Critical discourse analysis in analysing European Union foreign policy: Prospects and challenges

Introduction

Critical theorising in international relations that assumes a role for language in the construction of social reality has taken varying views on the types of methodology to be employed in the study of foreign policy. Some approaches, mostly among poststructuralists, have refused the employment of a rigorous and systemic methodology on the grounds that it would run contrary to challenging rationalist and positivist approaches to international relations that are perceived to be obsessed with methodology (Milliken, 1999: 26–27). Others have argued that attention to method is both possible and crucial for critical approaches, and that it would not necessarily imply the sharing of common ontological and epistemological grounds with rationalist studies (Hansen, 2006; Laffey and Weldes, 2004: 28–30). This article takes the second road and argues that a discursive approach to social reality does not necessarily require the refutation of methodological tools. In fact, this study embraces the understanding of discourse analysis as ‘the retroduction of a discourse through the empirical analysis of its realisation in practices’ (Laffey and Weldes, 2004: 28). In turn, specific methods of discourse analysis offer key analytical tools for making this retroduction possible.

At a general level, studies employing discourse analysis in foreign policy have opted to use the Foucault and Derrida-inspired poststructuralist tradition, adopting genealogy, deconstruction or analyses of the articulation of key foreign policy notions such as the ‘state’ and the ‘nation’ as macro methods of approaching texts, or have focused on the systems of signification (i.e., predicates, metaphors) that are utilised in texts in referring to key selected subjects. This has also largely been the case for discursive studies of EU foreign policy. For instance, Neumann (1998, 1999) has employed the genealogical method in analysing representations of Europe in its historicity with respect to its constituting Others such as Russia and Turkey. In a similar vein, Rogers (2009) has taken up the analytical tools of discourse theory to explicate the shifting discourse in EU foreign policy, from Europe’s representation as a ‘civilian power’ since the 1970s to its representation as a ‘global power’ from the end of the 1990s onwards. The Copenhagen School (albeit differing from poststructuralism in its intention to explain foreign policy actions) has analysed the articulations of the concepts of ‘state’, ‘nation’ and ‘Europe’ to explain nation-state policies towards Europe, as well as how they can shape EU foreign policy (Larsen, 1997a, 1997b; Wæver, 2005).

Among other approaches that rest more on the operational tools of linguistics in EU foreign policy, Hülsse (2006), for example, has chosen to focus on a specific type of signification, namely metaphors, used in EU enlargement discourse in constructing European identity. This was in the tradition of earlier works such as that of Chilton and Ilyin (1993), who used metaphor analysis in studying the articulation of the concept of the ‘common European house’ in several European states towards the end of the Cold War. Much of the aforementioned work has been grounded in the poststructuralist theorising of international relations in which discourse is conceptualised as constitutive of social reality, where there exists no separation between the discursive and the non-discursive realms. Thus, the major goal (with the exception of the Copenhagen School) is not to explain EU foreign policy, but to demonstrate the means through which it is being discursively constructed. Other than these, there have also been studies branded as discursive in EU foreign policy in which speech act theory has been used to explain the causes of EU decisions from a social constructivist framework. These works have largely studied EU enlargement, with a focus on the driving factors of enlargement policy (Schimmelfennig, 2003).

Discourse-analytical approaches to EU foreign policy have been particularly valuable in shedding light on the identities and subjects constructed through EU foreign policy discourse, albeit with certain shortcomings. For instance, while discursive methods focusing on systems of signification are particularly useful in tracing how subject identities are constructed through discourse, they do not sufficiently address the question of how discourses are naturalised in texts by marginalisation of alternative interpretations or, for that reason, have a substantial impact on general debates on EU foreign policy (Wæver, 2009: 167). Poststructuralist discourse theory is especially useful in showing the dominant representations of the social world, as well as its alternative interpretations. However, in studies employing this method, often very little attention is paid to the ‘linguistic’ dimension through which subject identities are created. Hence an analysis of systems of signification such as predicates and metaphors is sidelined in order to focus on broader representational practices consisting mainly of ‘events’ and ‘actions’. One may be inclined to argue that a combination of these approaches with systematic analyses of systems of signification such as predicates and metaphors may remedy this lack of focus on the ‘linguistic’. However, there are multiple linguistic mechanisms through which one can have an improved understanding of the means used in discourse to create subject identities, as well as a more in-depth view of the types of subject identities constructed.

For example, an investigation of ‘who speaks’, and thus representations of Selves, requires attention to the ways in which ‘we’s are used in discourse. Similarly, personifications, euphemisms and similes serve important purposes in creating subject identities. Linguistic tools aimed at observing framing and perspectivation can also tell us a lot about how Selves and Others in discourse are positioned with respect to particular events and other utterances. Nominalisation, agent deletion and footing are only a few means through which one can view the involvement or detachment of discursive actors in various representational practices. Linguistic tools, as such, are not only valuable for delving into subject identities; various argumentation patterns drawn from critical linguistics can also prove helpful in enhancing our understanding of the play of practice in discourses, especially in demonstrating the means through which certain representations of the world are deemed ‘natural’ and ‘uncontested’. Reliance on inferences from implicit assumptions, assumed continuity in historical tropes and the usage of numbers in particular ways are just a few among many argumentative moves that ‘normalise’ certain representations in discourse.

This article suggests that Critical Discourse Analysis (CDA) could provide a valuable approach that would combine the macro and micro analyses of texts in the context of EU foreign policy by resorting to refined linguistic and argumentative tools (Wæver, 2009: 167). In doing that, it takes a broader methodological view than the other articles of the volume (see Carta and Morin, 2013) by presenting an approach that can be utilised in both the EU’s institutionalist foreign policy discourse and member states’ foreign policy discourses, as well as incorporating various discursive strategies and linguistic tools that can also be separately taken up in discursive studies of EU foreign policy (see Carta, 2013; Larsen, 2013). It first outlines the main theoretical premises of CDA and its one particular variant, the discourse-historical approach, followed by a discussion on its analytical and methodological toolkit and its potential for the study of EU foreign policy. The article concludes with the theoretical challenges posed by CDA approaches.

Critical discourse analysis

CDA does not merely comprise a methodological toolkit. In fact, CDA provides both theories and methods for the empirical study of the relations between discourse and social and cultural developments in different social domains (see Carta and Morin, 2013). The influence of the Frankfurt School is particularly significant in the way in which the analysis has adopted Habermas’ notion that critical science has to be self-reflective. Such a theoretical standing leads to a focus on the role of language in power relations, processes of exclusion, inequality and identity building in works that place themselves under the CDA umbrella. CDA approaches in general view discursive practices as an important form of social practice which contributes to the constitution of the social world, including social identities and social relations. The aim is thus to shed light on the discursive dimensions of social and cultural phenomena and processes of change. While doing this, CDA approaches make a distinction between the discursive and the non-discursive in social life, albeit accounting for exchange between the two realms.

Within this framework, CDA seeks to investigate ‘opaque relationships of causality and determination between discursive practices, events and texts and broader social and cultural structures’ with the aim of disclosing the role of discursive practice in the maintenance of the social world, specifically with respect to power relations (Fairclough, 1995: 132). This leads to the emergence of an ‘emancipatory’ mission in CDA for radical social change geared towards empowering oppressed groups. Hence it is common for CDA researchers to make their own positions on a certain subject explicit while also trying to retain self-reflectivity during the course of the research.

These theoretical assumptions behind much CDA work immediately point to certain divergences with the poststructuralist notion of discourse. CDA’s theoretical premises render it closer to social constructivism in the sense that CDA views language as more than just a mirror of reality, by also accounting for non-discursive practices that help constitute social reality. Poststructuralist discourse analyses do not conceptualise such a distinction between the discursive and non-discursive realms of social life. Hence they do not share the goal of emancipatory critique in CDA, which involves the comparison of various representations with an implicit version of the way things really are (or should be). Nonetheless, there is also important common ground that is occupied by both approaches. For instance, poststructuralist discourse analysis broadly shares CDA’s concerns for a critical approach to taken-for-granted knowledge, the historical and cultural specificity of discourse and the role of social interaction in the construction of the world. It also acknowledges that no social scientific work can aspire to a fully bias-free analysis. In the words of Hansen,

post-structuralism’s critical political understanding of discourse implies, first, that as there is no place outside of language, there is no analytical place that does not make a political incision, and, second, that as there is no place outside of language, there is no analysis that can completely dispense with the vocabulary already in place (Hansen, 2006: 213).

Although rigid emancipatory foundations are not established as bases in the interpretation of discourses, this should not come to mean that – as in the well-known phrase utilised in the critique of poststructuralism – ‘anything goes’ in poststructuralist interpretation of texts. It has for example been argued that a commitment to heterogeneity, plural and non-nationalist conceptions of political community or the principle of not inducing harm in foreign policy can serve as yardsticks in the interpretations of texts upon the condition that self-reflexivity (of the researcher’s own normative assumptions) is constantly present throughout the analysis (Campbell, 1998; Diez, 2013).

Methodologically speaking, it can be argued that certain variants of CDA, particularly in relation to the linguistic and argumentative tools that they employ, can also be utilised in poststructuralist studies. For instance, Torfing (2005: 9) has argued that many of CDA’s ‘analytical notions and categories for analysing concrete discourse and distinguishing between different types and genres of discourse can be used in conjunction with concepts from poststructuralist discourse theories’, and Larsen (IN PRESS) demonstrates how a specific linguistic tool of CDA (namely the use of the ‘we’ pronoun) can be employed in a poststructuralist study of nation-state foreign policies. This article focuses on a major variant of CDA, namely the discourse-historical approach (DHA), as a wider methodological tool that can be utilised in both social constructivist and poststructuralist works on EU foreign policy, particularly in studies that aim to explore the various constructions of European identity through EU foreign policy and/or the EU foreign policy implications of the articulations of different Europe(s) in discourse at the level of the member states or the EU.

DHA is a type of CDA that is particularly distinguishable by its specific emphasis on identity construction, where the discursive construction of ‘us’ and ‘them’ is viewed as the basic fundament of discourses of identity and difference (Wodak, 2001: 73). This approach has been used in the analysis of national identities (Wodak et al., 2009) and has more recently been utilised in analysing the construction of European identities (Krzyzanowski, 2010; Krzyzanowski et al., 2009; Krzyzanowski and Oberhuber, 2007; Wodak, 2009). In fact, DHA is the only strand of CDA that has so far, albeit on a limited scale, been used in European integration studies.

This research tradition employs the principle of triangulation, which refers to the endeavour to work interdisciplinarily, multi-methodically and on the basis of a variety of different empirical data as well as background information (Reisigl and Wodak, 2001: 35). In addition to providing an analytical toolkit for the analysis of texts, DHA incorporates the central concept of intertextuality in the analysis. However, it does not treat intertextuality in the broader sense of the term, as observed in some poststructuralist analyses in which the concept accounts for the linkage between texts as well as between discourses (Hansen, 2006). Instead, it differentiates between intertextuality and interdiscursivity. In more concrete terms, intertextuality is used to refer to the ways in which a text draws explicitly or implicitly from other texts in the past or present ‘through continued reference to a topic or main actors; through reference to the same events; or by the transfer of main arguments from one text into the next’, while interdiscursivity accounts for the ways in which discourses are connected to and draw from one another (Wodak, 2007: 206). This is based on the conceptualisation of discourse as ‘patterns and commonalities of knowledge and structures’, where a text refers to a ‘specific and unique realisation of discourse’ through various genres (Wodak, 2007: 207).

The analytical apparatus of DHA consists of three main steps. The first step involves outlining the main content of the themes and discourses, namely the discourse topics in the narrative on a given subject (Van Dijk, 1984: 56). The second step involves the exploration of discursive strategies deployed in the narrative to answer the following empirical questions directed at the texts (Reisigl and Wodak, 2001: 44): How are the chosen subjects named and referred to linguistically? What traits, characteristics, qualities and features are attributed to them? By means of what arguments and argumentation schemes are certain representations of the subjects justified, legitimised and naturalised in discourse? From what perspective are these nominations, attributions and arguments expressed? Are the respective utterances intensified or mitigated? These questions all relate closely to how various ‘we’s are constructed and naturalised in discourse. In discourse-historical works of CDA, the totality of discursive practices that undergo analysis to answer these empirical questions are referred to as discursive strategies. This step requires a particular emphasis primarily on referential/nomination strategies and predication in responding to the first two empirical questions, a closer look at argumentation strategies in the case of the third empirical question and a focus on strategies of perspectivation and intensification/mitigation in the case of the fourth and fifth questions, respectively. The third step of analysis explores the linguistic means that are used to realise these discursive strategies.

Referential/nomination strategies can use various linguistic means, such as the use of tropes, substitutions, certain metaphors and metonymies, with the effect of creating ingroups and outgroups in discourse. For example, uses of ‘we’ and ‘they’, and metaphors such as ‘family’ or ‘home’, can be cited among the many linguistic means that involve referencing. They are very closely linked with the strategy of predication, which is the process and result of linguistically assigning qualities to subjects. This can be realised through attributes, collocations, predicative nouns/adjectives and various other rhetorical figures. For example, rhetorical devices such as flag words and stigma words can be considered as implicit predicates in discourse. While flag words such as multiculturalism, integration, freedom and democracy have positive connotations, stigma words such as racism and anti-semitism carry negative associations.

Argumentation strategies used in justifying attributions can take various forms. Among the most common is the employment of topos, defined as ‘parts of argumentation which belong to the obligatory, either explicit or inferable premises in the shape of content-related warrants that connect the arguments with the conclusion’ (Reisigl and Wodak, 2001: 74). For example, in the discursive construction of national identities, one often encounters the topos of culture and history. Another frequently used topos in the discursive construction of national identities is the topos of threat, which implies that if a certain course of action entails dangerous consequences, one should refrain from taking it, or that if certain threats are present, one should take the necessary precautions (Reisigl and Wodak, 2001: 77).

The major strength of DHA is the way in which this three-step analysis allows for the integration of both macro and micro analysis of texts into a discursive study. While the identification of discourse topics provides an overview of the main patterns and structures of discourses on a given topic, the focus on discursive strategies and linguistic devices entails the use of rigorous discursive/linguistic tools in displaying the ways in which certain representations of the social world, including its constituent subjects, acts, events, policies and processes, are naturalised and justified through discourses. The next section will attempt to explore the potentials of this approach for the study of EU foreign policy.

DHA and EU foreign policy

EU foreign policy is often taken to denote the foreign policy of the EU as a whole, rather than the separate foreign policies of the EU member states. Nonetheless, member states’ policies towards Europe are also relevant insofar as they have repercussions on the formulation of foreign policy at the EU level (Larsen, 2004: 63). In that sense, DHA can be used in both the analysis of EU institutionalist discourse and the foreign policy discourses of the EU member states. Debates on EU foreign policy intensified after the signing of the Lisbon Treaty, upon which the EU established an external action service and appointed a High Representative for EU Foreign Affairs and Security Policy, as well as a President of the European Council. Furthermore, the rising challenges in its southern neighbourhood, with the Arab Spring movements, have brought the issue of the EU’s external power further into the spotlight.

Texts on EU foreign policy that can be subject to DHA can be drawn from a variety of genres, such as parliamentary debates (pertaining to the European Parliament or the national parliaments of EU member states), official declarations/foreign policy documents (e.g. Declarations by the High Representative, European Council decisions, EU Presidency reports), political speeches (e.g. by EU leaders, Commission officials, the High Representative) and interviews (e.g. with the Commission’s civil servants in the external action service or member state bureaucrats). In Lene Hansen’s (2006: 85) typology of genre in discourse analysis, set on the three criteria of articulation of identity/policy, the degree of formal authority and the extent to which the text is read and attended to, parliamentary debates are classified as a type of genre that articulates both identities and policies and that carries high formal authority due to the elected nature of the politicians and the existence of an electoral platform and a constituency. Although parliamentary debates are not widely read and attended to, it is in fact the case that politicians are in constant interaction with society via various means such as the media and pressure groups, leading to the constant (re)articulation of their discourses in various settings where exposure to a wider audience is possible. Official declarations/foreign policy documents carry high formal authority and can be widely read and attended to, yet they can also score low in terms of articulating identity since they are the end products of negotiations involving various actors (particularly so in the case of the EU), and thus need to be combined with other genres to have a better sense of the ‘full discourse’.

Unlike parliamentary debates and official declarations/foreign policy documents, political speeches meet all three of Hansen’s criteria by entailing high political authority, articulating both identities and policies and reaching a wide audience (Hansen, 2006: 82–87). In the case of the European Commission, the speeches can also be characterised as a specific type of a new sub-genre of political speech – what Wodak and Weiss (2004: 235–242) refer to as ‘visionary/speculative speeches’ on Europe. In line with the distinguishing features of this genre, they are in general consensus-oriented, with a high reliance on argumentative strategies geared towards ‘making meaning of Europe’ (‘idea, essence, substance’), ‘organising Europe’ (‘institutional forms of decision making and political framework’) and ‘drawing borders’ (inside/outside distinction), where the interaction of these three dimensions forms the basis of the talk.

While qualitative interviews are rare in discursive works (particularly of the poststructuralist type), they are widely employed in DHA and pose particular advantages in discourse research that are not made available by other genres. The narratives and orientations of speakers are most often best revealed in interview data (Howarth, 2005: 338). This is largely due to two main factors. First, interviewees often enact their identities through recounting their experiences to the interviewer (Wagner and Wodak, 2006). Second, the genre’s dialogic nature allows moving beyond a specific utterance of the respondent towards an extended narrative that sheds light on patterns of (constructed) identities. Nonetheless, it is also this dialogical nature that endows the interviewee with the role of producing the discourse through interaction with the respondent. The principle of triangulation in DHA, by requiring the combination of different genres in the analysis, thus aids in countering the subjectivity of the interviews by the higher degree of formality in debates, speeches and/or official legal/policy texts. This also provides for double-checking the (ir)regularities across discourses in a way in which one can see whether or not similar discursive patterns can be discerned across different realms through which discourses are (re)produced, or whether alternative constructions occur in more unofficial, private and flexible settings.

In terms of the three-step methodological toolkit of DHA, after specific discourse topics on selected texts pertaining to EU foreign policy are identified, undertaking an analysis of the discursive strategies that utilise specific linguistic devices can provide substantial insight in responding to the following three key questions on EU foreign policy identified by Larsen (2004: 68) as among the main empirical questions on EU foreign policy from a discursive perspective: Is the EU constructed as an international actor? If it is, what kind of an actor is constructed? What kind of values is this actor based on? The empirical questions above relate closely to the discursive construction of European identity – through EU foreign policy – which overlaps with the central concern of identity construction in most studies that employ DHA. In a similar vein, these questions also entail responding primarily to the first three empirical questions identified earlier by DHA on the nomination and the predication of chosen subjects; the traits, characteristics, qualities and features that are attributed to them; and the argumentation schemes through which representations of the subjects are justified and naturalised in discourse.

Analysis of nomination and referential strategies in text can prove to be specifically useful in observing whether the EU is discursively constructed as an international actor. For instance, the various instances of the deictic expression ‘we’, which is used to indicate sameness, can provide insight regarding the actor-based content of a discursively constructed Europe. The use of the ‘we’ in EU foreign policy texts can show where Europe stands with respect to individual member state(s) and can point to the boundaries of the discursively articulated Europe in relation to its various geographic Others, as well as to other discursively constructed international entities such as the ‘West’. Similarly, the referential strategy of anthromorphisation (personification), which entails the attribution of human qualities to the entity in question, can play a decisive role in animating imagined ‘collective subjects’ such as Europe, the frequent usage of which can signify a strong international ‘actorness’ on the part of the EU. Other metaphorical expressions used in denoting Europe can also be illustrative of the actorness that is ascribed to the EU (see also Carta, 2013). For example, the use of the ‘magnet’ metaphor in situating Europe vis-à-vis its wider neighbourhood can construct it as a civilian/normative international actor with the capacity to impact on the domestic governance of non-member countries through civilian/normative means.

An analysis of predication and argumentation strategies, however, is better suited to identifying the ‘type’ of foreign policy actor that the EU is, as well as the values that it is based on. For instance, the repeated predication of the EU as an upholder of democratic values, principles and standards found in textual collocation with other predicates of mild mechanisms of change such as ‘influence’, ‘propagate’, ‘pass on’ and/or ‘peaceful means’ can be considered to help constitute the Normative Power Europe discourse, through which Europe is constructed as a normative power that is capable of attaining democratic change in countries through the spread of European democratic norms (Manners, 2002). As opposed to this, one might possibly find in texts predications of Europe as a more interventionist ‘global power’ or a ‘global player’ on a par with its global competitors such as the USA and Russia, the rising frequency of which could then lend support to the argument that the discourse on Europe as a ‘global power’ is replacing the one on Europe as a civilian/normative power (Rogers, 2009).

Argumentation strategies (in terms of the justifications of positive and negative attributions) used in DHA could hereby prove useful in the further scrutiny of such predications. Diez (2005), for instance, warned that the very danger inherent in the Normative Power Europe discourse is a potential lack of self-reflexivity, the presence or absence of which can only be substantiated through systematic discourse analysis. A closer look at the argumentation strategies of DHA can provide insight into the presence of claims to universality, whereby democratic values that are defined as particular to the EU are also branded as universal, which would in turn discursively reinforce the superiority of Europe/the West against the rest of the world and fortify the borders between the two. Chouliaraki (2005: 6) defines this argumentation strategy as the ‘topos of orientalisation’, where the equation of European/Western values with universal values leads to the ‘annihilation of the cultural weight of Other(s)’, reproducing the Eurocentric manner in which the superior European Self relates to its Others. Similarly, identifying the use of the topos of history, geography and culture in establishing essentialist boundaries between ‘Europe’ and its outside in texts would also imply that the self-reflexivity required by the Normative Power Europe discourse is being undermined.

In a related fashion, the argumentation strategy of securitisation, if found regularly in texts, can also be used to argue for the erosion of the (discursively constructed) normative bases of the EU. Securitisation refers to the ways in which security discourse utilised in justifying actions contributes to the naturalisation of a given community by constructing threats to its existence. Thus the strategy of securitisation lends a legitimising and authoritative dimension to border drawing and identity construction. This has been widely documented in the case of the nation-state, where the security discourse contributes to the naturalisation of the national community by constructing existential threats to the state (including government, territory and society) (Buzan et al., 1998: 21). Security threats are not only perceived as ‘potentially undermining the state’, but in fact constitute the state itself by posing fatal risks to certain invoked ‘national interests’ (Hansen, 2006: 34). In the case of the EU, construction of migration and human trafficking as security threats emanating from the EU’s Southern neighbourhood can aid in construing Europe as a ‘deeper and more tightly unified geopolitical space’ (Rogers, 2009: 846). Diez (2006), for instance, has argued that the EU enlargement policy creates a paradox in the sense that it creates new borders at the EU’s new outer edges that are securitised through references to the smuggling of goods, trafficking of people, illegal immigration or terrorism. Pace (2006) has similarly demonstrated how the Mediterranean region has been discursively constructed and securitised by the EU. Yet, boundary drawing as such does not just contribute to the construction of Europe as a geopolitical space that is securely protected from violence; it in fact constructs a form of violence itself via the ascription of a homogenous identity with respect to both the stable and peaceful inside and the crisis-ridden outside of Europe.

Securitisation need not necessarily be confined to the EU’s outside, but can also concern its inside regarding, for instance, migrant populations living in the EU. In this discourse on ‘societal security’ (Wæver, 1995), ‘threats are less likely to be associated with aggression from other states, with violations of state sovereignty, but instead with challenges to society, and in particular, social, cultural and national identity’ (Walters and Haahr, 2005: 96). In this framework, securitisation of not only migration but also drugs and organised crime can be considered as a part of the discourse on ‘societal security’ (Huysmans, 2000, 2006). It needs to be highlighted that this is an area in which DHA probably offers the most advanced analytical toolkit in discourse studies, where a wide range of discursive strategies can be learned from an ample number of works that deal with the Othering of migrants in EU countries (see, for instance, Reisigl and Wodak, 2001) and sought in EU texts that may not per se be on EU foreign policy, but can have significant implications for the construction of an international identity for the EU.

This brings us to the use of the notions of interdiscursivity and intertextuality in the application of DHA to the study of EU foreign policy. Following from the last point made above, it can be argued that using DHA in works of EU foreign policy also requires a focus on the interdiscursive and intertextual links between different texts on the object of study. For instance, a declaration on EU foreign policy can draw elements from a speech by the High Representative in an act of intertextuality. As for interdiscursivity, a discourse on EU foreign policy can also refer to topics or sub-topics of other discourses such as EU justice and home affairs.

Interdiscursivity is not only confined to the same genres of texts or texts pertaining to a single institution. An official declaration, a political speech or a parliamentary debate on EU foreign policy can also be discursively linked to academic/policy texts, such as Cooper’s (2004) book on liberal imperialism and Manners’ (2002) articles on Normative Power Europe, through the way in which Europe is accorded the role of a more interventionist global power or construed as a foreign policy actor that functions through the spread of norms and values. In a similar vein, even in situations where explicit references to Huntington (1994) or his views on the clash of civilisations are not incurred, the underlying patterns of knowledge and structures of his arguments could be visible. Such interdiscursivity might be discerned in cases where Islam and/or the Muslim world is juxtaposed against Europe and the West, in the over-encompassing role attributed to Islam in determining political and social life in Europe’s southern neighbourhood or in the presumed incompatibility between Islam and democracy.

At a broader level, interdiscursivity can be discerned with national identity constructs whereby (discursively constructed) national identities can infiltrate, for example, EU leaders’ discourses on EU foreign policy. One could also possibly discern interdiscursivity with constructed institutional identities, such as that of the European Commission which, through its discourse on EU foreign policy, could reproduce its role as a key player in managing Europe’s relations with the outside world or as a vanguard of its democratic identity. Interdiscursivity can also be found with respect to linkages with conceptual histories or certain identified discourses that extend beyond the confines of a given text, such as the neo-orientalist discourse or the modernisation paradigm.

Concluding remarks

This article began with a brief introduction to discursive approaches to EU foreign policy, followed by the presentation of DHA as a discursive approach that can be utilised in rigorous discourse analyses in this field. It has outlined the main contours of DHA as a main variant of CDA, and discussed how the main analytical tools of DHA can be put to work in studying EU foreign policy. Although the focus of the article is mainly on the opportunities that DHA offers in a field where it has been employed on a very limited scale, the challenges that it poses also need to be taken into consideration.

Probably the most important challenge in this endeavour is a theoretical one. CDA approaches (including DHA) generally take a theoretical standpoint that acknowledges the existence of other social phenomena alongside discursive phenomena, which places them in a theoretically compatible position with social constructivist approaches, yet on incompatible theoretical terms with poststructuralist perspectives. Although this chapter has argued that it is possible to employ DHA as a methodological toolkit also in poststructuralist studies, this requires an analyst to make certain qualifications at the outset of the research.

One such qualification concerns the notion of ‘history’ in the analysis of texts. DHA argues that ‘the background of the social and political fields in which discursive “events” are embedded’ needs to be integrated in the analysis. This rests on CDA’s theoretical underpinnings that conceptualise a ‘dialectical relationship between particular discursive practices and the specific fields of action (including situations, institutional frames and social structures)’ (Wodak, 2001: 66). Since poststructuralist approaches deny the existence of such a distinction between discourse and social/institutional structures, while they can present background information such as the timing and place of discourses, discursive participants and the actual material ‘events’ such as the signing of treaties, they require the contextual narratives of events to be presented in a critical light.

A second point that requires an analyst’s attention with regard to DHA’s theoretical compatibility with poststructuralist approaches concerns the treatment of the linguistic tools that the approach uses. CDA in general has been subject to criticism with regard to its Habermasian underpinnings, which are argued to lead to a conceptualisation of discourses as distortions of the way things really exist as ‘truths’, where the analyst enters the picture with the mission of revealing actors’ manipulative goals (Blommaert, 2005: 32–33). This is a point that is particularly relevant for the interpretation of discourses, where the poststructuralist is required to commit to self-reflexivity in studying discourse as a subject in its own right (see also Diez, 2013). For instance, the term ‘discursive strategy’ utilised in CDA can be criticised from a poststructuralist point of view for possibly denoting a deliberate ‘intentionality’ or ‘instrumentality’, since from a poststructuralist standpoint, it is not possible to reveal the ‘true’ motives of the actors concerned. However, the discursive tools utilised by these actors, at whichever level of cognition or with whichever aim – which is impossible for the analyst to truly discern – help to construct a discursive space that enables certain actions. In turn, identifying discursive strategies can in fact correspond to the exploration of historically contingent discursive practices through which policies are formulated. Thus, the methodological tools as such can also be adopted by poststructuralist studies as long as they are used to discern broader discursive structures and properties of discourse – the organisations of discourse that make particular talk or writing seem plausible and natural – rather than to search for the intentions, beliefs and perceptions of individual speakers and authors.