Title Menu - The International Journal of Peace Studies.  
Return to Table of Contents Return to I J P S Home

IDENTITY AND THE OUTBREAK OF WAR:
OR WHY THE COPENHAGEN SCHOOL OF SECURITY STUDIES SHOULD INCLUDE THE IDEA OF 'VIOLISATION' IN ITS FRAMEWORK OF ANALYSIS

Iver B. Neumann

Introduction: Politics and Understanding

Let me be explicit about what kind of understanding I am trying to bring to the concept of the political. This paper is mainly about understanding the outbreak of war, not about explaining the causes of war. Causes beg for causal explanations. Most studies of the role of identity in politics, however, tend to look at the understanding of war, the imagery of war, the narrativity of war, the symbolic technologies of war and the like. That is, they do not necessarily offer causal explanations which are cast in terms of the outbreak of war as a 'dependent variable' at all, but simply go about the task of theorising the political in other ways. One could perhaps argue with Roxanne Lynn Doty (1996) that, instead of asking 'why questions', they ask 'how possible-questions' about the outbreak of wars. In peace and conflict studies, such an approach may list a long heritage (Wallensteen, 1994).

Another way of highlighting this difference is to start out from the two very different views of the concept of the political which infuse these two different approaches. If, as mainstream political science tends to take for granted, the political is a question of who gets what when, then war becomes one of a number of mechanisms by means of which a certain distribution can be realised. Crucially, the 'whos' which are supposed to get the something at a particular point in time are treated as fixed entities, which are not themselves changed by the distributional games in which they participate, war being one of them. As Sir Michel Howard (1983: 7, 1) puts it, 'War has been throughout history a normal way of conducting disputes between political groups. [...Now,] war is only a particular kind of conflict between a particular category of social groups, sovereign states.' Stephen van Evera (1994) provides an example of how questions of identity may be brought to the analysis of war in this spirit. Identities are treated as fixed, externalised and hence non-negotiable.

The approaches which are engaged below, on the other hand, treat collective identity as being always in a state of formation, as ever-lasting negotiations about who is who - how that who comes about, how individuals become party to it and how it is reproduced over time. If viewed in this way, a causal approach is not a felicitous one, since it actually excludes the question of identity formation ipso facto. Thus, it is not a coincidence that the causal approach to the outbreak of war has neglected the question of identity. On the contrary, the very way in which it phrases the question actually excludes the possibility of reflexive analysis, and hence of the role constitutive political labour performed by identity.

Instead of seeing the political simply as a question of how already fixed actors decide between themselves who gets what when, one may of course see it as an ongoing negotiation of who 'we' are. Since a 'we' is unthinkable outside relations to a set of 'theys', the political, understood as the question of who 'we' are, is a question of separating us from them. To use Carl Schmitt's formulation, it is a question of separating friend from enemy. Let me draw on a recent identity-inspired elaboration of Chantal Mouffe's work in order to clarify this point.

In order to make her argument, Mouffe starts off by evoking a distinction between the political, which describes the ineradicable and ever-changing antagonism and hostility which characterises human interaction, and politics, which, taking note of the permanent antagonism characterising the political, seeks to establish a certain order and to organize human co-existence. She then lambasts liberal politics for misunderstanding the very nature of the political: 'This view, which attempts to keep together the two meanings encompassed by the terms "politics" - that of "polemos" and that of "polis" - is totally foreign to liberal thought; that, incidentally, is the reason why liberal thought is powerless in the face of antagonism.' (Mouffe 1994: 108). Politics is constituted by its outside (its extérieur constitutif), and inevitably bears the marks of its own exclusions. In standard poststructural fashion, Mouffe then proceeds to suggest that the way to alleviate the impact of this inevitable exclusion is to celebrate how, since the 'we' is constituted by its outside, that 'we' must also somehow be that outside:

On a general philosophical level, it is obvious that if the constitutive outside is present inside every objectivity as its always real possibility, then the interior itself is something purely contingent, which reveals the structure of the mere possibility of every objective order. This questions every essentialist conception of identity and forecloses every attempt conclusively to define identity or objectivity. Inasmuch as objectivity always depends on an absent otherness, it is necessarily always echoed and contaminated by this otherness. Identity cannot, therefore, belong to one person alone, and no one belongs to a single identity. We would go further, and argue that not only are there no 'natural' or 'original' identities, since every identity is the result of a continuing process, but that this process itself must be seen as one of permanent hybridization and nomadization. Identity is, in effect, the result of a multitude of interactions that take place inside a space whose outlines are not clearly defined (Mouffe, 1994: 109-110).

Mouffe draws two conclusions for political practice, one about the unwanted and in any case impossible forging of a European self, and one about democratic politics in general. On the former, she writes that

Contrary to what is popularly believed, a 'European' identity, conceived as a homogeneous identity which could replace all other identifications and allegiances, will not be able to solve our problems. On the contrary, if we think of it in terms of 'aporia', of double negative, as an 'experience of the impossible', to use Derrida's words from his L'Autre cap [The Other Heading], then the notion of a European identity could be a catalyst for a promising process, not unlike what Merleau-Ponty called 'lateral universalism', which implies that it is inscribed in respect for diversity. If we conceive of this European identity as a 'difference to oneself', as 'one's own culture as someone else's culture', then we are in effect envisaging an identity that accommodates otherness, that demonstrates the porosity of frontiers, and opens up towards that 'exterior' which makes it possible. By accepting that only hybridity creates us as separate entities, it affirms and upholds the nomadic character of every identity (Mouffe, 1994: 111).

Where democratic politics is concerned, she stresses the need to maintain a politics of antagonism centered on certain traditional identity nexuses in order to foreclose the possibility that other identities may be inscribed with paramount political meaning and thus made the defining foci of essentialist identity politics:

Unclear dividing lines block the creation of democratic political identities and fuel the disenchantment with traditional political parties. Thus they prepare the ground for various forms of populist and anti-liberal movements that target nationalist, religious and ethnic divides. When the agonistic dynamism of the pluralist system is unable to unfold because of a shortage of democratic identities with which one can identify, there is a risk that this will multiply confrontations over essentialist identities and non-negotiable moral values (Mouffe, 1994: 109).

However, one has to add a dimension which has been at the core of Mouffe's earlier work (Laclau & Mouffe, 1985), namely that exactly because class distinctions (in the old Marxian, not Bourdieuan sense) are evaporating as possible identities, political space is opening up for the plethora of social identities around which social movements have congealed over the last 25 years or so. And it is exactly as a reaction to this mushrooming of the number of available identities, which cannot easily be strung together in one overarching narrative of self, that there is a rush to defend the story of self which revolves around the nation. In these circumstances, the permanent antagonism which characterises human interaction may activate the ever-present possibility of violence rather than the transformation of antagonism into agonism which Mouffe sees as desirable.

Linking the literatures on identity and the outbreak of wars, then, must mean to reflect on how war is implicated in establishing friend and enemy. Perhaps the three first tasks to be tackled are a) going to war as a means of including and excluding subjects of world politics (still a statist concern); b) outbreak of war as a way of re-presenting an identity; c) how certain symbolic economies work to produce war as an outcome of ever more sharply defined friend/enemy relations. These key questions of identity are of course overlapping, but the first comes at war from an intentional angle, the second from a psychoanalytically informed angle, and the third from a functional angle.

Inclusion and Exclusion: Ringmar

First, how war may be used as a way of including and excluding certain human collectives as recognised players in world politics. We are talking here of studying intended, instrumental war making. The literatures on secession, nation building and the expansion of international society are of particular relevance here. The work which most directly tackles this job, however, comes at it from a narrativist angle. The major example of such work is Erik Ringmar's book on Sweden's entry into the Thirty Years' War in 1630.

Ringmar's starting point is a feeling that there is something deeply constrained and hence alarming about explanations which postulate a number of interests, point to a specific outcome like the initiation of a specific war, and then sees the explanatory work as being about picking out the most relevant interests. Let me quote Ringmar's specification of the problem at length:

We cannot merely be satisfied with the fact that a certain explanation explains something, but we must also find out how this explanation is achieved. We need to know which causal variables go with which others, why certain factors are brought in and not others, and under what circumstances certain assumptions can be expected to hold. By asking these more basic questions we are, however, no longer engaging in a historical or a scientific investigation of facts, but instead in an investigation of how those interpretations are constructed which the existence of facts presupposes. As I will argue, it is only through an investigation of the preconditions which guided the work of previous scholars that we can criticise those preconceptions and come up with new, alternative, ways in which to organise our data. As a pre-empirical investigation of this kind will show, historians and social scientists - despite their many differences - generally subscribe to one and the same theory of action: the notion that human beings are 'rational' and that their actions can be explained by reference to the 'interests' of the person or group who perform them (Ringmar, 1996: 12).

Having thus taken an antifoundationalist (no givens) and narrativist (there are only stories) position regarding the possibility of writing history, it comes as no surprise that Ringmar finds fault with the efforts made by the social sciences when it comes to explaining action, investigated specifically as the action of going to war. Ringmar chops up the efforts made by the social science guild in two. First, there are those who take their cue from 'the logic of the situation' - some general logic regarding the relative power of states is mustered in order to give a structural explanation. The examples given are Choucri and North's lateral pressure model, Gilpin's theory of hegemonic wars and Doran's power cycle theory, but it seems a fair reading that these are simply examples of most pre-1960s social scholarship as well as most non-behavioral post-1960s scholarship on the causes of war and indeed other action as well. This body of theorising is dismissed with the following words:

The fundamental problem with theories of situational logic is [...] that the entire causal relationship between material factors and human actions is misconstrued. We do not, as these theories imply, live in a material world which 'presents' us with various more or less constraining options; instead we live in a material world which we interpret, and it is on the basis of these interpretations that we present various options to ourselves. Hence it follows that a mere description of material factors will never tell us much about what actions a person will undertake. What an outside observer should study are not material factors, but instead the interpretations given to material factors; the way in which human beings make sense of their world (Ringmar, 1996: 37).

That is to say, the theories under his gaze are found wanting because they jump directly from general conditions to specific outcomes without considering the crucial role of the in-between factor of human intentionality.

Enters the other main chunk of social theorising on why wars break out, namely the fruits of the so-called behavioural revolution in the social sciences which culminated in the 1960s. By focusing not on material factors, but on decision-making processes, Ringmar acknowledges the 'vast improvement' inherent in actually introducing humans into the explanations. However, these theories are also found wanting because they also somehow look at a backdrop to the actual decision; if not the logic of the situation, then on the logic of decision-making as process. This, Ringmar insists, still amounts to ducking the issue:

The focus is always on the 'hardware' of the decision-making process, as it were, not on matters of 'software': it is not what people think about their worlds - how they give it meaning - which concerns them, but instead how they go about making decisions. [...] The theory presupposes the prior existence of a meaningful world, but it remains agnostic regarding how this world was created (Ringmar 1996: 39).

So, taking the lot of historical and social scientific theories together, Ringmar concludes that they may tell us things about what the general background to a decision was and how a decision was reached, but what they cannot do is tell us why that decision was more meaningful to the decision-makers than others.

Ringmar's second crucial move is to introduce a setting to the stories, that is, other story-telling entities. These 'others' about whom the self tells stories and who tell stories about the self, is thus a constitutive part of story telling. They are key audiences of the stories, and as such they participate actively in the formation both of identity and interests, making both these concepts relational: AIn order to find out whether a particular constitutive story is a valid description of us, it must first be tested in interaction with others (Ringmar, 1996: 80). And again, ASince all stories require audiences, it follows that we cannot formulate notions of interests in isolation from other people - we simply cannot want things alone (Ringmar, 1996: 79).

Leaving stories about actions behind and focusing on constitutive stories (where actions still play a crucial part), Ringmar goes further into how stories are confirmed. Confirmation cannot be given by just anybody, but only by those others which the self recognises and respects as being of a kind to itself. This set of others are referred to as 'circles of recognition'. To a state, the circle of major importance will therefore be made up of other states.

An instance which is worthy of particular theoretical attention is of course the one where others deny recognition to the self's constitutive stories. In this case, the storied self has three options: to accept stories told about it by others, to abandon the stories which are not recognised in favour of others or to stand by the original story and to try to convince the audiences that it in fact does apply. AThus while the first two options mean that we accept the definitions forced upon us by others, the third option means that we force our own definition upon someone else (Ringmar, 1996: 82; also 185). And typically, the way to do this is to act.

The need to obtain recognition for constitutive stories, Ringmar insists, will be greater at so-called 'formative moments,' periods of 'symbolic hyper-inflation - times when new emblems, flags, dress, codes, songs, fêtes and rituals are continuously invented.' It will also be greater for social upstarts - newcomers approaching a circle of recognition - than to others. Thus, 'Social upstarts are likely to be very good rule-followers, not primarily because they fear punishment in accordance with the rule if they fail, but because they want to be identified as members of the group where a particular rule applies' (Ringmar, 1996: 86).

Having sketched out his narrative theory of action, Ringmar is ready to analyse particular actions in terms of stories. The point of the analysis will be to look at the story which a certain self, for example a state, tells about the action and to do this in terms of the meaning which goes into the story. Then that particular action - for example the question of why a state goes to war, for example why Sweden went to war in 1630--will have to be subsumed either under the category of 'stories about interests and actions' or 'constitutive stories about identity and action'. In order to determine this, Ringmar (1996: 90-91) presents a check-list of four points along the following lines:

1. Traditional explanations phrased in terms of interests should produce ambiguous, highly contested or perverse results;

2. The period in which the action took place should correspond to what we have called a "formative moment." It must be a period when new metaphors were launched, when individuals and groups told new stories about themselves, and when new sets of rules emerged through which identities were classified;

3. The particular person or group whose action we want to explain must be engaged in a process of identity creation. It must be someone who tells constitutive stories and tries to establish a presence in both time and space; someone who constructs an affective geography of friends and enemies; someone who pays careful attention to the rules of the social system to which he or she or it seeks to belong;

4. We must identify an occasion, or a series of occasions, on which recognition was denied under humiliating circumstances. We need to prove that our person or group suffered as a result and that the failure of recognition was indeed experienced as a loss of dignity, worth and "face."

If these criteria are fulfilled, as Ringmar demonstrates that they were in the case of Sweden in 1630, then an explanation of going to war in order to be recognised is convincing.

Ringmar's approach is nice not least because it presents itself as an explanation, and so may more effortlessly slide into the literature on the 'causes' of war than most other identity scholarship. His work is particularly angled towards cases where aspiring states want to break into the system consisting of states which already recognise themselves as such. If, however, we want to study the outbreak of war more generally, as an outcome of concrete political processes where one particular 'cause' such as a King's intention to have himself and his state's status recognised by other Kings and other states is not what we are interested in, then we need a more process-oriented understanding of how the representations of self and other are actually forged. The politics of representation is the home turf of poststructural identity scholarship, and it is to parts of this body of work that I turn next.

Representation of Identity

Looking at this body of scholarship, however, we find a number of interesting studies of how representing of self and other go hand in hand with the waging of war, but there is no one book or even article which, like Ringmar's, actually applies identity scholarship to the outbreak of war. Typically, however, a poststructural analysis demonstrates how in Der Derian's phrase the construction of self is implicated in the killing of the other. As the Schulte-Sasses (1991: 72) put it, 'a society that uses representations of war as a means of unifying the body politic in an imaginary fashion needs an elaborate network of signs representing Oneness and Otherness.'
I pick this text by the Schulte-Sasses for the reason that they highlight how, for a country such as the US, 'nationalism is part and parcel of a production of signs that can best be described in cultural, historical, and psychoanalytic terms' (1991: 94). The explicit reference to psychoanalysis is of the essence here. They read the Gulf War as 'a means of simulating a unified body politic' (1991: 68). The postulate is that there are two contradictory drives in modern society: one (which we have already discussed in connection with Mouffe elaboration of the political) which runs to agonistic politics, and one which is to do with the desire of each individual to imagine the body politic as unified, as socially cohesive. Governments, they argue, may try to shift attention away from the agonism of domestic politics towards aesthetic pleasure in the body politic by going to war. This tendency will be even stronger when traditional state strategies for mustering support such as representations of utopias have lost much of their lustre:

Media images become the postmodern heirs to narratively constructed utopias. [...] wars are at least in part propelled by their power to unify the body politic and to instill in the state's subjects the illusion of being masterful agents of history. This illusion of being a historical agent is again an aesthetic phenomenon; it demands the aesthetic representation of substitutes with which we can identify and depends on images that contain nothing messy or confusing, such as dead bodies or a humanized enemy. [...] While experiencing ourselves, collectively and individually, as a unified body, we simultaneously fall prey to the illusion that we can decipher and master the world. The danger is, of course, that in experiencing what 'theory' likes to call the jouissance of mastering the world as text and the text as world, we cover up our actual impotence as agents, which in turn worsens the nation's material situation (economy, infrastructure, educational system, etc.) and increases its dependence on images of superiority. (Schulte- Sasses, 1991: 70-71).

Now, this insight from psychoanalysis -- which in IR scholarship is seldom made as explicit is it is here but which makes its presence felt particularly in a Lacanian tapping (for an early example see Shapiro 1992)--is of course very interesting, and analyses of how such sequences play themselves out during war fighting may be very informative. The major problem with this approach for our task, however, is that we have here a topic which may be ever-present in politics and which may even be of growing importance. However, if it cannot be specified more clearly how this growing importance is actually happening, and how this will vary across time and space, this approach will remain as one of the many which may throw some general light on the outbreak of war, but which is not specific enough to lend itself to specific readings of how outbreaks occur in certain places and at certain times, and not others. One must also ask the more general question of whether scholarship which sets out to be anti-foundational, but which nonetheless puts so much store in an arguably foundational category of desire as does this one, can still be considered to be anti-foundational.

An unwillingness to be specific enough in asking the 'how-possible' question when it comes to outbreak of war is also a feature of poststructural scholarship which draws less heavily on the psychoanalytical tradition. Among abundant examples (Behnke, 1997; Hansen, 1997; Neumann, 1996) I choose to illustrate the point by casting a glance at David Campbell's book Writing Security: United States Foreign Policy and the Politics of Identity. The main reason for choosing this work is that in my view the overall most successful empirical reading of identity that we have in IR. The book is a thick description of US foreign policy as a seamless web of discourse and political practice which has played itself out through a series of engagements with others from the time of Cortés and up to the Gulf War. The US self is understood as a narrative structure, and it is argued that 'For a state to end its practices of representation would be to expose its lack of prediscursive foundations; stasis would be death' (Campbell, 1992: 11). Due to the role played by immigration in its genesis, the United States is presented as the imagined community par excellence, and this is seen as an additional factor which increases its need of having its representational practices recognised and confirmed.

Campbell's is an ethical concern; he argues that the knack is for a human collective to be able to carry out its practices of representation while living in difference; that is, without 'othering' other collectives. This, however, is exactly what the United States has failed, and is still failing, to do. One of the consequences is that it is perpetually on the look-out for new collectives to other:

If we take the cold war to be a struggle related to the production and reproduction of identity, the popularly heralded belief that we are witnessing the end of the cold war embodies a misunderstanding: while the objects of established post-1945 strategies of otherness may no longer be plausible candidates for enmity, their transformation has not by itself altered the entailments of identity which they satisfied (Campbell, 1992: 195).

Campbell gives a detailed reading of how foreign policy, with its focus on border maintenance, is a particularly apposite practice for identity formation, but also stresses the internal consequences of this. His reading of early US cold war diplomacy and the work of the Washington State Legislative Fact-Finding Committee on Un-American Activities, for example, stresses how

concomitant with this external expansion was an internal magnification of the modes of existence which were to be interpreted as risks. Danger was being totalized in the external realm in conjunction with its increased individualization in the internal field, with the result being the preformative reconstitution of the borders of the state's identity. In this sense, the cold war needs to be understood as a disciplinary strategy that was global in scope but national in design (Campbell, 1992: 172-173).

Once again, Campbell's approach demonstrates how poststructural work throws up very interesting general insights, but do not engage directly the question of how the outbreak of specific wars is to be analysed in detail and compared to other cases and non-cases of outbreak of war. We need more formalism, and I should like to suggest that one way of getting it (which brackets some poststructuralist concerns and highlights others) is to start off from the work done by the Copenhagen School of security studies.

Introducing the Concept of 'Violisation' to the Copenhagen School

The third task to be tackled, then, is to look at how certain symbolic economies work to produce war as an outcome of ever more sharply defined friend/enemy relations. If the focus in task one is on actor intentions and in task two on re-presentation, the focus here is not on actors or re-presentations as such, but first and foremost on the markers or diacritica of the boundaries of the actors. In other words, the focus is on the stuff which delineates and bounds the actors. The relevant task is to say something about how these diacritica are bound up with the outbreak of war. I should like to suggest that this may be studied most effectively by adding a small but crucial supplementary factor to the work already done by Barry Buzan, Ole Wæver et al. on the concept of securitisation.

Indeed, if the exercise undertaken in this paper is to link literatures, in this case the innovation called for seems to be a rather incestuous one. I suggest that we splice the work done on what has been called the 'Copenhagen School' of security studies (McSweeney, 1996; Buzan & Wæver, 1997) with the work which has appeared above some of the same signatures in what I have previously referred to as the 'Copenhagen coterie' of identity studies (Neumann, 1996: 162) and extend it into the literature on the outbreak of war. Specifically, this means adding a category of 'violisation' to the Copenhagen concept of 'securitisation.'

As Wæver (1995: 67) put it in the published version of the 1988 paper which launched the concept of securitisation, 'State security has sovereignty as its ultimate criterion, and societal security has identity. Both usages imply survival. A state that loses its sovereignty does not survive as a state; a society that loses its identity fears that it will no longer be able to live as itself'. For our purposes, the major problem with this dichotomisation, which I am quick to point out definitely has its uses when it comes to opening up the debate on the referents of security, is actually detrimental to the debate about the outbreaks of wars. To declare war, after all, is still an activity where states play a crucial role. Societies and society-lever groups may of course be active in a number of ways, both before and after the declaration of war. There may be interesting gray areas where state collapse begs the question of whose name the war is fought, there may be bands operating without it being clear what kind of status if any the state has conferred on them, etc. The issue of civil war complicates the picture further. The point for our purposes, however, is that the differentiation between state and society which as it were relegates questions of identity to the sphere of society is not only unhelpful, but downright detrimental to attempts to link the issues of identity and war. Dedifferentiation seems to be called for if the issue of outbreak of war is to be studied within a Copenhagen School framework.

If these words may seem harsh to an outsider, I should think that (other) members of the Copenhagen School would readily accept them. After all, in a preview of the next major co-authored book by the School, Barry Buzan (1997: 13) writes that

The Copenhagen School [...] argues against the view that the core of Security Studies is war and force [...] Instead, it constructs a more radical view of Security Studies by exploring threats to referent objects, and the securitization of these threats, that are non-military as well as military.

Which, it seems to me, is to say that the School has (so far) consciously avoided confronting the issue of outbreak of war, because it has seen the most pressing issues for immediate theorisation as lying elsewhere (see esp. Wæver, 1996: 113n9). When it has been in a situation where the issue has proved unavoidable, as was the case in its central work so far, Identity, Migration and the New Security Agenda in Europe (Wæver et al., 1993), which indeed has a separate chapter called 'Societal security and the explosion of Yugoslavia', it has tackled the question by drawing up what is referred to as a checklist of causal components rather than attempting fully to integrate the issue into the proposed theoretical framework. It is indicative of this indirect approach that the reference is to the 'explosion' of Yugoslavia rather than to the outbreak of war.

I would argue that this exteriorisation of the issue of outbreak of war from the extant work of the Copenhagen School is relatively easily amended by a de-differentiating move which highlights that identity is also an aspect of sovereignty, and for this and other reasons has pertinence not only for society, but also for the state. Having suggested where the conceptual stop for linking identity and the outbreak is located, I now turn to the issue of how to dislodge it.
The Copenhagen School argues that securitisation can be thought of as an extension of politicisation:

issues become securitized when leaders (whether political, societal, or intellectual) begin to talk about them -- and to gain the ear of the public and the state -- in terms of existential threats against some valued referent object. [...] Securitization can thus be seen as a more extreme version of politicization. It is the intersubjective establishment of an existential threat with a saliency sufficient to have substantial political effects. In theory, any public issue can be located on the spectrum ranging from non-politicized (meaning that the state doesn't deal with it, and it is not in any other way made an issue of public debate and decision); through politicized (meaning that the issue is part of public policy, requiring government decision and resource allocation or more rarely some other form of communal governance); to securitized (meaning that the issue is presented as an existential threat requiring emergency measures, and justifying actions outside the normal bounds of political procedure). In principle, the placement of issues on this spectrum is open; depending on circumstances, any issue can end up on any part of the spectrum (Buzan, 1997: 14).

In terms of identity, then, politicisation is a matter of inscribing certain differences between self and other with meaning as the defining diacritica of self and other. That is, certain differences which have so far not been activated as part of the political, are being politicised, which makes it possible retrospectively to talk about them as having been 'non-politicised' before they were 'politicised'.

Now, securitisation, it will be recalled, is about inscribing diacritica of the political with the extra burden of defining what stuff which should make up the security politics of a certain human collective. Wæver, who like other poststructuralist subscribes to a Schmittian definition of the political, tends to think about security as existential or ontological politics - that is exactly the political at its most political, the questions of telling friend from enemy, of defining who 'we' are and, functionally, the threats to who 'we' are.

The theorem proposed is thus that stuff can be non-politicised, politicised or securitised, and that there exist four processes known respectively as politicistion, de-politicisation, securitisation and de-securitisation. Stuff, for example an identity, may be securitised by the performance of the speech act of inscribing that identity with meaning as part of security politics, and de-securitised by the speech act of unsubscribing to such a representation.

By highlighting the importance of the speech act of securitization, the Copenhagen School performs the very important task of highlighting the constructed, intrasubjective character of the concept of security and hence also of the modus operandi of security politics. By using the concept of discourse in order to de-differentiate words and action and analyse them separately, they have been able to carry out some very convincing work. The outbreak of war may of course also be conceptualised as a speech act - if one's institutional station and one's preparations are the right ones, pronouncing the words 'I hereby declare war' is going to war. The waging of war, however, adds something more to the speech act of declaring war, and the crucial thing added is the use of force - that is, a violisation of politics. Waging war is by definition not only a question of speech acts, but also of actions. A crucial role is played by the action of killing, and more generally, by acts of violence which literally inscribe the will of one collectivity onto the physical bodies which make up the body politic that is another human collective. The number of people killed - be that 317 (Richardson) or 1,000 (Correlates of War Project), is for good reason almost always made a defining trait of what war is (Wiberg, 1976).

This, of course, is in and of itself not news to anybody: for example, Wæver quotes Clausewitz to the effect that 'War is an act of violence pushed to its utmost bounds' (quoted in Wæver, 1995: 53). When war-like activity does not include acts of violence, they are referred to by modifiers: wars of position, cold wars etc. So, in order to link the work of the Copenhagen School to the outbreak of war, I suggest that we differentiate the concept of securitisation so that securitisation is reserved for speech acts which perform the tasks which Buzan, Wæver and others have defined it as performing, and then add a new category for the cases where large-scale violence is actually in evidence.

Of course, identity is already violised for example when an asylum centre is arsonised and people die as an effect, but what is at issue here is the outbreak of war, and therefore I suggest we fix the threshold betwen securitistion and violisation at the point which is outbreak of war rather than at the point where one individual dies. At issue, after all, is not societal security and the identity/society nexus, but identity and war, which I argued still implicates the state very directly. Thus, for this purpose societal violence which is not intended to impinge on the question of state borders may be bracketed, and attention traced on cases where the issue is the re-presentation of states as such. In this way, Clausewitz' formulation of war being the continuation of politics by other means can be slotted directly, and not only indirectly, into the Copenhagen School framework of politicisation and securitisation. If we add a category of 'violised' to the three categories already in circulation in the Copenhagen School and also the two corresponding processes of 'violisation' and 'de-violisation,' we get the following extended continuum:

Non-politicised - politicised - securitised - violised

Two questions then present themselves: first, is there any reason why one should not modify the model in this way, and second, how do we draw on this in empirical reasearch.

The main argument against differentiating the idea of securitisation as it already stands may be that this would detract attention from the speech act of security by once again in a traditional fashion pointing to the material factors at work in cases of large-scale violence. Even more problematically given that the work of the Copenhagen School has until now made perhaps most headway in opening up the question of the referent of security, by highlighting how identity pertains to states and not only to society, my proposed move would take some attention away from societal questions. For a number of purposes, this may indeed be a loss. The mirror image of such a critique would be complaints that reserving violence for actions which have a material character about them actually downplays the violence wrought by structural factors and by speech acts. As already noted, AThe Copenhagen School [...] argues against the view that the core of Security Studies is war and force. Adding a category of 'violisation' at the extreme end of the continuum does of course have the effect of once again highlighting war and force. I should think, however, that this would be worth while, since Security Studies must of necessity tackle the issue of outbreaks of war. The wish to widen the agenda and play down the absolutely overshadowing place traditionally taken up by this issue is definitely both understandable and laudable, but it cannot be taken so far that the war-baby is thrown out with its bathing water. If it is granted that the issue of outbreak of war is not integrated in the Copenhagen School framework as it stands, if it is granted that it should be, and if it cannot be integrated in a better way than that proposed here, then I suggest that the concept of violisation be adopted.

As for the question of empirical research, I should think that an obvious first case to which the proposed extended continuum could be applied would be the outbreak of wars in former Yugoslavia (Eide 1997). We have here a number of cases where national identities became not only securitised, but also violised. Serb and Croatian national identities, Bosniak political identity and Muslim religious identity come to mind as examples. By contrast, Macedonian and Albanian ethnic identities were securitised but not violised (I repeat that 'violised' is suggested used only when violence takes place on a certain scale). If it can be demonstrated that the outbreaks and non-outbreaks of war can usefully be analysed in terms of violisation of identity, then that would be very interesting indeed. If it turns out that the way identities were structured was not a crucial factor in the outbreaks of these wars, then that would be very interesting too.

Conclusion

Scholarship on identity does throw up a number of insights which seem useful for a study of the outbreak of wars. For various reasons, however, most extant scholarship has largely been preoccupied with other forms of social interaction than those which have led to war or have come close to leaing to war. Once this focus is present, furthermore, the focus is not on the action of going to war itself, but on a plethora of other forms of human interaction which are to do with war. Between the specific and explanatory focus of Ringmar's narrativist study of a King's intention and the general and critical focus of poststructural studies of how the self is implicated in war, I suggested that the formalised and process-oriented framework of the Copenhagen School stands out as a particularly promising pad from which to launch empirical studies of identity and the outbreak of war. The prerequisite, it is argued, is that the School adopt a concept of violisation, understood as the process whereby an already securitised issue such as identity becomes a casus belli over which blood must run.

Notes

1. This, Mouffe (1994: 108) argues, is because 'liberal thought employs a logic of the social based on a conception of being as presence, and which conceives of objectivity as being inherent to things themselves. This is why it is impossible for liberal thought to recognize that there can only be an identity when it is constructed as a 'difference', and that any social objectivity is constituted by the enactment of power.'
2. All this is to say that Ringmar brackets what may perhaps be called the 'second wave' of cognitive literature on ideas, frames, logic of action etc. which began to blossom as his own project neared conclusion, for examples see Goldstein & Keohane (1993), Risse-Kappen 1994; for a critique of this literature see Laffey & Weldes (forthcoming).
3. The choice of society rather than community may have been coincidental - in mainstream political theory of the 1980s and 1990s, 'communitarianism' has approached some of the same problems, and have highlighted the issue of identity as well (cf., for example, Mouritzen forthcoming). With the exception of Nick Rengger, however, IR has stuck to 'society', and this alone is probably the reason why society was chosen instead of security.
4. Let me be totally clear that the issue here is either the chapter as such, which is very interesting, or the general thrust of the research done by its author Håkan Wiberg, who actually devoted an entire chapter to an early book on Peace Research to 'war and its causes' (Wiberg 1975: 99). These two factors, it seems to me, weaken other explanations of the failure so far of the Copenhagen School to integrate the issue of outbreak of war and thus actually strengthens my argument.
5. Of course, what turns out retrospectively to have been the outbreak of war can happen without there being a previous overt speech act of declaration of war, as when Japan attacked the US in 1941 without previous notice. This need not detain us further here.

References

Bosnia and the Problem of Security. Paper presented to the 38th annual ISA convention, 18-22 March 1997, Toronto, Canada.

Buzan, Barry. 1997. ARethinking Security after the Cold War. Cooperation and Conflict, Vol. 32, No.1, pp. 5-28.

Buzan, Barry & Ole Wæver. 1997. ASlippery- Contradictory- Sociologically Untenable- The Copenhagen School Replies. Review of International Studies, Vol. 23, No. 2, pp. 241-250.

Campbell, David. 1991. Writing Security. United States Foreign Policy and the Politics of Identity. Minneapolis: University of Minnesota Press.

Doty, Roxanne Lynn. 1996. Imperial Encounters. The Politics of Representation in North-South Relations. Minneapolis: University of Minnesota Press.

Eide, Espen Barth. 1997. A`Conflict Entrepreneurship: A Few Post-Yugoslav Reflections on the `Art` of Waging Civil War. In Joint PRIO/NUPI Pilot Project,
Engaging the Challenges of Tomorrow: Adjusting Humanitarian Interventions to the Character of Future Conflicts, Oslo: NUPI/PRIO.

Goldstein, Judith & Robert O. Keohane. Eds. 1993. Ideas and Foreign Policy: Beliefs, Institutions, and Political Change. Ithaca: Cornell University Press.

Hansen, Lene. 1997. APast as Preface: Civilization and the Politics of the Third Balkan War. Paper presented to the 38th annual ISA convention, Toronto, March
18-22, Toronto, Canada.

Howard, Michael. 1983. AThe Causes of War. In The Causes of War and Other Essays. London: Temple Smith, pp. 7-22.

Laffey, Mark & Jutta Weldes. Forthcoming. ABeyond Ideas. European Journal of International Relations, Vol. 3.

Laclau, Ernestoand Mouffe, Chantal. 1985. Hegemony and Socialist Strategy: Towards a Radical Democratic Politics. London: Verso.

McSweeney, Bill. 1996. AIdentity and Security: Buzan and the Copenhagen School. Review of International Studies, Vol. 22 No. 1, pp. 81-93.

Mouffe, Chantal. 1994. AFor a Politics of Nomadic Identity. In George Robertson et al., Eds. Travelers' Tales. Narratives of Home and Displacement. London:
Routledge.

Mouritzen, Per. Forthcoming. The Idea of Civic Participation: Problems and Promises of Political Community. Doctoral Thesis presented to the European
University Institute, Florence, Italy.

Neumann, Iver B. 1996. Self and Other in International Relations. European Journal of International Relations, Vol. 2, No. 2, pp. 139-174.

Ringmar, Erik. 1996. Identity, Interest and Action. A Cultural Explanation of Sweden's Intervention in the Thirty Years War. Cambridge: Cambridge University Press.

Risse-Kappen, Thomas. 1994. AIdeas do not Float Freely: Transnational Coalitions, Domestic Structures, and the End of the Cold War. International Organization,
Vol. 48, No. 2, pp. 185-214.

Schulte-Sasse, Jochen & Linda Schulte-Sasse. 1991. AWar, Otherness, and Illusionary Identification with the State. Cultural Critique, Vol 19, (Fall), pp. 67-
95.

Shapiro, Michael. 1991. AThat Obscure Object of Violence: Logistics, Desire, War. Alternatives, Vol. 17, No. 4, pp. 453-477.

Van Evera, Stephen. 1994. AHypotheses on Nationalism and War. International Security, Vol. 18, No. 4, pp. 5-39.

Wallensteen, Peter. 1994. Från krig till fred: om konfliktlösning i det globale systemet. Stockholm: Almqvist & Wiksell.

Wæver, Ole, et al. 1993. Identity, Migration and the New Security Agenda in Europe. London: Pinter.

Wæver, Ole. 1995. ASecuritization and Desecuritization. In Ronnie D. Lipschutz,
Ed. On Security. New York: Columbia University Press.

---------. 1996. AEuropean Security Identities. Journal of Common Market Studies, Vol. 34, No. 1, pp. 103-132.

Wiberg, Håkan. 1976. Konfliktteori och fredsforskning. Stockholm: Scandinavian University Books.