Sending “our brothers” back “home”: Continuity and change in President Erdoğan’s discourse on Syrian refugees

Abstract

Considerable research has examined Turkey’s discursive governance of the Syrian refugee crisis, identifying the central themes and metaphors in top officials’ refugee-related messages. However, since they tend to rely on qualitative analyses based on convenience or purposive samples, prior studies have failed to assess the relative frequency of these themes and fall short of reliably gauging the shifts and continuities in the official discourse on refugees. Moreover, while several studies have noted the growing emphasis on the repatriation of Syrian refugees in recent years, no research has yet explored how the Turkish government has sought to reconcile this with its pro-refugee posturing. This paper addresses these limitations via a mixed methods analysis of 382 speeches President Erdoğan gave from September 2014 through December 2022. Quantitative findings show that Islamist and neo-Ottomanist themes have played a major role in Erdoğan’s refugee discourse throughout his presidency. However, since 2018, there has been a sharp increase in Erdoğan’s remarks about repatriating Syrian refugees. A critical discourse analysis of these remarks indicates that Erdoğan has appropriated the language of international law and standards on refugee returns so that he can continue to claim the moral high ground while simultaneously advocating mass repatriation of the Syrians.

Introduction

Shortly after Syria descended into civil war, the Turkish government adopted an “open-door” policy for Syrians fleeing violence, providing them with shelter and humanitarian assistance. According to data from the United Nations Refugee Agency (United Nations High Commissioner for Refugees; UNHCR), at the time, there were only about 10,000 refugees in Turkey – a tiny fraction of the roughly 10.5 million refugees then under the agency’s mandate (UNHCR 2011, 40). However, by the end of 2014, just a few years after the outbreak of the Syrian civil war, Turkey had become home to nearly 1.6 million Syrians who were registered under temporary protection (TP) status – the largest refugee population covered by UNHCR’s mandate (UNHCR 2015, 2). According to official figures from July 2023, the country now hosts over 3.3 million Syrian TP beneficiaries, who make up about 9.4 percent of the world’s 35.3 million refugees (UNHCR 2023).

The sheer volume and rapidity of this refugee movement, as well as its prolonged character, have brought immense political, economic, and social challenges for Turkey, making it critical for the administration to carefully frame and justify its asylum policies. Moreover, the transregional impact of the Syrian conflict has drawn international attention to Ankara’s response to the refugee influx, providing the Justice and Development Party (Adalet ve Kalkınma Partisi; JDP) government with a watchful audience beyond Turkey’s boundaries. It is against this background that considerable research has addressed JDP’s discursive governance of the Syrian refugee crisis.

This literature can be grouped under three distinct, though partly overlapping, research clusters. First, a good deal of scholarship scrutinizes JDP’s hospitality discourse vis-à-vis Syrian refugees, a discourse that draws on the popular mythology about traditional Anatolian–Turkish generosity toward guests. Most of these works criticize the guest metaphor and its presumption of temporariness regarding the refugee situation for hindering Syrians’ long-term integration into Turkey (Daǧtaş 2017; Özden 2013; Şirin Öner and Genç 2015; Toğral Koca 2016). Others go further and discuss how the notion of guesthood has enabled the Turkish government to discriminate among diverse refugee groups in accordance with its domestic and foreign policy considerations (Abdelaaty 2021). Second, various studies have investigated the connections between JDP’s discourse on Syrian refugees and its religious conservative nation-building project. This line of research is concerned with how the JDP leadership has exploited the refugee crisis to accentuate Turkey’s Ottoman–Islamic heritage while also asserting moral superiority over its domestic opponents and international counterparts (Balkılıç and Teke Lloyd 2021; Devran and Özcan 2016; Karakaya Polat 2018; Kloos 2016; Morgül 2022). Finally, the third group of studies highlights the changes in Ankara’s refugee discourse since the termination of the open-door policy in late 2015. Some of these studies document the emergence of a utilitarian logic based on skilled refugees’ potential economic contributions to Turkey (İçduygu et al. 2017; McCarthy 2021; Özdemir Taştan and Çoban Keneş 2019), whereas others report an increased emphasis on the Syrians’ eventual return to their country (Aydemir 2023, 669–670; Balkılıç and Teke Lloyd 2021, 371; Yanaşmayan et al. 2019, 41–42).

Overall, the existing literature provides important insights into JDP’s strategic framing of the refugee crisis by identifying the central themes, metaphors, and shortcomings in the party’s public discourse on Syrian migrants. However, since they tend to rely on qualitative analyses of convenience or purposive samples, prior studies fail to assess the relative frequency of the themes and metaphors they have observed in JDP leaders’ refugee-related messages. For the same reason, these works also fall short of reliably gauging the shifts and continuities in JDP’s refugee discourse – a significant limitation given the protracted nature of the refugee crisis and the fluctuations in the government’s policy responses to it. Furthermore, while several studies have noted Ankara’s growing emphasis on the eventual repatriation of Syrian refugees, scholars have not explored how the JDP government has sought to reconcile this emphasis with its pro-refugee political discourse.

I address these limitations via a sequential (QUANT → QUAL) mixed methods analysis of 382 speeches that President Recep Tayyip Erdoğan gave from September 2014 through December 2022. I focus on Erdoğan’s speeches because, as the most powerful and popular JDP leader, he is uniquely positioned to convey the party’s political and ideological coordinates to the masses. My quantitative findings show that Islamist and neo-Ottomanist themes have played a central role in Erdoğan’s discursive governance of the refugee crisis ever since he became president in 2014. However, starting in 2018, there has been a sharp increase in Erdoğan’s remarks about repatriating Syrian refugees. A critical discourse analysis (CDA) of these remarks reveals that Erdoğan has appropriated the language of the international legal framework on refugee repatriation, asserting that his government works to ensure the voluntary, safe, and dignified return of displaced Syrians to their homes. In doing so, Erdoğan has represented the JDP government’s unilateral efforts to repatriate Syrian migrants as a natural extension of Ankara’s humanitarian stance on refugees. This discursive move, in turn, has allowed the Turkish president to maintain his claims of moral superiority over his domestic opponents and international counterparts while simultaneously advocating mass repatriation as the desired solution to Turkey’s refugee crisis.

Prior research on JDP’s refugee discourse

Social scientists have long recognized the crucial role that strategic use of language plays in politics. Political discourse analysts, for example, have convincingly argued that social problems, collective identities, and political objectives are not pre-given inputs of partisan struggles; rather, they are constructed through symbolically mediated interactions between leaders and their varied audiences during such struggles (Chilton 2004; Fairclough and Fairclough 2013).

In keeping with this social constructionist insight, considerable research has addressed JDP’s discursive governance of the Syrian refugee crisis. These studies can be divided into three main clusters according to their thematic focus, with the first one examining JDP’s discourse of hospitality toward Syrian refugees. Many studies in this cluster emphasize that, during the early phase of the refugee flows from Syria, Turkish officials referred to the newcomers as “guests” rather than “refugees” to avoid the legal implications of the latter term (Abdelaaty 2021, 2833–2834; Daǧtaş 2017, 661–662; Makovsky 2019, 5; Özden 2013, 5; Şirin Öner and Genç 2015, 257). In fact, some scholars suggest that the tendency to prefer the guest metaphor over legally established categories such as “refugee” or “asylum seeker” continued well beyond the first few years of the refugee influx. Özdemir Taştan and Çoban Keneş (2019, 14), for instance, claim that Syrian refugees were “still defined as ‘guests’ in the rhetoric of the AK Party leaders” by the end of “their fifth year in Turkey.”

This body of research recognizes that the notion of guesthood was instrumental in JDP’s efforts to prevent an anti-refugee backlash among citizens, for it not only invoked the revered tradition of generosity toward guests but also conveyed to the public that the refugee situation and its associated costs would be temporary (Devran and Özcan 2016, 45–46; Özdemir Taştan and Çoban Keneş 2019, 13–19). That said, quite a few scholars have criticized this presumption of temporariness and its institutionalization in Turkey’s asylum regime for failing to provide long-term security to the refugees, thereby hindering their social and economic integration (Baban et al. 2017; Carpi and Pınar Şenoğuz 2019; Özden 2013; Şirin Öner and Genç 2015; Toğral Koca 2016). Moreover, some have argued that the legal ambiguity created by the notion of guesthood has enabled the Turkish government to vary its treatment of different refugee groups in line with its domestic and foreign policy considerations (Abdelaaty 2021). Consistent with this argument, several scholars have suggested that the Turkish government has been less welcoming toward Kurdish, Alawite, and Yazidi refugees than it has been toward Syrian Arab migrants of Sunni Muslim background (Balkılıç and Teke Lloyd 2021, 369–370; Karakaya Polat 2018, 512–513; Kloos 2016, 544–545; Toğral Koca 2016, 217–218).

A second cluster of studies probes the connections between JDP’s refugee discourse and its Islamist and neo-Ottomanist nation-building project. Here we can discern two major themes. The first theme concerns how JDP leaders have sought to establish a fraternal bond between Turkish citizens and Syrian refugees through their remarks about the shared history and religion between the two communities. Previous studies on this theme pay particular attention to the ensar-muhacir (ansar-muhajir) analogy, which not only justifies the government’s asylum policies on religious grounds but also reaffirms the centrality of Islam to Turkish national identity (Devran and Özcan 2016, 43–44; Karakaya Polat 2018, 505–506; Kloos 2016, 546–548; Korkut 2019, 670; Morgül 2022, 15–16; Özdemir Taştan and Çoban Keneş 2019, 12–13). The second theme, on the other hand, concerns JDP leaders’ repeated references to the alleged Ottoman tradition of giving shelter and protection to the oppressed regardless of their ethnic or religious backgrounds (Devran and Özcan 2016, 46; Karakaya Polat 2018, 506; Morgül 2022, 13). Some scholars suggest that this narrative has allowed the JDP government to couch its neo-Ottomanist foreign policy in humanitarian terms (Özdora Akşak 2020). Others argue that it has also represented Turkey as an emerging global power, thus appealing to the nostalgia for Ottoman greatness among conservative and nationalist voters (İçduygu et al. 2017, 459; Karakaya Polat 2018, 507; Korkut 2019, 667–668).

These studies additionally show that the JDP government has used the above-mentioned Islamist and neo-Ottomanist themes to denigrate its domestic opponents and international counterparts. To discredit the domestic opposition, JDP representatives have employed a populist language, portraying criticisms of Ankara’s refugee policies as a reflection of traditional political elites’ alienation from the Turkish people’s historical and religious traditions (Aydemir 2023, 666; Karakaya Polat 2018, 510–511; Morgül 2022, 18–19). At the same time, JDP leaders have targeted their European and American counterparts via a civilizationist language, one that contrasts the purported benevolence of Turkey’s Ottoman–Islamic heritage with the West’s deep-rooted racism and xenophobia toward international migrants (Balkılıç and Teke Lloyd 2021, 365–368; Karakaya Polat 2018, 507–510; Korkut 2019, 675–676; Morgül 2022, 19–20; Yanaşmayan et al. 2019, 47).

Finally, the third group of studies highlights the changes in JDP’s refugee discourse since late 2015, when Ankara took steps to completely close Turkey’s border with Syria to migrant crossings. Here we find discussions about two main changes. First, several researchers stress that a utilitarian perspective has emerged in JDP’s refugee discourse, especially after Erdoğan’s July 2016 announcement that Syrian migrants might be granted citizenship (İçduygu et al. 2017; McCarthy 2021; Özdemir Taştan and Çoban Keneş 2019). Prioritizing national economic interests, this perspective justifies the settlement and integration of Syrian refugees with needed skills in Turkey on account of their potential contributions to the country’s development. Since this pragmatic attitude pays no heed to historical, religious, or moral concerns, some scholars view it as a major transformation in JDP’s discourse on Syrian refugees. Analyzing statements by prominent JDP members, for instance, İçduygu et al. (2017, 464) identify a “shift” in the party’s rhetoric “from one form of selective humanitarianism based on shared culture and religion” to another based on “the need to integrate skilled, educated Syrian refugees.”

Second, some analysts note that JDP officials have placed more emphasis on the return of Syrian refugees in recent years, especially due to the surging public hostility toward migrants under rapidly deteriorating economic conditions. For example, Yanaşmayan et al. (2019, 42) observe that, unlike its 2015 election manifesto, JDP’s manifesto for the 2018 presidential and parliamentary elections allocated ample space to the issue of refugees, referring to “the safe return of a considerable number of migrants currently under temporary protection” as a “fundamental aim” for the party. Comparably, Balkılıç and Teke Lloyd (2021, 371) argue that JDP officials began to underline the Syrians’ temporary status in Turkey much more strongly after the ruling party suffered dramatic losses in the 2019 municipal elections – losses that were partly attributed to the growing public discontent with JDP’s “overly accommodating” stance on Syrian refugees. Examining this discursive shift more closely, Şahin Mencütek (2021, 2815–2817) contends that, in addition to addressing electoral concerns, JDP officials have also utilized the issue of refugee returns to legitimize Turkey’s military operations in northern Syria.

The existing literature makes important contributions to our understanding of JDP’s strategic discourse on Syrian refugees by identifying its central themes, metaphors, and shortcomings. However, previous studies are not without limitations. First of all, most research in this area relies exclusively on qualitative methods, using either small convenience samples of speeches by different JDP leaders or a particular class of texts (e.g., parliamentary debates) purposively selected from specific periods corresponding to major domestic and international events. Consequently, these works fail to assess the relative frequency of the themes and metaphors they have identified in JDP leaders’ political messages on Syrian refugees. Second, and relatedly, due to their methodological designs, prior studies also fall short of reliably gauging the shifts and continuities in JDP’s refugee discourse. This is a significant limitation given the protracted nature of the Syrian refugee crisis and the fluctuations in the JDP government’s policy responses to it. Lastly, while some studies have highlighted Ankara’s growing emphasis on the repatriation of Syrian refugees and connected it to JDP’s shifting domestic and foreign policy considerations, no research has yet explored how this emergent repatriation discourse is integrated into, or juxtaposed against, the Erdoğan regime’s Islamist and neo-Ottomanist narratives on refugees.

Data and methods

The analyses presented in this article are based on a corpus of 382 speeches that President Erdoğan gave from September 2014 through December 2022. To produce the corpus, I first extracted all speeches containing the keyword Suriye* (Syria*) from the website of the Presidency of the Turkish Republic. ¹ Using this keyword as a filter served to remove irrelevant speeches from the corpus and yielded a total of 599 texts. To further eliminate speeches that only mention Syria in passing but do not talk about Syrian migrants, I implemented a dictionary-based automated coding method on these 599 texts. The dictionary, which was created in light of prior research on JDP’s refugee discourse, included the following terms: Suriyeli* (Syrian*), göç* (migra*), sığın* (tak*/took refuge), sığınmacı* (asylum seeker*), mülteci* (refugee*), muhacir* (muhajir*), ensar* (ansar*), mağdurlar* (victims), mazlumlar* (the oppressed), geçici koruma* (temporary protection), açık kapı (open-door), and güvenli bölge* (safe zone*). Speeches that do not include any of these terms were dropped from the corpus, along with speeches that returned only false positives.

To analyze the corpus, I relied on a sequential mixed methods design whereby qualitative analysis builds on the initial quantitative findings. The quantitative phase itself proceeded through several steps. First, to reveal the ways in which Erdoğan has referred to Syrian refugees, I examined how often the term Suriye* occurs in the same sentence with a range of labels commonly applied to the Syrians such as mülteci*, sığınmacı*, misafir* (guest*), and kardeş* (brother*). In the Results section, I report both the raw co-occurrence frequency for each pair (e.g., Suriye* + misafir*) and the total number of speeches involving that pair.

Then, I employed dictionary-based automated content analysis to uncover the prevalence in Erdoğan’s rhetoric of four main discourses identified in the above-reviewed literature: (i) the discourse of religious and historical fraternity; (ii) the discourse of protecting the oppressed; (iii) the discourse of repatriation; and (iv) the discourse of national economic interests. Below are the terms used in each dictionary:

• Fraternity: ensar*, muhacir*, Suriye* + kardeş* pair.

• Protecting the oppressed: mazlumlar*, mağdurlar*, ezilenler* (the downtrodden), garipler* (the wretched).

• Repatriation: geri/evine/evlerine/ülkesine/ülkelerine/yurduna/yurtlarına/Suriye’ye dön* (return* home/to their homes/to their country/to their countries/to their homeland/to their homelands/to Syria), geri gönder* (send/t back), güvenli bölge*.

• National economic interests: eğitimli (educated), okumuş (learned), kalifiye (qualified), beceri* (skill*), kabiliyet* (talent*), meslek* (occupation*), mimar* (architect*), mühendis* (engineer*), avukat* (attorney*), hukukçu* (lawyer*), doktor* (doctor*), hemşire* (nurse*), öğretmen* (teacher*).

For each of these four discourses, the speeches were coded “1” if they contain any of the terms in the corresponding dictionary, and “0” if otherwise. ² The Results section presents the percentage of speeches involving these discourses across four distinct periods. The first period, 2014–2015, involves a dramatic surge in refugee flows from Syria and a growing recognition by government officials that the asylum of Syrian migrants in Turkey was turning into a protracted refugee situation (İçduygu et al. 2017, 452). In this period, we also see the formalization of the TP regime along with a gradual tightening of Turkey’s Syrian border in response to security concerns (Toğral Koca 2015). The second period, 2016–2017, involves a major shift toward restrictionism in Ankara’s asylum policies, with the fully fledged securitization of the Syrian border (Okyay 2017), the European Union–Turkey deal to curb the flow of migrants to Europe (Okyay and Zaragoza-Cristiani 2016), and the start of military interventions aimed at creating a “safe zone” in Syria (Oztig 2022). Beginning with this period, however, we also observe tentative steps toward integrating Syrian refugees into Turkish society (Makovsky 2019). In the third period, 2018–2019, the Turkish economy starts to falter, and public resentment toward refugees soars (Erdoğan 2020; Morgül et al. 2021; Morgül and Savaşkan 2021). Concurrently, the government intensifies its cross-border operations to prepare the ground for a large-scale return of Syrian refugees (Şahin Mencütek 2021). Finally, in the last period, 2020–2022, the refugee situation becomes entangled in, and is somewhat eclipsed by, a complex web of intersecting crises, from the COVID-19 pandemic to hyperinflation to political turmoil.

As the last quantitative step, I ran three binary logistic regressions that respectively predict the odds of a speech involving the discourse of fraternity, the discourse of protecting the oppressed, and the discourse of repatriation. ³ The regression models included period dummies (reference category: 2014–2015) as their key predictors, while controlling for audience type (1: domestic, 0: international), the overall length of the speech (total number of words), and the weight given to the refugee issue in the speech (total number of occurrences for the dictionary terms used to down-sample the corpus).

Building on the quantitative findings, I conducted a CDA of thirty speeches from the 2018–2022 period – the period in which President Erdoğan began to put a growing emphasis on the repatriation of Syrian refugees. To ensure that the selected speeches involve a thorough discussion of the refugee issue and are evenly distributed across time, I selected six speeches from each year with the highest value for the weight given to the refugee issue in the speech variable. This yielded a total of thirty texts for close textual analysis. In addition, I read all passages from the 2018–2022 period in which Erdoğan mentions the repatriation of Syrian refugees.

CDA, in a nutshell, entails an in-depth analysis of texts within their context (Flowerdew 2018). Scholars in this tradition integrate their knowledge of the social and political background into the analysis of texts to uncover how power relations are manifested in verbal or non-verbal communication (Wodak 2001). They also examine texts within their intertextual and interdiscursive context to highlight how they draw upon, revise, or reframe other texts and discourses (Fairclough 2003, 47–55, 127–133). In carrying out these analyses, critical discourse analysts pay particular attention to how social actors construct collective identities and draw symbolic and moral boundaries between ingroups and outgroups (van Dijk 1997).

Following these general principles, I examined Erdoğan’s remarks on the repatriation of Syrian refugees within their sociopolitical and interdiscursive context. I collected information on the sociopolitical context of refugee returns from both primary and secondary sources, including public opinion polls and reports by various UN agencies and international and national non-governmental organizations (NGOs). I collected information on the interdiscursive context by examining the key legal and policy documents that define the international normative framework on refugee returns. This allowed me to investigate how Erdoğan has strategically appropriated the language of international law and standards to legitimize his government’s unilateral efforts to repatriate Syrian refugees.

All coding procedures were carried out using MAXQDA Analytics Pro 2022. The statistical analyses were performed using STATA/SE 17.

Results

Quantitative results

I start the analysis by exploring the labels that Erdoğan has employed to refer to displaced Syrians. Table 1 presents the results. Here we can see that the label that Erdoğan has used most frequently for Syrian migrants is by far “our brothers.” The president has used this kinship metaphor to denote Syrian refugees 319 times in 154 different speeches. In second and third place are the terms “refugees” and “asylum seekers,” which Erdoğan has employed a total of ninety-two and eighty-six times, respectively. Table 1 also shows that, consistent with his efforts to portray Turkey as a refuge for persecuted people, the Turkish president has referred to the Syrians as “the oppressed” or “victims” seventy-eight times in fifty-one speeches. Notably, Erdoğan has labeled Syrian migrants as “guests” relatively infrequently: sixty-four times in fifty-three speeches. The two labels at the bottom of the list are “migrants” (twenty-nine times), and “muhajirs” (twenty-two times).

Table 1. Labels used by Erdoğan to refer to Syrian migrants

As Table 1 makes clear, the guest metaphor has not been particularly salient in Erdoğan’s rhetoric on Syrian refugees, at least since he became the president in 2014. Nor is there any indication that Erdoğan has avoided the terms “refugees” or “asylum seekers” when referring to displaced Syrians. These results contradict earlier studies which argue that JDP officials have refrained from using internationally recognized legal categories to define Syrian migrants. Nevertheless, it is true that Erdoğan’s preferred label for the Syrians, i.e., “our brothers,” is a legally empty category, for it does not invoke any rights and protections afforded to migrants under domestic or international law.

Next, I present my findings regarding the prevalence and trajectory of various discourses in Erdoğan’s remarks about Syrian refugees. As can be seen in Table 2, both the discourse of fraternity and the discourse of protecting the oppressed have been highly prominent in Erdoğan’s strategic framing of the Syrian refugee crisis: The former discourse is present in roughly 46 percent of his speeches on refugees, and the latter in about 45 percent. Moreover, neither discourse exhibits a clear upward or downward trend over time; rather, they show a relatively continuous pattern across the four periods. These results point to the enduring salience of Islamist and neo-Ottomanist themes in Erdoğan’s discourse on Syrian refugees.

Table 2. Prevalence (%) of four basic discourses in Erdoğan’s rhetoric on Syrian refugees, 2014–2022

Note: The 2014 data begin with September.

By contrast, the discourse of repatriation displays a striking upward trend. In the 2014–2015 period, only about 6.6 percent of Erdoğan’s speeches on Syrian refugees mentioned their eventual return to Syria as the desired outcome, whereas in 2016–2017, close to 13 percent did. More dramatically, this figure increased by a factor of almost 3.5 in 2018–2019, exceeding 44 percent. Despite a modest decline in the last period, the discourse of repatriation still featured in over 35 percent of all speeches. In other words, compared with the 2014–2015 period, Erdoğan was about 5.3 times more likely to mention the repatriation of Syrian migrants in his recent speeches. These results corroborate previous studies that highlight the increased emphasis on repatriation in JDP’s refugee discourse. It should be stressed, however, that this emergent repatriation discourse has not superseded but rather coexists with the discourse of fraternity and the discourse of protecting the oppressed. Hence, what has happened is not a U-turn in JDP’s refugee discourse but rather a reconfiguration of its elements.

Finally, we can see in Table 2 that the discourse of national economic interests has occupied only a marginal position in Erdoğan’s statements on Syrian refugees. In fact, I was able to find only six speeches (1.57 percent of the total) in which Erdoğan used an economic rationale to justify the long-term settlement and integration of at least some Syrian migrants in Turkey. While having a quantitatively marginal position in the full corpus does not mean that a particular discourse is qualitatively insignificant, these results cast doubt on the argument that economic justifications for the local integration of skilled Syrian migrants represent a “major shift” in JDP’s framing of the refugee issue.

Overall, these findings indicate continuity in President Erdoğan’s emphasis on the religious and historical fraternity between Turkish citizens and Syrian refugees, as well as in his assertions about Turkey’s restored willingness and ability under his leadership to protect the oppressed around the world. At the same time, we see a remarkable increase in his emphasis on the Syrians’ eventual return to their country, especially beginning with the 2018–2019 period. Table 3 reproduces these findings with binary logistic regressions that control for the variables audience type, the overall length of the speech, and the weight given to the refugee issue in the speech.

Table 3. Logistic regressions predicting fraternity, protection, and repatriation discourses

Note: Standard errors are in parentheses. Continuous variables are standardized to have a mean of 0 and a standard deviation of 1.

*p < 0.05, **p < 0.01 (two-tailed).

^† p < 0.10 (two-tailed).

As Table 3 shows, none of the period indicators predicts the likelihood of a speech involving either the discourse of fraternity or the discourse of protecting the oppressed. The only variables that yield statistically significant results for these two outcomes are audience type and the weight given to the refugee issue in the speech: Erdoğan was more likely to use these two discourses when he addressed domestic audiences and when his speeches allocated a greater space to the refugee question.

Conversely, two of the three period indicators are strongly associated with the likelihood of a speech involving the repatriation discourse, even after controlling for audience type, the overall length of the speech, and the weight given to the refugee issue in the speech. As the insignificant coefficient on the dummy variable 2016-2017 indicates, the first two periods (2014–2015 and 2016–2017) become statistically indistinguishable once we account for the control variables. However, there is a large and statistically significant difference between the first period on the one hand and the third (2018–2019) and fourth (2020–2022) periods on the other. It should also be noted that the difference between the third and fourth periods does not reach statistical significance at the conventional levels (z = 1.64, p > |z| = 0.101), and that both the third (z = 4.48, p > |z| = 0.000) and the fourth (z = 3.17, p > |z| = 0.002) periods are significantly different from the second period.

To facilitate the interpretation of these findings, Figure 1 gives the predicted probability for the presence of the repatriation discourse in a speech by period, while holding all other variables at their means. As the figure indicates, the predicted probability that a speech involves the repatriation discourse drastically increases after 2016–2017, exceeding 0.46 (46.35 percent) in the 2018–2019 period.

Figure 1. Predicted probability (Pr) of the repatriation discourse by period.

Note: All covariates are held at their means. Bars represent 95% confidence intervals.

Qualitative results

The enduring salience of Islamist and neo-Ottomanist themes

The quantitative findings discussed in the previous section suggest that Islamist and neo-Ottomanist themes remained salient in Erdoğan’s refugee discourse even after he began to push for a large-scale repatriation of Syrian refugees. My qualitative findings corroborate this claim. Indeed, even as late as December 2022, Erdoğan was stressing the religious bonds between Turkish people and their “Syrian brothers” and criticizing Kemal Kılıçdaroğlu, the leader of the main opposition Republican People’s Party (Cumhuriyet Halk Partisi), for taking anti-refugee positions:

We know the Ansar; we know the Muhajir, too. We’ll make no concessions on this. Mr. Kemal may not know, but we do. He has no such concerns. He neither understands the Ansar nor the Muhajir. My Prophet was a muhajir, wasn’t he? As a muhajir, he went to Medina from Mecca. Now, as his ummah, we know those who came to us as muhajirs. As Ansar, yes, we know how to open our arms. We’ve never considered the identity of the oppressed, and we do not today. ⁴

Similarly, despite Ankara’s intensifying efforts to repatriate Syrian refugees, Erdoğan did not stop claiming that under his leadership Turkey had once again become a powerful guardian of oppressed and downtrodden peoples – one that stood in stark contrast to Western apathy and xenophobia toward victims of humanitarian crises:

If fire has fallen on our neighbors, we cannot close our doors and windows and say, “It’s none of our business.” As the descendants of ancestors who distributed justice to the world for centuries, we cannot say, “We don’t care about Jerusalem, Palestine, Syria, Iraq, Yemen, Turkestan.” In particular, we cannot ignore the suffering and tears of others like Westerners do. Because we are a great state that is renowned for its history, culture, and values. ⁵

How has Erdoğan sought to reconcile his growing emphasis on sending Syrian refugees back home with these Islamist and neo-Ottomanist themes? How can he continue to claim moral superiority over his domestic opponents and foreign counterparts while the JDP government is actively pressing for a mass repatriation of displaced Syrians? The analysis presented below reveals that Erdoğan has strategically appropriated the language of international law on refugee returns to achieve these goals.

The discursive appropriation of international norms on refugee returns

International law recognizes everyone’s right of voluntary return to their country of origin or citizenship. Included in the Universal Declaration of Human Rights as a fundamental norm, the right of return has also been inscribed in several binding international legal instruments, such as the International Covenant on Civil and Political Rights. ⁶ Building on this principle, numerous resolutions adopted by the UN Security Council and General Assembly have upheld the right of refugees and internally displaced persons to return home. As such, the right of displaced persons to return to their homes or places of habitual residence is well established in customary international law.

International law also provides a framework of rights and obligations for refugee repatriation. Accordingly, three main conditions must be met for refugee repatriation to be humane and lawful: voluntariness, safety, and dignity. These conditions or principles have been highlighted by many UN resolutions and policy statements since the early 1990s. For instance, in a 1994 resolution on the former Yugoslavia, the UN Security Council affirmed “the right of all displaced persons to return voluntarily to their homes of origin in safety and dignity with the assistance of the international community.” ⁷ In a 2006 resolution on assistance to refugees, returnees, and displaced people in Africa, the UN General Assembly reaffirmed the principle of voluntary repatriation, noting that it “can be accomplished in conditions of safety and dignity.” ⁸ Recently, UNHCR (2018) has reiterated that any plans for large-scale repatriation of Syrian refugees should be grounded in these principles.

The principle of voluntariness follows from Article 33 of the 1951 Refugee Convention, which prohibits states from expelling a refugee to any territory “where his life or freedom would be threatened on account of his race, religion, nationality, membership of a particular social group or political opinion.” ⁹ For UNHCR, voluntariness does not only mean the absence of physical coercion; it also means lack of psychological and material pressures. According to a handbook on voluntary repatriation issued by the organization, this condition is met when “the positive pull factors in the country of origin are an overriding element in the refugees’ decision to return rather than possible push factors in the host country or negative pull factors, such as threats to property, in the home country” (UNHCR 1996, 11–12). Additionally, voluntariness requires informed consent: “Refugees and displaced persons should be provided with complete, objective, up-to-date, and accurate information, including on physical, material, and legal safety issues in countries or places of origin.” ¹⁰

Like voluntariness, the principle of safety has multiple dimensions. As explained in another UNHCR handbook issued in 2004, it includes legal and material safety as well as physical safety. Among other things, legal safety entails the “removal of legal and administrative barriers to return” and the presence of “mechanisms to redress human rights abuses, including independent judiciary.” Material safety, on the other hand, involves access to “means of survival and basic necessities in early stages of return” and the “promotion of economic self-reliance and income generating activities” in the longer term (UNHCR 2004, 4). Unlike the other two principles, dignity is not well defined. It is generally agreed, however, that voluntariness and the various facets of safety are necessary for dignified returns. Some have argued that dignity also requires the involvement of refugees in decision-making processes as autonomous agents so that they can influence the nature, timing, and conditions of their repatriation (Long 2013, 169–174).

In his remarks on the repatriation of Syrian refugees, Erdoğan frequently references these legal norms and principles, starting with the right of displaced persons to return home. The following example is from his speech at the first Global Refugee Forum held in Geneva in December 2019. Here Erdoğan frames Turkey’s efforts to repatriate Syrian refugees as a contribution to not only the normalization of Syria but also the realization of an “indisputable” human right:

Formulas should be put into effect to keep refugees in their own lands and to return those in our country back to their homeland. The right of refugees to return to their homelands is indisputable. Returns are as important as the fight against terrorism for establishing permanent stability and normalization in Syria. ¹¹

When talking about refugee returns, Erdoğan also regularly refers to the principles of voluntariness, safety, and dignity. For example, after he met with Russia’s Vladimir Putin and Iran’s Hasan Rouhani in Ankara in September 2019, the president said this:

It is obvious that Turkey cannot carry the refugee burden on its own. Our country cannot handle a new migration influx. Now we need to focus on the safe and voluntary return of Syrians to their countries. The peace corridor east of the Euphrates will also be a sheltered harbor for refugees. We think that at least 2 million of our Syrian brothers who took refuge in our country can be resettled in this region. ¹²

Likewise, when Erdoğan asked for support from the international community for Ankara’s plans to repatriate displaced Syrians in his address to the UN General Assembly in September 2021, he made sure to depict the refugee returns from Turkey as “voluntary, safe, and dignified”:

We welcome the twelve-month extension of the United Nations humanitarian aid mechanism delivered to northwest Syria via Turkey. We hope that the conciliatory approach displayed on this issue will also be put forward in order to advance the political process and ensure the voluntary, safe, and dignified return of asylum seekers. ¹³

Erdoğan makes similar remarks also when addressing domestic audiences. For instance, answering student questions in a program organized as part of the World Human Rights Day at Bilkent University, he assured the audience that his government was determined to repatriate Syrian refugees but also stressed that the returns would be on a voluntary basis:

What have 360,000 Syrians done on a voluntary basis? They have returned and are now living in their own houses in Jarablus. But unfortunately, they [the opposition parties] are spreading the propaganda that [Syrian refugees] will not be sent back, that they will remain on our lands. There is no such thing, but we’re in favor of sending them back on a voluntary basis, and so far, this has been successful in Jarablus. ¹⁴

Erdoğan’s repeated references to international norms on refugee returns serve a strategic purpose: They represent Ankara’s efforts to repatriate Syrian refugees as a natural extension of Turkey’s altruistic stance on displaced persons and other victims of humanitarian crises. This in turn allows Erdoğan to claim the moral high ground while simultaneously advocating mass repatriation of the Syrians. Notice how the correlative conjunction “Not only … but also” in the quotation below constructs a moral equivalence between hosting refugees and repatriating them:

Not only have we opened our doors to save the lives and dignity of the oppressed, but we’ve also made every effort to help them return to their homes. Since 2016, when Turkey started its cross-border operations in the face of the deepening humanitarian tragedy in Syria, approximately 500,000 Syrians have returned to the safe zones we’ve created. ¹⁵

In the example below, Erdoğan similarly creates a moral equivalence between his government’s various policies toward displaced persons – from welcoming asylum seekers to repatriating refugees to preventing new waves of migration – by listing them in an additive fashion:

We’ve ensured the return of 411,000 Syrian brothers who took refuge in Turkey to their hometowns. With our presence in Idlib, we’ve averted a new humanitarian tragedy and a great migration wave. We’re hosting 4.5 million refugees in our country, and we’re caring for as many in Syria. Only Turkey is doing this; there’s no other country like it in the world. ¹⁶

Are the returns from Turkey voluntary, safe, and dignified?

President Erdoğan uses two main discursive strategies to claim that Turkey’s repatriation of Syrian refugees meets the principles of voluntariness, safety, and dignity. First, he periodically updates the number of returnees, each time increasing the number he gives. In this way, he communicates both the willingness of Syrian refugees to return to “their homes” and the safety of the territories under Turkey’s control. The following quotations, one from April 2018 and the other from February 2021, illustrate this point:

160,000 Syrians have returned to the Euphrates Shield region. We see that hundreds of thousands of our Syrian brothers will return there after we completely clear the areas we control in Afrin from explosives and terrorist residues. Hopefully, we’ll also make Idlib, Tel Rifat, Manbij, Ayn Al Arab, Tel Abyad, Ras al-Ayn, and Qamishli safe and ensure that all Syrians reunite with their homes. ¹⁷

As the security environment in northern Syria improves, the safe and voluntary return of the Syrians in our country to their homelands accelerates. So far, 420,000 refugees from our country have returned to the regions we’ve cleared of terrorism. As the political solution efforts in Syria progress and the atmosphere of stability and peace gets stronger, these returns will increase even more. ¹⁸

Second, Erdoğan regularly describes Turkey’s housing and infrastructure investments in northern Syria to demonstrate his government’s determination to ensure that the returns take place in conditions of material safety and dignity. See, for instance, the following remarks he made after a cabinet meeting in May 2022:

Last year, with the support of non-governmental organizations, we started a project to build 100,000 briquette houses for people living in very difficult conditions on Syrian territory. … With the financing of international aid organizations, we hope to ensure that 200,000 houses with all the necessary infrastructure – including schools, hospitals, industry, agriculture – are built in thirteen different places in Syria. We aim to ensure the return of 1 million Syrians still living in our country to these cities with all humanitarian conditions. ¹⁹

Are the returns from Turkey to Syria really voluntary, safe, and dignified? The answer is a resounding “No.” To begin with, multiple issues undermine Erdoğan’s claim that the returns are taking place on a voluntary basis. First, it should be underlined that Erdoğan offers a very narrow definition of voluntariness, reducing it simply to the absence of physical coercion. As noted above, however, UN agencies interpret voluntariness more broadly to also include the absence of psychological and material pressures on refugees. Given the exceedingly high levels of public hostility toward refugees in Turkey (Erdoğan 2021), as well as their precarious legal and socio-economic situation (Baban et al. 2017), it is hard to conclude that the returns are truly voluntary. Moreover, since Ankara operates unilaterally to repatriate Syrian refugees, UNHCR is not fulfilling its legally mandated role of verifying the voluntary character of returns and ensuring that refugees are provided with accurate information on physical, material, and legal safety issues in Syria (Şahin Mencütek 2019, 30). This raises doubts as to whether the Turkish government honors the informed consent requirement in its refugee repatriation initiatives.

Second, reports by various international and domestic NGOs suggest that Turkey has violated even the most basic element of voluntariness – the absence of physical coercion – on multiple occasions, especially after JDP’s stinging losses in the 2019 local elections. A report by Amnesty International (2019, 5), for instance, estimated that in mid-2019 “hundreds of people across Turkey were swept up, detained, and transported against their will” to northwestern Syria. In the same year, a press release by Human Rights Watch (2019) drew attention to the rising numbers of arbitrary detentions and deportations in İstanbul and Antakya, citing allegations by some deportees that they were pressured into signing “voluntary return” forms through the use or threat of force.

And third, reliable surveys of the Syrians in Turkey indicate that the vast majority do not want to go back to Syria in the near or distant future. Most notably, relying on a sample of 1,414 refugee households from fifteen cities, the Syrian Barometer 2020 shows that interest in voluntary repatriation among Syrian refugees has declined dramatically over the past few years. In a question probing their views on returning to Syria, a whopping 77.8 percent of the respondents picked the option that reads “I do not plan to return to Syria under any circumstances,” whereas only 16.7 percent had picked the same option in 2017. Moreover, in a question about plans for return within the next twelve months, only 3.7 percent said, “I plan to return” (Erdoğan 2021, 230–232). These findings refute President Erdoğan’s recurrent assertion that most Syrian refugees look forward to going back home.

The case for safety and dignity is even weaker than that for voluntariness. Indeed, studies conducted by UN agencies and various international NGOs all emphasize that Syria remains unfit for safe and dignified returns. A report by UNHCR (2018, 3), for example, states that “present conditions in Syria are not conducive for voluntary repatriation in safety and dignity as significant risks remain for civilians across the country.” In the same manner, a large-scale study carried out under the auspices of the Voices for Displaced Syrians Forum concludes that “conditions in Syria are not safe for return and are unlikely to become safe in the foreseeable future” (Operations and Policy Centre 2021, 9). Importantly, this study documents grave socio-economic problems and human rights violations in all of the four major control areas in Syria, including territories controlled by the Turkey-affiliated Syrian Interim Government and its military force, the Syrian National Army (SNA). Similar issues are also highlighted in various reports of the Independent International Commission of Inquiry on Syria, which was established in August 2011 by the UN Human Rights Council. In its latest report, the Commission accuses the SNA (like other key players in Syria) of committing war crimes and observes that “Turkish forces may have violated their obligations” under international law by participating in SNA-led interrogations and failing to prevent the ill-treatment of detainees. ²⁰

In summary, although Erdoğan repeatedly mentions international law and standards on refugee returns, there is ample evidence to conclude that Turkey’s efforts to repatriate Syrian refugees do not meet the principles of voluntariness, safety, and dignity. It could thus be argued that Erdoğan’s references to these principles have been motivated by strategic calculations, rather than a genuine concern for the well-being of displaced Syrians.

Conclusion

In this study, I investigated the official discourse on Syrian refugees in Turkey through a mixed methods analysis of 382 speeches that President Erdoğan gave between September 2014 and December 2022. Examining basic lexical co-occurrence patterns for the term Syria*, I first showed that Erdoğan has referred to the Syrians most commonly as “our brothers,” which is a legally empty category despite being a kinship metaphor of high sentimental value. I also showed that the “guest” metaphor has not been particularly salient in Erdoğan’s rhetoric on Syrian migrants, and that he has not avoided internationally recognized legal categories such as “refugees” and “asylum seekers” when referring to displaced Syrians.

I then turned to the examination of continuity and change in Erdoğan’s strategic framing of the refugee crisis. Combining dictionary-based automated content analysis with binary logistic regressions, I reached three main conclusions. First, Islamist and neo-Ottomanist themes, which not only highlight the religious and historical fraternity between Turkish citizens and Syrian refugees but also portray Turkey as a powerful guardian of oppressed peoples, have played a central role in Erdoğan’s discursive governance of the refugee crisis throughout his presidency. Second, since 2018, there has been a remarkable increase in the frequency with which Erdoğan invokes repatriation as the preferred solution to the refugee crisis, though this has not led to a weakening of the discourse of fraternity or the discourse of protecting the oppressed. And third, only on a handful of occasions has Erdoğan used a discourse of national economic interests to justify the long-term settlement and integration in Turkey of Syrian refugees who have higher levels of education, desirable skills, or financial resources.

Finally, I carried out a CDA of Erdoğan’s remarks from the 2018–2022 period in order to examine how the president has sought to reconcile his growing emphasis on repatriating Syrian refugees with his continuing efforts to exploit Islamist and neo-Ottomanist tropes. This analysis revealed that Erdoğan has strategically appropriated the language of international law and standards on refugee repatriation through repeated allusions to the right of displaced persons to voluntarily return to their homes in safety and dignity. In doing so, he has portrayed Ankara’s unilateral efforts to repatriate displaced Syrians as a natural extension of the JDP government’s humanitarian stance on refugees. This discursive move, in turn, has allowed Erdoğan to maintain his claims of moral superiority over his domestic opponents and foreign counterparts while simultaneously advocating mass repatriation as the desired solution to Turkey’s refugee problem.

Like other research, this article has its limitations. For one, although Erdoğan is by far the most influential JDP leader, the official discourse on Syrian refugees in Turkey cannot be reduced to his speeches. It would, therefore, be informative to probe whether Erdoğan’s refugee rhetoric diverges from that of other JDP leaders at the national and local levels, and if so, why. For another, while this study alludes to several discrepancies between Erdoğan’s refugee discourse and his government’s actual refugee policies, it does not fully explore those discrepancies and their political implications. To give an example, the content analysis shows that economic utilitarianism is the weakest element in Erdoğan’s refugee discourse; yet since 2016, over 200,000 Syrians have gained Turkish citizenship, mainly on the basis of their socio-economic qualifications (Şimşek 2022). Future research should examine such gaps between JDP’s discourse and practice on refugees more closely and elucidate the various political and economic interests underpinning them.

Despite these limitations, however, this article contributes to the literature on JDP’s discursive governance of the Syrian refugee crisis through a mixed methods analysis of the speeches given by Erdoğan throughout his presidency. Implementing dictionary-based automated coding on a systematically generated corpus, the article offers a reliable account of continuity and change in the president’s discourse on Syrian refugees. Carrying out a CDA of carefully sampled texts, the article then provides a detailed explanation for how Erdoğan has sought to resolve the tension between his increasingly loud calls for repatriating Syrian refugees and his continued appeals to Islamist and neo-Ottomanist themes.

The findings of this study have broader implications for the literature on populism and migration. First of all, whereas existing research highlights nativist hostility toward refugees and immigrants as a core characteristic of the populist right in Western countries (Mudde 2007), the present study demonstrates that this does not necessarily apply to right-wing populists in other parts of the globe. In fact, the case of Erdoğan shows that populist leaders with authoritarian and nationalist tendencies may adopt a relatively inclusionary stance toward some migrant groups, especially if such a stance is aligned with their national narratives or domestic and foreign policy objectives.

Second, the shifts in Erdoğan’s refuge discourse, especially his increased emphasis on refugee returns in the context of rising public resentment toward migrants, reveal that even the most authoritarian and powerful populist leaders have to consider popular demands when articulating a particular story of peoplehood and its constitutive “us” versus “them” dichotomies. This draws attention to the relational and dynamic nature of populist communication, reminding us that “people-making” is an open-ended process which takes place within evolving social, political, and economic circumstances (Smith 2003).

Finally, this paper illustrates how opportunistic leaders in host countries may appropriate the language of the international legal framework on refugee returns in order to legitimize premature and politically motivated repatriation initiatives. Since we live in an era of protracted refugee situations and rising anti-refugee movements, it is quite likely that many host governments will resort to similar discursive strategies to justify mass returns despite the absence of basic safeguards. It is thus incumbent upon UNHCR and the wider international community to take a more assertive role in upholding internationally recognized legal and ethical principles regarding refugee repatriation.