Securitization as a context-changing speech act

Abstract

Securitization theory is the most well-known Nordic contribution to the study of security. A key division within its 35-year-long history of debate is its bifurcation into two main interpretative communities, those of the ‘philosophical’ and ‘sociological’ interpretations of securitization theory. The present article engages with this division by examining the issue of context and the transformative capacities of speech acts, and discusses what it would entail for the philosophical interpretation to take context seriously. In this way, the article contributes by disrupting the by now received reading of contexts and generations within securitization studies. The way the sociological approaches have viewed context in securitization risks the merger of the conventional and natural efficacies of speech acts. It is, therefore, necessary to keep the illocutionary effects of securitization separate from its perlocutionary effects. Securitization moves may be about swinging audiences and producing emotions in them, but securitization may also be about transforming the rights and duties that relate to an issue and the relationship between the speaker and audience. Studying these separately leads to different research interests and requires different methods: illocutionary forces can be discerned from the acts alone whereas perlocutionary effects cannot.

Introduction

Securitization theory has inspired the largest discussion within security studies within the last three decades, yet there is no grand view that would bring all the numerous empirical studies conducted with it into a synoptic image of what we know about securitization. This is partly due to the many variations of the theory that have been formulated over the decades. One of the key dividing lines here has been whether the focus should be on who or what does securitization (what ‘triggers’ it) or on what securitization does (what it ‘triggers’) (Guzzini, 2011: 336– 337). How ‘context’ has been viewed here has been vital and matters greatly for the potential of bringing empirical studies of securitization into a comparative setting. Indeed, if context is considered as determining over speech acts of securitization, comparisons become virtually impossible due to the myriad of political, social, and historical factors that bear on such processes: contexts are never ‘absolutely determinable’ or ‘saturated’ (Philipsen, 2020: 144–145). On the other hand, if what securitization speech acts do to their contexts is the focus, comparison becomes more feasible.

Many empirical studies of securitization suggest that securitization has to follow a certain order of events, and that one has to follow the other. The idea is that these kinds of consistencies can be detected with research, which may even enable the prediction of events with certain probabilities. An ‘if {a,b,c}, then securitization happens, and with it the defined effects {x, y, z}, typically involving some exceptional measures’ causal diagram of securitization (Patomaki, 2015: 129) is however problematic in Ole Wæver’s (2015) view: political status transformations should not be studied ‘neither in or by single actors nor as socially determined and thereby unpolitical’ (Wæver, 2015: 123); ‘politics is a sphere of re-presentation of the social – it is not the social, nor can it ever be’ (Wæver, 1989: 11). While there are consistencies in what ‘triggers’ securitization, and what it ‘triggers’ (Guzzini, 2011; Patomaki, 2015) – otherwise we would not have a model of securitization to begin with – these triggers are not deterministic. This is because securitization is neither necessary, nor sufficient, to achieve ‘security’ understood as a policy or some means to repel an existential threat: threat perceptions, securitizations, and security actions are indeterminate (Vuori, 2011). Nevertheless, it is easier to discern what securitization does to a context than to discern what it was in a context that brought securitization about.

This division has remained perennial because, as it stands, the concept of context is under-explicated in the original presentation of securitization theory and subsequent developments of it. At the same time, context and the need to take it into account are mentioned in most studies of securitization, yet what is meant by ‘context’ is not made explicit or thoroughly discussed. This has hindered empirical comparisons. Accordingly, I argue that explicating the meaning and role of context for the illocutionary variant of securitization theory is necessary for doing comparative studies of securitization (cf. Karyotis et al., 2025), and for the evaluation of pre-existing case studies through a common criterion that would go beyond reviews of the corpus (Baele and Jalea, 2023).

While such tasks are of vital importance, the stakes regarding the role of context in securitization theory are higher than just the facilitation of empirical comparison. Indeed, the role of context concerns the basic approach and application of the theory as such. It concerns what the theory is about and what it is for (cf. Wæver, 2015). Accordingly, the present article contributes to securitization studies by elaborating on the key division regarding contexts that lays at the heart of ‘philosophical’ and ‘sociological’ approaches to securitization theory, and by disrupting the by now received story of these generational veins of the literature. It provides philosophical clarity, and an opening for empirical comparison and evaluation by presenting a consistent take on speech act theory that fits the original purposes of the theory. The explication I lay out below sits firmly within the ‘philosophical’, or illocutionary interpretation of securitization theory, and as such, joins Wæver’s position. It shows though, how the approach can remain consistently within speech act theory without having to neither rely on Jacques Derrida nor be reduced to causal social mapping when taking context seriously in the illocutionary study of securitization.

I focus here on how securitization speech acts can alter sociopolitical contexts. This focus makes it possible to deem what needs to be considered when context is taken into account in the illocutionary interpretation of securitization theory. It allows for maintaining the responsibility of actors that speak security while remaining consistently within speech act theory. It also means that some detours are required: how intentions and success are understood relate closely to context and its transformation. To achieve a clearer understanding of the issues involved, I begin the article by revisiting the division between the ‘sociological’ and ‘philosophical’ variants of the theory. This clarifies the stakes of the interpretations.

I proceed then to discuss social transformations in various interpretations of speech act theory. Such excursions beyond International Relations (IR) and Security Studies are necessary because the theory of securitization ‘was built from the start on speech act theory’ (Wæver, 2015: 122). This is why the latest discussions there should be examined from the viewpoint of securitization theory, and they should also inform its potential development. The most relevant basic distinction within speech act theory here is the one between illocutionary and perlocutionary speech acts, or acts in saying something and acts by saying something (Austin, 1975 [1962]). Speech act theory suggests that people interact with the language they use by infusing it with illocutionary forces, which are used to produce (perlocutionary) effects in other people that can affect the feelings, attitudes and subsequent behaviour of the hearer(s). Unlike perlocutions, illocutions are conventional: they are done conforming to conventions that are historicized and dependent on social and cultural factors (Austin, 1975 [1962]). In this way, illocutionary acts have the capacity to transform the deontology ¹ of certain situations (Sbisà, 2007; Searle, 2011 [2010]); illocutionary acts can affect change in states of affairs that belong to the same level of reality as norms (Sbisà, 2009: 45). This basic distinction between speech act types is also at the heart of the dividing line between the philosophical and sociological interpretations of securitization theory: the philosophical approach emphasizes on what is done in the illocutionary act of securitization whereas the sociological approach favours what is done by the perlocutionary act of securitization.

This leads my argument to the explication of what context means in various readings of speech act theory. For securitization theory, the issue of context boils down to responsibility and agency: if context makes securitization, there is no agency or responsibility to be had for securitizing actors. For Thierry Balzacq (2019), ‘emphasizing speech acts means that internal validity trumps external validity, as performativity is endogenous to speech act rules’ (p. 335). This misconstrues the issue, as if focusing on the speech act would leave out the audience and the situation of making the act. In contrast, I show how the ‘external’ has a necessary role to play regarding both even when illocutionary speech acts are emphasized. Some have pointed to ‘the combined effect of discourse, practices and context’ (Vigneau, 2019: 191) in how security becomes constituted intersubjectively, or how ‘a combination of language and privileged social position can “shift” or “move” the situational context’ (Rythoven, 2020: 485). My discussion shows how such elements are considered in speech act theory, and how they are relevant for the illocutionary version of securitization theory too.

Indeed, contexts are not irrelevant for the ‘illocutionary’ or ‘internalist’ (Stritzel, 2007) version of securitization theory: securitization moves transform contexts irrespective of their success, and contexts are part of all speech acts, including the elementary ones that form the complex speech act of securitization (Vuori, 2011): ‘all security speech acts have the power to affect their context’, and failed acts enter the context too (Philipsen, 2020: 152). Like Lise Philipsen (2020), I take the potential of creating a new reality with performative speech acts seriously, as ‘inherently political’ (p. 150).

The final step is to delve into the crux of the matter, namely, the distinction between illocutionary and perlocutionary speech acts as they relate to securitization. This is necessary because of how the sociological approaches have viewed context in securitization risks the merger of the conventional (illocutionary) and natural (perlocutionary) efficacies of speech acts. While some securitization speech acts can have both aspects, they need to be kept theoretically and analytically separate, lest one falls into the Derridean fallacy of misrepresenting speech act theory. I close the article by presenting the resolution to the issue and paving the way for future securitization studies.

What is the problem? Conceptualizations of context in securitization theory

For Wæver (2011: 470), the theory of securitization combines a Schmittian concept of security with an Arendtian concept of politics, as it is ‘strung between Schmittian (anti)political exceptions and an Arendtian co-creation’ (Greenwood and Wæver, 2013: 501). The theoretical mix that inspired Wæver also includes post-structuralist elements from Derrida (Stritzel, 2014). Wæver deployed these veins of thought in order to frame politics as a realm of the unexpected and undetermined, where something new could happen. This overall view of politics is made operational through John L. Austin’s speech act theory: ‘the political conception of securitization theory is inspired by Arendt, implemented through speech act theory’ (Wæver, 2015). Holger Stritzel (2014) has argued that drawing on such multiple theoreticians makes the theory of securitization inconsistent. My argument here though is that the transformative power of speech acts is consistent within speech act theory, which can serve the purposes of multiple theories of politics.

The so-called ’sociological approach’ (Balzacq, 2005, 2011) has emphasized context in social processes of securitization. For Balzacq (2005, 2011), both distal and proximate contexts are dominating for acts of securitization, albeit not determining (Balzacq, 2015). Along similar lines, Stritzel (2007, 2014) who phrases the division as being between ‘internalist’ and ‘externalist’ approaches to securitization has also discussed pools of resonant values, and the contextualization of speech acts that pertain to securitization. Considering securitization to be a translation of pre-existing issues or notions of security makes sociolinguistic contexts crucial (Stritzel, 2014).

On the side of the ‘philosophical’ (Balzacq, 2011), ‘internalist’ (Stritzel, 2007), or the ‘illocutionary’ and ‘political’ (Wæver, 2015) approach to securitization theory, Wæver has not considered context to be as primary as the sociological approaches have. For him, to view securitization purely as ‘practice’ or as ‘process’ would lead to the inclusion of all interactions between people and, in effect, a 1:1 mapping of social situations rather than theorization and explanation; doing so would concomitantly replace responsibility and politics with causality (Wæver, 2015: 124). As in the study of history, too much emphasis on context may risk implicit or explicit circularity where context is ‘constructed as a function of the explanation it is called on to provide’ (Boureau, 1999: 221–222). For Wæver, of importance are the acts of securitization, what they do politically, and what they bring about, rather than the social context that would produce them (Buzan et al., 1998):

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. (p. 46)

Indeed, context in Wæver’s approach is most pre-eminently present as ‘facilitating conditions’ (Buzan et al., 1998: 32), which ‘are the conditions under which the speech act works, in contrast to cases in which the act misfires or is abused’ (Buzan et al., 1998: 32). These conditions fall into two categories: ‘1) the internal, linguistic-grammatical – to follow the rules of the act, and 2) the external, contextual and social – to hold a position from which the act can be made’ (Buzan et al., 1998: 32–33). These have been conceptualized in the subsequent literature in various ways, by for example pointing to the legal and political distinctions in the audiences of securitization (Wertman and Kaunert, 2022). Contexts have also been viewed as being in a co-dependent relationship with securitizing agent’s actions where such actions can be both facilitated and constrained by contexts, and where successful moves modify the context (Vigneau, 2019: 194).

The role of facilitating conditions has been discussed at length within securitization studies. For example, Stritzel (2007: 379, 2014) has noted how Wæver has confused felicity conditions with facilitating conditions in their various formulations. For Austin (1975 [1962]), felicity conditions were conventions that regulate the appropriate use of utterances and they were necessary conditions to achieve successful speech acts; felicity conditions were not entirely identical with (social) conditions of success, but dealt with issues of uptake, for example. ² In contrast, facilitating conditions are part of the more general social or political context. At the same time, facilitating/impeding conditions are about the politics of securitization rather than about the felicity of elementary speech acts that comprise the more complex speech acts of securitization. Indeed, felicity conditions are those types of conditions that could be termed necessary or sufficient, whereas facilitating/impeding conditions are conditions of the situation or context of the utterance. Juha A. Vuori (2011) has argued that it may be more appropriate to term these two types of things as felicity conditions and facilitation/impediment factors to avoid confusing felicity and facilitation.

Beyond Wæver, there have been a number of propositions for how to engage context in the securitization studies literature. These include sets of questions for identifying relevant dimensions for revealing how security is constructed in context (Ciuta, 2009: 317), and how the settings of on- and back-stage discussions of experts have a bearing on which kind of language is used (Salter, 2008). Most prominently, Balzacq (2005, 2010: 61) has criticized Wæver’s approach of reducing securitization to the acts of the speaker, to the illocutionary aspect of speech acts, and of not leaving any role for the audience of securitization, for the perlocutionary effects of securitization. ³ Balzacq (2005: 172–173, 2010: 63–65) subsequently argues that securitization is better understood as a strategic (pragmatic) practice that occurs within, and as part of, a configuration of circumstances, including the context, the psycho-cultural disposition of the audience, and the power that both speaker and listener bring to the interaction. This conceptualization has been subsequently refined to point to state rules and political perceptions as sources of legitimacy (Wertman and Kaunert, 2022: 73), and even branded the ‘audience turn’ in securitization studies (Soares, 2022).

The sociological approaches have emphasized context as a remedy for the perceived universality of securitization theory and its understanding of what security is. In this vein, Stritzel (2007: 369–370) separates the sociolinguistic reservoir of analogies and contracts from the sociopolitical context of more sedimented structures and positions of power. In turn, Matt McDonald (2008: 571) argues that to develop a universal framework to study the construction of security issues by analysing speech acts, would mean to downplay contextual factors.

The problem in general is that contextual factors cannot be defined in too much detail within theoretical models that are intended for wide use as such factors depend to a large extent on the political order, as well as other sociopolitical factors of various societies. This has for example been noted in regard to the multiple forms audiences have taken in various securitization processes (Cote, 2016; Jarvis and Legrand, 2017; Potenz, 2019), and how conceptualizations of securitization audiences need to take the cultures, norms, and rules of particular countries into account (Wertman and Kaunert, 2022: 72). It even seems that calls to ‘contextualize’ securitization speech acts within a model (Stritzel, 2014: 48–49) risk confusing theories with their application: to use a theoretical model in empirical research is to contextualize it; ‘all theories are analogs when applied to the world’ (Kaplan, 2014: 48), and no process of securitization is identical to another.

Indeed, a ‘theory on its own never predicts or explains’ (Wæver, 2015: 125), it is used as a basis for comparing empirical instances and to generate specific hypotheses; ‘when theories are applied to the world, their elements require specification’, and no conclusions of the social world can ‘be derived from a theory in the absence of such intermediate steps’ (Kaplan, 2014: 52). A theory names the relations of the qualities of its various elements, but does not mention what these particular elements are, as this would ‘reduce theory to an unlimited assembly of unrelated propositions’ (Kaplan, 2014: 56). When a theory ‘is used to take real-world initial and boundary conditions into account, it directs attention to the most relevant factors’ (Kaplan, 2014: 188). The identification of practically relevant conditions is the task of the applier of the theory. Initial conditions (like features of the political system where the study takes place) and boundary conditions (like domestic and international political situations and historical events) affect what is existentially valuable, the plausibility of linking threats and referent objects, and whether security issues can be dismantled. ⁴

What context means has not been explored in depth by the illocutionary strand of the literature. Lise Philipsen (2020: 141, 151) comes closest to delving into this issue, but even she focuses on how Derrida and Judith Butler can be used to point to ‘iterability’ as a defining aspect of securitization (only acts that iterate security count as securitization), and how securitization alters the concept or logic of security. When following Derrida, a context is never ‘absolutely determinable’ or ‘saturated’, whereby it cannot be a criterion for success in speech acts either (Philipsen, 2020: 144–145).

As even such post-structuralist approaches have not discussed how the transformation of context takes place, it is prudent to explicate what context means when the philosophical approach to securitization theory takes it seriously: how can context enter into the theorization and explanation of illocutionary status transformations without becoming the mapping of social situations and their causalities? In order to flesh out and unpack the implications of context for the illocutionary version of securitization theory, it is necessary to read and discuss views on context within speech act theory proper. Accordingly, I turn to investigate the implications of context for securitization as it is taken within speech act theory. Here, of importance are the two main interpretative veins in regard to Austin’s initial work, that is, the speech act philosophy of John R. Searle and Marina Sbisà.

Social transformation of contexts in speech act theory

Variations of speech act theory are not without their own debates and disagreements. They can be so vehement, that later developments in someone’s theorization are not cited, or other’s work is not cited at all. Such silences require third party interventions to show where the disagreements are real, and where they are merely the result of misreading. This is also the case for the speech act theories of Searle and Sbisà that I deploy here to show how the transformative power of speech acts and the role of context is conceptualized by both.

First, for Sbisà (2002: 421), speech acts are context-changing social actions: an act is ‘something that changes a state of affairs into another, initiating a new state of affairs (Sbisà and Fabbri, 1980: 306), ‘the responsibility for which is assigned to an agent’ (Sbisà, 1984: 94). Agreement and ascription of responsibility are taken as vital because the participants ‘implicitly agree on what the speaker counts as having done’ (Sbisà, 1999: 10). Illocutionary acts depend on ‘socially accepted conventional procedures for achieving certain conventional effects’ in ‘appropriate circumstances’ (Sbisà, 1999: 10).

Second, for Searle (2011 [2010]: 13), speech acts can take the form of ‘status function declarations’ (SFD), that is, ‘all of human institutional reality is created and maintained in existence’ by SFDs. Accordingly, speech acts have the power to create or alter things: they ‘can represent states of affairs that do not exist but which can be brought into existence by getting a community to accept a certain class of speech acts’ (Searle, 2011 [2010]: 85).

Indeed, as Derrida (1982) has noted, the performative ‘produces or transforms a situation, it operates’ (p. 321). Accordingly, for both Sbisá and Searle speech acts alter the ‘rights, obligations, licenses, commitments’ (Sbisà, 2007: 470) of the speaker and addressee, or bring about deontic powers such as ‘rights, duties, obligations, requirements, permissions, authorizations, entitlements, and so on’ (Searle, 2011 [2010]: 9). Searle (2011 [2010]) suggests that people have ‘the capacity to create states of affairs with a new deontology; you have the capacity to create rights, duties, and obligations by performing and getting others to accept certain sorts of speech acts’ (p. 85). Some of such capacities are more formal than others:

There are (uncodified) rights and obligations of friendship and dinner parties, just as there are (codified) rights and obligations of citizenship and employment. There are deontologies without institutional facts [. . .] but there are no institutional facts without some form of deontology. (Searle, 2011 [2010]: 91)

For Sbisà (2001: 1798; see also Sbisà, 1984: 97), deontic modalities bring about or cancel obligations, which ‘can be legal or moral, and more or less binding’. Here, the social fine-tuning of human relationships can take the form of both the assignment and elimination (elsewhere defeat) of deontological modalities, that is, ‘the creation of commitments, the assignment and elimination of rights and entitlements or obligations, the legitimation of expectations, and the like’ (Sbisà, 1999: 11). Power and deontic entitlement are at the core of illocutionary effects (Sbisà, 2001: 1801); such forces have shades that range from ‘authority or mere influence, to legitimation, capacity, or competence’. For Searle (2011 [2010]) too, legitimation becomes a key element when it concerns deontic powers that relate to politics: ‘When the institutions involve power relations that can be threatening, as in the case of governments and political power generally, the question of legitimation becomes crucial’ (p. 108). Indeed, legitimation was the initial focus of securitization studies too (Wæver, 1989, 1995), also termed justification (Vuori, 2003), and has been a recent point of discussion as well (Balzacq, 2019; Olesker, 2018).

As both Searle and Sbisà emphasize, speech acts have the capacity to transform social situations. While speech acts can bring about deontic powers, their abuse can also defeat and cancel them (cf. Sbisà, 2007: 470). Speech acts can alter the context of subsequent acts: the content of assertions is added to forthcoming presuppositions and the cognitive context of interlocutors, which reduces the set of possible worlds participants may want to distinguish (Stalnaker, 1978). The perception of sincerity can be crucial here. For Searle (1969; Searle and Vanderveken, 1985), sincerity conditions are among the elements of the successful and non-defective use of speech acts. Sbisà (2002) suggests that sincerity is assumed in real speech situations, unless there is some reason to doubt it. Speech acts may then change the context in this sense: previous insincerity may be taken as the assumed norm in future speech act situations too.

To emphasize the transformative power of securitization and the accountability of acting towards desired transformations (Soares, 2022) makes the notions of intentions, ‘success’, and consequences relevant too. Some have pointed to the ‘unintended consequences’ of securitization (e.g. Bello, 2017; Karyotis, 2011; Rythoven, 2020). Here, securitization is considered an ‘intersubjective process, intentional or unintentional’ (Vigneau, 2019: 191). Speaking security is not a neutral act, and thereby bears responsibility: it is expedient in political mobilization, yet can bear perverse or tragic consequences (Rythoven, 2020: 479). This makes it necessary to also explicate the role of intentions in speech act theory, particularly as they have been a point of confusion within its debates beyond securitization theory.

Indeed, we should not overplay the intentions and motives of securitizing or other actors. This is because a distinction between intentions and acts, vital for Austin’s (1975 [1962]) presentation of speech acts, is only possible if the context of enunciation is not reduced to intentional presence (Moati, 2014: 43). Our ‘words’, not our intentions are what accomplish our acts; the conventions and the grammar of speech acts are what commit enunciators to their enunciations, not their intentions or their sincerity. This issue of the relation between intentions and context is what needs to be examined next in order to get to the role of context in illocutionary speech acts.

Context and intentions: what do you mean?

When discussing context, it is good to examine its meaning, in context (Burke, 2002). The term itself comes from the Latin word for weaving (contexere), while the noun contextus referred to connections. The issue of hanging a text, an act, or an event together (ger. Zusammenhang) with other relevant features in relation to them has also been depicted with other notions, such as (lat.) circumstantiae, or the things that stand around (Burke, 2002: 153–154). Even though ‘situation’ was becoming the most fashionable term for the problematic in the 1920–1930s, it seems that Austin favoured the early modern notion of ‘circumstance’ to deal with the issue at hand. Indeed, he speaks of the circumstances, situations, and contexts of speech acts, but only ‘circumstance’ has a proper technical meaning, and is the term used in the definition of felicity conditions for example (Austin, 1975 [1962]). Perhaps following general trends in social sciences (Burke, 2002: 158), subsequent speech act theorization has used the notion of context (e.g. Sbisà, 2002; Searle, 1969), as well as background and network (e.g. Searle, 2011 [2010]) rather than circumstance to conceptualize the various elements that bear on the accomplishment of speech acts.

In Austin’s (1975 [1962]) speech act theory, performative force presides over linguistic meaning and allows for action in the world through the intermediary of words (Moati, 2014: 18). When they abide by conditions of success, or felicity, the force of illocutionary acts can have effects on the world and people in it. Yet, the ensemble of circumstances in regard to an enunciation is greater than its intentions, which means that intentions do not have complete mastery over the circumstances (Moati, 2014: 42–43). Indeed, as Searle (2011 [2010]: 41) has noted, intentions have a role in the initiation of speech acts, but they do not have a transcendental role: the mere existence of ‘intention-in-action’ is not enough to carry an action through; there is a gap between reasons and decision, a gap between decision and onset of action, and a third gap between onset of action and its continuation to completion.

In securitization theory, success is often considered in terms of the effectiveness of achieving security and of its concomitant policies, yet transformations of social processes are more relevant for some even here. This is why it is necessary to separate ‘success’ as the ‘happiness’ or ‘satisfaction’ of securitization speech acts (i.e. accomplishing non-defective securitization moves) and the concomitant status transformation of an issue (if the moves succeed in producing an effect) from the ‘success’ of the politics of securitizing something (the consequences of the status transformation): happy securitization and the establishment of a security issue may yet lead to very unfortunate political outcomes. At the same time, securitization speech acts can always fail: securitization is ‘equally constituted by its possible success and its possible failure – one is not primary and the other derived’ (Wæver, 1989: 45). Failure, or infelicity, is an inherent possibility in any speech act situation (Austin, 1975 [1962]; Derrida, 1988; Moati, 2014).

In practical terms, successes and failures of such acts are on a continuum (as in speech act theory): it is highly unlikely that entire audiences will ever be fully and uniformly convinced by any political speech acts, including securitization; a theoretical dilemma for speech act theory (Sbisà, 2009: 50), but a commonplace for politics. This has led to the proposition of gradients of failure and success in securitization processes: securitization can be an overwhelming failure, precarious success, conflicted success, durable success, and overwhelming success (Karyotis et al., 2025: 12). These are further divided into forms of political success and programmatic success (Karyotis et al., 2025: 18). Others consider internal conditions necessary for successful use of securitization, while the external conditions depend on the context (Philipsen, 2020: 142), which would mean that ‘a definitive assessment of the success of any security speech act’ cannot be made (Philipsen, 2020: 150).

The introduction of ‘satisfaction’ into the conceptual apparatus of securitization theory could reduce confusion here. To follow the conventions of securitization is to satisfy the status function declaration of security: speaker meaning ‘is the imposition of conditions of satisfaction on conditions of satisfaction’ (Searle, 2011 [2010]: 74). Success, however, suggests that the securitization works in regard to why it was brought to bear on the context (criterions of satisfaction and success vary with the political functions of securitization), that is, does the satisfied securitization serve its purpose. Felicity conditions are necessary and sufficient conditions for satisfying what counts as securitization (x counts as y in context c; Searle, 1969); success is more of a consequentialist view of what takes place after securitization has been satisfied. This also enables the distinction of the satisfaction of various securitizing acts and the performative force that enables such acts to transgress their contexts (cf. Philipsen, 2020: 148). It also keeps the responsibility for securitization in the hands of the securitizing actor while remaining consistently within speech act theory.

The roles of intentions and presence have been major points of contention within debates between different interpretations of speech act theory too. As often is the case, these debates have muddled the original issues at stake. For example, the debate between Derrida and Searle (Derrida, 1988) misconstrues Austin’s (1975 [1962]) examination of enunciation as being about communication (Moati, 2014: 13–14). Here, Derrida approaches intentionality and communication through the tradition of phenomenology whereas Searle approaches them through Gricean pragmatism. This initial distortion leads the whole debate some way off from what Austin was actually presenting and interested in. Raoul Moati (2014: 17) emphasizes that Austin sought to describe enunciation as not reduced to the transmission of meanings. What constitutes the pragmatic value of enunciations is their force, not a detached semantic meaning consisting of a sense and a reference (Austin, 1975 [1962]: 100). Vital here are the circumstances of an enunciation instead (Austin, 1975 [1962]: 15). Meaning is not so much conditioned by intentions, but by the adequacy of words in light of conventions and the circumstance of their enunciation (Moati, 2014: 42–43).

Using the notions of ‘situation’ or the ‘circumstances’ of securitization would put too much emphasis on the ‘indexical presence’ of securitizing actors and audiences. The assumed requirement of such presence was one of Derrida’s (1988) lines of attack on speech act theory. While Derrida missed with his polemic (Moati, 2014), the theory of securitization should not now fall prey to what he was criticizing. Indeed, too much emphasis on situations would repeat the oft presented criticism of securitization theory that it focuses on single actors and acts (e.g. Salter, 2008), which does not really hold. Securitization is not about single actors, but about the in-betweenness of politics (Greenwood and Wæver, 2013). Furthermore, deontological transformations in regard to securitization are rarely accomplished with single acts; securitizations that consist of singular speech act situations are quite specific and need to be strongly codified (Guzzini, 2011: 335), such as in the case of declaring war.

The way Searle approaches speech acts avoids the above collapse. Context has been a vital aspect of Searle’s speech act philosophy from its beginning. For him, social facts come about in the form ‘X counts as Y in context C’ (Searle, 1969). ⁵ The notion of the context of an utterance is necessary because the same utterance can be used to perform various illocutionary acts in different circumstances; context is ‘one of the determinants of the illocutionary act’ (Searle and Vanderveken, 1985: 27). The context of an utterance consists of five elements and sets of elements, namely, (1) speaker, (2) hearer, (3) time, (4) place and (5) those various other features of the speaker, hearer, time and place that are relevant to the performance of the speech acts (Searle and Vanderveken, 1985: 27–28). Particularly important are the psychological states – for example, intentions, desires, beliefs – of the speaker and hearer. These other features form the ‘world of the utterance’. Here, the notion of physical possibility is also relevant, as the abilities of the speaker and the hearer often enter in the preparatory conditions of the illocutionary act (Searle and Vanderveken, 1985: 30).

Searle uses two other concepts that relate to the conceptualization of context: background and network (Searle, 2011 [2010]: 32–32; see also Searle, 1969). The network consists of the various beliefs, intentional states, desires, and so on that enable us to make sense of our sentences, speech acts, and other actions. Yet this network of beliefs, hopes, fears, and so on is not enough to account for the functioning of sentences and speech acts. What is needed is what Searle calls the background of taken for granted things, practices, and know-hows. The ‘background’ is not purely ‘interpretative’ (counter to what Sbisà, 2002: 423 suggests): while intentions and intentional states are vital for Searle, they are not enough to explain behaviour which can be done with the inclusion of the background of abilities. The network then is the set of intentional states (many of which are unconscious) needed to have some intentional state, while the background is the set of abilities, capacities, and so on that are needed to apply the intentional states. ⁶

For Searle (1969), it is enough to understand a sentence in order to understand its force, unless the illocutionary act is implicit or equivocal. ⁷ The conventional explication of intentions reduces the importance of context for enunciations, reduces the importance of indexical presence (that Derrida mistakenly took as the focus of his attack on Searle); when phrases are not indexical, the context can provide for proper understanding (Moati, 2014: 91), it can ‘disambiguate the illocutionary force of an utterance’ (Searle and Vanderveken, 1985: 28). Intention is not behind speech acts, it is at the same level as the conventions of speech acts; intentions are assumed to conform to the conventions of speech acts (see also Sbisà, 2002: 424–426). In this way, context and indexicality are part of the background of utterances. There is no need to refer to a context once the conventional intentions of speech acts are made explicit. In other words, Searle does not propose a psychological model of intentions in regard to speech acts, but intertwines intentions in the conventional rules of discourse (Moati, 2014: 95). Context also relates to the imbuement of deontic powers: ‘We make it the case by Declaration that X has the status Y and thus is able to perform the function F in C’ (Searle, 2011 [2010]: 99). This brings the issue of status functions in. The purpose of such functions is to ‘create and regulate power relationships between people’ (Searle, 2011 [2010]: 106).

Sbisà (2002) has also discussed the issue of context for the performance of speech acts (see also Sbisà and Fabbri, 1980). She criticizes the ‘received’ reading of speech act theory that has been influenced by Searle and Greimas. She (Sbisà, 2002: 430) presents three dualities in regard to the context: given or constructed, unlimited or limited, and objective or subjective and notably cognitive, where her own position is that the context of speech acts should be taken as constructed, limited, and objective.

First, for Sbisà (2002: 424–426), the context is constructed, in contrast to her portrayal of the received view of given contexts. Felicity conditions are not checked before accepting an act, but the default is to assume they are in order (and subsequently that they remain not in order in the case of a proven liar). She draws on sociolinguistic analysis of verbal interaction to present her case that the context of interactional events is set up by the participants as the interactive process goes on. Second, contexts in regard to speech acts relate to whether they are limited or unlimited (Sbisà, 2002: 426–427). While Sbisà grants that unlimited contexts may yield means for the interpretation of speech acts, for her, we either have to give up assigning context any evaluative role (as Derrida, 1988 does; see also Philipsen, 2020), or we have to assume the context as limited. The goals of conversation determine against what a speech act is to be evaluated; ‘the situatedness of the speech act goes hand in hand with delimitation of its context’ (Sbisà, 2002: 426).

Finally, Sbisà (2002: 427–430) divides contextual aspects of speech acts between objective and cognitive contexts. For her, a context has an objective nature, if it is deemed as determined by relevant states of affairs that occur in the world and of which participants are not necessarily aware. In contrast, ‘cognitive contexts’ consist of participants’ ‘intentional states’ that involve representations (Sbisà, 2002: 428). For her, this would mean that certain intentional states suffice to bring something about (Sbisà, 2002). To avoid such a view, Sbisá (2002) argues that context needs to be taken into account as either material or social external circumstances, which means that ‘speech acts as social actions needs to rely on an objective conception of context’. This is in effect what Searle’s background does.

Still in Sbisà and Fabbri (1980: 305, 316), Sbisà argued against an objective context in favour of symbolic interactionism: it is up to the hearer to select an acceptable interpretation of the speech act on the basis of force-indicating devices and the context. ‘Context, therefore, will no longer appear as an objective final criterion for the interpretation of speech acts and strips of activity’ (Sbisà and Fabbri, 1980: 316).

Indeed, the demand for an objective context would tilt the theory of securitization to the Balzacq variant, even though Balzacq (2015) presents himself as following Searle’s reading of speech act theory, and Wæver (2015) as following Sbisà.

A key issue here is that Sbisà seems to misconstrue the issue of cognition and objectivity when it comes to the issue of contexts. Indeed, as Searle (2011: 18) maintains, institutional facts are epistemologically objective (i.e. intersubjective) yet ontologically subjective. Concomitantly, ‘the speech act is more than just the expression of an intention or the expression of a belief. It is above all a public performance’ (Searle, 2011 [2010]: 83; italics in original). Furthermore, Searle’s (2011 [2010]: 85) context is not just cognitive, it is about states of affairs that come about by getting others to accept speech acts. For Sbisà too, conventional effects of speech acts come ‘into being by being agreed upon by the relevant members of a social group’ (Sbisà, 2009: 49), and can change the objective context in terms of entitlements and the like (Sbisà, 2002: 433). This means that ‘the very utterance of a sentence affects the context’, which makes speech acts ‘full-fledged actions yielding results’ (Sbisà, 2002: 433).

This brings us to the crux of the debates in securitization theory: the distinction between illocutionary and perlocutionary speech acts.

The illocutionary and perlocutionary aspects of securitization speech acts

For Austin (1975 [1962]), illocutionary force was not an emanation of a will or a natural effect of language on the will of others; conformity to conventions is what provides utterances with performative impact (Moati, 2014: 60). In other words, illocutionary acts have proper conventional efficacy, whereas perlocutionary acts have natural efficacy (Moati, 2014: 19; see also Sbisà, 2001, 2007). The two have been confused many times, as in the debate between Derrida and Searle (Derrida, 1988), where Derrida fused these two aspects into a single natural force and completely abandoned the conventionality of illocutionary force. Yet, for Austin (1975 [1962]), they were irreducible: the illocutionary force of speech acts is intrinsic to the conventional value of such acts; conventions are what allow us to identify illocutionary values and thereby the force of speech acts.

Some illocutionary acts have effects that are defined by convention (Moati, 2014: 61; Sbisà, 2007); enunciating such an act commits the speaker to precisely those conventional effects and transforms the relational situation of the speaker and the addressee. The perlocutionary act that may happen at the same time as the illocutionary act operates on a distinct ontological plane from the conventional one: the natural force of perlocutionary acts have nothing conventional about them, they are about the extrinsic results of the act (Moati, 2014: 19, 61), or the invitation of a response (Sbisà, 2001: 1796). The perlocutionary act is ascribed by the interlocutor, and the natural effects of the act also happen there (Sbisá, 2001: 1797, 1801). While we can deduce illocutionary effects in certain circumstances from their conventions, perlocutionary effects are more unpredictable as they are more about the feelings they engender (Moati, 2014: 19). But this does not mean that perlocutionary effects could not be intended: the perlocutionary intention of an insult is to produce insult, a threat is to deter, and so on. Such results are beyond the will and intentions of the enunciator though.

Comparing deterrence and security may clarify the distinction here. Deterrence is a perlocutionary act: no convention can bring about the effect of determent; it happens as a natural effect and is ascribed by the addressee (Vuori, 2016). Securitization is not a purely perlocutionary act in the same way as deterrence is, even though deterrence can be a perlocutionary effect of securitization (Lupovici, 2019, 2024). Deterrence cannot be brought about through convention, yet some forms of securitization can. This means that securitization is an illocutionary act like deterrentification. Securitization reconfigures the (necessary) relationship of the speaker and audience (Wæver, 2015: 122–123).

In Searle’s (2011 [2010]: 82–83) terms, speech acts can take the form of status function declarations with certain commitments and deontic powers, which come about due to their acceptance: ‘there is no way you can make explicit speech acts performed according to the conventions of a language without creating commitments. [. . .] All types of speech acts contain an element of commitment’. This commitment takes place irrespective of whether the utterance act is felicitously performed or not. Thereby, already the illocutionary act may change or alter the social situation, as it brings the social institution/rationality/modality of security to bear on it. The illocutionary act on its own already evokes the possibility of the status transformation, even if it ‘fails’. What the transformation that happens in the illocutionary act is depends greatly on the sociopolitical context and the possible conventions present in the circumstance. But it is important to note that for the various perlocutionary intentions to succeed, the constitutive change that takes place in the illocutionary act of securitization also has to be successful. Securitizing actors can choose to perform securitizing moves, but they cannot decide on the effects these will elicit, that is, securitizing actors cannot decide on the success or failure of their illocutionary acts, nor can they ascribe responsibility for their perlocutionary acts. Thereby, securitization is an open social process which can always fail, and which can be refused.

To reiterate, the illocutionary act and the commitment to the convention the act entails are within the remit of the locutor, but the perlocutionary act is deemed by the interlocutor (the perlocutionary act is not a gesture, but responsibility ascribed by the addressee; Sbisà, 2007), as is the perlocutionary effect. Sbisà (2002: 423) suggests that Austin’s infelicities such as abuses or breaches do not impede the conventional effects of the purported act. The inappropriate circumstances of the utterance are what make the act misfire. Thereby, the illocutionary act of securitization brings the status function to bear on the context, yet whether this brings benefits for the actor is not up to the locutor. In addition, the ‘end game’ of the securitization may not be about the particular audience or even the deontic powers that it can imbue: the securitization may only be an intermediate action.

To use the convention of securitization in a situation that does not match the assumption can have consequences for the securitizing actor in later situations of using the convention. Indeed, for Sbisà (2001) the ‘assignment of deontic modal values is not a unilateral matter: a speaker who modifies the status of his or her addressee cannot remain unaffected by his or her own action (p. 1805)’. This also suggests that the audience is not necessarily inactive or without consequence, even for the illocutionary aspect of securitization (contra Cote, 2016: 554; see Philipsen, 2020: 149). Both securitizing actors and audiences matter. Here, it is necessary to distinguish between enabling audiences that provide moral support in response to securitization speech (and its ‘natural’ perlocutionary effects) and audiences that have the capacity to provide deontological transformations (and conventional effects for securitization speech) (cf., Balzacq et al., 2016). This is essentially what Ori Wertman and Christian Kaunert (2022) refer to with the notions of political and legal audiences of securitization.

What is the solution?

Securitization describes the process for creating social facts of security. Indeed, in philosophical terms, ‘security’ can be viewed as a ‘Status Function Declaration’, whereby the social construction of ‘security’ brings about certain deontic powers. Transcribed in more general Searlean (2011 [2010]: 99) terms, an object X can have the status function Y (of security) and thus is able to perform function F in context C. In other words, speech acts of securitization have the capacity to create states of affairs with a new deontology of rights, duties, obligations, requirements, and authorizations that come about by performing and getting others to accept certain sorts of speech acts. Such status transformations, or SFDs change the context, they bring conventions of commitments to bear on the social situation that would not have otherwise been there. Securitization then is an act, although responsibility for it is ascribed by the interlocutor (à la Sbisà), and the deontic powers work only when recognized by them (à la Searle). Even illocutionary acts of securitization are intersubjective, as they depend on the intersubjective conventions of such acts, the circumstances of the enunciation, and the recognition of the act (contra Balzacq). Both failed and successful acts can have implications for subsequent acts and their contexts (Wæver, 1989). To the disappointment of those who seek easy indicators, this is why change in behaviour cannot be a criterion for the success or failure of securitization (Vuori, 2011: 136–140).

How the sociological approaches have viewed context in securitization risks the merger of the conventional and natural efficacies of speech acts. It is necessary to keep the illocutionary effects of securitization separate from its perlocutionary effects. Securitization moves may be about swinging audiences and producing emotions in them, but securitization may also be about transforming the rights and duties that relate to an issue and the relationship between the speaker and audience. Studying these separately leads to different research interests and requires different methods: illocutionary forces can be discerned from the acts alone whereas perlocutionary effects cannot.

When the illocutionary variant of securitization theory takes context seriously, there are a number of elements that need to be considered. First, the context is part of what makes a certain speech act function as a particular illocutionary act; context is one of the determinants of speech acts. Second, the speech act affects, or even constructs the context it is part of (or activates relevant parts of the network and background), and can have a bearing on the contexts of subsequent acts. At the same time, such features make the contexts of securitization limited (contra Philipsen), and make speech acts full-fledged actions with results. Third, intentions are not behind speech acts, but on the same level as the conventions of speech acts; intentions are assumed to abide by conventions unless there is something to suggest otherwise. Intentions do not have complete mastery over the circumstances of speech acts. Finally, speakers are affected by attempts at modifying the status between speakers and addressees; the assignment of deontic modal values produces commitments and is not a unilateral matter. Together, these elements make context a necessary part of illocutionary acts of securitization, yet the interest of studying securitization through them is in the power of securitization – irrespective of its success or failure – to affect its current and subsequent contexts.

The above philosophical stand raises the question of what the primary functions of ‘security’ are in real political terms. Accordingly, empirical research should investigate particular x:s, y:s, and c:s in order to test assumed effects of securitization processes and to discover the particular f:s of security. This is a necessary basis for comparative studies of securitization, as well as one for evaluating pre-existing studies in order to collate what we know about securitization. This kind of evaluation would go beyond quantitative corpus-based reviews of the literature (Baele and Jalea, 2023).

Part of such investigations should be how securitization speech acts change the contexts they become part of or produce, that is, what do the deontic powers of F do to the context. For Searle (2011 [2010]), ‘We make it the case by Declaration that X has the status Y and thus is able to perform the function F in C’ (p. 99). The philosophical problem for this formulation that Sbisà’s reading of Austin creates is that F can alter C, or that in subsequent speech acts the F becomes C. Furthermore, the alteration of C can defeat F (wedding, marriage, and nullification vs divorce). This leads us to the issue of the dismantlement of the social fact of security, or desecuritization, but that requires another theory clarifying article.