The Practice of Soft Power

: Turkey exerts signiﬁcant inﬂuence over Balkans Muslims. While some of this has to do with Turkey’s military and economic power, much relates to their shared religiosity and common history. Some may characterize this as “soft power”; however, this term struggles to completely explain these dynamics. Many Balkans Muslims have an ambivalent attitude towards Turkey, even as they accept its inﬂuence. Moreover, Turkish inﬂuence comes not from passive qualities it possesses but active steps it takes to maintain its image. We argue that this can be better explained through the practice turn in international relations; Turkey follows commonly accepted religious practices that Balkans Muslims recognize, granting Turkey inﬂuence even if they do not internalize its dominant position. We demonstrate this with the results of interviews conducted among Balkans Muslim political, religious and media ﬁgures. The article provides insight into the strategies non-state religious actors, states and the media implement in world politics, while also expanding our understanding of soft power in the world.


Introduction
Turkey's soft power is different from that of other countries in its form and content. The potential of Turkey's soft power, which extends from the Balkans and the Middle East to inner parts of Central Asia, emerges from the cultural and historical experience it has inherited. (Kalin 2011, p. 10) It is difficult to say directly that Turkey is a foreign actor, with its characteristics reaching back from its Ottoman history . . . however, its influence is quite variable. While sometimes it has very positive charm, other times it is an actor regarded with suspicion and concern due to domestic politics. (Besnik Mustafaj, 21 April 2017) 1 These two quotes, one by a spokesperson of President Recep Tayyip Erdogan in Turkey and the second from a former Albanian Foreign Minister, illustrate the unique nature of Turkey's soft power in the Balkans. Turkey's long history in the region, and its record of aid and diplomatic outreach, have granted it significant influence over its countries and societies (Bechev 2012;Candar and Fuller 2001, p. 22). At the same time, there is a distinct wariness of Turkey in the region, even as Balkans Muslims accept its presence (Demirtaş 2013). This raises important questions about Turkey's foreign policy, as well as broader questions about soft power in international relations. What are the sources of Turkey's soft power in the Balkans? And how can it exist simultaneously with distrust of Turkey among Balkans states and societies? These questions have also taken on significant media attention over a decade regarding the role of Turkish state in the Balkans and its soft power position. 2 In this sense, the nexus of journalism, religion, and international politics is an important angle from which we can study the nature of Turkish influence.
Conventional soft power analyses would struggle to answer these questions and clarify different assumptions in the popular media. Many academic studies on soft power leave

The Debate about Soft Power
Joseph Nye defined soft power as power that occurs "when one country gets other countries to want what it wants", succeeding in its preferred outcomes because other states want to follow it or have agreed to a situation that produces such effects (Nye 1990(Nye , 2004. He argued this soft power was more effective than traditional "hard" power. Part of soft power arises from setting the agenda, such as through international institutions and multinational corporations. It also arises from the attractiveness of a country's culture. Nye also argued that public diplomacy is one way states can increase their soft power; that is, public diplomacy-an "instrument that governments use to mobilize cultural resources to communicate with and attract the publics of other countries" can expand soft power, and, consequently, a state's status (Nye 2008). Nye applied it to debates over America's international standing in the Bush years, arguing the Bush Administration foreign policy was undermining America's soft power (Nye 2004). Scholars have applied this concept in several studies. Some have focused on the United States (Chiozza 2007;Kroenig et al. 2010). Others look at a variety of countries including China, Brazil and Turkey (Kurlantzick 2007;Oguzlu 2007;Lai and Lu 2012;Huang and Ding 2006;Gamso 2019;Bry 2017). Still others have extended soft power to study religious politics (Haynes 2016;Köse et al. 2016;Sandal and Fox 2013;Mandaville and Hamid 2018;Ozturk 2021a). Generally, scholars approach soft power as a resource states acquire through various policies, which then grants them deeply internalized influence over other states.
However, there are some issues with the use of this concept. Some have criticized the concept for being vague or overly expansive, with an unclear understanding of what power is (Bially Mattern 2007). Others have argued it ignores the strategic elements of states' public diplomacy (Lock 2010). Still others claim there is an overly Western focus in soft power discussions (Suzuki 2009). Additionally, some call for greater attention to the relationship between soft power, "hard" power and "smart power" (Wilson 2008). Finally, some have called for a broader approach to soft power that encompasses its varying relationship with hard power (Gallarotti 2010).
Additionally, there are two specific limitations in this body of literature that prevent it from being useful to analyse Turkey's Balkans outreach. The first is that the literature provides incomplete explanations. They tend to focus on the impact of soft power once it has been attained, either passively through a state's culture or actively through its public diplomacy. Yet, Turkey's regional status is an ongoing process relating to a distinct set of policies (Parlar Dal 2016). Soft power analyses need to more directly address the policies states enact to enhance soft power, and how they grant influence through interaction with target states. The second limitation is that the incomplete explanations are also inaccurate. Turkey has all the appearances of soft power in the Balkans, but this is not accompanied by sincere acceptance of Turkey's position in the region, or signs that Turkish outreach has persuaded Balkans Muslims of the state's intentions. Moreover, the arguments that states such as Turkey gain soft power over fellow Muslims due to a shared identity are overly simplistic and run the risk of imposing essentialist analyses on Muslim states. There has been some effort to understand the nuances of religious soft power, but these studies remain in their early stages (Mandaville and Hamid 2018). Analysing this aspect of Turkey's Balkans outreach requires a new approach to soft power, which the paper will present in the next section. Furthermore, this new way of scrutiny will contribute to the popular semi-academic and media-based discussions regarding whether Turkey is a perfect example of soft power in the Balkans under the AKP rule.

Practice Theory and Soft Power
The practice turn in international relations has opened up many new avenues to understand the role ideas play in world politics. While it has not been used explicitly to study soft power, several developments in practice theory are relevant to this topic. Using these developments, this article argues that soft power can be the result of a state using commonly accepted practices to build up capital, which grant it symbolic power to wield influence over states. In the specific case of Turkey's Balkans outreach, this takes the form of Turkey acting in line with the expected role of a responsible Balkans power.

International Hierarchy and the Practice Turn
Practice theory "takes competent performances as its main entry point in the study of world politics" (Adler and Pouliot 2011). These are "socially-structured" behaviours that grant recognition of competency to those who perform them, and are based on shared "background knowledge" in a social situation (Adler and Pouliot 2011). The various practices in international relations help to structure and guide state behaviour; as Pouliot notes, the "stock of inarticulate knowhow learned in and through practice . . . makes conscious deliberation and action possible" (Pouliot 2010). These practices do not occur in isolation, but are instead part of a "field", or a "social configuration" including "relations of power . . . taken-for-granted rules" and "unequal positions" (Pouliot 2010). Practices can also become part of geopolitical struggles as the "politics of practice concern the ways in which agents struggle to endow certain practices with political validity and legitimacy" (Adler and Pouliot 2011). Pouliot, for example, focuses on the way diplomatic practices generate influence, while Morgan has studied deterrence as a form of practice (Pouliot 2010;Morgan 2011).
Practice theory can also discuss the ways states gain influence, through the concept of capital. Capital is "any type of resources that are recognized . . . in a social context" as "allowing a player to play the game" (Pouliot 2010). Pouliot discusses these as materialinstitutional capital and-important for this study-cultural-symbolic capital, or the "narratives and symbols that define the meaning of the world . . . and legitimize it" (Pouliot 2010). States that follow practices expected of them in a social situation acquire symbolic capital, which grants them influence among the other states with which they interact. States can then exert "symbolic power" that legitimizes and perpetuates their status (Pouliot 2010(Pouliot , 2016. However, as the dominant state's practices appear to be out of sync with the "rules of the game", formerly subservient states will begin to question the dominant state's position (Pouliot 2010).
Thus, practice theory can address soft power as the symbolic power states gain through international practices. Yet, these insights are distinct from the approach of most work on soft power as discussed above. Soft power studies tend to focus on attitudes towards a state, rather than interactions. The practice turn focuses on the social implications of culture; as Adler and Pouliot put it, "culture . . . is not only in people's minds", it is "also in the very performance of practices" (Adler and Pouliot 2011). Additionally, soft power studies tend to be rather static. But practices are both "individual . . . and structural"; they arise from collective belief systems like identity, but are deployed by individual states (Adler and Pouliot 2011). Finally, soft power studies often present their analysis as opposed to material power. By contrast, practices are not in competition with material factors; they occur through interaction with them (Adler and Pouliot 2011).
Additionally, the practice turn provides an alternative to the assumption that states reflect on a powerful state's attractiveness and internalize its status as soft power. Pouliot discusses the tendency of much work in international relations to "focus on what agents think about", leading to a "representational bias". Instead, he argues that practices operate according to what actors "think from", the aforementioned background knowledge (Pouliot 2010). Indeed, Pouliot explicitly contrasts his approach with the "norm internalization" explanation of constructivism (Pouliot 2010). While Pouliot focuses on constructivism-and much work on soft power does not explicitly engage with constructivism-the implications of the practice turn are similar. Rather than states reflecting on and internalizing a powerful state's attractiveness, the practice turn would expect states' status to arise from habit and background knowledge informing their actions.
This presents an alternative way to understand soft power. Instead of it being power one state exerts over another through the attractiveness of its culture, it is a hierarchical relationship between the two states. One state has greater status than the other, leading the subordinate state to be drawn to the first state and accept its authority. This arises not through a formal institutional relationship, but through the social situation created by the two states' interactions. In the next section, we will draw on these works to present a practice theory of soft power before applying it to the specific issue of Turkey's outreach to Balkans Muslims.

Practice Theory of Soft Power
Following practice theory, we argue that soft power needs to be understood in the context of a specific field of social interactions. In this field certain practices have become salient for demonstrating states' competence. These practices accomplish two things for states. They help them to acquire cultural capital, increasing their attractiveness to others in the field. And they allow them to translate this into symbolic capital in the form of influence over other states.
First, all examples of soft power occur in specific fields of international relations that include commonly accepted practices. A field is a "social configuration structured along . . . relations of power, objects of struggle and taken-for-granted rules" (Pouliot 2010). Such fields include distinct "positions" for various states based on their "accumulation of fieldrelevant capital", which we will discuss below (Nexon and Neumann 2018). That is, a state's soft power over another is part of a broader set of interactions, values and exchanges between the states. It does not arise in a vacuum.
Additionally, the shared beliefs in these fields include the acceptance of a set of practices states are expected to undertake to competently implement foreign policy. Fields, and the values that inform them, include a "practical sense" for how states should act; this includes the specific practices states are expected to follow, which "derive from their performers' locations in a field's hierarchical structure" (Pouliot 2010). That is, in the social configuration through which states interact, there are certain behaviours and rhetoric that signify a state is responsible and in line with shared values.
States that follow these practices build up cultural capital in their interactions with other states. As Pouliot (2010, p. 33) defines it, capital is "any type of resources" that "allow a player to play the game more or less successfully". Cultural capital, in turn, is the "artefacts, narratives and symbols that define the meaning of the world . . . and legitimize it" (Pouliot 2010). Similarly, Neumann and Nexon define cultural capital as "specific material tokens of . . . cultural standing" and "the knowledge of prestigious cultural codes" and customs (Nexon and Neumann 2018). Thus, the practices in a field signify cultural codes for states; as states perform these practices, they demonstrate their competence in the field and acquire cultural capital.
Wielding this cultural capital creates a hierarchical structure between the state and others. That is, the powerful state becomes recognized as having a higher status than the others through the cultural capital it acquires through the relevant practices. As Neumann and Nexon argue, "those whose dispositions reflect a better feel for the rules of the game gain advantages over others" (Nexon and Neumann 2018). Likewise, Pouliot (2010, pp. 47-48) claimed "the order of things is established through the iterated practices performed by capital-endowed players"; similarly, "dominant players become masters whose higher position in the game and control over its rules are self-reinforcing assets". And as we noted above, this enhanced status comes not from states internalizing the powerful state's position, but out of habit.
This dynamic process creates the set of relations commonly categorized as "soft power". In any set of interactions among states, there will be certain practices that states are expected to follow, especially involving salient cultural symbols. States that follow these practices will build up recognition among other states as a responsible and competent power. This translates into greater status in interactions, and a superordinate position in the social arrangement. Weaker states will be attracted to the powerful state due to this status as long as it continues to perform the requisite practices. Yet, this is not an automatic identification due to similar cultures or a principled embrace of the powerful states; the weaker states are operating out of habit.
Looking specifically at the case of Turkey, Turkey's interactions with the Balkan states are part of a clearly delineated social context, defined by historical legacies and common identity. Balkan states, in turn, recognize the significance of these practices, and accordingly recognize Turkey's unique position for the region in terms of history, religion and language. They do so begrudgingly, however, out of deference to Turkey's culturally prescribed status rather than internalization of Turkey's superiority. Thus, what appears to be Turkish "soft power" over Balkans Muslims is something more dynamic and complex. The history of Turkeys interactions with the Balkans has produced a set of behaviours that are expected of superordinate powers, based on historical legacies and common Muslim identity. Turkey has focused much of its foreign policy towards the region on performing these practices, gaining the recognition of Balkans Muslims.
In this regard, in the cases we will demonstrate why our practice theory of soft power is valid, and alternatives are not effective in explaining this case. This study expects to find that Turkish "soft power" arises out of the interactions between Turkey and Balkan states, both current and historical. Moreover, Turkish influence will be dependent on Turkey following the practices that are expected as part of this relationship. This is in contrast to material explanations, which would expect Turkish influence to be based on economic support to some of the Sunni Muslim groups in the region (Ozturk 2021b). It is also in contrast to constructivist explanations, which would expect Turkish "soft power" to result from a shared identity producing internalized positive attitudes towards Turkey among Balkans Muslims.

Methodology
This study analyses the practice of soft power in Turkish foreign policy through a qualitative study of attitudes towards Turkey among Balkans Muslims. This article uses process tracing, specifically the variant Pouliot presented for use in the study of international practices. We use a Bayesian version of process tracing to undertake this study. As defined by Bennett and Checkel (2014, p. 7), process tracing is "the analysis of evidence on processes, sequences, and conjunctures of events within a case for the purposes of either developing or testing hypotheses about causal mechanisms". They call specifically for a Bayesian version, in which scholars uncover evidence that would be unlikely if the theory being tested was incorrect. Pouliot's (2015, p. 243) variant involves researchers understanding "what the practice under study counts as in the situation at hand". That is, a study must understand what a set of behaviours "count as" in a social setting, and how these practices "structure the interactions and cause practitioners to do a number of things" (Pouliot 2015). While ethnographic studies are ideal, interviews can also capture the dynamics of practices (Pouliot 2015).
The study focuses on a single "case"-Balkans Muslims-but includes significant variation in country and position of interview subjects, ensuring this is not based on an unrepresentative sub-set of the region in question. There may be questions about generalization beyond the Balkans, which we discuss in the conclusion. The data this study draws on come from three sources. First are the statements that relevant Turkish Justice and Development Party (Adalet ve Kalkinma Partisi, herein after AKP) government officials, not the least of whom was Turkish President Erdogan, issued between 2002 and 2019 regarding the Balkans, as well as the reactions of Balkan leaders and political actors to these statements. Second, we examine activity reports and documents that Turkish transnational state apparatuses active in the Balkans prepared specifically for the region. Apart from consulates and embassies, we scrutinise the activities of the Diyanet (Presidency of Religious Affairs, Diyanet Işleri Başkanlıgı), TIKA (Turkish Cooperation and Coordination Agency, Türkİşbirligi ve Koordinasyon Ajansı Başkanlıgı), TOKI (Housing Development Administration of Turkey, Toplu Konutİdaresi Başkanlıgı), Yunus Emre Institute and the Maarif Foundation, which was established to break the regional influence of the Gülen Movement 3 .
Finally, this article relies on a total of 82 semi-structured elite interviews conducted in Turkey (13), North Macedonia (18), Bulgaria (21), Albania (19), Serbia (6) and Bosnia Herzegovina (5) between April 2016 and January 2020. 4 While some interviews were conducted with prominent members of Diyanet, the Turkish ambassadors in Balkan countries and former foreign ministers, Balkan countries political party representatives and political elites, others were held with figures and representatives from other Turkey's transnational state apparatuses, and the Gülen Movement. These interviews explored two main issues. First, we asked how and in what manner did the changing and transforming structure of Turkey's domestic politics influence Balkan policy. Second, we investigated whether Balkans Muslims perceive the religious, cultural, historical and other normative elements of Turkey as elements of power and how they define this power. Five of the interview subjects were regional journalists who could provide their perspectives on Turkey's influence among Balkans Muslims as well as insight into how they handled these issues in the media products they produced. We also draw on primary and secondary studies on Turkey and the Balkans states.

The Practices of Turkish-Balkans Relations
First, what is the social context of Turkey's "soft power" in the Balkans? Turkey and the Balkans are linked by a long historical legacy and cultural ties in which Turkey is in a dominant position particularly among the Sunni Muslim groups in the region. Both Turkey and the Balkans are culturally, religiously, historically and politically engrossed in a relationship of loyalty and dependence (Öktem 2010). As Bechev (2012) noted, Turkey and the Balkans Muslims have always existed at a historical convergence, due to their geographical proximity as well as their common ethnicity (Turkic background) and religious identity (being Sunni Muslim). The depth of these connections can be seen in the fact that the birth places of most parliamentary members from Turkey's founding assembly were at the time in Balkan nations not within the borders of Turkey. This does not take the form of a static common identity, however. Instead, this identity includes certain expected behaviours by both Turkey and Balkan states. For example, Turkey's religious influence in the Balkans during the 1990s involved not common religiosity but Turkey "watching out for" Balkans Muslims on religious matters. This is particularly apparent in concerns about the spread of Wahhabi and Salafi influence in the region, which Turkey endeavoured to counter. For example, during an interview we held in March 2018, a prominent official from the Islamic Religious Community of North Macedonia (Иcлaмcкaтa Bepcкa Зaeдницa вo Maкeдoниja) said: We were lucky in 1990s, because Turkey was prepared to assist us, and we desired this assistance from Turkey. Turkey's ongoing presence in our region from the 1990s until today as-in a cultural, historical and religious sense-one of us ensured that no deviant religious ideologies would exist in the Balkans.
These ties have deepened under the AKP. Some of this grew from expanded economic interactions: the relatively strong economic growth that Turkey exhibited, especially between 2002 and 2010, when the AKP ruled on its own (Öniş 2012), shifted towards the Balkans as an investment. But it also related to the foreign policy orientation of Erdogan, Ahmet Davutoglu, and other AKP figures, as they placed a much greater value on the Balkans than their predecessors as part of their neo-Ottoman vision (Yavuz 2016). This also involved cultural connections, both by non-state Turkish actors-such as the Gülen Movement-and Turkish state bodies such as the Diyanet which is one of the most active transnational religious state apparatuses of Turkey (Çitak 2018).
Again, this did not involve mere common identification as Muslims. Instead, Turkey was following a set of behaviours based on the historical social ties between Turkey and the Balkans, and Muslims in Balkan societies responded accordingly. This is apparent in a summary of the period by former Albanian Technology Minister Gec Pollo's offered a key summary of the period in April 2017: In its first years of Erdogan, Turkey took steps as if to prove that Muslim identities and democracy could coexist. Its goals at the time were the European Union and a fully developed democracy. What happened later and how it happened are separate topics of debate, but over time, Turkey has grown distant from its values and, in some way, has deviated from its path toward the west. This separation-I don't know if they are aware of it-it fundamentally means in a sense to deviate from the Balkans, because our path and our desire is for the European Union and Western values. 5 This sentiment was echoed by a Turkish Ambassador serving in one of the biggest and important Balkan countries in April 2016: The Balkans are a very important place for Turkey. That's why, if we are to claim that we are a great nation, a great civilisation, we must have a presence here. Moreover, this place is crucial for us politically. It is, on the one hand, a door opening to the European Union and, on the other hand, a geography where great powers, not the least of which being America and Russia, seek to gain influence. Here, it would be absurd not to consider Turkey's sway in the region. 6 This shifting attitude towards the Balkans is also apparent in the writings of Ahmet Davutoglu, who served as Turkish foreign minister between 2009 and 2014 and as prime minister between 2014 and 2016. The fundamental cause of this was that Davutoglu, himself an academic, was the main determinant of foreign policy until 2016. In his view, Turkey possessed the power to embrace the legacies of both the Ottoman Empire and pre-Ottoman Turkish and Islamic states. It could become a power through the use of Islam, its history and its geostrategic position in its own region, in former Ottoman territories and around the world. Although this notion was defined as pan-Islam in the academic sphere (Özkan 2012), Balkan nations began to respond positively to it. Significant here is what a senior official from the Macedonian Turkish Movement Party noted in April 2018: The AKP's Turkey, although sometimes a bit challenging and imaginative, had plans for the Balkan region. It began to make cultural, commercial and religious investments here. They began for the first time to facilitate relations with the public along with political relations, something I think was very important. 7 These new circumstances prompted Turkey to re-examine, with a different perspective, Turkey's Balkan policies during the AKP era (Brljavac 2011;Demirtaş 2015). However, a majority of these studies produce their analyses by accounting for the internal dynamics of neither Turkish nor Balkan politics, similar to many different media comments. Despite this, the Balkans and Turkey, in a state of interaction, experience domestic political changes both independently and jointly, and these changes affect policy preferences and Turkey's perception in the region.

The Practice of Religious Outreach, and Turkey's Balkan's Influence
Thus, Turkey's close ties to the Balkans involved not just a static identity, but a set of behaviours both sides expected Turkey to follow. These practices are what granted Turkey influence over the Balkans. This can be seen in the response of a senior-level official from the Islamic Community of Kosovo (Bashkësia Islame e Kosovës) in March 2018 to the question of how to interpret Turkey's evolving influence.
Turkey was, for us, always important. Even if we do not go and see, existing there is a nation that stands for us and is of our lives, our blood and our religion. But we felt its existence first during the Özal period; Turkey was later lost partly, but the Turks were always there. Now, Turkey is everywhere with Erdogan. 8 Some of this undoubtedly had to do with the extraordinary investment that Turkey made in Kosovo after its independence in 2008. Of these investments, the largest was an enormous mosque the Diyanet planned to construct in Pristine, a city with a population of close to 2 million and more than 800 mosques, in 2018 and 2019 despite the deteriorating Turkish economy. Just like the largest mosque in the Balkans, constructed in the Albanian capital of Tirana, and other mosques sponsored by the Diyanet, this too was designed with architecture symbolizing the Ottoman era. This is partly economic support from Turkey, but it is also clearly religious outreach, building up "cultural capital" by appealing to shared beliefs and history (Luke 2013).
Balkans Muslims recognized the cultural and symbolic element of these efforts by Turkey. That is, Turkey was not just buying goodwill by spending money in the region; it was gaining influence through engaging in the practices expected of it. For instance, Ferdinand Xhaferraj, who served as Albanian Minister of Culture between 2009 and 2011, interpreted Turkish investments and its relevant religious influence differently than Muslims. In an interview we held with him in May 2017, he stated: Turkey is, for us, a very important actor. There is no debate on this issue, but we have recently begun to see the perpetual existence of Turkey in religion in a manner not previously seen. Yes, it is true, Turkey's commercial investments, our collaboration are growing, but Turkey in a sense is now a nation defined through its religion. 9 The influence of these practices is also clear from Balkans Muslims' reactions to transnational Turkish state apparatuses such as TIKA, the Yunus Emre Institute and the Diyanet. During an interview in April 2017, Dedebaba Hajji Mondi (Haxhi Baba Edmond Brahimaj), the Bektashi leader in Tirana, said: We certainly received no direct aid from Turkish Diyanet, but we began to establish relations. And most of all, we observed. I can say this based on my observations: Turkey's institutions, I believe, work in coordination under the leadership of the Diyanet, although it may be another institution according to someone else. For instance, mosques are being built; TOKI and TIKA are doing this. The Diyanet is cultivating and appointing imams; they are providing religious books. The Yunus Emre Institute engages in cultural activities. This is a collective policy. 10 The work of the Gülen Movement, when it had good relations with the AKP government, was also part of this influence. The Gülen Movement had a somewhat more widespread and powerful network in the Balkans than Turkey did. They operated in cooperation in some manner in the region until 2013, based on the observations of the regional public and Gülen Movement representatives. In May 2018, an Albanian representative of the Gülen Movement noted: We worked together with the AKP and its institutions in the past. If only we had not made this error, but we did it at one time. Primarily, we vouched for them in the region. We vocalised that they would introduce here institutions and activities that would not clash with the values of Muslims, of those who believe in democracy, of the Balkan people. After all, we have been here for close to 30 years with our schools and other institutions in the region. 11 However, the Turkish side conveys the situation after 2013 much differently. Then, Turkish Ambassador to Bulgaria Süleyman Gökçe explained the relation in April 2017: I was not here before the 2010s, but the situation at the time was relatively the same in the Balkans as it was in other places around the world. If there was a request coming to us from the organisation, we would help as much as we could. They were going places thanks to us, establishing a form of relations with our credit. 12 Despite these different explanations the regional elite interpreted the relationship between the Gülen Movement and the AKP much differently, and this interpretation fundamentally harbours elements of soft power and religion. Former Macedonia Divinity Faculty Provostİsmaili Bazdi summarized in greater detail during our May 2017 interview what each Balkan political elite expressed during our field work, without exception: There may be two things about the relationship between the Gülen Movement and the AKP. First is the desire to establish hegemony under a soft influence, and second is to reveal the most Turkish and Muslim identity of themselves. While they maintained operations over these two things when they were together, they began, after separating, to strike one another over these two things and to influence the Balkans directly. So, what were they doing while together? They were working together like two halves of an apple. While Gülen was much more influential over society, the AKP began to engender influence over the political elite in Balkan states. 13

Turkey's Failure to Follow the Rules of the Game and Its Declining Soft Power
Although Turkey has maintained its role as one of the dominant external or international actors in the Balkans since 2013, some continue to regard Turkey and its regional activities with suspicion. The greatest source of this is the religious oriented changes in Turkey's domestic politics, which reflect on its foreign policy and, consequently, on Balkans Muslims.
As Ismail Bazdi mentioned above and originating from changes in domestic Turkish politics after 2013 (Selçuk et al. 2019), the evolving perception in the Balkans could not be reduced to merely an interest-based battle between the Gülen Movement and the AKP. We can summarize these changes, starting in domestic politics and reflecting in foreign policy, with four points: (a) Turkey's rapid departure from democratic values, (b) its deviation in its domestic political choices in a more ethno-religious direction, (c) its adoption of more reactive and aggressive preferences in foreign policy due to this deviation, and (d) its modification of its system of governance by shift.
These four undoubtedly interrelated facets of change did not occur spontaneously. The 2011 Arab revolutions (Akkoyunlu and Oktem 2016), starting after the 2008 economic crisis (Aydın 2013), spawned a degree of uncertainty in the AKP's self-confidence. The Gezi Protests, which began in 2013, and the clash between the AKP and the Gülen Movement triggered a hardening of the regime's policies and its more widespread utilization of religion and religious institutions in its depictions of itself (Taş 2018). We can comfortably observe these policy changes and the Gülen-AKP conflict in the Balkans, such that it is reasonable to claim that substantial changes have emerged in Turkey's regional soft power.
The conflict between the Gülen Movement and the AKP fundamentally demonstrates that the influence of the two structures, which have operated in the Balkans for many years, changed dramatically. Causing this most significantly was the desire of the two structures to quickly break the influence of the other using religion, essentially damaging the soft power that Turkey had crafted. In May 2018, a senior official in the Bosnia Herzegovinian Foreign Ministry explained this influence: The Gülen-AKP conflict, or I would easily call war is felt here in all its force. An unbelievable battle continued, especially until the coup. Both sides perceived the other as traitors and accused them of not being Muslim. This caused the political choices flowing in from Turkey to be much more religious. Anymore, the mosques under Turkey's control, with its assistance, have become fields of political propaganda beyond religion. Gülen activities were like that as well, and this is not something that we wanted. 14 As we noted before, the changes in Turkey's domestic politics were among the issues closely monitored in the Balkans. Apart from the Gülen-AKP struggle, the democratic erosion that Turkey suffered under Erdogan's control (Esen and Gumuscu 2016) negatively impacted Turkey's image in the Balkans. The 15 July 2016 coup attempt in particular culminated with Turkey's implementation of a religion-based policy-to an extent not seen before in the Balkans-and its more active use of transnational state apparatuses in its pursuit of this policy (Ozturk 2021b). This made the recommendations regarding a more extensive use of religion more visible (Demirtaş 2015).
It is in Albania that we most clearly see Turkey's intervention in domestic politics and functions in the Balkans and the reactions exhibited in opposition to this. Recognizably, Turkey constructed the largest mosque in the Balkans in the Albanian capital of Tirana, through the Turkish Diyanet (Öztürk and Baser 2022) and the Turkish Co-operation and Coordination Agency. The ceremony to break ground was held in Tirana in 2015. And during the ceremony, Erdogan articulated that the Albanian Diyanet's (The Muslim Community of Albania, Komuniteti Mysliman i Shqipërisë) Director Skënder Bruçaj, whom he believed was a member or a sympathizer of the Gülen Movement, should be removed from office, and refused to shake his hand. As it was presented in the media numerous times, the Albanian parliament protested in response, perceiving Erdogan's demeanour as actions interfering with the nation's domestic affairs. 15 Although Skënder Bruçaj continued in his position for a while afterwards, the Turkish Diyanet terminated its entire economic relationship with the Muslim Community of Albania for this reason.
Turkey occupies an increasingly active position in North Macedonia, as it does in other Balkan countries, through its transnational state apparatuses in such areas as Gostivari (Гocтивap) and Tetova (Teтoвo), where there are dense Muslim populations. Turkey builds mosques, trains imams and thus seeks to affect some degree of influence over society in the region. While this and similar activities have not encountered a directly negative reaction in North Macedonia or in Muslim-majority nations, such as Bosnia and Herzegovina and Kosovo, some of Turkey's more recent actions have transcended descriptions with the concept of soft power (Öztürk 2020). For example, Turkey signed an action in 2018 that reverberated around the world as a result of a collaboration agreement between the transnational state apparatuses in the country and the intelligence service. For instance, Turkish intelligence service kidnapped six Gülenists from Kosovo and brought them to Turkey and this took a significant media attention. 16 Again, there is a sense that Turkey is not playing by the rules of the game in its interactions with Balkans Muslims, harming its appeal.
Another institution Turkey actively mobilized in the Balkans, especially after 2016, was the Turkish Maarif Foundation, which sought to seize the educational facilities that belonged to the Gülen Movement and open alternative education centres. It is apparent that, while the institution portrays itself as a structure which aims to construct new bridges of education, culture and civilization, its fundamental objective is to decimate the regional influence of the Gülen Movement. It managed to establish numerous educational institutions over a short period of time yet has been unable to completely incapacitate the Gülen institutions in the region. For instance, the Yahya Kemal Prep School, which has been active in North Macedonia for 20 years, is still among the region's most influential institutions that provide Turkish education, and the regional politicians we interviewed expressed that its influence would not easily be eliminated for years. Some countries, such as Slovenia and Bulgaria, have still not conceded to allow this institution into their countries, the reason for which is a lack of trust.
Turkey's evolving and transformative policies prompted a reduction in the loans of some institutions, not only those that wished to set up operations in the region, starting from scratch, but also those that had long maintained a regional presence. In this context, the Diyanet, which has operated in the region for close to 20 years, has suffered through somewhat challenging times because it has exceeded its assigned duties. In 2017, Germany published a report about the international espionage activity of the Turkish Diyanet through its imams. This report covered both continental Europe as well as Balkan countries such as Bosnia and Herzegovina and Serbia. The media was named the "spying imam case". 17 Such circumstances were not limited merely to the report. For example, Ugur Emiroglu and Adem Yerinde, who served as religious officials for Turkey in Bulgaria in 2017, were deported because they interfered with Bulgarian domestic policy on Turkey's behalf. 18 This indicates the degree to which Turkey's changing preferences of political formation have harmed the structures of religious institutions that previously appeared as positive actors in the Balkans.
However, despite all these relatively negative situations, no Balkan country is opting to entirely sever its ties with either Turkey or its transnational state apparatuses. The underlying reason for this is historical, economic, political and cultural proximity. However, behind this is a situation that pertains to religion. More pressing is the concern about a possible invasion of Arab Islam should Turkey withdraw completely (Koppa 2020). Overall, Turkey still wields normative and positive influence over the region's Muslim and Turkishspeaking communities. This, accordingly, does not represent a normative rejection of Turkey's position but, rather, a wariness over its increasingly unacceptable practices-at least from the perspective of the Balkans.

Turkish "Soft Power" as Practice
The above discussion provides significant support for this article's argument that Turkish soft power is best characterized as an extra-territorial international practice. Turkey's relationship with the Balkan states is part of a common identity, but one based on commonly accepted behaviours rather than a more static identification. Historical legacies and religious beliefs have tied the two together, creating expectations for each side in their interactions. By living up to these expectations in its foreign policy, especially in recent decades, Turkey managed to gain some influence over Balkans Muslims. Yet, as Turkey's behaviour diverged from these standards, its influence waned. Thus, Turkey did acquire "soft power" over Balkans Muslims-getting them to "what Turkey wanted"-however, it was not a fixed asset but rather a result of ongoing, dynamic interactions. These elements of Turkish influence were apparent in conversations with Balkans Muslim political and religious elites, as well as among regional journalists covering these issues.
The discussion can also reject alternative explanations. Even though much media coverage is about Turkey's economic impact in the region, Turkish influence in the Balkans is not just about economic incentives. As the interviews with Balkans Muslim elites indicated, Turkey was not just "buying off" Balkans Muslims to increase its influence. It arose from historical, cultural and religious ties and from Turkey acting in line with expected behaviour. To ignore this is to miss an important element of Turkey's foreign policy. At the same time, a constructivist expectation that the Balkans and Turkey are united in a common identity is not valid. Balkans Muslims did not identify with Turkey because they were both Muslim and formerly part of the Ottoman Empire. Balkan elites and the Balkan based journalists we interviewed were not reflecting on norms guiding their behaviour. They were responding to Turkey acting in line with commonly expected practices. In this regard, we can go beyond the limits of the classical media-based arguments and note that the phenomenon known as religious soft power does not and cannot remain soft to the extent necessary or intended by state actors, because they employ economic and other sanctions in the name of power through instruments that utilise religion.
This study focuses on a single case, but its characteristics-a powerful state interacting with weaker states in the context of long historical relations and shared identity-is common enough to allow for some generalizable insights into both soft power and how to study media coverage of religious influence. Instead of looking for ways states can "build up" soft power, scholars should focus on the interactions between states and the social situation in which these interactions occur. Additionally, studies of soft power should try to connect their research to useful theoretical developments in international relations, such as the practice turn. Moreover, when attempting to understand the extent of a state's influence over other's religious affairs-as in the case of Turkey and Balkans Muslims-it is not sufficient to look in the media for positive or negative mentions or track opinion polls. Instead, scholars should pay attention to the religious practices shared on each side, and whether the actors involved are following those practices.
This study can also provide some insights for the international community as it tries to make sense of Turkey's status as an emerging power. Under AKP rule Turkey has been turning increasingly to religious nationalism; this began in the domestic realm but is spilling over in Turkish foreign policy. This has granted it greater influence over nearby Muslim populations but has also created an ambivalent situation. As the AKP government has wielded its religious nationalism to undermine political opposition domestically, this has increased suspicion among Balkans Muslims. Thus, the very instrument through which Turkey has expanded its influence in the Balkans may also prove its undoing; this could be the case for other regions, such as the Middle East. Regarding Turkey's soft power activities, it is better to check some popular media outlets; https://www.euractiv.com/section/ global-europe/news/turkey-steps-up-soft-power-efforts-neglecting-domestic-issues/ (accessed on 12 June 2022), https:// www.duvarenglish.com/turkeys-limited-soft-power-in-the-balkans-article-57505 (accessed on 12 June 2022), https://www. intellinews.com/beyond-the-bosporus-soft-power-in-the-balkans-hard-power-in-the-caucasus-200459/ (accessed on 12 June 2022), https://www.dailysabah.com/op-ed/2018/06/14/turkey-projects-soft-power-in-the-balkans (accessed on 12 June 2022). 3 The Gülen Movement is referred to by its members as Hizmet (service); however, the Turkish state classifies the Movement as a terror organization that played a role behind the 15 July 2016 coup attempt These two opposite emic and etic depictions reflect the polarized context that also informs the scholarly works on the subject. In order to maintain academic neutrality, we employ a rather neutral term, the Gülen Movement. 4 We anonymised many of our interviews because the political atmosphere in both Turkey and the Balkans is exceedingly tense. For Erdogan's full speech during the ceremony please, see: https://www.youtube.com/watch?v=quUD5t_XN-8. Last accessed 5 May 2020. 16 Erdogan's Long Arm: The Turkish Dissidents Kidnapped from Europe, by Michale Colborn and Maxime Edwards, Haaretz, 30 August 2018, https://www.haaretz.com/middle-east-news/turkey/.premium-erdogan-s-long-arm-the-turkish-nationals-kidnappedfrom-europe-1.6428298. Last accessed 5 May 2020. 17 Does Turkey use "spying imams" to assert its powers abroad? by Ahmet Erdi Öztürk, The Conversation, 15 April 2017; https: //theconversation.com/does-turkey-use-spying-imams-to-assert-its-powers-abroad-75643. Last accessed 5 May 2020. 18 The Turkish Diplomat, Consulate in Burgas, Declared Persona Non-grata, by an anonymous reporter, GuideBulgaria.com, http://news. guidebulgaria.com/a/10137/the_turkish_diplomat_consulate_in_burgas_declared_persona_non_grata_.htm. Last accessed 19 May 2020.