuss each of these in turn.
Constitutional law and discourse can be community-generative to the extent that the establishment, maintenance and development of a constitutional framework may be deemed to be embody a series of acts or even a continuous process29 of self-legislation and of democratic will formation on the part of the members of the polity. Alongside and, at least from some perspectives,30 shading into this procedural dimension, constitutional law and discourse performs substantive functions to the extent that it privileges certain norms of representation, of institutional capacity and design, and of individual and collective entitlements and obligations vis-à-vis the polity as the guiding or dominant modes of governance in the day-to-day operation of the polity. Next, constitutional law and discourse perform a technical function in so far as they provide a clear and co-ordinated statement of these governing norms in a particular document or other legal source, or in an ordered series of documents and other legal sources.
Finally, constitutional law and discourse perform a polity-affirming function, which, in the case of a non-state polity such as the EU, might take one of at least three different forms. In the first place, the very fact of its engaging in a self-conscious constitutional discourse might be seen to have some symbolic value in affirming the status of the European Union as a separate polity, if only because of the strong tradition of state polities possessing their own authoritative and autonomous constitutional law and discourse.31 Certainly, some of the support for the idea of a European Constitution, and perhaps even more of the opposition to such an idea, has turned implicitly or explicitly on the recognition (or fear) of just such a symbolic link. Yet, as the gradual acknowledgement amongst many Eurosceptics of the possible value of a European Constitution and Charter of Rights as a means of limiting European governmental power indicates,32 this point should not be pressed too far. A constitutional discourse may be seen as a necessary accompaniment and constituent condition of statehood, and by extension perhaps of `polity-ness' more generally, but it is clearly not a sufficient condition. In the absence of other significant preconditions, the mere assertion of a constitutional identity cannot do the symbolic work of transforming general perceptions about the existence or degree of the polity status of an entity in an affirmative direction. Secondly, given our earlier proposition about the intimate link between performative legitimacy and broader polity legitimacy in a post-state polity, constitutional discourse may be polity-affirming to the extent that it seeks to direct, encapsulate or reinforce the broad purpose or purposes of the polity. Thirdly, given our earlier proposition about the intrinsically relational character of the post-state polity, constitutional discourse can be polity-affirming and polity-defining to the extent that it seeks to specify the nature of the relationship between the post-state polity and other extant polities, most significantly, the Member State polities. That is to say, it may be polity-affirming to the extent that key questions concerning the articulation of the boundaries of the polity and its mode of relating to other polities are addressed and resolved, including questions of supremacy, mode of legislative applicability (direct or indirect), institutional linkage and competence delimitation.
We are now in a position to trace some of the connections between constitutional function and legitimacy. The community-generative dimension of constitutionalism is crucial here in two respects. It is important not only as an independent factor in fostering the identity element of polity legitimacy, but also in promoting the internal legitimacy of the constitutional framework generally, and so in feeding the legitimating roots of the other three constitutional functions - substantive, technical and polity-affirming. That is to say, the legitimate authority of the constitutional norms which supply the substantive framework of governance, the capacity of the body of constitutional norms to present itself as a clear, definitive and coherent framework with a plausible claim to perform an ordering function, and also its capacity to affirm the status, mission and relative position of the polity, all depend. in no small measure, upon the legitimacy of the generative process.
In turn, each of these three additional types of constitutional function can make its own contribution to legitimacy. The substantive norms clearly contribute to regime legitimacy, but can also, on account of the internal relationship between regime legitimacy and the other two types of legitimacy, provide a structural framework which facilitates performance legitimacy on the one hand, and which complements the identity-forming and reinforcing work of the polity-generative rules on the other. The polity affirming function clearly also connects with wider polity legitimacy, most directly with regard to its authority dimension. The technical function is less significant as a direct source of legitimation, but in so far as transparency, publicity, accessibility and comprehensibility are all formal `rule of law' virtues which provide added value to the constitutional framework generally, the technical function has the capacity to enhance, or, in its absence or deficiency, to undermine, the effective performance of the other three functions and the forms of legitimacy which flow from them.
If this analysis helps us to identify something of the legitimating potential of constitutionalism, it also points towards both the overall or external limitations upon this potential and the internal factors that may frustrate or curb what legitimating potential constitutionalism does possess. As regards the overall picture, there are definite limits, albeit controversial in their extent, to the potential of a legally-coded community-generative process to `pull itself up by its own bootstraps' and generate the degree of trust, solidarity or loyalty associated with community in the absence of other prior or simultaneous processes and features of common identity-formation. Furthermore, on account of the important foundational role of the community-generative aspect of constitutionalism, any limits in the foundational framework will lead to limits `all the way up', so to speak; that is to say, in the legitimating potential of the substantive, technical and polity-affirming functions of constitutionalism.
Each of these three functions also has its own intrinsic limits which reinforce the derivative limits. As regard the substantive governance-design function, while, in principle, this bears upon each of the performance, regime and identity-constitutive elements of wider polity legitimacy, it also underdetermines each of these. It underdetermines performance legitimacy in the sense that it may provide a facilitative framework for the generation and effective implementation of widely approved and beneficial policy norms, but it cannot guarantee these. It underdetermines regime legitimacy in the sense that it can lay a blueprint for an institutional complex which instantiates key governance values, but it cannot guarantee that, in the actual operation and inter-relationship of these institutions, these values will be effectively articulated or optimally balanced. Finally, it underdetermines that dimension of wider polity legitimacy with which it engages, namely, the identity dimension, for the same `boot-strapping' reasons as the identity dimension is underdetermined by the polity-generative rules.
The overall limitations of the technical function, which, as we have seen, is an important lubricant in the operation of the other functions, have to do with the incapacity of formal `rule of law' values to generate the conditions for their own effective implementation. In other words, there is more to predictability, accessibility, comprehensibility, etc., of the constitutional order than a thin commitment to, and exemplification of, these within the key texts of that constitutional order. Rather, there must already exist or be generated by other means a cultural propensity amongst all relevant constituencies, including legal professionals, key interest groups and institutional actors, as well as the general publics of Europe, to treat the constitutional framework in such a way as to promote, enhance and take advantage of these values.
Finally, the intrinsic limitations of the polity-affirming function in certifying the authoritative dimension of polity legitimacy have to do with the general limitations of law as a declaratory or exhortatory device. Each of the three modes of polity affirmation discussed above turns, in one way or another, on law's symbolic function.33 As noted earlier, the significance of the very existence of a constitutional framework for the perceived status of an entity turns on the strength of the symbolic association of constitution and polity. The significance of a broad constitutional statement of the polity's mission turns on the capacity of law to capture or to be an active agent in promoting something as broadly conceived - some would say ineffable34 - as an overall telos. The importance of a constitutional demarcation of the polity's relation with other polities turns on the capacity of the internal law of one legal order to define or influence its relationship with other legal orders against a complex and shifting macro-political backdrop and in circumstances of constitutional pluralism35 - that is, where there is no super- or meta-constitutional guarantee that the internal constitutional conceptions of the relation between polities espoused by each of the relevant polities will be in harmony or that any differences will be capable of definitive resolution. In each case, we see law removed from its normal context of authoritative justiciability and, instead, compelled to draw upon a broader and less assured symbolic capacity for persuasive ordering, a capacity which (especially in the absence of the deep-rooted tradition typical of national legal orders) is in significant measure dependent upon the fund of legitimacy accrued from the precarious success of its foundational polity-generative procedures.36
Alongside these overall external limits on the potential of constitutionalism, there are other internal factors which interfere with its capacity to realise its potential. As suggested earlier, the two types of limits are closely connected in as much as some of the very problems that constitutionalism `cannot reach' tend to interfere with the dynamics of constitution-building. In the context of the European Union, we can identify five interconnected problems associated with the internal dynamics of constitution-building.
First,. there is the problem of constitutional denial that is peculiar to the category of putative post-state polities to which the EU belongs.37 As we have seen, many sceptical voices which question the identity and authority claims of the EU in turn question the legitimacy of dignifying the EU with constitutional status, and this threshold objection continues to blight the development of a shared explicit constitutional discourse about the EU.
Secondly, there is the problem of self reference,38 which affects the generative rules of all constitutions, but which is particularly acute in the context of a polity whose generative rules are statist in origin and which, despite the gradual growth of constitutional awareness within the EU in its case law, institutional self-understandings and public attitudes,39 continues to lack a self-conscious self-legislative basis, or indeed, as the problem of constitutional denial indicates, even an agreement that, in principle, such an aim is legitimate. The problem of self-reference concerns the doubtful legitimacy and impartiality of responding to perceived inadequacies in the constitutional framework by seeking to reform the constitutional framework - including the procedural rules for generating that constitutional framework - by means of utilising the existing rules for generation and reform, where these existing rules may be viewed as being responsible for or associated with the perceived inadequacy in question. To exacerbate matters, the phenomenon of constitutional denial and the doubtful continuing legitimacy of an original process in highly altered circumstances which make the problem of self-reference so acute also make it difficult to reach agreement on a process of constitutional renewal or initiation other than by means of self-reference.
Thirdly, there is a more general problem of ideological disagreement over the substantive content of any European constitutional settlement. The ever-shifting institutional balance between the three main organs of the Commission, the Council and the Parliament; the gradual accretion of institutional novelties such as the European Council and the Comitology network; the incremental assumption of exclusive or shared substantive competences in new policy areas; the complexity of discovering satisfactory institutional solutions to predicaments which arise precisely because of fundamental disagreement over the nature and scope of the classical Community method, as in the Three Pillar structure and the flexibility provisions; the uneven history and tentative progress of the EU's fundamental rights commitment, most recently reaffirmed by the merely declaratory status of the new Charter of Fundamental Rights; each of these bears witness to the unstable and provisional nature of any agreement on substantive constitutional norms in a relatively new political entity in which the nature of political identity and the limits of political authority are fragile and contested. In turn, this problem is exacerbated by two other internal constraints on constitutionalism in the EU considered below.
Fourthly, then, there is the absence of a default constitutional template derived from the state model as a reference point for addressing disagreements associated with the governance structure of the new post-state polity. These "problems of translation"40 are manifold; for example, whereas most states "work on the basis of some division between executive, legislature, administration and bureaucracy"41 in which there is a reasonable correspondence between function and institution, in the European Union there is a much more complex, and ever shifting pattern of functional roles within and across institutions and no tried and tested conception of institutional design to guide reform. Fifth and finally, there is the problem of explicit constitutional engagement. One of the ironies of the continuing and inconclusive struggle to develop an explicit constitutional discourse for the European Union, whether in the traditional sense of a documentary Constitution or in the broader sense of seeking to attribute constitutional status to the continuing debate over its rules of governance, is that the very effort to strengthen and legitimise the governing base can expose difficulties which had been fudged, compromises that had been made, and "abeyances"42 which had been left under the less harsh spotlight of the long, incrementally adjusting, `pre-constitutional' stage of development of European government. Once the game becomes an explicitly constitutional one, the stakes may be raised and all previous bets cancelled, as, for example, in the case of the key polity-affirming function of specifying the relationship between the EU and its Member States, where deep constitutional exploration of concepts such as supremacy, direct applicability and effect and residual competence would inevitably expose a minefield of contestable interpretative possibilities.43
28 P. Craig, "Constitutions, Constitutionalism and the European Union," (2001) 7 European Law Journal 125, 126.
29 See, for example, R. Bellamy and D. Castiglione, "The Communitarian Ghost in the Cosmopolitan Machine: Constitutionalism, Democracy and the Reconfiguration of Politics in the New Europe," in R. Bellamy (ed) Constitutionalism, Democracy and Sovereignty: American and European perspectives (Aldershot: Dartmouth, 1996).
30 Ibid. From certain normative standpoints, the procedural, community-generative dimension must be in a continuous process of renewal and (re)vindication, in the sense that the constitution-making procedures themselves should not be firmly entrenched but subject to revision, or that the substantive results of constitution-making should not be firmly entrenched but subject to revision, or that the ambit of `higher' constitutional provisions should be narrowly circumscribed and there should be significant scope for the generation of governance norms using other flexible self-legislative procedures - or any combination of the above. Indeed, to the extent that a more flexible self-legislative element is endorsed, this will argue - whether categorically or as a matter of degree - against a traditional conception of constitutionalism in which there is a strict demarcation between a documentary foundation of higher law and other substantive norms dealing with governance aspects of the polity.
31 Walker, n.4 above, 278-79.
32 See, for example,. The Economist, 4 November 2000.
33 See, for example, R. Cotterrell, "Some Aspects of the Communication of Constitutional Authority", in D. Nelken (ed) Law as Communication, (Aldershot: Dartmouth, 1997) 129.
34 See, for example, I. Ward, "Beyond Constitutionalism: The Search for a European Political imagination" (2001) 7 European Law Journal 24.
35 See, for example, Walker n.12 above: N. MacCormick, Questioning Sovereignty: Law, State and Practical Reason, (Oxford: OUP, 1999) esp. ch.7.
36 See N. Walker, "Flexibility within a Meta-constitutional Frame: Reflections on the Future of Legal Authority in Europe" in G. de Búrca and J. Scott (eds) Constitutional Change in the EU: From Uniformity to Flexibility? (Oxford: Hart, 2000) 9. 29-30.
37 Walker, n.4 above, 278-279.
38 See, for example, P. Eleftheriadis, "The European Constitution and Cosmopolitan Ideals," (2001) 7 Columbia Journal of European Law, 21, 22-24.
39 See, for example, Craig, n.28 above, 128-30.
40 Weiler n.23 above, 270.
41 Craig, n.28 above, 139.
42 M. Foley, The Silence of Constitutions: Gaps, "Abeyances" and Political Temperament in the Maintenance of Government (London: Routledge, 1989).
43 Craig, n.28 above, 143-45.