Examining the ethos of conflict by exploring lay people’s representations of the Kurdish conflict in Turkey

Abstract

Societies in conflicts develop an ‘ethos of conflict’, a set of socially shared beliefs about the conflict. We argue that the ethos of conflict can be based on different representations of the conflict, and exploring such representations helps to analyze similarities and differences between and within conflict parties. We explored representations of the Kurdish conflict among 45 laypeople in a multi-ethnic city in Turkey based on comprehensive models of conflict analysis using an approach based on Q-methodology, which is suitable for uncovering socially shared viewpoints. Representations of conflict were conceptualized along three domains: causes and issues; relationships between the groups, processes and dynamics; and possible solutions. An integrated analysis across these domains revealed five qualitatively distinct viewpoints toward the conflict. The meaning of these viewpoints, their possible links to the ethos of conflict, differences and commonalities within and between conflict parties, the usefulness of our methodological approach to explore conflict viewpoints, and implications for conflict resolution are discussed.

Keywords

Conflict analysis ethos of conflict Kurdish conflict Q methodology

As identity plays a crucial role in intergroup conflicts, many scholars and conflict resolution practitioners use the concept of identity in order to explain how people subscribe to dominant conflict narratives or representations. Often these types of studies result in two dominant conflict narratives representing two conflicting groups. However, these studies might represent social reality in an oversimplified way as there might be alternative conflict representations among (and within) conflicting groups. In the present study, rather than approaching groups as if they were homogenous entities, we aim to explore alternative conflict representations among conflicting groups, namely Kurds and Turks, together with another group (Arabs), in the Kurdish conflict context in Turkey.¹

Identity and intergroup conflict

Satisfying one’s needs in a conflict context can be achieved through the functions of identities. According to Fisher (2006: 180), “[i]dentity groups are seen as the primary vehicle through which these necessities are expressed and satisfied, thus leading to intergroup conflict when one group’s basic needs are frustrated or denied.” Such an identity perspective is also in line with social identity theory, which states that part of the self-concept derives from membership in social groups such as being a woman or being homosexual, and people are generally motivated to identify with and feel positive about their groups (Tajfel, 1974; Tajfel and Turner, 1986).

Group dynamics based on shared identities play a key role in the outbreak and development of a conflict. For example, Fisher (2006: 182) argues that “[t]he functioning of each group, in terms of identity, cohesiveness, conformity pressures, and decision making, has a significant impact on how conflict is played out and ultimately resolved or terminated.” As soon as the group members recognize a specific situation as a conflict, they try to understand the conflict situation and to find answers for the sources of the conflict, the forces responsible for the conflict, and characteristics of the opposing conflict party (Bar-Tal, 2000). At the end of this cognitive processing, conflicting parties form conflict beliefs, not in an objective, but a rather biased way (Bar-Tal, 1990).

From ethos of conflict to conflict representations

According to Bar-Tal (1998, 2007), societies adapt to the challenges of long-term violent intergroup conflicts by developing an “ethos of conflict”, a configuration of central, socially shared beliefs that explain and justify the conflict. These beliefs refer to eight interrelated themes: justness of one’s group’s goals; concern about security for one’s group; positive image of one’s group; victimization of one’s group; delegitimization and dehumanization of the adversary; positive emotional attachment to one’s group; unity within one’s group; and concern about peace. The ethos of conflict is related to the dominant or “master narrative” about the conflict (Hammack, 2006). It provides society members with a positive social identity and a clear view of the conflict and the conflict parties, which allows them to understand the current reality and to predict future developments of the conflict.

To measure empirically to what extent society members endorse the ethos of conflict, Bar-Tal et al. (2012) have developed a scale which is one-dimensional and internally consistent among Israeli Jews. It can distinguish between society members in terms of the extent to which they endorse (or reject) the ethos of conflict. However, the one-dimensional structure of the scale may suggest greater unanimity in society members’ understanding of the conflict than there actually is. It is an empirical question whether the ethos of conflict is based on only one dominant narrative, or representation, of the conflict. We suggest that it can be based on alternative representations, to which significant segments of the population in society subscribe to different degrees. Such multiple representations of conflict may capture the complexity of the conflict better than the ethos of conflict resulting from them.

Understanding how social categories or groups in conflict are constructed and perceived in the first place should be prioritized in conflict research (Reicher, 2004). Speaking about the “Turkish–Kurdish” or the “Israeli–Palestinian” conflict, for example, suggests that there are two (ethno-nationally defined) groups in conflict. Social psychological research on intergroup conflict (and intergroup relations more generally) has tended to focus on the two parties to a conflict as if they were two homogenous groups (e.g. “ingroup” and “outgroup”, or “majority” and “minority” groups). Variations within each group have usually been studied only in terms of individual differences; the (potential) presence of subgroups within each conflict party, or cutting across the conflict parties, has been neglected. This likely represents a large oversimplification of the actual social reality (Elcheroth and Spini, 2012). To avoid this, Cohrs (2012) proposed identifying, in the first place, the “ideological groups” that are relevant in a particular context (see also the concept of “opinion-based groups”; Bliuc et al., 2007). These groups are based on shared beliefs rather than, for example, common ethnicity, religion, or language. Taking into account the psychological meaning of the social categories in relation to the issue at hand (i.e. the subjectivity involved) might give important and alternative insights for the conflict context because it is these ideological or opinion-based groups that matter most for political action (Bliuc et al., 2007) rather than the conventional categories (e.g. being Kurdish or Turkish; see also Brown, 2002). This strategy is in line with insights from the conflict resolution literature that there often are different perspectives about the conflict within each conflict party (e.g. Cheldelin et al., 2003; Kelman and Fisher, 2003).

Therefore, the present research explored socially shared viewpoints toward a conflict using a methodological approach that allows for alternative representations, shared within ideological or opinion-based groups, to emerge from the data. The study was carried out among laypeople, namely Arabs, Kurds and Turks, in Turkey with regard to the Kurdish conflict. Our study is unique in several respects: (a) it makes voices of laypeople heard which are generally unheard in a conflict situation (Lederach, 1999; Uluğ and Cohrs, 2016a); (b) socially shared viewpoints of the Kurdish conflict that are suggested in the literature are not based on empirical studies (see, e.g. Çelik and Blum, 2007); however, approaching the Kurdish conflict empirically with a systematic conceptual model is crucial to understand laypeople’s priorities and concerns in conflict; (c) it illustrates conflict understandings from three different groups (Arabs, Kurds and Turks) who live together in a multi-ethnic region in Turkey, thereby going beyond the usual approach in social psychological research to study relations between two main groups in conflict; (d) it explores alternative representations of conflict rather than approaching conflict with a top-down concept such as the ethos of conflict in order to better capture the complexity of the conflict; and (e) it points out some socio-psychological factors to explain tentatively how people subscribe to these socially shared conflict viewpoints. In the following, we further describe the background to our research by introducing approaches to conflict assessment and conflict mapping as well as providing information about the Kurdish conflict.

Conflict assessment and conflict mapping

Different taxonomies and typologies have been suggested to describe conflicts between groups in a more structured way (Sandole, 2003). Such taxonomies and typologies rest on conflict analysis: conflict assessment and conflict mapping. Conflict assessment aims to obtain a comprehensive understanding of the conflict by eliciting different perspectives within and between the conflict parties (Shmueli, 2003). After having assessed a conflict, a conflict map can be produced. A conflict map provides a system to illustrate the scope of a conflict, and to identify possible interventions for conflict resolution (Shmueli, 2003; Wehr, 1979, 2006).

In our research, we used a combination of the approaches to conflict analysis by Coleman (2003) and Wehr (1979, 2006), and defined the scope of possible representations of conflict with reference to: (a) causes and issues in the conflict; (b) relationships, processes and dynamics of the conflict; and (c) possible solutions for the conflict.² These three domains were used as a compromise between comprehensiveness (to ensure a broad coverage of different conflict dimensions) and simplicity (to be practical empirically). The first two combine dimensions specified by Coleman (2003) and Wehr (1979, 2006); the third one was added to more directly capture views on conflict resolution (see also Cheldelin et al., 2003). The three domains are described below with reference to different perspectives on the Kurdish conflict. We assume that representations of the Kurdish conflict are characterized by systematic combinations of viewpoints across these domains.

The Kurdish conflict

The Kurdish conflict has led to the death of as many as 45,000 people since 1984. Over a million people have been displaced from east and southeast Turkey to cities in the south and the west (e.g. Mersin and Istanbul). To start with, Kurds were the targets of human rights abuses, mass killings, and unsolved murders in the southeast of Turkey, especially in the 1990s. One nationally representative survey conducted by KONDA (2011) indicates that 23.4% of the Kurds reported that they could not express their ethnic identity freely, whereas among the Turks it was only 3.2%. Turks, in turn, have claimed that Kurds have not shown enough solidarity with the Turkish state. Formally, Kurds have the same position as Arabs, the third largest ethnic group after Turks and Kurds (Koç et al., 2008), in being an ethnic minority without official minority status. However, despite this similarity, the modern history of Turkey has witnessed uprisings of Kurds, but not of Arabs, possibly because of different geographical, historical and linguistic characteristics.

Many different perspectives on the Kurdish conflict have been suggested by various scholars and political actors. There is no consensus on the definition of the problem, the role of different actors in the conflict, or strategies for conflict resolution. In the following, we describe the main perspectives with their commonalities and differences in terms of the three aforementioned domains.

Causes and issues in the conflict

Tan (2008) distinguishes between five main perspectives on the causes and issues in the conflict: denial of the conflict, terrorism, economy, identity, and nation state (see also Beriker-Atiyas, 1997; Çelik and Blum, 2007; Tocci, 2005). The “denial of the conflict” perspective argues that there is no conflict because people of Kurdish ethnic origin have the same rights as all other citizens. According to the “terrorism” perspective, the conflict is a terrorism and separation problem arising from an infringement by the UK and the US on the territorial integrity of the Turkish state. These “foreign powers” are believed to support the PKK (Partiya Karkeren Kurdistan/Kurdistan Workers’ Party; the “armed wing” of the Kurdish national movement) and terrorist activities. The “economy” perspective sees the conflict as an economic problem arising from the Kurds’ feudal structure and regional underdevelopment. From the “identity” perspective, the conflict is an ethnic identity problem owing to Turkey’s problems of democratization. The Turkish state adopted an assimilation policy toward the Kurds, denying their Kurdish identity and prohibiting the free expression of their language and culture (Cerrah, 1999; Gunter, 2000). Finally, the “nation state” perspective sees the conflict as a national problem where some Kurds advocate for an independent Kurdistan.

Relationships, processes, and dynamics of the conflict

Concerning the conflict parties and their relationships, some scholars have claimed that the conflict is between the Turkish state and the Kurdish people in Turkey (Tocci, 2005) while Yavuz and Özcan (2006) have suggested that the conflict has more recently developed into a conflict along ethnic lines, between the Turks and the Kurds. A reference to “foreign powers” as actors in the conflict has already been made above. In addition, third parties involved in the conflict and its resolution include non-governmental organizations, community groups, and civil society organizations (e.g. İnsan Hakları Derneği/Human Rights Association).

In terms of processes and dynamics of the conflict, Turkey’s long-lasting EU membership process has affected the conflict because Turkey’s EU membership is partly dependent on solving the Kurdish problem and demands on Turkey regarding human rights reforms and democratization have led to an increase in importance of the conflict for the current AKP government (Adalet ve Kalkınma Partisi - Justice and Development Party) (see Çelik and Rumelili, 2006).

Possible solutions for the conflict

Although the conflict has affected every part of society and the country, there is still no clear and agreed solution. The Turkish government launched a Kurdish “Initiative” in 2009 (Çandar, 2009) and a peace process in 2013 to pursue democratic reforms. One can infer from governmental statements that the aim of this initiative and peace process was to end the PKK insurgency and broaden the democratic rights of the Kurds. However, there is no clear roadmap for the initiative and peace process, and strong criticisms have been made against them, not only by Turkish nationalists, but also by Kurdish people owing to a deep mistrust of the Turkish state. As of 2016, the future of the peace process in Turkey remains vulnerable as attacks from both sides have been continuing.

As a step toward a solution, Çelik and Blum (2007) mention improving dialogue and trust between Kurds and Turks, which in turn would help to develop healthy communication. Çandar (2011) suggests that the development of trust is dependent on the PKK laying down their arms, and the Turkish Army subsequently stopping military operations against the PKK and their camps. The government has already approved an amnesty law to facilitate the integration process of former PKK members into Turkish society (Yavuz and Özcan, 2006). However, it is still debated whether the law will (and should) apply to Abdullah Öcalan, the founder and leader of PKK who was captured in 1999 and sentenced to lifelong imprisonment. Solutions could include the active inclusion of Öcalan in the solution process or granting him amnesty (Çandar, 2011).

In addition, economy-oriented solutions focusing on economic development in the Kurdish-populated regions have maintained importance (Çelik and Blum, 2007). Economic measures taken by the state may increase living standards and decrease unemployment rates (Karakaya-Polat, 2008).

The present research

The aim of our research was to identify and describe representations of the Kurdish conflict among a sample of laypeople in Turkey. We did not adopt the theoretically based definition of the ethos of conflict as a one-dimensional concept, as exemplified in the research by Bar-Tal et al. (2009, 2012). Instead, we chose an exploratory approach that allowed multiple, alternative representations of conflict to emerge from the data. We relied on comprehensive models of conflict analysis (Coleman, 2003; Wehr, 1979, 2006) as well as empirical material from various sources about the Kurdish conflict to define the scope of alternative representations of conflict, and used an approach informed by Q methodology (Brown, 1980; Watts and Stenner, 2005a). Although Q methodology can be applied in many areas such as social psychology (e.g. close relationships; Watts and Stenner, 2005b) and health (Stainton Rogers, 1991), it is ideally suited for our purposes and has been used in conflict contexts before (e.g. on national identity in Taiwan; Wong and Sun, 1998). This approach allowed us to explore how laypersons in Turkey, from different ethnic backgrounds, view the conflict; how these individual viewpoints toward the conflict are shared across individuals in different segments of the sample; how these viewpoints relate to one another in terms of their similarities and differences; and, tentatively, what socio-demographic factors are related to the different viewpoints. We did not make predictions as to how many and which representations of conflict would emerge as we used Q methodology for exploratory purposes. In addition, Q methodology does not aim to statistically compare group differences (Watts and Stenner, 2012).

Method

Participants

Q methodology requires purposive sampling of participants (typically, between 40 and 60 as Q sorting as a data collection method is both labor-intensive and time-demanding; Brown, 1980) in a way that maximizes the chances of finding multiple socially shared viewpoints, owing to different social milieus, socio-economic positions, etc. For our study, we selected participants based on their ethnic background, while maintaining diversity in terms of age, gender, educational level, and residential area. Participants were 15 Turks, 15 Kurds, and 15 Arabs from Mersin, an ethnically diverse region in the south of Turkey. The data were collected in 2011 in Mersin, first, because Mersin has received significant numbers of migrants from the east and southeast of Turkey owing to the conflict in that region and has become a highly polarized city with frequent political protests by both Kurds and Turks. There are some regions in Mersin where the Kurdish population is quite dense (e.g. Gündoğdu district). In these regions, there have been tensions between the residents and the police. In addition, there are also social tensions between Turks and Kurds that might suddenly escalate, as happened with the burning of the Turkish flag in March 2005 (Çelik, 2010). Second, Mersin is a multi-ethnic city that also has a large share of ethnically Arab citizens. Arab participants were chosen in addition to Turkish and Kurdish participants to assess what viewpoints toward the Kurdish conflict would emerge in members of an ethnic group that has played a less focal role. Another reason was to explore how another group (i.e. Arabs) position themselves towards the two focal groups in the conflict (see, e.g. Glasford and Calcagno, 2012, for a study on political solidarity between minority groups).

Participants were from 18 to 65 (mean 43) years old; 23 were female and 22 male. Twenty-four participants had completed a university degree, 12 high school, 5 secondary school and 4 primary school. All but one spoke Turkish at home, 14 Turkish and Kurdish (all Kurds), 8 Turkish and Arabic (all Arabs), and 1 Turkish, Kurdish, and Arabic (an Arab). Twenty-two participants indicated that they or their family had been affected by the conflict (9 in terms of death; 13 in terms of harm, pressure, or imprisonment); 16 indicated that they had not been affected; and 7 did not give an answer. The political views of the participants ranged from 1 (left) to 10 (right; mean 4.13, standard deviation 1.90).

Participants were recruited through snowball sampling (see Online Supplementary Table 1 for participants’ characteristics). Participants gave their informed consent to participate in the study. For each participant, a convenient setting for data collection (e.g. the participant’s home) was negotiated.

Procedure and materials

The data collection phase in Q methodology involves a process of “Q sorting.” Participants receive a set of statements, with each statement printed on a separate card (Brown, 1993; Watts and Stenner, 2005a). Participants are first asked to read all statements and sort all statements into multiple categories along a quasi-normal distribution, usually ranging from “most agree” to “most disagree,” with the middle category for statements of least psychological significance. At the end, participants are given the option to move cards around until they are satisfied that the Q sort accurately reflects their viewpoint. The resulting Q sort is thus a highly informative, holistic, contextualized representation of the individual’s viewpoint on the issue in question.

In our study, in line with the models of conflict analysis (Coleman, 2003; Wehr, 1979, 2006), we used separate Q sorting tasks for the three conflict domains: causes and issues in the conflict; relationships, processes, and dynamics of the conflict; and possible solutions for the conflict. For each domain, we formulated statements (in Turkish) based on material from the scholarly and political sources referenced above. To cover alternative representations of the Kurdish conflict more comprehensively, we searched additional sources for “pro-” and “anti-Turkish” and “pro-” and “anti-Kurdish” materials such as blogs, magazines, newspapers, and books (both in Turkish and English).³ The resulting lists of statements were then reduced iteratively to minimize redundancy while maintaining comprehensiveness. To avoid researcher bias in deciding which statements to keep, the authors consulted three other political psychology and conflict resolution scholars who were familiar with and had conducted research on the Kurdish conflict. Additionally, five laypersons were asked to evaluate the comprehensiveness of the statements and to nominate any potentially missing issues.

Causes and issues in the conflict

For the “causes and issues” part, we generated an initial list of 42 statements, which was subsequently reduced to 31 statements. Participants were asked to sort the statements according to how important versus unimportant they considered them in terms of causing and defining the Kurdish conflict. There were seven response categories, from most important (+3) to most unimportant (−3). A fixed quasi-normal distribution was required for each Q sort.

Relationships, processes, and dynamics of the conflict

For the “relationships, processes, and dynamics” part, we generated an initial list of 68 statements, which was reduced to 56 statements. Participants were asked to sort the statements according to how strongly they agreed or disagreed with them as representations of particular events, parties, relationships, and dynamics of the conflict. There were 11 response categories, from most agree (+5) to most disagree (−5).

Possible solutions for the conflict

For the “possible solutions” part, we generated an initial list of 41 statements, which was reduced to 30 statements. Participants were asked to sort the statements according to how effective versus ineffective they considered them as possible solutions for the conflict, with seven response categories, from most effective (+3) to most ineffective (−3).

The main researcher was with the participants during the data collection and noted down any comments of the participants during the Q sort process. These comments are particularly useful in Q methodology as they facilitate the interpretation of the factor-analytical findings. After each Q sorting task, participants were also asked to comment on those statements they sorted into the extremes, and to suggest additional statements if they felt any significant content was missing.

Analysis

In Q methodology, data analysis usually involves a factor analysis across participants (rather than a factor analysis across items as in the standard variable-focused approach). Each resulting factor represents a perspective on the phenomenon in question that is shared by the participants whose Q sorts load on the factor. After “flagging” individual Q sorts that define the factors, “idealized” Q sorts are generated by aggregating across the flagged Q sorts, weighted by their loadings, for each factor. The meanings of these aggregated Q sorts are then inspected and interpreted through comparison and contrast, with the additional help of participants’ open comments (Brown, 1993; Watts and Stenner, 2005a).

There is no clear methodological strategy for how to handle data from separate Q sorts. The most common procedure would be to analyze the Q sorts separately, and then examine whether there are interrelations between the solutions for the three domains, either statistically through correlations (e.g. Conover and Feldman, 1984), or by visually inspecting whether the same participants load on the same factors across the separate Q sorts (e.g. Stainton Rogers and Stainton Rogers, 1990). However, both variants represent an analysis at the between-participants level and thus the findings would be influenced by the variation that happens to exist in the sample of participants.

A potentially superior strategy would be to rotate the factors of the separate Q sorts jointly; however, there is no established technical procedure for this. As an approximation, we opted for an integrated analysis that combines the Q-sort data across the three domains into one dataset (see Klaus et al., 2010, for a similar approach). Note that the combined sorting patterns can only be meaningful if there are good reasons to assume that a participant’s “voice” can be heard consistently across the different domains (Stainton Rogers and Stainton Rogers, 1990). We have good reasons for this: theoretically because of the models of conflict analysis that we employed, and empirically because factors derived from separate analyses for the three domains could be rotated to similar solutions (see also Stahel and Cohrs, 2015). Also, note that this approach does not assume that equally ranked statements from different domains are of equal psychological significance, nor that the combined Q sorts are identical to those that would have emerged had all statements been sorted together in one large Q sort.⁴

To prepare the data for an integrated analysis, we combined the three separate Q-sort data into a 117 (statements) × 45 (participants) matrix. To avoid the “relationships, processes, and dynamics” part (for which more response categories were used, ranging from −5 to +5 rather than from −3 to +3) dominating the factor analysis, we transformed the separate Q sorts in such a way that they had the same range (i.e. from 0 to +30).

Factors were then extracted using Principal Component Analysis and Varimax-rotated (Watts and Stenner, 2005a). The Varimax-rotated solution was modified manually to further improve interpretability. The data were analyzed using the free PQMethod 2.32 package (Schmolck and Atkinson, 2012). We explored solutions with different numbers of factors and decided on the final number using a combination of criteria (Webler et al., 2009): simplicity (i.e. all else equal, a solution with fewer factors is better than a solution with more factors), clarity (i.e. the more Q sorts load highly and uniquely on one factor, the better), distinctness (i.e. the lower the correlations between the idealized Q-sorts representing the factors, the better), and stability (i.e. certain subsets of participants load together across different solutions). In addition, each factor should have at least two or even three defining Q sorts to qualify as a socially shared viewpoint (Webler et al., 2009).

Results

A five-factor solution appeared optimal, distinguishing between five shared viewpoints toward the conflict. The five factors explained 53% of the variance. Of the 45 sorts, 36 loaded clearly on only one factor. Most of those who did not load clearly on one factor had multiple loadings, and thus subscribed to a mix of the viewpoints described below. Only two participants did not have any factor loading of at least 0.40, and thus had more idiosyncratic views.

In the following, we examine the meaning of the factors by inspecting the statements that ranked highly positively and highly negatively in the respective idealized Q sorts as well as the statements that distinguished most between the different factors. For presentation purposes, the order of factors was rearranged so that the first factor represented the most “pro-Turkish” and the fifth the most “pro-Kurdish” viewpoint. The correlations between the idealized, prototypical Q sorts for each factor (Table 1) showed that the first three factors were all positively interrelated, as were the last two factors. The third and fourth factors were modestly positively related, and the fifth factor was negatively related to the first two factors.

Table 1.

Correlations between factors (idealized Q sorts)

	F1 Terrorism	F2 Economy	F3 Ambivalent	F4 Democracy	F5 Pro-Kurdish
Factor 1	—	0.61	0.40	0.11	−0.37
Factor 2		—	0.44	0.10	−0.49
Factor 3			—	0.31	−0.15
Factor 4				—	0.39
Factor 5					—

Viewpoint 1: “We are fellows; PKK and separatists are traitors and ruin the harmony!”

Factor 1 had six defining sorts, with factor loadings from 0.47 to 0.73 and a mean inter-sort correlation of 0.41. Five participants were men and one was a woman (see Online Supplementary Table 1). Three were Turkish, two Arab, and one Kurdish. Three participants had graduated from university, two from high school, and one from secondary school. Four participants indicated that they or their families had been affected by the conflict (one of them by a death). Loadings on this factor correlated negatively with speaking Kurdish at home, r = −0.50. Among the Turkish participants, loadings correlated positively with right-wing political view, r = 0.73, and educational level, r = 0.46. At this point, however, we should note that, owing to the small sample size and purposive sampling strategy, the analyses of correlations between the factor loadings and other variables are purely descriptive. The relationships presented here cannot be generalized and can only cautiously be used to contextualize the different viewpoints.

This viewpoint resembles the “denial of the conflict” perspective described by Tan (2008), with additional elements of the “terrorism” perspective. Foreign powers, terrorism, and economic underdevelopment of the region are seen as important causes of the conflict (statement causes and issues 17, 18, and 20; henceforth c17, c18 and c20; see Online Supplementary Table 2), which threaten the fellowship between Turks and Kurds (relationship, processes and dynamics 17, henceforth r17) and harmony among all ethnic groups (r18). Domination and discrimination of the Kurds are rejected as causes (c1 and c8). Consistent with the emphasis on harmony, the development of dialogues and trust are seen as effective solutions (solutions 18 and 20; henceforth s18 and s20). The idea that the conflict is between the Turkish state and Kurds is rejected (r30). The PKK and Kurds are seen as different from each other; the PKK is not seen as a legal voice for Kurds at all (r27, r25, and r34) and should stop its activities that aim to make Kurdish voices heard (r41). Therefore, the PKK giving away their weapons (s28) is seen as a precondition for conflict resolution. With terrorism and the political campaign related to PKK considered important factors, an amnesty for Abdullah Öcalan and his inclusion in the solution process are strongly rejected (s21 and s13). However, reintegration of PKK members into the society is agreed on (s25). As solutions, reducing unemployment (s19) and economic development of the region (s1) are considered promising (see Table 2 for acceptable and unacceptable solution strategies for each of the five viewpoints).

Table 2.

Acceptable and unacceptable solution strategies for each of the five viewpoints

Viewpoints	Acceptable solutions	Unacceptable solutions
1. “We are fellows; PKK and separatists are traitors and ruin the harmony!”	1. PKK giving away their weapons2. Development of dialogues between Kurds and Turks3. Solution of the unemployment problem of the young population	1. Inclusion of Abdullah Öcalan in the solution process2. Amnesty for Abdullah Öcalan3. Implementation of certain constitutional changes in order to allow some form of self-rule in the southeast
2. Turkish unity threatened by foreign powers, PKK, and economic problems	1. PKK giving away their weapons2. Solution of the unemployment problem of the young population3. Designing projects to heal traumas of the society	1. Amnesty for Abdullah Öcalan2. Prevention of uncontrolled population growth of Kurds3. Inclusion of Abdullah Öcalan in the solution process
3. State authoritarianism and ignorance of the problem	1. A new constitution involving civic and democratic laws2. Guarantee of all ethnic diversities by constitution3. Implementation of democratization	1. Amnesty forAbdullah Öcalan2. Inclusion of Abdullah Öcalan in the solution process3. Giving rights to Kurds for a public school system
4. Democratization problem	1. Implementation of democratization2. A new constitution involving civic and democratic laws3. PKK giving away their weapons	1. Amnesty for Abdullah Öcalan2. Implementation of liberal democracy3. Re-integration the PKK members to society
5. Oppression of Kurds (pro-Kurdish perspective)	1. Amnesty for Abdullah Öcalan2. Prevention of uncontrolled population growth of Kurds3. Solution of the unemployment problem of the young population	1. Implementation of land reform2. Implementation of liberal democracy3. Development of trust between groups

Note: These solution strategies are the three highest and three lowest-ranked solution statements in each factor (except for Factor 5 where the solution statements are not sorted very highly).

Viewpoint 2: Turkish unity threatened by foreign powers, PKK, and economic problems

Factor 2 had 12 defining sorts, with factor loadings from 0.48 to 0.72 and a mean inter-sort correlation of 0.41. Three participants were men and nine were women. Six were Turkish and six Arab (of whom three did not speak Arabic at home). Seven participants had graduated from university, four from high school, and one from secondary school. Six participants indicated that they or their families had been affected by the conflict (three of them by deaths). Loadings on this factor correlated negatively with speaking Kurdish at home, r = −0.78. Among the Turkish participants, loadings correlated with gender, r = 0.55 (higher loadings for women than for men).

This viewpoint resembles a combination of the “terrorism” and “economy” perspectives described by Tan (2008). Outside influence (c16 and c17), terrorism (c18 and c3), and economic problems (c20 and c4) are seen as important causes of the conflict, whereas humiliation and lack of political representation of Kurds (c14 and c15) are rejected as causes. In line with the emphasis on economic issues and terrorism, reducing unemployment (s19) and the PKK giving away their weapons (s28) are seen as effective solutions. Owing to the belief that foreign countries aim to divide Turkey (r36 and r32) through the PKK (r21, r19, and r29), there is a strong emphasis on the unity of Turkey in terms of territory and language (r14 and r8) and, accordingly, opposition to the foundation of an autonomous Kurdish region or an independent Kurdish state (r3 and r9). Also, an amnesty for Abdullah Öcalan and his inclusion in the solution process are strongly rejected as possible solutions (s21 and s13), as are granting particular cultural and political rights to Kurds (s27 and s15).

The distinction of this viewpoint from the first one lies mainly in this viewpoint’s emphasis on outside influences that threaten Turkish unity (r21, r32, r36, and c16). This viewpoint also has a negative view on Kurds such as perceiving Kurds as being ignorant and aiming to divide Turkey (c24 and c31). Viewpoint 1, in contrast, emphasizes the fellowship between Turks and Kurds, and sees the PKK as traitors that ruin the harmony among different ethnic groups.

Viewpoint 3: state authoritarianism and ignorance of the problem

Factor 3 had three defining sorts, with factor loadings from 0.55 to 0.65 and a mean inter-sort correlation of 0.34. One participant was a woman and two were men. Two were Arab (who spoke Turkish and Arabic at home); one was Kurdish (who spoke Turkish and Kurdish at home). One participant had graduated from university, one from high school, and one from secondary school. The Kurdish participant indicated that he or his family had been affected by the conflict. Among the Arab participants, loadings on this factor correlated positively with speaking Arabic at home, r = 0.47, and negatively with educational level, r = −0.47. Among the Kurdish participants, loadings correlated with gender, r = −0.74 (higher loadings for men than for women).

Causes of conflict are seen in foreign powers (c17), misbehavior of the gendarmerie in the Kurdish region (c28), the Turkish people’s ignorance of the problem (c30), and terrorism (c3). As solutions, democratization is seen as an effective solution to the conflict (s5), along with a new constitution that grants civic and democratic rights independent of ethnicity and guarantees the right to express one’s ethnic identity (s6, s7, and s11). The implementation of socio-economic policies that create compatible interests for Kurds and Turks (s12) is also seen as an effective solution. Similar to Viewpoints 1 and 2, granting an amnesty to Abdullah Öcalan and his inclusion in the solution process are rejected (s21 and s13). Granting particular political or social rights to Kurds is also rejected (s15 and s29). Additionally, this viewpoint emphasizes the unity of Turkey (r14 and r8) and rejects the foundation of an independent Kurdish state (r9) and adopting multiple official languages (r13). Yet, despite the emphasis on the unity of Turkey, giving Kurdish names to villages, towns and cities is not seen as damaging (r16). There seems to be a particular awareness of recent conflict dynamics, as reflected in references to the Taksim bombing in 2010 (r46), the formation of an autonomous Kurdish region in Northern Iraq (r5), and rising expectations of Kurds in response to the Kurdish initiative (r43 and r38). These dynamics are seen to have changed the existing harmony (r17 and r48) and created tensions between Turks and Kurds (r4). This viewpoint is also different from Viewpoints 1 and 2 in recognizing (rather than denying) some misbehaving of state officials in the region towards its ethnically Kurdish citizens.

Viewpoint 4: democratization problem

Factor 4 had five defining sorts, with factor loadings from 0.52 to 0.71 and a mean inter-sort correlation of 0.35. Four participants were men; one was a woman. Two were Turkish, two Kurdish (both speaking Turkish and Kurdish at home); one was Arab (speaking Turkish and Arabic at home). All five participants had graduated from university. One participant indicated that he or his family had been affected by the conflict. Among the Turkish participants, loadings on this factor correlated positively with age, r = 0.45, and with gender, r = −0.49 (higher loadings for men than for women).

This viewpoint resembles the “identity” perspective described by Tan (2008). Although there are some overlaps with the first three viewpoints in terms of seeing PKK as illegitimate, there are also clear differences. The most important causes of the conflict are seen in the denial of Kurds for many years (c5), suppression of Kurds (c12), and Kurds not having had some cultural rights in the past (c13). In line with this, violations of human rights are also seen as an important cause of the conflict (c26). Accordingly, causes such as Kurds aiming to divide Turkey (c24), Kurds being a separatist and ignorant group (c25 and c31), and Kurds and Turks being historically different from each other (c21) are strongly rejected. Owing to the emphasis on long-term suppression and denial, there is a strong agreement with Kurdish language rights in education (r11). In this viewpoint, PKK is seen as an illegitimate group (r34) that has direct links to the pro-Kurdish party BDP (r24). Consistent with the emphasis on PKK’s illegitimacy, the most effective solution is seen in the PKK giving away their weapons. Also, implementation of democratization (s5), the development of the application of human rights (s16), and a new constitution involving civic and democratic laws that guarantee of all ethnic diversities (s6 and s7) are seen as effective solutions to the conflict. However, the official language of Turkey should not be changed (r14). Amnesty for Abdullah Öcalan is also strongly rejected (s21). Viewpoint 4 is different from Viewpoint 3 in blaming the Turkish state for the conflict and supporting the Kurds being given more rights.

Viewpoint 5: oppression of Kurds (pro-Kurdish perspective)

Factor 5 had 10 defining sorts, with factor loadings from 0.49 to 0.75 and a mean inter-sort correlation of 0.39. Three participants were men and seven women. All were Kurdish (all but one speaking Turkish and Kurdish at home). Three participants had graduated from university, three from high school, two from secondary school, and two from primary school. Eight participants indicated that they or their family had been affected by the conflict (four by deaths). Loadings on this factor correlated positively with speaking Kurdish at home, r = 0.80.

This viewpoint resembles a combination of the “identity” and “nation state” perspectives described by Tan (2008), although the demands are for an autonomous Kurdish region rather than an independent Kurdistan. It is very different from the first three viewpoints, but has some similarities to Viewpoint 4. The most important causes of the conflict are seen in long-term denial, suppression, and humiliation of Kurds (c5, c12, and c14) as reflected in Turkey’s assimilation policy toward Kurds (c7). Turkish people’s ignorance of the problem (c30) is also seen as an important cause. Accordingly, the PKK and Kurds being a separatist group that tries to divide Turkey are rejected as important causes (c18, c24, and c25). Owing to the emphasis on long-term suppression, denial, and assimilation policies, there is a strong agreement with Kurdish language rights in education and speaking in public places (r11 and r1), as Turkey can accommodate multiple official languages (r13). There is also insistence on the right to found an autonomous Kurdish region (r3). The conflict is seen as existing between the Turkish state and ethnic minority groups (r30), rather than between Turks and Kurds. However, it is also emphasized that there are social tensions between Turks and Kurds on the streets (r4). According to this perspective, because all Kurdish people support the PKK (r27), it can continue its activities until Kurds’ voices are heard (r41). The most effective solutions are seen as an amnesty for Abdullah Öcalan (s21) and his inclusion in the solution process (s13). Also, solutions such as the development of the application of human rights and solving the unemployment problem are seen as effective (s16 and s19). However, it should be noted that the preferences for solutions of the conflict in this viewpoint are not very pronounced as compared with the other viewpoints, as the solution statements are not sorted very highly. Viewpoint 5 is different from Viewpoint 4 because the views in Viewpoint 5 are more “radical” (e.g. supporting an amnesty for Abdullah Öcalan and an autonomous Kurdish region).

Discussion

Our analysis highlighted five different viewpoints on the Kurdish conflict among Turkish, Kurdish, and Arab laypeople from a multi-ethnic city in the south of Turkey. Two similar “pro-Turkish” viewpoints emerged. The first of these could be identified as a “right-wing” perspective that emphasizes the political nature of PKK’s campaign as a threat to harmony between Turks and Kurds, which is similar to Tan’s (2008) denial and terrorism perspectives. The second perspective points more to external powers and economic issues as factors in the conflict, and has some overlaps with Tan’s (2008) terrorism and economy perspectives; also some Arabs endorsed it. These two viewpoints might be seen as different varieties of the “Turkish State Discourse” described by Yeğen (2007), or of a Turkish “master narrative” (cf. Hammack, 2006). Somewhat larger differences exist to the third viewpoint, which expresses an ambivalent view of the Turkish state and an awareness of recent changes in conflict dynamics. Although our findings are only suggestive in this regard and, therefore, cannot be generalized, one could still argue that this perspective might be less appealing to Turks and more to Arabs, in particular those who speak Arabic at home. Finally, two viewpoints could be identified as “pro-Kurdish” perspectives. The fourth perspective, which is similar to Tan’s (2008) identity perspective, expresses a democracy orientation that emphasizes the democracy and human rights dimensions of the conflict. It was defined by Turkish, Kurdish, and Arab participants. The last viewpoint, which was defined solely by Kurdish participants, emphasizes long-term oppression of Kurds and asserting Kurdish identity and political rights. This viewpoint has identity and nation-state themes as described by Tan (2008). It can be argued that the five viewpoints are clearly differentiated despite their similarities.

We suggest that, with the exception of the more ambivalent Viewpoint 3, all identified viewpoints are compatible with, and thus may form the basis for, an “ethos of conflict” (Bar-Tal, 1998, 2007). They seem likely to reflect, in somewhat different ways and to different degrees, the elements of the ethos of conflict: believing in the justness of one’s own groups’ goals; concern about one’s own group’s security; a positive image of and emotional attachment to one’s own group; seeing one’s own group as victimized; emphasizing unity in one’s own group; being concerned about peace; and delegitimizing the adversary. For example, on the pro-Turkish side, regarding delegitimization of the adversary, in Viewpoint 1 it is mostly PKK which is seen as an organization that ruins the harmony, whereas in Viewpoint 2 foreign powers are the primary target of blame. Likewise, in terms of victimization, in Viewpoint 1 the harmonious fellowship between Turks and Kurds is threatened, whereas in Viewpoint 2 it is more “the unity of Turkey” which is under attack. Correspondingly, on the pro-Kurdish side, regarding delegitimization of the adversary, in Viewpoint 4 it is the Turkish state that is blamed for the conflict, whereas in Viewpoint 5 it is mainly the Turkish Army (i.e. gendarmerie) that is blamed as a “murderer of Kurds.” In terms of victimization, in Viewpoint 4, all ethnic identities (Turkish identity to a lesser extent) are victims of the conflict because of lack of democracy, whereas in Viewpoint 5 it is mainly the Kurdish identity that is victimized.

Thus, the existence of the ethos of conflict does not imply that there is no variation within a conflict party in how the conflict is represented. Rather, there may be different alternative representations of conflict that can underlie the ethos of conflict, shared within different ideological or opinion-based groups (Bliuc et al., 2007; Cohrs, 2012). This is the case both for Turks and Kurds, where multiple factors and substantial ideological variation emerged. Also, what “one’s own group” means can differ. It seems to be more subjective categorizations and identifications rather than “objective” ethnic group membership that matter. Our reliance on comprehensive models of conflict analysis and on a Q-methodological approach allowed for this greater ability to grasp the multiplicity of representations of conflict (see also Cohrs et al., 2015).

It should also be kept in mind, however, that the intensity of the Kurdish conflict is different from that of the Israel–Palestine conflict, where the concept of the ethos of conflict has been developed. Although there is also evidence for the existence of the ethos of conflict in the Palestinian society (Shaked, 2016), one cannot simply assume that the ethos of conflict can be transferred to other conflict-ridden societies. We do not know whether a similar level of multiplicity would emerge in a more prototypical “intractable” conflict (Bar-Tal, 1998). However, Gayer (2012), in a Q methodological study on national identity constructions among Israeli Jews and Palestinians, also found two Q factors on each side of the conflict.

Our results also speak to the question of whether the perspectives of the adversaries in a protracted intergroup conflict are “mirror images”. This concept has been invoked to account for the apparent similarity of American images of the Soviet Union and Soviet images of the US in the context of the Cold War (Bronfenbrenner, 1961). In Salomon’s (2004: 276) words, “by necessity, the collective narratives of groups in conflict contradict each other and mirror each other, providing interpretations of events that negate those of the other side”. Somewhat contrary to this notion, the “pro-Turkish” factors and the strongly “pro-Kurdish” factor were not very strongly negatively correlated. Thus, different ethnic groups tended to endorse qualitatively different representations of conflict which are not so fundamentally contradictory. Relatedly, we also obtained qualitative evidence that the “pro-Turkish” viewpoints and the moderate “pro-Kurdish” viewpoint were not exclusively (though mostly) endorsed by Turks and Kurds, respectively. The most pro-Turkish viewpoint, Viewpoint 1, also included a Kurdish participant, and Viewpoint 4, reflecting a pro-Kurdish viewpoint, also included two Turkish participants. The Arab participants were also spread out across all viewpoints, with the exception of the strongly pro-Kurdish one. The analysis thus revealed similarities and differences both between and within ethnic groups. This shows that aggregating viewpoints into “Turkish”, “Kurdish”, and “Arab” groups, simply according to objective group membership, would be misguided. Thus, rather than relying on objective group membership, research on conflict should focus on opinion-based or ideological groups in order to understand alternative perspectives. This might allow alternative shared perspectives to be explored and captured even across conflicting parties.

In addition, the loading pattern of the Arab participants’ Q sorts might indicate that Arabs as a minority group in Turkey do not show a high level of solidarity with the Kurds. One can further speculate that this might be due to Arabs having more contacts with Turks as a majority group and less contact with Kurds, and not seeing a lot of commonality with Kurds (cf. Glasford and Calcagno, 2012).

Our methodological approach proved useful to uncover socially shared viewpoints toward conflict among laypeople. Q methodology may thus represent a promising alternative to latent class analysis, which has been used to explore mental representations of the Israel–Palestine conflict (Kempf, 2010). The “standard” scale-construction approach in social psychology would have been another alternative. However, scale construction assumes that participants have the same understanding of each item and that each item is a fixed indicator of an underlying factor. Both assumptions are unlikely to be met given the systematic similarities and differences across viewpoints observed in our study (see also Brown and Ungs, 1970).

Implications for conflict resolution

Our findings can usefully contribute to conflict resolution in various ways. In general, they contribute to a comprehensive conflict assessment, which aims to elicit multiple, alternative perspectives on a conflict (Shmueli, 2003). A comprehensive conflict assessment, in turn, facilitates long-term learning and understanding of important concepts for the groups involved in a conflict. It also makes it more likely to grasp the “whole issue” in conflicts (Kriesberg, 1998; Sandole, 2003). Most approaches to conflict intervention involve third parties (Cheldelin et al., 2003). For example, third parties can help conflicting parties with unofficial meetings in which the roots of the conflict, different viewpoints, and possible solutions are explored with the conflicting parties (e.g. Kelman, 2008; Miall, 2004). Our study might be seen as playing a similar role (although more limited in scope). It can inform conflict resolution attempts such as negotiation between the conflicting parties without third parties (Cheldelin et al., 2003) by feeding in insights on laypeople’s understandings of the conflict.

More specifically, our findings may help to facilitate integrative negotiations by highlighting the beliefs that are either common to the different viewpoints or are of high importance for some viewpoints, but irrelevant for others (cf. Pruitt, 1986). This includes support for the PKK laying down arms, implementing democratization and human rights, resolving the unemployment problem in the region, and rejecting certain constitutional changes to allow some form of self-rule of Kurds in the southeast of Turkey.

Our results also indicate that almost all viewpoints (Viewpoint 1 to a lesser extent though), despite differences in how the problem is defined and what solutions are preferred, agree that there are social tensions between Turks and Kurds throughout Turkey, especially in bigger cities in western Turkey. This acceptance of the existence of social tensions across different groups is like a mirror from society suggesting to policy-makers and politicians that they should focus on this immediate problem and that people are unlikely to object to serious attempts at addressing this problem. The problem definition that there are social tensions has emerged bottom-up (from laypeople), and bottom-up processes are important in conflict resolution and peace building (Lederach, 1999). So far, laypeople’s perspectives have perhaps not been considered enough in the conflict resolution process. For example, approximately 140 workshops have been conducted on the Kurdish conflict, but the participants were academics rather than lay representatives of society (see Başaran, 2010).

Dugan (1996) argues that in some conflict cases not only should the issues be addressed, but also the relationship between groups should be restored (see also Miall, 2004). The results showing agreement that there are social tensions also indicate that the relationships between Kurds and non-Kurds are tense and should be addressed by policy-makers and politicians. In the conflict resolution and peace process, in addition to focusing on long-term goals such as legal or constitutional changes, this issue should be a priority because in conflict-ridden societies social tensions between groups can easily escalate.

Finally, our findings give some insights into which approaches to conflict resolution are seen as ineffective by the different groups, and thus may be less worth focusing on by policy-makers. The study also made explicit the existing diversities among the perspectives on the conflict. For example, whether Abdullah Öcalan should be included in the solution process and what rights can be granted to the Kurds are highly controversial issues. Policy-makers need to take into account these diversities as much as the similarities (see Table 2 for a summary of acceptable and unacceptable solutions for each viewpoint).

Limitations and future directions

Our study is limited in several ways. The participants noted a few problems with some statements. For example, some statements were not entirely clear for some participants owing to jargon or technical language. Although this was not very problematic as these cards were usually sorted in the middle of the Q-sort distribution, ambiguities and/or misunderstandings may have affected the reliability of the resulting Q factors.

A strength of our study is that it represents voices from three different groups who live together in a multi-ethnic region in Turkey, including those of another ethnic group living in the same society, which have too often been ignored in social psychological research on intergroup conflict. However, further research could also aim to recruit members of other ethnic minority groups and people from different regions in Turkey, since the political situation of Mersin might have affected the viewpoints of the participants. Mersin is a highly polarized city with frequent political protests by both Kurds and Turks. Therefore, participants who indicated they had not been affected by the conflict may still be very aware of severe conflict outcomes. Such broader sampling would not affect our conclusion that the ethos of conflict may be based on different alternative representations of conflict. However, it might lead to the identification of additional viewpoints, perhaps also additional forms of “pro-Kurdish” viewpoints (e.g. held by subgroups within Kurds such as Zazas; Özdağ et al., 2011). Relatedly, another limitation of the study is that no material in Kurdish was used. Even though the majority of the Kurdish population in Turkey speaks Turkish, publications in Kurdish might represent different (and possibly more “radical”) ideologies and perspectives of the Kurdish population.

Further research could also use our results and develop quantitative measures from the Q factors for use in large-scale surveys to investigate their distribution and correlates (e.g. Baker et al., 2010; Danielson, 2009; for an example in relation to the Kurdish conflict context see Uluğ and Cohrs, 2016b). As a small-sample approach, Q methodology is not well suited to examining why different people subscribe to different viewpoints that exist in a population. Therefore, we cautioned readers above that the correlations between factor loadings and socio-demographic variables cannot be generalized. Large-scale survey studies could also investigate the influence of additional socio-demographic factors (e.g. whether participants had experienced internal displacement or not) as well as psychological constructs (e.g. ingroup attachment and ingroup glorification as well as intergroup contact as predictors of viewpoints; Glasford and Calcagno, 2012; Roccas et al., 2006).

In addition to large-scale surveys, analyses of historical and political narratives or media materials might be useful to explore conflict viewpoints; however, through the Q methodological approach, we were able to identify intra-group differences in greater detail (e.g. differences within Kurds or Turks) and to explore the Kurdish conflict viewpoints holistically and systematically (see also Uluğ and Cohrs, 2016a, c).

Conclusion

Laypeople’s understandings can have an impact on the course of a conflict. Yet there are only some initial attempts to investigate these representations systematically (e.g. Bar-Tal, 1998, 2007; Bar-Tal et al., 2012; Kempf, 2010). Our study, exploring representations of the Kurdish conflict among laypeople in Turkey, contributes to this emerging literature in various ways. It shows that there is less unanimity in the perspectives on the conflict among groups members than suggested by the concept of “ethos of conflict” (Bar-Tal et al., 2012); that there is less diametrical opposition between adversary groups than is suggested by the concept of “mirror image” (Bronfenbrenner, 1961); and that Q-methodological approaches are promising to investigate socially shared representations of conflict and to contribute to conflict resolution.

Footnotes

Acknowledgements

We would like to thank Ahmet Çoymak, Rezarta Bilali, Reşit Kışlıoğlu, Önder Bakırcıoğlu, Emma O’Dwyer, Lea Stahel, and Hüseyin Çakal for their help in developing the statements, the reviewers for their constructive comments on an earlier version of this manuscript, and Peter Schmolck for his kind methodological advice. We would also like to thank our participants who were willing to participate in our study.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

References

Baker

van Exel

Mason

Stricklin

(2010) Connecting Q and surveys: Three methods to explore factor membership in large samples. Operant Subjectivity 34: 38–58.

Bar-Tal

(1990) Israel–Palestinian conflict: A cognitive analysis. International Journal of Intercultural Relations 14: 7–29.

Bar-Tal

(1998) Societal beliefs in times of intractable conflict: The Israeli case. International Journal of Conflict Management 9: 22–50.

Bar-Tal

(2000) From intractable conflict through conflict resolution to reconciliation: Psychological analysis. Political Psychology 21: 351–365.

Bar-Tal

(2007) Sociopsychological foundations of intractable conflicts. American Behavioral Scientist 50: 1430–1453.

Bar-Tal

Raviv

Dgani-Hirsh

(2009) The influence of the ethos of conflict on Israeli Jews’ interpretation of Jewish–Palestinian encounters. Journal of Conflict Resolution 53: 94–118.

Bar-Tal

Sharvit

Halperin

Zafran

(2012) Ethos of conflict: The concept and its measurement. Peace and Conflict: Journal of Peace Psychology 18: 40–61.

Barrett

(1986) Factor comparison: An examination of three methods. Personality and Individual Differences 7: 327–340.

Başaran

(2010, November 20) Uzlaşı yolunu “Ortak Akıl” açıyor. Radikal. Available from: http://www.radikal.com.tr/politika/uzlasi_yolunu_ortak_akil_aciyor-1029602 (accessed 23 August 2014).

10.

Beriker-Atiyas

(1997) The Kurdish conflict in Turkey: Issues, parties and prospects. Security Dialogue 28: 439–452.

11.

Bliuc

McGarty

Reynolds

Muntele

(2007) Opinion-based group membership as a predictor of commitment to political action. European Journal of Social Psychology 37: 19–32.

12.

Bronfenbrenner

(1961) The mirror image in Soviet–American relations: A social psychologist’s report. Journal of Social Issues 16: 45–56.

13.

Brown

(1980) Political subjectivity: Applications of Q methodology in political science. New Haven, CT: Yale University Press.

14.

Brown

(1993) A primer on Q methodology. Operant Subjectivity 16: 91–138.

15.

Brown

(2002) Structural and functional information. Policy Sciences 35: 285–304.

16.

Brown

Ungs

(1970) Representativeness and the study of political behavior: An application of Q technique to reactions to the Kent State incident. Social Science Quarterly 51: 514–526.

17.

Çandar

(2009) The Kurdish question: The reasons and fortunes of the “Opening”. Insight Turkey 11: 13–19.

18.

Çandar

(2011) Dağdan iniş—PKK Nasıl Silah Bırakır: Kürt sorunu’nun şiddetten arındırılması [Coming Down from the Mountain—How the PKK Can Disarm: the de-violencing of the Kurdish problem]. TESEV Kürt Sorunu Raporu.

19.

Çelik

(2010) Kürt meselesinde aktörler ve süreçler: barışa giden farklı seçenekleri keşfetmek [Actors and processes in the Kurdish question: Exploring different alternatives to peace]. Available from: http://myweb.sabanciuniv.edu/bcelik/su_yayinlar/ (accessed 8 August 2014).

20.

Çelik

Blum

(2007) Track II interventions and the Kurdish question in Turkey: An analysis using a theories of change approach. International Journal of Peace Studies 12: 51–81.

21.

Çelik

Rumelili

(2006) Necessary but not sufficient: The role of the EU in resolving Turkey’s Kurdish question and the Greek–Turkish conflicts. European Foreign Affairs Review 11: 203–222.

22.

Cerrah

(1999) Ethnic Identity versus national identity: An analysis of PKK terror in relation to identity conflict. Turkish Journal of Police Studies 3: 1–12.

23.

Cheldelin

Druckman

Fast

(2003) Conflict: From Analysis to Intervention. London: Bloomsbury.

24.

Cohrs

(2012) Ideological bases of violent conflict. In: Tropp

(ed.), Oxford Handbook of Intergroup Conflict. New York: Oxford University Press, pp. 53–71.

25.

Cohrs

Uluğ

ÖM

Stahel

Kışlıoğlu

(2015) Ethos of conflict and beyond: Differentiating social representations of conflict. In: Halperin

Sharvit

(eds) The Social Psychology of Intractable Conflicts: Celebrating the Legacy of Daniel Bar-Tal, Vol. 1. New York: Springer, pp. 33–45.

26.

Coleman

(2003) Characteristics of protracted, intractable conflict: Towards the development of a framework—I. Journal of Peace Psychology 9: 1–37.

27.

Conover

Feldman

(1984) How people organize the political world: A schematic model. American Journal of Political Science 28: 95–126.

28.

Danielson

(2009) Q method and surveys: Three ways to combine Q and R. Field Methods 21: 219–237.

29.

Dugan

(1996) A nested theory of conflict. Women in Leadership 1: 9–20.

30.

Elcheroth

Spini

(2012) Political violence, intergroup conflict, and ethnic categories. In: Bar-Tal

(ed.), Intergroup Conflicts and their Resolution: A Social Psychological Perspective. New York: Psychology Press, pp. 175–194.

31.

Fisher

(2006) Intergroup conflict. In: Deutsch

Coleman

Marcus

(eds), The Handbook of Conflict Resolution: Theory and Practice. San Francisco, CA: Jossey-Bass, pp. 176–196.

32.

Gayer

(2012) Gendered Intractability: National Identity Constructions and Gender in the Israeli–Palestinian Conflict. Baden-Baden: Nomos.

33.

Glasford

Calcagno

(2012) The conflict of harmony: Intergroup contact, commonality and political solidarity between minority groups. Journal of Experimental Social Psychology 48: 323–328.

34.

Gunter

(2000) The continuing Kurdish problem in Turkey after Öcalan’s capture. Third World Quarterly 21: 849–869.

35.

Hammack

(2006) Identity, conflict, and coexistence: Life stories of Israeli and Palestinian adolescents. Journal of Adolescent Research 21: 323–369.

36.

Karakaya-Polat

(2008) The AKP and the Kurdish Issue: What went wrong? SETA Policy Brief 14: 1–6.

37.

Kelman

(2008) Evaluating the contributions of interactive problem solving to the resolution of ethnonational conflicts. Peace and Conflict: Journal of Peace Psychology 14: 29–60.

38.

Kelman

Fisher

(2003) Conflict analysis and resolution. In: Sears

Huddy

Jervis

(eds), Oxford Handbook of Political Psychology. Oxford: Oxford University Press, pp. 315–353.

39.

Kempf

(2010) Mental Models of the Israeli–Palestinian Conflict (Diskussionsbeiträge der Projektgruppe Friedensforschung, No. 64). Konstanz, Germany: University of Konstanz.

40.

Klaus

Wingreen

Blanton

(2010) Resistant groups in enterprise system implementations: A Q-methodology examination. Journal of Information Technology 25: 91–106.

41.

Koç

Hancıoğlu

Çavlin

(2008) Demographic differentials and demographic integration of Turkish and Kurdish populations in Turkey. Population Research and Policy Review 27: 447–457.

42.

KONDA Research Report (2011) Kürt Meselesi’nde Algı ve Beklentiler Araştırması [Research on the Kurdish Issue in terms of perceptions and expectations]. Available from: http://www.konda.com.tr/tr/raporlar/2011_06_KONDA_Kitap_Basin_Bulteni.pdf (accessed 17 March 2013).

43.

Kriesberg

(1998) Constructive Conflicts: From Escalation to Resolution. New York: Rowman and Littlefield.

44.

Lederach

(1999) Building Peace: Sustainable Reconciliation in Divided Societies (3rd edn). Washington, DC: United States Institute of Peace Press.

45.

McCrae

Zonderman

Costa

Bond

Paunonen

(1996) Evaluating replicability of factors in the Revised NEO Personality Inventory: Confirmatory factor analysis versus Procrustes rotation. Journal of Personality and Social Psychology 70: 552–566.

46.

Miall

(2004) Conflict transformation: A multi-dimensional task. Berghof Handbook for Conflict Transformation. Available from: http://www.berghof-handbook.net/ (accessed 8 August 2014).

47.

Özdağ

Vurucu

Akbaş

(2011) Doğu raporu: Bölgede Türk kimliği ve Türklük algısı [East Report: Perception of Turkish identity and Turkishness in the Region]. Ankara: Kripto Kitaplar.

48.

Pruitt

(1986) Achieving integrative agreements in negotiation. In: White

(ed.), Psychology and the Prevention of Nuclear War. New York: New York University Press, pp. 463–478.

49.

Reicher

(2004) The context of social identity: Domination, resistance, and change. Political Psychology 25: 921–945.

50.

Roccas

Klar

Liviatan

(2006) The paradox of group-based guilt: Modes of conflict identification, conflict vehemence, and reactions to the in-group’s moral violations. Journal of Personality and Social Psychology 91: 698–711.

51.

Salomon

(2004) A narrative-based view of coexistence education. Journal of Social Issues 60: 273–287.

52.

Sandole

DJD

(2003) A Typology. In: Cheldelin

Druckman

Fast

(eds), Conflict: From Analysis to Intervention. London and New York: Continuum International, pp. 39–55.

53.

Schmolck

Atkinson

(2002) PQMethod 2.32 [computer software]. Available from: http://schmolck.userweb.mwn.de/qmethod/ (accessed 18 October 2012).

54.

Shaked

(2016) Ethos of conflict of the Palestinian society. In: Sharvit

Halperin

(eds), A Social Psychology Perspective on the Israeli–Palestinian Conflict: Celebrating the Legacy of Daniel Bar-Tal, Vol. 2. New York: Springer, pp. 133–148.

55.

Shmueli

(2003) Conflict assessment. Beyond Intractability. Available from: http://www.beyondintractability.org/essay/conflict-assessment (accessed 26 June 2016).

56.

Stahel

Cohrs

(2015) Socially shared representations of the Israel–Palestine conflict: An exploration among third parties. Conflict and Communication Online 14(1): 1–19.

57.

Stainton Rogers

(1991) Explaining Health and Illness: An Exploration of Diversity. London: Harvester Wheatsheaf.

58.

Stainton Rogers

(1990) What the Brits got out of the Q: And why their work may not line up with the American way of getting into it! Electronic Journal of Communication 1(1). Available from: http://www.cios.org/EJCPUBLIC/001/1/00113.html

59.

Tajfel

(1974) Social identity and inter-group behavior. Social Science Information 13: 65–93.

60.

Tajfel

Turner

(1986) The social identity theory of inter-group behavior. In: Worchel

Austin

(eds), Psychology of Inter-group Relations. Chicago, IL: Nelson Hall, pp. 7–24.

61.

Tan

(2008) Kürt sorunu [Kurdish question]. Istanbul: Timaş Yayınları.

62.

Tocci

(2005) Conflict resolution in the neighborhood: Comparing the role of the EU in the Turkish–Kurdish and Israeli–Palestinian conflicts. Centre for European Policy Studies Working Documents 221: 1–32.

63.

Uluğ

ÖM

Cohrs

(2016a) An exploration of lay people’s Kurdish conflict frames in Turkey. Peace & Conflict: Journal of Peace Psychology 22(2): 109–119; https://dx-doi-org.web.bisu.edu.cn/10.1037/pac0000165

64.

Uluğ

ÖM

Cohrs

(2016b) “If we become friends, maybe I can change my perspective”: Intergroup contact, endorsement of conflict narratives, and peace-related attitudes in Turkey. Peace & Conflict: Journal of Peace Psychology, in press.

65.

Uluğ

ÖM

Cohrs

(2016c) “Who will resolve this conflict if the politicians don’t?”: Members of Parliament’s understanding of the Kurdish conflict in Turkey. International Journal of Conflict Management, in press.

66.

Watts

Stenner

(2005a) Doing Q methodology: Theory, method and interpretation. Qualitative Research in Psychology 2: 67–91.

67.

Watts

Stenner

(2005b) The subjective experience of partnership love: A Q-methodology study. British Journal of Social Psychology 44: 85–107.

68.

Watts

Stenner

(2012) Doing Q Methodological Research: Theory, Method and Interpretation. London: Sage.

69.

Webler

Danielson

Tuler

(2009) Using Q Method to Reveal Social Perspectives in Environmental Research. Greenfield MA: Social and Environmental Research Institute.

70.

Wehr

(1979) Conflict Regulation. Boulder, CO: Westview Press.

71.

Wehr

(2006) Conflict mapping. In Beyond Intractability. Available from: http://www.beyondintractability.org/essay/conflict-mapping (accessed 26 June 2016).

72.

Wong

Sun

(1998) Dissolution and reconstruction of national identity: The experience of subjectivity in Taiwan. Nations and Nationalism 4: 247–272.

73.

Yavuz

Özcan

(2006) The Kurdish question and Turkey’s justice and development party. Middle East Policy 8: 1–7.

74.

Yeğen

(2007) Turkish nationalism and the Kurdish question. Ethnic and Racial Studies 30: 119–151.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.06 MB

0.00 MB