The theory act: Responsibility and exactitude as seen from securitization

What kind of theorizing of politics?

The tragic irony of political science is that exactly scholars of politics tend to erase politics. Two versions exist. The dominant one explains away politics by causally making it derivative of other factors. A second destructive move dissolves political action into political theory, which in this context means philosophy. Neither causal social science nor political philosophy is as such misplaced in relation to understanding politics, but in their heavy handed versions, they fail to respect the centrality of politics as action. Causal explanation from non-political factors is probably the more obvious case, but political theory can have the same effect when politics is placed as a philosophical question to be solved intellectually, implying that this provides a secure guide for taking the correct approach to politics. Then, action is transposed into contemplation and human interaction becomes inner thought.

Hannah Arendt has, in my view, compellingly argued that at least one crucial dimension of politics is action that goes on between men, corresponding to the human condition of plurality. Because politics takes place among people, in-between us, because power only emerges when people act together, it basically consists of action directed to and dependent on the reaction of others, not doing things directly. History is not ‘made’ in the sense that one actor has a plan and then carries it out. Politics is always more uncontrolled – action leads to other acting and so forth in chain reactions. Theories should never explain away that irreducible, open element that is the in-betweenness of politics. ² It is not enough to work out safe, progressive meta-positions – politics demands the wager of action with sometimes unpredictable effects, and whether one’s action was good or not will be established only later by the storyteller. ³ Theory should be viewed as action similarly – both by those who invent/design one and by every user who performs the act of adding this theory to a study. ⁴ Given that a theorist cannot know every specific case in advance, the main responsibility of the theorist is to design the theory structurally so that a truly political understanding becomes unavoidable.

However, it is not evident how an empirical research agenda on securitization could be constructed directly on this basically philosophical premise. Enter, speech act theory. Securitization theory was built from the start on speech act theory, because it is an operational method that can be designed to protect politics in Arendt’s sense. ⁵ Put in short form, the political conception of securitization theory is inspired by Arendt, implemented through speech act theory.

Securitization theory can both deliver explanatory social science and respect politics as open space by partaking in development of a general, emerging approach that might be called ‘the collective theory of political speech acts’ – with potential relevance far beyond security studies. Within speech act theory, it has been articulated by Marina Sbisà, and I have attempted a similar argument especially in various ‘replies to critics’ on securitization’s meta-theory. Sbisà has argued that Austin’s theory of speech acts entails that the illocutionary effect (‘done in saying’) is co-produced by the audience in a more extensive sense than pure uptake, and the status transformation entailed in, for example, securitization is a redefinition of the rights and responsibilities of actors, not just a form of communication (as has become ‘the received view’ of speech act theory, through especially Searle). ⁶ Similarly, I have insisted on analyzing securitization as an illocutionary and not a perlocutionary act in order to organize the theory around the constitutive, transformative event of actors reconfiguring the relationship of rights and duties rather than seeing an external cause–effect relationship between speech and effects. ⁷ Marina Sbisà has demonstrated why ‘deontic modal competences’ should be tied to the illocutionary act, and why Searle waters out Austin’s speech act theory when he (as most other mainstream philosophers and linguists) re-creates a divide within speech act theory between language and action. The radical potential of Austin’s theory is lost when these traditionalists separate everything social from the illocutionary, and de facto reduces the illocutionary act to communication, and thus – against Austin’s original intent – removes the possibility that speech can be really action. We witness surprisingly, ‘the contrast between saying and doing, reborn within the very context of speech-act theory’. ⁸

In the co-produced political illocutionary event, a social situation goes through a ‘phase transformation’ to become defined and regulated by security (or in other cases, for instance risk). This has taken securitization theory beyond a focus on discourses and rhetoric (which it originally shared with a general trend toward ‘constructivism’ and ‘language’ within the new security studies) to a theory of political co-production between multiple actors of social states – ‘the mutual shaping of the agents of the interactional event’. ⁹ Such changes in the participants’ ‘deontic modal competences’ has the character of ‘event’ in the sense explored by Jacques Derrida, ¹⁰ where something happens that is irreducible, and therefore also irreducible to the tempting interpretation in terms of ‘performativity’.

With this theory, political events can be studied empirically as social phenomena. The event of such changes between ‘political states’ has to be investigated as political co-production in the relationships among actors, neither in or by single actors nor as socially determined and thereby unpolitical. ¹¹ The effects of these changes of the political situation can, however, be studied in more classical, causal language as effects on society, process, and polity. Conditions for these events include material as well as discursive and institutional matters to be studied. With this distinct version of speech act theory at the center, it becomes possible to integrate three distinct stages of analysis where different methodologies are appropriate for each. In the first stage, the aspirations of multiple actors are related to societal conditions. In the second stage, speech act theory allows for analysis of complex arrangements in terms of the political codification that constitutes them as particular relationships. Third, effects of a specific form of organization (in casu securitization) have been suggested in various literatures and expected effects on political, legal and socio-psychological life could be tested in case studies. Political theory plays a key role in interrogating the deontic modal constellations politically and ethically.

Interestingly, (as demonstrated also by this forum) the different strands of securitization theory convergence on combining multiple methods and forms of analysis. A crucial task will now be to investigate what combinations are internally consistent. My ‘illocution focused’ version of securitization theory claims it can integrate causal explanations, social mechanisms, hypothesis testing and political theory, by systematically organizing the different parts around securitization as specific kind of political event. The consistency and consequences of this set-up should be probed, and similarly for the other strands.

Insistence on illocutionary force as the center of securitization studies is exactly motivated by the resulting kind of theory, not least in terms of how it handles politics. Therefore, the so-called sociological version of securitization theory ¹² is problematic. It defines securitization as ‘practices’ and ‘processes’, and its general explanatory ideal is to capture all the interactions whereby people influence each other – ultimately drifting toward a reconstruction of individual dispositions and reactions, a sender–receiver view of communication, a mentalist conception of meaning and a cause–effect understanding of social relationships, including politics. As social science, this is unproductive because it leads to 1:1 mapping rather than theory and explanation, and politically, it replaces responsibility with causality. ¹³ This is not a politically neutral choice between theories, not purely analytical – but partly a political choice, partly about the kind of knowledge one aims at.

For instance, the concept of ‘audience’ becomes, in the sociological version, an empirical question, but that is only possible if securitization has been defined in a way where it can happen without an audience, whereas the ‘political’ (illocutionary) version of the theory studies securitization as a relationship, and therefore the question becomes how speaker and audience jointly reconfigure their relationship. Consequently, if the audience is found (cf. Balzacq, this forum) to participate less explicitly than originally expected, this has to be re-theorized as a form of political interaction. When the sociological version of securitization theory allows for ‘securitization’ to occur without political co-production between securitizing actor and audience, this implies that securitization no longer means the same thing, and the whole ideal-type procedure collapses.

So, what kind of theory is securitization? It is a theory of something political – security – that is structured around a core concept designed to preserve a truly open, political, constitutive space of human inter-/co-action (specifically handled through a specific version of speech act theory of collective illocutionary acts) with other forms of research ordered into designated roles around this center: causal and philosophical analysis as well as discourse and conceptual analysis.

The role of theory

Theory has knotty relations both ‘upwards’ (to meta-theory and paradigms) and ‘downwards’ to empirical analysis.

On the first front, we should center debate on theory, not paradigms (International Relations (IR)) or meta-theory. Debates keep bringing up the weird argument that securitization theory is inconsistent because it combines elements from, for example, realism and post-structuralism, or causal and constitutive logic. This criticism presupposes that these larger groups are internally consistent and mutually isolated. On the contrary, we all know numerous examples of internally consistent theories that draw on several traditions – and many more examples of theories that stay within their ‘box’ and yet are horribly inconsistent. Therefore, investigations of the internal consistency and productivity of research traditions should focus on distinct theories, not loose collections hereof. IR paradigms are most important for the sociology and history of the discipline, less for actual analysis. Meta-theory is important, but it should support, not replace theory of – in this case – security. Patrick Jackson’s powerful plea for philosophy of science actually points in this direction too. ¹⁴ It does not point to one form of theorizing, but argues that with a higher level of training in philosophy of science, we could discuss our theories more competently.

More recently, I have attempted to clarify with philosophy of science from Robert Giere, Nancy Cartwright, Kenneth Waltz and others, ¹⁵ especially the relationship between theory, research design and empirical analysis. A theory is here a model that does not in itself explain, but it forms a coherent system in relation to which it is possible to both compare instances and formulate specific hypotheses. Therefore, an ‘empirical’ case study always involves much theoretical work to specify and ‘attach’ the theory to a specific interest. A theory on its own never predicts or explains; a set-up for a specific purpose, often involving several theories, might be able to do so. Reporting from empirical analyses should therefore include much more attention to how theories were posited and utilized in order to carefully draw more general lessons, where too often empirical findings are directly taken to prove what is ‘missing’ or ‘wrong’ in the theory. Empirical exactitude is an ideal for the analysis, and the question to the theory is how it helped to produce that analysis.

In this forum, Balzacq has helpfully brought Weber’s ideal types in play. ¹⁶ The difference between ‘model’ in philosophy of science and ideal type is too big a question for now, but two problems with the way Balzacq uses ideal type can be briefly mentioned. First, it is not possible to construct an ideal type outside of perspective. According to Weber, one necessarily abstracts with the intent to ‘exaggerate’ a particular dimension or dynamic and thus, from a theoretical angle, optimize intellectual coherence. An ideal type is constructed from ‘a theoretical and hence “onesided” viewpoint’. ¹⁷ Several different ideal types can be constructed for the same concept. Balzacq draws up a more disjointed list of features that unify cases of securitization. This comes close to what Weber strongly warns against: mixing up ideal types with classificatory concepts.

Second, by attaching ‘model thinking’ (in the shape of ideal types) to the concept of securitization, but not the theory/ies, Balzacq leaves it open what kind of theory his version of securitization theory is. This allows the sociological ‘theory’ to become an unstructured mapping of actors’ perceptions/predispositions and their exchanges of communication – with no theoretical focal idea, in particular no organizing theoretical tool at the collective, political level. This is why the classical Copenhagen School version of securitization has a theory that takes the form of a model.

This raises the question about sociology, philosophy and political science, which also Williams point to in his important contribution. Recent debates on securitisation ¹⁸ contrasts ‘sociology of securitisation’ to (classical Copenhagen style) ‘theory of securitisation’. There is some truth to this, if the former means pragmatism and micro-sociology and the latter means illocutionary logic. Then, it is an important choice because this kind of ‘sociology of securitization’ dissipates into unending chains of cause–effect, where everything has to be taken into account at each stage to understand how each unique actor interprets and reacts to each input. ¹⁹ But more broadly, sociology could and should mean more than this. The expression ‘sociology of’ has generally become a popular move these days (replacing ‘the social construction of …’?). It makes a play on something being more empirical and attentive to local variations, while simultaneously anchored in a social science. Probably, it would be more productive to think of ‘sociology of’ and ‘politics of’ as different kinds of theory. Political theory similarly has achieved a confusing status in these debates, often as an ethical reflection external to the theoretical concepts, where it would play a stronger role as the political theory of those operations represented by the key concepts in the social theory. Securitizsation is that kind of theory, organized around a theoretical conception of a distinct political move, allowing for causal analysis of its consequences, sociological analysis of social patterns that condition political possibilities and political theory exploring human life under different arrangements.

Conclusion … if any

Has the research program been cumulative? Do we need the theory for the coming voyages? On most conventional measures, ‘securitization’ is an academic success story. However, some might speculate whether it is the (classical) theory or just the concept that has been catchy? Many references are to the ‘idea’ or the ‘slogan’. However, numerous dissertations and other studies have been made with this ‘framework for analysis’, so it seems that more than the concept has proven useful. The specific meta-theoretical explications were not available at the time, but probably present implicitly. Whether they (i.e. ‘the kind of theory’) made the usage more or less productive would be an interesting second-order research question.

Despite the Forum’s ‘if any’ question, it seems that the classical Copenhagen version of the theory has certainly been a theory and of a very specific kind. The critical question is rather whether it has been too much of a theory – whether it is necessary and/or helpful to play the theory card that hard or more is gained by a ‘less theoretical’ approach such as, for example, the so-called ‘sociological’ version. To assess this, the discipline needs to cultivate a more elaborate terminology and publication format for assessing how a theory participates in specific studies – what exactly does it do?