Norms and Ethics in International Relations

Andrew Hurrell. Handbook of International Relations. Editor: Walter Carlsnaes, Thomas Risse, Beth A Simmons. Sage Publication. 2002.

This chapter examines the study of norms and ethics in the field of international relations (IR). In particular, it explores the links between normative theory (What would a just world order look like? How should it constructed?) on the one hand, and the historically created normative practices embedded within the institutions of international and global society on the other. For all the ‘norm-talk’ inspired by constructivist scholarship, the status of normative political theory remains somewhat elusive. The first section therefore seeks to give a general overview of how the sub-field of international ethics or international normative theory has developed; the main approaches adopted; the reasons for its continued importance; and the different ways in which normative theorists have engaged with the empirical study of norms and institutions in world politics.

Building on this last point, the second section of the chapter explores the work of those writers (political theorists, historical constructivists, international lawyers and those working within the tradition of the English School) who have sought to understand how the normative structure of international and global society has changed and evolved through the period of the classical European state system and into the age of globalization. It also seeks to underline the importance of focusing on normative structures in order to understand better the nature and meaning of specific rules, how they affect behaviour, and how they change through time. The final section examines the difficult and contested subject of cultural difference and diversity. Cultural diversity and value pluralism have been central themes in both historical and contemporary normative analysis of what a just world order might look like. And they are equally central to all attempts at evaluating the breadth, depth and degree of consensus that exists in international society at the present time.

Normative Political Theory

One can approach the subject of norms and ethics in IR from three distinct perspectives: the first considers the role that normative ideas play in the practice of politics (‘How have ideas about what should be done influenced political behaviour?’); the second seeks to engage in rational moral debate as to the nature of ethical conduct (‘What ought we to do?’); and the third examines the extent to which moral behaviour is heavily constrained by the dynamics of political life (‘Given the realities of political life, what can be done?’). Mervyn Frost argues that the academic study of international relations remains resistant to normative theory, especially in this second sense:

The post-positivists have given us rich accounts of the complex social structures within which we are constituted and within which we act. They have shown how these structures are not static but are in constant flux. They stress the important role which normative considerations play in the practices which we study and in our practices of scholarship themselves. But, once again, following a long tradition of IR scholarship, they stop short of engaging in detailed normative theorizing about what ought to be done. In particular they fail to take on that most important normative question of all: in what would a just world order consist? (Frost, 1998: 129; see also Smith, 1992)

Nevertheless, one of the most striking developments of the past fifteen years has been the re-emergence of normative political theory. This has involved the forging of closer links between IR and mainstream political theory, a more sophisticated reworking of classical traditions of thought on international relations, and the development of a rich and impressive literature on topics ranging from nationalism and national self-determination, to intervention and just war, to global distributive justice.

In part this resurgence has followed from attempts to rethink the way in which we mark off international politics from domestic politics. As the academic study of international relations developed, it became common to stress the uniqueness of the international–either as a realm of anarchy and conflict or as a society but one structured around institutions that were distinct from those that existed within the state (Bull, [1977] 1985; Schmidt, 1998). For Wight this meant that ‘international theory’ differed fundamentally from ‘political theory’ and looked to its own history and to its own traditions (Wight, 1966).

And yet as Boucher (1997, 1998), Brown (1992), Linklater (1998), Rengger (2000), Walker (1993), and others have argued, this approach involved a strangely limited view of politics and of political theory. If we shift the focus of the question and if we alter the way in which we conceive of the divide between the domestic and the international, then we get a rather different set of theoretical questions and a rather different set of historical debates and traditions that might help enlighten them (Mapel and Nardin, 1992, 1998). A concern with the nature and extent of community, for example, places the moral basis of the state in question and opens up a fundamental division between cosmopolitan and communitarian answers to normative issues (Brown, 1992; Frost, 1996). For others it is that very division that is problematic and normative theory is intrinsically bound up with the principles and practices of state sovereignty and with the recognition that ‘Sovereignty is a claim to particularity that can be meaningful only in relation to something more general’ (Walker, 1999: 154–5, 1993). Or, to quote Boucher:

political theory and international relations theory become genuinely united in focusing upon the issue of identity, that is, answers to the question ‘Who am I?’ The consequent politics of recognition, inclusion and exclusion bring into focus not only the politics of the nation, state and cosmopolitanism, but also of gender, race, religion and ethnicity. (Boucher, 1997: 193)

The revived interest in normative inquiry also followed from broader doubts as to the purposes of social theory. In part, this reflected the power of Robert Cox’s simple but powerful statement that ‘Theory is always for someone and for some purpose’ (Cox, 1986: 207), and his influential development of the distinction between critical theory and problem-solving theory. In part, it followed from the attacks by a wide range of critical and postmodern theorists on positivism–the idea that the observable regularities in the social world can be analysed with the same methods that have proven successful in analysing the natural world and that social interaction is governed by objective forces the causal workings of which can be formulated in terms of general laws which hold independent of human subjectivity. On this view, then, the return of the normative reflects the failure of social science to provide definitive answers as to how the political world operates and the continued role of moral choice even in those cases where we think we know how the political world operates. Post-positivist routes into the normative have been especially important for critical theorists such as Linklater, concerned with the possibility of expanding the character and range of political community (Linklater, 1998); and for anti-foundationalist postmodern approaches to both political theory (for example Rorty, 1993) and IR (Campbell and Shapiro, 1999; Der Derian, 1987; George, 1994; Walker, 1993).

However, the expansion of normative writing on international relations also reflected changes within the mainstream of Western political theory. In part this had to do with broader ‘resurgence of political theory,’ as David Miller characterized it in 1990 (Miller, 1990: esp. 427–34; see also on this topic, Parekh, 1996) and, in particular, with the torrent of work that followed from the publication of John Rawls’s A Theory of Justice in 1971. The intense debates between individualist and communitarian liberals inevitably involved competing understandings of global politics. On the one hand, communitarians were concerned to provide moral justification for a world of separate units and for the problems that followed from the different character of these units (peoples, nations, states, nation-states). On the other, cosmopolitans were inevitably drawn to question that very division and to delineate more carefully the range of our ‘duties beyond borders’ (Hoffmann, 1985). Thus, for example, Beitz and Pogge took up Rawls’s arguments in favour of egalitarianism within the state and argued that the changing conditions of justice now required distributive justice to be applied on a global scale (Beitz, [1979] 1999b; Pogge, 1989).

But engagement was also forced upon political theorists by the intensity and complexity of the moral dilemmas of world politics. This had of course been true of the Cold War period which saw major work on just war (Walzer, 1977), on the ethics of nuclear deterrence, and, for a rather brief moment in the 1970s, on distributive justice (for example Singer, 1972). But the inevitability of engagement grew stronger in the post-Cold War world–as a result, for example, of the frequency of ethnic and nationalist conflict; the so-called ‘new interventionism,’ the increased activism of the UN and the revival of debates around the concept of humanitarian intervention; the degree to which increasingly intrusive international rules (for example on human rights) combined with renewed US hegemony and Western dominance sharpened the issue of cultural diversity; or the continued growth of global inequality and the manifest failure of the global system to provide even the most minimal conditions of human welfare and well-being for a very large proportion of humankind (for details see UNDP, 1998). It would be an odd kind of political theory that was not led to ponder on the manifest injustices of contemporary world politics.

Yet lurking behind these specific examples has been a more general and potentially important shift. For many theorists (and for much political and public opinion), increased economic interdependence, an increasingly dense and activist transnational civil society and an expansion in the number and scope of international institutions have shifted understandings of global justice. In the first place, they have necessarily altered the scope of justice. It is, for example, no longer possible to accept Martin Wight’s classic distinction between domestic society as the arena within which understandings of the good life might be debated, developed and, potentially, realized, whilst international relations are condemned to remain the arena of ‘mere survival.’ To take only the most obvious example, ‘mere survival’ in relation to the protection of the global environment depends fundamentally on how societies are organized domestically and on how their various conceptions of what the good life entails can be brought together and reconciled. Material and moral circumstances have therefore pushed international society inevitably beyond Nardin’s practical association–‘an association of independent and diverse political communities, each devoted to its own ends and its own conception of the good’(Nardin, 1983: 9).

Second, integration and globalization have undermined the boundedness of political communities whose particular cultures, traditions and ways of living are given so much weight by communitarians. For many they have given a new reality to the sense of sharing a single world and to the nature of plurality, connection and finitude (O’Neill, 1996: ch. 4). In an integrated world, after all, what possible sense can one make of Rawls’s taking as his starting point the idea of a basic social structure built around the ‘self-sufficient schemes of cooperation for all the essential purposes of human life’ (Rawls, 1993: 301; emphasis added)?

This is not to suggest that the tide is flowing in a clearly cosmopolitan direction. Indeed, it is striking to note that a more or less traditional vision of international society is defended by some of the most important contemporary political theorists, including most notably John Rawls (Rawls, 1999). Other examples include Robert Jackson’s powerful reworking and extension of the case for normative pluralism and of the rules and institutions of the state system as providing the framework for that pluralism (Jackson, 2000). But it is to note how even more pluralist and communitarian theorists such as Miller or Walzer have been forced to defend their view of the limits to global justice more carefully (David Miller, 1995b, 1999), or to accept the thin obligations that have already been established in the practices and principles of contemporary international society (Walzer, 1994).

The range of issues and problems addressed by normative political theorists is therefore extremely wide. So too is the range of approaches–consequentalist (Singer); Kantian constructivist (O’Neill); Rawlsian (Rawls, Pogge, Beitz, Teson); pluralist or communitarian (David Miller, Walzer); Habermasian (Habermas, Linklater); constitutive (Frost, Brown); pragmatic (Cochran). It is impossible in a short chapter of this kind to enter into a substantive discussion of or even to survey the main schools of thought (for such a survey see Brown, 1992; Thomson, 1992; also see Cochran, 1999; Frost, 1996). Instead I will develop one particular theme: the relationship between normative theory and the role of norms in the practices of world politics. The conventional position is, of course, to insist on the fact/value distinction and to see normative and empirical theory as separate enterprises. As Charles Beitz puts it: ‘The analytical and the normative, although related, are different domains, and it is a virtue, not a shortcoming to observe the distinction’ (Beitz, 1999a: 270). Yet, even leaving aside the foundational doubts of critical and postmodern theorists, this distinction is far from clear-cut and involves a number of difficult issues.

One dimension concerns feasibility. At one extreme are those who stress the constraints of international politics. Thus classical realists (who were of course themselves passionately committed to a moral project and to uncovering and exposing the dangers of utopianism) believed that moral politics were closely tied to rational statecraft and to the ethics of responsibility that would inevitably dominate the choices of political leaders. This was itself both a moral project and, together with the idea of the national interest, provided the moral basis for realism. In a related vein, English School theorists argued that normative theory should concentrate on the ways in which the law of the jungle might be at least marginally deflected; on the kinds of international society that it might be just possible to establish; and on the principles of prudence and moral obligation and the consensus of shared values that actually existed within international society and that might form a part of achieving this goal. The difficulty with these positions is that they risk becoming an apologia for state practice and that they forswear any clear argument as to what principles ought to guide the practice of international politics and on what sorts of foundations they rest. Prudential ethics are all very well, but, although they speak to the difficulties of going anywhere, they do not give a very clear sense of where we ought to be trying to go. (For a critique of the alleged greater ‘realism’ of an ethics of responsibility, see McCormick, 1999: ch. 1.)

For those at the other extreme, political theory is in the business of stating clearly what justice requires, and of pushing out the normative boat at least to the margins of what conventional opinion takes to be plausible. For some, there is no great concern with the relationship between what is normatively desired and the chances of it coming about. As Brian Barry puts it: ‘most truths are pointless in that sense [i.e. efficacious] and none the worse for that. It rebounds to the honor of the human race that there should be people going about saying that slavery is an absolute evil, even if this cuts no ice with the beneficiaries’ (Barry, 1990: 368). For others, the aim should be to delineate a grounded utopia or a realistic utopia. As Rawls puts it: ‘Political philosophy is realistically utopian when it extends what are ordinarily thought of as the limits of practical political possibility’ (Rawls, 1999: 6; see also 11–16). What might constitute the realism of this realistic utopia depends partly on the time-frame and partly on judgements and assessments that derive from the empirical study of politics.

Here the more empirically minded political scientist is likely to enter two caveats. The first concerns the important distinction drawn by O’Neill between abstraction and idealization and between ideal and idealized theory. ‘For example, if human beings are assumed to have capacities and capabilities for rational choice or self-sufficiency or independence from others that are evidently not achieved by many or even by any actual human beings, the result is not mere abstraction; it is idealization’ (O’Neill, 1996: 41; see also discussion in Hoffmann, 1998: 40–1). The second concerns the way in which empirical evidence is incorporated into normative argument. In some cases the problem might lie with the weight placed on a particular finding (as, for example, when Rawls uses Sen’s work on famines to conclude that it is domestic political culture rather than, say, external vulnerability, that determines the conditions for minimal economic justice). In other cases the problem lies in the uncritical elevation of a contingent and contested argument–as, for example, when Rawls talks in The Law of Peoples of the ‘Fact of Liberal Democratic Peace’ (Rawls, 1999: 125; capitals in original).

The other dimension is more fundamental and relates to the foundations of the normative arguments being presented. Again, let us take two extreme positions whilst acknowledging the reality of many intermediary possibilities. At one extreme, it is precisely human reason that provides both the foundation of moral argument and the hope that it can be acted upon. The normative theorist begins with his/her best considered judgement based on reasons that are suitably coherent and generalizable. Whatever people may actually believe, he or she seeks to find good reasons why they should alter their beliefs and their patterns of behaviour. Showing that certain values are widely accepted in social practices is not the same as providing valid arguments as to why they are justified. But even many of those who wish to start with their own ‘best considered judgements’ as to what justice requires and who seek to build theories of justice around principles that would be chosen by any rational individual do not end the story there. Thus, for example, Rawls insists that the theorist’s considered judgement be related to the values that are available within the political or moral culture of a given society (the idea of a so-called reflective equilibrium) and that valid principles of justice must be publicly justifiable. (See the discussion in Miller, 1999: ch. 3.)

For other theorists, normative theory should begin with the norms and values that exist within particular communities. The critical side of this position is to stress that both the concerns and the methods of political theory are themselves the product of particular historical traditions and are rooted within a particular cultural tradition. Thus, Freeden, for example, seeks to contextualize contemporary American philosophical liberalism and to uncover the way in which its inherently ideological character is disguised by the methods and style that it adopts. ‘Liberal principles are consequently stated in such a way as to blur the distinctions between the theory and the ideology; in particular–again, as with Marxism–to disengage the theory from a diachronic tradition of thought against which its identity has been subject to continuous appraisal’ (Freeden, 1996: 227, but see especially ch. 6). Or Rogers Smith analyses the ways in which Rawls’s work fits within and grows out of broader patterns of American liberal political thought (Smith, 1997: 481–5). From this perspective, abstract theoretical analysis, whether normative or empirical, can never be wholly divorced from the historical circumstances against which it arose. It can never be treated as though it had neither a past nor a future–nor, of course, that it is connected to particular structures of political power.

The more positive aspect is to seek to ground normative theory in the existing normative practices of particular communities. As Michael Walzer puts it, ‘principles of justice are … the inevitable product of historical and cultural particularism’ (Walzer, 1983: 5–6, quoted in Barry, 1990: 365). Theory should uncover, interpret and critically develop understandings of morality that exist within specific international historical and cultural contexts. The theorist or social critic ‘gives expression to his people’s deepest sense of how they ought to live.’ On this view, as David Miller puts it: ‘There are no universal principles of justice. Instead, we must see justice as the creation of a particular political community at a particular time, and the account we give must be given from within such a community’ (Miller in Miller and Walzer, 1995: 2).

This is an approach to grounding normative debate that ties in closely with English School perspectives on IR: the need to nurture (but not uncritically accept) the normative consensus that has come to develop within international society; to build arguments and proposals for greater justice out of the values and modes of reasoning that have already begun to take root; and to pay very close attention to the balance between substantive principles of global justice on the one hand and the principles of fair institutional process that underpin the reality of some kind of international community on the other. This conventionalist or internalist position has been central to communitarian theorists. But there is no reason in principle why such accounts of justice cannot provide a fuller account of the pluralist case for the state system (as with Jackson, 2000), or indeed push beyond pluralism. Thus, for example, changes in the practices of international institutions or of transnational civil society may well move different states and societies towards ‘shared understandings of the meaning of social goods,’ to use Walzer’s phrase (Hurrell, 2001).

In a different idiom, critical theorists also seek to push towards emancipatory reform but on the basis of tendencies and that are already potentially immanent in current practices. To quote Andrew Linklater:

No critical theorist influenced by Marx would deny that normative theory must engage with actual social structures and real social conflicts … critical theory in the broadly Marxian tradition avoids the extremes of resignation to international political fate and retreat into utopian fancy. Normative theory in this tradition focuses on what is immanent within existing social structures, and intimated by the decisive social struggle in any epoch–not on the requirements of disembodied reason. (Linklater, 1999: 165)

It is well beyond the scope of this chapter to try to defend a single resolution of these intrinsically difficult and contested issues. Although it remains important to try to separate out questions of normative rightness or plausibility from questions relating to the sociology of norms (How did such an idea arise? What function does it serve?), finding stable criteria for adjudicating these distinctions is fraught with difficulty. And yet we might still wish to deny the claim, made by both realists and Marxists, that all moral discussion can be reduced to simple-minded notions of self-interest or viewed as simply reflective of material forces. Nor is it correct, as so much postmodern ethical argument suggests, that all moral debate can be reduced to a clash of incommensurable contending discourses.

No doubt in a politically and culturally fragmented world there are examples of true conflict. But there will always be scope for rationally grounded moral debate and discussion. Much moral argument involves issues of facts and the application of shared moral principles to particular empirical circumstances. Equally, a good deal of moral debate is concerned with the definition of rules and their logical relationship to one another. When we move beyond these shared premises and into the realm of true incommensurability, then there is no alternative but to strive to nurture the political institutions through which normative consensus might be developed and extended; and, in particular, through which respect for pluralism and coexistence might be reconciled with the need for more active cooperation and for an overlapping moral consensus. This is, after all, part of the specificity of politics: that arena in which the practical results of different moral visions are to be negotiated and rationally constructed if they are not to be imposed by coercion and unequal power.

Normative Structures

As is evident from the previous section, normative theory has for a long time (and necessarily) interacted closely with ideas and arguments about the changing character of international society and capacity of that society to provide a satisfactory basis for international or global order. What kinds of ordered cooperative relations have states and their governments sought to establish amongst themselves? What kinds of shared purposes and values (if any) can be discerned in the norms, rules and institutions that structure and regulate their interaction? How far have these been successfully achieved and what explains the gap between normative ambition and political realization? Such questions have been explored most directly in the classic studies of international order (Aron, 1966; Bull, [1977] 1985; Falk, 1972, 1975, 1999; Hoffmann, 1987, 1998; Mayall, 2000; Lynn Miller, 1990; Vincent, 1974). It also connects with more recent discussion of global governance (Diehl, 1997; Keohane, 2001; Nye and Donahue, 2000; O’Brien et al., 2000; Reinecke, 1998; Woods, 1999; Young, 1999). There is an obvious overlap with the literature on institutionalism surveyed elsewhere in this volume. But the focus here is less on theoretical understanding how particular institutions or regimes emerge and develop, and more on assessing the overall character of institutionalization in world politics, the normative commitments of different varieties of institutionalism, and the adequacy of existing institutions for meeting practical and normative challenges.

There are many aspects of change that are of potential importance for normative theory. Equally, the range of writing on the broad subject of change and transformation in world politics has been enormous. We have seen major works of synthesis (for example Buzan and Little, 2000; Ruggie, 1998), as well as a flood of work examining the links between globalization and transformation (see in particular Held et al., 1999; Held and McGrew, 2000; Scholte, 2000). But the concern here is with a particular subset of the analysis of change, namely change in what may be called the normative or constitutional structure of international society, as this is manifest in international legal rules and practices, in international political norms and in the dominant ideologies and practices that animate them.

Many of the proponents of deep change offer a stark and highly implausible choice: between a grossly simplified image of a Westphalian past and an invariably complex, but usually underspecified, post-Westphalian present and future. Even the most suggestive and sophisticated of those arguing for change tend to skirt rather quickly around juridical and normative structures, to conflate claims about power and control with claims about law and authority, and to rely on a range of plausible but elusive concepts. Ideas about ‘post-sovereign states’ or ‘multilayered geo-governance’ do indeed point to potentially very important changes but they are embedded in a discourse of normative transformation that is in most cases extremely difficult to pin down. On the other side, those who stress continuity rely on such a one-dimensional view of the role of norms and such a very thin notion of the legal order that it becomes impossible to make sense of the tremendous changes that have indeed taken place, above all in the period since 1945 (for example, Krasner, 1999).

One question, then, involves seeking to delineate and trace models of the international normative order, both in terms of the different goals to be promoted and the different means of achieving them. Thus Falk distinguished between Westphalian and Charter conceptions of the international legal order (Black and Falk, 1969), a theme developed by Cassese (1986). Bull distinguished between ‘pluralist’ and ‘solidarist’ models of international society (Bull, [1977] 1985; Alderson and Hurrell, 2000), a way of thinking developed more recently by Wheeler (2000). Nardin (1983) distinguished between purposive and practical associations. Reus-Smit (1999) sought to analyse different international societies by distinguishing between issue-specific regimes, fundamental institutions and constitutional structures. Philpott (2000) has traced the development of ‘revolutions in sovereignty.’

Others have engaged with the immensely difficult task of trying to think through the possible nature of a post-sovereign normative order.

Thus political theorists such as Held have sought to chart the contours of cosmopolitan democracy (Archibuggi and Held, 1995; Archibugi, Held and Köhler, 1998; Held, 1995), whilst others such as Linklater have delineated the possible character of the post-sovereign state (Linklater, 1998). Within international law, Chayes and Chayes (1995) and Koh (1996) have developed transnational process and managerial models of the emerging legal order; whilst Slaughter has provided an important depiction of how a liberal international legal order would differ from traditional understandings (Slaughter, 1995, 1997 and 2000; for two very useful collections on the changing roles of International Law see Byers, 2000 and Goldstein et al., 2001).

This branch of international relations has brought together political theorists, analytically minded international lawyers, historical constructivists, and those working within the tradition of the English School. There are many differences and shades of opinion, but also a series of general commitments. From this perspective, the international system cannot be viewed solely in material terms as a decentralized, anarchic structure in which functionally undifferentiated units vary only according to the distribution of power. Central to the ‘system’ is a historically created, and evolving, structure of common understandings, rules, norms and mutual expectations. The concepts of state sovereignty, international law or war are not given by the game of power politics. Rather shared and historically grounded understandings of war or sovereignty define what the nature of the game is and how it is to be played and, very critically, how it might change or evolve. IR cannot therefore be taught or analysed as part of an abstract ‘game’ independent of its human or historical origins. It grows out of its past, although never fully outgrows it. The system is the result of historical processes of change and development and therefore cannot be understood in terms of timeless, immutable laws.

Material structures matter, but these material structures cannot be understood outside of the shared knowledge and shared understandings held by the actors themselves. Amongst these shared understandings and intersubjective knowledge, norms and institutions play a number of fundamental roles. They may well serve as regulatory rules designed to constrain choices and/or as parameters within which individual agents pursue their own preferences. But the critical point is that they do far more than this. In the first place, they help explain how actors are constituted. As Wendt puts it, this is ‘one of the most important things a structure can explain’ (Wendt, 1999: 16). Second, they help us make sense of the identity of actors and hence of the sources of their preferences. Norms can be understood as expressions of what states are, where they belong and of the kinds of roles they play. And third, norms do not simply constrain but also enable and empower action. In Martin Hollis’s words: ‘They enable not only by making collective action easier but also by creating forms of activity’ (Hollis, 1991: 13).

A great deal of recent work has involved more detailed historical understanding of how the institutional structure of the classic European state system emerged and developed. Although bland (and often inaccurate) references to ‘Westphalia’ still blight many textbooks, the past two decades have seen a sustained attempt to get beyond the textbook stereotypes. Thus the study of classical theories of international relations has grown significantly (for general studies see Boucher, 1998; Clark and Neumann, 1996; Doyle, 1998; and Knutsen, 1992) and there have been important reassessments of the major traditions of thought in international relations (Dunne, 1998; Nabulsi, 1999; Walker, 1993). Westphalia has been demythologized by Osiander (1994 and 2001), Philpott (1999) and Krasner (1999). Biersteker and Weber (1996) and Bartelson (1995) have emphasized the importance of seeing state sovereignty as a social construct. Work on the history of international law remains underdeveloped but important works for the study of international relations include Koskenniemi (1989) and Grewe (2000). And finally there has been a very important move into the area of ‘international relations’ on the part of those working within the history of political thought–a move which has expanded immensely the degree of historical sophistication in the study of the subject (see in particular Hont, 1994; Pagden, 1995; Rothschild, 1995; Tuck, 1999).

Understanding the historical evolution of the normative structure matters for the obvious reason that it is only by understanding the past that we can gauge and assess claims for change and transformation. But understanding normative structures matters for other reasons and shapes our understanding of the role of norms in international life: how we interpret and understand specific rules, how they affect outcomes, and how they change through time.

What Norms Are

All normative analysis revolves around the two classic meanings of the term ‘norm.’ On the one hand, norms are identified by regularities of behaviour among actors. Norms reflect actual patterns of behaviour and give rise to expectations as to what will in fact be done in a particular situation. On the other, norms reflect patterned behaviour of a particular kind: a prescribed pattern of behaviour which gives rise to normative expectations as to what ought to be done. Thus regularity is combined with an internal attitude involving criticism of oneself or others on the ground that a particular norm is being violated. Norms can therefore be defined (and will be so used in this chapter) as ‘a broad class of prescriptive statements–rules, standards, principles, and so forth–both procedural and substantive’ that are ‘prescriptions for action in situations of choice, carrying a sense of obligation, a sense that they ought to be followed’ (Chayes and Chayes, 1994: 65).

Deciding which norms matter and in which circumstances will be difficult, if not impossible, without being able to systematically distinguish levels of authority or normativity among the different principles or rules within a particular regime (Brunneé and Toope, 2000). Equally, compliance cannot be reduced to following a static set of clear sharped-edged rules (see Chayes and Chayes, 1995; Kingsbury, 1988; Koh, 1997). But for the sceptic, it is the very ubiquitousness of norms in international life, their frequent ambiguity, and the variety of interpretations to which they may give rise that is the problem. Powerful actors can always find a norm to support their consequentially based choice. Norms are often in conflict. As Krasner puts it: ‘International rules can be contradictory … and there is no authority structure to adjudicate such controversies.’ Or again: ‘The international environment has been characterized by competing and often logically contradictory norms, not some single coherent set of rules’ (Krasner, 1999: 6, 54). The strong can always pick and choose amongst existing rules or set new ones, through formal agreement or through custom–a source of international law which places disproportionate weight on the actions and arguments of powerful states (Byers, 1999).

It is important to recognize the seriousness of the problem and to avoid a stylized or idealized view of the legal order. But it is one thing to accept that indeterminacy and instability are problems but quite another to argue that anything goes, that anything can be claimed and said, and that powerful states can simply pick and choose the rules that suit their purposes.

In the first place, it is in the very nature of norms that there will always be questions as to which norms are relevant and how they should be interpreted. The critical issue is the degree to which interpretation reflects shared practices. The theoretical point is well made by Kratochwil: ‘As a practice, a rule not only tells me how to proceed in a situation that I have never faced before, it is also governed by certain conventions of the community of which I am part. To this extent, my interpretations of a rule as well as my uses of words are monitored and reinforced by a group of competent speakers’ (Kratochwil, 2000: 52). The institutional point is that international law is built around an interpretative community of lawyers (and others who argue in and through law) who inhabit a more or less unified conceptual universe and who share a common set of discursive practices. It is by dint of this that questions relating to the determinacy of rules or the meaning of contested concepts can be rationally debated, if never fully resolved. Second, international law has to be understood as an integrated institution and as an interconnected normative system in which historical development and the evolution of specific legal doctrines or concepts over time play a crucial role. Thus the content of a particular norm and the degree of obligation that attaches to it is related to its place within this broader normative order. Doubtless there are many cases where competing principles animate controversy. Similarly, many rules are capable of widely differing interpretations. But the integrity of law sets limits to the range and influence of eligible principles, and to the range of legitimate interpretations. If indeed ‘organized hypocrisy’ is a useful term, we still need to reflect on the crucial fact that it is just that: organized hypocrisy.

Something similar is true of moral and normative argument. All human societies rely on historical stories about themselves to legitimize notions of where they are and where they might be going. Over time, understandings of normative problems and categories of normative arguments become organized into intelligible patterns, traditions, or ideologies. Thus to take the case of war: the continual involvement of individuals and societies in war and conflict, the moral and political necessity of trying to make sense of what war involves, and the limited range of plausible arguments have led over time to the creation of intelligible patterns, traditions and ideologies. As Michael Walzer puts it: ‘Reiterated over time, our arguments and judgements shape what I want to call the moral reality of war–that is, all those experiences of which moral language is descriptive or within which it is necessarily employed’ (Walzer, 1977: 15).

How Norms Affect Outcomes

For neorealists norms do not matter in themselves but only to the extent that they reflect, or are backed by, the power of powerful actors. For rationalists, norms and institutions matter to the degree that they affect actor strategies (but not their underlying preferences) by reducing transaction costs, by identifying focal points for coordinated behaviour, and by providing frameworks for productive issue-linkage. Moving beyond rationalism, cognitive approaches highlight the role of knowledge, and especially of scientific or technical knowledge, in shifting state understandings of interest in ways that foster cooperation.

Attempts to go beyond these kinds of approaches (for a survey see Hasenclever et al., 1997) rely on two well-established steps. The first is to challenge the rationalist account of human choice and human agency. Rationalists like to claim that they adopt an ‘actor-centred’ theory and that, if this is rejected, then one is necessarily committed to a view of human agents blindly following internalized norms. Much recent writing within both political science and social theory has relied on this stark and, to many, implausible dichotomy (March and Olson, 1998). Thus Elster constrasts instrumentally rational action that is hypersentive to consequences with norms understood as internalized Kantian imperatives (‘blind, compulsive, mechanical or even unconscious’; Elster, 1989). Krasner distinguishes between consequential action on the one hand and ‘taken for granted’ ‘deeply embedded’ ‘internalized’ norms on the other (Krasner, 1999). And Keohane has written on the differences between instrumental and normative ‘optics’ (Keohane, 1997).

As against this, the critics of rationalism highlight the complexity of human choice and the inevitability of deliberation, of moral conflict and of moral purpose (O’Neill, 2000). As Jackson puts it: ‘People are not automated or mechanical things. People are human beings who make choices, and whose policies and actions must be justified to other people’ (Jackson, 2000: 8). It was, after all, for this reason that Carr believed that political realism had to accept the importance of utopian ideas (Carr, 1939). Or, as Robert Musil (1995) puts it: ‘If there is a sense of reality, then there must also be a sense of possibility.’ The critics also point to the difficulties of any clear-cut distinction between a rational logic of consequences and a norm-following logic of appropriateness. How we calculate consequences is often far from obvious and not easily separable from our understanding of legal or moral norms. Think, for example, of how states sought to calculate the consequences of alternative courses of action in the Kosovo bombing campaign and how these calculations were deeply shaped by the existing legal and moral normative framework. Moreover, time and process are especially important in understanding how logics of consequences and of appropriateness interrelate. At any point in time it may indeed to be helpful to think of actors making choices between consequentialist calculations and normative appropriateness. But over time the obviousness of certain sorts of norms (for example against slavery or military conquest) becomes such an accepted part of the international political landscape that it becomes part of how actors routinely calculate consequences.

The second well-established challenge to rationalism is to question the rationalist account of how norms are related to causality. The issues are complex and murky but the basic point is a simple one–that there are many routes to explaining social action.

Thus, contrary to Popper and the predominant epistemology, ‘explaining’ does not seem to involve simply the procedure of the ‘subsuming’ of a single case under a general law, but comprises a rather heterogeneous set of procedures by which we try to understand actions and events. Explaining often means providing a context, such as when we make a series of actions and events part of a wider narrative. However, explaining an action might also involve us in elaborations and justifications of the choices made, or of our reasons for choosing certain beginnings and endings. (Kratochwil, 2000: 64)

Or, in the earlier classic statement with John Ruggie:

Norms may ‘guide’ behaviour, they may ‘inspire’ behaviour, they may ‘rationalize’ or ‘justify’ behaviour, they may express ‘mutual expectations’ about behaviour, or they may be ignored. But they do not effect cause in the sense that a bullet through the heart causes death or an uncontrolled surge in the money supply causes price inflation. (Kratochwil and Ruggie, 1986)

If these two points have been frequently made, there are two further areas on which much work remains to be done. Both concern the way in which normative structures (as opposed to individual norms) may affect political outcomes. The first relates to constitutive norms. ‘Constitutive rules define the set of practices that make up any particular consciously organized social activity–that is to say, they specify what counts as that activity’ (Ruggie, 1998: 22). The distinction between regulatory norms and constitutive norms has a long history within philosophy and jurisprudence (Rawls, 1955; Searle, 1995)–although it is by no means universally accepted (for example by Raz, [1975] 1999: 108–11). Constructivist writing has laid great emphasis on the importance of constitutive rules and the theoretical and general debate has been intense and often illuminating (see especially, Kratochwil, 1989; Onuf, 1989).

Yet many puzzles remain. What actually are the most important constitutive rules in international relations and what are the criteria for assessing their importance? How are we to understand the normative force of constitutive rules? How do believers in the importance of constitutive rules answer the sceptical response that powerful actors can find new constitutive norms to suit their interests (as in the case of revising sovereignty in the post-Cold War world)? Answers to these questions may well involve sharper distinctions between different sorts of rules that can loosely be placed under the ‘constitutive’ banner: transactional or facilitative norms as, for example, in the law of treaties or diplomatic law; ‘power-conferring norms’ (to use Raz’s term) which determine the capacity to participate in political or legal processes; membership norms that govern the participation of actors in individual institutions or in international society more generally; and finally those international norms which both reflect and seek to shape the domestic constitution of major actors–their corporate as opposed to social identities to use Wendt’s rather problematic distinction (Wendt, 1999: ch. 5). This last category is especially important. Constitutive norms are powerful not only (or indeed primarily) because they formally create or enable certain kinds of activity; but rather because they foster group identification and therefore allow for the coordination of great social power. It is for this reason that national self-determination is the most important constitutive norm of the modern era.

The second set of on-going puzzles concerns the many different mechanisms by which norms are diffused across the international system and internalized by actors. It is easy to suggest in broad terms that past institutional choices affect the present choice of norms and the framework within which strategic interaction takes place. But, as Krasner argues with his usual clarity of purpose, it is much harder to establish how institutions become embedded: ‘that is, dictate actual behaviour and endure over time or across different environmental conditions’ (Krasner, 1999: 226). Within the rationalist agenda, the systemic dimension is incorporated through hard notions of path dependence (and maybe also of sunk costs) that ‘lock’ actors into patterns of behaviour. Patterns are followed because the costs of divergence are high. Within a sociological account the explanatory work will have to be done by some version of individual socialization.

Whether individual-level socialization is important is an empirical question and not one that can be settled by theoretical fiat. But, very importantly, it is by no means the only mechanism by which actors may come to ‘internalize’ norms that originate elsewhere in the system. Three other mechanisms deserve brief mention. The first can be labelled discursive enmeshment and highlights the importance of argumentation, deliberation and persuasion. Words are not always cheap. The reasons for making normative claims or framing interests in normative language may well be purely instrumental. Such a course of action legitimizes and makes universal claims that would otherwise seem partial and particular. It may also reflect straightforward calculations of interest (talking the language of human rights in order to secure aid or investment or, as with China, its acceptance as a legitimate Great Power; Foot, 2000). But entering into a particular debate and accepting particular principles, ideas and arguments shapes and constrains the sorts of arguments that can be made in the future and provides institutional and normative platforms for different forms of political mobilization. On the importance of legitimacy see Brunneé and Toope, 2000; Franck, 1990; Hurd, 1999.

A second mechanism concerns bureaucratic enmeshment. Individual leaders may well remain completely unsocialized into international norms. And yet they may find that external norms have been incorporated into bureaucratic structures in ways that preclude or raise the costs of certain sorts of policies. There may, for example, be very powerful situational pressures to deviate from the laws of war. Yet politicians and generals may find themselves effectively constrained by the incorporation of these rules into bureaucratic operating procedures and national military codes. Similarly, national and international bureaucratic procedures in the field of human rights (to compile reports or to attend review meetings) help explain the continuity of human rights policies despite the changing attitudes and commitment of individual leaders. Third, there is legal internalization. As should by now be well known, international legal rules cannot be understood in any straightforward sense as simply ‘international.’ At one level, rule-governed behaviour does reflect the dynamics of inter-state politics and of inter-state institutions. But that inter-state dynamic is very often heavily influenced by the degree to which international rules have been incorporated into national legal systems and also by transnational groups and pressures that act across the domestic/international divide (Goldstein et al., 2001).

How Norms Change

Change matters. As Hollis put it: ‘Norms are no less effective for being fluid and no less real for being negotiable. Both ideally and actually the stuff which binds societies is more like mastic than cement’ (Hollis, 1991: 13). But how are we to understand normative change? Realists and neorealists argue, often correctly, that changes in the normative structure are closely bound up with power and the distribution of power–within the state system but also within the global economy and transnational civil society. Rationalist institutionalists are inclined to see institutions as responses to the increasingly serious collective action problems generated by growing societal, ecological and economic interdependence. Such interdependence creates huge scope for joint gains and hence a powerful ‘demand’ for institutions (although their supply is an altogether more difficult matter).

More recently, liberal pluralists have emphasized the role of transnational civil society. Transnational civil society refers to those self-organized intermediary groups that are relatively independent of both public authorities and private economic actors; that are capable of taking collective action in pursuit of their interests or values; and that act across state borders (Scholte, 2000). Their role in norm formation and norm development is important (although by no means obviously greater than, say, in the late nineteenth and early twentieth centuries): first, in the formal process of norm creation, standard-setting and norm development; second, in the detailed functioning of many international institutions and in the processes of implementation and compliance; and third in direct participation in many governance activities (disbursing an increasing proportion of official aid, engaging in large-scale humanitarian relief; leading efforts at promoting democracy or post-conflict social and political reconstruction). In all of these areas the analytical focus has been on transnational networks–for example, knowledge-based networks of economists, lawyers, or scientists; or transnational advocacy networks which act as channels for flows of money and material resources but, more critically, of information, ideas and values (Finnemore and Sikkink, 1998; Keck and Sikkink, 1998).

Four points deserve to be highlighted. First, as the density and complexity of the international legal system increases and as globalization opens up new channels of transnational political action, so the process of norm creation becomes harder for even powerful states to control. Thus transnational NGO networks have carved out important roles and relatively weak states have been able to promote new and often far-reaching legal rules and institutions (as with the International Criminal Court). In part this success has depended on material or ideational resources. But it also reflects the ability of non-state groups to exploit institutional platforms. State action may be shaped by NGO lobbying but it is often state action that is crucial in fostering the emergence of civil society in the first place and in providing the institutional and normative framework that enables it to flourish. In addition, success will often depend on the capacity to build on already established patterns of legal argument and to import and adapt already accepted principles from other areas–in other words, to exploit internal logics of norm development.

All of this points to a broader change. In the traditional legal order, dominant norms were created by states and depended directly on the consent of states. Over recent decades, the process of norm creation has been opened to a wider range of actors, both states and non-state groups, and there is an easing of the degree to which states can only be bound by those rules to which they have given their explicit consent. Lest this appear too cosy and liberal, two important caveats are in order. First, civil society is an arena of politics like any other in which the good and thoroughly awful coexist and in which outcomes may be just as subject to direct manipulation by powerful actors as the world of inter-state politics. And second, the degree of effective pluralism varies greatly across issue areas. It is most apparent in fields such as human rights and environment but far less so in the hierarchically structured institutions and institutionalized practices that shape the global economy.

Second, there is the inherent tendency for all normative systems (especially reasonably well-institutionalized judicial systems) to expand from within and to enmesh actors within certain patterns of discourse, reasoning and argumentation. As Stone Sweet put it: ‘norms … develop in path-dependent, self-reinforcing ways, one mechanism of which is the ubiquity, and naturalness, of normative reasoning itself. Normative systems are inherently expansionary to the extent to which they enable people to reason from one situation to another, by way of analogy’ (Stone Sweet, 1999: 157). In complex legal institutions, such as the ECJ or the WTO dispute settlement procedures, norm development does not simply reflect periodic bargains amongst states. It often takes place internally through the practices of the institutions themselves: filling in gaps in treaties, developing answers to new problems, establishing precedents (even where precedent is not formally admissible).

Third, institutions act as platforms for normative debate, for the mobilization of concern, and for debating and revising ideas about how international society should be organized. However much social scientists insist on analysing international institutions solely in terms of the provision of international public goods, normative issues cannot be kept out of the actual practices of those institutions. It is, for example, politically and theoretically significant that this should be true even of economic institutions that define their own purposes in instrumental terms. Thus arguments about the effectiveness and efficiency of the World Bank or the IMF are increasingly having to be made in the language of legitimacy or even of moral purpose.

Finally, normative change reflects changes in the organization of domestic society and in the powerful transnational ideological forces that have shaped those changes. Thus the legitimacy of governments (democratic and authoritarian) has come to depend on their capacity to meet a vastly increased range of needs, claims and demands. In part this has involved increased expectations of the role of the state in economic management (something that remains substantially true even in an era of deregulation, privatization and globalization). In part it reflects changed notions of political legitimacy and broadened understandings of self-determination, of human rights and of citizenship rights. In addition, although it may be true, as realists tell us, that the international system tames and socializes revolutionary regimes, it is also true that each of the great social revolutions of the modern era has left an indelible mark on the dominant norms of international society (Armstrong, 1993; Halliday, 1999). In one sense this may be obvious. But it is obscured by the very common tendency in IR to distinguish too sharply between ‘thick’ domestic norms and ‘thin’ international norms. Many international norms (national self-determination, economic liberalism, sustainable development) are powerful precisely because of the way in which they relate to the transnational structures within which all states are embedded and to the broad social forces that have transformed the character of states and altered the dynamics of the state system.

Much work has been done on non-state groups (both firms and NGOs) either as lobbyists within individual states or as participants directly within international regimes and institutions. But much more needs to be done, for example, on the emergence of convergent state norms and forms of ‘world law’ that develop not through explicit or formalized inter-state bargaining but rather through regulatory competition, pressure from international bodies or mimesis (Meyer et al., 1997); or the emergence of private authority structures that exist largely autonomous from the framework of both municipal and international law: private systems of arbitration and dispute settlement, privatized rule-production resulting from technical standardization, internal regulations within transnational firms; and private regimes governing particular sectors of the global economy (Cutler et al., 1999; Ronit and Schneider, 1999; Teubner, 1997).

Norms and Cultural Diversity

What Sen calls ‘the empirical fact of pervasive human diversity’ (Sen, 1992: xi) has long been seen as a major issue confronting normative theorists: What value should be accorded to the traditions and practices of particular human communities? How might obligations owed to such communities be balanced with obligations owed to humankind in general? Cultural diversity has also long been a central problem for those who ask: How broad and how deep is international society? How strong is the consensus on the nature of a desirable world order and the means by which it might be achieved? A deep-rooted tradition of Western thought has associated social order and political institutions (both domestic and international) with the existence of common values and with the shared culture within which those values are embedded. It was, for example, axiomatic to the classical European writers on international society that the society of states was in some way related to shared values (Bull, [1977] 1985). Some stressed the shared international political or diplomatic culture of élites; others emphasized shared societal values whether those of Christianity, of Europe or of civilization (Watson, 1992; Wight, 1977). On this view, the interests and identities of actors are shaped by particular histories and cultures, as well as by historically constituted relationships and ongoing processes of interaction with other states. It is perhaps for this reason that English School approaches resonate so strongly outside the world of Western political science. But certainly for Bull (if less so for Wight), the role of values is an empirical question to be investigated, not an analytic assumption. Although shared values may have been important at particular times, Bull did not believe that international society necessarily rested on the existence of a common value system as, for example, suggested by Krasner (Krasner, 1999: 47; see discussion in Alderson and Hurrell, 2000: 6–7 and 69–70).

This way of thinking about the role of culture in international relations has attracted significant criticism. There are many problems with the view (often associated with Parsonian sociology) that simply equates order with the existence of a common value system and with normative consensus, and which sees norms in terms of the social functions that they serve. Value systems, after all, do not just happen. They are created by social agents for particular purposes and they are maintained because it will often pay people in some way to ensure that they are (Barry, 1970: 75–98; and Coleman, 1990). Functionalist accounts of social norms run the risk of circularity, of post hoc reasoning, and of reducing human agents to automata blindly following internalized norms. In addition, if cooperation does not necessarily follow from shared values, conflict need not necessarily be the result of their absence (see Cederman in this volume and Hardin, 1995). More generally in political science, after a brief flourishing in the 1950s, concern with culture as an analytical concept faded away. (The emphasis on strategic culture formed one partial exception, see Booth, 1979 and Snyder, 1977).

The 1990s, however, witnessed a renewed interest in culture. Three aspects can be highlighted. In the first place, changes in the international system seemed to underline the renewed importance of cultural factors. Thus trends towards globalization appeared to many to make cultural differences more salient and indeed to act as a spur to the reassertion of cultural traditionalism and particularism. The resurgence of nationalism and ethnic conflict seemed to push questions of identity and culture centre-stage and to focus attention on the construction of boundaries and the production of meaning (Lapid and Kratochwil, 1996). And the increasing emphasis placed by Western states on human rights and democracy gave new life to old arguments about cultural relativism and ‘Asian values’ (Bauer and Bell, 1999; Dunne and Wheeler, 1999).

Second, the concern with culture formed part of a more general interest in the role of ideas and non-material forces in explaining social phenomena (Goldstein and Keohane, 1993). Theorists argued, for example, that the models that dominated mainstream IR were incapable of dealing with the complex dynamics of integration and disintegration that were producing many different political forms and political affiliations, both above and below the state. Moreover, those models were themselves socially and culturally constructed so that, for example, even realism could be traced to culturally specific roots (Johnston, 1995). It was also argued that rationality is not static and that experience and shifting identity can change or modify behaviour even where the strategic environment remains constant (as in the cases of Germany and Japan; see Berger, 1996; Katzenstein, 1996a). Equally, actors may be culturally unaware of the range of potential options open to them or may be culturally inhibited from adopting particular paths (Katzenstein, 1996b). Although the range of arguments has been very wide, culture was seen to matter in terms of its effects on the definition of interests; the way that it shapes links between interests and identity; and its impact on normative structures such as identity (Kowert and Legro, 1996).

It is not possible to provide here a full analysis of these claims and the interested reader is referred to other chapters in this volume. But a couple of points need to be made. First, within this literature it is not always easy to pin down what is meant by culture. The terms ‘culture’ and ‘identity’ have often been used interchangeably and with a significant degree of imprecision. And second, the links between culture and language remain underdeveloped. It is certainly true that linguistic constructivists have added a very great deal to our understandings of how social meanings and intersubjective structures are produced and reproduced (Fierke, 1998; Kratochwil, 1989; Onuf, 1989). But it is not just language in general that matters; it is also the specificities of particular languages and linguistic systems. Here it is not necessary to accept determinist claims about the way in which different languages map the world differently or predetermine certain modes of observation and interpretation. The point is rather to stress the complex ways in which language affects the way people think and act; the role of particular languages in both symbolizing culture and producing culture; the difficulties of translating ideas across linguistic boundaries (Steiner, 1992; Wierbicka, 1997); and the very different value connotations of apparently similiar words (for examples of how this matters politically see Cohen, 2000; Inoue, 1991; Komatsu, 1999).

The third dimension of the return of culture has to do with broad characterizations of the post-Cold War international order. In the influential view of Samuel Huntingdon the world is fundamentally divided along cultural or civilizational lines and it is these divisions that will determine the shape of global politics in the coming century (Harrison and Huntingdon, 2000; Huntingdon, 1997). Minimizing conflict for Huntingdon involves accepting the reality of civilizational difference, forsaking vain attempts to promote Western liberal values globally, and, implicitly at least, returning to a pluralist order of coexistence, based, in the post-Cold War world, not on balanced power but on hegemonic self-restraint.

The problems with this image of clashing civilizations are well known. It is difficult to identify clear territorial boundaries to civilizations, and harder still to imbue them with the capacity to act as a coherent unit. It places far too much in the undifferentiated category of ‘civilization,’ reifying and essentializing cultures and downplaying the multiplicity of trends, conflicts and contradictions within any particular cultural tradition. Cultures are not closed systems of essentialist values and it is difficult to see the world as consisting of a limited number of cultures each with its own indestructible and immutable core, particularly given the expansion of channels, of pressures and of agents through which norms are diffused across the world. Of course there are cultures and communities that mil-itantly refuse to be globalized. But there are also many examples of powerful normative diffusion; and still more cases in which deep-rooted, inherited local traditions come up against powerful globalizing pressures. The result is neither fusion nor fragmentation but rather continued eclectic and syncretic diversity.

And yet despite these difficulties, human diversity and value conflict remain important. All communities and polities have to find ways of dealing with diversity and value conflict. Conflict is, after all, intrinsic to all morality and even within a single value system conflicts arise: how different principles are to be related to one another; how shared principles are to be applied to the facts of a particular case. For international society the problem has always run much deeper. The creation of any kind of universal society of states or any other kind of world society has to face up to the existence of fundamental differences in religion, social organization, culture and moral outlook. Diversity is a basic and common feature of humanity. The clash of moral, national and religious loyalties is not the result of ignorance or irrationality but rather reflects the plurality of values by which all political arrangements and notions of the good life are to be judged.

It is certainly true that great care needs to be taken as to what should be placed under the heading of ‘culture’: how far the problems of cultural diversity reflect differences in national histories, in social and economic circumstances and conditions, and in political contexts, rather than in culture per se. And yet the question of how to mediate amongst opposing moral and political worlds remains one of the greatest and most difficult challenges to normative discussion of what might constitute a feasible and desirable international order.

One aspect of this problem concerned the transition from a European to a global international society and the incorporation on at least slightly more equal terms of a much wider range of cultural, religious and social traditions (Bull and Watson, 1984). Here the optimists can point to the alacrity with which those Third World leaders who had so vehemently denounced imperialist international law came to appreciate its benefits (as had French, Soviet and Chinese revolutionary leaders before them). Being mostly weak it made great sense to buy into the political advantages that sovereign statehood provided and such protection as international legal rules might afford. Moreover, as Robert Jackson and Christopher Clapham have shown, external support and access to the instruments of juridical statehood played major roles in the battle for domestic survival (Clapham, 1996; Jackson and Rosberg, 1982a, 1982b).

And yet, as the international legal order moves in more solidarist and transnational directions and as the ‘waterline of sovereignty’ is lowered, so the political salience of societal difference rises. International rules relating to human rights, to the rights of peoples and minorities, to an expanding range of economic and environmental issues impinge very deeply on the domestic organization of society. Divergent values also become more salient as the legal order moves down from high-minded sloganizing towards detailed and extremely intrusive operational rules in each of these areas and towards stronger means of implementation. And divergent values come into sharper focus as inequalities of power grow more extreme and as the debate on ‘Western values’ meshes with debates on US hegemony.

Finally, the stakes have risen substantially. The capacity to opt out of what was previously a largely consent-based legal system has declined. Refusal to accept either non-derogable core legal norms or those norms that are particularly prized by powerful states run the risk of being branded a ‘rogue’ or ‘pariah’ state. And the post-Cold War period witnessed a revival of formal or informal criteria by which full membership of international society is to be judged. All of these developments have magnified a fundamental tension in the character of the international legal order: between sets of rules that seek to mediate amongst different values and those that seek to promote and enforce a single set of universal values. The political problems are, of course, still more acute when the purportedly universal values are those of the most powerful states.

Values also enter the picture because of the role that they play in explaining cooperation. One aspect has to do with the degree to which successful cooperation may depend on some prior sense of community. Rationalist models of cooperation may indeed explain how cooperation is possible once the parties have come to believe that they form part of a shared project or community in which there is a common interest that can be furthered by cooperative behaviour. They assume the accepted legitimacy of the players; a common language for bargaining; a shared perception of potential gains; and some mechanism for at least potentially securing contracting. Once there is a common identification of, and commitment to, some kind of moral community (however minimalist in character) within which perceptions of potential common interest can emerge, then there may indeed be prudential reasons for the players collectively to cooperate. But rational prudence alone cannot explain the initiation of the game and why each player individually might choose to begin to cooperate. Rationalist approaches neglect the factors that explain how and why contracting is possible in the first place and the potential barriers that can block the emergence of any such a shared project.

The role of values in rational action is complex and difficult. As Michael Hechter argues, values are ex ante reasons for taking action, or relatively general and durable internal criteria for taking and evaluating action (Hechter, 1994). Whereas rationality is conditional and future oriented, values are the product of history. They are what differentiates one society from another and one international system from another. They might indeed be incommensurable, or, more often, they may shape the meanings of rational action in ways that can either help or hinder cooperative endeavours. Offer has, for example, demonstrated how values and rational action can be combined in his analysis of how codes and notions of honour pervaded the cultural script of pre-1914 Europe and how this script in turn shaped understandings of how the game of power politics should be played and what constituted ‘rational action’ (Offer, 1995: 213–41).

In some areas, such as human rights, the potential importance of differing societal and cultural values has been extensively debated and analysed. But the relevance and frequent intractability of these problems extends well beyond human rights. Thus the politics of security is not only driven by problems of trust and credible contracting but also by deep disputes as to which values are to be incorporated into understandings of security and as to whose security is to be promoted (states? nations? regimes? individuals?). Equally, liberal governance approaches to global environmental negotiations can easily overlook the absence of a shared cultural or cognitive script that allows the largely rhetorical consensus value of ‘sustainability’ to be translated into stable and effective operational rules. Or again, arguments about international trade in GMOs are not about how scientific standards of risk assessment might be applied but rather about the intrinsic validity of those standards and the criteria by which international institutions might decide upon the permissible sphere of legitimate societal difference. Culture does not necessarily matter but difference and diversity do. Understandings of world order vary enormously from one part of the world to another, reflecting differences in national and regional histories, in social and economic circumstances and conditions, and in political contexts and trajectories. Why is international order so difficult to achieve and why are common interests so hard to capture? First, because of the intractability of the problem of power in international relations. But second, because of the persistent difficulty of accommodating human diversity and the continuity of pervasive value conflict.

Conclusion

What is distinctive about IR is its concern with what makes the world hang together, as Ruggie puts it so nicely (Ruggie, 1998). It is a zone of academic endeavour united not by particular methods or by the insights of a single academic discipline, but rather by a concern with certain central questions. The appropriate methods and disciplinary approaches follow from these central questions and from the particular puzzles and research questions to which they give rise, and not the other way around. This chapter has focused on a series of questions related to the role of ethics and norms in international relations: How have ideas on international ethics and on the nature of a just world order evolved? How has the normative structure of international society developed and how do normative structures influence what norms are, how they operate and how they change? And how has the issue of culture and the role of cultural diversity affected the possibility of international order and conceptions of what a just world order should be like? This chapter has identified some of the major lines of enquiry that have emerged in answer to these complex issues. In all of these areas we can look back to a range of rich traditions of thought in which links between political science and other disciplines (especially international law, history and political theory) have always been apparent; and where those links are likely to gain ever greater importance in the future.