Revisiting ‘identity’ in International Relations: From identity as substance to identifications in action

Introduction

In the past 25 years, the concept of identity has received wide attention from constructivist and, more broadly, non-foundationalist International Relations (IR) scholars. Identity scholarship has successfully argued that preferences, interests and norm structures are inseparably tied to actors’ identities, which need to be taken seriously in order to account for state action and international security dynamics. Today, the concept of identity is central to research agendas that seek to move beyond rationalist and materialist assumptions of state action. Yet, despite the attempt of identity research to leave rationalist and materialist IR behind, it oftentimes continues to model identity in methodologically individualist and causal terms and thereby fails to account for the performative dimension of actors’ self-understandings within social arrangements. More concretely, this gives rise to two interrelated problems, which we seek to address here. For one, contemporary identity research runs the risk of treating identity as a property of the secluded individual (see Bucholtz and Hall, 2004: 376). This is partly surprising since identities are argued to be constructed in relation to others. However, while identity research underscores the relationality of identity, it tends to focus on narratives of the self in domestic discourse and does not sufficiently situate actors and their actions within their complex networks of interdependencies (Elias, 1978; Guillaume, 2007; Onuf, 1998). Hence, by focusing away from the historically contingent social embeddedness of actors, identity research inadvertently stabilizes and reifies articulations of a core identity.

Taking such a perspective is, second, tied to treating identity in terms of a causal variable that gives rise to, and explains, action. It seems that it is now largely considered to be (ontologically) unproblematic to argue that state actions can be understood by linking them to underlying notions of identity. As such, identity is treated as the functional equivalent of independent variables in most constructivist scholarship. In practice, even post-structuralist identity research understands the concept of identity as something that precedes state action and that lets us grasp why states act in certain ways (e.g. Hagström, 2015; Hansen, 2006).

This substantialist way of thinking about identity leads to asking research questions aimed at uncovering what the true or essential (even if multiple and/or fragmented) identity of something is and how it consequently gives rise to actions. In addition, and more basically, conceptualizing identity in this way inadvertently undermines the non-foundationalist starting point of constructivist and post-structuralist thinking. This effectively draws into doubt whether contemporary identity scholarship has successfully left the confines of rationalist and materialist approaches of IR behind.

Given these difficulties, we suggest replacing the concept of identity (multiple, fragmented) with a focus on ‘acts of identification’. This conceptual shift allows us to systematically study empirically observable articulations that make reference to ‘identity’ (acts of identification), without reproducing the substantialist view that social entities have identities that precede and explain action. Clearly, not all acts of identification become central to political debates. Some articulations of ‘what identity is’ will remain marginal, while others are made prevalent. We suggest referring to what is commonly termed individual or state identity as bundles of identifications that acquire a temporarily privileged status within a specific discourse. ¹ However, centrally, we do not view the temporary and incomplete privileging of identification bundles to be caused by a deeper underlying ‘identity’ of the social entity. Rather we suggest that acts of identification are bundled together through action. Bundles of identifications can consequently be understood as the temporal stabilization of a specific set of acts of identification in a contested political discourse. ² As these acts are always geared towards something/someone, identification practices have an irreducibly relational dimension. We thereby hope to make visible not only the developmental character of identification bundles, but also their social embeddedness.

Accordingly, we aim to refocus our research from an analysis of identity as a substance that causes, to studying acts of identification in relations (figurations). In doing so, we combine processual-relational thinking (p-rt) (Jackson and Nexon, 1999) and figurational sociology (Elias, 1978). This provides us with an alternative to ahistorical and methodologically individualist conceptions of identity, which necessarily downplay the evolving and politically contested character of identification bundling. Competing identifications, especially in the form of alternative narratives of the ‘self’ and diverging articulations of relations to ‘others’, play a central role in the way identifications are temporarily bundled and stabilized in action. The processual character of our approach focuses on the contestedness of identification practices and makes the continual development of identification bundles visible.

The shift to identifications aims to move beyond treating identity as ontologically separate from social relations, which ultimately makes it difficult to understand the historical development of ‘articulations of the self’. We hope that the mode of thinking that we will outline here can complement and add to post-structuralist perspectives, with an approach that highlights (the unintended consequences of) agency in social processes as it focuses on the temporal outcomes of actions.

Our argument proceeds as follows. We will first take a closer look at the existing literature and outline in more detail why the conceptualization of identity in IR and Security Studies should consider moving towards studying acts of identification based on processual-relational sociology and figurational thinking. To present our argument, we will develop a figurational framework of identification analysis that builds on a short introduction to p-rt in the second part of this article. Since we believe that meta-theoretical debates should ultimately be aimed at making practical differences in research, we will discuss the methodological dimension of working with a p-rt model of identification analysis in a third step. To illustrate how identification analysis differs from identity scholarship, we will subsequently apply our approach to the case of Switzerland. More specifically, we will present an analysis of two foreign policy episodes: first, and in some detail, we will focus on Switzerland’s ‘flirtation’ with the acquisition of nuclear weapons in the 1950s and 1960s; and, second, and more briefly, we study its long road towards becoming a member of the United Nations (UN) in 2002. While Switzerland’s long tradition of neutrality would seem to suggest a rather stable identity that could explain behaviour, we will show how political processes are continually characterized by competing bundles of identifications, which temporarily and incompletely acquire a privileged status in (foreign policy) decision-making. In closing, we will reflect on the broader implications of our approach for understanding a range of issues pertinent to both current international relations and IR as a discipline.

Reconsidering identity in IR and Security Studies

Following a prolonged period in which identity as a category of analysis was broadly neglected, constructivists and post-structuralists have, in recent years, revived the notion that identities play a central role in international relations. Challenging the rationalist assumption (Mearsheimer, 2001; Moravcsik, 1998; see also Snidal, 2002) that actors’ identities are given prior to and independent of interaction (Fearon, 1999; Price and Reus-Smit, 1998; for a critical discussion, see Lebow, 2008, 2012), their research findings underscored that identities might not only be situated at the ‘core of national and transnational interests’ (Adler, 2002: 103, emphasis added; see also Gourevitch, 2002; Tickner, 1996), but are continually subject to change on the basis of interaction (Adler and Barnett, 1998; Anderson, 1991; Lynch, 1999; Wendt, 1999).

Constructivists have subsequently studied the mechanisms through which identities change over time, and what this implies for interests and actions (Adler, 1997; Checkel, 1999, 2001; Crawford, 2002; Fierke, 2007: 75–98; Gurowitz, 2006; McSweeney, 1999). Broadly speaking, one can distinguish between two strands of conceptualizing identity in constructivist research. While ‘norm-focused’ constructivists stress how identities develop through contestation and alternation, and dissociation or adherence to prevailing norms (e.g. Finnemore and Sikkink, 1998; Klotz, 1995; Tannenwald, 2007), more relationally inclined constructivists place emphasis on the formative role of others for identity formation processes (e.g. Jackson, 2013; Jackson and Nexon, 1999). While this research highlights the conceptual importance of identities for understanding the dynamics of international relations, Kowert and Legro’s (1996; see also Zehfuss, 2001) critique that constructivists all too often fail to account for the sources and the genesis of identities remains valid. Similarly, Brubaker and Cooper (2000: 6) are correct to point out that constructivist studies are often characterized by ‘an uneasy amalgam of constructivist language and essentialist argumentation’ (for a critique of essentializations, see also Epstein, 2010). This is problematic because socially constructed identities cannot serve as static foundations for theorizing (Bucholtz and Hall, 2004: 382; Joas, 1989: passim), or as analytical primitives that explain social action (Epstein, 2013; Guzzini, 2005; Kratochwil, 2008). Essentializing identity and treating identity as something that precedes (and explains) action at a basic ontological level (or replacing the notion of one identity with the concept of multiple roles (McCourt, 2012)) leads to empirical research that cannot grasp the constitutive relationship between actors’ identities and the social arrangements in which they emerge. As such, we argue that both the empirical and theoretical shortcomings of identity research originate from ahistorical individualist conceptions of identity.

Post-structuralist research takes issue with the shortcomings of mainstream constructivist identity research at a theoretical level. It avoids essentializations on the basis that the ‘constitution of identity is achieved through the inscription of boundaries that serve to demarcate an “inside” from an “outside”, a “self” from an “other”, a “domestic” from a “foreign”’ (Campbell, 1998: 9). Hence, identity is known ‘by what it “is not”, that is difference’ (Hagström, 2015: 125; see also Strömbom, 2014). Rather than conceptualizing identity as a unitary entity that precedes actions ontologically, it is understood performatively, that is, as an ‘ongoing, always incomplete series of effects of a process of reiteration’ (Laffey, 2000: 431; see also Agius, 2013: 243; Weber, 1998). Consequently, post-structuralist approaches seek to avoid notions of fully structured subjects (Laclau, 1990: 44) that would amount to ‘mere dupes acting according to what particular sets of practices dictated’ (Doty, 1997: 380). This is accomplished by substituting notions of structure as ‘fully constituted objective wholes’ (Torfing, 1999: 148) with the concept of ‘playful structure’, which is ultimately undecidable (Doty, 1997: 383–384). The necessary incompleteness of structural identity thus ‘constitutes the subject as the locus of a decision about how to establish itself as a concrete subjectivity with a fully achieved identity’ (Torfing, 1999: 149). Identity, then, is understood not as a positive presence, but as a constitutive lack (Epstein, 2010: 336).

Post-structuralism and the approach we introduce in the following share a number of similarities, especially with regards to ontology and non-foundationalist commitments. However, two significant differences, which have implications for research strategies and empirical analyses, stand out. The first difference concerns the post-structuralist focus on broad discursive structures, which is directly linked to the notion of constitutive lack outlined earlier. In decentring their analysis, post-structuralists tend to discard individual action. Although this is unobjectionable in principle, agency and the highly contingent unintended consequences of agents acting in the world (see Elias, 1978) are often eclipsed from analysis. This might be more or less desirable, depending on the specific research objectives at hand. If one focuses on conditions of possibility, structural conditions for action and prevailing power relations at a macro-level, the eclipse of agency might be welcome. However, it might be less desirable if one seeks to grasp how articulations in decision-making are tied to individual and societal developments.

Second, and relatedly, we argue that some post-structuralist identity scholarship substitutes a (realist) meta-narrative of perpetual power struggles with an equally compulsive, trans-historical struggle to secure identity (Campbell, 1998; Cha, 2015; Hansen, 2006). The ‘drive to fix the state’s identity and contain challenges to the state’s representation’ (Campbell, 1998: 12) subsequently takes the place of a new master narrative. History is thereby ‘reduced to a single theme and variations, to singular “moments” of identity formation and re-formation’ (Tuathail, 1996: 651). More importantly, by primarily situating identity construction at the domestic level and by subsequently asking how identities shape foreign policy, post-structuralists treat identity as a cause of action. This ultimately boils down to the assumption that ‘[w]hat a state is … is thought to translate into how it behaves’ (Hagström and Hanssen, 2015: 4). However, focusing on domestic identity discourses that ‘generate’ foreign policy resembles essentialist approaches in which identity (often through the internalization of norms) causes actions. The primary focus on domestic discourses consequently runs the risk of reproducing identity as something that is prior to action.

The approach that we introduce in the following builds on relational constructivism and post-structuralism, and adds a process-based figurational sociology perspective. We thereby seek to show how identifications become privileged and enacted within constitutive networks of interdependencies, that is, within ‘figurations’ (Elias, 1978; Linklater, 2011).

Taking a processual-relational approach: From identity to identifications

P-rt replaces the concept of identity with (multiple) acts of identification, which are bundled and temporarily and incompletely privileged (or marginalized) in the act of decision-making. Identification bundles thus appear as a multidimensional process that is stabilized through the performance of political action vis-a-vis others. Such a conceptualization calls for a fundamental shift away from a substantialist ontology (Elias, 1978) in order to grasp what is commonly referred to as ‘identity’ in terms of a cultural-political artefact that is permanently re-established. Such a perspective sidesteps the ahistorical essentialism of identity research by dissolving the necessity to posit a core or substance of identity underlying action in the first place. P-rt begins with the counter-intuitive notion that processes and relations, not substances or things, are the most basic and most important elements of (scientific) ontologies (Bucher, 2015; Jackson, 2011; Jackson and Nexon, 1999). Where substantialist thinking tends to see separateness, stability and passivity, p-rt sees constitutive interdependence, development and activity (Rescher, 1996: 35). As relations and processes of boundary-drawing and individuation are considered fundamental to our understanding of social objects/reality, p-rt sees unity and stability (‘thingness’) as temporal achievements, that is, as artefacts of socio-political processes.

From this perspective, social entities are not given apart from social contexts, but are conceptualized as ‘complex bundles of coordinated processes’ (Rescher, 2000: 9) that exhibit varying degrees of stability. ‘Identity’, then, no longer appears as a unified (given) thing, a ‘substratum’ underlying actions, but as a ‘unity of functioning’ (Rescher, 2000: 16), that is, a bundle of identifications that emerges through discursive practices and that, in turn, shapes social ties. Analysing discursive practices, then, aims at grasping how bundles of identifications are tied together and how they are temporarily privileged or relegated within specific contexts, and in relation to other actors who themselves are continually re-imagined in narratives of the ‘self’ (Guillaume, 2007: 750–751). As such, we focus on the processual link between actors (persons/states) and society (human/of states). From a figurational perspective, ‘“the individual” and “society” [are understood as] “two different but inseparable levels of”’ social reality (Bauman, 1989: 39). The privileging of identification bundles is consequently continually achieved through altercation and in relation to other actors and significant symbols (Mead and Morris, 1934; see also Dunn, 1997). This shifts analysis from a focus on a self–other dichotomy (in the singular) towards studying identifications in figurations, or, more precisely, in foreign relations (in the plural). As actors in processual-relational terms are viewed as continually constituted intersections within social networks, the individual (person or organization) in the singular is not privileged ontologically, but constitutively embedded within societal interdependencies. The added value we discern is that actors within social formations become visible not as an original presence, substance or cause, but as an intersection of constitutive interdependencies (Elias, 1978). This ties together the development of actors and their social arrangements, or, in figurational terms, their psycho- and socio-genesis. It thereby allows us to grasp how micro- and macro-sociological phenomena develop in an intertwined way. This establishes the basic ontological link between individuals and their social environments, which is largely missing in contemporary literature.

This processual-relational perspective also underscores that acts of identification contain references to the past (where we come from), to the present (who we are) and to what we might become in the future (Berenskoetter, 2014; Emirbayer and Mische, 1998). It is consequently central to emphasize that acts of identification relate actors not only to other actors, but also to imaginations of past and future versions of the ‘self’.

Given this basic understanding of what a non-essentialist concept of identifications entails, it is important to flesh out the implications that such a perspective has in terms of empirically studying how identifications are bundled and privileged (or marginalized) in practice. Clearly, inquiries cannot aim to uncover some true or stable ‘identity’ underlying ‘more superficial or artificially imposed “selves”’ (Hall, 1990: 223). Rather, identifications and how these are bundled and temporarily and incompletely privileged in actions must take centre stage. A focus on identification practices at base studies the temporal fixing of meanings and the political practices of boundary-drawing, which characterize political (and especially) foreign policy decision-making processes. In contrast to essentialist notions of agency, which are tied to what one might call interior identity, acts of identification (as discursive traces) are out in the open and empirically accessible. This has the distinct advantage that we need not seek to uncover motives, intentions or convictions, but can instead retain a focus on actions and their (unintended) consequences.

This move from identity to (bundles of) identifications is conceptually demanding, but well-suited for IR. When speaking about international actors, one is not immediately tempted to take recourse to ‘the great myth of interiority’ (Foucault, 1989: 21). States as organizations are recognized as developing sets of subject positions, rules and resources, loci of decision-making, and so on. They can be seen as multidimensional processes that cannot be reduced to a unity of intention, or underlying core agency. Rather, states and other international actors are easily linked to multiplicity, contingency and development. IR therefore allows for the application of processual-relational modes of thinking in a surprisingly intuitive fashion.

However, irrespective of the specific social entities under investigation, studying acts of identification is central to understanding the ongoing construction of social reality. Identifications are acts of demarcation and differentiation and are consequently part of an irreducibly relational process that intertwines actors and society. Consequently, our focus on relations over time does not seek to simply add another layer to studies in which domestic identity formation shapes foreign policy. Rather, we look at how bundles of identifications are temporarily and incompletely tied together and privileged in foreign policy decision-making processes. As such, we do not merely view ‘the foreign policy arena as a locus of national identity contestation’ (Hintz, 2015: 5), but rather view it as the decisive theoretical entry point to understanding identification politics. As this approach is not applied widely in IR (for a rare exception, see Guillaume, 2007), we will provide a methodological toolkit that makes such a conception manageable for practical research.

Studying (bundles of) identifications: Methodological considerations and research procedure

Given these theoretical considerations, research needs to be attentive to identifications in specific foreign policy relations. As ‘identities’ are enacted in relation to others, it is central to trace how bundles of identifications are moved centre stage (or sidelined) in specific (foreign policy) decisions, and to make visible how these articulate the relations between ‘the self and others’ (be they state or non-state actors), as well as how they articulate the relations between ‘past, present and future selves’ (spatial and temporal relatedness).

But how can these processes be studied empirically? As ‘research methods … should be aligned with the researcher’s style of reasoning’ (Pouliot, 2007: 360), we base our empirical studies on the methodological principles of Grounded Theory (GT) (Glaser and Strauss, 1999; Strauss, 1987; Strauss and Corbin, 1994; Strübing, 2004; Jasper, 2013; Wilson, 2012). ³ GT methodology is characterized by an abductive, iterative-dynamic research process that facilitates the elaboration of intertextual linkages. This dovetails with our attempt to uncover how different acts of identification are interwoven and privileged (while others are relegated) in actions. Moreover, rather than subsuming new empirical data under given categories and analytical concepts, this methodology calls for a continuous interplay between data collection, analysis and concept formation. It demands us to repeatedly relate preliminary results to new material, thereby generating a grid of interpretations. This implies that we do not establish a priori the relevant ‘speakers’ (nor the significant others) in identification processes, but rather identify them during the analysis.

In specific methodical terms, this leads to a twofold research procedure. On the one hand, GT theorists ‘comb’ through the textual material, seeking to abductively collect ‘pertinent observations’ (Friedrichs and Kratochwil, 2009: 709), that is, references to a state’s identity. Such ‘combing’ processes can be guided by questions like: Who speaks? What are the concepts informing identifications (like ‘Western’, ‘progressive’ or ‘socialist’)? Do certain identifications exclusively correlate with specific ‘others’ or specific issue areas? Can argumentative incompatibilities or discursive tensions be identified? If so, how are these dealt with? How do identifications in the context of one dyad correlate with identifications in other relations? Do identifications co-vary across time and relations? Naturally, addressing questions of this kind produces a vast set of references. Since not all of these will be equally discursively significant, it is important to collate the most frequently recurring and arguably most important references that take centre stage.

The second and interwoven research level encompasses the explication of intertextual linkages and cross-references within the discursive corpus. The goal is to establish a preliminary web of identifications among the different documents, reflecting both temporal and relational patterns. This allows uncovering the temporal shifts in identifications over time, and how different identity interpretations relate to (or perhaps contradict) each other. Importantly, the two parts of the research process do not take place in subsequent phases, ultimately moving towards the establishment of one or a few core categories, but stand in a reciprocal relationship with each other.

In order to facilitate data interpretation, we additionally suggest disaggregating research projects into three dimensions of analysis (for a similar approach, see Wæver, 2002). This analytical tripartition reduces complexity and systematically directs attention to different but interrelated aspects of discourses, yet does so without introducing pre-established explanatory variables.

A first dimension of analysis centres on identifications that express ‘self-understandings’ (narratives of the ‘self’). These types of articulations situate ‘the self’ in regard to, for example, religious, economic or political principles, and can include references to cultural, historical or foundational narratives. This focus emphasizes the role of temporal relations in identifications since it reveals how a contemporary narrative of the ‘self’ is situated in relation to the past and the future. Clearly, past imaginations of the ‘self’ cannot serve as objectively given foundations, but are themselves subject to continuous interpretation and discursive contestation (Carr, 1961). When analysing these types of identifications, one can view them as being located on a continuum ranging from continuity, on the one hand, to fundamental re-conceptualization (change), on the other. Identifications that relate to past and future (short of actor/state dissolution) will always involve the construction of continuity (Wodak and De Cillia, 2007). At the same time, attempts to construct continuity will necessitate interpreting the past in the light of contemporary and future challenges, making simple reproduction a problematic notion. While such processes can be located at very different points along the ‘continuity versus fundamental re-conceptualization’ continuum, they are inevitably characterized by a double movement that ties together continuity and change. As such, we hold that continuity and change are not opposed in acts of identification, but are fruitfully conceptualized as two sides of the same coin.

A second dimension of analysis aims to uncover which ‘others’ take a central position in identification practices. As such, this dimension of analysis situates narratives of the ‘self’ within a figuration of multiple different actors. Importantly, it will ask about the different categories and concepts informing identifications in the specific relations that take centre stage in foreign policy discourse. These might include depictions of relations as being antagonistic, cooperative, friendly, strategic, traditional or the like. Actors may also be categorized as belonging to different status groups, civilizations or imagined communities like the West, or as being on different sides of history (cf. Clark, 1989; Dunn, 2003; Jackson, 2006; Klein, 1990). This dimension of analysis will inquire into the ways in which identifications map onto specific ‘I–we–they’ logics. This possibly includes taking into consideration how respective (state) actors ‘articulate state identity’ in taking the perspective of those considered to be significant others (Mead and Morris, 1934: 140). It will therefore be important to systematically include (where possible) articulations that take the point of view of the ‘other’, or from the ‘generalized standpoint of the social group as a whole’ (Mead and Morris, 1934: 138). Clearly, the uncovering of the most significant relations in identification practices — and the respective categories of identification — need to be arrived at empirically and cannot be established in advance. As the selection of relations is guided by the analysis of primary sources, a specific focus (e.g. geographic, economic or security) cannot be privileged prima facie. Also, the analysis should provide for the possibility that non-state actors might be central to identification practices. Changes over time in the significance attributed to different types of actors promise to reveal shifts in international discourses on sovereignty and legitimacy.

Building on the first two dimensions of analysis, a third dimension asks about identifications in specific policies. This also concerns the range of policy options that are taken into consideration and the justifications underlying political decisions. The policy dimension is of interest for at least two reasons. First, it allows for the uncovering of argumentative incompatibilities or discursive tensions in identifications across policies, while making visible that identity is not prior to actions. Second, the justifications of actual policies are central elements in the process of temporarily fixing meanings and consequently for the temporal and incomplete privileging of (bundles of) identifications. Political decisions will move certain (bundles of) identifications centre stage and marginalize or relegate contending narratives of the ‘self’. This third dimension of analysis ties together foreign policy statements and actions on the ground and allows us to empirically grasp the processes of political contestation (Wiener, 2009) underlying the privileging of specific identifications over others.

Taken together, this analytical framework allows the tracing of identifications over time and their bundling and privileging in action. It enables us to grasp how identifications continually reconstruct imaginations of the self in relation to others (states, international organizations (IOs), non-governmental organizations (NGOs), international society), as well as in relation to the past and future. Although we analytically separate temporal and spatial identifications in order to simplify the analysis, it is important to remember that both are ontologically intertwined.

Connecting these dimensions of analysis does not aim to find a unified notion of identity or a cohesive self. While we expect that inquiries will show that some relations and categories will be more important and more stable across identification practices than others, we are clearly not looking for a substantialist core of identity. We seek to make the complexity and development of identifications visible in order to stress that ‘identity’ does not precede and/or cause action, but rather that identifications are continually tied together in political decision-making processes.

In the following part of the article, we will apply our model to the case of Switzerland, drawing on two episodes: the debates on acquiring nuclear weapons in the 1950s and 1960s; and the UN accession issue after the end of the Cold War. Several considerations guided our case selection. First, in contrast to other cases, such as the USA (Campbell, 1998; Cha, 2015), Germany (Banchoff, 1999; Kirste and Maull, 1996) or Japan (Hagström, 2015; Yamada, 2014), Switzerland has not received much attention in the (English-language) IR literature, nor in the more specific literature on identity. This is regrettable as the Swiss case is particularly rich and abundant. Questions related to ‘Swissness’ have long been (and continue to be) highly contested issues on the country’s political agenda. This provides a broad reservoir of empirical traces. At the same time, and puzzlingly, a high degree of contestation over identifications is observable despite Switzerland’s long history as a neutral state. Indeed, one might expect that a state that has anchored its foreign policy in neutrality for two centuries would exhibit a more consolidated and static ‘identity’. Yet, counter-intuitively, this is not the case, as our analysis is able to demonstrate. Rather, the Swiss socio-political discourse in general, and in the presumably ‘challenging field’ of security policy more specifically, is inextricably interwoven with competing identifications. The debate on the nuclear question in the 1950s and 1960s exemplifies this vividly, as the following quote underlines: ‘The nuclearization of our military is not merely a military question, but to a much larger extent a political and moral one which touches upon the principle ideas on which our country is built’ (Schweizerische Bewegung gegen die atomare Aufrüstung, 1962a). ⁴ The Swiss case is therefore a theoretically challenging, empirically rich and yet under-studied case that allows us to analyse a broad array of identification processes over time.

Swiss identifications and the nuclear question

The question of whether to acquire nuclear forces (policy dimension) was highly contested in Switzerland after the Second World War, both in regard to how military nuclearization would square with different notions of ‘being Swiss’ (competing temporal narratives of the ‘self’) and in regard to what either strategy would imply for ‘what being Swiss meant’ in relation to others (identifications in specific relations).

Switzerland had not participated in the Allied war against Germany, leaving the country somewhat isolated politically once the war was over. Its bilateral diplomatic relations with the major powers were severely strained (Altermatt, 1992; Spillmann et al., 2001). It was against this background that the Swiss began re-evaluating their security situation in the arising bipolar world order (Möckli, 2000a; Trachsler, 2011). In this context, members of the military and of the centre-right political spectrum called for a serious assessment of the potential role of nuclear weapons, which culminated not only in a long and fierce public debate, but also — historically unique — in two public referenda regarding the acquisition of nuclear weapons in the early 1960s. Asked whether or not the government should be prohibited from acquiring nuclear weapons, the majority of voters answered in the negative, thereby giving the decision-makers a free hand in defining the state’s nuclear future. This ‘carte blanche’ notwithstanding, decision-makers ultimately abstained from military nuclearization and, instead, signed (in 1969) and ratified (in 1977) the Nonproliferation Treaty. Large parts of the proliferation literature account for this decision by reference to overriding security calculations (Paul, 2000; but see also Metzler, 1997; Winkler, 1981; Wollenmann, 2004). It was only more recently that several theoretical and empirical studies pointed at the crucial role of identity considerations for states’ nuclear decision-making (in general and with regard to the Swiss case) (Abzhaparova, 2011; Hymans, 2006; Jasper, 2014). At the same time, most studies reproduce identity as substance and as a preceding cause of the decision to (non-)proliferate.

Focusing on the three, often overlapping, analytical dimensions introduced earlier — narratives of the ‘self’, relations towards ‘others’ and specific policies — we seek to illustrate in the following that it is misleading to posit a core of Swiss identity that determined the state’s nuclear course. Instead, our GT-based analysis of the historical contestation process over acquiring nuclear weapons reveals three different, often conflicting, strands of identifications, all of which largely revolved around different interpretations of neutrality. These differing identifications (mostly upheld by three groups of actors: the initially pro-nuclear government, pro-nuclear non-governmental groups and the anti-nuclear movement) were continuously rearticulated. Rather than proposing that some specific notion of Swiss identity preceded the decision to abstain from nuclearization, we hold that the decision process itself temporarily and incompletely privileged specific bundles of ‘what it meant to be Swiss’ in its decision not to acquire nuclear weapons. We will also show in a brief second empirical episode that this did not mean a closure of the struggle over identifications.

Dimension 1: Temporal narratives of the ‘self’ and the nuclear debate

The decisive part of the debate over nuclear weapons began in July 1958, when the Swiss government took a public stand on the nuclear question for the first time, immediately leading the New York Times to opine that ‘The Swiss will seek atomic weapons’ (NYT, 1958). In response, the government explicitly framed the issue of nuclearization as a necessary means for the continuation of Switzerland’s traditionally neutral attitude in the international system (Eidgenössisches Militärdepartement, 1958). The argumentation identified neutrality and independence as historically situated at the heart of Swiss ‘identity’. The Ministry of Defence argued that ‘In accordance with our centuries-old tradition of resistance … the army should be equipped with the most effective weapons for the preservation of our independence and the protection of our neutrality. This entails nuclear weapons’ (Eidgenössisches Militärdepartement, 1958).

This way of linking contemporary ‘articulations of Swissness’ with a tradition of neutrality and independence not only resonated with the contemporary audience (see for example Bindschedler, 1956; see also Gabriel, 1997: 19–27), but also inscribed nuclear weapons with an almost transcendental value that went far beyond their purely strategic purpose. They were not understood as a deviation from the state’s traditional foreign and security policy, but, to the contrary, were argued to embody a high degree of continuity with both historical policies and what these identifications expressed ‘Swissness’ to be. Nuclear weapons were framed as apposite, modern technological means to preserve the country’s historically developed ‘identity’.

At first sight, it appears that non-governmental pro-nuclear voices simply followed the governmental advocacy for nuclear weapons by articulating identifications that portrayed Switzerland as a long-standing neutral international actor. However, this group predicated neutrality in a distinctly different manner. By asserting what they depicted as the country’s defiant attitude and non-submissive stance in the past (e.g. in relation to the threat posed by the Habsburg army in the 14th century), non-governmental proponents of nuclearization portrayed the Swiss Federation as a model of fortitude and virtue. As we will outline in more detail later, this characterization culminated in identifications that depicted Switzerland as being virtually the antithesis of a ‘normal’ actor in the international system (Schweizerisches Aktionskomitee gegen die Atominitiative, 1962; see also Breitenmoser, 2002: 39). ‘Armed neutrality’ was thus identified as the central historical category of Swissness and elevated to an almost mythical backbone of the country’s self-proclaimed identity. This historical paragon of Swiss ‘identity’ should consequently be upheld and defended through the acquisition of nuclear weapons, the non-governmental proponents of nuclearization argued.

The third strand of identifications, articulated by the anti-nuclear movement, promoted a very dissimilar interpretation of Switzerland’s historical ‘identity’. While pro-nuclear advocates tied together identifications focused on armed neutrality and independence/resistance, and linked these to the past, the anti-nuclear movement articulated identifications that were less confrontational and less antagonistic, and that were related to a different historical vantage point. They centrally associated their understanding of ‘what it meant to be Swiss’ with the humanitarian heritage of the Red Cross (and its founder Henri Dunant), characterizing the country’s historical ‘identity’ as that of an entrepreneur of ‘worldwide humanity’:

Switzerland is the country of respected neutrality; the country that has — through the deeds of Henri Dunant — countered the cruelty of war with the enterprise of worldwide humanity. The moral credit that Switzerland has gained thanks to its humanitarian mission in the world would be questioned by its own nuclearization. (Schweizerische Bewegung gegen die atomare Aufrüstung, 1958)

Thus, while also focusing strongly on a historical narrative that centred on neutrality, the anti-nuclear movement linked neutrality to a set of values that differed starkly from the focus on resistance and virtuous self-preservation.

Concentrating on the temporal relations underlying competing narratives of the ‘self’ in this case reveals that neutrality served as a broadly shared historical concept. At the same time, this concept was interpreted in dramatically different ways and associated with distinctly different identifications (see also Ludi, 2005). These diverging imaginations of the past were explicitly linked to what the three groups dominating the debate considered to be ‘Swiss’. Moreover, their interpretations of Switzerland’s past ‘identity’ were clearly intertwined with visions of what Switzerland should be in the future.

Dimension 2: Relations towards others

A close reading of the identifications that characterized the nuclearization debate indicates that not only temporal identifications, but also spatial identifications, were highly contested. The administration sometimes portrayed Switzerland as a passive and weak actor and described the country as a small, non-influential member of the international system that lacked the means necessary to effect political change. Here, we cannot analyse the more specific relations with other states like the US or the USSR, and especially the neighbouring states of Italy, France or Germany. However, in terms of relating Switzerland to a generalized other, Switzerland was characterized as a ‘non-threat’ (see also Bundesrat, 1961) to any other political actor. At the same time, the administration evoked the fear of further marginalization: ‘The unarmed and weak will not be able to survive in the long run; such a country will go from being a subject to being the object of world politics’ (Bundesrat, 1960: 322).

The analysis of non-governmental sources reveals, moreover, that this positioning of Switzerland within its social environment was challenged by two perspectives. While the government sought to portray the country as a weak but ‘normal’ small statewithin the international system that was to remain detached from the tensions inherent in the bipolar order, the non-governmental nuclear advocates’ articulations were discernibly different. Here, Switzerland was not primarily depicted as a neutral and benign, peace-loving, defensive state, but as clearly juxtaposed to an antagonistic other, namely, the ‘communist threat’. In contrast to the government, the documents released by non-governmental nuclear proponents reveal a far more unambiguous and explicit stance regarding the perceived threat environment and the global strategic situation, describing communism as an inherently expansive, predatory, totalitarian and violent force that needed to be actively opposed (Erklärung der 35, 1958; Schweizerisches Aktionskomitee gegen die Atominitiative, 1963). They also related to several other qualitative characteristics — for example, bravery and heroism — that were said to define not only the Federation’s exceptionalism, but also its — almost recalcitrant — relations towards others (e.g. Verein zur Förderung des Wehrwillens und der Wehrwissenschaft, 1962). As a consequence of this antagonism, the country was argued to have every right to acquire the weapons it needed in order to guarantee its continued existence. Thus, rather than merely referring to a historically evolved and broadly shared concept of neutrality, the invoked notion exhibited a more irreconcilable stance towards the communist enemy and a determinedly militaristic subtext.

Proponents of nuclear abstention by contrast portrayed Switzerland as much less of an outsider willing to resist international pressure. Rather, they underscored the country’s role as a neutral broker that actively seeks to contribute to a functioning international order. They, too, alluded to notions of exceptionalism and uniqueness and stressed Switzerland’s detachedness from traditional forms of power politics, yet emphasized the opportunities granted by neutrality. Switzerland was portrayed as an atypical international actor — but one that was clearly part of the script of international politics and that had an active external role to play. Moreover, while the advocates of nuclearization envisioned a state that withdraws as much as possible from external affairs, the peace movement’s outlook was explicitly international. It stressed notions of shared humanity and shared responsibility, referred to nuclear debates in other (European) states, and aimed to integrate the Swiss decision into the broader context of domestic and international politics (Schweizerische Bewegung gegen die atomare Aufrüstung, 1962b, 1963b). This common notion strongly shaped the way Switzerland was related to others and, in turn, how being Swiss was understood. Consequently, focusing on the figurational relations of Switzerland suggests that the relations — and, consequently, the ensuing foreign policies — towards others were understood in three different, even contradictory, ways: detached neutrality versus antagonistic resistance versus active involvement.

Dimension 3: Policy perspectives

An analysis of the official governmental position reveals that the administration maintained that the procurement of nuclear weapons would not undermine Swiss ‘identity’, but would conform to the country’s century-old tradition of neutrality. By semantically linking nuclear weapons with predicates such as ‘historical’, ‘legal’ or ‘defensive’, the policy of ‘going nuclear’ was represented as a means for the continuation of what was argued to be Switzerland’s traditionally neutral, non-offensive attitude towards other actors. Non-governmental advocates of nuclearization, moreover, justified the acquisition of nuclear weapons by emphasizing Switzerland’s role as a resistant and defiant neutral in a largely hostile environment. While these two perspectives converged, the peace movement’s identifications entailed a different foreign policy agenda, namely, one that aimed to maintain Switzerland’s ‘moral credit’ by actively promoting good offices (Probst, 1963; see also Stamm, 1974) and foreign policies based on humanitarian considerations rather than realpolitik. The renunciation of nuclear weapons was thus seen as a proactive, ‘humanitarian’ foreign policy in line with what they argued Swiss ‘identity’ to be. This pointed to an activist and constructive conception of foreign policymaking and a depiction of Switzerland as an apparently influential, though militarily self-limiting, player that had the duty to influence and shape international relations according to normative considerations:

Our struggle is a creed to the accepted platform of traditional power-political self-limitation that the small state has to take as its starting point in order to contribute in an adequate manner to the constitution of international understanding. Its goal is: to win the majority of the Swiss people for new, constructive politics in the service of détente and understanding among nations. (Schweizerische Bewegung gegen die atomare Aufrüstung, 1963a)

While this frame evoked a sense of Swiss exceptionalism, Switzerland did not here appear to be detached from, but rather a part of, international society, which can function as a force for common goals.

Privileging bundles of identifications in decision-making

What is striking about Swiss decision-making with regard to nuclearization is its temporal-processual dimension of change and alteration, despite a high degree of political-institutional continuity in the relevant time frame. While the depiction of Switzerland as an armed, resistant, neutral state originally dominated the debate, this reading lost its preponderance in subsequent years. The early documents issued by the government had linked Switzerland to a stance of non-involvement and armed neutrality. According to this argumentation, the Swiss state was forced to acquire a sufficient, possibly nuclear, defence capacity in order to be able to further pursue the ‘ethically correct’ policies of a neutral state — that is, without being dominated by any other power. The acquisition of nuclear weapons was thus justified on the grounds of Switzerland’s neutral stance and its peculiar position within the international system. In later years, however, such a rationale came to be articulated less frequently. Instead, governmental documents began to be based more strongly on humanitarian identifications. In a report detailing the Swiss position towards the Partial Test Ban Treaty (PTBT), the Bundesrat (executive organ) wrote:

Given the strong and so far always disappointed aspirations of all peoples for détente and peaceful co-existence, it befits Switzerland with its humanitarian tradition to welcome every concrete step towards a reduction of the menace of total war and to support such steps with the modest means it commands. (Bundesrat, 1963: 620)

Acceding to the treaty was thus interpreted not only as a means of implementing a specific policy preference, but also as an enactment of a ‘Swiss humanitarian identity’. In other words, rather than causing a shift in policy, this particular identity interpretation was enacted through foreign policy.

A number of further events and developments fed into this change of policy that we cannot discuss here in detail. One of these concerns what came to be known as the Mirage Affair of 1964, when it was revealed that the defence department’s procurement of fighter airplanes massively overran the budget. In addition to causing far-reaching changes at the military institutional and political level, the scandal also triggered a fundamental transformation in the civil–military relationship and seriously eroded voters’ faith in the military. The deeply engrained trust in the Swiss military and its widespread role in politics and everyday life was contested and supplemented by a more critical reading that emphasized the dangers inherent in a culture of exalted militarization and expanding armament.

Another, broader process at the international level concerned the rise of a global non-proliferation norm. While the government had previously tried to systematically counter the peace movement’s objections to the nuclear weapons option — for example, by downplaying the hazards triggered by radioactivity — it now acknowledged the markedly anti-nuclear arguments in the process of signing the PTBT. The potential detrimental effects of nuclear weapons for both the environment and human health became increasingly recognized by the broader public. The demand for nuclearization subsequently lost much of its appeal, for domestic as well as international reasons, while a narrative of the neutral’s humanitarian obligations gained hold in governmental documents.

This change also became tangible in key strategic and security documents issued in the late 1960s. The Defence Posture published in 1966, for example, renounced much of what had been argued to be militarily and technologically feasible in the previous years. This signalled not only a shift in military strategy, but also an alteration of dominant identifications. Rather than upholding identifications that were linked to technological optimism and militaristic omnipotence, the administration gradually moved away from pursuing grand procurement projects, arguing that ‘it is inevitable to adjust military planning more strongly to the needs of the civilian population’ (Bundesrat, 1966: 870).

Moreover, in relational terms, we see a re-examination of the state’s positioning towards the international system — entailing a revised role both vis-a-vis the superpowers and other middle powers. Easing Cold War tensions led several actors within the Swiss administration to emphasize the country’s opportunity to actively function as a ‘moral institution’ and a ‘bridge between east and west, north and south’ (Interpellation Furgler, 1965: 551). Some went on to argue that in light of superpower ‘nuclear complicity’, Switzerland — ‘a small and neutral state particularly committed to a peace policy’ — should pursue active collaboration with other ‘like-minded’ and ‘highly developed industrialized states’ in order to influence the elaboration of a non-proliferation treaty (Interpellation Binder, 1967: 594). Also, with regard to other neutrals such as Sweden and Austria, Switzerland now competed for the role of international broker and proponent of humanitarianism. ⁵ Thus, the previous notion of an abstaining or even antagonistic actor made way for a concept of proactive involvement. This discursive shift signalled a larger transformation in the argumentative pattern eventually leading to the official renunciation of the weapons option in 1969 (Wollenmann, 2004).

In sum, the analysis indicates that for more than a decade after the end of the Second World War, multiple conflicting identifications were linked to neutrality (being itself virtually an empty concept), each entailing different political outlooks and different foreign and security policy agendas (Hug, 1999: 75–79). It was only during the course of the 1960s that the predominant bundles of identifications changed from emphasizing the country’s ‘resistant neutrality’ to primarily revolving around notions of ‘humanitarian neutrality’. Thus, the narrative that became dominant in the mid-1960s stressed Switzerland’s obligation to actively promote peace and stability for humanitarian reasons.

Switzerland’s long-ambivalent nuclear attitude highlights contestation and makes tangible that specific bundles of identifications were temporarily and incompletely privileged in the process of deciding against nuclear armament. What one might call the ‘humanitarian narrative of Swiss identity’ became temporarily dominant in the political decision-making process, yet it did not precede the decision to forgo nuclear armament. Likewise, it would be misleading to treat competing identifications as a merely epiphenomenal by-product of politics. They are integral parts of policymaking since identifications underlie justifications of policy. While we suggest that the decision to forgo nuclear armament privileged the ‘humanitarian’ bundle of identifications, we are not arguing that identification practices were thereby arrested in a final sense. Quite to the contrary, the humanitarian identifications that emerged in the nuclearization discourse would not determine Swiss behaviour in the major political debates that followed. Rather than preceding policy, ‘what it meant to be Swiss’ was again rearticulated, renegotiated and re-enacted. This is exemplified by a second — brief — episode: Switzerland’s two referendums on UN membership in 1986 and 2002, which related the country towards international society.

Switzerland and the UN

In 1986, neutrality again became central to identification practices around which discussions converged. However, like before, the question of what neutrality actually signalled was highly disputed. As the notion of neutrality as non-involvement in international affairs (and not as humanitarian obligation) set the terms of the debate in the mid-1980s, the burden of proof was on those who argued that Switzerland should join the UN (Bundesrat, 1982). Especially, the prospect of having to engage internationally in cases that would not find domestic support was viewed with scepticism across the electorate (Schweizerische Gesellschaft für Praktische Sozialforschung, 1990: 18). This was widely considered to be incompatible with enduring and armed neutrality (Moos, 2001). Switzerland was depicted as deriving its legitimacy from its specific form of political organization — and not from being part of a society of states. As a result, Swiss sovereignty, independence and neutrality were given ontological primacy over sovereignty, independence and neutrality as institutions of international society. While the identification of ‘Swissness’ as being exceptional and outside of international society (Einzigartigkeitsdiskurs) became dominant (cf. Kopp, 1985), it clearly contrasted with the discourse in the latter phase of the debate over acquiring nuclear weapons.

While these identifications were privileged in rejecting UN membership in 1986, a very different bundle of identifications (although also linked to neutrality) began to move centre stage prior to Switzerland’s second UN referendum in 2002. At this point in time, it was successfully argued that the UN no longer represented specific interests, but had become an embodiment of international society as a whole, seeking to uphold international law. States in opposition to the UN were depicted as being outside of the law. This, in turn, gave rise to the notion that neutrality (as exceptionalism) became untenable since there could not be neutrality between lawbreakers and legitimate law enforcers. Second, it was argued that Swiss neutrality would not be negatively affected by Security Council decisions because the UN was no longer viewed as a warring party (Möckli, 2000b; Bundesrat, 2001: 1214–1215). Rather than linking neutrality to unwelcomed entanglement, neutrality was now associated with international solidarity and with actively shaping international relations. It was, indeed, argued that the guiding principles of Swiss foreign policy were practically identical to the goals formulated in the UN charter (Bundesrat, 2001: 1184). Hence, neutrality was no longer viewed as something that primarily originated in Swiss domestic politics. Instead, it was argued that neutrality could no longer be exercised outside of international society. Being neutral was now interpreted as the obligation to be a responsible part of the whole of international society. This again markedly changed the set of privileged identifications and altered how Switzerland was situated figurationally, that is, in relation to other states and international society.

The UN membership debates in the 1980s and early 2000s briefly illustrate that although bundles of identifications become privileged, they do not remain stable, so as to inform future decisions. Both cases suggest that ‘identity’ cannot be argued to precede action. More broadly speaking, the recurring theme of neutrality and the ways in which neutrality was continually renegotiated underscores that change and continuity are tied together in identification practices. Conceptualizing identity in terms of a substance consequently runs the risk of neglecting this relational process.

Conclusions

Despite continuous attempts to move beyond rationalist and materialist theory, identity research remains wedded to essentialist thinking and conceptualizes identity as a unified, substantivist property of the individual social entity. While some authors acknowledge that identities shift and alter or that actors can have several identities simultaneously, even post-structuralist scholarship ultimately reduces identity to a substratum that informs action. The argument that we have presented here, in contrast, suspends substantialist ontological assumptions and argues that processes and relations are the basic components for studying identification processes, rather than ‘identities’. Instead of searching for a core ‘identity’, we examine how actors articulate, bundle and (temporarily and incompletely) stabilize interpretations of the ‘self’, thereby privileging some identifications over others. We therefore build on, but move beyond, the claim that identities are polymorphic, contested and discursively constructed, and turn to the very processes of identification in historically specific, contingent relations. By doing so, the approach conceptually underscores agency, and grasps that continuity and change are not opposed in these dynamics. At base, we aim to connect ‘shifting articulations of the self’ to the development of broader social arrangements. Such a theoretical framework seeks to overcome both the barrier between individual and society and the barrier between domestic and foreign policy that characterize contemporary identity scholarship.

Our study of identification practices in Switzerland has a number of broader implications. For one, a p-rt approach denaturalizes notions of ‘identity as a cause’ empirically. This makes it possible to question taken-for-granted assumptions in cases in which the constitutive relatedness of identifications is not commonly acknowledged. This specifically concerns cases of radical othering in which actors are depicted as being outside of international society, or as being illegitimate interlocutors per se (Crawford, 2002: 424–425). Take, for example, the notion of rogue states or pariahs in IR. Here, p-rt can uncover the constitutive relations among rogues and non-rogues by empirically engaging acts of identification in policy formulations and how these lead to highly antagonistic self–other relations (Bucher, 2015). At a policy level, our approach opens new avenues along which to think about deviance in IR more generally (Wagner et al., 2014). P-rt models can also add to the existing literature on security alliances and institutional choice, which continues to treat identity as a causal variable that originates in a purely domestic context (Hemmer and Katzenstein, 2002: 593). While these approaches argue that identities explain choice, our model highlights that ‘identities’ should not be taken to inform decision-making. Rather, focusing on identification practices allows us to grasp identification dynamics that emerge in institutional practices of policymaking.

On a more general level, our model also helps to avoid state centrism by conceptually allowing for non-state relations to play a (potential) role in the processes in which bundles of identifications are privileged. The degree to which non-state actors will serve as relevant others in the future is a fascinating empirical question. General shifts in regard to the salience of non-state actors in identification processes would arguably signal changes in the identity–legitimacy nexus underlying notions of ‘we-ness’ in international relations.

In our view, (meta-)theoretical considerations should ultimately be geared towards providing tools that make a practical contribution to the way we study international relations. We hope that leaving behind essentialist and individualist notions of core identity in order to study identifications from a processual-relational perspective is a move in this direction and will generate policy-relevant empirical studies that add to our understanding of international dynamics.