Introduction
One of the questions, which has concerned Security Studies since its inception, is whether states – or other collectivities – might ever be in a state of such security, that they no longer think of themselves as facing external or internal threats. The classical answers to this question are well-known: realists hold that the anarchical system implies, that the risk of threats is inevitable; liberalists are more susceptible to an improvement of inter-state insecurity relations, for instance through the democratic peace; and non-traditional or ‘widening’ approaches from constructivism to post-structuralism allow for a lessening or even a full move out of the logic of insecurity. The reason why such diverging answers are delivered – and why it is highly unlikely that agreement would arise – is that security approaches make different assumptions about the identity of the state and its capacity to change its views of other actors. Yet, even those realists, who are sceptical that states could exist in a system of amity, recognise that what falls into the category of threats and dangers changes across time and place. The concept of desecuritisation, launched by Ole Wæver and the Copenhagen School fifteen years ago, focuses precisely on the attractions and dangers of moving ‘out of security’, and the concept has turned out to be a popular one. Coined as the conceptual twin to securitisation, desecuritisation refers to ‘the shifting of issues out of emergency mode and into the normal bargaining process of the political sphere’, and it is, according to Barry Buzan et al., the preferred ‘long-range option’.Footnote 1
For those exploring the political and philosophical implications of security theory, it is through the preference for desecuritisation that the Copenhagen School taps into larger normative and political issues.Footnote 2 Others have used desecuritisation to cast analytical light on cases as diverse as Soviet/Russian policies from the late 1980s onwards, the framing of environmental security within the US Department of Defense, minority rights in Eastern Europe, and female soldiers in post-conflict Sierra Leone.Footnote 3 The call made by Wæver in 1995 for a heightened theoretical and empirical attention to ‘de-securitizing politics’ can thus be said to have been heeded.Footnote 4 Yet, the attention bestowed upon desecuritisation notwithstanding, the concept is far from viewed as unproblematic or consistently empirically applied.
That there is still further work to be done on desecuritisation is indicated, first, by the claim of security theorists, such as Jef Huysmans, who holds that the preference for desecuritisation is technical, managerial and instrumental, rather than genuine political or ethical.Footnote 5 Huysmans's critique has been seconded by Rita Floyd and Claudia Aradau, who argue, that the concept is ‘largely under-theorised and open to interpretation’, that it is undermined analytically or politically by an ‘insufficient attention to politics in the theory of securitization’, and that there are no normative connotations to the Copenhagen School due to its repudiation of the concept of emancipation.Footnote 6
Second, due in part to desecuritisation's underdeveloped status, the securitisation-desecuritisation nexus has been read through a wide range of political theorists. Huysmans, Michael C. Williams and Andreas Behnke identify an influence from Carl Schmitt in the Copenhagen School's emphasis on the exception; Williams, Rita Taureck and Thomas Diez and Atsuko Higashino see an affinity between Jürgen Habermas's discourse ethics and the requirement, that an audience must be persuaded to accept a securitisation; Aradau raises the possibility of ‘a Foucauldian-inspired desecuritisation move’; Holger Stritzel and Floyd discuss the ontological and methodological significance of Jacques Derrida; and the impact of Hannah Arendt is noted by Matti Jutila and Floyd.Footnote 7 Schmitt, Habermas, Michel Foucault, Derrida, and Arendt offer very different understandings of politics, and the Copenhagen School has not been particularly clear on where it locates itself, as its concept of politics is said to derive from ‘a middle ground’ between as diverse and complex theorists as Arendt and David Easton, Schmitt and Habermas, and Max Weber and Ernesto Laclau, who are only briefly discussed.Footnote 8
A third indication of the need for further work on desecuritisation is that the concept is applied in ways that, when compared, seem unsystematic or even contradictory. Some, like Bülent Aras and Rabia Polat identify desecuritisation, when one set of enemies is replaced by another; in Kristian Åtland's analysis, it captures the shift of an issue from the field of security to that of normal politics, with no other issues appearing as ‘replacement threats’; and Megan MacKenzie adopts desecuritisation to criticise, how female ex-combatants in Sierra Leone are cut off from the funding and attention, that male ex-combatants receive.Footnote 9 These multiple conceptualisations of desecuritisation at the level of empirical application tells us something important, not only about the analytics of desecuritisation, but about the desire among securitisation theorists for foregrounding the political-normative.
Thus far, the international political theory and the applied analysis literatures have by and large run on separate tracks. These literatures are related, however, in that they all engage – explicitly or implicitly – with the concept of politics in securitisation theory: the concept is said to be absent, there are too many political theory ancestors for one coherent concept to crystallise, or desecuritisation unfolds empirically through a set of disjunct political dynamics. This article holds in response, that we can get a better understanding of the concept of desecuritisation – and the political and normative choices it entails – if we read the political theory debates and the empirical applications in the light of one another. More specifically, the article makes three contributions. First, it counters the claim, that securitisation theory has a thin conception of politics. It clarifies, second, how the readings of the Copenhagen School as Schmittian, Habermasian, Derridarian, Arendtian and (latently) Foucaultian fit into securitisation theory. Third, it shows, how the empirical applications of desecuritisation can be seen as falling into four forms of desecuritisation, each of which is distinguished by a particular conception of politics.
In terms of the concept of politics adopted in this article, I start from the Copenhagen School's own account of politics as ‘a continuous struggle to establish the quasi-permanence of an ordered public realm within a sea of change’.Footnote 10 Politics is about providing stability to social relations, at the same time as it entails openness, as to what kind of ‘stable’ solutions should be provided.Footnote 11 To politicise something is thus to do two things: to claim that this is of significance for the society in question and to make it the subject of debate and contestation. Taking this meta-concept of politics to the realm of security, there is a distinction between the securitised and the politicised, as ‘security’ ‘takes politics beyond the established rules of the game and frames the issue either as a special kind of politics or as above politics’.Footnote 12 The politicised, by contrast, allows for deliberation, discussions and ‘normal bargaining processes’.Footnote 13 Securitisation is, on the one hand, a move out of the logic of normal politics, but it can also be seen as ‘a more extreme version of politicization’.Footnote 14 ‘Politics’ thus has a double status: it refers to a political, public sphere of engagement, and it refers at the meta-level to the moves – and choices – between the politicised and the securitised. As Williams has suggested, there are two themes that are particularly central to the meta-choices that surround security politics: the friend-enemy distinction and the extent to which a public sphere of engagement and mediation of difference can be justified.Footnote 15
The article falls in two parts. The first part consists of four readings that recover the political and normative richness of desecuritisation from within securitisation theory itself. First, the concept of securitisation draws upon Schmitt, is highly political, and implies an understanding of (security) politics as emergency and exceptionality. As securitisation and desecuritisation were formulated in tandem, with desecuritisation working as, in Derrida's terms, the supplement, desecuritisation is necessarily a political concept as well. Second, the concept of the public sphere in securitisation theory has affinities with Habermas, but there are more post-structuralist elements, too. Third, the relationship between securitisation and desecuritisation is constituted as one of choice and responsibility. This conception of choice is highly political and derived not only from Schmitt,Footnote 16 but from Derrida, Hans J. Morgenthau and Arendt. Fourth, desecuritisation arose from Wæver's political analysis of Cold War détente.
The second part of the article connects the political theory driven discussions of the first part with the empirical applications of the concept of desecuritisation. Building on the understanding of security politics as concerned with two issues, namely the status of enmity and the possibility of a public sphere, four forms of desecuritisation are identified. Change through stabilisation is when an issue is cast in terms other than security, but where the larger conflict still looms; replacement is when an issue is removed from the securitised, while another securitisation takes its place; rearticulation is when an issue is moved from the securitised to the politicised due to a resolution of the threats and dangers, that underpinned the original securitisation; and silencing is when desecuritisation takes the form of a depoliticisation, which marginalises potentially insecure subjects.
Desecuritisation as the supplement to securitisation
If there has been scepticism as to the political status of desecuritisation, the concept of securitisation has been widely recognised as political. Huysmans, Williams, and Behnke hold more specifically, that Schmitt's understanding of the Political lies at the core of the Copenhagen School's definition of ‘security’ as the sovereign authority's ability to legitimate the use of emergency measures in the face of exceptional threats. ‘Security’ is thus a speech act with particular political implications. Schmitt held that a community is unable to exist without a distinction between friend and enemy, and that making decisions on that distinction may require ‘a strong, dictatorial political leadership’.Footnote 17 Schmitt's decisionism was, moreover, formulated against, what he saw as liberalism's formalism, rationalism, and technical view of the political. This in turn echoes the status of the security speech act within the Copenhagen School as a decision and thus as radically ‘unfounded’.Footnote 18 As the mobilisation of the security speech act is always a choice, it is, holds Williams, always ‘explicitly political’.Footnote 19
There are strong affinities between Schmitt's the Political and the logic of securitisation, but the Copenhagen School also holds, that the Political is not identical to politics as such: ‘In terms of Schmitt versus Habermas, politics cannot be reduced to the friend-enemy distinction.’ Wæver has also made clear, that while his conceptualisation of security draws on Schmitt, his understanding of politics follows Arendt in its stress on inter-action, plurality and constellations of decisions.Footnote 20 As Patricia Owens further explains, ‘The ultimate expression of Arendt's idea of politics is not Schmitt's struggle to the death between enemies. It is the ability to appear before plural equals and to debate and act to build a common world … When individuals come together to debate and to act they create this political realm between them.’Footnote 21 There is also an affinity between Arendt's understanding of the political and Williams's observation that the influence of Schmitt on securitisation theory is reigned in by the need for security speech acts to be accepted by an audience, which in turn opens for dialogue and the ‘transformation of security perceptions both within and between states.’Footnote 22
Having established that the concept of securitisation has strong affinities with Schmitt's concept of the Political, let us return to the concept of desecuritisation. The fact that we have ‘securitisation theory’, rather than ‘desecuritisation theory’, illustrates that ‘securitisation’ has a (seemingly) superior status. Desecuritisation is derivative of securitisation semantically (modified through ‘de’), and in terms of the political modality, the concept identifies: desecuritisation happens ‘away from’ or ‘out off’ securitisation. Moreover, there is a theoretical inferiority attached to desecuritisation in that it lacks securitisation's grounding in popular language. One cannot desecuritise through speech acts such as, ‘I hereby declare this issue to no longer be a threat’, as this would be invoking the language and logic of security.Footnote 23 Desecuritisation happens as a result of speech acts, but there is not, strictly speaking, ‘a’ desecurity speech act.Footnote 24
Yet, the fact that the terms were coined together through Wæver's 1995 article on ‘Securitization and Desecuritization’ indicates, that ‘securitization’ was invented in tandem with ‘its’ opposite concept. As Derrida famously argued, all signs are constituted through hierarchical juxtapositions to something they are not, but some signs enter into particularly striking hierarchical pairs, as one term is seen as the real, original, or essence, and the other as the supplement.Footnote 25 The supplement is derivative of and inferior to the privileged ‘real’, but the supplement is at the same time, ‘that which is required to complete or fill up some existing lack’.Footnote 26 An often used example is that of the ‘supplement’ to a lexicon: the supplement is unthinkable without ‘the real lexicon’, but ‘the real lexicon’ is also incomplete without ‘its’ supplement. In Christopher Norris's account of Derrida's logic of supplementarity, the ‘apparently derivative or secondary term takes on the crucial role in determining an entire structure of assumptions.’Footnote 27 Taking the idea of supplementarity to securitisation theory, in the same way as the lexicon is incomplete without its supplement, securitisation relies upon desecuritisation.Footnote 28 Not only does this invert the superiority-inferiority relation between the two terms, it also shows the political status of both. First, if securitisation is about the fundamental distinctions that found human life, there is no way that decisions about leaving this logic through desecuritisation would not be political, too. Second, were there only securitisations, there would be only hyper-politicisation and no ‘normal politics’ for securitisation to separate itself from.Footnote 29 Securitisation needs, in other words, the desecuritised as its constitutive, and equally political, outside for it to achieve analytical and political meaning.
The public sphere and the friend-enemy distinction
Another way to interrogate the political status of desecuritisation is to ask how the concept ‘reworks’ the two components that make up security politics, that is, the public sphere and the friend/enemy distinction. Let us first clarify what kind of space desecuritisation refers to. The definition of desecuritisation as the ‘shifting of issues out of emergency mode and into the normal bargaining processes of the public sphere’ suggests a move from the securitised to the politicised (‘meaning the issue is part of public policy, requiring government decision and resource allocations or, more rarely, some other form of communal governance’), rather than to the non-politicised (‘meaning the state does not deal with it and it is not in any other way made an issue of public debate and decision’).Footnote 30 A move that takes us from securitisation to politicisation is different from one that takes us to non-politicisation, and given that the Copenhagen School defines desecuritisation as the movement of issues ‘into the ordinary public sphere’, it might be best to delimit desecuritisation to the former move.Footnote 31
Desecuritisation thus creates or restores a genuine public sphere, where humans can, in an Arendtian fashion, ‘debate and act to build a common world’.Footnote 32 Securitisations are, to be clear, also a mode of ‘bargaining’ in that securitising actors need to convince their audiences that emergency measures are required and suspensions of normal rules and rights thus legitimate. This, according to Williams, shows a Habermasian commitment to communicative action and discourse ethics within securitisation theory.Footnote 33 Yet, the acceptance of the securitising speech act is also a bargaining ‘out of’ the political sphere which radically transforms it. This makes an audience's acceptance of a securitisation a leap of faith, in that the suspension of normal political rules can prevent issues from becoming part of the public domain. Securitisation implies a right to prevent information from reaching the public sphere and ‘the public’ might thus be unable to mobilise in favour of desecuritisations, it would have wanted to, had it had the information in question. Thus while Williams is right to identify a communicative rationality and a dynamic of persuasion in the securitising move, it is simultaneously a move that challenges – or even eradicates – a Habermasian assumption of what politics should be, namely that ‘participants in a discourse are open to being persuaded by the better argument and that relationships of power and social hierarchies recede in the background’.Footnote 34 The extent to which an audience retains a capacity to revoke its accept of a securitisation and take it back into the politicised thus becomes a crucial issue.
Securitisations, which have become institutionalised to such an extent that they no longer are in need of explicit articulations to justify their status, pose a particular challenge for desecuritisation attempts.Footnote 35 Viewed through the concept of the public sphere, the absence of speech acts seems to make them less susceptible to ‘being taken back’ through desecuritisation. Yet, whether this is seen as a political-normative problem is a different question. Institutionalised securitisations might rely upon repression and violence or they might be the product of repeated audience acceptances atrophying into a set of institutionalised practices. The potential of securitisations to become embodied within unproblematised institutions establishes a complex relationship between what we might call ‘institutionally sanctioned practices’ and public discourse. The Copenhagen School does not assume, that there is ‘a rather simple one-way flow’ between public discourse and policy practice, as Christina Boswell suggests, but sees the political sphere as a dynamic space where actors seek to justify their policies and destabilise those of their opponents.Footnote 36 Situations where it looks as if ‘administrative agencies may adopt securitarian practices without a prior green light from political discourse’ are thus to be analysed either as indications of ‘prior green light’ having been given to such an extent that they are no longer needed.Footnote 37 Or, if no light is indeed discernable, such practices are susceptible to being brought to the public attention in which case political actors, who are formally responsible, confront the choice to lend their legitimacy – and political capital – to the securitisations upon which those practices are based or curtail such practices so that the audience no longer sees them as securitising. As Roxanne Lynn Doty's study of border vigilantes on the US-Mexican border illustrates, security practices can be carried out by non-state actors ‘in diverse and amorphous ways, long before officials or elites utter securitizing speech acts’.Footnote 38 The point is that securitising actors are held accountable for their securitisations in different ways depending on their status vis-á-vis formal political institutions and that ‘political actors’ constitute – or avoid responsibility for – the practices of ‘their’ administrators or citizens through dynamic processes.
The reinvigoration of the public sphere, that desecuritisation implies, facilitates the engagement of a wider range of actors than if an issue is constituted as one of securitisation. But a desecuritising move might not ‘only’ expand the number and kind of agents, but transform the identities and interests of Self and Others.Footnote 39 Desecuritisation requires a loosing of the friend-enemy distinction possibly to the point of a whole-scale transformation where ‘the enemy’ shreds its identity, as was the case with the Soviet Union as the Cold War ended. Another possibility is a less fundamental or sudden reconstitution of identities that may facilitate change over time. Wæver's conception of politics as one of interrelations implies, that the process of desecuritisation is itself one of shifting interrelatedness: it transforms not only who the self and its enemy are, but what they are. Desecuritisation is, in short, performative: it must instantiate the non-threatening identity of the Other for desecuritisation to be possible.Footnote 40 Desecuritisation is not, in other words, a linguistic or political two-step procedure where first ‘we’ have to agree that X is no longer threatening and then, ‘we’ agree to stop speaking security.
One should keep in mind that speech itself is not transparent or devoid of power and that ‘the security speech act is not defined by uttering the word security’.Footnote 41 Securitising actors may reconstitute an issue such that it avoids the high-pitched nodes of radical, barbaric, blood-thirsty Others, while still situating it within a modality of securitisation. To give an example, immigration discourse might be couched in ‘civilised’ terms where ‘immigrants’ are not ‘threats’, but for instance ‘better helped in their own environments’. Yet, the institutional structures and ways in which anti-immigrant control is practiced might reveal a much more ‘securitised’ political terrain. Such cases of ‘strategic self-moderation’ raise the question why securitising actors appropriate this form of discourse rather than a more linguistically overt securitisation. One answer would be, that there are certain ‘civilising tropes’, that the audience in question is less likely to rebel against – or at least securitisation actors believe this to be the case.Footnote 42 If we read the status of the public sphere in the light of Wæver's post-structuralist writings, particularly of the late 1980s, we might further add, that subjects are not given prior to their entrance into the public sphere, but that they are constituted in discourse. This implies that the ability of politicisation to transform relations between Selves and Others is not only a matter of dialogue and persuasion in a Habermasian sense, but of questioning the very subject constructions that are on offer. For instance, debates over how ‘Europe’ can dialogue with ‘its Muslims’ presuppose, that there is a Muslim subject and that individuals want to be recognised as such, rather than through other forms of subjectivity.
Desecuritisation as responsibility
A different way that desecuritisation reveals its political status is through the emphasis on responsibility and choice in decisions related to desecuritisation. Wæver states a preference for desecuritisation on the grounds, that it ‘would be more effective than securitizing problems’, yet, one should be hesitant to conclude, as does Huysmans, that ‘effective’ equals an understanding of desecuritisation as technical, managerial, and instrumental.Footnote 43 ‘Effective’ could be instrumental, but it might also refer to what is ‘effective’ in terms of pursuing a political-normative project. For one thing, the Danish word effektiv has the connotation not only of efficiency, but of something being enabling. Wæver is, moreover, adamant that ‘we’ are responsible for our securitisations and desecuritisations, and he calls for a morally committed form of agency.Footnote 44 The intensity with which ‘our responsibility’ is invoked is for instance supported by Wæver's description of his own words as ‘alarmist’.Footnote 45 ‘We’ – academics, politicians, citizens, all who speak – must actively work through security to get to desecuritisation.Footnote 46
The reconstitution of the friend-enemy distinction, that desecuritisation entails, also imply that while desecuritisation is ceteris paribus the preferred long term option, the Copenhagen School refrains from an embrace of ‘universal desecuritization’, that sees desecuritisation as ‘the preferable ethico-political strategy for scholars and societal actors, regardless of time and space’.Footnote 47 As Wæver puts it, although ‘our reply’ is usually to aim for desecuritisation, ‘occasionally the underlying pessimism regarding the prospects for orderliness and compatibility among human aspirations will point to scenarios sufficiently worrisome that responsibility will entail securitisation in order to block the worst’.Footnote 48 Wæver's insistence on there being instances, where we would indeed want to securitise, also leads him to criticise ‘hard-core post-structuralists’, who seek to open up, avoid making choices, and who recommends ‘living dangerously’.Footnote 49
The responsibility for ‘closing down’ the political domain that securitisation entails raises the question, whether a move from the securitised to the depoliticised might ever be desirable. Considering that desecuritisations take place around threats that have been explicitly articulated, the likelihood of these ‘disappearing’ from the securitised directly into the depoliticised is probably quite small. Yet, as Floyd's analysis of the Bush Administration's move of defence environmental issues from the securitised to the depoliticised shows, it is not impossible.Footnote 50 As actors securitise and politicise can they also depoliticise in ways that facilitate the long term goal of a political bargaining sphere? One might think of this as a possibility, when an issue has been so deeply securitised, that locating it within the politicised brings so much attention that it risks slipping back into the securitised. Depoliticisation might in such cases appeal as ‘strategic time-out’. The dynamics surrounding securitisation of diseases such as HIV/AIDS might illustrate this situation.Footnote 51 Assume we have a securitisation of those afflicted by HIV/AIDS followed by a medical-technological break-through, where an affordable cure is introduced. This, in turn, facilitates a constitution of ‘disease management’ as a technical-bureaucratic issue. Not only might one argue that the ‘security problem’ of the decease has been ‘solved’ such that it no longer has to be the subject of lengthy public debates, one might see this as a ‘pre-emptive strike’ that protects the funding allocated to disease management by not making it the subject of political debates, debates that might then move HIV/AIDS back into the modality of the securitised.Footnote 52
The danger of constituting something as a technical-nonpoliticised issue is, however, that it – like securitisation – implies a move out of the political bargaining sphere and an institutionalisation of contestable political decisions and priorities. How we assess the potentials of non-politicisation thus depends on whether we see public deliberation – and restitution – as the (only) way to ameliorate conflicts. A case that shows the complexities of this issue is that of post-World War II Yugoslavia, where atrocities committed by Serbian and Croatian forces during the war were repressed until they were securitised in the late 1980s and early 1990s. Although the claims that older conflicts were ‘frozen’ during the Cold War should be treated with caution – as this presupposes essential identities just waiting to resume ‘their’ true form – the silencing of intra-Yugoslav massacres did provide securitising actors with an apt material around which to rally an audience.Footnote 53 Hence what in securitisation terms can be characterised as a move from securitisations during World War II to a depoliticisation of Croat-Serbian relations/atrocities during the Cold War might in the long run have facilitated the return of mutual securitisations.
The choice between securitisation and desecuritisation must – considering the political status that both concepts have – be inherently political. As a consequence, we cannot derive decisions on whether to (de)securitise from any objective status that threats might have. Nor should we, holds Wæver, assume that we could identify a set of universal principles, that would give us the answer to whether securitising, say the environment, terrorism, or immigration is a good or a bad thing.Footnote 54 A truly universal set of (de)securitising directions would, as all universalisms, be both depoliticising and inherently particularistic, as no ‘universalism’ can exist without spatial or temporal exceptions.Footnote 55 Drawing on Arendt and Morgenthau, Wæver calls instead for seeing the political situation as ‘always unique and concrete, a field of forces, a situation demanding a choice, a choice that has consequences’.Footnote 56 This accentuates the demand for personal responsibility, as politics is essentially unpredictable and ‘can never be reduced to meta-questions’.Footnote 57 In the absence of meta-answers one has to ‘make ultimately unfounded ethical and political choices’ in a world of ‘unsolvable real-world dilemma’.Footnote 58
The political history of desecuritisation
The fourth and final element in the recovery of the political status of desecuritisation consists of a reading of the historical context within which desecuritisation was and has been developed. Interestingly, in the numerous discussions of securitisation theory, few pay much attention to the way in which securitisation theory was developed in the shadow of the Cold War.Footnote 59 This is puzzling given that securitisation theory has been developed – and presented – through continuous engagement with contemporary events.Footnote 60 Yet, as Wæver writes in 1995, it was in fact ‘European security between 1960 and 1990, the period of change and détente, which provided the framework for developing the speech act interpretation of security’.Footnote 61 Wæver's key concern – and the reason for inventing the concept of desecuritisation – was to find a way in which East European social movements and dissidents could modify their societies without triggering ‘security responses’ from ‘their’ elites.Footnote 62 The use of concepts ‘other than security’ such as ‘détente’ and ‘human rights’ within the Helsinki/CSCE process were part of a desecuritising discourse, where Western states and non-state actors worked to reassure Eastern elites, that transformations could occur without generating ‘insecurity’. Theorising détente was thus part of a political project, that strove ‘to open up domestic space for more political struggle’, and this clearly made desecuritisation more than an analytical or technical concept.Footnote 63 As is evident from the footnotes in ‘Securitization and Desecuritization’, the formulation of the political position that ‘desecuritization’ encapsulated was not a free-standing incident, but the culmination of years of politically engaged writings, usually from a deconstructivist-post-structuralist position.Footnote 64
Going back to these writings of the late 1980s, the view of détente as capable of generating a full-scale transformation of the Cold War is less optimistic than ‘Securitization and Desecuritization’ might perhaps lead one to expect. Wæver held, that ‘The East-West conflict is not a conflict to be solved’, but to be managed through the development of a non-violent conflict culture.Footnote 65 Change was only possible, if both parties considered their position recognised by the adversary. ‘A dynamic process is only possible when the actors concerned engage in a certain amount of mutual stabilization and assurance … Political change is only possible in an atmosphere of détente. And détente is only possible when the fundamental structure is stable.’Footnote 66 Wæver observed that he could be criticised for being ‘socially status quo oriented by wanting (both sides) to respect the “security interests” of the power holders on the other side’, although his call for change through stabilisation did not endorse either of the political establishments.Footnote 67 Rather, given that the Cold War was an unsolvable conflict – at least for the foreseeable future – détente was the only strategy through which relations could be eased.Footnote 68 Although the terms securitisation and desecuritisation are not yet used, security is identified as a speech act,Footnote 69 and one might see the genesis of desecuritisation – as a concept and a political preference – in Wæver's definition of the long term goal as to channel
the East-West conflict into forms of competition which are not only non-violent but also non-security. The task is to ‘push back’ the use of the emergency right to label challenges as ‘security interests’. ‘Security’ is a label used by a regime when it needs or wants to claim a special ‘state’ right to use all means necessary to hinder a particular development that allegedly threatens ‘security interests’. More and more trans-national activities should become purely ‘economic’, ‘social’ and ‘political’.Footnote 70
In the light of the debates over whether desecuritisation has political sting, it is worth stressing, that Wæver saw his analysis as linked to him being a ‘peace activist’ and ‘a kind of security expert in Denmark’, and that he called for ‘a higher tolerance of criticism and politics’, when debating East-West security arrangements.Footnote 71
In hindsight, it is easy to fault Wæver for his failure to predict the end of the Cold War – although in this, of course, he was far from alone – and for the conservatism of some of his predictions: it would be highly negative for détente, if countries unilaterally left their alliances and governments should not directly support oppositional forces on the opposite side.Footnote 72 Yet, it is important to acknowledge that conservative or not, this was a highly political analysis based on the capacity of a conceptual shift to generate – or at least support – qualitative improvements in global (in)security. Although the concept of desecuritisation is not yet invoked, it is present in the analysis as Wæver couples the desirability of a move away from security and into the realm of politics to ‘détente’. ‘Détente’, moreover, is both a term through which political and academic debate is generated, and a process that describes relations between the two blocks. Comparing ‘détente’ to later examples of desecuritisation, it might be said to retain some element of ‘security’ insofar as it is a concept/process through which long term ‘non-security’ might eventually be attained. Détente opens up space for politics, both within and between the two blocks, but the fact that the conflict still looms in the background implies, that (re)securitisations remain a possibility. If one day, conflict completely disappears, so would, one assumes, the concept of détente. And so it did.
After the Cold War, as desecuritisation comes into its own as a concept, it is applied by Wæver to new contexts and conflict constellations while the concept retains its political status. First, the idea that conflict is avoided, not by great powers agreeing on one definition of ‘Europe’, but through different, yet compatible, ‘Europes’, entails an ethos of recognition similar to that of détente: it is through accommodating others' visions of themselves rather than enforcing one's own, that one avoids securitisations.Footnote 73 Second, the use of desecuritisation to capture how the EU encouraged prospective East European member countries to solve cross-border minority problems without resorting to force or high pitched security discourse echoes a similar political-normative preference for peaceful conflict resolution.Footnote 74 And third, the invocation of Europe's past as ‘the Other’ – and with that a securitisation of the possibility that intra-European conflicts might return – is an example of a ‘responsible securitization’, that preempts ‘numerous instances of local securitization that could lead to security dilemmas and escalations, violence and mutual vilification’.Footnote 75 The possibility of desecuritisation being applied to different contexts and dynamics is further heightened by the way in which desecuritisation has been applied by scholars beyond Copenhagen, applications to which we now turn.
Applying desecuritisation: four political forms
In the light of the recovery of the political and normative substance of desecuritisation above, the second part of the article now turns to how desecuritisation has been constituted through empirical applications. I suggest, more specifically, that one can identify four forms of desecuritisation each of which invoking different understandings of the public sphere and the status of the amity-enmity distinction. The elaboration of each form starts with a presentation of analysis that deploys it, then the particular political dynamic that come with this form of desecuritisation is discussed before the specific political and normative questions that arise within this form of desecuritisation are brought up.
The four forms of desecuritisation can be applied at three different levels. First, they can be used to characterise the ontological, political positions, that theories adopt. At this level the four forms of desecuritisation constitute deeply held normative-political views about the ontological status of security, politics, community, authority, and so on, and as such they describe not only what has happened in the past or the present, but what is possible and might become. This implies further, that the forms cannot meaningfully be tested against the empirical. The deliberation of the four forms of desecuritisation at this deeper normative-political level underscores, that the goal of this article is not to arbitrate between them, or to decide, for instance, whether a Schmittian understanding of desecuritisation as replacement is superior to a Habermasian one, but rather to sort out what positions are adopted and how they differ. Second, the four forms of desecuritisation might be used analytically to identify the trajectory of a given instance or case of desecuritisation. Third, the four forms of desecuritisation each generate more specific political and normative questions. The four forms are ideal types, which means that individual security theorists might articulate understandings of (security) politics at level one which does not fit completely into one of the four forms. Or, empirical analysis of specific cases at level two may combine two or more of the four forms. Yet, given that the typology is drawn from the debates over and applications of securitisation theory, my claim is that we are unlikely to find – at least at the moment – instances of desecuritisation, which cannot be analysed through these four forms of desecuritisation.
Change through stabilisation
Let us begin with the form of desecuritisation for which the concept was originally designed, namely that of détente during the Cold War. First, however, it might be worthwhile considering whether this form of desecuritisation has relevance beyond the context for which it was developed. As argued above, this form of desecuritisation implies a rather slow move out of an explicit security discourse, which in turn facilitates a less militaristic, less violent and hence more genuinely political form of engagement. It also requires that parties to a conflict recognise each other as legitimate. Hence Wæver's description of détente as change through stabilisation. Surveying the desecuritisation literature, détente seems to have died so completely that it has disappeared from political and academic use, not only as a concept, but as a way to understand desecuritisation. The irony is thus, that while desecuritisation has its genesis in détente, this form of desecuritisation has been little invoked subsequently. One obvious explanation of this absence is that the political context has changed to such an extent that it is no longer empirically relevant. Change through stabilisation took place on the backdrop of the Cold War, a global conflict which was widely seen as unsolvable, and the only ‘macro-securitization’ that might be considered a post-Cold War replacement, namely that of the Global War on Terror, does not entail a similar space for recognition and accommodation as did détente.Footnote 76 The War on Terror might be receding, but this does not evolve around an acceptance of ‘terrorists’/Al-Qaeda, but through a discourse that holds that Afghanistan – and Iraq before it – has been sufficiently stabilised, that ‘the West’ can leave the responsibility for security to local forces.
Yet, while there is no macro-desecuritisation equivalent to that of détente, this form of desecuritisation might still be identified, for instance in cross-border minority conflicts. Take Wæver's case of the desecuritisation of Hungarian minorities.Footnote 77 One interpretation of this case is that security problems have been solved, possibility so successfully that the issue has disappeared altogether.Footnote 78 Another, less optimistic, interpretation is that the issue was desecuritised in a manner akin to détente: it receded into the background, but it staying there depended on continued recognition and accommodation on both sides.Footnote 79 The tensions between Hungary and Slovakia in early 2011, as the former granted citizenship to thousands of ethnic Hungarians in neighbouring countries – a move which was seen by the latter as nationalist or even irredentist – might support such an interpretation.Footnote 80
Détente, we might recall, was also a concept used in political discourse as a ‘counter-concept’ to deterrence – for instance by German Social DemocratsFootnote 81 – and it thus allowed for a self-reflexivity at the level of the concept itself: ‘we need to rethink nuclear deterrence, and doing so through the concept of détente shifts our understanding of the parties, their dynamics, and hence of the conflict itself’. Although she adopts the concept of emancipation rather than desecuritisation, one might read Aradau's call for shifting the discourse on sex trafficking from one of ‘victims’ or ‘migrants’ to ‘workers’ in a similar vein.Footnote 82 Both ‘victims’ and ‘migrants’ are established terms within discourses that securitise ‘sex trafficking’, whereas that of ‘worker’ facilitates a move to a political-economic domain. There is also a political-normative resonance between Aradau's ‘worker’ and Wæver's ‘détente’ in that both seek, through a move out of the securitised, to create political space for individuals and groups threatened by security elites, be that the East European dissidents of the Cold War or the trafficked women of the twenty-first century.
The most pressing political-normative issues that arise from change through stabilisation concern this desecuritisation's conservative, system-stabilising character. Even if an embrace of this desecuritisation is strategic, rather than built on a deeper ethical acceptance of an illegitimate regime, the immediate and medium-term consequences might be hard to justify. One's assessment of this issue takes us back to one of the key elements of the concept of politics discussed above, namely the extent to which politics is understood as an ‘unexpected process which implies changes of identity and interests’.Footnote 83 If politics is the unexpected, there is a chance that détente might indeed do more than ‘just’ push existing conflicts into the background, yet, whether such more dramatic change is to be applauded is then another question.
Replacement
Replacement theorises desecuritisation as the combination of one issue moving out of security while another is simultaneously securitised. Pinar Bilgin for example holds that Turkey's post-1999 transformation was not one of universal desecuritisation, but rather ‘was rendered possible through claiming “security-speak” by certain societal actors who framed other issues … in security terms, and who pointed to Turkey's EU accession as a solution to its security problems’.Footnote 84 Bilgin does not claim that this was an inevitable process and thus does not subscribe to replacement as a basic political ontology on level one. Since most security theorists would agree that we can find examples of actors who have removed one issue from their security agenda while other issues have been included, the key question is whether replacement is seen as inevitable, that is as a political ontology on level one. The most thorough argument in favour of seeing desecuritisation as necessarily followed by replacement is made by Behnke, who holds that the fading of one issue or actor from security discourse will lead to others being securitised. In his words, ‘At some point, certain “threats” might no longer exercise our minds and imaginations sufficiently and are replaced with more powerful and stirring imageries’.Footnote 85 Behnke builds this position on Schmitt and his claim, laid out above, that politically communities cannot exist without the distinction between itself and an enemy. This position is, holds Behnke, further supported by post-structuralists like David Campbell and Michael Dillon who have demonstrated ‘the way in which states constantly produce and reproduce their national identities through discourses of in/security’.Footnote 86 Several empirical studies of desecuritisation have supported this understanding of desecuritisation. Paul Roe holds that minorities require securitisations in order to exist, at least in Central and Eastern Europe,Footnote 87 and Aras and Polat's analysis of Turkey shows that relations with Syria and Iran have been desecuritised, while other issues, Northern Iraq for instance, are becoming securitised.Footnote 88
As shown above, the Copenhagen School holds that Schmitt's understanding of the Political is not the only form of politics possible, which might suggest that desecuritisations are not necessarily followed by new securitisations. On the other hand, as the Copenhagen School is not explicitly saying that a society might exist without (wanting the option of) securitisation, replacement is not overruled as an empirical phenomenon or an ontological position. Wæver's claim that Europe's Other is its own past might, for instance, be read as subscribing to the view that Others are required to keep communities and political projects together.
Assuming that Others are required generates a set of more specific analytical and normative questions. For one thing, one needs to look more closely at what forms of otherness appear and disappear, and what this implies for the public sphere. In Wæver's example, for instance, the securitisation of Europe's past facilitates the prevention of a (re)securitisation of relations between states. This is a securitisation that blocks violent conflicts and hence invokes a rather different friend-enemy distinction than that of inter-state war. It is also a form of securitisation that is centred on the past of the Self which facilitates – perhaps even requires – a self-reflexivity that an inter-state securitisation may not. While a securitisation of one's past is thus a form of securitisation, it is one that intersects with both the friend-enemy distinction and the public spheres in ways that sets it apart from a purely Schmittian position. This, in turn, raises the question whether such a ‘past as Other’ securitisation is in fact sufficiently ‘friend-enemy’ like to ‘work’ as a foundational move. In terms of the normative and political issues that replacement raises one should, in other words, critically interrogate whether shifts from one securitisation to another involve a qualitative shift, as with Europe's past, or merely a move within the same threat-otherness dynamic, for instance by substituting one state or domestic group for another.
Because Schmitt and post-structuralism play central roles in debates over securitisation theory, it should be added that it is possible to see the two as more distinct than suggested by Behnke. For one thing, while post-structuralists like Campbell, Richard K. Ashley, and R. B. J. Walker stress the difficulty of moving beyond the dichotomies of Self/Other and inside/outside, they simultaneously identify openings, ambiguities, or movements that defy them. Walker, for example, calls for a ‘triple move against Schmitt’ that includes a questioning of his view of the modern subject as ‘able to draw the line cleanly around itself’ and Walker refuses ‘the assumption that the world of modernity that Schmitt takes for granted is indeed the world’.Footnote 89 Another indication of the possibility of a world without securitisation is Costas Constantinou's use of ‘poetic exploration’ to develop ‘narratives of security that resist securitization; i.e. narratives that do not offer rhetorical legitimation to different regimes of power or justify the intervention of security experts and practitioners.’Footnote 90 Or, we might turn to Behnke himself who in his earlier work invokes Derrida and Machiavelli in a call for a ‘de-naturalization of political reality’ which opens up for ‘effective agency and the possibility of ethical intervention’ and moving beyond the ‘defensive, static logic of the Fortress’.Footnote 91 More could be said about the positive content of post-structuralism, here the point is merely that it envisions a move out of the logic of replacement, at least as defined by Schmitt.
Rearticulation
Rearticulation refers to desecuritisations that remove an issue from the securitised by actively offering a political solution to the threats, dangers, and grievances in question. Comparing rearticulation with change stabilisation through and replacement as forms of desecuritisation at level two (the trajectory of a given issue), rearticulation suggests a more direct, radical form of political engagement: there is no conflict looming in the background as with change stabilisation through, and the issue is rearticulated rather than just replaced. At level one, rearticulation refers to fundamental transformations of the public sphere including a move out of the friend-enemy distinction. Put differently, rearticulation as a basic political ontology makes a claim about the ability of system-wide securitisations to be resolved, which sets it apart from the inevitability of a Schmittian replacement or the more pessimistic – at least in the short run – view of change through stabilisation. It is thus possible to combine a level one ontology of replacement with rearticulation at level two: one identifies the transformation of a specific issue as a political solution is found, and this in turn is believed to trigger the need for new securitisations. Rearticulation as a political ontology at level one entails by contrast such fundamental transformations of the public sphere, and of the identity and interests of Selves and Others, that this in itself offers an antidote against new securitisations at level two.
Åtland's account of Gorbachev's Murmansk Initiative, launched in 1987, is one example of a rearticulation where a politician pursued a new agenda in the attempt to remove non-military issues ‘from his country's national security agenda and reintroduce[d] them to the sphere of “normal politics”’.Footnote 92 That this was part of a fundamental rearticulation was evidenced by Gorbachev's recasting of the entire East-West conflict. To take another example, one might point to the successful resolution of the Danish-German border and minority conflict. A referendum in 1920 and subsequent minority policies on each side of the border have facilitated a move from the securitised (epitomised by the wars of 1848–50 and 1864–4) to the politicised (around and in the decades following the referendum) to what is today arguable a depoliticised border.Footnote 93 This is also an example that shows the possibility of rearticulation as a wider societal transformation in that the resolution of the border problem took place without German and Danish minorities turning to new securitisations of ‘their’ minority identity. Both the public sphere and the friend-enemy distinction were, in other words, fundamentally changed.
The examples of Gorbachev and the Danish-German border show a desecuritisation dynamic that is undertaken voluntarily – or at least this is how events are narrated in hindsight. In this respect, they might sound like a Habermasian's dream come true: previously antagonistic actors realise that their own and others' survival and interests are better served through collaboration, accommodation, and negotiation than by securitising the other side. In reality, however, the process might be less smooth or parties' voluntarism might be the product of more complicated power dynamics. Wæver's example of the EU applying pressure on prospective member countries such that they resolve minority issues through desecuritisation is one indication of the complicated ways that power, persuasion, and self-reflexivity might intersect. One might also see Gorbachev's moves in the mid-1980s as propelled by the economic and political bankruptcy of the Soviet regime, and the referendum in 1920 as the least-worst scenario for a Germany defeated in World War I. An ontology of rearticulation can, in other words, encompass not only a Habermasian position, but also more conflict and power-oriented perspectives.
As a vision of politics, rearticulation comes across, at least at first, as inherently positive: it does not have the conservatism-problem of change through stabilisation or the challenge of new securitisations materialising as does replacement. The most pressing normative and political questions that arise for rearticulation are thus related to the stability and desirability of this form of desecuritisation. As politics is dynamic, one cannot in principle ever define a conflict as inherently solved. As shown by numerous cases, history is a dynamic material that political actors mobilise, and conflicts that have appeared as settled may be brought up again and securitised. In this respect, there is an uncertainty that rearticulation is always up against and which introduces an ontological contradiction at level one: it claims a finality, yet finality is inherently impossible. Rearticulation may thus be in need of supporting discursive practices that invoke specific understandings of past conflicts, while suppressing – or addressing – the desire for new securitisations. Rearticulations also face a political-normative question, that is, do ‘we’ actually want a particular issue to be presented as resolved rather than securitised? As noted above, there is no universalism that securitisation theory can invoke in defence of desecuritisation, which means that instances of rearticulation must be justified on substantial grounds. One specific form that challenges to the desirability of rearticulations could take would be to question whether the basic grievances on which a conflict was built have in fact been addressed. Or, is a rearticulation, in other words, more akin to a silencing than a genuine resolution?
Silencing
The final form of desecuritisation that one encounters in the literature applying securitisation theory is that of silencing, that is when an issue disappears or fails to register in security discourse. In MacKenzie's study of female soldiers and the post-conflict reconstruction process in Sierra Leone, we find desecuritisation as a strategy of exclusion. MacKenzie documents the way in which the international community and local partners have constructed the female subjects as victims, as ‘abductees, camp followers, domestic workers, and sex slaves’.Footnote 94 The silencing of female soldiers that ensues is defined by MacKenzie as desecuritisation, and her analysis shows that male combatants are securitised as they are linked to a securitisation of the return of conflict.Footnote 95 Male combatants receive, as a consequence, funding through programmes aimed at facilitating a return into post-conflict society, thus it is quite obvious that ‘silent desecuritizations’ work – at least in this case – to the disadvantage of those-not-securitised.
Considering the ontological and epistemological status of the speech act in securitisation theory, silence as desecuritisation might well seem the trickiest form of desecuritisation to fit into the theory. MacKenzie's study traces the disappearance of female combatants, thus a move from the securitised to the non-politicised rather than the political proper. As a consequence, desecuritisation is constituted as normatively and politically problematic rather than – as is usual within securitisation theory – as desirable. A key question is whether MacKenzie's usage shows a move ‘out of security’ or the prevention of security speech, before it would ever arise. As we recall Buzan et al.'s definition, desecuritisation indicates ‘the shifting of issues out of emergency mode and into the normal bargaining process of the political sphere’.Footnote 96 If desecuritisation becomes defined as everything that does not register as securitisation, then the concept risks loosing its analytical and political purchase: the world is (fortunately) full of issues and events that never have and never will be even close to ‘security’. Yet, a more critical-Foucaultian reading of MacKenzie's case study suggests that female combatants fail to fall into the category of ‘never securitized’, but rather that they have a history of securitisation that runs through their inclusion in the wartime category of ‘combatants’ and ‘armed forces’. Their active exclusion from the general subject of ‘combatants’ (which was securitised during conflict and which is securitised today) thus constitutes both an analytical and a political-normative link to securitisation theory.
Of the four forms of desecuritisation, silencing is the one that stretches securitisation theory the most. Analytically, because while the Copenhagen School acknowledges that securitisations might be institutionalised to such an extent that there are no explicit speech acts, it does not offer a theory or a methodology for how such ‘silent institutionalisations’ might be identified.Footnote 97 As a consequence, there are also unanswered political and normative questions. First, to what extent is the understanding of the political sphere that securitisation theory adopts sufficient to allow subjects like the female ex-combatants of Sierra Leone to enter? What are, in other words, the assumptions about the individual subject that underpin the understanding of the political sphere as one of bargaining and persuasion? And, what happens when such forms of politics fail? Second, desecuritisation as silencing straddles the securitised, politicised and non-politicised. In MacKenzie's study, there is a securitisation of the return of conflict, which facilitates a politicised engagement with former (male) combatants whose reintegration into society is supported, yet there is also a depoliticisation of female combatants. As such, it calls for further theorisation of the relationship between the three spheres of the securitised, the politicised and the non-politicised.Footnote 98
Conclusion
Securitisation theory is one of the most popular widening approaches to security, particularly in Europe, and its concepts of securitisation and desecuritisation have generated political theory debates as well as a long list of applied studies. These two bodies of literature have to a large extent run on separate tracks, yet, both are in fact concerned with the status of politics, particularly with the relationship between security politics on the one hand and ‘normal’ politics on the other. As the concept of desecuritisation describes the move from the securitised to the politicised, it is not surprising that it has been the subject of heated theoretical debate and that the empirical use has featured numerous conceptualisations. This article has counted the argument that securitisation theory has little or no conception of politics – and hence no normative stance – through four readings. The first reading clarified that securitisation theory relies upon a Schmittian understanding of the Political, but that this understanding underpins Buzan and Wæver's conceptualisation of security, rather than politics as such. Nevertheless, this reliance makes securitisation a highly political concept. The political status of securitisation was then combined with a reading of desecuritisation as a Derridarian supplement which brought out the constitutive and political status of desecuritisation. The second reading examined how desecuritisation reworked the public sphere and the friend-enemy distinction in ways that resonated with a Habermasian discourse ethics, but which also entailed a more critical post-structuralist understanding of subjectivity and power. The third reading emphasised the role of responsibility, choice and decisions, particularly in Wæver's work, and brought out the inherently political status of the choice to (de)securitise. It was shown, moreover, that Wæver's position draws upon Derrida's critique of universalism which further accentuates the unpredictability of political decisions. The fourth reading traced desecuritisation to the historical context from which the concept rose, namely that of Cold War détente. Not only did this show Wæver's engagement with concrete political issues, it allowed for a discussion of the political dynamics attached to détente or ‘change through stabilisation’. The last part of the article turned to the empirical applications of desecuritisation reading these through the discussion of the concept of politics in desecuritisation laid out in the first part of the article. Four forms of desecuritisation were presented, each form presenting us with different ontological views of politics and a set of more specific political-normative questions.
It has been an underlying premise of this article that desecuritisation is a concept worth engaging. An important explanation for the concept's success is undoubtedly that it zooms in on one of the quintessential questions of political life: should we treat this as a matter of danger and exceptionality or is it not worthy of this labelling or better dealt with if we conceive of it in less fearful terms? This article has attempted to provide further signposting for those engaged in desecuritisation debates as well as for those who wish to use it empirically. This is not to suggest that there is no more to say about desecuritisation. The debates over Schmitt and the concept of the Political/political are likely to continue, one might envision further discussions of the specific questions that each form of desecuritisation engenders and one might interrogate whether there are sector specific forms of desecuritisation. Turning to the more applied literature, further studies might trace whether desecuritisation occur empirically in ways that blur the four ideal-types or generate new ones.