“Take it outside!” National identity contestation in the foreign policy arena

Introduction

In the early 2000s, Turkey’s foreign policy orientation swung enthusiastically toward the European Union (EU), just at the moment when a moderately Islamist party assumed power. It then turned sharply away, in favor of an orientation toward former Ottoman territories and other Muslim states. Turkey’s cooperation with Iran in attempting to broker a nuclear deal and its prolonged support of Bashar al-Assad and Muammar Gaddafi during the Arab uprisings, for example, are often assumed to be manifestations of Turkey’s increasingly Islamic political identity. While many of these arguments are less rigorous than rhetorical (for exceptions, see Fischer Onar, 2009; Yanık, 2011), such explanations gain plausibility with the surge of Islam- and Ottoman-based public discourse at both elite and society levels within Turkey. Pointing to the Islamist roots of the ruling Justice and Development Party — (Adalet ve Kalkınma Partisi (AKP)), recent identity-based arguments cite “Middle-Easternization” (Oğuzlu, 2008) or “Neo-Ottomanism” (e.g. Daloğlu, 2013) as driving Turkey’s apparent reorientation toward its Muslim neighbors. How, though, could domestic Islamism explain both the Western and the Eastern shifts?

Identity is, indeed, deeply important for understanding foreign policy, but existing accounts fail to effectively link Turkey’s domestic discursive struggles with its multiple, varied international outcomes. Static conceptions of identity such as those underlying Islam-based arguments cannot explain the initial pro-European swing followed by deeper engagement with Muslim countries all under the same AKP. Nor can monolithic explanations citing Islamic ties account for the recent conflicts between Turkey’s Sunni-led government and Alawite rule in Syria, or Iran’s Shiʿi regime. Further, the Ottoman and Islamist identities claimed to be reorienting Turkey’s policies are fundamentally at odds with the state’s own previously dominant, Western-modeled understanding of Turkishness, often referred to as “Kemalism” (Taşpınar, 2008). The now-frequent references to Turkey’s glorious Ottoman past and conservative values of Islam are antithetical to the Kemalist concept of the “model citizen”: Western, modern, and secular (Ciddi, 2009: 13–30). What accounts for these rapid shifts in the contours of Turkey’s identity debates and its foreign policy?

The answers to these questions lie in the complex role of foreign policy in domestic struggles over identity politics. Using Turkey as a case study, this article engages both International Relations (IR) and comparative politics literatures that typically do not adequately interact to tease out the complex link between national identity contestation and foreign policy. I examine contestation as a mechanism for change, accounting for the rise and fall of support for particular “proposals” (Abdelal, 2001) for national identity within states. Rather than being a phenomenon bounded within the domestic arena, I demonstrate that contestation among identity proposals both shapes and is shaped by foreign policy gambits whose domestic identity components might not be immediately observable. To elucidate this relationship, I develop a theory of identity hegemony, positing that supporters seek hegemony by spreading their own proposal across a population, but may be blocked by those who fundamentally object to their proposal for who “they” should be. Namely, I argue that elites who face obstacles or threats to their attempts to spread their own proposal for national identity across the population at home choose to take their struggle outside to the foreign policy arena. These international struggles, in turn, have distinctive, often counter-intuitive, effects on the contours of domestic identity discourse.

The analysis of identity contestation and foreign policy presented here derives from an in-depth case study of Turkey, whose national identity has long been hotly contested in scholarship as well as in practice (most recently, see Ahmad, 2014; Kedourie, 2013; White, 2013; Keyman and Gümüşçü, 2014). While empirically rich, such discussions generally lack clear delineation of what Turkish identity actually consists of (i.e. how we know it when we see it) and what specific issues are at stake in identity debates. Further, while it seems clear that some version of “Ottomania” has swept through Turkey’s population over the past five to six years, domestic politics arguments cannot account for this: the AKP came into power with an unprecedented parliamentary majority in 2002. On the foreign policy side of identity orientation, the AKP’s EU advance-and-retreat supports neither economic nor security-based arguments. EU-based reforms slowed years before the institution’s financial crisis, and in its third term, the AKP has demonstrated flagrant disregard for its alliance commitments and willingness to instigate conflicts on its own borders (Hürriyet, 2014a).

To fill these gaps, I systematically break down proposals competing to define Turkishness, and account for shifts in support for them over time. In doing so, I demonstrate how foreign policy preferences are organically, intrinsically linked with conceptions of national identity. I specify the particular red lines of contestation between Republican Nationalist and Ottoman Islamist proposals on domestic and foreign policy issues, illuminating how institutions imbued with the former proposal served as an obstacle to the spread of the latter. Finally, I show how by engaging in the EU accession process and selectively applying its democratization criteria, the AKP was able to reconfigure the domestic playing field, weakening Republican Nationalist obstacles and thus making it safer for spreading its own identity proposal back home. In brief, I demonstrate how Turkey’s EU-oriented foreign policy made the subsequent rise of Ottoman Islamism possible.

As a North Atlantic Treaty Organization (NATO) ally, EU candidate country, Organization of Islamic Cooperation member, and aspirant regional power broker, Turkey offers an excellent empirical window onto how identity debates spill over into foreign policy, but these implications extend far beyond Ankara. Some version of these dynamics pervades the foreign policy of all states; Benjamin Netanyahu’s recent address to US lawmakers, which emphasized Iran’s “oldest hatred” of “the one and only Jewish state” (The Washington Post, 2015), serves as a topical example of the foreign policy arena’s appeal to elites for advocating their version of identity given not only its timing shortly before Israel’s national elections, but also Netanyahu’s deeply ingrained perception of Iran as an existential threat that he fears not enough Israelis recognize. As explored in greater depth in the following, elites facing obstacles to spreading their proposal for identity back home can circumvent these obstacles through foreign policy in several ways; viewing foreign policy as an arena for identity contestation — rather than solely as a source or a product of identity — facilitates understanding how. In addition to its empirical insights into Turkey’s identity and foreign policy shifts, the article therefore makes two theoretical contributions by specifying the characteristics of the foreign policy sphere that make it a conducive forum for elites’ politicization of identity debates, and by conceptualizing identity red lines as the ideational basis of the obstacles driving them to do so.

Identity and foreign policy

While the “constructivist turn” in IR scholarship sought to unpack the concept of identity (Checkel, 1998), such attempts have been problematic. The use of identity as a variable has been criticized as fraught with analytical obstacles (Zehfuss, 2001), normative objections (Erskine, 2012), and “definitional anarchy” (Abdelal et al., 2009: 18). Even if all these obstacles can be overcome, one also must question whether identity is actually doing any constitutive work in the cases we seek to understand. As Hale’s (2008: 25–30) discussion of “ethnicity-as-epiphenomenal theories” explains, arguments pointing to causal mechanisms ranging from elite manipulation (e.g. Mueller, 2000; Snyder, 2000) to political action coordination (e.g. Hardin, 1995; Laitin, 1998; Posner, 2005) assert that group identities such as ethnicity have no intrinsic value, but rather serve as tools in the pursuit of (largely) material interests, such as power, resources, and security.

As an analytical heuristic, conceptually spare treatments of identity are tidy, but such approaches omit elements essential for understanding behavior, such as the ontological and emotional importance of identities and the actor-specific interests they generate. Constructivist arguments account for factors that fundamentally shape ideas about what is tolerable, desirable, and necessary — prescribing and proscribing behaviors based on understandings of identities and interests that purely rationalist arguments do not problematize. That a gun in the hand of a friend differs from a gun in the hand of an enemy is commonsensical. The explanation for this lies not in distributions of material capabilities, but in the intersubjective understandings that give meaning to relationships between actors (Wendt, 1994: 389). Rationalist arguments explicitly incorporating information variables offer much more nuanced explanations of interactions in international politics than previous theories (for a review, see Glaser, 2010), but cannot account for the shared cultural understandings that may shape states into being revisionist or security-seeking in the first place. Neoclassical realist (e.g. Kitchen, 2010), liberalist (e.g. Moravcsik, 1992), and various International Political Economy (IPE) approaches (e.g. Lake, 2009), in turn, acknowledge that domestic institutions and actors’ interests shape states’ options at the international level, but their focus on material factors occludes the powerful identity-based constraints on what domestic actors are willing to tolerate, what they desire, and what they see as necessary in terms of foreign policy.

Notable constructivist works address the challenges just enumerated while directly engaging the role of domestic identity as a constraining and constitutive element in foreign policy. Bozdağlıoğlu (2003), Hopf (2002), and Tsygankov (2014), for example, demonstrate how conceptions of identity and domestic schools of thought shape international behavior; Finnemore (1996), Johnston (2008), and Lynch (1999), in reversing the direction of argument, illustrate how international dynamics shape states’ own conceptions of themselves. Telhami and Barnett ‘s edited volume (2002) comprises an aggregated assortment of both. Building on this scholarship, this article develops a more comprehensive, constructivist account of changes in both domestic and foreign policy orientation by newly examining the foreign policy arena as a locus of national identity contestation. Synthesizing so-called “second image” and “second image reversed” approaches (Gourevitch, 1978), I fill a gap in the existing literature by identifying a specific, foreign-policy based mechanism of contestation by which the contours of such menus can be strategically altered — in brief, how levels of public support for and institutionalization of particular identities can increase and decrease via actions at the foreign policy level.

Identity hegemony theory

In accounting for the link between shifts in the contours of domestic identity debates and foreign policy orientations, I argue that the practice of foreign policy provides an arena — a term I select deliberately to convey its battleground connotation — in which elites can compete against their domestic rivals as much as their international counterparts. While important work has been done examining education systems (Kaplan, 2006), the military (Krebbs, 2006), museums (Anderson, 2006: 167–191), and other “nationalizing” institutions (Brubaker, 1996) as sites for defining and spreading a particular understanding of national identity in the domestic context, cases of how identity can be contested when groups lack access to — or are threatened with censure or extermination by — such institutions remain under-theorized. How can groups spread their proposal for identity when faced with barriers posed by a constitution, judicial system, security apparatus, or other institution that was installed specifically to prevent the promulgation of that identity, for example, racial definitions of citizenship, a court that bans ethnic parties, a military that topples religious-leaning governments? How can we account for the spread of an identity that had previously been staunchly blocked? Proposing that elites seek to circumvent such obstacles by “taking the fight outside,” this article explores foreign policy as a counter-intuitive site for the contestation of domestic identity debates. Particularly, this conceptualization challenges rationalist IR understandings which assume that states’ interests are fixed, and that foreign policy is relatively immune to the vagaries of politics — much less the politics of identity.

Also, moving beyond theories of domestic politics explaining competition over resources or rationalist electoral calculations, identity hegemony theory analyzes struggles to define and disseminate the appropriate identity for a particular social group. This focus on identity provides a deeper level of insight into the standards of prescribed and proscribed behavior for supporters of a particular identity proposal, as well as the motivations behind their efforts to spread this, in their minds, more appropriate version. The theory is grounded in traditional constructivist understandings of identity as malleable, formed through social interaction, and subject to change due to numerous actor- and structure-based factors. While not attempting to explain the initial emergence of identities, this article’s conceptualization of these understandings as proposals serves to illustrate the strategic contestation that groups of supporters engage in to persuade others to accept their proposed understanding of identity. Highlighting seemingly incompatible definitions of self across supporters of competing proposals — civic versus ethnic, capitalist versus communist, secular versus pious — reveals identity-based red lines of resistance to this contestation that can pose obstacles to their struggles.

These struggles carry immense ontological significance as groups compete against each other to delineate standards such as the boundaries of inclusion and exclusion, friends and enemies, and desired goals of the group — essentially who “we” are and how we should behave. Drawing on insights from social psychology regarding the function of social identities, identity hegemony theory assumes that supporters of a particular identity proposal — specifically, an understanding of the appropriate content of a national identity for citizens of a state — seek to increase support for their proposal by spreading its acceptance to more individuals. Social identity, defined by Tajfel (1981: 255) as “that part of an individual’s self-concept which derives from his knowledge of his membership in a social group … together with the value and emotional significance attached to that membership,” provides individuals with a mechanism for generating self-esteem by fulfilling a need not only for distinctiveness, but also for positive distinctiveness (Clunan, 2009; Welch Larson and Shevchenko, 2010). These ontological needs for group membership and positive distinctiveness serve as the “motivational underpinnings” of social identities (Huddy, 2004: 956). Individuals seek to fulfill these needs through intergroup evaluation of and competition among — treated here as processes of contestation — various social identities. National identity proposals, as social identities especially “thick” (Geertz, 1973) with historical and political meaning, particularly function to provide legitimacy and esteem to Ingroup members (supporters of the same proposal), and thus are expected to be highly contested among supporters of different proposals.

The theory also assumes that Ingroup members share a desire to realize in practice the goals envisioned by their identity. In constructivist IR terms, this is equivalent to an actor attaining the interests generated by his identity (Wendt, 1992: 398). Individuals thus believe that the more widely supported their identity proposal is, the more likely they will be able to gain access to institutions advantageous to achieving the goals (interests) that their identity prescribes. Individuals thus strive for hegemony of their proposals in order to satisfy personal needs and realize group interests.

I select the term “hegemony” here to connote the power wielded and legitimacy enjoyed by an identity proposal that is widespread, in terms of support across a population, to the point of being “relatively stable and unquestioned” (Cox, 1993: 42). A hegemonic proposal’s supporters therefore possess the capacity to dictate political and societal rules within the state, as well as the behavior of the state, in line with ideas about appropriate behavior stemming from its identity. This formulation of identity–power relations shares some assumptions with Gramsci’s (1971) theory of cultural hegemony in that identity hegemony also involves support among Ingroup members for an overarching social structure of norms and rules that shape everyday life. However, it departs significantly from Gramsci’s assumption that hegemony comes about by a passive revolution (see also Lustick, 2002; Tuğal, 2009). In contrast, the theory assumes that the struggle to achieve hegemony is often met with active and sustained resistance to an identity proposal ascending by acquiring widespread support, and that acquiescence — required to constitute a hegemonic structure of relations among actors (Cox, 1986: 219) — is nearly impossible to attain due to the ontological importance of identities to individuals, as noted earlier. Hegemony is thus a goal for which supporters of competing identity proposals strive, but may rarely achieve.

Externalizing contestation in the foreign policy arena can serve as a highly successful strategy for advancing identity proposals when there are major obstacles to these proposals back home, such as those institutions suggested earlier, because the rules and contenders in the foreign policy contest differ. First, engaging in foreign policy provides a means of circumventing opponents who vehemently defend their own understandings of identity and thus pose obstacles to the pursuit of the interests defined by a particular proposal. The ontological commitments to particular components of a national identity are much weaker for interlocutors in foreign policy because they are generally not part of that Ingroup. Second, those who do constitute part of that Ingroup but live abroad, such as diaspora communities, can serve as sources of financial support or external legitimacy for contestation back home (Han, 2013). Third, and relatedly, targeting transnational civil society as a foreign policy strategy can generate naming and shaming mechanisms (Keck and Sikkink, 1998; Klotz, 1995) that persuade domestic actors to remove identity-based obstacles, such as bans on ethnic or religious groups, in line with prevailing international norms. Fourth, international organizations can provide the resources and justification for adjusting the roles and power of domestic institutions (Checkel, 2001; Lavenex and Schimmelfennig, 2009) in line with accession criteria. Elites can strategically use these constraints to their own advantage, having conditionality at their disposal to weaken institutional obstacles to their proposal. Finally, foreign policy is treated here not solely as the realm of governmental elites, but as a forum in which leaders of an identity-based movement (e.g. minority or civil rights) may attempt to spread international awareness of their proposal when their contestation attempts are often — given their relative disadvantage vis-a-vis institutional access — blocked at home.

Having outlined the mechanism of contestation used to advance the spread of identity proposals across a population, I now present a framework for analyzing the identity content specific to each proposal. Identity content represents the “stuff” of identity — that is, the elements that supporters of an identity proposal share that form an Ingroup among them and differentiate them from other Outgroups. Difficult to determine empirically and capture analytically, identity hegemony theory utilizes a framework that breaks down identity content into four components (Abdelal et al., 2009: 17–32), thus making the nebulous concept of identity easier to grasp. First, constitutive norms provide guidelines for membership within and appropriate behavior for the Ingroup, defining who “we” are and how we should behave. Second, the social purpose component defines group interests — that is, the goals that the Ingroup believes it should achieve. The third component of relational meanings defines the Ingroup’s relation to various Outgroups; some of these relations may be friendly, while others may be hostile, fearful, and so on. Finally, the cognitive worldview component provides an overarching sense of the group’s role in the world. Collectively, these components delineate who the group is and how it should behave.

Importantly, this framework does not assume that any of these elements are fixed; rather, through processes of contestation both within an Ingroup and among an Ingroup and various Outgroups, these components can change. This framework therefore facilitates the analysis of topics such as changes in how an Ingroup views its relations with a particular Outgroup (e.g. a shift from viewing Europe as an enemy to a partner), and changes in the relative distribution of proposals across a population (e.g. a shift from high support for a “Western” proposal to high support for an “Eastern” proposal).

This framework also facilitates identifying “red lines” of seeming intolerability among proposals, indicating identity components whose prescribed behavior is not only inappropriate, but also unacceptable to supporters of a rival proposal. ¹ The qualifier “seeming” is used here to emphasize the constructed and malleable nature of identities and their constitutive components, no matter how intractable contestation over particular components may appear at a given time. Some components of identity content may be shared across proposals (two proposals with a common cognitive worldview of global norm entrepreneur) and others may differ but not necessarily conflict (e.g. Proposal A’s constitutive norm of religious freedom and Proposal B’s constitutive norm of piety), allowing for collaboration, or at least peaceful coexistence, among groups. Proposals with components that appear to be intolerable among groups (e.g. social purposes of spreading capitalism versus spreading communism), however, pose ontological threats to supporters’ sense of who they are, leading supporters to prevent such proposals from becoming more widespread. Identifying intolerable components among proposals thus provides insight into conditions under which supporters of a proposal may find their pursuit of hegemony blocked in the domestic arena, and therefore choose to pursue their identity contestation through foreign policy.

In sum, the framework developed here specifies content components of identity proposals, highlights points of seeming intolerability among proposals, and identifies a mechanism of contestation that allows for both the content and distribution of identity proposals to change.

Methodology

The first step in employing a theory of identity hegemony is being able to identify and distinguish among competing identity proposals. Using an explicit framework of identity content to delineate the components that make up each identity proposal allows for the assembly of the bits of identity “stuff” once located. The task of determining how to collect and interpret these bits, however, remains. Rather than engage in a priori, ascriptive classifications that could prejudice the findings, I employ an intertextual analysis approach that seeks to extract existing proposals for national identity content from discourse. Intertextual analysis involves inductively recovering existing identities by comparing oral, written, and image-based texts, organically constructing collective proposals that cohere around shared understandings.

The sources used to extract various identity proposals in this study include texts found in news and social media, as well as governmental archives. In line with Hopf’s (2002: 35) assumption that novels depict background daily practices in which “incidental asides to identity” are likely to be found or in whose plots overt discussion of identity may play out, ² I also drew on texts of best-selling novels. I expanded the breadth of texts to analyze the content of television series and films. Samples chosen a) had high ratings/box-office sales and b) were cited frequently in an initial survey of 175 university students, who might be more likely to access such media through online sources. I further broadened the collection of texts through semi-structured interviews, questionnaires, and participant observation. ³

From this cumulative body of material, the identity content framework outlined earlier provides a guide for making analytical sense of the discursive practices extracted from these texts. This facilitates creating composite understandings of identity that cohere around shared constitutive norms, relational meanings, social purposes, and cognitive worldviews. This framework thus functions here in reverse of the order that was originally intended by Abdelal et al. (2009) — that is, to parse out content among identity proposals that have already been determined. In contrast, these four content elements serve as a useful framework on which to drape the discursive practices observed to discern whether they do, in fact, flesh out into coherent identity proposals. Bearing in mind the complexity of identity in practice, this extraction of identity proposals does not claim to be able to capture perfectly neat boxes into which all content elements discretely and uniformly fall, even in the necessarily finite number of texts analyzed here. There is, for example, some overlap among elements of the three identity proposals that emerged from this process. The relational meaning of belonging within the Western world and the social purpose of modernization, for example, may be shared among texts that differ greatly in constitutive norms of human rights and various freedoms of expression. ⁴

Despite such factors, general patterns of correspondence among content elements clearly emerged — for example, texts containing both the constitutive norms of piety and belief in a strong ruler also tended to exhibit the social purpose of increasing space for Islam in the public sphere, the relational meaning of brotherhood and solidarity with other Muslim countries, and the cognitive worldview of Turkey as a revered cultural and military power in line with its imperial legacy. Further, after these patterns of identity content began to coalesce around four general sets of elements that can be considered as workable identity proposals — in that each contains clear and consistent content elements for all four categories — examinations of new texts conducted while adhering to the principle of ontological openness produced similar alignments of identity content.

In brief, this process of identity extraction using texts collected from multiple sources generated three main composite proposals for the content of national identity in Turkey. Western Liberalism, Republican Nationalism, and Ottoman Islamism collectively capture the most widespread proposals within Turkey’s society today. ⁵

National identity content in Turkey: Three proposals ⁶

Taking Ottoman Islamism outside

The previous section’s discussion of elements of the Ottoman Islamist proposal articulated clear interests in terms of foreign policy and domestic politics derived from its supporters’ understanding of Turkishness. Why, then, did the AKP not immediately act to pursue these interests and reorient itself toward Turkey’s former Ottoman neighborhood when taking power in 2002? Further, why did it not act quickly to expand the permissible space for Islam in the domestic public sphere once apparently gaining the institutional power to do so?

From an identity hegemony perspective, the AKP could take neither of these steps as prescribed by Ottoman Islamism because it faced serious obstacles within the domestic arena. Examining the various proposals for identity content fleshed out earlier and identifying points of intolerability facilitates understanding what constituted these obstacles. Western Liberalists could be expected to object to the advancement of any identity proposal that contains membership norms based on religion and submission to an authoritarian leader, which are directly opposed to their own constitutive norms of diverse inclusivity and the protection of liberties. Similarly, the Republican Nationalist social purpose of protecting the founding principles of Atatürk — and particularly the principle of secularism — fundamentally conflicts with the Ottoman Islamist social purpose of increasing space for Islam in the public sphere.

While both of these identity proposals contain points of intolerability vis-a-vis Ottoman Islamism, Republican Nationalism posed the greatest threat to potential AKP attempts to pursue identity hegemony upon coming to power. Republican Nationalism was, in essence, the far stronger contender in the fight. Since the founding of the Republic, Republican Nationalism had been embodied in institutions such as the Turkish Armed Forces (Türk Silahlı Kuvvetleri (TSK)), the Constitutional Court, university rectorships, and numerous media outlets, each self-tasked to uphold Atatürkist principles. Unlike relatively less widespread Western Liberalism, Republican Nationalism’s supporters therefore had institutional tools with which to obstruct attempts to advance the spread of Ottoman Islamism throughout the population and achieve the interests it prescribes. As a result, even though the AKP received enough votes to form a single-party government, the party was far from free to realize the interests generated by its identity.

The successor to a string of parties banned on charges of threatening the principle of secularism only to reopen under a new name, the AKP leadership initially acted pragmatically and cautiously. This cautiousness arguably stems from the constitutive effect that the period of the Welfare Party’s (Refah Partisi (RP), one of the AKP’s predecessors) closure had on former members. The RP, led by Necmettin Erbakan, was Turkey’s first party from the Islamist National Outlook (Milli Görüş) tradition to lead a coalition government; Erbakan became Turkey’s first Islamist prime minister in 1996. Following Erbakan’s personal appeals for jihad and the instatement of sharia law (Yalçın, 1994: 223), and his explicit rejection of a Western orientation in favor of membership in an international Islamic Union — all points of intolerability for Republican Nationalists — the TSK forced the RP from power in 1997; the party was closed by the Constitutional Court in 1998. Erdoğan arguably felt the constitutive effects of this Ottoman Islamist defeat personally: in December 1997, as an RP member and mayor of Istanbul, Erdoğan was arrested for reciting a poem claimed to incite Islamist hatred (Hurriyet Daily News, 1999).

In a 180-degree departure from the “Islamic Union” foreign policy orientation pursued by Erbakan, and seemingly in direct contradiction to the behaviors prescribed by Ottoman Islamism, the AKP immediately declared Turkey’s membership in the EU to be a primary foreign policy pillar. The AKP also began working diligently to implement political and economic reforms that were in line with the accession criteria of the EU’s Copenhagen Agreement, as if trying to prove to skeptics citing the Milli Görüş heritage of the party’s founding members that the AKP was completely different and that its intentions were genuine. Republican Nationalists and some Western Liberals doubted the AKP’s commitment to EU accession, just as they doubted its professed commitment to democratization, pointing to Erdoğan’s speeches during his time as RP mayor of Istanbul only five years earlier. His repeated statements that “for us, democracy can never be a goal,” but is merely a “vehicle” (speech in 1997; see YouTube, 2011), is exemplary of the sources of these doubts (Haber7, 2013). Other Western Liberals were willing to give the AKP the benefit of the doubt, accepting its declarations that it was a conservative democratic party rather than an Islamist party and that it was fully dedicated to achieving EU membership. The well-known journalists and public intellectuals among these were later dubbed “Yetmez Ama Evetçi’ler” — “Those who say ‘It’s not enough but yes’” to the AKP’s reforms — to indicate that they were happy with the path that the AKP was on but expected more.

Despite this support, there are many legitimate reasons to doubt the AKP’s commitment to both democratization and EU membership, which came to light in its third term (2011–2015); nearly all of the Yetmez Ama Evetçi’ler would come to criticize the AKP. A ministerial official made the disconnect between the reforms that the AKP had been pushing through and the EU accession process quite clear, stating that a constitutional amendments package had little to do with the EU: “it was already on the ministry’s desk and needed to be justified as part of the EU process.” ³¹ In the same way, the AKP became able to target obstacles to its pursuit of Ottoman Islamist hegemony by shifting the arena of contestation to the EU process.

By citing the need to demonstrate Turkey’s commitment to accession, the AKP was able to push through civil–military reform, thus taking on the most powerful obstacle to Ottoman Islamism in an arena in which the military could not compete. Through its reforms, the AKP worked to significantly weaken the military’s political role, reducing its ability to block the spread of Ottoman Islamism with threats of intervention. Pointing repeatedly to reforms necessitated by the EU’s Copenhagen Agreement, the AKP institutionalized civilian authority over the military, removing its “special status” by legislating nine harmonization laws between 2002 and 2004 (Michaud-Emin, 2007). In its 2004 Regular Report (European Commission, 2004: 22), the EU noted that Turkey had “shown great determination in accelerating the pace of reforms,” showing enthusiastic approval of reforms targeted toward “civilian control of the military.” The most effective step toward reducing threats from an autonomous, staunchly Republican Nationalist institution was taking control of the National Security Council (Milli Güvenlik Kurulu (MGK)), the body that forced the RP to step down in 1997 (Eldem, 2013). The influence of this previously powerful body was constricted through EU-mandated reforms to making recommendations to the Council of Ministers solely in a consultative function. Before these reforms, Article 118 of the 1982 Constitution had stipulated that the government would give priority (öncelik) to decisions made by the MGK. ³² Further, the MGK was removed as a member of the Council on Higher Education (Yükseköğretim Kurulu (YÖK)) and the Higher Council of Radio and Television (Radyo ve Televizyon Üst Kurulu (RTÜK)), shrinking the power it wielded over university life and curricula, and the nature of content broadcast by the media, respectively. In criticizing the impact that these changes would have on the military’s influence, a blogger warned that the “army was being liquidated along the path to EU democracy” (TürkSolu.org, 2003).

The 2007 EU Progress Report (European Commission, 2007) — published at the beginning of the AKP’s second term in power — praised the MGK’s “new role,” the drastic reduction in the overall size of the MGK and the halving of the number of military personnel on the committee. Tellingly, those areas in which the report criticized a lack of progress, including civilian control over the Gendarmerie and civilian oversight over defense expenditures, involved issues that did not constitute direct threats to the AKP’s tenure. That nothing came of the famous 2007 “e-memorandum” — in which the TSK website expressed the military’s great disapproval of Islamist candidate Abdullah Gül’s nomination to the presidency — testifies to the AKP’s success in reducing the role of a Republican Nationalist institutional obstacle through foreign policy channels. Emboldened by this success, supporters of Ottoman Islamism went further in applying EU democratization criteria to its military obstacle. The Ergenekon and Balyoz trials, for example, charged many Republican Nationalist opposition members with forming a terrorist organization and plotting a coup to overthrow the AKP. Initially heralded as a step forward in the democratization of civil–military relations by applying the rule of law even to former Chief of General Staff and President of Turkey Kenan Evren, the trials came to be seen as a way of removing the threat of potentially powerful individuals, as well as tarnishing the institutional credibility of the armed forces. According to the results of a multiple-year survey, the percentage of respondents stating that they did not trust or definitely did not trust the military rose from 19.6% in 2011 to 31.9% in 2013 (Kadir Has University, 2015: 17). In the same survey, the percentage of respondents who stated in 2013 that the influence of the military in politics had decreased or definitely decreased in the recent period was 54.9% (Kadir Has University, 2015: 18). The weakened influence of the military in Turkey was also demonstrated by a general continuation of politics as usual following the resignation of the chief of general staff, as well as the leaders of the army, navy, and air force, in protest in 2011. Erdoğan’s immediate acceptance of the resignations and appointment of his own choice for chief of general staff served to make the domestic arena a safer one for the AKP’s contestation of national identity in its third term.

Also as part of its EU foreign policy, the AKP pushed through judicial reform that ended the jurisdiction of military courts over civilians. EU-mandated reforms were also used to abolish the State Security Court, which had sentenced Erdoğan to prison and temporarily banned him from politics for violating the principle of secularism (Yavuz, 2003: 84). The Constitutional Court, another looming obstacle to Ottoman Islamism as the body responsible for banning the AKP’s predecessors, also became a target of judicial reforms carried out in line with EU accession criteria. The AKP began to move forward with these reforms after seeing its amendment intended to guarantee the right of university students to wear headscarves overturned by the Constitutional Court in 2008, and after barely surviving a closure case the same year thanks to the Constitutional Court’s new president and (then) AKP sympathizer Haşim Kılıç (Yeni Şafak, 2008). While the AKP gained an automatic advantage when Gül became president, as the president selects all the members of the Constitutional Court, the AKP strove to quickly change the make-up of the Constitutional Court by proposing 17 regular justices rather than the existing 11 regular and four substitute justices. While in harmony with the EU standard of delimiting justices’ term limits to 12 years, this simultaneously sets the AKP up to rotate out justices that were nominated by previous Republican Nationalist President Ahmet Necdet Sezer and place AKP-friendly justices in their stead. These institutional reforms, ostensibly taken in pursuit of EU membership, also greatly advanced the AKP’s prospects for being able to spread Ottoman Islamism on the inside by weakening domestic obstacles.

Conclusion: Challenges to Ottoman Islamism

This article’s analysis of identity hegemony as pursued by the AKP’s strategy of taking its fight “outside” highlights the role of foreign policy as an arena in which proposals of national identity can be contested when they face domestic obstacles. In doing so, the theory of identity hegemony developed here provides comprehensive understanding of what appears at first sight to be a series of rapid and even erratic shifts in Turkey’s foreign policy. The theory also accounts for the meteoric rise of an Ottoman Islamist identity proposal, whose content elements clash directly with those of Republican Nationalists. The article’s breakdown of identity content provides understanding of the goals (interests) that supporters of particular identity proposals wish to see realized, while its dynamic model of identity contestation illustrates how proposals for particular identities gain and lose support due to both domestic and international factors. The AKP’s swerve away from its initially ardent pursuit of EU accession thus becomes comprehensible when viewed through a lens of identity contestation. With threats from traditional bastions of Republican Nationalism thus weakened through a foreign policy initiative, the AKP’s Ottoman Islamist understanding of national identity — including its constitutive norms of piety and acceptance of strong rulers, and Turkey’s legitimate role as leader in the Islamic world — could be spread throughout the domestic public sphere.

Major recent challenges to this pursuit of hegemony in both domestic and foreign policy arenas, however, indicate that although the AKP may have believed that it was operating with the hegemony of its identity proposal, the social structures that the party attempted to put in place were by no means met with acquiescence. Apparent brothers in the struggle against Republican Nationalists’ suppression of religious expression, Islamic cleric Fethullah Gülen and members of his Hizmet (Service) movement — who had taken their own fight against Republican Nationalism outside by establishing international charter schools and sponsoring events such as the Turkish Olympics — began to reject Erdoğan’s sultanesque governance following a series of identity- and financial-based disputes (Gülen, 2015). More broadly, a swell of opposition in the form of the 2013 Gezi Park protests catalyzed formerly apathetic supporters of rival identity proposals who found the increasingly widespread, institutionally powerful, and lifestyle-encroaching identity proposal of the AKP intolerable in the first place. Gezi represented both a staunch rejection of Ottoman Islamism and a previously unthinkable uniting of strikingly diverse groups. Prominent among the protesters’ many references to identity ³⁵ was the rejection of Ottoman Islamism’s social purpose of spreading Sunni Islam in the public sphere; “Türkiye laiktir laik kalacak!” (“Turkey is secular and will remain secular!”) was chanted hundreds of times over at various scenes of protest. ³⁶ In one of many uses of humor used to challenge power, a protestor poked fun at Erdoğan’s call for women to have three or more children and his aspiration for a newly pious generation by spray-painting “Are you sure you want three kids like us?” (Milliyet, 2013).

This use of the word “us” effectively underscores the of “Us” versus “Them” discourse that constitutes the recently catalyzed contestation by those proposing elements of Western Liberalism ³⁷ and Republican Nationalism ³⁸ as a new understanding of national identity. Indeed, the diverse make-up of Gezi protesters and the amicable interactions among them demonstrates, from a constructivist perspective, the potential for a new, hybrid identity proposal to emerge among previous rivals. However, the breakdown of identity content presented in this article, and particularly the current red lines of intolerability that it identifies, make clear the challenges to be faced — recognizing Kurdish rights, for example — in developing and disseminating an overarching identity. While certainly not insurmountable, these red lines serve as useful points of focus for those seeking the sources of, and solutions to, Turkey’s seemingly intractable identity polarizations.

Moving forward, the theoretical framework developed in this article can generate improved insight into identity struggles and foreign policy in cases such as: Israel, as suggested in the introduction; Ukraine’s increasingly conflictual West–East conundrum; and India’s potential reorientation under the Hindu nationalist Bharatiya Janata Party. Extending the framework outside of the EU context could also challenge suppositions that the EU constitutes a sui generis not only in its institutional make-up, but also in its ability to mold domestic structures and identities (e.g. Checkel, 2001; Subotic, 2011). Further, the framework’s specification of the unit of analysis as identity proposals facilitates its extension to non-state cases, including cultural and civil rights movements. Utilizing this framework could therefore identify specific conditions under which sub-state movements take their contestation outside (for example Turkey’s Kurdish and Gülen movements, South Africa’s anti-apartheid movement); contestation within movements over appropriate goals and means for achieving them; and the diaspora, transnational civil society, and international institutional dynamics shaping this contestation.