- 1 Rules are generalisable and adaptable mental procedures in relation to other contexts (Giddens 198 (...)
1In the perspective shaped by the introduction of this special issue, we are concerned with the transformation of kinship usages, or, more specifically, clans, in a village in Turkey. We will adopt a dualist viewpoint, inspired by Giddens, and discuss the relationship between rules and resourcesin defining kinship as a sector of social life involving rules1 and their associated practices, notably relations of kin, relations between age class, conceptions of childhood etc. The contexts not being univocal, the actors have to decide, with regard to their system of classification of interactions, if a situation raises kinship issues or not. The use of certain rules of preference to others implies, as a practical consequence, the mobilisation of specific resources or constraints, a point to which we shall return.
- 2 We thank our sources, who gave us access to the village, identified here by another name. Elsewher (...)
2Our data2 originates from a survey carried out in Kirazbahçe, a Kurdish village in the administrative department of Mardin (Turkey). We are concerned first of all with the 1970s and 1980s as it was a period of brutal change in the clan system. The population therefore presents a strong socio-economic coherence, all inhabitants living in the fields and possessing a few animals, with the exception of the mullahs, who cultivate gardens. The absence of commerce or of large property allowed the maintenance of a more or less economic equality. Elsewhere, the village was a relatively independent and coherent social unity, in the sense that most observable social relationships, conflictual, or of cooperation, took place in this space, which did not prejudge the very real effects of the encompassing in the Turkish society. The village was not very open to the market until the 1970s, subsistence farming being dominant and valued. For example, something which was not a local produce like tea was not normally used until the 1970s (in the 1950s in most cities). The physical presence of the state was discrete: a school and a dispensary. The villagers spoke Kurdish and most men, especially those involved with the military service, adopted Turkish notions, while women did not master the national language. The relations of cooperation or confrontation with neighbouring villages were rare.
- 3 A part of the population left in 1990 and 1993 due to the fight between the korucu (militia origin (...)
3Our information essentially comes from interviews carried out since the year 2000, during short stays in the village, and at Diyarbakır. Most of these interviews, often collective, took place in an informal manner, generally with men belonging to clans (men of the Suleymani and their allied clans) to whom my key source is acquainted with, notably with former muhtar who know village life in detail spanning back many decades. The information at our disposal is however fragmentary for several reasons. The village was evacuated by the army in 19933 and only a minority of inhabitants have returned since 2000, giving a particular aspect to the village and limiting the number of interlocutors in situ. Besides, it was not possible to consult the population register of the administration and those of the muhtar has disappeared. Many marriages are not legally registered and a significant proportion of women do not have an identity card. Concerning deaths caused by the confrontation of clans, the information collected during the interviews clans cast doubt on the official statistics, which indicate very low figures for the whole of Turkey, including the Kurdish regions (Tezcan 1981: 151). From an economic point of view, the fields were never registered in the village and we do not have sufficiently reliable quantitative information on the economic situation. The object of this work not being to recount the history of the village, we are primarily concerned with the 1980s, when the clans disappeared as a dominant institution.
4Indeed, the survey demonstrates a rapid change in the place of clans [merev] in the 1980s. The term ‘merev’, which translates as ‘clan’ is employed by our source. I have translated the term aşiret, which for our sources designates the encompassing institution, which disappeared in the 1920s, by tribe (see § 7). The term merev can moreover signify kinship in a broad sense, which indicates that our representatives have regard to the central difference between the aşiret, which supposes a leader or a mechanism of possible mobilisation, and the merev of the village, which precisely offers no model of going beyond the segmentation. Today, if the villagers continue to acknowledge their clan, their behaviour does not depend on their affiliation anymore. Access to economic resources or security is no longer linked to the clan institution. How can we explain how the clans could disappear as an effective framework of mobilization in such a brutal manner? How can we explain that certain people, young people and single women, could rejoin an organisation like the PKK, in rupture with the existing social model? These changes and their rapid rhythm go in any case against the idea of a functional and integrated village universe. The rapid disappearance of clans as a framework on mobilisation in the urban area, a phenomenon which is not limited to this village, demonstrates that the efficiency of an institution as a framework of mobilisation is not fundamentally concerned with the more or less profound interiorisation of an identity.
- 4 ‘En 1963, j’organisais le premier enseignement d’anthropologie économique en France, avec l’idée d (...)
- 5 ‘Pour qu’une activité sociale – et avec elle les idées et institutions qui lui correspondent et l’ (...)
5In order to attempt to explain these changes, we will start with a Marxian stance, today rejected by its author4: ‘A social activity – with its ideas and institutions which correspond to it and organise it – plays a key role in the functioning and the evolution of a society (so in the thought and action of individuals and groups which compose this society), it does not suffice that it assumes several functions, it necessary that it directly assumes, as well as its finality and its explicit functions, the function of production relationship’5. Godelier’s thesis was not verified and this village, after other examples, show that the clan played a dominant role in the life of the village without organising production relationships. On the other hand, one retains the idea that a specific rule organises access to resources, leading us to the duality of the rules and practices seen above. We propose the hypothesis that different rules are in competition within the village and those who allow a more efficient mobilisation of the resources tend to evince others, possibly reduced to the state of survival (Sewell 1992). In the case studied, the clan is the dominant framework for the mobilisation of resources for several decades. In the 1970s and 80s, for reasons that we will expose, other rules, until then available but not as relevant, fulfil the functions until then reserved for clans. The complex relationship between schemes of thought and the mobilisation of resources introduces a distance in relation to the rules and provides for a split with a functionalist approach to socialisation. The actors are not ‘rational’ in the sense of methodological individualism, but the distance of practices in relations to designs of though allows a form of social invention (Bensa 2006). Generally, sociology is marked by opposition between non-modern societies, which are marked by personal relationships, and modern societies (western to begin with), characterised by impersonal institutions from where we find the strategic role of the distinction Gemeinshaft/Gesellshaft. Kinship is the central social structure for social anthropology as opposed to sociology which works on classes. We follow Goody (1990) when he observes that it is seriously necessary to nuance this opposition in noting that the game with multiple rules and characteristics does not only concern modern societies. The significantly different trajectories of the surrounding villages, which we will not develop here, shows that other collective strategies, giving a more important place to the clan, were possible.
6To verify these hypotheses, we will first show how the clan constituted the predominant institution in the village. The second part will describe the observable rupture in the 1980s with a redefinition of the uses of clans linked with the politisation of a part of the population and with economic change.
- 6 Essentially, the account of the villagers corresponds to that of Kamil Bey’s nephew (Diken 2005: 2 (...)
7Far from being an ‘immemorial’ institution, clans – without major structural change between the 1930s and 1970s – are a product of the Turkish Republic, and more precisely of the disappearance of the Sürgücü (formerly known as Avelanı or Awina) tribe [aşiret] of which the clans [merev] were the components. At the end of the First World War, the village of Kirazbahçe was controlled by Hacı Ahmed, an ağa owner of tens of villages, installed in Sürgücü market town and whose imposing house overhung the locality. This ağa imposed an ‘oppression’ [zülum] – a term used with a clear negative connotation used by our representatives – through his brotherMehmet Eyüp ağa installed in Kirazbahçe itself. The Sürgücü tribe had a very hierarchical organisation: the ağa collected taxed, collected a part of the harvest and designated the muhtar. It also carried out arbitrary roles, for example in the case of settling a kandavası. At the time of the Republic, Hacı Ahmed and his son Kamil Semin aligned themselves with the government of Ankara. But, during the insurrection of Sheikh Said (1925), the government suspected an agreement between Kamil Semin and the rebellion, although the Sürgücü tribe helped the government6. Kamil Semin was sentenced to death but, due to the intervention of his father-in-law, he was pardoned and exiled with all his family to Istanbul. His removal led to the disappearance of the tribe as a framework of mobilisation. The villages previously under the authority of the ağa became independent bodies and the merev were no longer encompassed in a tribes. The evolution of the institutions in this village, and others belonging to this tribe, contrasts with the regions where the government reinforced the roles of local notables (sometimes after some years in exile) to use them especially in the period of the one-party system.
- 7 Most of the names of the clans are family names.
- 8 Anthropological literature have much evolved on the so-called stability of segmentary models. We w (...)
8In the village studied, there are seven merev: Suleymani, Mukri, Alikoli, Mahmi, Sirahmadi, Zeydini and Ambeli, forming two alliances since the 1930s7. On the one hand, the Suleymani unites the Mahmi, the Sirahmadi, the Zeydini and the Ambeli. On the other hand, the Alikoli and the Mukri form an alliance against the others clans. The two alliances are in part territorialized as the village is informally divided in the lower district and the higher district, corresponding largely to the grouping of the clans: to the East for the Mukri and to the West for the Suleymani. The conflicts led to a progressive territorialization of the groups and one can suppose that, without the intervention of the state, the dominant clan would expel the minorities8. In the 1970s, the merev constituted the framework of interaction and of preferential competition: they defined the matrimonial sphere, organised the security of individuals, the every day social interactions, the alignment of votes, and, in certain limits, the organisation of agricultural work.
9The merev did not stem from tribal political organisation, where the ideology of kinship was concerned with an often reconstructed genealogy, but formed on the basis of a real kinship that certain individuals at least could effectively retrace for several generations. The merev was therefore an alliance of families, resting on the unity of habitation and production. Children were generally inserted in an independent house when they were married, but remaining in the part of the village where their clan was dominant. The merev defined the matrimonial sphere, contributing to their reproduction. Marriages preferably took place between allied merev, exceptionally between those opposed to each other, and almost never surrounding villages. Indeed, the links subsisting between brother and sister after the marriage, constituting a marriage between adverse clans, would potentially be the source of tensions as married women did not remain under the principal responsibility of their brothers, as is the norm for Arabs, but not entirely under the responsibility of the husband according to the Turkish model (Meeker 1976). The merev is a social construct since the links of solidarity were created by positive procedures (an obligation to participate in certain activities of the clan) and negative ones (prohibition of contact with other clans).
10Participation in collective activities and everyday sociability took its place in the circle of the merev. Contrary to other villages in Turkey, there was no communal room [köy odası] for the whole of the village or, the opposite, a multiplication of the oda on the basis of individual initiative (Yalçın-Heckmann 1991: 142), but seven oda, one per clan. One can put forward the theory that the recent implantation of the population and the belonging to a larger tribe [aşiret] initially plays a role against the emergence of a strong village identity. In the 1980s, the attempt to open a café, of which the localisation showed moreover that it was open to the Suleymani and their allies, was short-lived. Men were obliged to participate every day – for a duration which varied with regard to the agricultural work – in meetings of the oda, which mobilised them a large part of the day during winter. As Stirling says (1965: 238): ‘Attendance at a guest room in mid-winter is the only alternative to sitting in one's own household with wife and children. Not only is this uncomfortable and undignified, it automatically cuts a man off from male company since his male kin and neighbours will not visit him at home’.The participation in an oda worked like an open affiliation and a commitment of solidarity, it was the place where the social capital of each person was publically evaluated, notably by the skew of the gossip. At school, children were grouped in bands with regard to their clan affiliation and a certain competition or hostility characterised the relationships between the groups.
11The merev were the institution which allows man (women and children were never the target of violent action) to ensure their security and that of their property. The solidarity between men of the same clan implies a collective responsibility and adolescents of about 15 years were given rifles, contributing to the regular presence in the oda, to mark their passage in the virile age and their affiliation to the clan. Competitive relations between merev led to various confrontations since the 1920s. The villagers thought that that until the 1980s there was, on average, one death a year with strong variations from year to year – that is to say more than 40 deaths under the Republic and many injured. According to our sources, this figure would be equal to or above the average of neighbouring villages. The armament, and therefore the tactics of confrontation, would change with time. The dagger fights would disappear with the introduction of firearms in the 1950s. Shootings from one hill to another or more murderous ambushes were the common tactics.
12The causes of these conflicts are diverse according to our sources and recounted events. Certain villagers justify the regularity of the kandavası with the fact that the village is more populated and more deprived of water than its neighbours, leading to fighting for land. Contrary to this general proposition, the factual accounts show that the conflicts were not directly provoked by delimitations between grounds (despite the absence of a register) and there are few contestations on the subject. Furthermore, the land or abandoned houses by a clan were not occupied by another. The fields remained unexploited if inaccessible by a clan because of the fighting. In the same way, when houses near the centre of the village and belonging to the Alikoli were destroyed, the area stayed untouched.
13In the accounts, the abduction of women are generally put forward as the immediate origin of the conflicts. In the 1970s, a group of Alikoli and Mukri families installed within the Suleymani territory, was accused of the assassination of a Suleymani boy, responsible for having abducted a girl from the Mukri clan. These families were forced to moved to the outskirts of the village, abandoning their houses, which were then destroyed.
14The presence of members of a clan in an area normally belonging to another clan could equally cause tension, even a confrontation. In 1972, in a context marked by strong tensions between the merev, the Suleymani went to the plateau overlooking the village, normally belonging to the Mukri, at the time of the harvesting of the grape. A quarrel led to the exchange of fire, injuring a Mukri. In reprisals, an ambush held by the Mukri killed three Suleymani, and another Mukri died in the confrontation. Around twenty people involved in the incident took the hills leading to the formation of two armed bands patrolling the hills surrounding the village. After this day where four villagers were killed, the police intervened in a memorable way. Indeed, the police made all the men strip before the gathered people in the village, threatening the women to do the same if the suspects taking refuge in hills did not surrender. As a consequence, seventeen men surrendered (10 Mukri and 7 Suleymani) and were taken to Mardin where they spent a year and a half in prison before the Ecevit amnesty in 1974.
- 9 Notably from the fact that the ziarat is in fact constructed on communal land.
15No institution permitted arbitration between the clans, encouraging us – in accordance with local terminology – not suppose the formation of a new tribe. In the 1970s, a saint man, member of the brotherhood of the Naqshbandis, would have been able to play the role of arbitrator, as he did not belong to any clan (his family being from outside the village). However, his role as negotiator between clans remained marginal. The efforts of his family to create a ziarat around his tomb, and therefore capitalise a notoriety, are contested9 or meet a relative indifference. Besides, two mosques, held by two brothers, regrouped men with regards to their affiliation to their clan. It was in fact the hardening of relations between the clans which led to the creation of another mosque. This shows that the religious space depended on the clan more than the inverse, even if the mollah did not belong to a clan as they are derived from outsider families. During the fighting between clans, the mollah could try to stop the violence with a Koran on their head, but this was not effective (however, no one would target them). Some area existed however, notably a ziarat near the village, of which the sacrality put them outside of conflict and in any case outside of the competition for economic exploitation.
16Agricultural production organised itself partly according to the clans, notably for collective work and access to water. Certain common works formed part of the framework of the merev, notably the harvesting of grapes and the cooking of the bulgur. The shepherds held all the animals and, before 1980, the herds were regrouped according to the merev of the landowners. However, the merev is only a marginal element for the definition of property rights, for example, inheritance. In the 1960s and 1970s, the land was generally worked by brothers indivisibly, but a share was possible in case of a disagreement.
17The rare economic resource is not the land, but water. There is no river nearby, but only sources, with the result that the ground is not irrigated in a permanent way, contrary to the situation of nearby villages. A ‘traditional’ agreement (undated) between families organised an alternated irrigation of the field every two years. A guardian ([kerax] in kurmandji) oversaw the distribution. He had to ensure the respect of the rules regarding water-sharing (He woke the landowners in the night) and ensure also the absence of fraud. Furthermore, disagreements were frequent and could lead to quarrels, even violence if the mediation by the mollah or the muhtar was not accepted. The distribution of the water depended on the merev who negotiated it. Indirectly, the use of sources also depended on the muhtar, who generally favoured the clan system. In this way, a fountain was installed by a muhtar Mukri in the 1970s at the top of the village, further from the sources. The installation of the fountain was until then prevented by the muhtar Suleymani, as it was supplied to a distant source of three kilometres which provided for the part of the village dominated by the Suleymani, decreasing the resources of water of the latter.
- 10 Six Suleymani, two Mukri and one Alikoli. The chronology of the village is often done by the perso (...)
18There are institutions catered to the village, notably the muhtar and the council members [aza] who assisted it, but they were closely dependent on the clans. One elder [makul] chosen in each merev, or about fifteen for the whole of the village in the 1970s, designated the muhtar, who represented the village as regards the administration. The appointment of the muhtar depended on the solidarities of the clan, and, in practice, the latter was the representative of an alliance of merev. The neat predominance of the Suleymani with the muhtar can be explained by their demographic significance10. Elsewhere, the work for the common parts of the village was organised with regards to these affiliations. In this way, each merev held a part of the wall of the cemetery, as a result it was veritable collective work. Conversely, in at least one case, the clans of the village made alliances with those of the neighbouring village – equally structure by the opposition of clans, recreating at a superior level the same games of opposition. Finally, very few circumstances allowed individuals of merev not allied to get involved. To demonstrate this point, in the case where collective interests could be identified, the muhtar represented the village. For example, in the 1970s, a violent conflict opposed the villagers to the administration to make sure that the layout of the route would be in their favour. All the villagers, without regard to clans, blocked the route in an air-shooting.
19During the legislative elections, one can note that the votes corresponded to the solidarity of clans, but political participation was weak. Generally, an affiliation to the Turkish Republic was rarely demonstrated and no republican festival was celebrated in the village, the two festivals observed, l’Eid-e ramazan and l’Eid-e kurban, were religious. Political men, uniquely present at electoral campaigns did not introduce a clientelist system as they did not offer anything concrete, apart from a small amount of help to those at Savur or an intervention in the administration. This possibility explains why few villagers moved: about 10 out of hundreds of potential voters. The votes were case on the basis of your familial belonging. In the 1950s and 1960s, the Suleymeni voted for the CHP, locally represented by Kamil Boran, an Arab, re-elected many times. The Mukri and the Alikoli however voted for the DP and then the AP.
- 11 We often have a struggle to impose one rule against another: for example a bureaucratic relationsh (...)
20We saw that the clan system organised certain important aspects of the life of the village. Is it possibly to speak of a coherence of the schemes of thought which would participate in an ‘enterprise of integration’ ‘logic and moral’ of the society (Lenoir 2003: 47) ? Could the clan, and more generally kinship be seen as a ‘matrix of schemes structuring a vision of the social world’ (Lenoir 2003: 91) ? A coherence was effectively established by the skew of linguist procedures, homology representing an ‘analogical operator of differences’ (Lenoir 2003: 50-52 citing Bourdieu 1979: 55 note 13). In this sense, the metaphors are neither simply rhetorical figures, nor pure rational models, but at the same time means of action and schemes of thought. In the case concerning us, the actors maintained the indissociability of the clan, of family, of Islam, and of individual and collective honour. In this sense, the village experienced a coherence, if not of practices, of the dominant discourse – that of the oda11. However, this coherence was carried out at the price of the dissimulation of the dependence of the local order in relation to the encompassing society and its institutions.
21This expression could not dissimulate the instability of this ideological construction, as most legitimate expressions, as this put forward the necessities and the natural functions at the heart of their argumentation. The fragility of homologies constructed to legitimise each institutions by others – a ‘horizontal legitimation’ – opened up the possibility of crisis when the context changed. The ideological work to present honour, family and the clan as indivisible came into conflict with different rules as the village is an englobed entity in Turkish society. Thus the villagers were in contact with institutions diffusing other models: the state, Islam, the market etc. In the 1970s and 1980s, those alternative rules could mobilise resources: politisation, family, nationalism (Kurdish of Turkish) and the market.
22How did the merev gradually lose their functions, leading to a redefinition of clan uses, and more largely, of kinship? Socio-demographic evolutions led to the appearance of a population less integrated in the clan system. According to a former muhtar, the population which was of about 1000 in the 1960s reaches 3500 in 1981 and the peak was probably reached at the moment of the evacuation of the army in 1993. What were the manifestations of this demographic push? One added another floor to one’s house (the date of the work being registered on the house) and the village became extended towards the hills, new houses being further away from sources. In this way, the district of Carısefer was constructed in the 1970s, Kabla and Beroj in the 1980s. Elsewhere, the village was situated in a surrounded basin of hills, which were progressively deforested by the inhabitants to collect wood for heating (to make coal), indicating lower wood supplies. Because of weak output, land revenue earned by each person began to decrease in the 1960s and 1970s, our sources noting cases of malnutrition among children of the poorest families. Because of demographic pressure, cultivated land was further and further away from the village, up to a walk of one and a half hours away.
23Migration (temporary or otherwise) to the cities is one response to the relative over-population. From the 1960s, seasonal migration proliferated: in general, one or two people per family had experience of temporary work in the cities. Göç-Der (2002) has published an eye-witness account which follows the same lines as those collected by us, of a village resident who speaks of his departure for Istanbul in 1972 and his return in the autumn for the grape harvest. In the city, he sells old clothes with his friends, and soon his family would come to settle with him in Istanbul because of the lack of work in the village. These young men (between 15 and 20 years) make up a geographically mobile population and one which is less economically and socially integrated in village life. In the city, their clannish identity does not enable them to mobilise resources (places to stay, mutual help), in contrast with the more widespread affiliations, such as Kurdishness or geographical origin [hemşehirlik]. Through migration, the young people from the village, confronted with the same experiences, came together to share the same places of work and accommodation. Some representatives highlight the fact that the young people from different clans continue to club together after their return to the village.
24Though it is difficult to presuppose the existence of a simple causal link with what precedes, our sources underline a declining importance of the oda in the 1970s. For a section of young people, the economic capital or other things that their parents can pass on are devalued, whilst urban migration becomes the likely pathway for many. In particular, the land and social capital linked with family or clannish honour are devalued outside the village boundaries. Consequently, social pressure is less effective against the deviants, the village loses its coherence as an economic unit; all because a growing portion of income comes from the city and, in terms of a living space – because the experience of life is so markedly different.
25In the 1980s, the military coup, the PKK guerrillas from 1984 radically alter village life by offering new possibilities of engagement stemming from the presence of political parties or governmental organisations. Membership of the PPK by a number of the inhabitants would be answered by the engagement of other villagers in the government militia [korucu]. These mobilisations occurred outside of, and sometimes openly against, the clan solidarities, which makes it all the more surprising since there was no politicisation before 1980 – the affiliation of two teachers to a radical leftist group having no influence and more besides not being known to most of the villagers.
26The PKK would mobilise the Kurdish identity through its propaganda, but through a scheme of nationalist thinking. During the 70s, the Kurdish identity is known, valued, and Mustafa Barzani is a hero throughout the whole of the Kurdish regions – his fights often being notorious from individual narrative accounts or Kurdish language radio broadcasting from Erevan. Yet, locally, the ethnic difference is ultimately more evident in relation to the Arabs (there is no local Turkish population). For instance, the Kurds, even administratively linked to the city of Mardin, referred to themselves as being from Diyarbakır in order to draw a distinction with Mardin city which is mostly Arab (until the 1990s). The migration experience has probably played a part in the advancing of the Kurdish identity as against the Turkish one, but the key is that it became possible to mobilise the Kurdish identity from within an organisation, which offered resources to a category of individuals.
27Dozens of young men and women from the village would join the PKK at the end of the 1980s and in the 1990s. According to sources, the number of deaths in the guerrilla ranks is about 10 or 15, of which four or five were female. Recruitment begins from the mid 1980s at the moment where governmental pressure begins to make itself felt. The PKK fighters spend the night in the village, sometimes reunite the inhabitants and organise meetings in order to convince the inhabitants to support the movement. Who becomes a PKK militant? The militants recruit, apparently without many exceptions, from amongst the young people who are familiar with migration. The young men and women (often around 15 to 16) become PKK sympathisers and some go all the way and become involved in the organisation. Is militant activity also a manifestation of clan affiliation? The answer is in the negative, but it is possible that the Alikoli clans were engaged a little more frequently inside the movement than others. These young people belong to different clans and are integrated in the same organisation which unequivocally rejects clannish affiliations both in its ideology, and according to witnesses, to all intents and purposes just as equally in practice. The witnesses concur on the fact that clannish identities do not survive the integration in the organisation. Equally, young single women engage themselves in the movement in open contradiction with village norms, family honour being de facto reserved for the men (fathers and brothers), even though the movement had a puritanical ideology and even if the female camps are in principal separated from the men. The PKK thus appears as an alternative to the familial and clannish order, which prohibited women from studying or having remunerated employment. Within the families, the daughters threatened (without necessarily carrying out the act) the fathers with joining the guerrillas.
- 12 A PPK commander, Zeki, originally from a nearby village, was famous in the region in the late 1980 (...)
28The military pressure of the PKK led the government to form militia [korucu] in the context of an anti-insurgency strategy. The korucu were formed in villages by the government at the end of the 1980s. About forty people were members, originally from all clans12. There were diverse motivations for joining sometimes responding to a financial incentive (the pay of the government), sometimes to opposition to the PKK, which is described as an atheist party – at least for its leaders. For the muhtar, the participation in the militia is above all the means of protecting the integrity of the village, a strategy which would end in failure. Lastly, for some inhabitants, an affiliation with the Turkish nation made sense. If the state was barely present, the majority of the villagers had been called for military service, where they had leaned basic Turkish, and some were called to serve in Korea in the 1950s. But whatever the initial reasons for joining, membership of militia was not merely a formality because the militiamen patrolled together and therefore found themselves exposed to the same dangers. Furthermore, they shared a communal house which served as a meeting place, a little akin to the oda, and developed personal bonds. The PKK attack against the headquarters of the militiamen, which led to the death of several people, is seen as a moral crisis. The pressure of their wives led the militiamen to relocate in Mardin and, indirectly, caused the korucu to leave the village.
29Due to the very presence of armed groups (PKK or korucu), the clans lost their central place of security. Thus, from the 1980s, we observe the disappearance of the vendetta [düşmanlık], even thoughweapons flowed abundantly in the region and confrontations between villages and within villages multiplied. The fact of accompanying a guest with weapons clearly survives today (as the actors themselves say), a role more than a response to the lack of safety. This evolution has the indirect consequence that the inhabitants now have a propensity to resort more easily to the arbitration of official institutions. In 2006, some Mukri cut the water pipe supplying the school and the mosque. This unilateral act did not result in reprisals or threats, but gave rise to the opening of legal proceedings, which proves the recognition of the power of the state and the disappearance of an alternative regulatory system. Formerly, legal proceedings were very rare and amongst the inhabitants; arbitration was carried out by the mullah, or inside the clans. The military operations and, later on, the return to the village have enabled the penetration and permeation of state administration.
30The information gathered indicated that the merev disappeared in the migration. On the one hand, the organisation of the migration, notably the choice of future residence, did not depend on a collective decision of the clan, but of individual or familial decisions. The presence of parents could also play a role, explaining for example the frequent migration to the district of Bayrampaşa in Istanbul, though the village was not reconstituted. The inhabitants tended to regroup but did not carry out a particular role, although many worked with textiles.
31From 2000, villagers were progressively authorised to return. In 2006, the population stood at more than 100 families in winter. However, the village lived largely according to an artificial economy, as the principal resources did not come from agriculture but of retirement and the sending of money from those families living in the cities. In practice, the land was no longer an important a stake as before. The merev remained importance in local, political life, notably for the election of the muhtar and the distribution of water. This was perhaps due to the fact that the more politicised part of the population (korucu or militants) left the village, the elderly constituting the majority of those who returned. Marriages took place in a preferential manner within the clans, but the matrimonial sphere was not defined as strictly: a marriage with a family of the neighbouring village did not cause as much controversy. We can recall the observations of Fliche (2007) when he underlines the fact that the village was seen by migrants as a complex space, with a part of nostalgia proper to patrimonialization, but also as a place of conflict. The absence of any synthetic study leads us to finally become more prudent in our conclusions on the disappearance or the change in clan relations in the urban area.
32The redefinition of clan usages implies a redefinition also of the situations concerning the kinship sector. The unity between the family and the clan (or tribe, as before) disappears and the family and represents the only institution belonging from the kinship sector. This sector, with one sole institution, from now on has a reduced importance. Furthermore, there is a change in family models, between nuclear families and the middle classes in Istanbul or Diyarbakır and the large, poor families.
33The relationship between schemes of thought and resources is complex and the actors have limited room to manoeuvre, but real nonetheless. Other villages, with similar structures, experienced a different evolution. The villages which korucu benefited from state help and the clan system could be reinforced to a certain point, even if rural migration remained a general phenomenon and if we remain septic in relation to the idea of a significant retribalisation. In certain cases, clan affiliations served as a framework of mobilisation, the resources of the state being use in this framework, this in turn reinforcing and legitimising the existing system.