Village guards as “in between” in the Turkish-Kurdish conflict: Re-examining identity and position in intergroup conflict

The current paper utilizes new approaches in intergroup conflict studies to examine the village guard system and its role in the Turkish-Kurdish conflict in Turkey. Recent work suggests that a two-group paradigm in researching intergroup conflict leaves out important contextual factors that influence trajectories and outcomes of conflict. The current paper is based on a project examining the views of 63 active and retired village guards in five provinces in eastern Turkey. Participants were asked how they became village guards, their experiences while holding the position and after they’ve retired, their relationship with neighbors and neighboring villages, as well as their views on the peace process and whether they believe a lasting peace is possible. Results indicate that village guards became guards either because their tribe took arms as a whole, they felt economically there were few other options, or were pressured by the state. Participants also reported feeling otherized by both non-village guard neighbors as well as state actors, and were generally positive about a peaceful outcome to the conflict but were concerned about the sincerity of the government.

and even in cross-border warfare into Northern Iraq -expanding their duties well beyond those stipulated by law (Balta & Akça, 2013).
Of particular note is that duties of the village guards have changed substantially over time. Until the 1990's, village guards primarily dealt with providıng local security, but the perception of the PKK threat began to change, so that village guards were now charged with determining who was or was not loyal to the State (Balta, Yüksel, & Acar, 2015). For local authorities, rejecting the system (i.e., rejecting the government, police, military) meant supporting the guerrillas, either actively or passively, which often resulted in oppression and evacuating and burning villages.
Thousands of people were displaced in the 1990's due to the conflict, fear of violence, and economic activity becoming unsustainable. A number of studies have revealed that many victims were forced to evacuate their homes because they refused to become village guards (e.g., Balta, Yüksel, & Acar, 2015;Kurban, Yükseker, Çelik, Ünalan, & Aker, 2006). Both village guards and victims of village guards have similar stories regarding the pressure to become village guards and the evacuation of the villages: security forces enter the village and tell the villagers that they have to become village guards or leave their village. They're given some time to make their decision. If villagers are unable to choose between rejecting or accepting becoming village guards, the soldiers evacuate their village.
Despite the intensity of fighting decreasing after the 1999 capture of PKK leader Abdullah Öcalan and the Council of Ministers' statement in 2000 that the recruitment of village guards would end, village guards continued to be hired in the 2000's and 2010's. National and international criticism of the village guard system continues, but claims that the village guards' lives would be in danger if they lay down arms before the PKK were put forth.
Village guards continued to be hired even during the relative calm of the peace process in 2013-2015, and national and international criticism of the village guard system continued. In July of 2015, though, the peace process effectively came to an end as the Turkish military resumed large-scale attacks on the PKK. In September 2016, under the current Justice and Development Party's government (Adalet ve Kalkınma Partisi, AKP), the Ministry of the Interior began talks to create a department for village guard affairs, hire thousands of new guards, and train guards in the use of heavy artillery. In November 2016, the law was implemented, making the guard system permanent and changing their title from "village guards" (köy korucusu) to "security guards" (güvenlik korucusu), effectively ending any discussion of abolishing the village guard system. i

Re-Examining Two-Group Approaches to Intergroup Conflict
The village guard system is an essential component of the ongoing Turkish-Kurdish conflict; under-researched and under-discussed, it is nevertheless essential in better understanding the way conflict occurs and how it is continued. Like many other conflicts, this one is by no means dichotomous. Other ethnic and religious minorities within the country, special interest groups globally, and other factors are all in play in such large-scale and longlasting conflicts. Relying on social identity and self-categorization theory (e.g., Tajfel & Turner, 1979;Turner, 1985Turner, , 1991Turner, Hogg, Oakes, Reicher, & Wetherell, 1987; also see Hogg & Abrams, 1988;Abrams & Hogg, 2010;Hogg, 2006), we know that the self is hierarchically organized, context-specific, and variable (Subašić et al., 2008). People categorize themselves in different social situations based on representations of the self as personal or social, but also at a certain level of social -whether that be as part of a superordinate or subcategory of a superordinate identity. Therefore, in different situations, a person can be seen as ingroup or outgroup based on the salience of a particular identity.
Recognition of group identity is another central factor in terms of who is considered ingroup or outgroup. Recognition, broadly understood as people feeling that their own sense of identity is affirmed (Hopkins & Blackwood, 2011), and non-recognition have an important influence on intra and intergroup relations. Non-recognition of identity can influence how people feel and behave in their everyday interactions in the public sphere -when people feel their identity is not affirmed, they may avoid or feel constrained, threatened or harmed by the interaction. (Hogg, Fielding, & Darley, 2005;Hopkins & Blackwood, 2011).
The ability to categorize and recategorize oneself and others is essential to understanding intergroup relations.
Especially in contexts of conflict, it is useful to know how and when people see themselves in contrast to others, rather than as part of the same group. While the importance of categorization in this way has been extensively studied in research on prejudice reduction and conflict resolution (e.g., Brewer, 2000;Brown & Hewstone, 2005;Gaertner et al., 2000;Wenzel, Mummendey, & Waldzus, 2008), comparably little attention has been paid to how recategorization complicates intergroup conflict. That is, as intergroup conflict is oftentimes perceived as existing between two immutable and mutually exclusive groups, it does not often recognize the fluid nature of categorization. It can, for example, often be the case that a person is both a victim and a perpetrator in the same conflict. As noted by Vollhardt and Bilewicz (2013), what is often referred to as the perpetrator-victim-bystander triangle in research on genocide is often much more complex than the literature suggests. As they note, perpetrators may perceive themselves as victims, victims may violently retaliate against perpetrators, and in some cases, individuals can variably occupy all three roles. Seeing roles as static and intergroup relations as solely bipolar prevents us from having a full understanding of the conflict, and removes the conflict from the social and historical context within which it takes place (Eggins, Haslam, & Reynolds, 2002). Recognizing the importance of context-dependent variability in identity gives us a richer understanding of the conflict.
Recent perspectives in social psychological work on group dynamics (e.g., Kerr, Durrheim, & Dixon, 2017) have played an important role in questioning the relevance of a dichotomous approach, and have eluded to the fact that intergroup relations and intergroup conflicts in particular don't happen between two opposing groups. Especially in contexts where parties have been colonized, a dichotomous approach overlooks what happens when groups are faced with social relations on multiple fronts, the dynamics of which "entail expressions of loyalty, solidarity, favoritism, collusion, and betrayal that are ultimately irreducible to a binary logic" (Kerr et al., 2017, p. 47). While there may be advantages to producing a more streamlined model, in overlooking these complicated factors, the two-group paradigm may be over-simplifying a multi-faceted situation.

The Current Study
Keeping in mind the recognition (or lack thereof) of social identity and the need to allow for nuance in traditionally two-paradigm perspectives on intergroup conflict, the current paper attempts to shed light on an often over-simplified conflict by examining perspectives on the origins of village guards' active participation in the conflict, how this participation has influenced their engagement in the public sphere, and their particular perspectives on the peace process and reconciliation.

Method
Participants were 63 village guards in Siirt (1), Şırnak (12), Hakkari (21), Mardin (15), and Van (14), as part of a larger project on the role of village guards in the peace process conducted with the research group Süreç Analiz, an NGO that conducted research and disseminated information on domestic and international conflict. Participants were all male; ages ranged from 34 to 68 (M = 46.54). Provinces were chosen based on their location at the border, the high amount of conflict seen in those areas, and research experience and connections in the area. Three researchers coordinated data collection teams in each of the provinces, with one researcher and their team in one province at a time. Data collection was conducted sequentially in all five provinces, and took place in November and December 2014. In each province, village guard associations iii were first approached and informed about the research. A number of village guards were reached with their cooperation; a snowball method was utilized to reach remaining participants. All participants were interviewed face-to-face in a place of their choosing; though interviews alone were preferred, due to the sensitivity of the subject and the situation, if participants preferred to be interviewed in a more public place or with family or friends present, the researcher allowed it.

Procedure
Before starting the interviews, all participants were informed about the purpose of the study, which was to gain the perspective of village guards on what was at the time the ongoing peace process. Participants received a consent form, which was discussed in full by the researcher. Participants were told that their responses could be used in subsequent reports or articles, but that their personal details would be kept confidential. They were informed that they could stop the interview at any point, and could refuse to answer any question they felt uncomfortable responding to. They were not asked to sign or return the consent form; rather, they received the form and kept it as informative of the project as well as the contact details of members of the research group in case they had further questions or later changed their minds about participation.
Interviews typically began with introductions on the part of the researcher and assistants, followed by opening questions about the participant, their families, and where they live. Initial interview questions were read verbatim to participants; as the interview continued, the questions were used to verify that all topics were covered through the flow of conversation. Participants were then asked to reflect on their self-identification, how they became a village guard, their subsequent relationships with military and other state representatives, their neighbors and other villagers, and also their views on the peace process.
Interviews were generally conducted in Turkish, though if necessary, Kurdish-speaking assistants would confirm meaning of questions for participants. Participants were asked if they would allow audio recordings of their interviews; due to security concerns, participants requested not to be recorded. The researcher and assistant therefore took simultaneous notes throughout the interviews, and would confirm immediately after the interview that notes were accurate. While every effort was made for notes to be as verbatim as possible, some words and non-verbal cues were lost in the process. Notes were translated by the bilingual researcher; a native English speaker was consulted on the fluency of translation. Length of interviews ranged from 45 to 120 minutes, with an average of approximately 70 minutes.

Analysis
All materials were analyzed based on Qualitative Content Analysis (QCA), a method used to systematically describe the meaning of qualitative material (Schreier, 2012). All interviews were noted by the researcher and research assistants, with the coding frame and analysis conducted by the researcher. The entire interview, rather than answers to particular questions, was analyzed based on the research questions, with categories being refined as coding went on.
With QCA, the research questions guide analysis, and allows the researcher to systematically focus on selected aspects of the material. For the current research, the results related to three areas of focus were extracted from the overall data and are categorized here as 1) accounts of becoming a village guard, 2) group-based relationships with non-village guards and the government (including representatives of the government, such as police and military), and 3) perspectives on the peace process and postconflict relationships. Interview questions related to these three categories can be found in the Appendix. After specifying the main categories, subcategories were identified and stated in the material in relation to the main categories. All subcategories were developed inductively.
During this process, data were labeled and defined, and are here illustrated by means of examples from the raw data (Schreier, 2012; see also Uluğ et al., 2017). Key illustrative quotes from each of the main categories and subcategories are presented below. As interviews were not recorded, the provided quotes are taken directly from the notes, and while they may be close to verbatim, they should be taken as approximate reconstructions made by the research team.

Becoming a Village Guard
When asked about how they became village guards, accounts fell mostly into one of three categories. Participants either stated they became village guard because their tribe joined as a group, because of economic hardships making becoming a village guard the only option, or because they were forced by the Turkish state to do so. Examples and illustrative quotes of each are provided below.

Joining Based on Group Affiliation
Especially in the first years of the village guard system, state representatives, rather than approaching individuals, would approach tribal leaders to decide who would become village guards. In doing so, they used the existing social structures in place to create the guard system, meaning that rather than hiring individuals, state representatives would negotiate with tribal leaders, who would then make all eligible individuals in the tribe guards. Inadvertently or not, this act legitimized the power and authority of the tribes in the region, solidifying existing hierarchies.
Participant 62 describes his experience as follows: Village Guards in the Turkish-Kurdish Conflict 930

Extract 1
The state called together the heads of families. We all met, as the elders of the tribe…At the meeting, [the governor] offered us the position of village guards. We…accepted the armaments. At first 31 of us became guards. We didn't even know if there was a salary. We were volunteers. Of course later we shaped the village guard system. We created and fitted the village guard system in [town]. We decided who would be guards. I said "come, be guards." Then I told others. We offered. That's how we developed the system here. Some individuals within the tribe refused to become village guards. We convinced them to become guards.
At the time, all issues about what the village guard system as well as the duties therein were learned from the tribal leaders. Through the village guard system, tribal leaders were able to maintain their economic, political, and social control in the region. Salaries were given to tribal leaders to be passed out to guards.

Joining Due to Economic Hardship
Especially with very little economic opportunity, having the guard system in place became the source of income for many. Participant 58:

Extract 2
In our village the tribal system is important; if the ağa iv becomes a village guard, most people will become a village guard. Our village was on the border, it had a strategic location in the mountains. The State wanted our village to become guards, so we became guards. There's no livelihood, no trade. There was no better alternative, anyway.
In both cases, there is a sense of necessity to become a guard (either from the tribe or economic concerns), but no direct pressure from the state or its representatives. In many cases though, guards reported being forced or pressured into becoming a village guard.

Joining Due to Pressure
Like many stories coming from the region at the time, village guards were presented with two options by the military: evacuate the village completely, or stay and become village guards. Participant 7:

Extract 3
In 1993 there was an operation and raid and they took the whole family. Even the children. We remained in custody for 8 days.

Relationships With Others
Participants were asked about their relationships with non-village guards and the Turkish State, both tangibly (through its representatives, such as the police and military), and intangibly (the Turkish State as a construct).

Relationships With Non Village Guards
When village guards were asked about their relationships with non village guards, they generally stated that relationships were quite poor. On a basic level, they stated that the public did not care for them, and saw them quite negatively. Participant 56:

Extract 4
Everyone in our village became guards, but our relationship with the people was affected. When they see us it's like they've seen the enemy. We can't walk around freely.
This sentiment was shared by most village guards, who also stated that it is their identity that places them squarely in the middle between the PKK and the Turkish State. Participant 18:

Extract 5
I'm a Kurd, the PKK doesn't like me, the State sees me as a terrorist. We are stuck in the middle.
Some stated that the relationship wasn't just bad, but dangerous, and that the problems of the past continued to follow them. Guards mentioned that events that occurred to previous generations affect their relationship with non village guards who see them as the cause of violence in the past. Participant 40:

Extract 6
We haven't been able to leave home for six years; a lot of village guards experience this and say this…we are the targets of the children of the unsolved murders, v because when those murders were committed they were two years old, now they're 20-27 years old.
The problems also took on tribal dimensions, as in the 1980's and 1990's it was primarily tribal leaders who made the decision about becoming village guards. Participant 34: Extract 7 [Our two] tribes were like two divided countries. Before 1992, we used to intermarry. But after [1992] there was a border between us. They wouldn't give us supplies when we came to the city center. Our relatives wouldn't even greet us.

Relationships With the State
Many guards specifically feel caught in the middle in the conflict between the State and the PKK. While some express openly that they stand by the Turkish State, they still feel that they are not taken care of by the state or the system. For others, the system itself is the problem, rather than the guards. Participant 42:

Extract 8
It's an empty system. We are forced to be caught in the middle. It shouldn't be our job to ensure safety.
We're stuck between the public and the state. If only this system didn't exist. I never defended this system. Some families got stuck between the state and the PKK and they chose the village guard system. Some of them needed it, some of them were forced. Then problems arose between those who chose the village guard system and those who didn't.
In some cases, being a village guard provided "special privileges" on the part of the military or local authorities.

Extract 9
Smuggling became like a gang activity. And soldiers would turn a blind eye to village guards. That's how smuggling was done.
In this case, then, becoming a village guard means that one can provide for one's family when they would otherwise have no other option.
Guards feel that they have been forced between two competing political forces, and that neither really supports them or sees them as part of their "whole." Participant 57:

Extract 10
We're again stuck between the state and Kurdish politics. They meet. One day they'll come to terms. But neither side trusts us. For Kurds, we are enemies who have betrayed them. For the state, at the end of the day, we are still Kurds.

Village Guards, the Peace Process, and the Future
Of the village guards interviewed, almost all saw the peace process as beneficial not just to Kurds but to the whole of Turkey. Participant 13:

Extract 11
People don't care about money, property…people just want peace and to live freely. Sometimes they say the process has been ruined and we lose our morale. This public has lived through this war for forty years.
Those coming from the outside haven't experienced it. They come for a short time and then they leave.
The public have experienced this cruelty. That's why this public wants peace, they don't want blood to flow.
Others described the newfound peace by comparing their current situation to what they'd experienced in the past. For some, the peace process is something they view positively, but they remain concerned about the way that the AKP government approaches the peace processes, and even Kurds in general. Participant 11:

Extract 15
The process is going well, but we don't have any expectations from the AKP. How many years has it been since a ceasefire? They still haven't taken any steps, there are still sick prisoners, the KCK ix prisoners haven't been let go. My thought is that AKP isn't taking any steps [forward]. It's like the resolution process is a distraction. I think that if the PKK disarms and comes down from the mountains, the state will do something even worse to the Kurds [than they have done before].
Participant 47 shares a similar concern: Extract 16 We want the resolution process to conclude with peace with all our hearts. The AK Party government has to be sincere…In this resolution process, the state is preparing soldiers for war. Why are they building new police stations [kalekol]? Why are they building these dams? They want us to immigrate. The state is telling the international public that it is making peace with the Kurds, but it is actually preparing for war.
The state has to be careful what it is saying in this process. They have to stay away from harsh language.
The State is all of ours and we have to stay loyal to it. And the State needs to know this, and behave accordingly. It shouldn't be marginalizing us. We village guards are for peace, if they tell us to disarm we will do it immediately. We won't resist. But in general the State looks at the Kurds differently; it has to look at them more warmly. While overall the village guards saw the peace process in a positive light, there was some concern about the sincerity of the government in its approach to the conflict. Participants were positive about their abilities to cope in a postconflict scenario, if they felt they had the support of the government behind them, both economically and in terms of security.

Discussion
Results suggest that accounts of becoming a village guard generally fell into one of three categories: participants either became a village guard because their tribe joined all together, due to economic hardships making becoming a village guard the only viable option, or reported they were forced by the Turkish state to do so. In line with previous research, accounts of how guards joined the conflict indicate that at various times throughout the process, village guards describe themselves in different positions within the conflict, as either perpetrator or as victim (e.g., Vollhardt & Bilewicz, 2013).
It is perhaps relevant that the timing and reason for becoming a village guard may influence the way participants Results also indicate that village guards are particularly aware of the way they are viewed by others. They intermittently refer to other Kurds as well as to the Turkish State as ingroup, while recognizing that one or the other, and sometimes neither, accepts them as such. As important identifications shifted from tribal affiliation to the larger, politicized Kurdish identity, the question often turned to which "side" the village guards are on. In line with previous research on identity non-recognition (e.g., Hopkins & Blackwood, 2011), participants' relationships with non-village guards colored their behaviors and interactions in the public sphere, and indeed felt unsafe in these interactions. Participants also noted the effects their position had on their children. Some sent their children to other parts of the country for school, so they would not be negatively impacted by their status as the children of village guards, while others brought their children into the guard system, passing it along as a family trade of sorts.
It is likely, therefore, that the effects of being a village guard also have profound impacts on their families, including later generations.
As seen with criticisms of the two-paradigm approach to intergroup conflict (e.g., Kerr et al., 2017), participants also reflect on the opposing space they occupy in comparison with the Turkish military and with the PKK, and even their Kurdish non-village guard neighbors. As one participant summarized, they felt that neither side saw the village guards as ingroup or allies, but rather as an "other." Village guards also tended to see themselves collectively as both perpetrators -recognizing their role, or at least the role of "other" village guards, in unsolved murders -as well as victims, both in that they don't receive the support they need from the state, as well as victims of non-village guards who hold them responsible for previous crimes. Importantly, across the board, village guards felt positively about a resolution to the conflict. It's possible that their unique role as both "insider" and "outsider" in the conflict makes them especially willing to engage in peace. While apprehensive about the AKP's sincerity in ending the conflict, they still felt hopeful about the future. Many participants indicated that they would lay down their arms and live peacefully so long as the government were able to provide work for them in the postconflict period.
It is also relevant to note the many ongoing contextual factors that have influenced the data collection. First of all, data was collected at the end of 2014, just after the Kobanê protests had ended. The protests were especially charged, and 43 people were killed in the process. It is possible that these protests were still salient in the minds of participants as they responded to questions about a possible peace. It is also important to note that data collection occurred while the peace process was ongoing. It is unclear to what extent participants' sentiments would be the same now, after the latest peace process has ended and the village guard system has attained a more permanent place in the Turkish government. It is also important to note that the place and manner of data collection is not as systematic as it could have been in a less conflictual situation. Village guard associations were the usual first step in each province in reaching village guards. It is unclear to what extent these associations influenced which guards we spoke to or how freely those guards spoke. However, considering the nature of fieldwork in conflict (see e.g., Moss, Uluğ, & Acar, 2019), it is still notable that guards were reached at all.

Implications and Future Directions
Future research should examine how changes to the village guard system have impacted guards as well as their relationships with others. With the creation of a branch of government that has been specified to look after the needs of the guards, and with the conflict once again active, the way individuals may conceive of their roles as village guards may change, and identification as a village guard may be more salient. It is possible that with the peace process having ended, their views on peace and on postconflict scenarios could also likely be impacted.
This study in particular focuses on paramilitary forces who are recruited and maintained by a branch of the government. In particular, how is the violence perpetrated by village guards perceived, as compared with violence committed by the PKK? Since village guards have the right to bear arms, and also receive immunity, are their actions, no matter the cost, legitimized? Previous research by Vollhardt and Bilewicz (2013) discusses the various roles individuals may play in a conflict; future research could further examine to what extent roles change when there is an institutionalizing force such as the government behind one side.
It is also particularly worthwhile to examine the intergenerational effects of conflict -as previously noted, the families of village guards are also impacted when one member of the family takes up arms. Notably, some participants also had parents who were village guards, or who planned for their children to become village guards, meaning there is likely to be some effect on the way future generations see and understand their role in the conflict.
While there has been previous work on the effects of political socialization around intractable conflict (Nasie, Harrison Diamond, & Bar-Tal, 2016), I do not know of work based on the role that previous generations took in the conflict, and how this influences future generations.

Conclusion
The peace process effectively came to an end in the spring of 2015, when a number of attacks in the east and southeast began, and the Turkish State, ostensibly to secure peace, began a series of curfews and military operations across the region. In just a few months, conflict spread from rural areas into urban ones, something un- precedented in the many years of the conflict. In addition to the thousands of deaths since the conflict began again, the village guard system has undergone immense changes; until now, the question was if and when the village guard system would end. But, in November 2016, the Ministry of the Interior created a department specific to village guards, thousands of new guards have been hired, and preparations are in place for guards to be trained in the use of heavy artillery, effectively ending that speculation.
Over the course of the almost 40-year conflict, there have been a number of ceasefires and peace talks, and all, thus far, have failed. But the end of peace talks in 2015 is by no means an indication that a lasting peace cannot occur. One way to ensure that this and other conflicts truly address the way locals are impacted is to re-examine traditional approaches to intergroup conflict research. Categorization as it relates to conflict and prejudice reduction have been extensively studied, but the fact that conflict-generated categories are neither immutable nor mutually exclusive is often lost in these discussions. As we move forward in developing more nuanced approaches to intergroup conflict, understanding the many roles that individuals within these conflicts take is crucial to understanding possible pathways to peace.

Notes
i) As of May 2018, the total number of village guards peaked at approximately 72,446. http://www.kpsscafe.com.tr/kpss-haberleri/2018-de-guvenlik-korucusu-alimi-yapilacak-iste-sartlar-astsubay-h84554.html ii) Is it important here to mention that not all village guards are Kurdish. Some village guards are Arab, Kyrgyz, Azeri, and other ethnic groups. However, the vast majority of village guards are Kurdish, and as such, for the purpose of this study, I focus on Kurdish village guards.
iii) Local associations created for and by village guards to systematically work with government representatives to address the needs and demands of village guards. Associations were contacted by the researchers through extended contacts. Associations were generally receptive to the research and researchers, and encouraged us to speak to village guards they were associated with, though I do not know to what extent they may have guided us to speak to particular individuals (and not to others).
iv) lord, landowner v) Thousands of Kurdish citizens were murdered in the 1990's by soldiers and civil servants. While some were caught, most weren't and the incidents were played down by the Turkish government and many remain unsolved to this day.
vi) Plateau-like area where animals are taken for their summer feeding grounds.
vii) Large-scale protests by Kurds in Turkey as a reaction to events in the Northern Syrian city of Kobanê which left 43 protesters killed.
viii) Mountains in Northern Iraq notable as the sanctuary and headquarters for the PKK.
ix) Group of Communities in Kurdistan (Koma Civakên Kurdistan); since 2009, thousands of people have been detained and tried for membership in the KCK.

Funding
This research is part of a larger project conducted by Süreç Analiz, which was funded by the Istanbul Policy Center and the Open Society Foundation.