In the contemporary taxonomy of modern International Relations (IR) theory, inter-state anarchy and a hierarchical ordering of political authority are considered the hallmarks of international and domestic politics, respectively (Waltz Reference Waltz1979). Within this framework of two opposed structural principles of political organization no middle ground is typically considered possible, although there is a recent trend towards highlighting the degree of hierarchy present in certain inter-state relations (Lake Reference Lake2007; Reference Lake2009). The question of the possibility of cooperation under anarchy thus remains a fundamental point of departure for IR theory (Lake Reference Lake2007, 47). Yet there are examples of political organization that call into question this supposed antinomy between hierarchy and anarchy. At least such is the case in two instances of ‘states union’ (Forsyth Reference Forsyth1981): the European Union (EU) today and the United States before the Civil War. In these, admittedly rare, systems, sovereign states opt to form a union of both individual states and citizens collectively (Fabbrini Reference Fabbrini2007). The novelty of such an arrangement is that – unlike formal and informal empires – it is deliberately anti-hierarchical: the objective is not to replicate a monopoly of legitimate coercion at a level of government beyond that of the individual territorial units (Deudney Reference Deudney1995).
Traditionally, the concept of state sovereignty has served to police the boundary between anarchy and hierarchy in international politics even though states have consistently violated the tenets of sovereignty itself (Krasner Reference Krasner1999). But in the EU and the antebellum US, the exact sovereign status of the units remains unclear because there is no undisputed and unequivocal hierarchy of political authority established at the level of the union, whilst there is also no condition of simple anarchy between the units as they have agreed to share decision-making and common rules. Consequently, there is no replication of the standard domestic hierarchy characterized by a pyramid of internal sovereignty atop of which stands an institution or actor with ultimate decision-making authority over a territory and with the ability to deploy legitimate means of coercion. This incongruity has been understood by scholars focusing on both the early American republic (Deudney Reference Deudney1995; Hendrickson Reference Hendrickson2006; Reference Hendrickson2009) and the contemporary EU (Keohane Reference Keohane2002; Collard-Wexler Reference Collard-Wexler2006) as a challenge to IR theory by creating a situation of overcoming anarchy without substituting hierarchy.Footnote 1 Less noted though is a second fundamental ambiguity, one which this paper argues sustains the very ability to straddle the middle ground between anarchy and hierarchy in a democratic fashion. The US and EU states unions share indeterminacy over who exactly can exercise popular sovereignty, that is, is the people the aggregation of individuals across the union, those of the separate territorial units or a combination of the latter acting at the same time. The presence of this dual foundational ambiguity – over both the nature of sovereignty and the ‘constitution of the people’ (Näsström Reference Näsström2007) – allows for a comparison between both states unions that can shed further light on the constitutive relationship between sovereignty and democracy, an enduring subject of theoretical and empirical debate within international relations (Archibugi Reference Archibugi2008).
Scholars of international relations as well as normative theorists point out the notable lack of democratic participation in creating the thickening set of rules that bind states today (Moravcsik Reference Moravcsik2002; Archibugi Reference Archibugi2008; Keohane et al. Reference Keohane, Macedo and Moravcsik2009). The consequent received wisdom is that whilst democracy is a form of government that must somehow be congruent with sovereignty (Moravcsik Reference Moravcsik2002), multilateral institutions can nevertheless enhance certain features of the democratic process even when constraining participation (Keohane et al. Reference Keohane, Macedo and Moravcsik2009), although mechanisms for articulating popular sovereignty directly across borders might still be preferable for solving international policy problems (Archibugi Reference Archibugi2008). Such accounts reveal the dominance of theorizing democracy as a process that necessarily mobilizes citizens at the level of a certain bounded community. This identification of a level where democracy can be exercised matches the presumed antinomy between hierarchy and anarchy by assuming that democracy is possible only under a hierarchical arrangement that sets the boundaries of a political community.
However, the nature of states unions challenges precisely this assumed isomorphism between sovereignty and democracy by virtue of what will be shown to be the mutually sustaining ambiguity over both sovereignty and the constitution of the people. This paper connects these two constitutive elements of a states union to demonstrate how they render possible a blurring of sovereignty that confounds IR's traditional antinomy. To do so the analysis problematizes the ability to sustain this kind of constitutional ambiguity as an issue of how democracy and the people or peoples that form a democratically organized community are understood to be structured.
Existing treatments of the consequences of the EU's unusual sovereignty arrangements mostly debate the validity of the democratic deficit critique. This means debating whether there is a need to connect EU authority to popular sovereignty, either through direct democracy or representation (Bellamy Reference Bellamy2010). Yet focusing on the level of government where popular majorities can express policy preferences overlooks how the ambiguous nature of sovereignty within the EU is related to uncertainty over the form democracy can take where there is no settled definition of the people. Although some scholars have identified the significance of having multiple demoi within the EU constitutional architecture (Nicolaïdis Reference Nicolaïdis2004; Fabbrini Reference Fabbrini2007; Eriksen and Fossum Reference Eriksen and Fossum2011), there is little systematic work exploring the inter-relationship between democracy and sovereignty and how this serves to blur the boundary of anarchy and hierarchy. This lack of attention can be explained by the tendency to treat the EU's two foundational ambiguities separately. IR scholars with different perspectives on the exact meaning and importance of sovereignty nevertheless agree that the EU's overlapping sovereignty claims between member states and Brussels, have prevented re-creating national sovereignty writ large (Keohane Reference Keohane2002; Slaughter Reference Slaughter2004). Ambiguity over who the people to be represented actually are and how this will be done has given rise to a separate yet extensive literature on a ‘democratic deficit’. Scholars in this latter tradition argue over whether existing indirect forms of legitimacy via national democratic practices are sufficient (Moravcsik Reference Moravcsik2006), whether the injection of majoritarianism will produce less efficient regulatory policy (Majone Reference Majone2005), or whether democracy requires an existing historical political community or not (Habermas Reference Habermas2000).
Hence these two major literatures on the democratic problem that stems from an ambiguous sovereignty arrangement fail to problematize explicitly the degree to which the two are mutually sustaining – a point made by Bartolini (Reference Bartolini2006) to pour cold water on proposals for majoritarian democracy in the EU. Thus the broader question of why and how democracy as a principle of political organization can explain the nature of sovereignty arrangements between states remains overlooked. To understand this relationship, the paper juxtaposes the case of the antebellum US states union with the contemporary EU (the first section). This allows for an examination of why democracy, understood as a form of legitimacy as well as a principle of political action exercised at the level of a particular bounded community, can be a central element in determining relations between states. In this way, democracy is revealed to be a constitutive feature of international relations (cf. Agné Reference Agné2010) by shaping the contours of sovereignty and holding the key to sustaining ambiguity over the nature of sovereignty in an anti-hierarchical states union.
Contrasting the antebellum US with the contemporary EU is useful not because the earlier experiment in overcoming anarchy without hierarchy ended in bloody conflict but because it demonstrates how the exercise of popular sovereignty in the course of the 19th century fundamentally challenged the absence of hierarchy in a constitutionally ambiguous system. From a Madisonian middle ground that suggested ‘a state acted in its “highest sovereign capacity” only when the sovereign people of a state acted in combination with the sovereign people of other states’ (Fritz Reference Fritz2008, 225), constitutional ambiguity bifurcated into two competing and exclusive visions (the second section). This debate involved settling the question of where popular sovereignty was located in order to decide upon the nature of sovereignty within the union. On these new terms, sovereignty could only be located in one of the two places – the federal or the state level – depending on where popular sovereignty was thought to reside. By seeking to establish what can best be termed a hierarchy of democratic legitimacy, these mutually exclusive constitutional visions made it impossible to sustain ambiguity over the exact sovereign status of the units.
Thus, the paper queries whether this same specter of exclusive constitutional visions threatens the EU and, if not, what makes the latter different (the third section). The analysis presented here ultimately suggests that the EU differs from the US states union thanks to its ability to sustain ambiguity over the constitution of the people, thereby preventing an exclusionary understanding that popular sovereignty must be located at some level where governed and governing are congruent. Sustaining this ambiguity is the product of the continuing multiplicity of overlapping claims by different political communities, via various channels, to hold some aspect of the EU to account. Extrapolating the implications for avoiding a hierarchy of democratic legitimacy in international cooperation (the fourth section) reveals that acceptance of pooled or shared sovereignty corresponds with and indeed is sustained by a plural debate over the constitution or boundaries of the people. Rather than expecting there to be a level at which popular sovereignty can be exercised to settle constitutional issues, blurring the boundaries of sovereignty between states means accepting the contribution of multiple peoples to accountability processes.
Ambiguities over sovereignty in the antebellum US and EU states unions
In order to establish the nature of the relationship between ambiguity over sovereign status and indeterminacy over the boundaries of the people, it is first necessary to outline how and why sovereignty arrangements were indeterminate in both states unions. Even though one has a treaty foundation and the other a constitutional founding both experienced multiple sovereignty claims between the units and the union.
The antebellum US
Notoriously, the US constitutional founding did not specify a single locus of sovereignty in the classic hierarchical sense of ultimate and indivisible political authority. As Walter Bagehot put it, the American founding fathers purposefully ‘shrank from placing the sovereign power anywhere’ (Bagehot Reference Bagehot1963, 218). In this way, the constitution symbolized the retention of the Tudor principle of a government of ‘separated institutions sharing powers’ (Huntington Reference Huntington1966, 393), which the colonists had fought to preserve in the face of the new-fangled British doctrine of parliamentary sovereignty. The establishment of two levels of government, federal and state, in dynamic tension, was the basis for a republican system designed to ‘avoid the extremes of anarchy and hierarchy’ (Deudney Reference Deudney2004, 342). In this republican tradition, popular sovereignty could be said to be located actively in the people of the individual states, in the people of all the states combined, or recessed, that is, represented in the institutions created by the constitution since this had been ratified by state popular conventions (Fritz Reference Fritz2008).
The dynamic tension over the separation of powers was dominated by vexing conflicts over what status the states retained within this compound, rather than hierarchical, system of government. Essentially, these disputes represented a continuation of the original cleavage over federalist and so-called ‘anti-federalist’ visions of how the US republic should be organized in the aftermath of the revolution (Storing Reference Storing1985). The anti-federalists who opposed the Philadelphian system had sought to maintain a purely confederal arrangement between the former colonies. These pamphleteers were hostile to the constitution's establishment of a direct relationship between citizens and the federal level through the creation of a federal administrative and judicial system as well as to the invention of a federal level of political representation in the form of the presidency. These proposed constitutional changes overhauled the existing confederal relationship, under the Articles of Confederation (1781–89), in which states were the sole political actors and legal subjects of the inter-state order. Anti-federalists believed novelties introduced by the federal constitution abolished the sovereignty of the states, thereby rendering the government unitary and hierarchical rather than federal. This is why the ‘anti-federal’ label is a misnomer: they considered themselves advocates of ‘true federal’ principles (Storing Reference Storing1985). Consequently, they interpreted the US Constitution as the blueprint for an overweening and distant government that diluted the republican bond between states and citizens.
The anti-federalists lost the struggle to reject the federal constitution but the resulting form of government was far from a unitary state based on internal hierarchy abolishing unit autonomy. However, as Publius explained, the new system also had to be distinguished from traditional inter-state cooperation under anarchy: a confederation or ‘league’ of sovereign states that could not recognize a direct legal and political relationship between individual citizens and the union. The compound republican alternative rested instead on appreciating the fact that ‘we must extend the authority of the Union to the persons of the citizens, the only proper objects of government’ (Federalist 15). In practice, however, the question of the states’ sovereign status was very much ambiguous and was invariably at the heart of constitutional debates about the relationship between the units and the federal union.
The disputed question of sovereign status during the antebellum period, in the guise of ‘sovereign immunity’ to suits in the federal courts, is perhaps best exemplified by a constitutional dispute immediately after ratification of the constitution. It arose because the founding document declared that ‘the judicial power of the United States shall extend to …Controversies between a State and Citizens of another State’ (Article 3, Section 2.1). Although this principle was upheld in a 1793 Supreme Court Case, Chisolm vs. Georgia, it caused such a stir that an amendment to the constitution – one of only two enacted between the Bill of Rights and the Civil War – was swiftly passed to restore states’ sovereign immunity. This 11th amendment was the product of Georgia's stubbornness in refusing to accept the Supreme Court's ruling in Chisolm vs. Georgia that federal courts could hear non-state and foreign creditors’ suits against a state. States feared that British and Tory creditors would seek redress for confiscated property in the federal courts (Orth Reference Orth1987). On the other hand, committed federalists believed state immunity to such suits would exacerbate fiscal irresponsibility by allowing states to repudiate much of their debt.
The Supreme Court's reasoning in recognizing Alexander Chisolm's right to sue another state for unpaid war debts was intimately connected to a theory of the sovereignty of the people. Justice Wilson grounded his argument rejecting state sovereign immunity in the republican notion that ‘the supreme power resides in the body of the people’. This allowed him to claim, in a Madisonian fashion, that ‘the citizens of Georgia, when they acted upon the large scale of the Union, as part of the “People of the United States”, did not surrender the supreme or sovereign power to that state, but, as to the purposes of the Union, retained it to themselves’ (Orth 1987, 16). In other words, if sovereignty lay anywhere, it was in the people of Georgia as part of the greater collective body of US citizens; thus a state could not claim that a certain sovereign status was a sine qua non of its existence. As a result, there were no restrictions on what powers Georgia's citizens could vest in the federal government because the allocation of powers to the states themselves was the creation of popular sovereignty exercised by the combined citizens of the union. To put it another way, contrary to Georgia's claim, there is no residual sovereign status required for the individual states to continue to exist as states; their status is simply the product of what their people, in concert with the people of the other states, have decided it should be. Hence as long as the constitution – accepted by the combined peoples of different states – affirmed the Supreme Court's jurisdiction to settle such constitutional disputes it could not be argued that this verdict emasculated states’ sovereignty.
Yet this constitutional reasoning and theory of popular sovereignty was not self-evident. Consequently, within a month of the final judgment both houses of Congress had proposed amending the constitution to uphold state sovereign immunity in suits of law or equity (Orth Reference Orth1987, 20). The amending text found sufficient support amongst state legislatures in under a year, even though the union had grown to 15 states by this stage, although presidential proclamation of ratification had to wait until 1798. In fact, the 11th amendment marked the beginning of several decades of conflict over the status of the states within the federal union. This uncertainty was connected to the fundamental ambiguity of the 10th amendment and its reference to powers ‘reserved to the States respectively, or to the people’. Supporters and critics of states’ rights alike, therefore, attempted to make sense of the implications of this amendment through rival interpretations of where popular sovereignty lay and the extent to which a certain sovereign status within the units was necessary to maintain a properly federal union. Moreover, it must be borne in mind that throughout the antebellum period the actual right of the Supreme Court to consider itself the final arbiter over the boundaries between state and federal government as well as the right to strike down state law remained contested (Goldstein Reference Goldstein2001). Indeed, often the Court was viewed solely as a mechanism for reviewing the constitutionality of federal law, that is, ‘with power over cases and controversies arising under the Constitution itself, laws enacted in pursuance thereof … but not cases arising from state laws, which were reserved to state courts’ (McDonald Reference McDonald2000, 78).
Hence in the antebellum there existed a multitude of claims to sovereign status articulated by different political actors. The existence of these claims was in keeping with foundational ambiguity over where sovereignty resided and how popular sovereignty might be expressed when there was no single definition of the people. The dual separation of powers (within the federal government as well as between the federal and state governments) reflected a notion of accountability premised upon ambiguity surrounding who the people were and how exactly they could act together.
The EU
Seen in the light of the antebellum US, the sovereign status of the member states in the EU is similarly subject to ambiguity whilst disputes over this residual status form a constant backdrop to the politics of the integration process. Given the enormous literature on how EU integration affects sovereignty, it is sufficient here to highlight three aspects of status ambiguity within the integration process: opt-outs, intergovernmentalism, and the jurisprudence of constitutional courts. These examples best evince the importance of contested claims to sovereign status in the EU constitutional system.
Ad hoc policy ‘opt-outs’, negotiated to allow certain member states to agree to incremental advances in integration without compromising certain ‘red lines’ of sovereignty, illustrate well the politics of claims to sovereign status within the EU order. Under this arrangement, recalcitrant member states agree not to block treaty reform on condition that they will not be bound by certain new arrangements that constrain sovereignty. The first opt-outs were brokered during negotiations for the 1992 Maastricht Treaty. The UK opted out of the third stage of economic and monetary union, the single currency, as well as spurning the Protocol on Social Policy intended to provide Europe's citizens with greater social rights. Denmark similarly refused to convert to the Euro and also turned its back on defense cooperation in the nascent Common Foreign and Security Policy. Although the opt-outs on the single currency were initially considered to be merely temporary derogations, nearly two decades later Denmark and the UK still remain beyond the Euro pale, whilst Sweden, not a member state at the time of Maastricht, unilaterally refused to join after entering the EU in 1995. Moreover, the trend of opting out has continued with the UK, Poland and the Czech Republic withdrawing from the Lisbon Treaty's provision for making the Charter of Fundamental Rights legally binding when applying EU law. In all these cases, member states have expressed their right to opt out as a fundamental attribute of their sovereign status, that is, that failure to opt out would constitute an irreparable loss of sovereignty, although empirical data suggests that opt-outs merely mask de facto coordination on issues such as asylum and immigration (Adler-Nissen Reference Adler-Nissen2008).
A second arena of integration politics where contestation over sovereign status abounds is intergovernmental decision-making. Decision-making within the EU has increasingly shifted from intergovernmentalism (where states decide by unanimity) to supranationalism (based on a supramajority of states). Despite this shift, a delicate calibration has proved necessary in order counterbalance pooled sovereignty with respect for residual member state sovereignty. Hence, unanimity continues to be the norm for foreign policy, taxation, and treaty reform. Indeed, as integration deepened following the Maastricht Treaty, a new category of legal instrument was developed in Justice and Home Affairs as well as Common Foreign and Security Policy to ensure that these would be subtracted from the oversight of the EU's supranational institutions: the Commission and the Court of Justice. By denying the Commission the power of initiating legislation in these areas as well as the Court's jurisdiction to monitor the resulting legislative instruments, member states were able to create a hybrid intergovernmentalized law (Hanf Reference Hanf2001, 17) to reflect their residual sovereign status. The 2009 Lisbon Treaty enshrines continued intergovernmentalism by retaining existing vetoes and preserving the separate legal basis of the Common Foreign and Security Policy as well as restrictions on access to EU courts for matters pertaining to immigration and asylum (unlike in ordinary Community law, cases in these areas can only be referred by the highest national court).
The third and final illustration of the ambiguous sovereign status of member states within the EU system concerns the jurisprudence of certain national constitutional courts. These are the institutions whose acquiescence to the EU legal order was absolutely vital to the establishment of constitutionalism beyond the state (Stein Reference Stein1981; Weiler Reference Weiler1991; Goldstein Reference Goldstein2001) not least because in many member states there is a process of constitutional review for all EU treaties. Yet, even these courts have at times sought to take a stand for the residual sovereign status that allows them to delimit under what conditions they can continue to accept the terms of EU treaties. Surprisingly, it is the German Constitutional Court that has been most vocal here, beginning with its ruling on the constitutionality of the 1992 Maastricht Treaty.
In a startling decision, following a case brought by German Members of European Parliment claiming the treaty violated Germany's Basic Law (Grundgesetz), the Constitutional Court of Europe's most pro-integrationist country declared that the Union's democratic deficit placed definite limits on the constitutional transfer of powers away from the member states (Wieland Reference Wieland1994). Rejecting the complainant's claim that the Qualified Majority Voting principle was antithetical to Germany's Basic Law provision on the democratic character of political authority, the court nevertheless found that ‘should the Bundestag transfer too many of its competences, too much state power would be legitimated only indirectly; as a result, the Democracy Principle would be violated’ (Boom Reference Boom1995, 183). Although it did not specify what might constitute an illegal transfer of sovereignty, the Court based its judgment on the fact that the Union's democratic credentials were too weak as things stood to permit an empowered federal Europe well beyond the qualified extension of powers under the Maastricht Treaty. In particular, the court singled out the absence of ‘a constant, free exchange of ideas leading to a common public opinion, transparent and understandable (to the ordinary citizen) objectives of public authority, and the possibility of every citizen to communicate in his native tongue with public authorities to whom he is subjected’ (Boom 1995, 183). Indeed, the German Constitutional Court revisited this very topic in deciding that the Lisbon Treaty was constitutional (Eriksen and Fossum Reference Eriksen and Fossum2011). However, in its decision, the Court referred specifically to the EU as an association of sovereign states (Staatenverbund). Under this interpretation, the EU is considered a treaty-based association which exercises public authority, but whose fundamental order is subject to the decision-making power of the Member States and in which the peoples, that is, the citizens, of the Member States remain the subjects of democratic legitimation (BVerfG, 2 BvE 2/08).
In this way, the German Constitutional Court – exactly like Justice Wilson – made an explicit connection between sovereignty arrangements within a states union and what exactly constitutes the people that should have a say in resolving the disputes arising from status ambiguity. Thus, it is no coincidence that in both systems sustaining inter-state cooperation without hierarchy raised the issue of where popular sovereignty is ultimately located. Indeed, as demonstrated by the antebellum US example discussed in the following section, the desire to link sovereign status to popular sovereignty resulted in a debate over what political community had the final say to determine sovereignty issues. Two mutually exclusive visions of the level at which popular sovereignty should be exercised emerged in this dispute over where this hierarchy of democratic legitimacy resided. This development came at the expense of the Madisonian understanding than no single political community had an overriding accountability claim within the multiple separations of powers that rendered sovereignty ambiguous. This kind of zero-sum contest over whether democratic legitimacy lies with the citizens of the entire union or else those of an individual state has, as the third section explains, limited parallels with recent developments in the EU. Whilst the latter has experienced a clumsy attempt to legitimize it through a connection with popular sovereignty, this has not proved sufficient to undermine indeterminacy over the constitution of the people in the EU. Hence in the EU an absence of hierarchy continues to correspond with ambiguity over who the people are, preventing a clash between rival claims over where the hierarchy of democratic legitimacy resides.
Contesting sovereign status in the antebellum US: the competition over the hierarchy of democratic legitimacy
In the antebellum US, finding a via media between hierarchy and anarchy required a great measure of ambiguity over the states’ retention of sovereign status. But claims to sovereign status arising from competence clashes invariably made reference to the republican doctrine of the sovereignty of the people when seeking to legitimize the constitutional arrangement between states and the union (Fritz Reference Fritz2008). What ensued was a contest over sovereignty characterized by competing attempts to invoke popular sovereignty as both units and the federal government struggled to determine their respective claims to sovereign status. Conducted against a backdrop of changing mechanisms of political representation, this competition to define the proper locus of popular sovereignty wreaked havoc with the states union and its anti-hierarchical foundations. In particular, the popular sovereignty debate undermined the pre-existing ambiguity over who the people were and how exactly they could act in common.
The whole debate over the republican basis for units’ sovereign status was encapsulated by the stand-off between Daniel Webster and Robert Hayne on the Senate floor in 1830. The situation arose as a result of Congress’ protectionist tariff on manufactured imports, which South Carolina thought unfairly targeted plantation states. Hayne and Webster disputed whether popular sovereignty lay with the states or at the federal level to justify South Carolina's nullification of the tariff and President Jackson's defense of its legitimacy, respectively. Both senators found it necessary to re-examine the history of how the constitution was originally adopted: in particular, the people(s)'s role in authorizing the change in constitutional structure (Fritz Reference Fritz2008, 220–34). Excavating the past in this fashion sparked a profound theoretical reflection on the proper connection between states’ sovereign status and popular sovereignty within the union that also found echo in popular mobilization.
Webster espoused the ‘popular’ or ‘people's’ conception of the founding. According to this interpretation, the constitution ‘was not the creature of the states’ (Fritz 2008, 224) but rather the product of the American people in the aggregate. South Carolina's attempt to justify nullification – the sovereign right to refrain from applying a federal law, in this case the tariff – was from this perspective mere ‘revolution or rebellion’ (Fritz 2008, 245) because it contravened the sovereignty of the American people collectively. Conversely, Hayne argued (following the republican theory of John C. Calhoun) that popular sovereignty was and remained in the possession of the people of the various individual states. Historical evidence for this latter claim was provided by the fact that the states had convened special conventions in order to adopt the constitution. Logically, therefore, nullification, if the result of a special convention as indeed occurred in the South Carolina case, was merely an expression of republican self-government: an instance of active rather than recessed popular sovereignty. Moreover, the existence and configuration of the Senate and Electoral College was taken as evidence that the US Constitution sought to preclude mobilization of popular sovereignty at the federal level.
However, fundamental changes in the structure of political representation in the antebellum states union meant that by the 1830s a case could be made for locating active popular sovereignty in the American people collectively. Two significant and interrelated developments in democratization made Webster's argument about the true locus of popular sovereignty – and thus the hierarchy of democratic legitimacy – compelling in practice. First, there was the rise of mass, cross-unit parties formed to contest the presidency as an increasingly national institution, itself the second fundamental shift.
In 1804, 8 of the 17 states provided for the direct election of presidential electors; by 1824, only 6 out of a total of 24 states did not allow for direct election. Only Delaware and South Carolina did not follow suit by 1828 (Aldrich Reference Aldrich1995, 106). In this way the state legislatures lost control over the selection of presidential electors, enabling politics to become both more populist and national. Furthermore, as the parties in the post-Jacksonian era organized to mobilize political support they turned the election of presidential electors from one based on congressional districts to a winner-take-all principle so that the winning candidate received all the Electoral College votes (Gienapp Reference Gienapp1996, 87). This made it much easier for a candidate to win a landslide of states’ Electoral College votes with only a relatively small percentage of the popular vote, as Lincoln did in 1859 when 54 percent of the popular vote in the free states was enough to give him 98 percent of the North's electoral college votes (Gienapp 1996). With these democratic developments the presidential office therefore became the scene for hotly contested elections pitting rival parties and candidates against one another in the race for winning enough Electoral College votes across the union. Unsurprisingly, therefore, Andrew Jackson's 1832 proclamation in response to South Carolina's nullification claim was to argue that, thanks to the increasing majoritarianism made possible by a move to the direct election of presidential electors, ‘We are ONE PEOPLE in the choice of the President and Vice President’ (Elliot Reference Elliot1836, vol. 4, 589).
Thus the antebellum states union had changed markedly by mid-century. Changes in the system of political representation unleashed a fierce intellectual and popular challenge to this altered republican system, which was deemed a hierarchical deviation from the original constitutional blueprint by the champion of states’ rights, John C. Calhoun. In the light of this unexpected shift – the presidency was initially designed to be anti-populist whilst cross-unit parties were supposed to be stillborn in an ‘extended republic’ – Calhoun ‘thought that it was essential to revise republican theory and constitutional arrangements to fit these new circumstances’ (Ford Reference Ford1994, 45). The American union had to adapt to a novel situation in which despite the size of the republic and the founders’ constitutional devices the federal government was now potentially the instrument of a partisan majority, at least over the slavery question. As a result, in his famous Discourse on the Constitution and Government of the United States (1850) Calhoun developed not only a theory of ‘concurrent majorities’ as the cornerstone of federalism but also set forth ex post anti-majoritarian proposals to counterbalance the development of a system of representation more centralized and majoritarian – by activating popular sovereignty at the federal level – than at its origin.
Calhoun's gambit only served to exacerbate the contest over defining where popular sovereignty was located. Indeed, the argument between Webster and Hayne was in effect replayed 30 years later in the confrontation between newly elected President Lincoln and his southern antagonists. In his inaugural address, Lincoln spelt out clearly his position – reminiscent of Jackson's attack on the nullification doctrine – that popular sovereignty was exercised at the federal level and embodied in the office of the president. In that speech he explained that as ‘unanimity is impossible; the rule of a minority, as a permanent arrangement, is wholly inadmissible; so that, rejecting the majority principle, anarchy or despotism in some form is all that is left’ (Lincoln Reference Lincoln1991, 58). In return, the southern states drew on their own understanding of the appropriate level where republican popular sovereignty was located and formed special conventions to withdraw from the union.
In other words, as the debate over units’ sovereign status invoked appeals to determine the proper locus for the exercise of popular sovereignty, the ability to maintain an ambiguous compromise over the boundaries of state and federal authority alike was put in jeopardy. Arguments to invoke a recessed popular sovereign at either the state or federal level no longer allowed for a via media that left ambiguous the constitution of the people and how they could ever express a preference together. This meant abandoning James Madison's conviction that in a republican states union ‘a state acted in its “highest sovereign capacity” only when the sovereign people of a state acted in combination with the sovereign people of other states’ (Fritz Reference Fritz2008, 225). The antinomy between hierarchy and anarchy thus reappeared in the American states union as a result of seeking to use popular sovereignty to resolve constitutional disputes. Moreover, this focus on the level where democracy should be exercised entailed ignoring the issue of what form accountability could take when the constitution of the people itself was ambiguous. By the time of secession, what mattered was the debate over the level democratic accountability would be exercised not the form it took, given different configurations of the political community to whom power was responsible.
Hence, it is instructive to turn towards the EU case to investigate potential similarities, particularly in light of recent attempts to invoke popular sovereignty during constitutional debates over negotiating sovereignty in this system. The following section demonstrates how foundational ambiguity has remained an essential part of the EU despite certain attempts to peremptorily define the constitution of the people. As with the antebellum example, such attempts stem from trying to use popular sovereignty to resolve constitutional issues, revealing the constitutive connection between sovereignty and democracy. However, there are clear limits to the ability to mobilize around the notion of popular sovereignty within the EU, thereby minimizing the threat this represents to its dual foundational ambiguity.
Enduring ambiguity in the EU: multiple accountability claims representing different notions of the people
In the early decades of European integration, ambiguity over sovereign status and who exactly can exercise popular sovereignty – that is, who is ‘the people’ – was largely uncontroversial owing to a ‘permissive consensus’ that gave latitude to Europe's elites when negotiating inter-state cooperation (Moravcsik Reference Moravcsik1998). This permissive attitude no longer holds as sovereignty arrangements and democratic shortcomings within the EU exercise elite and popular sentiment alike. Despite an increase in EU competences since the 1992 Maastricht Treaty, there remains a marked separation between mass political participation in national politics and a relative absence of channels for participation in politically remote EU-level decision-making (Mair Reference Mair2007; Bellamy Reference Bellamy2010). There have also been popular manifestations of dissatisfaction with the limited role popular sovereignty has had in determining constitutional arrangements: anti-EU parties have experienced success in both European and national elections (Mair Reference Mair2007); national referendums have been pledged as well as held to try to endow EU treaty reform with democratic legitimacy (Binzer Hobolt Reference Binzer Hobolt2009). At the same time, EU elites continue to flirt with the notion of fostering popular sovereignty at a European level. This can be seen in the continued expansion of the power of the European Parliament, the financing of the party groupings sitting in the parliament, and the move towards personalizing EU policies around the twin figureheads of the President of the Council and the High Representative for Foreign Policy.
The combination of a flirtation with EU-level popular sovereignty and demands for national direct democracy to settle sovereignty issues raise the specter of two mutually exclusive constitutional visions devoid of ambiguity over the community that can be called the people. As with the antebellum US, any move to establish a hierarchy of democratic legitimacy would make it harder to sustain ambiguity over the units’ sovereign status by making sovereignty congruent with the community able to express popular sovereignty. However, the evidence suggests that democratic legitimacy is not articulated in hierarchical terms sufficient to undermine the EU's foundational indeterminacy. In particular, the inability to construct a plausible mechanism for popular sovereignty exercised at a pan-EU level prevents the emergence of two rival and mutually exclusive constitutional readings. This allows for an enduring uncertainty over the constitution of the people, as expressed by the multiplicity of accountability claims from different actors representing different configurations of the people.
Constitutional ambiguity is threatened to some degree by mobilization around the notion of popular sovereignty exercised at the national level. Invocations of popular sovereignty at the national level, especially via the holding of referendums on treaty reform, are based on the claim that the people of the separate member states ultimately have the final say on sovereignty arrangements within the EU. Not all countries choose to exercise popular sovereignty in this fashion (some such as Germany and Italy are constitutionally unable to hold referendums on international treaties). However, the trend is increasing (Binzer Hobolt Reference Binzer Hobolt2009) to the extent that when the Constitutional Treaty was awaiting ratification by the Netherlands in 2005 a successful public campaign was mounted to hold a non legally binding referendum, which helped derail the entire EU constitutional project. Indeed, the furore over the 2004 Constitutional Treaty reflects the residual power of the constitutional vision that locates popular sovereignty exclusively at the national level. In this vein, important member states such as France and the United Kingdom have taken legislative measures to ensure national referendums are held on future enlargement or treaty reform, respectively.Footnote 2 These measures indicate that certain member states are comfortable with designating their own electorates as endowed with the hierarchy of democratic legitimacy to accept or reject constitutional change in the EU.
Nevertheless, it is also revealing to examine what has happened in practice when a country's citizens have used a referendum to disavow a new treaty. In the recent Irish example of a failed referendum on the Lisbon Treaty in 2008, as also occurred when Denmark rejected the Maastricht Treaty or Ireland refused the earlier Nice Treaty in 2001, this setback was surmounted by applying inter-governmental pressure to make the recalcitrant country vote anew. Solving the constitutional problem created by the exercise of popular sovereignty at the national level – for a negative vote in a referendum blocks the entry into force of a treaty for all member states – has thus been a matter of denying the legitimacy of national referendums as means to settle sovereignty issues. This is further evidenced by the fact that the Lisbon Treaty is a bowdlerized version of the Constitutional Treaty that was rejected by French and Dutch voters in referendums in 2005.
Yet the resulting inter-elite arrangement is not accompanied by a rival claim that locates popular sovereignty at the EU-level. Practical attempts to foster popular sovereignty at the EU level have been insufficient to credit this level as the proper locus of the hierarchy of democratic legitimacy. The ever more powerful European Parliament, whose composition is increasingly important for the composition and policy agenda of the European Commission (Hix Reference Hix2008), is a legislature that does not have a formal say on EU treaty reform. Whereas it might have a credible claim to represent the will of the people of the EU for legislative proposals – although low electoral turnout undermines such a claim – this is not the case for treaty changes that affect sovereign status. Moreover, proposals from the academy to require a pan-EU referendum with a Swiss-style ‘dual majority’ principle requiring a majority of states and citizens to approve constitutional amendment (Auer Reference Auer2007) have gained no public or elite traction.
Of course, national referendums could be the means for a country to withdraw from the EU, as officially provided for by Article 50 of the Lisbon Treaty. This is after all the objective of a party such as the United Kingdom Independence Party, which seeks to use popular sovereignty to make sovereignty congruent with the British political community. If successful, this would end ambiguity over not just the sovereign powers of Westminster but also over British participation within an overlapping political community that crosses national borders. Hence this exclusive constitutional vision is a threat to ambiguity only for the national community that might leave the EU in this fashion – this exit would not affect foundational indeterminacy for the rest of the union.
Enduring indeterminacy about both sovereignty and the constitution of the people is characterized by the prevalence and indeed increase in accountability claims expressed by various political actors representing different configurations of the people across the EU. In the absence of a bifurcated debate between two rival, mutually exclusive visions of the location of popular sovereignty, the EU exhibits rather multiple channels of accountability relating to heterogeneous notions of political community. The relationship between sovereignty and democracy is thus more a struggle to define the form accountability can take within the EU's competence framework than a constitutional debate over the proper level for exercising popular sovereignty.
Certain accountability claims are made in the name of the broadest possible notion of the EU political community. This is the case with the ‘convention method’ used to draft the Constitutional Treaty (2002–03). Convention delegates were chosen from national parliamentarians, NGOs, and even countries with official candidate status for future EU accession (Karlsson Reference Karlsson2010). The ambition behind this convention was to create an improved deliberative setting in which to discuss the future contours of the EU legal and political system. Although they did not have an explicit electoral mandate for revising the treaties, members of the convention produced a draft treaty that provided the basis for national diplomats to subsequently write the 2005 Constitutional Treaty.
Referendums are not the only means for expressing an accountability claim on behalf of a specific national political community. Although divorced from the expression of popular sovereignty, national constitutional courts have proved willing to act on behalf of their nations to scrutinize the evolution of the EU. Most notably, perhaps, the German Constitutional Court has formally declared itself competent to oversee whether continued treaty change is democratically legitimate. Asked to pronounce on the Lisbon Treaty's (2009) compatibility with Germany's ‘eternal’ Basic Law that guarantees the principle of democracy, the court ruled that the EU ‘lacks … a political decision-making body which has come into being by equal election of all citizens of the Union and which is able to represent the will of the people’. The verdict, Eriksen and Fossum (Reference Eriksen and Fossum2011, 154) explain, means the court sees ‘the EU [a]s in democratic terms a derivative of the Member States’. However, national courts have accepted the writ of EU authority – less obstreperously indeed than their state counterparts did with federal authority during the antebellum (Goldstein Reference Goldstein2001). Moreover, national courts’ willingness to participate in enforcing EU legal authority has not only been instrumental in the constitutionalization of the EU system (Stein Reference Stein1981) but also corresponds with national legislative self-restraint. That is, domestic courts have been greatly assisted in applying EU law by the fact that member states have so rarely expressly sought to challenge EU rules through national legislation (Phelan Reference Phelan2010).
Such self-restraint is indicative of a self-conscious participation in a broader political community, whose boundaries are wider than the nation-state and yet not the simple aggregation of EU citizens. Here the European Parliament is in a sense the exception as the representative of the entire EU citizenry. The boundaries of the people are rather more fluid and ambiguous in the accountability claims that arise from the new citizens’ initiative and the mechanism for involving national parliaments in policing the subsidiarity principle, which lays down that competences should be exercised at the most effective level of government. Both of these features are innovations of the Lisbon Treaty. The former consists of a petition system whereby a million EU citizens from across at least a quarter of member states can formally ask the Commission to tackle a particular policy problem within existing EU competences. National parliaments have been granted two ways of objecting to EU legislative proposals they consider ultra vires. One third of national parliaments can mobilize together to invoke a ‘yellow card’ to ask the European Commission to review a legislative proposal whilst the stronger ‘orange card’ allows a majority of national parliaments to act in unison to block a bill until the EU's legislatures settle its constitutionality (Cooper Reference Cooper2011). Although the citizens’ initiative has yet to be used and despite there being no established tradition for inter-parliamentary cooperation these very possibilities are a clear sign that political action in the EU is not tied to either the national community or the EU citizenry in the aggregate.
Furthermore, the EU contains another institutional feature that complicates still more the question of the constitution of the people and by extension the form accountability takes. This complication arises from the possibility of ‘enhanced cooperation’, whereby a sub-group of member states can choose to pursue further integration around enhanced shared competences. Few such initiatives have been carried out in practice, with the notable exception of the Euro – the EU's own currency union. This sub-grouping of 17 countries from within the EU 27 constitutes an internal differentiation between participants and non-participants, with the latter having a reduced say on Eurozone decision-making. Nevertheless, as amply demonstrated by the recent public debt turmoil, the EU's institution for representing national heads of state and government, the European Council, is a setting for countries outside the Eurozone to pressure that sub-grouping to reform. In return, countries outwith the single currency, such as the United Kingdom, have agreed not to impede the legal process for this reform, which requires in part acquiescence of all 27 member states. In this way, the boundaries of political community in the EU again fail to correspond to an either/or distinction between national or European.
The constitution of the people within the EU, therefore, remains highly ambiguous despite attempts to invoke popular sovereignty at the national level to settle competency issues. The very definition of the people in whose name democratic accountability functions is, at various times, national, cross-national, and pan-European. As a result, there has been no contest, like the Hayne–Webster debate, between two rival interpretations of where popular sovereignty resides as a prelude to determining the boundaries of sovereignty between the units and the union. Instead of debating democracy in terms of the level where popular sovereignty is exercised, the EU continues to debate the form accountability takes where not only sovereignty but also the definition of who the people are remain ambiguous.
Avoiding a contest over the hierarchy of democratic legitimacy in international organization
The EU is fortunate, therefore, to have avoided a stand-off over the level at which popular sovereignty should be exercised. Yet the discussion over how the EU has achieved a ‘post-sovereign’ condition (MacCormick Reference MacCormick1999) omits precisely this possibility of a bifurcated contest over the hierarchy of democratic legitimacy; the same is true of most of the ‘democratic deficit’ literature imploring a change in the system. The reason for this omission is that scholars from both traditions have largely failed to identify the constitutive relationship between sovereignty and democracy and hence the risk posed by a constitutional debate over where the hierarchy of democratic legitimacy resides (cf. Bartolini Reference Bartolini2006). In particular, IR theorists have hitherto failed to connect the EU's ability to straddle the supposed boundary between hierarchy and anarchy to the ability to sustain ambiguity over who the sovereign people actually are. This lacuna is suggestive of the need to rethink the place of democracy within IR's fundamental analytical distinction between anarchy and hierarchy.
Consequently, this paper represents an attempt to do more than just situate the EU states union in relation to its antebellum US counterpart. The contrast made between these two polities has illustrated explicitly how and why the EU differs by virtue of avoiding the kind of divisive struggle over the hierarchy of democratic legitimacy that fatally compromised the US states union. Yet this analysis has implications beyond the scholarship that engages with the EU as an anomalous form of political organization.
Extrapolating from the two cases of states union analyzed here, it becomes clear that the republican notion of popular sovereignty (Morgan Reference Morgan1988; Fritz Reference Fritz2008), is a political principle ultimately seeking to make the exercise of power accountable to a single, bounded sovereign political community. According to this rationale, democratic accountability is ultimately premised on establishing a level of government where sovereign powers co-exist alongside a direct connection between governed and governing. In this way, popular sovereignty implies a hierarchy of democratic legitimacy that precludes the democratization of international cooperation in a non-hierarchical fashion. Proposals to legitimize the writ of international organizations by relying on the mobilization of political participation across international borders thus need to acknowledge the potential this creates for a clash over where the hierarchy of democratic legitimacy resides. Such a clash will have destabilizing effects upon international cooperation between sovereign states similar to those liable to occur in a states union. However, this analysis by no means suggests that democratic legitimacy is an impossibility in such circumstances. Rather, the argument is that international organizations enmeshed in the thickening web of global governance ought not to expect that their authority can be established from the source of a single, bounded political community. Hence the growing concern for democratizing the plethora of international organizations that increasingly set the agenda of public policy (Archibugi Reference Archibugi2008) needs to correspond with a better understanding of what democracy entails for international organization and cooperation between states.
The analysis pursued here of the relationship between sovereignty and democracy underscores the importance of taking democracy seriously as a constitutive element in international relations. In particular, the argument of this paper distinguished concern over the level at which popular sovereignty is exercised from the question of the form democratic accountability takes between states. Whereas the former seeks congruence between sovereignty and the people in whose name it is exercised, the latter allows for claims by different actors using multiple notions of the people to whom power must be accountable.
From this perspective, an emphasis on designating the territorial level where accountability ultimately takes place will produce hierarchical articulations of democratic legitimacy. Political actors representing rival, bounded conceptions of the people will thereby be drawn into exclusive and rival visions of which level of government can hold power to account. Even if mechanisms for fostering transnational political participation can be developed (Archibugi 2008), the idea of finding a territorial level where popular sovereignty can be exercised appears an unpromising gambit for democratizing international organizations. This suggests continued caution about learning from the ‘domestic analogy’ in IR, which remains a seductive idea as with the recent call to democratize international law by convening ‘global citizens’ juries’ (Goodin and Ratner Reference Goodin and Ratner2011).
By contrast, an emphasis on articulating the multiple forms of accountability that constrain the discretionary power of an international organization is a far more promising avenue for democratization. This implies shifting the focus of institutional reform away from identifying a level where accountability is exercised in the final instance to promoting multiple channels of accountability. These channels in turn ought to relate to institutions representing different conceptions of the people affected by policy-making rather than representing a single, bounded political community. It is this requirement of promoting multiple accountability claims based on overlapping notions of the people that can permit and sustain the foundational ambiguity found not just in states unions but also within formally institutionalized international cooperation.
This ambiguity exists precisely because treaty-based international organizations have more than simply powers conferred by sovereign states – they also increasingly have representative institutions and courts that act in the name of a constituted political community. This is the case notably for regional organizations such as the African Union, the Andean Community of Nations, and Mercosur. Although international cooperation in these instances is conducted on an anti-hierarchical basis, given member states’ wariness toward circumscribing their sovereignty, these organizations face repeated calls to do more and to do so more democratically. Hence in this context, the issue of democratizing international organization must problematize the form democratic accountability should take, so as to avoid a possible contest over where the hierarchy of democratic legitimacy resides. This means recognizing the multiple claims that different institutions can stake to hold power to account even whilst the constitution of the people is not only open to question (Näsström Reference Näsström2007) but also contested or negotiated from outside recognized territorial borders (Agné Reference Agné2010).
Thus scholars of international relations, who have already learnt from the EU about ambiguity over sovereignty (Keohane Reference Keohane2002), also need to acknowledge the room for ambiguity over who the people are in a democratic system. This requires a better appreciation of what constitutes a cosmopolitan democratic order that draws on the participation of different peoples but without a hierarchy of legitimacy (Eriksen Reference Eriksen2009; Eriksen and Fossum Reference Eriksen and Fossum2011), something also termed a ‘demoi-cracy’ (Nicolaïdis Reference Nicolaïdis2004). Here theorizing the interaction of both representative institutions, such as elected assemblies and intergovernmental bodies, and courts that speak in the name of different jurisdiction, is crucial. Evidence of this connection – and its importance to strengthening democratic accountability – can be found in the European Court of Justice's 2008 Kadi ruling, which successfully challenged the application of UN Security Council sanctions on an individual's Swedish-based assets because of concerns for due process (De Búrca Reference De Búrca2009). This example suggests, notably, the possible presence of a broad network of actors capable of holding the power wielded by an international organization to account (cf. Slaughter Reference Slaughter2004). The challenge for IR theory, therefore, is to put to use conceptions of democracy beyond the parameters of a single, bounded political community. In particular, these conceptions are required to explore the space between anarchy and hierarchy that is made possible through indeterminacy over both sovereignty and who exactly are the people in whose name power is to be held accountable. Here the research agenda concerned with exploring the legitimacy of authority claims made by global governance actors (Lake Reference Lake2010) appears especially fertile ground for integrating a focus on the multiple forms of accountability claims made possible via ambiguity over the constitution of the people.
Conclusion
This paper probed the relationship between sovereignty and democracy in a states union. The objective was to show the operation of this mutually constitutive relationship in so far as a states union depends not just on ambiguity over sovereignty but also over who the people actually are. Contrasting the antebellum US and the contemporary EU revealed the significance of invoking popular sovereignty as a method for seeking to resolve constitutional issues. Foundational ambiguity in the antebellum was shown to have been undermined by this invocation, which sought to establish one level of government as democratically hierarchical to the other. Instead of sustaining an ambiguous reading of who the people who could actually exercise popular sovereignty were, political actors engaged in a bifurcated constitutional debate that sought to make sovereignty and the people congruent at either the state or the federal level.
The Madisonian constitutional design that was skeptical about allotting a constitutional role to popular sovereignty was also diffident about defining the people as a single, bounded entity capable of acting as one (Fritz Reference Fritz2008). However, by the 1830s, the Hayne–Webster debate revealed how US constitutional visions had strayed from the Madisonian middle ground. Calhoun's proposed constitutional innovations to remedy the union (concurrent majorities and constitutional conventions) can be interpreted as a last attempt to confound the divisive struggle over establishing a hierarchy of democratic legitimacy at one level of government (Ford Reference Ford1994). Yet changes in the system of political representation furthered the claims of supporters of popular sovereignty at the federal level and likewise encouraged their antagonists at the state level.
By contrast, the EU has witnessed limited attempts to invoke popular sovereignty to resolve constitutional disputes over sovereignty. More importantly, this has been uni-directional: national referendums have been used to settle ratification of EU treaties. There has not been an accompanying rival move to mobilize popular sovereignty at the EU-level. In fact, the latest legal-political reform of the EU, the Lisbon Treaty, enshrines dual foundational ambiguity based on avoiding a peremptory definition of who the people are and where sovereignty resides. Thanks to the new provisions of this treaty, accountability claims can be made, at various times, in the name of a national, cross-national, and pan-European people. Hence the key constitutional question remains centered on the form accountability can take given multiple definitions of the people – not the level where sovereignty and the people are both congruent.
Consequently, the EU has managed to avoid the baneful establishment of competing notions of where the hierarchy of democratic legitimacy resides. The EU might have stumbled upon this solution more by accident than by design. Yet this should not prevent other instances of cooperation between states learning the lesson that establishing a middle ground between anarchy and hierarchy requires more than just ambiguity over where sovereignty resides. It is ambiguity over who the people are, allowing for multiple accountability claims based on overlapping notions of political community, that is necessary to avoid a contest over the hierarchy of democratic legitimacy.
Acknowledgements
Earlier versions of this paper were presented at workshops held at Edinburgh Law School and the University of Michigan Law School, the latter as part of a SIAS Summer Institute funded by the Alexander von Humboldt Stiftung and the Andrew W. Mellon Foundation. I would like to thank the respective organizers, Cormac Mac Amlaigh, and the tandem of Christoph Möllers and Daniel Halberstam, for invitations to present this research as well as for their learned comments. The four anonymous reviewers provided acute and exceptionally helpful feedback that greatly improved the quality of the argument. Another intellectual debt is owed to the editors of International Theory, for providing encouragement and generous assistance in the revising process. Before submission, Patrick Jackson was a source of wise counsel when reviewing the original version of this paper. Finally, Michael Glencross was a patient and highly attentive proof-reader.
