European Journal of International Relations Copyright © 2007
SAGE Publications and ECPR-European Consortium for Political Research, Vol. 13(3): 357–383
[DOI: 10.1177/1354066107080128]
Towards a Theory of Securitization:
Copenhagen and Beyond
HOLGER STRITZEL
London School of Economics and Political Science, UK
One of the most important and controversial contributions to a vibrant
body of new security theories since the 1990s has been the idea of securitization.
However, rather than providing a consolidated position the
discourse on securitization has only just begun to transform the new
idea into a more comprehensive security theory. This article argues that
such a theory needs to go beyond the current reflections on securitization
by the Copenhagen School. Through internal critique and conceptual
reconstruction the article generates an alternative framework for
future empirical research and identifies two centres of gravity as a first
step towards a more consistent understanding of securitization as a
comprehensive theory of security.
KEY WORDS ♦ constructivism ♦ discourse ♦ organized crime
♦ securitization ♦ threat image
Introduction
For many years now, insights of social theory and linguistics have inspired
various streams of European approaches to International Relations and have
led to a broad range of alternative perspectives on international politics.
While being introduced as early as the 1980s, many thoughts became popular
to a wider audience only in the course of the 1990s. In security studies
traditional military strategists and European or American realists have been
supplemented and severely challenged by a vibrant body of theoretical innovations.
New security theories often share a more dynamic and, in general,
ontologically more reflective perspective on security. Moreover, they often
apply discursive methodologies and take a critical stance towards the taken
for granted ‘realities’ of security in the world.
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 357
One of the most important and controversial contributions to this theoretical
discourse has been the idea of securitization as it is articulated in the
works of the Copenhagen School. Ole Wæver and Barry Buzan, the core of
the Copenhagen School, define securitization as a successful speech act
‘through which an intersubjective understanding is constructed within a political
community to treat something as an existential threat to a valued referent
object, and to enable a call for urgent and exceptional measures to deal
with the threat’ (Buzan and Wæver, 2003: 491). Yet, rather than providing
a consolidated position, the discourse on securitization has only just begun
to transform the new idea into a more comprehensive security theory.
The attempts of the Copenhagen School to construct a more comprehensive
theory rest on two central concepts. First, on a triology of the speech act,
the securitizing actor and the audience and, second, on three ‘facilitating conditions’
that influence the success of a securitizing move. However, as this
article will show, these attempts suffer from several internal tensions in the
argument, an often too vague and undertheorized terminology and, in general,
the fact that too much weight is put on the semantic side of the speech
act articulation at the expense of its social and linguistic relatedness and sequentiality.1
In order to turn the basic idea of securitization into a framework
to study securitizations and securitizing moves systematically, an alternative
conceptualization is needed.
In principle, one can think of such an alternative as a consistent framework
or a comprehensive theory.2 An alternative framework would mainly require
to be more systematic and clearer than the Copenhagen School with respect
to the main dimensions and assumed dynamics of securitization. Apart from
a more concise conceptual statement, the main aim would be to provide better
guidance for systematic and comparative empirical analysis, yet leaving it
to the empirical studies themselves to work out in detail which element of
the framework is, when and why, most important. In contrast, comprehensive
theories are usually more specific about how their elements relate and
more demanding with respect to their metatheoretical (ontological and epistemological)
consistency. Consistent comprehensive theories are also explicit
with respect to their main ‘centre of gravity’, i.e. that part of the theory from
which the other parts gain much of their own power.
The first contribution of this article to the debate towards a theory of
securitization is pretheoretical in the sense that its task is an exercise in internal
critique and conceptual restructuring. The basic idea is similar to what
Hempel has called ‘conceptual explication’:3
Conceptual explication attempts to specify the logical structure of given expressions:
… explication aims at reducing the limitations, ambiguities, and
inconsistencies … by propondering a reinterpretation intended to enhance …
clarity and precision. (Hempel, 1952: 12)
European Journal of International Relations 13(3)
358
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 358
Thus, at this level the aim of this article is to make transparent the main tensions
and boundaries of securitization theory and to suggest an alternative
that is more systematic and clearer. With a clearer framework, communication
between scholars would be improved and insights from ‘real-world’
securitizations could be gathered and compared. This, in turn, could help
improve the theoretical reflection and conceptual restructuring towards a
theory of securitization. In this sense, even a mere conceptual critique and
an alternative framework would be of help, given the problematic range of
contradictory empirical applications of securitization theory we have experienced
so far. When reading ‘between the lines’, Wæver (2003) admits these
problems by writing:
There is by now a surprising amount of empirical studies done with full or
partial use of the securitization theory. These do not follow a standardised
format … The theory does not point to one particular type of study as the
right one … Optimistically, the diversity is a sign that the theory … can generate/structure
different kinds of usage and even produce anomalies for itself
in interesting ways. (Wæver, 2003: 16–17)
Less optimistically, however, theoretical contradictions, anomalies and inconsistent
empirical applications of securitization cannot only be celebrated as
‘diversity’, but they also have clear disadvantages. Most importantly, they prohibit
the improvement of existing concepts in the light of (comparative)
empirical findings.
Beyond these pretheoretical ambitions, the article also suggests first steps
towards a more comprehensive theory of securitization. At this level, the article
argues that there are two centres of gravity of securitization theory that
are currently both theoretically underdeveloped. While these centres could
be reconciled to some extent, ultimately they reflect two rather autonomous
readings of securitization and are based on two separate metatheoretical convictions.
To some extent they perhaps even reflect two opposing views of
what a theory (of securitization) can and should do, whether it should be
kept diverse and contradictory or whether a more coherent position is good
for the future of securitization theory and securitization studies. The first
understanding concentrates on the speech act event and is grounded in the
concept of performativity (or ‘textuality’). This understanding would correspond
with an ‘internalist’, more poststructural/postmodern (Derridarian
and/or Butler) reading of securitization and is by now only articulated in a
rudimentary form in the concept of ‘illocution’. An internalist reading of
securitization is stronger in Wæver’s early works but has become much weaker
in the joint works with Buzan. The second understanding theorizes the
process of securitization, based on, I would suggest, the central idea of
embeddedness. This understanding would correspond with an ‘externalist’,
more constructivist4 reading of securitization. Although there are some
Stritzel: Towards a Theory of Securitization
359
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 359
externalist elements in the joint works of Wæver and Buzan (e.g. ‘facilitating
conditions’, ‘authority of the speaker’ or ‘structured field’), in sum, such a
reading is strongly underdeveloped and has even been explicitly rejected by
Wæver (see e.g. Wæver, 2000, 2004). Towards a theory of securitization,
this article takes side with and elaborates theoretically an externalist position
by claiming that security articulations need to be related to their broader discursive
contexts from which both the securitizing actor and the performative
force of the articulated speech act/text gain their power.
The argument of this article will be developed in four steps. It starts with
the basic idea of security as a speech act. This part examines how the idea is
articulated by the Copenhagen School and how this articulation relates to
speech act philosophy. The second part analyses and criticizes the attempts
of the Copenhagen School to transform this basic idea into a more comprehensive
theoretical position. Based on this critique, the third part suggests
an alternative framework to study securitizations systematically. Fourth, and
finally, similarities with the Copenhagen School, boundaries of the suggested
alternative and implications for future research will be discussed.
Securitization: Basic Idea and Internalist Centre of Gravity
The defining feature of the Copenhagen School approach to security is the
fact that it proposes to study security practices by drawing on speech act philosophy,
assuming that the articulation of security is a crucial form of security
action. It is this articulation that has the potential to structure the social practices
that follow. The articulation of ‘security’ entails the claim that something
is held to pose a threat to a valued referent object that is so existential
that it is legitimate to move the issue beyond the established games of ‘normal’
politics to deal with it by exceptional, i.e. security, methods. This puts
an actor in a very strong position to deal with an issue as he/she thinks is
appropriate. As Wæver put it, ‘by uttering “security” a state-representative
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 defining textual feature of securitization is therefore a specific security
rhetoric which is marked by survival, priority of action and urgency
(Wæver, 2003: 10), a claim to a modus of exceptionality that is contained in
the meaning of (national) security.
The conceptualization of this distinct modus emerged out of Wæver’s
analysis of traditional (i.e. realist) concepts of security which centre around
the idea of national survival (see Wæver, 1997). In Wæver’s concept, the
realist meaning of security thus continues to exist as a rhetorical claim
marked by the grammar of security as a distinct field of practice that distinguishes
— according to realism — ‘security’ from other fields. Speaking
European Journal of International Relations 13(3)
360
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 360
Stritzel: Towards a Theory of Securitization
361
‘security’ successfully therefore draws an issue into this particular (realist definition
of a) field of exceptionality with all the consequences this may have.
The basic mechanism of this fundamental transformation draws on John L.
Austin’s concept of ‘performative utterances’ (see Austin, 1962). Applying
Austin, Wæver argues that the very utterance of ‘security’ is more than just
saying or describing something but the performing of an action. According
to Austin, performative utterances do not just describe but have the potential
to create (new) reality, for the Copenhagen School the modus of exceptionality,
of dealing with an issue in a new way. Moreover, performative
utterances fall outside the conventional true/false dichotomy. In Austin’s
words, they do not have ‘truth conditions’ but ‘felicity conditions’. If the
felicity conditions are met, the speech act — although perhaps not ‘true’ —
may still happen felicitously. By applying Austin to security studies the conceptual
focus thus shifts from the traditional threat–reality nexus (i.e.
whether a claim that there is a threat is actually true) to what a speech act
does (i.e. what happens because of its utterance). For the Copenhagen
School, the main effect of uttering security is its potential to let an audience
tolerate violations of rules that would otherwise have been obeyed (Wæver,
2003: 11).
Especially in his single-authored publications (see in particular 1989,
1995, 2000, 2003, 2004; but see also Buzan et al., 1998: 46–7) Wæver also
draws on insights from Derrida and Butler5 to stress the always political and
indeterminate nature of the speech act event6 whose meaning and performative
force is not related to its context:
In my case, some of the most important areas where this ‘general philosophy’
was worked out by Derrida was in relation to speech act theory … It points to
the centrality of studying in a text, how it produces its own meaning, rather
than relating it to a ‘context’ which is a doubtful concept because it tends to
imply the traditional sender–receiver view of communication where an original
meaning can be retrieved if only put in the proper context. (Wæver, 2004: 11)
This is an acknowledgement of Derrida’s famous claim that ‘there is nothing
outside the text’ and that meaning can never be fixed.7 Judith Butler has
extended such a poststructural/postmodern position — which Derrida has also
directed against Austin (see Derrida 1972, 1977) — to develop the idea of
speech acts having productive power (see Butler, 1996, 1997). According to her
concept of performativity, speech acts have the power to constitute new meaning
and create new patterns of significance in social relations. Since for Butler
actors and structures must be ‘performed’ to exist in a meaningful way, both
actors and structures are constituted ‘retroactively’ (see also Zizek, 1989:
100–2).8 Thus, according to Butler, it is the speech act itself which has the
power to create authority and bring about change rather than any pre-existing
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 361
context that would empower actors and/or speech acts in the first place.
Wæver acknowledges her work when he argues:
A speech act is interesting because it holds the insurrecting potential to break the
ordinary, to establish meaning that is not already in the context. It reworks or produces
a context by the performative success of the act. (Buzan et al., 1998: 46;
Wæver, 2000: 286)
In sum, Wæver agrees with Butler and Derrida — and in accordance with
Schmitt and Arendt — that a speech act has an indeterminate force of its own
that is not related to features of an existing context. Moreover, the authority
to speak is constituted by the performative power of the speech act itself which
constitutes actors and their social relations retroactively. As I will argue in more
detail later on, it is exactly this postmodern/poststructural root in the thinking
of the Copenhagen School — that also runs through their co-authored
texts — which hints at a possible (internalist) centre of gravity for a comprehensive
theory of securitization.
Towards a More Comprehensive Framework: Adding Externalism
The discussion so far mainly reflects the early conceptualizations of security
as a speech act that can be found in ‘Security, the Speech Act’, ‘Securitization
and Desecuritization’ and ‘Concepts of Security’ (see Wæver, 1989, 1995,
1997). Yet, the basic idea laid down in these publications itself is too limited
to also guide the study of ‘real-world’ securitizations. The Copenhagen
School has realized this problem. In their joint works on securitization they
have moved towards a more comprehensive understanding by outlining what
they call ‘facilitating conditions’ (Buzan et al., 1998: 25, 31–3; Buzan and
Wæver, 2003: 71–2, 74). In another recent paper on securitization, Wæver
is also more specific on the role of actors and audiences (Wæver, 2003). In
the analysis that follows I will therefore critically examine these attempts by
the Copenhagen School to turn their conceptual idea into a more comprehensive
theory. My analysis will start with their understanding of actors and
audiences.
Speech Acts, Actors and Audiences
The skeleton of a more comprehensive theory of security action by the
Copenhagen School itself is marked by three elements: (1) the speech act, (2)
the securitizing actor and (3) the audience. Unfortunately, they have not yet
conceptualized the exact relationship between the actor and the audience very
clearly. Generally speaking, it is clear that the Copenhagen School regards
securitization as an intersubjective act of a securitizing actor acting towards a
European Journal of International Relations 13(3)
362
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 362
Stritzel: Towards a Theory of Securitization
363
significant audience. Thus, rather than reducing threat assessments to a single
actor, most often ‘the state’, the Copenhagen School splits the actor into two
elements: the securitizing actor performing a securitizing move by uttering a
security speech act, and the relevant audience accepting or refusing this move.
With that a certain threat is no longer simply assessed but its interpretation
and representation is ‘negotiated’ between an actor and the relevant audience.
While the actor can only propose a certain recognition and representation, it
is the audience which decides whether this proposal is accepted as a common
narrative, i.e. whether the proposal will be intersubjectively held as real.
However, in empirical studies one cannot 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’.9 Similar problems
occur with the idea that a proposal is ‘intersubjectively held’, which implies
‘voluntarily held’, as opposed to aspects of coercion, repression and ‘silence’
(Hansen, 2000). What if a dictator is the securitizing actor? Is the speaker–
audience model the most appropriate model for non-democratic settings?
And how can this model be reconciled with the Derridarian and Schmittian
elements? As we have seen, Wæver explicitly refuses the ‘traditional sender–
receiver view of communication’ (Wæver, 2004: 11) — but at the same time he
believes in a speaker–audience model?
Although the concepts of speech act, actors and audience seem to be important
for the Copenhagen School, the approach is much too silent about the exact
mechanism of their triology and too unspecific about the substance of their
terms. In particular, the Copenhagen School should say more about the relative
status of the idea of a security utterance or speech act event as opposed to the
idea of an intersubjectivity of actor and audience or the process of securitization.
How does the power of the audience relate to the decisionism of the speech act
event? And how can the idea of the performative force of a security articulation
stressed in Security, The Speech Act be reconciled with the concept of a (social or
intersubjective) process of securitization developed in Security: A New
Framework of Analysis? There are numerous formulations in the works of the
Copenhagen School that reflect this tension. It would be most obvious if statements
in Security, The Speech Act and Securitization and Desecuritization were
compared to statements in Security: A New Framework of Analysis. However,
the tension can also be found in a single document such as Security: A New
Framework of Analysis and even on one page of this document:
This self-based violation of rules is the security act … For the analyst … it is to
understand the processes of constructing a shared understanding of what is to be
considered and collectively responded to as a threat. The process of securitization
is what in language theory is called a speech act … it is the utterance itself
that is the act. By saying the words, something is done. (Buzan et al., 1998: 26;
emphasis added)
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 363
As these lines make clear, Wæver and Buzan continuously fluctuate between
the terms process and speech act/utterance as if both were synonymous. The
act is at the same time defined as an intersubjective process of constructing
a threat and as just an ‘utterance itself’. Yet, I would argue that the (decisionist)
performativity of security utterances as opposed to the social process
of securitization, involving (pre-existing) actors, audience(s) (and contexts),
are so different that they form two rather autonomous centres of gravity.
The Role of Facilitating Conditions
In contrast to the reflections on the triology of speech act, actor and audience,
what the Copenhagen School calls ‘facilitating conditions’ offers a more specific
framework for analysing securitizations empirically. To some extent, they
can even be interpreted as a ‘nucleus’ provided by the Copenhagen School
itself for a more comprehensive externalist theory of securitization. From an
internalist position, however, the introduction of facilitating conditions is a
problematic move towards causality which undermines the Schmittian and
Derridarian legacy of securitization. Inspired by Austin’s concept of felicity
conditions10 (1962: 14–15) they refer to:
1. the demand internal to the speech act of following the grammar of security,
2. the social conditions regarding the position of authority for the securitizing
actor — that is, the relationship between speaker and audience and
thereby the likelihood of the audience accepting the claims made in a
securitizing attempt, and
3. features of the alleged threats that either facilitate or impede securitization.
(Buzan et al., 1998: 33)
In Security: A New Framework of Analysis, Buzan and Wæver introduce these
conditions with remarks that both power (Buzan et al., 1998: 31–2) and the
intersubjective establishment of a threat (Buzan et al., 1998: 25) are important
to understand securitizing speech acts. In this respect they even claim that ‘in
concrete analysis … it is important to be specific about who is more or less privileged
in articulating security. To study securitization is to study the power politics
of a concept’ (Buzan et al., 1998: 32; emphasis added). Moreover, they
claim to conceive of security as a ‘structured field’ because ‘some actors are
placed in positions of power by virtue of being generally accepted voices of
security, by having the power to define security’ (Buzan et al., 1998: 31).
With these lines Buzan and Wæver stress that they do not follow radical
poststructuralists who would only concentrate on the power structure of the
linguistic features of a speech act (Buzan et al., 1998: 46). Instead, they seem
to conceptualize the linguistic features of a speech act as only the internal
European Journal of International Relations 13(3)
364
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 364
Stritzel: Towards a Theory of Securitization
365
‘facilitating condition’ and add, contra radical poststructuralism and Wæver’s
own sympathy for the idea of performativity, an element of contextuality as the
external side of a successful speech act (Buzan et al., 1998: 32).
However, this would require a much more coherent grounding of securitization
theory in, for example, Bourdieu’s reflections on ‘authority’ and
‘field’. Indeed, Bourdieu could give securitization theory a contextual twist
with which one could also theorize what should be done ‘in concrete analysis’.
On the other hand, such a move has the potential to destroy the coherence
of their theory, the more Buzan and Wæver at the same time adhere to
the autonomous productive power of performative speech acts highlighted
by Derrida and Butler. And even now, their contextual ad hoc remarks create
a certain degree of tension with their ideas on actors, audiences and intersubjectivity:
at least, the stronger one emphasizes the idea that ‘some actors
are placed in positions of power’, the more problematic it becomes to uphold
a split between the securitizing actor and the audience where the audience is
placed in the position of ‘having the power to define security’.
On a more general level, one has to ask whether their import of the
Bourdieuian concept of authority still is not too narrow to understand the complex
and power-laden social dynamics of securitization. Substantively, they limit
the concept of power to the power to persuade and the idea of contextuality or
embeddedness to a context defining a sort of reputation (‘the authority of the
securitizing actor’). Temporally, they reduce the speaker–audience relationship
to the securitizing attempt: ‘the likelihood of the audience accepting the claims
made in a securitizing attempt’. As a result, more complex power relations as well
as more sedimented and sequential forms of the relationship between the speaker
and the audience get lost. Yet, because of a sedimented relationship between a
speaker and the audience, certain securitizing moves may be prohibited in the
first place or not taken at all; potential audiences, in other words, may be — to
paraphrase a term by Michael Mann (1985) — ‘discursively outflanked’. In the
concept of the Copenhagen School, any power of ‘non-decision’ (Bachrach and
Baratz, 1963) as well as more structural notions of power are not captured.
Yet, I would argue that a theory of securitization should go beyond an intrinsic
understanding of ‘social magic’ and take into account and conceptualize
the deep embeddedness of security articulations in social relations of power without
which its dynamics and non-dynamics cannot be understood. In other
words, what is missing in the current articulations of the Copenhagen School is
a better and more comprehensive awareness of the existence of a social sphere.
General Problems
The critical discussion of attempts by the Copenhagen School to move from
the conceptual idea to a theory of securitization has shown that their own
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 365
current proposals are confronted with several single contradictions. In the
following I will abstract from these specific aspects and turn to more general
problems and tensions of the Copenhagen School approach to securitization.
I will argue that the main problem of the current articulation of securitization
theory is their insufficient consideration of the situatedness of speech
acts and a too-static conceptualization of the speech act event in general.
The first tension concerns a certain degree of implicit objectivism as opposed
to a more reflectivist and relativist overall approach. This is particularly evident
in their concept of facilitating conditions. Wæver is aware of this when he
writes: ‘This part of the theory dealing with conditions is highly sensitive
(because it can lead to a re-introduction of objectivism into the theory) and it
is necessary to be very precise about the exact status of the different elements
of the theory’ (Wæver, 2000: 252; 2003: 14). In this context it must be particularly
perplexing for any more strongly reflectivist scholar to read the third
condition: ‘features of the alleged threats that either facilitate or impede securitization’
(Buzan et al., 1998: 33). Without any further explanations this reads
as if the Copenhagen School would attribute a causal role for a given external
materiality outside its being mediated through language. In the meantime,
Wæver has weakened this impression by arguing that this condition refers to
aspects ‘historically associated with a threat [and] generally held to be threatening’
(Wæver, 2003: 15). Nevertheless, a certain tension in the argument of
the Copenhagen School remains: the more poststructural one reads the
Copenhagen School, the more problematic is their concept of facilitating
conditions, and, conversely, the more emphasis is put on the facilitating conditions,
the more difficult it is to read the Copenhagen School as a poststructural
position.
The second tension concerns static elements of the theory as opposed to
a more dynamic and contingent overall approach. One example of a static
element concerns the separation of referent object and actor and the introduction
of neatly separated ‘sectors’ with many more reflectivist scholars
arguing that securitizations usually escape a static ‘actor–referent objectsector’
heuristic (see e.g. McSweeney, 1996, 1998, 1999; Buzan and Wæver,
1997). Another example is the very conceptualization of the speech act event.
As we have seen, the Copenhagen School claims that by speaking security
successfully, an issue is drawn into a particular (realist definition of a) mode
of dealing with it which is marked by exceptionality. This reduces, contra
Derrida,11 a securitization to a static event of applying a (fixed) meaning (of
security as exceptionality) to an issue rather than seeing it as an always (situated
and iterative) process of generating meaning, i.e. as a dynamic (social and
political) sequence of creating a threat text. Furthermore, many scholars
would criticize the strong (realist) emphasis on exceptionality in this conceptualization.
Taking the realist understanding of security as an intellectual
European Journal of International Relations 13(3)
366
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 366
Stritzel: Towards a Theory of Securitization
367
starting point (Wæver, 1989: 36–41), Wæver perpetuates the realist vocabulary
of exceptionality which many scholars would consider to be empirically
inadequate (and perhaps even ethically problematic). As risk scholars argue,
many current security practices deal with ‘threats’ below the level of exceptionality
(see e.g. Coker, 2002; Rasmussen, 2001; see also Abrahamsen, 2005).
The realist vocabulary tends to create a rather strict politization–securitization
dichotomy which prohibits making securitization theory more widely applicable
to ‘real-world’ securitizations.
My former point relates me back to the initial critique that the Copenhagen
School refuses to conceptualize securitizing speech acts and securitizing
actors as embedded in broader social and linguistic structures. By taking side
with an externalist reading of securitization, I would argue that an actor cannot
be significant as a social actor and a speech act cannot have an impact on
social relations without a situation that constitutes them as significant. It is
their embeddedness in social relations of meaning and power that constitutes
both actors and speech acts. This neglect of ‘externalism’ in the concept of
securitization is even more puzzling since Wæver himself has provided first
hints for a more embedded approach. In an article on foreign policy analysis
he states:
Discourses organise knowledge systematically, and thus delimit what can be
said and what not. The rules determining what makes sense go beyond the
purely grammatical into the pragmatic and discursive … Subjects, objects and
concepts cannot be seen as existing independent of discourse. Certain categories
and arguments that are powerful in one period or at one place can sound nonsensible
or absurd at others. (Wæver, 2001: 29; emphasis added)
Similarly, in ‘Defence of Religion’ — which is co-authored by Carsten Bagge
Laustsen — Wæver writes:
By exploring the structure of discourse constitutive of threats … we can show
what makes securitisation particularly attractive and under what conditions.
(Laustsen and Wæver, 2000: 706; emphasis added)
Yet, he then reduces discursive embeddedness to a proper definition of — in
this case — religion as a referent object. Again, this is a move towards a static
concept that does not capture the relational dynamics of the social and political
process of generating meaning.
In sum, it seems that Wæver is aware of the virtues of externalist (more structural
and discursive) readings in principle but has sided with an internalist
position when it comes to conceptualizing securitization. In his joint work
with Buzan he has then tried to move his initial idea of security as a speech
act some steps towards externalism — with the negative side-effect that severe
tensions have been created that make it more difficult to move coherently
towards a comprehensive theory of securitization. As a result, the current
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 367
state of securitization theory is theoretically vague and it does not provide
clear guidance for empirical studies. Therefore, systematic comparisons of
‘real-world’ securitizations which could help improve the theory cannot be
generated.
Embedding Securitization
Building on my critique of the Copenhagen School and Wæver’s own hint at
the crucial importance of discursive contexts, I will now outline central lines
of a more embedded understanding of securitization which I conceptualize
as a dynamic three-layered triangle of text, context and positional power. As
an exercise in conceptual reconstruction, the boundaries of my own concepts
are set by the basic idea of securitization, understood as the act versus process
of applying/generating meaning. In other words, the purpose of this section
is defined by the interest in a framework (and first steps towards a comprehensive
theory) of exactly this act/process and the relations of power involved
that can be identified through conceptual reconstruction. One could characterize
my reflections as a — broadly defined — constructivist move away from
Wæver which shares many insights with structuration theory and critical discourse
analysis (see e.g. Fairclough, 1992; Wodak and Meyer, 2001). To illustrate
the central concepts of the framework I refer to the securitization of
organized crime in the USA.12
Meaning, Power and the Relationality of Agents and Structures
The agent–structure problem is a controversy on the extent to which social
phenomena can be understood as the result of agents and/or structures,
with many social theorists favouring one element over the other and structurationist
scholars, most prominently perhaps Giddens (1984), Bhaskar
(1979) and Archer (1995), discussing synoptic concepts which integrate
both elements into a theoretically ‘elegant’ hybrid. In light of the intense
reflection on agents, structures and structuration over the past decades (see
e.g. Parker, 2000; Wight, 2006), it is nowadays almost trivial to state that
agential and structural elements of social context and power are related. In
other words, most scholars are nowadays aware and take into account that
actors always act within a structural context which constitutes them and provides
a frame of enabling and constraining conditions and that structures
need agents to translate their attributes into a dynamic of action and change.
Or, as Colin Hay has stated:
Agents are situated within a structured context which presents an uneven distribution
of opportunities and constraints to them. Actors influence the development
of that context over time through the consequences of their actions.
European Journal of International Relations 13(3)
368
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 368
Stritzel: Towards a Theory of Securitization
369
Yet, at any given time, the ability of actors to realise their intentions is set by
the context itself. (Hay, 2002: 116–17)
Among the many structurationist specifications of this basic insight/conviction
I personally find the strategic-relational approach (see e.g. Jessop, 2005) particularly
interesting. The strategic-relational approach, which extends Giddens’
reflections on structuration from a critical realist position and thereby integrates
the critique by Archer (1990), has recently been used to fuse mobilization
theories on (cultural) frames/framing, opportunity structures and leadership/
norm entrepreneurs (see Fumagalli, 2006). This is in many respects similar
to my own framework of a triangle of text, context and positional power.
Furthermore, by being able to integrate moderate discursive approaches, the
strategic-relational approach also provides an answer to the controversy on
discursive versus extra-discursive social practices related to the social construction
of threats (see e.g. Huysmans, 2006: 91–5). It is therefore a viable way to
socially embed linguistic discourse and thereby specify and improve existing
constructivist reflections on the construction of threats (see e.g. Weldes, 1996;
Fierke, 1998; Doty, 1993) which have so far mainly or only been focused on
linguistic dynamics (but see already Fairclough, 1992: 62–100). Nevertheless,
the conceptualization I propose is not explicitly strategic-relational but deliberately
abstract, favouring, on the basis of a broadly defined structurationist
understanding, a ‘grounded-theory’ approach towards a comprehensive theory
of securitization that would make use of insights gained from (comparative)
empirical studies to specify the theory consecutively.
A broadly defined structurationist understanding of social context would
stress the relationality of agents, structures and texts in the sense that their existence
is mutually constitutive and not reducible to the sum of structural, agential
or textual factors treated separately.13 Indeed, for a structurationist position
the three factors are so closely related that they can perhaps best be thought
of as different layers of agency (see e.g. Wight, 1999). Based on this core
assumption, I would propose to distinguish socio-linguistic and socio-political
dimensions of context. With the socio-linguistic dimension of context I refer to
the network of constitutive rules and narratives that surround a single linguistic
act. This dimension is essential to understand a speech act. Actors can
exploit linguistic contexts as a reservoir of analogies, similies and contrasts.
We can therefore often observe that securitizing actors speak to and from a
broader linguistic context by framing their arguments in terms of the distinct
linguistic reservoir that is available at a particular point in time. In contrast, the
socio-political context concerns the often more sedimented social and political
structures that put actors in positions of power to influence the process of
constructing meaning. Since a given social context is at best in some very rare
cases truly symmetric, actors are usually endowed with an unequal ability to
influence the evolution of an individually proposed meaning into a collectively
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 369
held representation. The concept of socio-linguistic contexts thus refers to
a fairly contingent and usually more fluid dimension of sociality which often
involves a high degree of human creativity and (unintended) dynamics. In
contrast, socio-political contexts represent a more sedimented dimension
which often hint more directly at the likelihood for a securitizing move to be
successful.
These dimensions correspond with a structurationist understanding of
power as the relatedness of (1) the existing discourse, constituting the performative
power and the meanings of security articulations, and (2) the positional
power of actors, influencing the process of defining meaning by
enacting particular threat texts and/or shaping the existing discursive context.
Conversely, the performative force of a threat text can help constitute
or change existing discourse coalitions and/or change an existing discourse,
thereby reconfiguring existing relations of power. Influencing the process of
defining meaning is always marked by acts of translation of a certain threat
text into an existing discourse. The better the compatibility of the articulated
text/textual structure and the existing discourse (i.e. its ‘resonance’) and the
better the positional power of securitizing actors, the easier it is for them to
establish their preferred individual text as a dominant narrative for a larger
collective.
This understanding of power is more limited than similar attempts in the
literature to use structuration to construct a comprehensive framework for
the analysis of power (see e.g. Barnett and Duvall, 2004, 2005; Guzzini,
1993). While the notion of power as such may indeed be overly complex and
‘polymorphous’ (Barnett and Duvall, 2004), i.e. impossible to capture by a
single formulation, an understanding of securitization as embedded in (historically
distinct) structures of meaning and power has a rather clear-cut centre
of gravity: existing discourse. It is, according to this reading, existing
discourse in its fluid socio-linguistic and more sedimented (socio-political
and/or socio-linguistic) form which privileges or disadvantages certain actors
(‘positional power’) and texts (‘performative force’) as opposed to others. In
this sense, existing discourse is always ‘strategically selective’ (Jessop).
As a framework of analysis these reflections create three central layers/
forces of securitization: (1) the performative force of articulated threat texts,
(2) their embeddedness in existing discourses and (3) the positional power of
actors who influence the process of defining meaning (see Figure 1).
The first element of the framework focuses on the structure of a security
articulation (‘text’) understood as a rather durable product of linguistic and/or
symbolic actions (Wodak, 2001b: 66), assuming that different semantic and/
or semiotic threat structures generate different performative forces. Understood
this way, the concept of text goes beyond mere speech to also include the
symbolic language of visuals/images and sound. Moreover, in contrast to the
European Journal of International Relations 13(3)
370
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 370
Stritzel: Towards a Theory of Securitization
371
Copenhagen School, the meaning of a threat text is not given (by the concept
of security as exceptionality) but generated — often as the result of a dynamic
social process. Its meaning and performative force is therefore never uniform
and perhaps impossible to figure out in abstract. Since textual structures,
according to an externalist understanding, are always temporally and spatially
constituted, one will hardly find texts that can be isolated from both the contingent
creativity of actors and their embeddedness in social and linguistic
contexts.14
The performative force of organized crime mainly stems from what has been
coined ‘the mafia mystique’ (Smith, 1975). It can be poetically and hermeneutically
unearthed by interpreting the textual structure and considering its
embeddedness in broader (cultural) discourses, in the case of organized crime
mainly in Italy and the USA. Organized crime then appears as a discursively
sedimented and condensed threat text that was being generated sequentially,
gaining influences from diverse cultural, cinematographic/fictional and political
sources. These influences have established a rich reservoir of associations
which continue to create a ‘sublime’ mixture of fear and fascination (Burke,
1958) whenever invoked. Politically, the rich cultural heritage of organized
crime was constantly invoked by securitizing actors to instil a process of intensification15
from the early 1950s to the late 1960s (see Moore, 1974; Smith,
1975; Lampe, 1999). Sequentially and increasingly, organized crime became
constructed as a macro-threat for the USA.
The second element concerns the power of the existing (linguistic) discourse,
i.e. a complex bundle of simultaneous and sequential interrelated acts
(Wodak, 2001b: 66). This dimension refers to the interactions between the
text and the larger discursive practices in which it is embedded. Texts often
Figure 1
Framework of Securitization
Layers Forces
Text Performative
Power
embedded
embedded
Discourse/ Power of
Context Discourse
changing
changing
Actor Positional
Power
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 371
evolve out of particular socio-linguistic contexts so that they often carry with
them the baggage of a historically evolved practice out of which and towards
which a particular text was articulated. Texts therefore sometimes even show
traces of differing discourses and ideologies struggling for dominance (Wodak,
2001a: 11).
The example of securitizing organized crime in the USA can illustrate this
dimension of the framework quite clearly. In particular, it also shows the
importance of considering the interrelatedness of ‘performative force’ and
‘embeddedness in (linguistic) discourse’. In Italy, organized crime was mainly
constructed as rites of masculinity, honour and violence which were in turn
embedded in broader mythical and anecdotal narratives about ‘Sicilism’ (see
Hess, 1973; Paoli, 2003). In the USA of the early 20th century this heritage
was taken up and supplemented by experiences of the Prohibition period as
well as several rumours and speculations about murders of policemen, politicians
and the criminal activities of ethnic (mainly Italian) groups (see e.g.
Smith, 1975; Kenney and Finckenauer, 1995). Importantly, also, Hollywood
and the US cultural industry discovered the topic and created powerful icons
of organized crime such as the Thompson machine gun, the ‘gangster suit’
and, perhaps most importantly for the political process of securitization, the
Italian accent as a ‘sound of racketeering’ (see e.g. Ruth, 1996). Furthermore,
the myth of the charismatic leader heading a hierarchically organized, businesslike
corporation was established for which ‘Al Capone’ became perhaps most
emblematic. These fictional representations, in turn, resonated strongly with a
distinct moralistic and religious discourse that represented New York and
Chicago as the sites of immorality, decadence and subversiveness, a form of discourse
on the ‘evilness’ of modern urbanity that has been identified as a typical
element of US exceptionalism (Lipset, 1996).
Finally, for the proposed third dimension of securitization I argue that more
sedimented forms of context can be amalgamated with an agential notion of
power into the concept of positional power.16 A certain power capacity can be
defined as positional to the extent that relevant actors are placed in different
positions within a given social environment to influence collective meaning
constructions. In extreme cases actors may have some sort of official, delegated
or enforced, ability to define meaning so that their power capacity may come
close to a monopoly. In other cases their ability to execute power may be more
indirect through their privileged position to influence — e.g. ‘behind the
scenes’ — a social and political structure or process that is particularly relevant
for the construction of meaning.17
In the case of organized crime in the USA, the importance of positional
power can also be clearly identified. For a very long time during the political
process of securitization, the main securitizing actor, the Federal Bureau of
Narcotics (FBN), operated in the background by influencing information
European Journal of International Relations 13(3)
372
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 372
flows over which they had superior power. In contrast, more respected and
publicly visible speakers such as the well-respected Federal Bureau of Investigation
(FBI) under Edward Hoover as well as several leading politicians
were very sceptical about whether a mafia-like organization really existed. Yet,
the FBN had the largest database on organized crime, and leading ministerial
agencies, such as the Organized Crime Section established under Robert
Kennedy, had to rely on their data (Schlesinger, 1978: 181). Also, they could
influence public hearings such as the so-called Valachi hearings by secret
prior questionings before the official hearings started (Maas, 1968: 29–30).
This way, they had a privileged position to shape the public perception and
understanding of the organized crime problematique in general, a position
of power ‘behind the scenes’ on knowledge structures that did not correspond
with their more visible authority.18
Reflecting on the Framework
In this last section I will conclude with a broader reflection on the proposed
framework by discussing (1) similarities and differences between the proposed
framework and current articulations of the Copenhagen School, (2) limits
and boundaries of the proposed framework and (3) consequences for a theory
of securitization.
As a consequence of re-conceptualizing securitization, elements of the
proposed framework can also be found in the current articulations of the
Copenhagen School. Overall, however, the new framework suggests a
less-decisionist, less-linguistic and more social/structural understanding of
securitization. Again, the agent–structure problematique creates the space
for the discussion.
Elements of actor-centrism of the Copenhagen School mainly reside in
the idea of the act of utterance bringing about change in conjunction with
an emphasis on decisions by the securitizing actor, often the political leader.
As Michael Williams (2003) has pointed out, this opens the theory for a
rather strong decisionist interpretation which shares central understandings
with Carl Schmitt. Their claim that the success of a speech act also depends
on the ‘authority of the speaker’ — and the speaker–audience relationship in
general — then introduces a more Bourdieuian and intersubjective understanding
— which, as I have argued before however, presupposes a more
reflected (e.g. structurationist) conceptualization of embeddedness. In contrast,
the concept of positional power — understood as the ability to influence
a process of meaning construction — is (1) more complex, capturing
‘hidden’ forms of influence, and (2) more structural. As such, it also departs
more clearly from a merely linguistic concept of power to give securitization
a stronger grounding in social theory.
Stritzel: Towards a Theory of Securitization
373
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 373
Structural elements of the Copenhagen School are rather weak and can at
best be found in their reflections on facilitating conditions, the speaker–
audience relationship and the ad hoc introduction of the term ‘field’ on p. 31
of Security: A New Framework of Analysis. Alternatively, as I have shown
earlier, structural elements can also be ‘re-imported’ from Wæver’s ideas on
foreign policy analysis, yet thereby creating more visible tensions with both
the decisionist and the Derridarian elements of securitization. Examined more
closely, their introduction of the term ‘field’, at least in the Bourdieuian
sense, is very problematic for the Copenhagen School because they thereby
introduce a rather objectivist element of Bourdieu’s theory. Regardless of
whether Judith Butler is right in calling Bourdieu’s concept of field an ‘inalterable
positivity’ (Butler, 1996: 34), it is hard to see how the Bourdieuian
field can be reconciled with the ‘intrinsic iterability’ of performative speech
acts and/or the Schmittian/decisionist elements. Thus, while the structural
elements of the Copenhagen School are either underdeveloped or very problematic,
this article is an attempt to give them a firm grounding by introducing
and elaborating the idea of embeddedness. In the proposed framework,
all other elements are defined in terms of their relatedness to existing (fluid
or sedimented) discourse.
Finally, the concept of (threat) text is, again, more complex than the concept
of speech act. While speech acts — if not understood as iterated through
social/political processes — are temporally and substantively limited to the
speech act event, the introduction of the term ‘text’ is to stress that linguistic
structures can evolve sequentially over longer periods of time, consisting
of more complex configurations which stem from various influences (as we
could see in the case of organized crime).
The main boundaries of the proposed framework result from the concept
of securitization itself and from the structurationist understanding of embeddedness.
As a reconceptualization of securitization, the framework is explicitly
discursive and does not address questions of intentionality, interests
or rationality that have been discussed intensively in political theories of
power (see e.g. Lukes, 1974, 2005). It also excludes extremely structural or
extremely actor-centric positions such as Wallerstein’s reflections on power
(Wallerstein, 2003) or many realist concepts (see e.g. Schmidt, 2005). Most
importantly, however, the externalist reading of securitization developed in
this article creates a more visible tension with the idea of performativity, i.e.
that actors and structures must be performed to exist, that they are constituted
‘retroactively’.
Ultimately, one can perhaps only have two separate, coherent readings of
securitization. The basic idea of uttering a state of emergency can be kept by
both readings. And I would argue, an externalist reading can also acknowledge
to some extent that threat texts can have a dynamic of their own which
European Journal of International Relations 13(3)
374
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 374
contains a certain degree of contingency or ‘magic’ which may, moreover,
contribute to change a given context. However, the more we search for a
comprehensive theory to understand the process of securitization, the clearer
we have to be with regard to the main centre of gravity from which the other
concepts gain their power. And as a result, the stronger the tension between
both readings becomes. At least in its most radical form, the internalist position
would hold that there is no social process involved but just the text and
its intrinsic iterability, no ‘con’-text, no positional power of actors being privileged
to enhance one text as opposed to others and no historical ‘pathdependency’
and intensification over time. Conversely, the externalist position
would always hold that it is misleading to think of an intrinsic iterability of
‘freely floating’ texts (or the definition of meaning as nothing more than a
decisionist act). Texts are always interwoven with relational dynamics of
power and meaning, rarely the work of any one person (Wodak, 2001a: 11)
and often even ‘historically intertextual’ in the sense that they
transform/translate past meaning structures into the present (Fairclough,
1992: 85).
To be sure, there would still be much room for communication between
the two wings. For example, Judith Butler’s critique of Bourdieu’s reflections
on ‘authorized language’ (Bourdieu, 1991: 107–16) in Performativity’s Social
Magic (Butler, 1996) can be read as such an attempt. Her reflections on
‘habitus’ in particular can be understood as an internalist position trying to
integrate elements of the idea of a social sphere into an essentially linguistic
position. Moreover, in Excitable Speech Butler refers to ‘sedimentation’ and
the existence of a ‘context’ throughout her book (see Butler, 1997). Similarly,
one can interpret her concept of ‘(heterosexual) matrix’ in Gender Trouble
(Butler, 1990) as an acknowledgement of the embeddedness of performances
in sedimented relations of power and meaning — yet, admittedly, for Butler
in a more Foucauldian sense than suggested in this article. Conversely, an
externalist position can move some steps from the social towards the linguistic
to investigate the embeddedness of utterances within spatially and temporally
contingent social contexts of meaning and power.
However, for an externalist position Butler overemphasizes the constitutive
role of language as an autonomous power at the expense of the social
sphere and its more fundamental influences on the very possibility of language
— especially in non-democratic environments. Moreover, an externalist
would be sceptical towards her strong claims that subjects are created,
that they are, only by virtue of constitutive utterances and thus cannot be
thought of as influencing the social struggle over meaning as pre-existing units.
As Terry Lovell (2003) could show, Butler’s reflections on performative
power and on ‘speaking with authority without being authorised to speak’
(Butler, 1997: 157) are based on a misrepresentation of the social sphere as
Stritzel: Towards a Theory of Securitization
375
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 375
something fully constituted, fixed and unequivocal (Lovell, 2003: 9). Yet,
since social contexts are never fully constituted, they do always allow room
for change, transgression and resistance. This is even more obvious the
stronger one follows a structurationist understanding of the social sphere
which stresses process and change through actor–action dynamics — rather
than any kind of structural determinism — as an essential conceptual feature.
This leads to the question of future theoretical steps towards a complex
understanding of securitization. Currently, scholars are confronted with the
problem that many concepts/terms of securitization are undertheorized. Yet,
this is not only true for an externalist reading but similarly for an internalist
one. The main problem for an internalist theory of securitization is that one
would need a clearer and more elaborated theory of the securitizing act.
A central question, and perhaps dividing line, for the concept of ‘act’ is to
what extent it should be thought of as an always political, contingent and
momentous move (as favoured by Wæver) as opposed to a more cultural and
sequential one, intensifying an already existing or even partly sedimented
practice (as favoured by this article). Following the latter position, securitization
studies would always need to be aware of the ‘poetics’ of the (threat)
text, the ‘hermeneutics’ of its context and their political embeddedness in
social relations of power.
While many scholars in favour of a more embedded approach to securitization
would probably share a certain scepticism towards a Derridarian
and/or decisionist approach to securitization — and Wæver’s fascination for
both — there isn’t one single way to specify ‘embeddedness’ and/or the securitizing
act. Several alternative suggestions have been made so far to better understand
the ‘act’ of securitization, e.g. to conceptualize it as an illocution
(Copenhagen School), perlocution (Balzacq, 2005), bodily performance
(Hansen, 2000) or as a technocratic practice (Bigo, 2000, 2002). The most
promising candidate for an explicitly internalist position would perhaps be to
elaborate the concept of performativity. An explicitly internal, retroactive
understanding of performative acts would then be clearly distinguishable from
the externalist position suggested in this article of conceptualizing the securitizing
act as an always situated articulation. Towards a more embedded reflection,
one may well follow an explicitly Bourdieuian and/or moderate
Foucauldian approach instead. The works by Bigo and Huysmans can be read
as such an attempt (Bigo, 2000, 2002; Huysmans, 2002, 2006) with their
focus on the role of security professionals and/or techniques of government/
governance for framing (in)security.19 In constrast to Bigo and Huysmans the
proposed framework stresses more strongly linguistic text–discourse dynamics.
Yet, metatheoretically, a broadly defined structurationist position is clearly compatible
with an explicitly Bourdieuian approach and I would say also with at
least a moderate Foucauldian one.20
European Journal of International Relations 13(3)
376
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 376
Conclusion
In this article I have suggested reconceptualizing the current articulation of
securitization posited by the Copenhagen School. The basic idea of security
as a speech act itself is too limited to allow a scholar to study ‘real-world’
securitizations. In reality, the speech act itself, i.e. literally a single security
articulation at a particular point in time, will at best only very rarely explain
the entire social process that follows from it. In most cases a security scholar
will rather be confronted with a process of articulations creating sequentially
a threat text which turns sequentially into a securitization. I have argued that
the attempts of the Copenhagen School to capture these dynamics with a
more comprehensive framework are hampered by too many contradictory
concepts creating several tensions and multiple centres of gravity in their
overall argument.
In light of this, I have suggested working with three layers of securitization:
(1) the performative force of an articulated threat text, (2) its embeddedness
in existing discourses and (3) the positional power of securitizing
actors. Using this revised framework, more empirical insights into the relative
importance and precise patterns of interaction can hopefully be generated
which can then be compared in a systematic way to reflect upon how to
improve securitization towards a more specific and comprehensive theory.
As a contribution to a consistent theory of securitization, I have argued for
distinguishing two main centres of gravity which emerge from an internalist
versus an externalist reading of securitization. Introducing the vocabulary of
internalism/externalism to the debate on securitization theory has two advantages.
First, it can be used as a lens directed towards the current writings of the
Copenhagen School. As a result, a more systematic understanding of many
otherwise isolated tensions and contradictions appears. In light of internalism/
externalism, it becomes clear that the Copenhagen School struggles with and
suffers from an attempt to have both, a social sphere (with ‘actors’, ‘fields’,
‘authority’, ‘intersubjectivity’, ‘audience’ and ‘facilitating conditions’) and a
(poststructural/postmodern) linguistic theory based on Derrida and performativity.
Yet, is not clear how one can have pre-established actors and prestructured
fields and at the same time claim that both are constituted retro–
actively by performative acts. Second, the vocabulary of internalism/externalism
directs more clearly towards future theoretical research. I have argued that a
possible route for an internalist wing of securitization theory would be to elaborate
the concept of performativity. For an externalist wing, this article has suggested
working with the idea of discursive embeddedness which should be
further specified in constant dialogue with a more consistent reservoir of
insights from (comparative) empirical studies. Yet, whether a comprehensive
theory, to which this article only attempts to make some suggestions, will
Stritzel: Towards a Theory of Securitization
377
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 377
indeed finally end up as internalist and/or externalist, and as Copenhagen or
beyond, is, ultimately, not for this article to decide.
Notes
I would like to thank Barry Buzan, Ole Wæver and the anonymous reviewers for comprehensive
and insightful suggestions (and disagreements). I have also received helpful
comments and encouragement from Michael Barnett, Felix Berenskötter, Christian
Büger, Karin Fierke, Anna Geis, Jef Huysmans, Matti Jutila, Anna Leander, Janus
Mortensen, Rita Taureck and Juha Vuori. Earlier drafts of this article have been presented
at the PhD workshops ‘Security Theory — Critical Innovations’ in November
2004 in Copenhagen, ‘Critical Approaches to Security in Europe’ in June 2005 in Paris
and ‘International Relations Theory’ in November 2005 in London. Last but not least
I would like to thank Otto Keck for his continuous trust and encouragement. I owe
much more to him than support for this project.
1. I owe the term ‘sequentiality’ to frame specific pitfalls of the Copenhagen School
to Michael Williams.
2. The notion of theory, and the related issue of causality versus constitution, are
highly controversial. They have in the past often led to seemingly incommensurable
understandings. A strictly empiricist/positivist understanding of testing ‘a set
of general propositions’ stands (or has been portrayed to stand) in stark contrast
to theory as merely a way of ‘structuring the empirical material’ — which is treated
by other scholars as a ‘framework’. Recent IR studies along the lines of critical
realism suggest, however, that questions of theory, causality and truth can escape
a stark ‘positivism–postpositivism’ dichotomy if a less empiricist understanding of
theory and a more demanding concept of causality is applied (see e.g. Kurki,
2006).
3. This similarity does not imply that I would buy (all of) Hempel’s strongly empiricist
research programme.
4. By constructivism I refer to approaches which stress (1) the social constructedness
of reality, (2) the ideational embeddedness of agency and (3) the focus on constitutive
research. The latter point makes me doubt whether to also subsume those
‘moderate’ constructivist studies under ‘constructivism’ which operate with a very
rigid understanding of causality and empirical testing and therefore appear indistinguishable
from soft versions of rationalism.
5. For a reconstruction of the influence of Derrida on Wæver, see e.g. Taureck
(2006) who elaborates Wæver’s own reflections in Security: The Speech Act
(Wæver, 1989). For a reflection on the role of Schmitt, see e.g. Williams (2003)
and Huysmans (2006). To my knowledge, no one has yet elaborated the influence
of Arendt; but see the very thoughtful article by Honig (1991) who compares
Arendt and Derrida in a way that is also relevant to better understand the
Copenhagen School’s insistence on the always political nature of the speech act
event.
6. This is why the procedural side of speech act philosophy is much less important
than Balzacq (2005) assumes. This is clearly laid down in Wæver (1989).
European Journal of International Relations 13(3)
378
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 378
7. It is controversial, however, whether Derrida really meant that all of reality operates
within ‘the text’ and that interpretation never ‘contextualizes’ (see e.g. Roberts and
Joseph, 2005: 109).
8. I would like to thank Janus Mortensen for introducing Zizek to me. For his own
critical appraisal of the Copenhagen School, see Mortensen (1998).
9. See Wæver’s own reflections on these problems (Wæver, 2003).
10. Strictly speaking, (linguistic) felicity conditions are conventions which regulate
the appropriate use of utterances. They are not — or not necessarily — identical
with (social) conditions of success. However, Wæver seems to conflate the two.
For example, in The EU as a Security Actor he writes: ‘On the basis of theories of
speech acts, we can say that there are three “felicity conditions” of a successful
speech act’ (Wæver, 2000: 252; emphasis added). These conditions then seem to
have the status of ‘necessary but not sufficient’ conditions since they always need
to be realized by a political act.
11. As we have seen before, Derrida has stressed that meaning is always disseminated
and can thus never be fixed.
12. My research on threat images shows that the framework is not limited to the
securitization of organized crime. It is here just used to illustrate the framework.
13. This sentence draws on Hay (2002: 127).
14. Yet, although the performative force of a particular text may therefore be the
most mysterious and contingent element of the three layers, a force that may
never be explained fully by its general features, it would be interesting to examine
empirically if some general patterns can still be identified by analysing which
linguistic structures generate which effects. There are first hypotheses for such
studies already, e.g. the assumption that more elusive structures are easier to
translate and are thus more powerful (Best, 2001; Hall, 1989).
15. I owe the term ‘intensification’ to Michael Williams.
16. I do not claim that the concept of positional power is original. There are several
examples both in the general literature on power and in the literature in IR which
have invented more relational and/or positional concepts of power. See e.g.
Baldwin (1989) and Guzzini (1993).
17. This draws a parallel to agenda-setting theories (Kingdom, 1995; Rochefort and
Cobb, 1994; Eriksson and Noreen, 2002) which have identified various reasons
why an actor may have the capacity to directly or indirectly influence issues that
are put on the political agenda as opposed to those that are ‘denied’ (Cobb and
Ross, 1997).
18. This connection between power and knowledge hints at a more Foucauldian
approach to securitization. I will come to this point below when discussing the
works of Didier Bigo and Jef Huysmans.
19. Unfortunately, both scholars have so far only applied their embedded reflections
systematically to internal security, especially to migration, and the EU context
where the lines between the domestic and the international are much more
blurred than elsewhere. My own comparative research on the representational
threat dynamics of organized crime and proliferating states does not support the
degree of similarity between internal and external security dynamics and discourse
that Bigo takes for granted (see Stritzel, 2006).
Stritzel: Towards a Theory of Securitization
379
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 379
20. To be compatible with a broadly defined structurationist approach, one would
mainly have to read Foucault as less (macro-)structural and more actor-centred
than is usually done.
References
Abrahamsen, Rita (2005) ‘Blair’s Africa: The Politics of Securitization and Fear’,
Alternatives 30(1): 50–80.
Archer, Margaret S. (1990) ‘Human Agency and Social Structure: A Critique of
Giddens’, in Jon Clark, Celia Modgil and Sohan Modgil (eds) Anthony Giddens:
Consensus and Controversy, pp. 73–84. Brighton: Falmer Press.
Archer, Margaret S. (1995) Realist Social Theory. The Morphogenetic Approach.
Cambridge: Cambridge University Press.
Austin, John L. (1962) How to Do Things with Words. Oxford: Clarendon Press.
Bachrach, Peter and Morton Baratz (1963) ‘Decisions and Nondecisions: An
Analytical Framework’, American Political Science Review 57(3): 632–42.
Baldwin, David (1989) Paradoxes of Power. Oxford: Basil Blackwell.
Balzacq, Thierry (2005) ‘The Three Faces of Securitization: Political Agency, Audience
and Context’, European Journal of International Relations 11(2): 171–201.
Barnett, Michael and Raymond Duvall (2004) Power in Global Governance. Cambridge:
Cambridge University Press.
Barnett, Michael and Raymond Duvall (2005) ‘Power in International Politics’,
International Organization 59(4): 39–75.
Best, Joel (ed.) (2001) How Claims Spread. Hawthorne, NY: de Gruyter.
Bhaskar, Roy (1979) The Possibility of Naturalism: A Philosophical Critique of the
Contemporary Human Sciences. Brighton: Harvester Wheatsheaf.
Bigo, Didier (2000) ‘When Two Become One: Internal and External Securitisation
in Europe’, in Morten Kelstrup and Michael C. Williams (eds) International
Relations Theory and the Politics of European Integration. Power, Security and
Community, pp. 171–204. London: Routledge.
Bigo, Didier (2002) ‘Security and Immigration: Towards a Critique of the
Governmentality of Unease’, Alternatives 27(1): 63–92.
Bourdieu, Pierre (1991) Language and Symbolic Power. Cambridge, MA: Harvard
University Press.
Burke, Edmund (1958 [1757]) A Philosophical Enquiry Into the Origins of Our Ideas
of the Sublime and Beautiful. London: Routledge.
Butler, Judith (1990) Gender Trouble: Feminism and the Subversion of Identity.
London: Routledge.
Butler, Judith (1996) ‘Performativity’s Social Magic’, in Theodore R. Schatzki and
Wolfgang Natter (eds) The Social and Political Body, pp. 29–47. New York:
Guilford Press.
Butler, Judith (1997) Excitable Speech. A Politics of the Performative. London:
Routledge.
Buzan, Barry and Ole Wæver (1997) ‘Slippery? Contradictory? Sociologically
Untenable? The Copenhagen School Replies’, Review of International Studies
23(2): 241–50.
European Journal of International Relations 13(3)
380
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 380
Buzan, Barry and Ole Wæver (2003) Regions and Powers. Cambridge: Cambridge
University Press.
Buzan, Barry, Ole Wæver and Jaap de Wilde (1998) Security: A New Framework of
Analysis. London: Lynne Rienner.
Cobb, Roger W. and Marc H. Ross (1997) Cultural Strategies of Agenda Denial:
Avoidance, Attack, and Redefinition. Lawrence, KS: University Press of Kansas.
Coker, Christopher (2002) ‘Globalisation and Insecurity in the Twenty-first Century.
NATO and the Management of Risk’, Adelphi Paper No. 345, IISS and Oxford
University Press.
Derrida, Jacques (1972) ‘Signature Event Context’, Glyph 1: 172–97.
Derrida, Jacques (1977) ‘Limited Inc a b c’, Glyph 2: 162–254.
Doty, Roxanne L. (1993) ‘Foreign Policy as Social Construction: A Post-Positivist
Analysis of US Counterinsurgency Policy in the Philippines’, International Studies
Quarterly 37(3): 297–320.
Eriksson, Johann and Eric Noreen (2002) ‘Setting the Agenda of Threats. An
Explanatory Model’, Uppsala Peace Research Papers No. 6.
Fairclough, Norman (1992) Discourse and Social Change. Oxford: Polity.
Fierke, Karin (1998) Changing Games, Changing Strategies: Critical Investigations in
Security. Manchester: Manchester University Press.
Fumagalli, Matteo (2006) ‘A Strategic-relational Approach to Post-Soviet Ethnopolitical
Mobilization’, paper presented at the PSA conference in Reading, April.
Giddens, Anthony (1984) The Constitution of Society. Cambridge: Polity.
Guzzini, Stefano (1993) ‘Structural Power. The Limits of Neorealist Power Analysis’,
International Organization 47(3): 443–78.
Hall, Peter (ed.) (1989) The Political Power of Economic Ideas: Keynesianism across
Nations. Princeton, NJ: Princeton University Press.
Hansen, Lene (2000) ‘The Little Mermaid’s Silent Security Dilemma and the
Absence of Gender in the Copenhagen School’, Millennium: Journal of
International Studies 29(2): 285–306.
Hay, Colin (2002) Political Analysis: A Critical Introduction. New York: Palgrave.
Hempel, Carl (1952) Fundamentals of Concept Formation in Empirical Science.
Chicago, IL: University of Chicago Press.
Hess, Henner (1973) Mafia and Mafiosi: The Structure of Power. Farnborough: Heath.
Honig, Bonnie (1991) ‘Declaration of Independence: Arendt and Derrida on the
Problems of Founding a Republic’, The American Political Science Review 85(1):
97–113.
Huysmans, Jef (2002) ‘Defining Social Constructivism in Security Studies: The
Normative Dilemma of Writing Security’, Alternatives 27(1): 41–62.
Huysmans, Jef (2006) The Politics of Insecurity: Fear, Migration and Asylum in the
EU. London: Routledge.
Jessop, Bob (2005) ‘Critical Realism and the Strategic-Relational Approach’, New
Formations 54(2): 40–53.
Kenney, Dennis J. and Jom O. Finckenauer (1995) Organized Crime in America.
Belmont, CA: Wadsworth.
Kingdom, John W. (1995) Agendas, Alternatives and Public Policies, 2nd edition.
New York: HarperCollins.
Stritzel: Towards a Theory of Securitization
381
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 381
Kurki, Milja (2006) ‘Causes of a Divided Discipline: Rethinking the Concept of
Cause in International Relations Theory’, Review of International Studies 32(2):
189–216.
Lampe, Klaus von (1999) Organized Crime: Begriff und Theorie organisierter
Kriminalität in den USA. Frankfurt a.M.: Peter Lang.
Laustsen, Carsten Bagge and Ole Wæver (2000) ‘In Defence of Religion: Sacred
Referent Objects for Securitization’, Millennium: Journal of International Studies
29(3): 705–39.
Lipset, Seymour Martin (1996) American Exceptionalism: A Double-Edged Sword.
New York: W.W. Norton.
Lovell, Terry (2003) ‘Resisting with Authority: Historical Specificity, Agency and the
Performative Self’, Theory, Culture & Society 20(1): 1–17.
Lukes, Steven (2005) ‘Power and the Battle for Hearts and Minds’, Millennium:
Journal of International Studies 33(3): 477–93.
Lukes, Steven (1974) Power. A Radical View. Basingstoke: Macmillan.
Maas, Peter (1968) The Valachi Papers. New York: Putnam.
McSweeney, Bill (1996) ‘Identity and Security: Buzan and the Copenhagen School’,
Review of International Studies 22(1): 81–93.
McSweeney, Bill (1998) ‘Durkheim and the Copenhagen School: A Response to
Buzan and Wæver’, Review of International Studies 24(1): 137–40.
McSweeney, Bill (1999) Identity and Interests: A Sociology of International Relations.
Cambridge: Cambridge University Press.
Mann, Michael (1985) The Sources of Social Power, Volume I. Cambridge: Cambridge
University Press.
Moore, William H. (1974) The Kefauver Committee and the Politics of Crime 1950–1952.
Columbia, MO: University of Missouri Press.
Mortensen, Janus (1998) Sikkerhed som talehandling: En kritisk gennemgang af
Wævers sikkerhedsbegreb. Copenhagen: Forlaget Politiske Studier.
Paoli, Letizia (2003) Mafia Brotherhoods: Organized Crime, Italian Style. Oxford:
Oxford University Press.
Parker, John (2000) Structuration. Buckingham: Open University Press.
Rasmussen, Mikkel Vedby (2001) ‘Reflexive Security: NATO and International Risk
Society’, Millennium: Journal of International Studies 30(2): 285–309.
Roberts, John A. and Jonathan Joseph (2005) ‘Derrida, Foucault and Zizek: Being
Realistic About Social Theory’, New Formations 54(2): 109–20.
Rochefort, David A. and Roger W. Cobb (eds) (1994) The Politics of Problem
Definition: Shaping the Policy Agenda. Lawrence, KS: University Press of Kansas.
Ruth, David E. (1996) Inventing the Public Enemy: The Gangster in American
Culture 1918–1934. Chicago, IL: University of Chicago Press.
Schlesinger, Arthur M. (1978) Robert Kennedy and His Times. London: Deutsch.
Schmidt, Brian C. (2005) ‘Competing Realist Conceptions of Power’, Millennium:
Journal of International Studies 33(3): 523–50.
Smith, Dwight C. (1975) The Mafia Mystique. New York: Basic Books.
Stritzel, Holger (2006) ‘The Möbius Ribbon of Insecurity: Myth or Reality?’, paper
presented at the conference ‘Critical Approaches to Security in Europe II —
Constructing Insecurity and the Political’, 29 September–1 October, Tampere, FL.
European Journal of International Relations 13(3)
382
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 382
Stritzel: Towards a Theory of Securitization
383
Taureck, Rita (2006) ‘Securitisation Theory — The Story So Far (Part One):
Theoretical Inheritance and What it Means to be a Poststructural Realist’, paper
presented at the 47th annual ISA convention, 22–25 March, San Diego, CA.
Wæver, Ole (1989) ‘Security, the Speech Act: Analyzing the Politics of a Word (and
the Transformation of a Continent)’, unpublished manuscript.
Wæver, Ole (1995) ‘Securitization and Desecuritization’, in Ronnie D. Lipschutz
(ed.) On Security, pp. 46–86. New York: Columbia University Press.
Wæver, Ole (1997) Concepts of Security. Copenhagen: University of Copenhagen,
Institute of Political Science.
Wæver, Ole (2000) ‘The EU as a Security Actor: Reflections from a Pessimistic
Constructivist on Post Sovereign Security Orders’, in Morton Kelstrup and Michael C.
Williams (eds) International Relations Theory and the Politics of European
Integration, pp. 250–94. London: Routledge.
Wæver, Ole (2001) ‘Identity, Communities and Foreign Policy’, in Lene Hansen and
Ole Wæver (eds) European Integration and National Identity, pp. 20–50. London:
Routledge.
Wæver, Ole (2003) ‘Securitization: Taking Stock of a Research Programme in
Security Studies’, unpublished manuscript.
Wæver, Ole (2004) ‘Ole Wæver’s 10,’ Tidsskriftet Politik 7(4), December; (http://
www.tidsskriftetpolitik.dk/index.php?idϭ125, December 2005).
Wallerstein, Immanuel (2003) The Decline of American Power. New York: The New
Press.
Weldes, Jutta (1996) ‘Constructing National Interest’, European Journal of
International Relations 2(3): 275–318.
Wight, Colin (1999) ‘They Shoot Dead Horses Don’t They? Locating Agency in the
Agent–Structure Problematique’, European Journal of International Relations
5(1): 109–42.
Wight, Colin (2006) Agents, Structures and International Relations: Politics as
Ontology. Cambridge: Cambridge University Press.
Williams, Michael C. (2003) ‘Words, Images, Enemies: Securitization and International
Politics’, International Studies Quarterly 47(4): 511–31.
Wodak, Ruth (2001a) ‘What CDA is About — Summary of its History, Important
Concepts and its Developments’, in Ruth Wodak and Michael Meyer (eds) Methods
of Critical Discourse Analysis, pp. 1–13. London: Sage.
Wodak, Ruth (2001b) ‘The Discourse-historical Approach’, in Ruth Wodak and Michael
Meyer (eds) Methods of Critical Discourse Analysis, pp. 63–94. London: Sage.
Wodak, Ruth and Michael Meyer (2001) Methods of Critical Discourse Analysis.
London: Sage.
Zizek, Slavoj (1989) The Sublime Object of Ideology. London: Verso.
357-384 EJT-80128.qxd 25/7/07 8:38 AM Page 383