Securitisation and causality

Much of the Copenhagen School’s rejection of causality in securitisation theory reproduces the ‘explaining/understanding’ or ‘causative/constitutive’ dichotomy introduced in the IR field by Martin Hollis and Steve Smith,Footnote ¹³ who fixed, on one side, the notion that causal analysis has to do exclusively with explanatory theories, and on the other side, the notion that theories concerned with understanding, rather than explaining, are by definition constitutive, and hence non-causal. While empiricist or positivist theorists have assumed the first position as if it were scientifically superior, the so-called reflectivist, post-positivist, or constitutive theorists have assumed the second position, arguing that causal analysis is not desirable and much less necessary for understanding international politics due to its social constructivist characteristic.Footnote ¹⁴ Wæver’s argument on the non-causality of the linguistic core of the Copenhagen School’s securitisation theory,Footnote ¹⁵ due to its constitutive characteristics, clearly reproduces the second pole of Hollis’s and Smith’s dichotomy.

For theorists of the Copenhagen School, therefore, any attempt to find causal relations explaining why securitisation happens makes no sense because of its constitutive characteristic.Footnote ¹⁶ If ‘why questions’ and their definition in terms of causality are meaningless within the constitutive moment of securitisation, authors engaged with sociological approaches to securitisation, like Thierry Balzacq, reject the poststructuralist discursive core of securitisation theory and advocate a more substantive empirical analysis of the social construction of security, assuming in a less controversial way a causative force besides the constitutive dimension of discursive action.Footnote ¹⁷ Thus, even if the question ‘What causes securitization?’ cannot be answered through a ‘precise causal link as the exclusive source of a securitized issue’, this does not mean that causality makes no sense in securitisation.Footnote ¹⁸ As Balzacq suggests, if one puts the question in terms of causal adequacy instead of causal determinacy, causality becomes plausible in securitisation theory, allowing inquiry into the degree of congruence among convergent causal forces towards the outcome of the process.Footnote ¹⁹

From this sociological perspective, where securitisation is understood as a strategic, pragmatic process primarily involved in practices, context, and power relations, discourse and action are linked by both constitutive and causative forces, which means, according to Balzacq, that securitisations ‘can be discursive and non-discursive; intentional and non-intentional; performative but not “an act in itself”’.Footnote ²⁰ This shift from a ‘poststructuralist’ to a ‘sociological’ version of securitisation has, according to the author, important methodological implications: instead of the exclusive focus on textual analysis advocated by the Copenhagen School, securitisation should go further to include ethnographic research, content analysis, and process tracing.Footnote ²¹ Despite the conventional association between process tracing and positivist causal explanations, the author suggests that process tracing can be useful to examine causal mechanisms – that is, according to his own terms, ‘the forces within the network of causality’ – and to answer the question ‘under what conditions does securitization occur?’.Footnote ²² In summary, while the Copenhagen School focuses exclusively on the constitutive dimension of securitisation theory, rejecting any attempt to find a causal relation explaining ‘what causes securitization’, Balzacq’s pragmatic or sociological approach advocates a more comprehensive theory of securitisation, where constitutive and causative forces can be investigated to answer that same question.

Instead of rejecting the poststructuralist discursive core of securitisation theory, as Balzacq does by proposing a pragmatic turn in securitisation studies, Stefano Guzzini relies on causal mechanisms to expose securitisation’s explanatory status in a way that is compatible with the poststructuralist meta-theory of the Copenhagen School’s securitisation theory. Guzzini looks at securitisation theory not only as a conceptual move and a political theory of security, but also as a framework of analysis (where securitisation defines the formal logic of the construction of security problems and specifies its conditions of possibility) and an empirical theory of security (where securitisation is approached in more substantive terms).Footnote ²³ While the formal framework for analysis is the perspective that most influences the research agenda on securitisation, it is the conception of securitisation as an empirical theory of security that attracts Guzzini’s attention for the purpose of his reflections on causality. According to him, this latter perspective allows us to explore the explanatory status, which although not openly assumed by the Copenhagen School, is implicitly present in securitisation theory.Footnote ²⁴

Unlike the Copenhagen School, Guzzini understands the performative, linguistic characteristic of securitisation as part of a broader process of social construction of security and not as a single event defining the whole process. When looking at the paradigmatic case in which the concept of securitisation and its opposite notion, desecuritisation, were originally framed by WæverFootnote ²⁵ – the end of Cold War in the context of German Ostpolitik – Guzzini notes that the USSR and the threat of communism were not desecuritised via ‘a single authoritative act’, but rather through ‘a policy of gradual change’ affecting the ‘logic of high politics’.Footnote ²⁶ From this perspective, securitisation and desecuritisation cannot be empirically conceived as the speech act in itself, understood as ‘a single bombshell event’, but refer to ongoing processes of policy change within the particular context in which the speech acts are realised.Footnote ²⁷

According to Guzzini, therefore, even if we recognise that ‘discourse of security does securitization’ by definition, it is also important to note that its effectiveness as a successful process depends on its internal coherence, its contextual and cultural legitimacy, and its resonation ‘with collective memories and the lessons of the past’.Footnote ²⁸ It is against the background of specific political processes, always contextual, temporal, and contingent, that discourses of security result in successful or unsuccessful securitisation processes. Thus, as an empirical theory, securitisation is not a generic ‘friend-foe’ linguistic articulation established by speech acts of self-interested elites, but a historical process, contextually situated, embedded in ‘self-other understandings that predispose political discourses, public opinion and hence the receptivity of the wider public to certain political moves’ that insert a particular issue in the logic of high politics.Footnote ²⁹ If we regard securitisation this way as an empirical theory, speech acts are components, not the entire process of social construction of security.

Based on Guzzini’s arguments, what is crucial for the discussion proposed in this article is that securitisation is a process that is not only causally linked to its external effects in the world, but also is linked endogenously to certain events or features of foreign policy and public opinion, as well as to specific collective political, temporal, cultural, and contextual notions of legitimacy that ‘cause’ a shift in political debate towards a security logic. For speech acts to be effective, they have to be integrated within a broad process of legitimation that creates the conditions for the possibility of successful securitisation. The social construction of security, from this perspective, is implicated in causal relations throughout the whole process, and not only outside its performative moment, as Wæver partially admits.Footnote ³⁰ In other words, causality in securitisation theory does not refer exclusively to its external effects in the world, but also refers to its internal, linguistic core, including the contextual conditions for its empirical realisation as a collectively legitimated discourse of security.

Given the discursive core of securitisation theory, in which a non-representational conception of language is implied, the kind of causality envisaged here cannot of course be captured by a rationalist or positivist version of causation. For this reason, Guzzini relies on causal mechanisms – understood as ‘recognizable causal patterns that are triggered under generally unknown conditions or with indeterminate consequences’Footnote ³¹ – as an appropriate alternative to solve the question of causation in a constructivist, post-positivist setting. For Guzzini, securitisation can be conceptualised as a causal mechanism whose ‘triggering capacity’ should be seen in two ways: first, securitisation is the process that is triggered by something else; second, securitisation is itself the trigger of certain effects, which means that securitisation is ‘both explanandum and explanans’.Footnote ³² Thus, it is perfectly plausible to reconcile both ways to causally study securitisation, seeking to analyse ‘what made securitization possible’ (the focus, here, is the causal mechanism that triggered securitisation) or ‘the effects of securitization in other processes’ (in this case, the focus is the effects triggered by securitisation in other social processes).Footnote ³³ It is also possible, Guzzini argues, to use the causal mechanism both ways, by encompassing the process as a whole, without having to define the ‘right’ use or ‘where exactly to temporarily locate a mechanism’ (at the triggering moment of securitisation, linking securitisation to its triggered effects, or within the process taken as a whole).Footnote ³⁴ In short, the entire securitisation process can be conceptualised as a causal mechanism and it is the researcher’s empirical question that drives the way the mechanism is understood.

Conceived outside a positivist correlational logic and defined as part of an interpretive process, the causal mechanism is therefore consistent with securitisation’s meta-theoretical assumptions, and could be used, according to Guzzini, ‘in a type of interpretivist process tracing to allow for more coherence in the use of securitization in empirical explanations, because ‘explanations they are’.Footnote ³⁵ With these concluding words, Guzzini presents the causal mechanism as a way to reconcile causal analysis with the poststructuralist linguistic core of securitisation theory. However, as the following section aims to show, poststructuralist discourse meta-theory has strong reluctance to accept causation, so that for discursive analysis to become compatible with causality, further discussions are needed.

Reconciling constitutive and causal analysis

This section complements and refines the discussion on causality presented in the previous section. Relying on critical realist innovations to rethink causality, I try to reconceptualise cause and integrate discursive practices and constitutive relations into causal accounts, going beyond the ‘causative/constitutive’ divide that prevents Copenhagen School theorists from accepting that securitisation’s discursive core part is also implicated in causal relations.

As a starting point, it is important to examine how poststructuralist discourse theorists articulate reality and discourse, and why they reject causality. A first point for poststructuralists is that despite the obvious reality and independence of objects, it is only through the socio-practical articulatory nature of discourse that ‘physical objects are brought into a system of meaningful relations’.Footnote ³⁶ From this perspective, all the social world is encompassed by discourse, and it is only through language, as the most elementary system of social interaction, that ‘objects become real to us’.Footnote ³⁷ Faced with the impossibility of a meaningful constitution of objects outside language, discourse becomes the real – hence Jacques Derrida’s famous phrase ‘il n’y a pas de hors-texte’ (‘there is no outside-text’),Footnote ³⁸ that is, there is no extra-discursive instance (empirical facts, objective criteria or real points of reference) capable of safeguarding the truth outside the endless meaning-making process of discursive practices.

It is because of these ontological assumptions, and their epistemological and methodological implications, that causal analysis seems taboo to IR poststructuralist theorists. As Lene Hansen stresses, it is impossible to ‘measure’ the causal impact of one discourse over another: ‘for discursive causality to be considered an actual causal effect, one needs to separate two variables and to observe each independently of the other’, which is impossible given the co-constitution characteristic of discursive practices and their objects.Footnote ³⁹ For David Campbell, contrary to empiricists’ search for material causes existing outside thought to which events and actions can be reduced, poststructuralists ‘embrace a logic of interpretation that acknowledges the improbability of cataloguing, calculating, and specifying the real causes’ explaining why events and actions happen.Footnote ⁴⁰ The Copenhagen School also rejects causality in securitisation theory’s discursive core part by arguing that, even with the concept of facilitating conditions (introducing some materiality in securitisation analysis), ‘the theory is not causal in a traditional sense’ because securitisation, conceptualised as a speech act, is never exhaustively explained by objective conditions and ‘the act can never be reduced to a transmission belt in causal chains’.Footnote ⁴¹

From a critical realist perspective, the arguments presented above derive from the fact that they all adopt a Humean account of cause as constant conjunctions and regularities. As Milja Kurki notes, the reluctance of poststructuralist theorists to deal with causality – because causation for them has to do exclusively with a deterministic and empiricist mode of explanation that does not conform to the discursive, constitutive characteristics of the social construction of the social field – has not contributed to open the debate on causal analysis beyond this narrow perspective. In fact, the post-positivist refusal to engage with the question of causality has reinforced, rather than weakened, the dominant Humean conception, as if it were the only possible route for explanatory analysis.Footnote ⁴²

As an alternative, critical realists propose rethinking the concept of cause beyond the dominant Humean conception. A key for this purpose is to recognise not only the limitations of positivism to ‘experience’ the deeper, not always observable, layers of reality, but also the poststructuralist tendency to limit the real to the discourse. For critical realists, the problem of poststructuralist discourse theory lies not in its constructivist social assumptions, which critical realism also shares, but in its view that the real is limited to discourse. As Colin Wight argues, the fact that we can only know things under discursive descriptions ‘does not negate the ontological status of that to which we refer’, which means that ‘things’ and ‘talk about things’ should not be confused.Footnote ⁴³ Indeed, as Wight notes, poststructuralist theorists, like empirical positivists, commit a philosophical mistake in the way both define the real.Footnote ⁴⁴ While empiricists define the real in terms of our experiences of the world, poststructuralist theorists define the real in terms of linguistic conventions. According to Wight, both perspectives ‘give an epistemological category an ontological task’.Footnote ⁴⁵ This is what Roy Bhaskar has termed the ‘epistemic fallacy’: the definition of ‘statements about being’ (ontology) in terms of ‘statements about our knowledge of being’ (epistemology), that is, the reduction of ontology to epistemology.Footnote ⁴⁶ By assuming the possibility of being experienced and the possibility of being conceptualised or expressed through language as the reality itself, both perspectives overlook the causal criterion for the definition of something as real.Footnote ⁴⁷

For critical realists, in contrast, there is more to know about reality that is not limited to what can be experienced or symbolically transmitted in discourse. Firstly, the Humean empiricist apparatus cannot access ‘transfactual’ aspects of reality, which Bhaskar calls ‘intransitive objects of knowledge’Footnote ⁴⁸ (that is, real objects, structures, or mechanisms that exist ‘beyond and behind the appearances’ and independently of our ‘ability to know or manipulate them’).Footnote ⁴⁹ Secondly, the poststructuralist perspective, by confusing the real with the limits of discourse, also fails to provide a deeper and broader account of reality. As Wight argues, ‘the totality of the social field is not exhausted or confined to the dimension of intersubjectively intended and symbolically transmitted meaning’, and ‘the linguistic structure of social life is but one element’,Footnote ⁵⁰ not the entire social field. Thus, even if language and discourse are our privileged means of accessing reality, the real is not confined in discourse, and language cannot be considered an obstacle to accessing extra-discursive things.Footnote ⁵¹

Based on what has been discussed up to this point, two aspects are crucial for causal analysis. First, if superficial phenomena and our experiences of them do not exhaust the real, the kind of explanation derived from the Humean causality as constant conjunctions and regularities is not appropriate.Footnote ⁵² Second, if the discursive condition of the social world does not prevent us from having access to non-discursive things, this should not prevent us from integrating ideas, rules, norms, meanings, concepts, discourses, and non-discursive things within a complex of causes, whose causal properties, even if not observable and measurable, have constraining and enabling powers that explain changes in the world. Important here is to note a shift from phenomenalism to ‘transfactualism’,Footnote ⁵³ that is, a move from superficial experiences to deeper ‘structures, powers, and tendencies’Footnote ⁵⁴ that give rise to what is empirically manifest. It is this shift that allows critical realists to go beyond empiricist constant conjunctions and regularities and start talking about causal powers.Footnote ⁵⁵

Claiming the need to go beyond closed systems and the surface of observable events in order to reach deeper layers of reality, where underlying multiple causes, not always observable, interact and counteract each other in complex and unpredictable ways,Footnote ⁵⁶ some critical realists have made contributions to rethink causation as ‘causal mechanisms’,Footnote ⁵⁷ that is, as certain aspects in the structure of a thing that enable this thing to have a generating power, an ability to produce some result. As Wight stresses, a causal mechanism should not be understood only in a mechanistic, physical sense. Rather, a causal mechanism can be an institution, a psychological condition or a discourse that integrates ‘the operative or motive part, process, or factor in a concrete system that produces a result’.Footnote ⁵⁸

It is important to stress here that the way critical realists conceive causal mechanisms does not exhaust all possible ways of articulating the mechanism metaphor from a causal perspective. When Alexander George and Andrew Bennett rely on causal mechanism to identify a sequence of potential causes in a process tracing procedure in positivist strategies of research,Footnote ⁵⁹ this obviously has nothing to do with what critical realists have in mind. Indeed, the only thing George and Bennett introduce into the analysis is a greater number of causal links leading to the covering-law,Footnote ⁶⁰ not adding anything new to causal explanation, since what they actually do is reduce causal mechanisms to a sequence of intervening variables.Footnote ⁶¹

Critical realists, in turn, rely on the mechanism metaphor aiming at a radical change in the concept of causality. For them, instead of isolating or tracing a sequence of specific causal factors, trying to determine their impact on the outcome, as George and Bennett misleadingly suggest, critical realism relies on the concept of mechanism to challenge positivist causation, preferring to operate with a ‘causal story’ in case-specific narratives rather than ones that seek to identify any kind of individual or ultimate cause explaining a given event. Thus, for critical realists it is the holistic causal narrative and not any individual causal factor that is significant for the explanation.Footnote ⁶²

In spite of the effort of some authors to dissociate the expression causal mechanism from its mechanistic connotation and its positivist use, this has not led to a convergent position on the use of this metaphor even among critical realists. As Kurki notes, critical realists have been criticised for employing the mechanism metaphor in a way that is too vague and imprecise. Moreover, the mechanism metaphor is strongly associated with a ‘pushing and pulling’ force, that is, an efficient moving mechanism, which is too narrow to reflect the holistic causation that critical realists have in mind. Thus, instead of using this problematic term, the most appropriate way of referring to the causality advocated by critical realists, as Kurki suggests, is simply to call it a ‘complex of causes’ or a ‘causal complex’.Footnote ⁶³ Considering that not only Kurki, but also other IR theorists committed to critical realism, such as Heikki Patomaki and Colin Wight, do not see the mechanism lexicon as necessary for a non-positivist account of causality, terms like ‘causal power’, ‘causal complex’, or ‘causal process’ have been preferentially employed in their works.Footnote ⁶⁴

Besides clarifying the scepticism about the mechanism metaphor and the preference for the term ‘causal complex’, one should highlight in Kuki’s work the effort to deepen the notion of cause to reach the causal power of ‘unobservable’ entities, and to broaden the assumptions associated with the notion of cause beyond the notion of ‘efficient cause’.Footnote ⁶⁵ In this effort, Kurki revisits Aristotle’s multifaceted notion of causality, in which various types of causes – material, formal, efficient, and final – constitute a broad spectrum of causality, each of them ‘causing’ in different ways.Footnote ⁶⁶

Why is this Aristotelian typology of cause useful for our understanding of causality in a constitutive setting? Kurki answers this question using several arguments. First, this multifaceted conception of cause is useful since it allows us to overcome deterministic modern conceptions of causation by outlining different ways of ‘causing’. Second, Aristotle’s insights allow us to assess causes as complex and interacting. For the Greek philosopher, it is the complex interaction of all these four types of causal forces that explains why any change or thing has come about in the world. Thus, when inquiring into any change or thing, the search for determinant singular causes is insufficient in most cases, since causes never exist in isolation from each other. Third, because the social world is shaped not only by material factors, but also by ideas, rules, norms, and discourses (often interpreted as ‘constitutive’ forces), the notion of formal cause helps us to understand these constitutive aspects as causes. As thought by Aristotle, formal cause is that which makes or defines a given thing, its ‘structure’, its qualities and its properties. From this perspective, ideas, rules, norms, and discourses can be understood as formal causes. Fourth, Aristotle’s concept of final cause, often referring to ends and purposes, brings intentions, motivations, or reasons to causal analysis. Kurki stresses, however, that intentions, motivations or reasons should not be conceived to ‘push and pull’ in the same sense as efficient causes (defined by Aristotle in a more traditional sense, as any active source of change). Unlike efficient cause, final cause is causal in the sense that it works as ‘a contributory cause, that “for the sake of which” something is done’.Footnote ⁶⁷

It should be noted from this alternative angle that the reification of the ‘causative/constitutive’ divide in IR starts to sound misleading. By rethinking causality outside the Humean box, constitutive relations can be reinterpreted as causal, paving the way for discourse theorists in general, and securitisation theorists in particular, to embrace causal analysis in a less problematic way. After all, one cannot fail to notice that when poststructuralists stress the role of discourse in the constitution of social life, they do so because such discourses, by constituting the perceptions and reasoning of agents, ‘have “consequences” for how agents perceive the world, themselves, others and, hence, their actions’.Footnote ⁶⁸ The same can be said about securitisation: when the Copenhagen School defines securitisation as a speech act, discourse matters not only because it constitutes a shared sense of threat, but because it constrains normal politics and enables exceptional procedures. Thus, discourse matters not only in the sense in which it defines meanings, but also in the sense in which it constrains and enables courses of action – this shows that ‘constitutive discourses are, in fact, causal’.Footnote ⁶⁹

With this extended and deepened notion of causation in mind, the next section and the subsequent discussion intends to show that securitisation, illustrated by the case of Somali piracy, can be seen as a causal complex composed of discourses, ideas, concepts, meanings, norms, rules, reasons, motivations, and extra-discursive aspects, all endowed with constraining and enabling powers in the process. From this perspective, the securitisation of piracy is not only triggered by these causal powers, but is also the triggering of a shift from a conventional fixed government polity to a more fluid and dispersed governance framework in the maintenance of order in the ocean space.

The securitisation process of Somali piracy

The International Maritime Bureau (IMB) – the maritime division of the International Chamber of Commerce – has been a key actor in the construction of maritime crime as a security issue, not only by calling attention to the most troubling outbreaks of piracy in the world since the early 1990s, but also by demanding militarised involvement of states to address the problem. This ability to influence the security agenda results not only from the role the Bureau plays as spokesman of the maritime industry on piracy issues and the pressure it exerts on a select audience of states and governmental organisations (such as the International Maritime Organization (IMO),Footnote ⁷⁰ where the IMB participates as a non-governmental consultative member), but also from its expertise as the main provider of regular and systematised statistics and analysis of worldwide piracy since the late 1980s.

It is worth noting that the provision of statistics and expertise has been the main tool and the main justifying base upon which the IMB constructs its securitising narrative against criminality at sea. In this regard, one cannot fail to note the implications of the reformulation of the concept of piracy carried out by the IMB in the first half of the 1990s. The UN Convention on the Law of the Sea (UNCLOS) defines piracy, following customary law, as acts of private violence committed against ships on the high seas or in waters not subject to the jurisdiction of any state.Footnote ⁷¹ All other situations of private violence at sea, such as criminal actions committed against vessels within the territorial sea of a particular state or in its ports, are defined as ‘armed robbery against ships’ according to a resolution of the IMO.Footnote ⁷² These legal aspects have crucial implications for internal and external management of crime at sea. On one side, the crime of piracy can only occur on the high seas or in international waters, and can be policed by warships or civilian ships belonging to any state transiting those waters.Footnote ⁷³ On the other side, armed robbery against ships occurs within territorial waters and ports of a particular state and should be handled by port police and national coast guards.

According to international law, therefore, the distinction between piracy and armed robbery against ships is not substantive, but merely jurisdictional. Moving away from this legal characterisation, in the early 1990s the IMB resolved to merge crimes of piracy and armed robbery against ships into a single definition for the purpose of compiling statistics and producing reports on piracy.Footnote ⁷⁴ Two aspects are crucial in this redefinition: first, it completely disregards the jurisdiction of waters where crimes occur (which is a key aspect in defining internal and external dimensions of private violence at sea according to international law); and second, it removes the term ‘private’ as a defining feature of piracy (which is crucial to distinguish piracy from acts of violence with political purposes, such as maritime terrorism or attacks perpetrated by militias in conflict). Consequently, all the attacks committed at sea, regardless of national or international jurisdiction over waters where incidents take place and their private or political ends, are treated indiscriminately under the label of piracy by the IMB.

Since the IMB is a private organisation, this expanded meaning of piracy obviously has no legal effects. But this does not mean there are no implications in terms of security governance. Some analysts have noted that this reconceptualisation is part of the IMB’s strategy to increase the number of incidents of piracy in its statistical reports with the aim of shocking governments and impressing them with the need for more vigorous action in combating criminality at sea.Footnote ⁷⁵ The overall number of cases of piracy, if considered strictly within its legal definition (that is, only attacks in international waters), is significantly smaller than the number of attacks in territorial seas or ports, and becomes negligible when compared to statistics of crime on land. J. N. Mak performs this type of empirical analysis in the specific case of Southeast AsiaFootnote ⁷⁶ and the IMB itself recognises this fact.Footnote ⁷⁷ Thus, by aggregating in its statistical data all the attacks committed at sea under the same label, the IMB shows in quantitative terms a much more alarming global situation of piracy. From a qualitative point of view, this conceptual move brings to maritime crime as a whole the symbolic charge historically tied to the term piracy, enhanced by its current association with maritime terrorism, which suggests a kind of ‘nineteenth century solution’, that is, the use of military force against sea criminals regardless of the motivation (private or political) and the jurisdiction of waters where attacks occur (national or international).

The crucial point to note, in short, is that the IMB’s reconceptualisation of piracy, as well as discourses, reports, and statistical data produced in function of this reconceptualisation, have contributed to the construction of a more menacing picture of the worldwide situation of contemporary piracy and the justification of a more robust response of states, including their military engagement in addressing the problem. Regarding the specific case of Somali piracy, the IMB and other maritime industry representatives advocated the need for a more effective response of states and international governmental organisations, effectiveness that could only be achieved through the mobilisation of military forces and the use of force against Somali pirates. Within this strategy of persuasion, addressed not only to the shipping industry, but mainly to an audience of policymakers endowed with power to decide on security issues (leaders of states and international organisations, such as the IMO, UN, and UNSC and their experts and security bureaucracies), the discourse was strongly supported by statistical data on the increasing number of attacks in the Horn of Africa (on the high seas and within Somali territorial waters) and some extra-discursive aspects, such as the protracted civil conflict in Somalia, the absence of state control within the Somali territorial sea, the fact that Somali pirates operated at great distances from the coast using heavy weaponry, and especially after 2007, the risk of alliance between pirates and Somali Islamic militias to carry out acts of maritime terrorism.Footnote ⁷⁸

In the mid-2000s, a sequence of hijackings and attacks on ships chartered by the World Food Program (WFP) precipitated the involvement of the UN with the question of Somali piracy.Footnote ⁷⁹ In this context, executives of the WFP and the IMO jointly adopted a dramatic speech on the humanitarian situation in Somalia, portraying pirates as an existential threat to the survival of thousands of poor people affected by drought and civil war in the country.Footnote ⁸⁰ The humanitarian argument was highly effective in convincing a target-audience, including the UN Secretary-General, the UNSC members and some states and regional organisations (in particular France, the European Union, and NATO), whose leaders and security bureaucracies (defence and navy authorities) began to justify the use of naval forces to protect humanitarian aid destined for Somalia in WFP chartered ships.Footnote ⁸¹

This process culminated in 2008 with the unprecedented involvement of the UNSC, which resulted in four resolutions against Somali piracy approved under Chapter VII of the UN Charter.Footnote ⁸² With regard to the securitisation moves conducted in the UNSC, a more complex form of operation was observed. The UNSC could be seen from two angles. First, it could be viewed as a forum where member-states acted as securitising actors and audiences, negotiating decisions on Somali piracy. Second, the Council ‘embodied’ the decisions taken by its members, and acted as a singular securitising actor trying to convince states and regional organisations about UNSC arguments and decisions.Footnote ⁸³

Seen from the first angle, the interactions occurred following a logic of negotiation among UNSC members, including the formation of coalitions and alliances between securitising actors and audiences. Within this process, two critical aspects stood out: the question of breaching the sovereignty of Somalia and the construction of the concept of pirate to be included in the resolutions (civilian criminal or combatant?). The speeches of some delegates, especially those from Indonesia, China, and South Africa, were very cautious on the question of sovereignty, arguing that their countries only agreed with the resolutions because they were applied exceptionally to the case of Somalia and did not include any provision that would ‘touch the sovereignty of other countries’.Footnote ⁸⁴ Regarding the concept of pirate, although the UNSC involvement with the question of piracy in principle had the main purpose of creating a legal device that merely projected in Somalia’s territorial sea the same legal rules of policing applied on the high seas, the draft resolution put under discussion gave margin to an ambiguous interpretation, according to which military hostilities against pirates could be taken as if they were combatants in similar lines applied to the war on terrorism.Footnote ⁸⁵ Thus, while on one hand the negotiations led to a shared understanding that Somali pirates should be seen as criminals and not as combatants, on the other hand the text approved in Resolution 1851 went beyond this understanding, especially when authorising actions on land in Somalia.

This resolution, proposed and negotiated by the United States, clearly reflected the American strategic plan to curb piracy in the Horn of Africa approved in the last moments of the Bush administration in 2008. This plan, which described Somali piracy as a global threat and advocated immediate action to prevent, disrupt, and punish acts of Somali pirate organisations, also argued that to achieve these objectives, actions on land were required to enhance measures taken at sea and deprive pirates of profits coming from ransom payments.Footnote ⁸⁶ Here lies a crucial ambiguity: although negotiations on resolutions configured Somali pirates as criminals instead of combatants, the provisions of Resolution 1851, which mirrored the American strategy, gave room for the entry of aircraft in Somali airspace with the purpose of bombing pirate bases on land (which was actually put into practice later by European Union military forces). This provision introduced a level of lethality in the use of force that did not seem commensurate with the type of treatment normally required by questions of criminality.Footnote ⁸⁷

Seen from the second angle, the UNSC acted as a singular securitising actor after the resolutions against piracy were passed. In this second moment, the Council articulated Somali piracy as a threat to the various referent objects defined in the previous securitisation moves (humanitarian aid to Somalia, commercial sea routes, international shipping, crew and passenger life, international peace and security), justifying an exceptional response to the problem (use of force against pirates by international military forces within the territorial sea and on land in Somalia). This securitising move at the scale of the international system, which synthesised and organised all the previous securitisation moves, sought to convince an audience of states, regional organisations, defence alliances and naval coalitions of the need to formulate coordinated policies to combat piracy in the Horn of Africa. The broad acceptance of this move could be empirically noted from late 2008 and early 2009 through the mobilisation of several states (Japan, China, Russia, India, Malaysia, South Korea, Saudi Arabia, Iran, and Yemen) and three multinational maritime coalitions (one led by the United States, another led by the European Union, and a third led by NATO), which then expressed willingness to participate in the planning, operationalisation, and coordination of military action to fight Somali piracy.Footnote ⁸⁸

This brief analysis shows that piracy, over the course of two decades, was taken off its normal political agenda – where it was treated as criminal activity within the national legal sphere of states or viewed as a matter of commercial loss to be managed through usual relations between ship owners and insurance companies – and raised to the top of the international security agenda. This shift from normal to exceptional politics, which reached its peak with the involvement of the UNSC in 2008, did not result from a single or linear process. On the contrary, attempts to construct an exceptional sense for Somali piracy involved multiple securitising moves, led by a range of governmental and non-governmental actors, targeting different audiences – who often reversed their roles and began to act as securitising actors in subsequent moves – and referred to objects that ranged from the individual level (the life of sailors, the survival of people in Somalia) and reached the global level (international maritime trade, international humanitarian assistance, international peace and security). These multiple and mutually reinforcing securitising moves produced a layered securitisation process, whose acceptance by the various audiences made Somali piracy a shared existential threat at the highest level of the international system.Footnote ⁸⁹

If this empirical narrative plausibly shows that Somali piracy was constructed as an existential threat through intersubjective relations significantly implicated in discursive practices, and if because of this it is possible to conclude, following the Copenhagen School’s framework for analysis, that this case meets the formal requirements of a successful securitisation process, crucial questions remain open: What is it all for? What is the relevance of this formal exercise, if not to understand its generative effects? Poststructuralist discourse theorists would argue that these are misleading questions. After all, inquiring into constitutive questions, from their point of view, necessarily excludes causal questions. Copenhagen School theorists would agree with this by adding that all one can say about the securitisation of piracy is exhausted within its constitutive discursive process.

From this point of view, the analysis of the securitisation of Somali piracy could stop here and nothing more would be said about this process. However, if one accepts the arguments advanced in this article, constitutive analysis can never be considered ‘an end in itself’: constitutive relations leading to discourses, rules, and norms matter ‘precisely because they are causal, that is, because they produce, contribute to, direct or constrain and enable ontological social realities’.Footnote ⁹⁰ Thus, if we want to take a step further, it is necessary to integrate the analysis above into a broader empirical framework, trying to connect this securitisation process with its constraining and enabling effects in the social world.

Connecting securitisation and governance

Although a theoretical discussion about governance is beyond the limits and purposes of this article, some basic considerations on this concept are necessary to better understand the kind of causal connection this section aims to show between securitisation and security governance. Traditional governance studies often treat governance in terms of a pre-given set of agents and structures with fixed and defined responsibilities and activities that take place within formal institutions.Footnote ⁹¹ Since the last decades of the twentieth century, however, governance researchers have noted a growing transformation of this traditional framework, which they describe as a shift from ‘government’ to ‘governance’. According to Allan Dreyer Hansen and Eva Sørensen, while the term ‘government’ has been employed in the traditional sense in situations in which governance is exercised by formal and hierarchical instances involving bureaucratic and public administrative powers, the term ‘governance’, in turn, has been increasingly associated with a situation in which governance is exercised though horizontal decentralised networks, involving a plethora of public and private actors.Footnote ⁹² Put in a nutshell, the contemporary shift from ‘government’ to ‘governance’ has increasingly challenged the traditional concept of sovereignty and the unity of the nation-state by transforming them into a fragmented governance structure, where public and private agents coexist in the maintenance of the social order.

Regarding maritime crime, some analysts have stressed that the impressive international involvement with the problem of Somali piracy does not aim to merely managing this specific problem. Instead, it involves a broader concern with the global governance of the ocean space. James Kraska and Brian Wilson, for example, note that the United States has used the problem of Somali piracy to advocate the idea of a Global Maritime Partnership.Footnote ⁹³ Christian Bueger and Jan Stockbruegger observe that some informal discussions in the UN General Assembly have referred to Somali piracy as a case study that offers the opportunity to rethink the governance of maritime security in a more comprehensive manner.Footnote ⁹⁴ It is worth noting, however, that these authors do not go much further than describing the elements and the structure of this governance mechanism, failing to develop a causal narrative that connects in some way the securitisation of piracy and its constraining and enabling effects in terms of security governance.

The broad empirical narrative outlined here seeks to fill this gap, by showing that securitisation has causal powers that enable changes in the governance of the ocean space because traditional regulatory, spatial and institutional maritime arrangements – such as defining norms of piracy, jurisdictional demarcation of national and international waters, inviolability of the principle of sovereignty, divisions of labour between domestic policing and external defence, and boundaries between public and private – are no longer seen as capable of presenting effective solutions for collective problems with transnational characteristics that emerge in peripheral zones of the world, especially in the so called ‘weak’, ‘fragile’, ‘failed’, or ‘collapsed’ states (as Somalia is construed in international political discourse). If one integrates securitisation within this broad narrative, it is possible to note that the social construction of piracy as an international security problem creates the conditions of possibility for governance at a supranational level, not only by enlarging traditional regulatory arrangements of the ocean space, but also by enabling the mobilisation of an expanded security community that merges non-traditional and traditional external and internal security actors and bureaucracies (such as navies, intelligence agencies, foreign affairs officials, coast guards, port police, national juridical institutions and agents, shipping industry agencies, private security contractors) to overcome national constraints and facilitate achievement of the stabilising purposes of global governance. In short, the successful securitisation of maritime crime explains why and how certain changes in spatial and institutional arrangements in the governance of the ocean space have become possible.

Several aspects contribute to this explanation. First of all, it is noteworthy in the process analysed above that the securitisation of Somali piracy has exerted pressure on spatial, institutional, and regulatory arrangements that have traditionally separated the internal and external dimensions of security in the ocean space. The first relevant aspect in this respect has to do with the effort to construct an alternative concept of piracy. This effort is clearly visible in attempts to redefine piracy by non-state agencies such as the IMB or states such as the United States, which share a similar point of view that broadly frames piracy as any kind of attack or robbery of ships, irrespective of the jurisdictional waters where the crimes occur. By placing under the same label incidents on the high seas, within territorial waters, ports and, as defined in the American Strategic Plan to Curb Somali piracy and some UNSC resolutions, even the connections on land (pirate bases, ransom negotiators, logistic providers, funding networks, and money launderers), these securitising actors clearly construct an expanded meaning of piracy in which the external dimension is merged with a domestic dimension that is unprecedented in contemporary regulation of criminality at sea.

The second point to make is that this process has produced international pressures on the sovereignty of states over their territorial seas. While in other piracy prone regions, such as in the Malacca Strait, the pressure (mainly from the United States and Japan) has met with strong resistance from Malaysia and especially Indonesia,Footnote ⁹⁵ in the case of Somalia the UNSC interference effectively eliminated any vestige of sovereignty in the name of combating piracy. Although the UNSC resolutions against Somali piracy have ensured that all states should request prior permission from the Transitional Federal Government in Somalia (TFG) to enter Somali territorial sea or land when engaged in anti-piracy operations, the fact is that this embryonic government created by the international community was far from a central government in any practical sense. In reality, the TFG could barely control a few blocks of the capital Mogadishu with the help of the African Union peacekeeping forces, without any influence on the rest of the country or recognition and legitimacy among the Somali people as a whole.Footnote ⁹⁶ Although the existence of the TFG has allowed the UNSC to act as if dealing with the legitimate authority of a centralised state, in practice this has not been the case,Footnote ⁹⁷ and although the transitional character of this government achieved permanent status in 2012, one cannot say that the situation in the country has evolved substantially.

The case examined allows observing another relevant aspect related to the merging of internal and external dimensions of security at sea: the increasing adaptation of navies to constabulary functions. While twenty years ago this was a contentious issue – rejected by most naval officers still under the aura of the glorious strategic role of the Cold War’s large blue-water fleets, and doctrinally influenced by the legacy of Mahan’s strategic thinking on sea power and the symbolic and instrumental significance of navies in the international systemFootnote ⁹⁸ – today the picture looks different. Maritime patrolling, prosecution and imprisonment of criminals at sea, suppression of terrorists, pirates or drug traffickers, training of port police and coast guards and other constabulary tasks are relevant missions assumed under the umbrella concept of maritime security even by great naval powers. One can clearly see this move in recent United States documents, such as the Naval Operations Concept 2010 jointly published by the US Navy, the US Marine Corps, and the US Coast Guard. This manual shows that, besides combating piracy, other constabulary tasks should be jointly conducted by the three forces, such as the protection of maritime resources, the support of free and open seaborne commerce, as well as the suppression of maritime terrorism, weapons proliferation, transnational crime, environmental destruction, and illegal seaborne immigration.Footnote ⁹⁹ Other military forums, such as NATO, have replicated this attempt to construct maritime security as an umbrella concept to accommodate new threats and justify the use of naval forces in operations aimed at enhancing and enforcing law and order at sea.Footnote ¹⁰⁰ Within all of these documents, the piracy situation, illustrated by the case of Somalia, serves as an argument to justify increasing naval engagement in constabulary tasks.

One cannot also fail to observe some remarkable bureaucratic features of the internal-external interface of maritime security in the realm of technological and operational cooperation. The Somali case clearly shows technological and operational exchanges between foreign navies, coast guards and police forces; anti-piracy training missions of national maritime police forces conducted by foreign navy personnel; installation of radar and satellite surveillance systems in piracy prone areas (usually provided by the major naval powers); and the growing convergence of data on piracy attacks in several piracy information centres. Besides a range of military, policing, and judicial initiatives conducted by states and international organisations, several frameworks of interaction between international naval forces and the shipping industry, as well as anti-piracy mechanisms of cooperation between regional states, have been created since late 2008. The Maritime Security Centre-Horn of Africa (MSC-HOA), Shared Awareness and Deconfliction (SHADE), Training Awareness and Deconfliction (TRADE), the Djibouti Code of Conduct and the Contact Group on Piracy off the Coast of Somalia are major examples of these kinds of initiatives. The growing technological and operational cooperation, already noted by Didier Bigo in the context of Europeanisation of military-technological assistance,Footnote ¹⁰¹ has created in the maritime context a community of experts on piracy and built linkages between these transnational security professionals, who now share and reproduce similar fears and insecurities in the ocean space.

Regarding specialisation, also noteworthy is the emergence of a community of private maritime security professionals. While the privatisation of security was a trend already noted in the Afghanistan and Iraq wars, Somali piracy has opened the maritime environment to this kind of entrepreneurship, leading to a multiplication of private maritime security companies (principally based in the United Kingdom or in the United States and composed mainly by former military personnel). The reliance of the shipping industry on private maritime security companies (which was a kind of taboo since privateering was abolished in the mid-nineteenth century) began to intensify in 2011 due to the worsening of piracy despite the huge military mobilisation in the region’s waters since late 2008. In that context, pressures of the maritime industry on governments to allow firearms on board merchant ships resulted in a series of IMO circulars by the middle of 2011, giving ship-owners guidance on the use of private security companies for self-defence. Subsequently, several flag states such as Norway, United Kingdom, Greece, and United States changed their policies on this matter, deciding to endorse the use of private armed guards aboard merchant ships. As a result, it is now estimated that about 40–60 per cent of merchant vessels crossing the Western Indian Ocean are protected by private security professionals.Footnote ¹⁰² It can be noted, therefore, that the securitisation of Somali piracy has greatly contributed to changes affecting contemporary maritime security through privatised activities, broadening the limits of what can today be considered the maritime security community.

To sum up, securitisation allowed the problem of piracy to be refocused through strategic and supranational lenses in order to open increasing room for regulation and governance of the ocean space in a coercive, emergency, and transnational basis. Within this process, the question about scale seems crucial, and attempts to construct piracy and armed robbery at sea as threats to international peace and security demonstrate a clear intention to relocate these issues beyond national constraints, to allow external interventions that meet the stabilising purpose of global governance. If one projects on the ocean space the conceptual shift from ‘government’ to ‘governance’ briefly outlined at the beginning of this section, one realises that the securitisation of piracy triggered significant changes from traditional government polity to more fluid and flexible governance arrangements for the maintenance of order at sea. Because of this, it is plausible to affirm that securitisation ‘causes’ in the sense that it has constraining and enabling powers that can produce changes in the social world.

Conclusion: Securitisation as a causal complex

If we examine the securitisation of Somali piracy in a comprehensive way, we cannot fail to note that discursive, ideational, and symbolic elements (reconceptualisation of piracy, historical meaning associated with the word piracy, dramatisation of the discourse on piracy, association between piracy and terrorism, UNSC textual label ‘threat to international peace and security’) and extra-discursive factors (production and manipulation of statistics, civil conflict in Somalia, Somali humanitarian crisis, lack of central government in Somalia, heavy weaponry used by pirates, Somalia’s lack of material capacities to impose order, strategic interests of states and regional organisations, interests of international military bureaucracies, economic interests of private contractors) interacted and counter-interacted within the constitutive process of securitisation, triggering the conditions for removing national polity constraints (in general related to the norm of sovereignty, legal impediments to the entry of foreign militaries into Somalia’s territorial sea and land, and the traditional ban on the use of private armed guards on merchant ships), thus enabling an expanded framework for security to operate. This process paved the way for a shift from traditional government polity to more fluid and flexible governance arrangements in the maintenance of order at sea.

This expanded framework suggests that, despite the importance of the securitisation discursive basis, the study of its meaning-making process cannot be seen merely as a formal exercise and much less as an end in itself. As Guzzini observes, viewed as an empirical theory of security, securitisation is implicated in causal relations throughout the whole process, involving not only its performative moment, but also the contextual conditions for its empirical mobilisation as a collectively legitimated discourse of security. Reinforcing this view, the securitisation of Somali piracy analysed in this article shows that speech acts are part of a long-term process, in which discourses interact with norms and concepts; statistics; reasons, motivations, and strategic interests of private and public actors; as well as contextual conditions. Within this process, speech acts are components, not the entire process, and something more than a ‘quantum jump’Footnote ¹⁰³ or a ‘social magic power of language’Footnote ¹⁰⁴ happens so that the securitisation became successful.

Indeed, what happens in this process is that a complex of factors endowed with causal powers not only triggers its realisation as a successful securitisation (which means that the constitutive process of securitisation is also causal), but also explains the effects of this process on other processes, particularly on the governance of the ocean space. Thus, when analysing the securitisation of Somali piracy, we cannot isolate the speech acts from the broader empirical framework in which they are realised. It is within this broader context, where ideas, concepts, reasons, meanings, discourses, and extra-discourse factors interact and counter-interact, that the speech acts resonate, predisposing relevant audiences to accept as legitimate the political moves that insert the problem of piracy in a logic of exceptionality and urgency. From this perspective, the constitutive process of securitisation is not merely descriptive and formal, but is deeply permeated by causal powers. It is because securitisation has triggering powers, which result in constraining and enabling conditions for security to operate, that agents construct a shared sense of piracy as an existential threat. In other words, agents securitise piracy because this meaning-making process has real effects in terms of maritime governance.

This allows recognising securitisation as a causal complex that explains crucial changes in the world. But what is key for this explanation is the whole causal complex – the process of securitisation – not each of its constituent factors considered in isolation. As discussed in the article, several elements interact and counter-interact within the causal complex, contributing to bring about given events or processes. Within this process, no systematic relationship determines the weight or the individual role of each factor in the production of the outcome. Critical realists highlight, and Guzzini also suggests in his reflections on the causal status of securitisation, that causality in a non-positivist setting should be conceived as underlying multiple causes, which are triggered under generally unknown conditions or with indeterminate consequences, and that interact and counteract each other in complex and unpredictable ways.

Thus, rather than trying to isolate or trace a sequence of specific factors or particular events within a causal mechanism – as George and Bennett suggest, and Balzacq seems to replicate by proposing the use of process tracing as a way of determining ‘the relative status of one of the forces within the network of causality’ in securitisation analysisFootnote ¹⁰⁵ – the conclusion here is that a better way to approach securitisation causally is to consider the ‘complex’ and not any ‘individual factor’ (speeches, ideas, concepts, meanings, norms, rules, reasons, motivations, and extra-discursive aspects) as the significant element for the explanation.

It is important also to emphasise that the explanatory status of securitisation cannot be equated through Hume’s notion of causal necessity. As an open intersubjective social process, securitisation is always subject to possibilities of success or failure, which makes the idea of predictability meaningless for securitisation studies. One must keep in mind that the causal narrative examined here leads to particular outcomes within the particular context that involves maritime crime in the Horn of Africa. It cannot be said that the same outcomes would be generated in different social contexts, such as in the Malacca Straits or in the Gulf of Guinea, which have also been considered maritime crime prone areas. Thus, prediction is not compatible with the type of causal analysis advocated here, and explanations are always contingent, not generalisations.

To conclude, one could inquire about the relevance of bringing causality to securitisation studies. In this respect, the argument advanced here – by resorting to an alternative notion of causality, and assuming the explanatory status of securitisation – contributes, firstly, to better position securitisation theory within a broader empirical framework of world politics and its complex causes. Secondly, by recognising securitisation as a causal complex, this move may help to challenge the dominant vision in the IR field that placed constitutive theories in a subordinate position, enclosed within the limits of mere descriptive/understanding concerns, while the most valued position, endowed with explanatory status, is reserved for positivist forms of theorising. Thirdly, the framing of securitisation as a causal complex opens the door of securitisation studies to epistemological and methodological pluralism, going beyond its core poststructuralist discourse foundation and the constraints this imposes in terms of methods by limiting the study of securitisation to textual analysis. This debate on methods, generally neglected by securitisation theoristsFootnote ¹⁰⁶ and not explored here due to the limits of the article, opens an important window for further examination.

One can argue, finally, that the notion of causal complex and the integration of discourse in causal explanations contribute to establish a bridge between internal disputes and controversies in securitisation studies that still compel analysts to choose between linguistic or sociological versions of securitisation, as if they were absolutely incompatible with each other.Footnote ¹⁰⁷ The aspects discussed in this article make securitisation as a whole compatible with a holistic notion of causality, understood as mutually reinforcing causal and constitutive forces that work to constrain or enable their security effects. This provides a common ground for integration between linguistic and sociological versions of securitisation, whose disputes have led to an increasingly sterile debate.