During the bitter war with the guerrillas of the Kurdistan Workers’ Party (PKK) in the 1990s, the Turkish army and paramilitary “village guards” depopulated and burned villages in southeastern Turkey on a systematic scale. Clearance of the countryside and resettlement of the rural population, from which the PKK drew membership, logistical support and intelligence, was considered crucial for a successful counterinsurgency. From 1993 onward, the evacuation and burning of villages were no longer simple reprisals against villages suspected of supporting the guerrillas, but part of a “draft or destroy” approach that imposed upon villagers a frightening dilemma. Either they could become members of the village guards and take up arms against the PKK themselves, or they could face punitive measures, including expulsion and destruction of their villages, which aimed at the forced migration of civilians to urban centers, where state forces were better able to control the population and battle the insurgents.
In the absence of reliable statistics, it is difficult to assess the exact dimensions of the village depopulation and destruction. Government figures indicate that between 3,000 and 4,000 villages and hamlets were “evacuated” and destroyed, displacing 55,000 households or between 350,000 and 380,000 people, most of them Kurds. The number of settlements depopulated and destroyed is not really in dispute, but the number of people affected is a subject of great controversy. Human rights organizations claim that Turkey deliberately presents low numbers to camouflage the magnitude of the displacement. These organizations variously estimate the number of displaced persons at 1.5 or 4 million. Both official and non-official figures indicate that some 93 percent of the displaced people wish to return to their villages, but clearly it makes a vast difference if this number is 93 percent of 380,000 or of 4 million. Successive Turkish governments have devised various resettlement plans, but failed to implement them. Urban resettlement plans failed because of lack of funds and lack of concern on the part of the authorities, but rural resettlement plans failed because of institutional discord.
The latest government resettlement scheme, the “Return to Village and Rehabilitation” program of 1999, came about as Turkey endeavored to meet the requirements for eventual accession to the European Union. As the EU accession debates continue, this scheme is also failing. Reading the fine print of the plan, moreover, reveals that it is fundamentally more concerned with consolidating the Turkish state’s control over the formerly war-torn provinces than with returning displaced villagers to their homes.
Clearing the Countryside
Before the late 1980s, the notion that a small group of highly motivated but poorly equipped Kurdish guerrillas could defeat the second-largest armed forces in NATO appeared to be a delusion. Previous attempts at armed struggle in the Kurdish areas of Turkey had failed badly, resulting in the deaths of militia leaders and cadres. But the “liberation of Kurdistan” — as the PKK called it — was close at hand by 1990. The army was rapidly losing control and was only able to retake the initiative in 1993–1995 after thoroughly revising their strategy and unit structure.
PKK strategy was to wear down the state security forces through constant, small-scale attacks. Through these operations, Turkish forces were steadily driven out of rural areas. The concomitant army strategy prior to 1991 to concentrate in the larger towns and cities, and to defend supply and communication lines, contributed to the PKK fighters’ success, as it left the smaller villages and hamlets under the guerrillas’ control. The army sweeps through the areas believed to be hiding guerrilla camps were ineffective with fighters slipping away as troops got into position. Gradually, the Turkish general staff came to believe that victory would be on the side of whoever was able to control both the population and the territory. They intended to prevail regardless of the cost to the population, if necessary by separating the population from its rural lands. 
In the spring of 1991, the Turkish general staff announced a reorganization of the army and the doctrine of “field domination.” The army shifted from a relatively cumbersome divisional and regimental structure designed to fight a war against the Soviet Union to a relatively flexible corps and brigade structure — better for a war of movement against the PKK guerrillas. Instead of fortifying garrisons and conducting periodic sweeps, the army would use its numerical superiority to “clear and hold” areas under the control of or penetrated by the PKK. The army began to approach its war with the PKK with guerrilla tactics, at a time when the PKK was preparing for a conventional war to carve out an area for the establishment of a provisional government. Commando brigades and special forces were deployed to live and fight in the field day and night. These units kept constantly on the move within their assigned zone, searching out and ambushing PKK fighters. Simultaneously, the army carried out its “draft or destroy” campaign in the countryside. As a result, the number of village guards increased from 5,000 in 1987 to 67,000 in 1995, and thousands of villages were left in smoldering ruins. The depopulation and destruction of villages caused a massive influx of people into the nearby cities of Van, Batman and Diyarbakir, as well as Istanbul and other western cities. There, the displaced rented accommodations, crowded in with relatives or constructed shanty dwellings.
The expulsion of the rural population was a violent army campaign marked by summary executions and the destruction of livelihoods. Resettlement of the displaced was not a tightly organized scheme taking the displaced to precise locations, but merely a track that determined the direction of forced migration to the cities, where the displaced had to earn a living without any support from the state. More formal plans for the urban resettlement of the rural populations failed. The most pronounced examples of these urban resettlement plans were the toplu kondu and toplu çiftlik plans of 1994.
The toplu kondu (collective shelter) project was designed as a hybrid of the shantytown dwellings known as gece kondu (literally, “shelters built overnight”) and the government housing program called toplu konut (collective housing). In the toplu kondu blocs the government would provide for a one-floor basic shelter of 538 square feet built on a piece of land varying in size between 2,100-4,300 square feet. In accordance with their own needs and means, inhabitants were supposed to construct additional floors, extra rooms or workplaces. Several of these toplu kondu blocs with a total of 8,000 shelters were planned outside the cities of Adana, Urfa, Diyarbakir and Gaziantep, far away from the depopulated areas. Another 2,000 shelters were planned in the proximity of several district towns. Turkey asked the World Bank for a loan of $50 million to commence the scheme, but the World Bank dropped out for reasons that were never explained.
A similar plan envisioned the construction of large settlements with an average population of 1,000 inhabitants. These settlements were referred to as toplu çiftlik modeli projesi (collective farm model projects), or simply toplu çiftlik (collective farms). Notwithstanding its name, the toplu çiftlik plan was meant to resettle the population in the vicinity of urban areas, where the resettled people would again be provided with shelters of 538 square feet. Additionally, the displaced would be given access to agricultural land, either to 50 acres of irrigated land or to 120 acres of rain-fed land. In a circular dated October 24, 1994, Prime Minister Tansu Çiller officially announced the collective farm plan, and in December of that year the government contacted the Social Development Fund of the European Council for support. Çiller requested $278 million for the implementation of the project, but Turkey was soon compelled to abandon its application. The main reason for the failure to raise funds, according to a Dutch diplomat, was European resistance to the idea that Turkey would shift the costs of village depopulation and destruction onto Europe. The Europeans considered the scheme the tail end of a policy of forced migration, devised to keep the displaced permanently away from their former villages.
Villages of No Return
Rural resettlement was first considered in 1995, when the coalition government of the Republican Peoples’ Party (CHP) and the Justice Party (DYP) proposed the “return to village project of the southeast restoration project.” This plan was designed to resettle people in rural areas that had been depopulated in the years before — and as such set an important political precedent. But the idea of allowing the displaced to return to their areas of origin was opposed by the military and the regional governors, who claimed that the project posed security risks. In 1997, Prime Minister Mesut Yilmaz of the Motherland Party revisited the idea, announcing that his government would support the return of the displaced population to their villages. Even though a return was made conditional on the ability of the army to provide security, the implication was that the depopulation was an anomaly and return was inescapable.
In spite of the announcements by the DYP-CHP and Motherland governments, not much happened until 1999, when Prime Minister Bülent Ecevit ordered the Regional Development Administration of the Southeast Anatolia Project to coordinate research into resettlement of people in the depopulated areas; in the end, research was carried out in collaboration with the Turkish Social Sciences Association. The research included surveys among inhabitants of 297 villages, said to be 9 percent of the depopulated villages.  The results of the study have been committed to writing in a 12-volume report, one volume for each province in which “return to village and rehabilitation” projects were supposed to be implemented. The “East and Southeast Anatolia Return to Village and Rehabilitation Project Sub-Region Development Plan,” as the report is called in full, is shrouded in secrecy and only distributed among the state institutions involved. These are the Office of the Prime Minister, the Southeast Anatolia Project, the General Directorate for Village Affairs, the governors in the affected provinces and several other departments. The secrecy gave rise to misinterpretations that the project is an ad hoc scheme.  Rather, the project is designed as a comprehensive scheme whose implementation involves several state institutions — from the village affairs directorate to the provincial governors to the Ministries of Education and Health.
“State Rehabilitation Doctrine”
The 12 provinces where the “Return to Villages and Rehabilitation” plan is to be implemented are Batman, Bingöl, Bitlis, Diyarbakir, Elaziğ, Hakkari, Mardin, Muş, Siirt, Şirnak, Tunceli and Van. All of these provinces had been directly affected by the war between the Turkish armed forces and the PKK guerrillas, and all saw extensive destruction of villages. The plan is not referred to as a resettlement scheme, but as a development plan.  The subject of the plan’s special attention is not the displaced, but the area’s settlement structure. The depopulation of villages is even looked upon as an opportunity for the development of an appropriate configuration of settlements. “Apart from social and economic problems, the event of evacuated villages in East and Southeast Anatolia has created new opportunities and dynamics for the formation of new standards that can accomplish a new rural settlement pattern; for the transition from dispersed and unsuitable settlement units towards settlements units of endurable size and potentials.”  The concept of “rehabilitation” appears in the document, but its definition differs from the generally accepted one. In the international literature, the concept of rehabilitation is used to refer to actions that aim at reestablishment of incomes, livelihoods and social systems. Yet in the Return to Village and Rehabilitation plan, rehabilitation is defined as something physical — such as improvements to roads, highways and buildings. 
For the sake of the design of a new settlement structure, the plan introduced two concepts: the sub-region (alt-bölge) and the center-village (merkez köy). The concept of a sub-region is defined as a cluster of settlements distinguished from other settlements by economic, cultural, administrative or social characteristics. Mutual affinity between people and villages in a sub-region is taken as a starting point for their joint development. The center-village is defined as the one settlement within a sub-region that by its size, location and infrastructure could be turned into a junction for the other settlements and become an intermediate layer between the district town and small villages and hamlets. In sum, the plan reveals a political doctrine for rural resettlement that may be called the “state rehabilitation doctrine.” It is concerned with a search for the optimal settlement pattern for the establishment of state institutions in the countryside.
The Return to Village and Rehabilitation Sub-Region Plan proposes 12 pilot projects, one pilot per province. The pilot projects are not about 12 settlements, but about 12 sub-regions, each of them including one center-village and a number of dependent settlements, both villages and hamlets, varying between 6 and 42. The total number of settlements involved is 12 center-villages, 77 villages and 105 hamlets, or 194 all told.
The pilot plan does not mean that 194 depopulated villages will be reconstructed, for not all the settlements included in the pilots are depopulated settlements. In one case — the settlements that are part of the Yeniköy pilot project in the Çüngüs District in the Diyarbakir province — the villages and hamlets are all inhabited. As the government study explains: “Just since the end of the 1980s and in particular since the 1990s, for reasons of terror village evacuation in East and Southeast Anatolia occurred on a wide scale. It is necessary to give priority to the issue [in these areas]. The solution to the problem is not to stay within the boundaries of a return to the evacuated villages…. Villages that have not been evacuated at all have to be included in the alternative models that are part of the framework of the sub-region development plan.”  Other pilot projects, such as those in Batman, Bingöl, Bitlis, Muş and Van, are composed of both depopulated and inhabited settlements.
Villages now inhabited by village guards recruited during the original scorched-earth campaign are included in the pilot projects, meaning that prospective returnees are to be settled next to village guards. Available information indicates that such is the case with regard to the projects in the provinces of Batman, Bingöl, Bitlis, Hakkari, Muş, Şırnak and Van. Finally, not all the “evacuated” settlements are eligible for repopulation. The pilot in the Yağızca sub-region in the Genç district in Bingöl shows that some of the depopulated settlements (Küçükbayırlı and Bayırlı) are not to be reconstructed. The population will be resettled at locations more suitable for the state.
In any case, implementation of the pilot projects has met an institutional backlash. The governors of Hakkari, Siirt and Van provinces overtly disallowed implementation of parts of the program for reasons of security. The governor of Siirt stated that the dispersed settlement pattern is “the major problem in the war on terror” and subsequently did not give permission for a return to nine depopulated villages.
The field domination doctrine is fundamental to understanding the military logic behind the destruction of villages and the forced migration of the inhabitants. The Turkish armed forces were neither willing nor able to garrison all the rural settlements, nor could they turn civilians in all these settlements into village guards. On a tactical level, resettlement of the rural population aimed to deprive the guerrillas from food, shelter, intelligence and recruits. On a strategic level, future domination of the area was linked with the accommodation of people in larger settlements. The authorities wanted to drive displaced migrants from guerrilla-controlled small rural settlements to government-controlled urban areas — though not in the random way in which the migration actually occurred. This strategic goal accounts for the almost complete absence of resettlement schemes since the “draft and destroy” campaign.
Earlier schemes laying out a possible return to rural settlements failed to materialize because of institutional discord. The “field domination doctrine” developed by the army and the “state rehabilitation doctrine” developed by civil authorities do not agree with each other. Where one values security absolutely and therefore the depopulation of parts of the countryside, where supervision and control are difficult to establish because of the dispersed settlement structure and low population density, the other contains a component of development and repopulation of the countryside. But even if implementation of the “Return to Village and Rehabilitation” project overcomes bureaucratic resistance and goes forward without additional constraints, that would not mean that all of the hundreds of thousands of displaced persons will be able to return. Only some of the villages scorched in the Turkish army’s counter-insurgency campaign of the 1990s are included in the project; others are simply villages of no return.
 Umit Özdag, The PKK and Low-Intensity Conflict in Turkey (Ankara: Frank Cass, 2003).
 Oğuz Oyan and Melih Ersoy et al. Dogu ve Güneydogu Anadolu Bölgesi Köye Dönüs ve Rehabilitasyon Projesi Alt Bölge Gelisme Plani (Ankara: Bölge Kalkinma Idaresi ve Turk Sosyal Bilimler Dernegi, 2001).
 Human Rights Watch, Displaced and Disregarded: Turkey’s Failing Village Return Program (New York, 2002).
 Oyan et al, vol. I, p. 1.
 Ibid., p. 7.
 Ibid., p. 5.
 Ibid., p. 7.