4.1 What is Security?

As noted above, constructivism gained momentum after the Cold War era when new methods of analysis seemed to provide more appropriate ways of understanding international politics than those delivered by traditional approaches. At the same time, years of scholarly debate had led to a broadening of the concept of security. Whereas security studies had hitherto largely been bound to the military sector, it now increasingly included other aspects of life that it had previously neglected, such as economy, society, poverty, human and minority rights, environmental and climate change.52 The Copenhagen School gained great prominence within this new realm of constructivist security studies. The School’s approach, which centers on the works of Barry Buzan and Ole Wæver, provided an innovative framework for the analysis of security (Wæver et al. 1993; Wæver 1995; Buzan et al. 1998; Buzan/Wæver 2003, 2009). The aim was to broaden the concept of security while ensuring that it did not lose its inner logic, as this would have rendered the concept meaningless. Although the Copenhagen School developed three central concepts, its most influential conceptual apparatus is that of ‘securitization’.53 This concept focuses on the question of how security issues become what they are: how and by whom they are constructed and how security invests actors with the power to take exceptional political action. The concept of securitization sparked a tremendous amount of security studies research in IR, and the vigor of this approach is demonstrated by the scope of publications in which the concept has been applied. Securitization lent a new dynamic to a field that had otherwise been regarded as rather reactionary. The importance and innovative character of the concept, however, also called for scrutiny, and in turn triggered vibrant debates about theoretical assumptions in security studies. This led the leading figures of the Copenhagen School, Barry Buzan and Ole Wæver, to acknowledge “that there are many ways to understand security” and that their approach had merits and drawbacks just like any other (Buzan/Wæver 1997: 249f). The following outlines the concept of securitization and mainly draws on the Copenhagen School’s central work, the 1998 book Security: A New Framework for Analysis, which sets out the most systematic formulation of the model.

4.2 The Concept of Securitization

In his scholarly piece from 1995 entitled Securitization and Desecuritization, Ole Wæver expressed his discontent with the basic premises of the traditional approach to security studies that argued, “security is a reality prior to language, is out there” (Wæver 1995: 46). After all, if language constitutes the social world, this must be true for security as well. Wæver’s criticism hints towards the cornerstone of the securitization approach: security is understood as a social, and in this sense primarily verbal, construction rather than an objective condition. Threats are not regarded as ‘given’ but as constructed by being labeled as such. The concept of security is thus viewed as discursive in character, and the Copenhagen School investigates both the rules of ‘speaking security’ and how the power of ‘spoken security’ unfolds. In other words, the focus is on how security issues are constructed and how they are put into practice in terms of extraordinary measures. The CS believes that there is a “specific rhetorical structure” (Buzan et al. 1998: 26) to security speech acts which gives security politics a characteristic pattern. Moreover, this approach highlights that the way people and issues are framed is vital if an environment of approval for certain policies is to develop.

In its analysis of how security threats are constructed, the Copenhagen School relies on Austin’s speech act theory (see section 2.7). As demonstrated above, the main insight provided by speech act theory is that speaking is rule-guided action that possesses reality-making power. Words do things. This performativity of speech is what the Copenhagen School refers to when accentuating the constitutive role of security language. What the authors focus on and claim that is special in this realm of speech acts is that “security is about survival” (Buzan et al. 1998: 21; Wæver 2003: 9). Security thus means presenting an issue “as posing an existential threat” (Buzan et al. 1998: 21) to a designated object – such as the sovereignty of a state, the identity of a societal group or the natural environment – and “by saying the words, something is done” (Buzan et al. 1998: 26). It is important to note that what the Copenhagen School points to is the linguistic representation of the threat, rather than its ‘real’ existence. In this understanding, “the utterance is the primary reality,” and this cannot be judged as true or false; it does not have to refer to “something more real” (Wæver 1995: 55, cf. 2003: 10; Buzan/Wæver 1997: 246), although speech acts can be fortified by ‘facilitating conditions’ (see below).

Through the representation of an issue as an existential threat, a shift of assessment takes place. The internal urgency of securitizations “takes politics beyond the established rules of the game” (Buzan et al. 1998: 21; cf. Wæver 2003: 10). Whereas politicized issues are viewed as the body of everyday politics with debates and decision-making processes according to common rules, a securitized issue is regarded as “above normal politics” (Wæver 2003: 10; cf. Buzan et al. 1998: 26) as it justifies actions beyond the usual limits of political procedures (Buzan et al. 1998: 23f, 29). By ‘speaking security’, a securitizing actor “moves a particular development into a specific area, and thereby claims a special right to use whatever means are necessary to block it” (Wæver 1995: 55). The invocation of security is thus key to legitimizing the breaking of rules and the use of force (Buzan et al. 1998: 21, 25).54 As this indicates, securitizations have political consequences. By turning an issue into a matter of existential threat, it is impossible not to deal with it; instead, the issue is granted precedence. Since security problems are threatening in “a particularly rapid or dramatic fashion” (Wæver 1995: 54), immediate action by top leaders is imperative, otherwise it will be too late and everything else will seem irrelevant (Wæver 1995: 53; Buzan et al. 1998: 26). Securitizations thus claim an “unavailability of choice” (Wæver 2003: 25) as they pose the necessity to act. In turn, this leads the majority of everyday political options, such as negotiation and diplomacy, to be left out. In the face of an allegedly lethal threat, the range of possible options is restricted to the most severe actions. Such scenarios are usually presented in dramatic fashion and are overstated. However, when “master signifiers” (Buzan/Wæver 2009: 267) are used, like ‘terrorism’, for example, or ‘war’, or other issues from the military sector, audiences understand without elaboration that the matter referred to implies danger, urgency and the need to combat it (Buzan et al. 1998: 26–29; Buzan/Wæver 2009: 267). Securitizations are regarded as successful when the verbal construction of an existential threat leads (at least) to deliberation about emergency action that affects ‘others’ (Buzan et al. 1998: 26).

In order to expose the Copenhagen School’s focus on “who can securitize what under what conditions and with what effects” (Wæver 2000: 252; cf. Buzan et al. 1998: vii), the following discusses the views of the School’s most influential theorists in detail.

• – Securitizing Actor: A securitizing actor is an enunciator who “performs the security speech act” (Buzan et al. 1998: 40), that is, who attempts to securitize an issue by declaring something – a referent object – as existentially threatened (Buzan et al. 1998: 36). Initially, the securitizing actor makes a ‘securitizing move’. Whether this turns out to be successful depends on the audience and on contextual factors (see below). In principle, the concept is open to securitizing actors of all kinds, however, the possibility is limited by the supposition that the securitizer has institutional authority (Wæver 1995: 57; Buzan et al. 1998: 32f) and that the referent object in security discourses is first and foremost the state (see below). In other words, securitizers do not have to be a member of the political elite, but they usually are. As the Copenhagen School’s authors put it, “common players in this role are political leaders, bureaucracies, governments, lobbyists, and pressure groups” (Buzan et al. 1998: 40). These securitizers can act as individuals or as groups. As they are representatives of public institutions, they are not viewed as private individuals but as authorized agents of the collectivity for which they speak (Buzan et al. 1998: 40f). Securitizing actors and referent objects can also correspond, such as when states are concerned. In these cases, “the referent object will in a sense speak for itself through its authorized representatives” (Buzan et al. 1998: 42).
• – Referent Object: The referent object is that which is said to be under existential threat and, since it has “a legitimate claim to survival” (Buzan et al. 1998: 36), in need of protection. In theory, this could be anything; effectively, however, it is primarily a state or nation, where state security means defending sovereignty and national/societal security to preserve identity (Wæver et al. 1993: ch. 2; Wæver 1995: 47–49; Buzan/Wæver 1997: 245; Buzan et al. 1998: 36–40).55 Although securitizations can occur on the individual and the system level, the Copenhagen School argues that in practice neither the micro nor the macro but “the middle scale of limited collectivities” is most accessible for securitization (Buzan et al. 1998: 36; cf. Buzan/Wæver 2009). The reason for this is the strong collective feeling within a state or nation that shares the same values and principles and thus “forms an interpretative community” (Buzan et al. 1998: 37). Just as classical security studies traditionally attributed a projecting role to the state, here, too, it often appears in the form of the referent object. In its subsequent theory, the Copenhagen School focused more on the macro level, though, and introduced macrosecuritizations to its model of overarching conflicts. Macrosecuritizations work in the same way as securitizations, but are more complex in structure (they bind together securitizations from lower levels) and are more universal in their claims. Macrosecuritizations refer to the principles of a ‘higher order’ that potentially refers to the whole of humanity, such as ‘civilization’ or ‘freedom’, and are apt for mobilizing other actors on the international level (Buzan/Wæver 2009).
• – Audience: The audience is the group of people that the securitizing actor addresses, such as parliament or the public. The role of the audience is decisive as it defines the success of a securitization. It is only if an audience accepts that something poses an existential threat to a shared principle or value, that securitization can be successful (Buzan et al. 1998: 31; Wæver 2000: 251). If this does not occur, the securitizer’s endeavors remain just a ‘securitizing move’.56 The acceptance of a securitizing move is thereby not necessarily embodied by the actual implementation of emergency measures but by the debate “from which it is possible to legitimize emergency measures” (Buzan et al. 1998: 25; cf. Wæver 2003: 11). Although the Copenhagen School does concede that there could be pressure on the audience during the decision-making process, it believes that the acceptance of a securitizing move cannot be reached by force. As the CS notes, “securitization can never only be imposed, there is some need to argue one’s case” (Buzan et al. 1998: 25). This also underscores securitization as an intersubjective process.
• – Facilitating conditions: As stated above, the audience is crucial for the success or failure of a securitizing move, but acceptance is also influenced by what are known as facilitating conditions. The Copenhagen School acknowledges that successful securitizations are more likely under certain circumstances (Buzan et al. 1998: 57). While the nature of the concept is in principle unlimited, the structure of securitization limits the scope of possibilities. In this respect, the Copenhagen School distinguishes two categories. The first category, the “internal, linguistic-grammatical” (Buzan et al. 1998: 32) refers to the form of the speech act that must be presented in a comprehensive and convincing way. In line with Austin’s speech act theory, the speech act must follow “accepted conventional procedures” (Buzan et al. 1998: 32). Moreover, the illocutionary act has to follow a certain dramaturgy, that is, the “grammar of security” which means, “construct[ing] a plot that includes existential threat, point of no return, and a possible way out” (Buzan et al. 1998: 33; cf. Wæver 2000: 252, 2003: 14).

The second category, the “external, contextual and social” (Buzan et al. 1998: 32) refers to the particular people and circumstances involved. The securitizer must be in a position of authority. In principle, the maxim holds true that the greater the authority of the enunciator, the greater the prospects for successful securitization. Wæver, in particular, emphasizes the decisive role played by elites: “By definition, something is a security problem when the elites declare it to be so” (Wæver 1995: 54), or elsewhere, “a problem would become a security issue whenever so defined by the power holders” (Wæver 1995: 56, emphasis in original). Consequently, although everyone can become involved in an issue, less powerful actors have fewer chances of successfully securitizing it (Buzan et al. 1998: 31). Besides the status of the securitizer, the circumstances surrounding the threat scenario also influence the prospects of success. Even if no ‘real’ danger has to be present, it is useful when something is ‘out there’ that corroborates the assertion that a situation is perilous. As such, if a securitizing move is enacted within a context of “conditions historically associated with a threat” such as “tanks, hostile sentiments, or polluted waters” (Wæver 2000: 253, cf. 2003: 15; cf. Buzan et al. 1998: 33), it is more likely to be successful. Taken together, whether securitizing moves are successful depends on conditions and context. Facilitating conditions can heighten their chances of success. However, these conditions also pose constraints that limit the scope of possible securitizations by confining who can act as a securitizing actor and which referent objects can be considered feasible. The Copenhagen School concludes that this predestines collectivities to be treated as units on behalf of which security action is taken (Buzan et al. 1998: 36).

Although much emphasis has been placed on the securitization concept, the Copenhagen School actually advocates the opposite: desecuritization (Wæver 1995, 2000: 253; cf. Buzan et al. 1998: 4, 207–209). The main aim of CS is to keep issues in the realm of ‘normal’ politics and to forestall securitizations in order to prevent escalation and security dilemmas that lead to violence (Wæver 2000: 251–254, 2003: 23). Therefore, security is not appraised as an outright positive value that should be maximized, because when doing so actors “eliminate other, potentially more useful ways of conceptualizing the problems being addressed” (Wæver 1995: 57). The reasoning is that securitizations, as discussed in section 3.4, are constructions of exclusion. They alter the political decision-making process in a way that certain options and remedies become excluded from the discourse. This, however, could worsen rather than solve a security problem. At the same time, securitizations confine who is authorized to speak. They are therefore considered anti-democratic as well as susceptible to manipulation and abuse, as they “silence opposition” and have “given power holders many opportunities […] to handle something with less democratic control and constraint” (Buzan et al. 1998: 29; cf. Wæver 1995: 55, 2003: 12). As a consequence, securitizations are seen negatively “as a failure to deal with issues as normal politics” (Wæver 2003: 12; cf. Buzan et al. 1998: 29). Desecuritization, on the other hand, is perceived as “ideal” (Wæver 2000: 251) and “the optimal long-range option” (Wæver 2003: 12; cf. Buzan et al. 1998: 29) because it moves an issue out of the logic of threat/defense and restores the possibility of dealing with the situation through normal political procedures. Desecuritizing and thus shifting a topic out of the emergency mode works by abandoning specific vocabulary: “Transcending a security problem by politicizing it cannot happen through thematization in security terms, only away from such terms” (Wæver 1995: 56, emphasis in original).

It is obvious that the Copenhagen School is normative in its approach. In order to circumvent securitization and strengthen politicization, the CS aspires to raise actors’ awareness of their responsibility for their actions. Those involved in securitization should not believe that it represents “an innocent reflection of the issue being a security threat” (Buzan et al. 1998: 29, emphasis in original). As speech acts generally include the option to say things differently, securitizations are always political choices (Buzan/Wæver 1997: 246; Buzan et al. 1998: 29; Wæver 2003: 24). The Copenhagen School considers them as intentional, strategic actions aimed at giving a discourse a specific direction. Uplifting an issue to a level above ‘normal’ politics is regarded as the very reason behind securitizing moves conducted by political leaders who seek justification and legitimation for projected emergency action (Wæver 1995: 63, 2000: 251; cf. Buzan et al. 1998: 21). If securitization embodies a choice for deliberate proceedings, actors can be made responsible for their conduct. Similarly, the Copenhagen School points to audiences and their responsibility for accepting and reproducing securitizations. Instead of doing so, however, the CS calls for securitizations – the framing of a situation as a lethal threat and the imperative to enact emergency measures – to be challenged (Buzan et al. 1998: 29; Wæver 2003: 24f). Finally, with regard to responsibility, the Copenhagen School also addresses the role of scholars. By conducting research, academics can increase the risk of furthering securitization, because “even when talking security in order to achieve de-securitization, it is possible that one contributes to securitization by the very fact of producing more security talk” (Wæver 2000: 252, cf. 2003: 29). Critical awareness, therefore, is also required from researchers, even if it is impossible to avoid this problem.

4.3 The Copenhagen School and its Critics

Although the concept of securitization was praised in IR for its significance to the field of security studies, it was also placed under scrutiny. The engaged and often sophisticated debate about various aspects of securitization has led to a great body of work, embossed with critical discussion as well as the attempt to amend the framework of analysis.57 A main point of criticism is that the approach is still too traditional in its conceptualization of security as it strongly focuses on states and their representatives. This model has been criticized for treating threats as external to the state and thus taking the state as the relevant unit; for emphasizing the role of political elites and their defining power in designating threats, and for interpreting emergency measures primarily as national military action (McDonald 2008a: 579). In more general terms, most critics view security as a much broader social construction, and securitization as a more dynamic and complex process than the Copenhagen School’s perspective indicates (Trombetta 2011: 136). Many points of criticism thus refer to aspects of discourse theory, especially in terms of contextual features, and power relations – aspects that have been neglected by speech act theory, the basis of the concept of securitization. The following discusses some of these main points.

The model developed by the Copenhagen School is regarded as ontologically inconsistent. Although the authors claim to be social constructivists “all the way down” (Buzan/Wæver 1997: 243, fn. 6), critics have highlighted objectivist positions in the School’s reasoning. McSweeney (1996, 1999), for example, leveled a widely noticed critique at the Copenhagen School by pointing to its conception of societal identity. He argued that the School composed both society and identity in a “near-positivist” manner as preexistent “objective realities”, something which runs contrary to the notion of social construction (McSweeney 1996: 83, cf. 1999: ch. 4). In line with this, societal identity would be viewed as unitary, but this would disregard its multiplicity and fluidity.58 This exposed the Copenhagen School to a debate similar to that discussed in the previous chapter over ‘moderate’ and ‘radical’ constructivism. The reason for this is that the CS includes both forms of constructivism in its model. Whereas Wæver (2003: 15) rejects resorting to “rump materialism”, when referent objects are considered, the CS is in part close to Alexander Wendt’s position. In response to McSweeney’s critique, the Copenhagen School pointed out that it viewed identities as socially constructed, yet that identities “petrify and become relatively constant elements to be reckoned with” (Buzan/Wæver 1997: 243, fn. 6; cf. Wæver 2003: 28; Buzan et al. 1998: 35). These sedimented identities are thus stable enough to become referent objects for security. However, “this does not imply that identities do not change, only that we should not expect everything to change all the time” (Buzan/Wæver 1997: 243). Despite this ‘moderate’ constructivist stance, the Copenhagen School openly acknowledges a ‘radical’ constructivist view with regard to security speech acts (Buzan/Wæver 1997: 246). In this view, a security problem does not exist in and of itself but must be communicated as such. On this point, in works that Wæver (1995, 2000, 2003) has written by himself, the stance is even closer to a post-structuralist position as Wæver’s analysis is similar to Derrida’s famous claim that ‘there is nothing outside of the text’.59 In joint works with Buzan and other authors, however, this position is replaced by a more comprehensive view of security that adds audiences and ‘facilitating conditions’ as external parameters to the speech act (Stritzel 2007: 361f; McDonald 2008a, 2008b; Williams 2011b: 215).

A related point of criticism is that the Copenhagen School considers the logic of security, its surrounding conditions and patterns, as relatively fixed (Balzacq 2005). With its “specific rhetorical structure” and the characteristic triad of “survival, priority of action and urgency” (Wæver 2003: 10; cf. Buzan et al. 1998: 26), securitizations, it is argued, would become predictable. In fact, the Copenhagen School bases its call for desecuritization and for dealing with issues in the realm of ‘normal’ politics on the assumption of foreseeability. Critical security theorists and the second generation of critics (Balzacq 2005, 2011b; Stritzel 2007), however, challenge the timelessness and fixed character of the logic of security. They emphasize that the character of security can change in particular contexts and that this means that security discourses can differ in form and in the manner in which they shape politics. According to Balzacq (2005: 172), there is no universal principle that works in every setting, and because of that, the analysis of context is vital. The point here is that by regarding the social, political, and historical context in which securitizations are embedded one can better understand how they become possible. The reason is that it is only in relation to these contexts that the securitizing actor has authority and his/her speech acts have performative power (Stritzel 2007: 360). This also includes the reaction of the audience and its potential acceptance of the securitizing move as this, too, will vary according to the context.

Critics acknowledge that – technically – the Copenhagen School considers contextual features as ‘facilitating conditions’, yet they point to the fact that this part of the model is undertheorized. Possibly, the silence on this point is another expression of the Copenhagen School’s theoretical indeterminacy. With Wæver’s inclination towards the post-structuralist notion that ‘there is nothing outside of the text,’ it is problematic to consider external contextual features and to frame them – as the Copenhagen School does – as ‘conditions’. In fact, Wæver acknowledges that “dealing with conditions is highly sensitive” as it could lead to an introduction of causality and (even more) objectivism into securitization (Wæver 2000: 252, cf. 2003: 14). He emphasizes, therefore, that securitizations are socially constructed and not reducible to causal factors; ‘conditions’ merely have a facilitating influence but cannot replace political interaction (Wæver 2000: 252, 2003: 14; Buzan/Wæver 2003: 71f). On this point, Balzacq (2005, 2011b) believes that the Copenhagen School over-emphasizes security as a discursive product fairly independent of what is ‘out there’ and suggests that the CS pays greater attention to how external extra-linguistic contexts affect securitizations.60 Sweeping calls are being made for a refinement of the Copenhagen School’s concept. They imply that securitization should be viewed “as a strategic (pragmatic) practice that occurs within, and as part of, a configuration of circumstances;” these circumstances are viewed as belonging to the “external context, the psycho-cultural orientation of the audience, and […] the differential power between the speaker and the listener” (Balzacq 2005: 172, 174). However, such an interpretation would shift securitization towards a sociological process (Balzacq 2005, 2011b). Thought further, this would also mean examining how political communities themselves are constituted and how narratives of national identity become dominant; and it would mean investigating how these communities handle security issues. Moreover, this approach would demand that alternative discursive framings of security be recognized, framings that are either overruled (and if so, why?), or that have been delegitimized (and if so, how and by whom?) (McDonald 2008a: 568). It is obvious that such an approach leads to a great refinement of the Copenhagen School’s model, the more so as it takes diverse actors, audiences and social factors into account (Salter 2011: 117f). Although there are good arguments for this approach as it permits more precision, it either tremendously broadens the complexity of study, or forces scholars into tiny niches of research.

In addition to the fixed character of the security logic, the Copenhagen School has been criticized for only considering a particular point in time when looking at securitizations. Some reviewers see the developing character of securitization as undermined by the speech act concept. In their view, although securitizations actually embody the culmination of a (long-term) process that generated meaning over time, the Copenhagen School focuses on a particular moment of intervention in which a fixed meaning is applied to an issue or person. There is a history to securitizations, it is argued, and therefore, there is no sudden shift from the realms of ‘normal’ politics to that of ‘panic’ politics. Rather, this is seen as a sliding scale where an issue or person gradually shifts into the realm of security. Moreover, this need not end in “reaching the category of existential threat” (Abrahamsen 2005: 59; cf. Williams 2003: 521; Stritzel 2007: 366). Consequently, critics challenge the idea of an either-or-concept that views issues as either politicized, and thus in the realm of ‘normal’ politics, or as securitized and located in the realm of ‘panic’ politics and emergency action. Rather, politicization and securitization can co-exist (Neumann 1998; Williams 2003; Abrahamsen 2005; cf. Bigo 2000, 2001, 2002b). Moreover, as Donnelly (2013) explains, securitizations might not involve breaking free of the rules as assumed by the Copenhagen School. Instead, in the course of securitizations rules can be changed and redefined.

The Copenhagen School has taken note of some of these points, such as on time spans of development and the possibility of gradation. Although the School uses the terms ‘act’ and ‘process’ interchangeably (Stritzel 2007: 364) – and this could qualify the statement – the CS does point out that securitization is understood as “processes of constructing a shared understanding of what is to be considered and collectively responded to as a threat” (Buzan et al. 1998: 26, my emphasis). With regard to the framing of certain topics, there is a spectrum ranging from the non-politicized to the politicized to the securitized, and “depending on circumstances, any issue can end up on any part of the spectrum” (Buzan 1997: 14; cf. Wæver 2003: 10). This is also implied by (de-)securitizing moves that end up somewhere in between these poles.

In line with the traditional imprint of securitization, the Copenhagen School has also been criticized for clinging to a narrow – and beyond that, negative – concept of security. In so doing, the approach reifies the traditional security concept, critics say, and neglects other perspectives like human security (Doty 1998/99).61 In this realm, the debate also turns to the normative implications of retaining classical security logic. For example, only considering elites as securitizing actors could further strengthen dominant voices while silencing social groups that have already been marginalized. Critics therefore stress the risk of cementing power relations that this approach brings with it (Hansen 2000; Huysmans 2002; McDonald 2008a: 573–581; cf. Milliken 1999: 244f). It is fair to say, though, that the Copenhagen School demonstrates critical awareness of this point. Even without referring particularly to the voiceless, the CS does address the researcher’s responsibility, and, as outlined above, expounds the problem of furthering securitization by studying it (Wæver 2000: 252). Besides that, one of the reasons why the CS prefers desecuritization is that securitization is prone to instrumentalization by the powerful and thus tends to be undemocratic. Yet, for some critics, this normative position is also problematic. Proponents of Critical Security Studies and the Welsh School claim that the logic of security need not be negative per se. On the contrary, under certain conditions securitizations could be morally right, if they provide underprivileged groups with a voice and help to improve their circumstances. Used this way, securitization could further progressive and emancipatory change (Doty 1998/99; Williams 2005: 144; Booth 2007: ch. 3; McDonald 2008a: 580, 2008b: 71).62 The question therefore is not simply who has the power to securitize and convince the audience, but rather who gets the power to do so.63 The call on researchers, therefore, is not to promote this bias but to give a voice to the powerless (Milliken 1999: 244f). Wæver also acknowledges the possibility of meeting progressive ends through securitization but nevertheless advocates desecuritization (Wæver 2000: 285).64

Many of the shortcomings discussed so far relate to the point that the Copenhagen School is too selective in its approach and that it does not embrace the complex and multilayered process of securitization. In this regard, further debate has taken place about the ‘act’ of securitization, what it is and whether it comprehends more than just speech. This has led much to be written about the impact of images on security discourse. Today, it is argued, political communication is increasingly being transmitted by electronic media and thus, especially with regard to televisual coverage, through images (Williams 2003; cf. Hansen 2000: 300f). The 9/11 attacks provide strong examples of this: first, when the planes crashed into the World Trade Center in New York, and second, when the towers collapsed. These pictures were shown on television in endless repetition and powerfully corroborated shock and fear. Other authors point to the significance of photos (Möller 2007) and cartoons (Hansen 2007), universalized via the internet, as influences on perceptions of threat. These visual representations convey meaning and account for the assessment and potential acceptance of political action. The recommendation, therefore, is to integrate them into the Copenhagen School’s concept as non-verbal aspects of securitization, even if, as Vuori (2010; cf. Hansen 2007) points out, they might not bring about securitizations themselves. Rather, they unfold their impact within a previously established securitization. The inclusion of images in securitization, however, might be more problematic than it appears. The reason for this is that images are ambiguous in meaning (McDonald 2008a: 569f; cf. Möller 2007: 180; Vuori 2010). This means that it could be difficult to assess their effect and whether they were deployed strategically. In a more general sense, this would also call for an investigation into the role of journalists and in how far media professionals deliberately participate in securitizing moves. According to this view, a greater consideration of institutional media structures and their linkages to the political realm would be needed as part of analyses of securitization.

Other authors suggest considering the role of non-verbal practices in securitization processes. Neumann (1998), for example, recommends that the Copenhagen School directly refers to the outbreak of war and includes violence as actions with “a material character” (Neumann 1998: 19) into the model. He proposes extending Buzan’s continuum to include four steps: non-politicized, politicized, securitized, and violized. Wilkinson (2007: 12) points to the impact of censorship by the powerful or protest actions by the powerless. Huysmans (2011: 377; cf. Bigo 2002b: 73) even suggests that today securitization is more ‘act’ than ‘speech’ as discursive dramatization is displaced by technological surveillance, profiling, and data collection, which turns objects of everyday life into security objects. Similar to what has been discussed above, in this respect securitization is understood as a creeping process rather than urgent crisis management and rule breaking; however, this approach focuses on securitization of a different kind, its characteristic changes here. It is an investigation into the role of institutional environments in producing and managing threats. Huysmans’ interest in study is thereby close to that of Bigo’s (2000, 2001, 2002b) and the Paris School, which works on the institutionalization of security and how this leads to securitizations.65 The Paris School examines the security apparatus and the impact of its professional and bureaucratic actors, and in particular “the systems of meaning they generate and the productive power of their practices” (c.a.s.e collective 2006: 458; cf. Huysmans 2006: 153–155). In short, whereas the Copenhagen School focuses on the rhetorical structure of securitizations, Huysmans and the Paris School conceptualize securitization as a technocratic practice that is effective on a transnational level to an increasing degree. These studies illustrate two points. First, bureaucratic actors are privileged in the securitization of issues since their routinized practices construct and communicate meanings of security. Second, these securitizations form a permanent technique of governance that promotes a political environment in which the ‘state of exception’ is less dramatic than in the Copenhagen School’s concept, but becomes the norm (Fierke 2007b: 112; cf. Huysmans 2002: 57; Wæver 2003: 30; McDonald 2008a: 570–576). In this realm, Balzacq (2008: 76) points out, securitizations can occur and produce “social and political consequences without the explicit assent of an audience.” Moreover, this form of the institutionalization of security poses a problem for the Copenhagen School’s concept of desecuritization. As Donnelly (2013: 49) correctly remarks, if security is internalized by people to such a degree that it no longer has to be stated, then ‘rewording’ the matter becomes problematic. In this case, desecuritization would have to start by raising critical awareness of the institutionalization itself, and, respectively, of that which is treated in security terms.

Finally, a large debate also developed around the role of the audience in securitization. The concept of the audience has been criticized for its vagueness and lack of definition (Balzacq 2005, 2011b; Stritzel 2007; McDonald 2008a; Roe 2008; Salter 2008; Vuori 2008; Léonard/Kaunert 2011). Much of the contestation has centered on the question in how far securitization is truly conceptualized as an intersubjective process. Initially, as McDonald (2008a: 566) points out, the performativity of speech acts was emphasized more strongly in the concept of securitization. Although the decisive role of the audience was later incorporated into the concept, as laid out in Security: A New Framework for Analysis in 1998, it stayed “so undertheorized as to ultimately remain outside the framework itself” (McDonald 2008a: 564). The discursive exchange and negotiation between the speaker and the audience remains neglected and the emphasis remains instead on the speech act of the securitizing actor. Balzacq (2005, 2011b) sees the reason for this problem in the Copenhagen School’s dependence on Austin’s speech act theory with its focus on illocution. As he notes, the Copenhagen School conceptualizes securitization as illocution while neglecting the perlocutionary aspect and thus the effect on the audience.66 For Balzacq, perlocution is not an inherent quality of speech acts. In his view, it is only in relation to external reality, including “brute threats” and “external objective developments,” that the securitizer manages to convince an audience (Balzacq 2005: 181f). As mentioned above, in order to include perlocution, he therefore suggests reconceptualizing securitization and viewing it as a ‘strategic practice’ rather than a ‘speech act’. He criticizes the Copenhagen School for emphasizing the performativity of speech by which security becomes a self-referential practice (Balzacq 2005: 177, cf. 2011b: 19f). This in fact is stated precisely by Wæver (2003: 11), and it once again demonstrates that the Copenhagen School’s model leads to direct contradictions. On the one hand, the Copenhagen School points out that the process of securitization is intersubjective (Buzan et al. 1998: 30; Wæver 2000: 252, 2003: 12) and that it is only through audience acceptance that a securitizing move becomes a successful securitization. On the other hand, it also claims that “it is the actor […] who decides whether something is to be handled as an existential threat” (Buzan et al. 1998: 34). As mentioned earlier, Wæver was very clear on this point when stating that “a problem would become a security issue whenever so defined by the power holders” (Wæver 1995: 56, emphasis in original). If the latter is true, the characterization of securitization as an intersubjective process is blurred. The Copenhagen School is clearly aware of this problem as Wæver and Buzan have pointed to the danger “that too much focus can be placed on the acting side, thus privileging the powerful while marginalizing those who are the audience and judge of the act” (Buzan et al. 1998: 41). Again, the problem with dissolving this tension might be rooted in contending theoretical understandings.

The Copenhagen School also acknowledges that the ‘audience’ needs “better definition and probably differentiation” (Wæver 2003: 26). However, despite its vital role, it remains unclear who “the relevant audience” (Wæver 2000: 251) is. Who has to be convinced? Whose acceptance of proposed actions is required? The main audience could be the parliament, the public, the media, or special societal or interest groups; however, it could also be an intersection of these groupings, and, in the case of macrosecuritization, it could even include several audiences on various levels across countries. Thus, in order to define the success of securitizations, it might be necessary to consider multiple audiences (Balzacq 2005; Stritzel 2007; Roe 2008; Salter 2008; Vuori 2008). Salter (2008: 322), for example, distinguishes between four types of audiences: popular, elite, technocratic, and scientific. These audiences alter the particular setting and influence the form and content of securitizing speech acts and their prospects of success; as mentioned earlier, threat representations resonate in different ways with different audiences (Williams 2011b: 213). Although Wæver (2003: 12) concedes that the audience will vary “according to the political system and the nature of the issue,” the Copenhagen School provides no clear criteria with which to identify target groups; nor does it set out how to evaluate the respective relevance of particular audiences whose actual capacity to influence policy-making may differ. As Balzacq (2005) and Roe (2008) indicate, in securitization processes the ‘formal support’ of policy makers (such as cabinets or parliaments) might be more important for the application of emergency measures than the ‘moral support’ of the public. The likeliness of a successful securitization rises, however, when both audiences grant their support (Balzacq 2005: 184). The analysis of ‘formal support’ touches on another difficulty in determining the relevant audience – an aspect that is hardly discussed in the literature: if securitization is understood as an intersubjective – in a constructivist sense, mutually constitutive – process, the repercussions of security discourse on the securitizing actor also needs consideration. If securitization is a discursive process and if there is debate – as one always needs “to argue one’s case” (Buzan et al. 1998: 25) – the securitizing actor and its apparatus itself become part of the audience. In the end, a clear division between securitizing actor and audience is impossible as both sides speak and listen to one another and influence each other.

In the face of the topic’s complexity, one cannot, as Stritzel (2007: 363) remarks, “always figure out clearly which audience is when and why most relevant, what implications it has if there are several audiences and when exactly an audience is ‘persuaded.’” The latter points to another lack of definition, that is, even if ‘the relevant’ audience could be determined, how could the acceptance of a securitizing move by the audience be assessed? How is the success of securitizing moves exactly defined? The Copenhagen School does not say much about this besides that there would have to be “enough resonance for a platform to be made from which it is possible to legitimize emergency measures or other steps that would not have been possible [otherwise]” (Buzan et al. 1998: 25). It is, however, exactly this resonance, “this politics of consent that must be unpacked further” (Salter 2008: 324). What is clear, though, is that audiences also bear responsibility for their contributions to discourses. They might face challenges in evaluating the framings the securitizer puts forth, yet even under these circumstances, the acceptance of securitizations is a political choice (Buzan et al. 1998: 29).

In the realm of institutionalized security in particular, where “constant drama” is not required (Buzan et al. 1998: 27), the role of the audience becomes even more problematic. Since in these cases the government has a particular degree of legitimacy “to act without the consent of the general public and/or other securitizing audiences,” the role of the latter is “at best marginalized or at worst excluded” (Roe 2008: 618). In a similar way, intersubjectivity is hampered when the security topic is deliberately insulated from public debate and dealt with in secret. If operations are kept secret to prevent information leaking to an enemy or for other reasons, the acceptance of the audience is at least postponed. This is not to say that the audience does not play a role as politicians in democratic systems will probably have to justify their positions to the public at some point. However, and despite what the Copenhagen School suggests, there might not always be a straight sequence leading from securitization to acceptance by the audience and the implementation of emergency action. These aspects remain underdetermined in the Copenhagen School’s concept (Williams 2003; McDonald 2008b; Roe 2008).

4.4 Constructing Self and Other in Security Discourse

Securitization is largely a discourse of identity.67 Although also circumstances like climate change or economic crashes can be securitized, in the majority of cases the alleged threat in securitizations is posed by humans in terms of groups, peoples or (state) leaders/representatives. Therefore, in principle, the threat scenario is a constellation of ‘self’ and ‘other’ where the latter is an (external) enemy that endangers the state sovereignty and/or (national) identity of the ‘self’. However, in security discourses of this kind, the framing of the ‘other’ is not only decisive in defining which policies are an appropriate means of confronting the threat the ‘other’ allegedly poses, it is also constitutive of the ‘self’. By framing the ‘other’ in a certain way, the securitizing actor defines the ‘self’ in relation to it. Identity formation is thus a kernel of security discourse and foreign policy. The following section outlines this relation and demonstrates the extent to which certain characteristics of language use contribute to contrasting ‘self’ and ‘other’.

As outlined above, identity formation is considered largely a matter of language; how then is identity formed? In approaching this question, a certain feature of language gains importance, namely, its structure of binary opposition. Across disciplines, scholars recognize that language is based on difference and that, therefore, we think, speak and structure our world in dichotomies. This finding was prominently put forward by the structural linguist Ferdinand de Saussure (1974), as well as the communication theorist Kenneth Burke (1969 [1945]) and the philosopher Jacques Derrida (1981 [1972]). According to their insights, meanings depend on opposites: if one thing is good, the other must be bad; if one stance is right, the other must be wrong (Fourie 2007: 249; cf. Llorente 2002: 39). Binary thinking is the “placement of one thought or thing in terms of its opposite” (Burke 1969 [1945]: 403). This finding is also viewed as applying to subjects and thus identities. ‘Self’ and ‘other’ are mutually dependent since the ‘self’ is that which the ‘other’ is not. Consequently, the constitution of identity is a relational process in which the collectively shared image of the ‘self’ is constituted against the image of the ‘other’, that is, in difference to the ‘other’, and vice versa (Derrida 1978; Connolly 1991; Harle 2000: ch. 1).68

With regard to international politics, a discursive differentiation between ‘us’ and ‘them’ is a normal procedure as the relation to and demarcation from other nations and cultures is constitutive of national identity (Hall 1991a: 20–22, 1991b; Torfing 2005: 14; de Fina 2011: 271). As Geis (2006: 154) points out, “it is inherent in modern statehood and sovereignty that ‘others’ are permanently and inevitably constituted.” The question, however, is how this is done as “the naming and framing of one’s counterpart very much shape the potential options in dealing with him” (Geis 2006: 154; cf. Doty 1996; Howarth/Stavrakakis 2000: 3f). In fact, thinking and speaking in binary oppositions easily becomes precarious as it is prone to political instrumentalization, especially in security discourse. Often, ‘othering’ implies a hierarchy, since the ‘other’ is not simply viewed as different but inferior (Connolly 1991: 40, 45–48, 64; Todorov 1992: 42, 146–167; Riggins 1997: 9f; Campbell 1998: 73f; van Dijk 2001: 362). As Derrida (1981 [1972]: 41) points out, there is no “peaceful coexistence” between these binary oppositions but a “violent hierarchy” in which “one of the two terms governs the other […] or has the upper hand.” Whereas the positive attributes remain with the ‘self’, the negative ones are ascribed to the ‘other’ who is devalued.69 Consequently, the ‘self’ claims predominance. There can be various degrees of ‘othering’ up to detrimental opposition where the ‘other’ is constituted as “absolutely-not-self” (Abdel-Nour 2004: 430–433) and less than human.70 This constellation easily permits double standards and suggests the legitimacy of humiliation, exploitation, and violence. In general, it paves the way for doing unto ‘them’ what should not be done unto ‘us’.

The alleged legitimacy of the use of force rises further when otherness is constructed as threatening; this marks the pivot of securitization processes. In these cases, securitizations can intensify these dichotomies towards antagonisms. In its most severe form, the ‘other’ is constructed in such a demoniac manner that – according to the purported security logic – it must be eradicated. Commonly, this constitutes a strong politics of fear as the ‘self’ is constructed as civilized and good, whereas the ‘other’ is represented as pure evil and ready to strike. The reasoning put forth in such cases is simple: because the ‘other’ is evil, it commits evil.71 Because of that, the implementation of extreme measures is appropriate in order to satisfy one’s own security needs. Above that, force against the ‘other’ is not only justified but a noble if not holy act as it means fighting for God and the good (Sluka 2009: 145; cf. Harle 2000: 12).

The dichotomy of ‘self’ and ‘other’ comes along with a homogenization of the respective identity (Hall 1994: 142). The confrontation of a good ‘self’ and an evil ‘other’ leads to a generalized classification of in-group and out-group and impedes the consideration of internal difference and individual characteristics; consequently, the regard for individual bearing, guilt or innocence, is lost. In line with this politics of inclusion and exclusion, all who belong to the ‘self’ are part of a “legitimate community” whereas all who belong to the ‘other’ are identified as “enemies of the state” (Hajjar 2002: 120). Part of this homogenization is also the tendency to imaginatively invigorate the ‘other’ by constructing it as monolithic, that is, as an internally united enemy whose acts are well coordinated – a circumstance that further intensifies the alleged dangerousness of the ‘other’ and the perception of threat (Jervis 1983: 173f). It is therefore apparent that “the issue of identity – […] what makes us believe we are the same and them different – is inseparable from security” (Booth 1997: 88, emphasis in original). Identity and foreign/security policy are mutually constitutive. Identity “shapes how ‘security’ is understood” and security practices shape the understanding of identity (Lock 2008: 68). Faced with an evil ‘other’, one’s own interests can be clearly defined and particular policies can be legitimized as necessary and in line with national interests (Finlay et al. 1967; Jervis 1976; Edelman 1988: 76; Kristol 1990: 16; Hansen 2006: xvi).72 Understood in this manner, foreign policy is a discourse of danger and fear that constitutes the ‘self’. David Campbell (1996, 1998) convincingly demonstrated this with regard to the United States where the representation of external evil ‘others’ is shown as constitutive of America’s identity and interests. In fact, even in liberal democracies ‘othering’ is a well-known element of securitization. As democratic peace theory has shown, democratic states are not inherently peaceful. These countries also consider force and war as options, in particular towards their non-democratic counterparts (Daase 2006). Although the Copenhagen School has been criticized for having tailored its concept too narrowly to Western democracies, scholarly work confirms that this is still a vivid field of study. There are manifold reasons why ‘othering’ is ubiquitous. As mentioned above, it helps to shape and unite the ‘self’, and it increases patriotism and strengthens the social acceptance of one’s own values and belief systems; it thus enhances the chances for corporate and effective action against opponents. Positively speaking, it increases the population’s loyalty towards its leadership, but there is also a great potential for repression as people can be forced to consent or else face stigmatization “as saboteurs and traitors of the fatherland” (Fiebig-von Hase 1997: 32). Enemy imagery can also be used as diversionary maneuvers to shift the public focus from internal problems to an outward enemy. Moreover, the antagonism of ‘us’ versus ‘them’ simplifies complex situations and helps to legitimize proposed actions that are framed in terms of self-defense (Finlay et al. 1967: ch. 1; Jervis 1976, 1983; Ostermann/Nicklas 1982; Lippert/Wachtler 1988; Flohr 1991: ch. 4; Fiebig-von Hase 1997; Weller 2000: 85–93; Geis 2006). This includes, as the Copenhagen School describes, actions that would otherwise be impossible. Hence, one of the central characteristics of enemy constructions and the politics of fear is that it empowers the political elite.73

‘Othering’ often builds on established concepts of the ‘other’ that have been reinforced over time. Reviving and adapting enemy images and other stereotypes is thus a matter of historicity and intertextuality. One of the most consolidated binaries noticeable in Western discourse with regard to international security politics is that of the ‘West and the Rest’ (Hall 1994; cf. Karim 1997). It is in comparison and demarcation to ‘the rest’ (the Muslim world), that the West developed its awareness and image of superiority. The alleged inferiority of Arabs, who were seen as primitive and irrational also led to a deep mistrust towards them (Said 1978; Abu-Lughod 2002; Parry 2013: 9). In general, ‘othering’ takes place in discourses of exclusion, and Orientalism certainly can be seen, as outlined above, as exemplary of a discourse where power relations have produced a particular kind of knowledge and truth. This is also a case in point for the perpetuation of enemy images. Due to its historical grounding, Orientalist discourse with its particular representations of, and feelings about, the ‘other’, can easily be invoked by Western securitizers. New information can be filtered and organized in a way that fits into and confirms the pattern (Fiebig-von Hase 1997: 7f). In other words, through recontextualization the enemy’s image can be adapted to an ongoing situation, and can thereby be instrumentalized in political processes. As mentioned above, this happens by homogenizing both the in-group and the out-group. The Western ‘self’ is thereby consistently constituted as good, modern, democratic, and peace loving, whereas the Arabic ‘other’ is defined as barbaric, primitive, despotic, and violent (Volpp 2002: 572). This antipodal understanding was certainly reinforced by the security discourse that developed in the aftermath of the attacks of September 11, 2001; this is analyzed in the book’s empirical part below.

4.5 Summary

This chapter presented the Copenhagen School’s concept of securitization whereby security is defined as a speech act. According to this model, issues and people are turned into security problems when framed as such and when this framing is accepted by the audience(s). When this occurs, the political leadership can break free of the rules and apply extraordinary measures that would otherwise be illegal under national and international laws or treaties. The School’s assumption is not that threat constructions and enemy images never correspond with what is ‘out there’. On the contrary, if something threatening does exist, then there are greater chances that a securitizing move will be successful. However, the School points out that this correspondence is not an indispensable prerequisite for securitizations. Rather, threats can be constituted or intensified through the way things are presented in discourse. Securitizations can thus carry a matter beyond the status quo. In line with this, securitizations constrain the range of possible action, since ‘normal politics’ no longer appears appropriate and instead, emergency action seems to be imperative.

As discussed above, there are theoretical shortcomings to this approach. These include inconsistencies in its ontological assumptions, the narrowness of the model with its wide disregard for contextual features and for the processual character of securitizations, the institutionalization of securitizations, which in fact alters their logic, and the lack of definition and theoretical consideration of what constitutes a ‘relevant’ audience. It has been pointed out that the audience – despite its alleged centrality to the model – slips somewhat out of sight and that this questions the intersubjective quality of securitizations. While this book takes the criticism leveled against the Copenhagen School into account, the School’s model of securitization still proves a valuable approach to understanding how political developments and deviations from ‘normal politics’ can come about. It thus provides a theoretical orientation for a broad variety of security studies as it promises insights into the designation of threats and the shaping of foreign policies. As mentioned above, a central part of securitizations is ‘othering’ – the construction of an evil ‘other’ who threatens the ‘self’ in terms of sovereignty or identity. These enemy images are a powerful tool with which to legitimize emergency measures. Analyzing these identity and threat constructions thus makes it possible to retrace how force comes to appear necessary and justified. Therefore, the relevance of the Copenhagen School’s concept also rests with its ability to heighten people’s awareness of the power of security language, which in turn imposes responsibility on securitizing actors and their audiences. The CS underscores the point that talk is not innocent because it constructs circumstances that might lead to violence or even the declaration of war. Although the securitization concept does not portray the whole picture and security is more complex than the model represents, it still captures a vital aspect of the issue at hand. After all, discourse is the most common mode of political practice. Therefore, securitization is a useful framework for the discourse analysis conducted in this book.