Constructions of ethnic identity in the late Ottoman Empire and Republican Turkey: the Kurds and their Others


Martin van Bruinessen
Utrecht University



Ziya and the Kurds 

In 1914, Gaspiralı İsmail wrote in Türk Yurdu that Turkish nationalism was a reaction to the earlier nationalisms of other peoples in the Empire, not their cause: that of the Kurds had, he claimed, begun twenty years before, that of the Albanians thirty years; Armenian nationalism was at least forty years old, Bulgar nationalism sixty, and that of the Greeks eighty. 

This would place the beginnings of Kurdish nationalism around 1894, the year in which Mehmet Ziya Bey (the later Gökalp) shot a bullet into his head. Among the explanations for this attempted suicide put forward by later authors, one particular form of Durkheimian anomie is relevant to our subject. Rohat, a contemporary Kurdish author, suggests that Ziya's search for identity and his inability to accept his Kurdish background lay at the roots of his mental crisis. 

Ziya may in fact have been one of the earliest thinkers in Eastern Anatolia to reflect upon his ethnic identity in a novel way — although his writings do not allow us to retrace the development of his thought as far back as 1894. An unreflective awareness of Kurdishness and of ethnic differences between Kurds and others has existed at least since the incorporation of the Kurdish emirates into the Ottoman Empire. The Sharafname, the history of these emirates by Sharaf Khan of Bitlis, completed in 1597, and Ahmad-i Khani's Kurdish romance Mem û Zîn of 1695 are clear expressions of Kurdish self-awareness, the latter even of sentiments one is tempted to call 'nationalist'. Such outsiders as Evliya Çelebi also wrote about Kurds as if it was unambiguously clear who were and who were not Kurds. The question as to whether townspersons, non-tribal peasants, Kızılbash and similar sects were also Kurds or something else was usually not posed. 

Gökalp still adopted much of this traditional view of the Kurds. In the study of Kurdish tribes that he wrote in 1922 at the request of Minister of Health Rıza Nur (but which was not published until Kurdish nationalists got hold of it and printed it in 1975), he simply adopted the Sharafname's inclusive view of who were Kurds and refers, among others, to the Kurdish poet Ahmed-i Khani, whose work Mem û Zîn had been printed a few years before. 

Rather then declaring some or all of the Kurds to be of Turkish ethnic or racial stock, as later ideologists were to do, Ziya only came to deny his own Kurdishness and to pretend that his family was of Turkish origin. He appears in his later years to have blamed his earlier mentor Abdullah Cevdet, however, for continuing to assert his Kurdish identity. By then, Ziya was convinced that ethnic identity can be a matter of conscious choice instead of birth. 

In Türkleşmek, İslamlaşmak, Muasırlaşmak (1918, but containing material published in Türk Yurdu from 1913 on), Ziya criticizes the Istanbul şehrî (city-dweller, almost used as a synonym of 'Ottoman'), whose (ethnic or regional) identity has been 'erased', and who is neither Turk nor Kurd or Laz or Arab or Arnavut: şehrî'nin milliyeti yoktur. This passage appears to mark his definitive break with Ottomanism and his opting for nationalism based on Turkish ethnicity. (Other ethnic groups had never let themselves be cheated by the Ottomanist ideology, Gökalp maintains, only the Turks had.) In this period, incidentally, the names Kurd, Arnavut, Laz, etc. appear to have primarily a regional rather than cultural-linguistic connotation for him: they refer to the Muslim populations of East Anatolia, Rumeli and the eastern Black Sea coast. 

Missionaries and the Kurds 

Some scholars place the beginnings of Kurdish nationalism even earlier than Gaspiralıİsmail did. In 1880, the American missionary doctor Cochran in Urmia received a letter from the Kurdish Shaykh `Ubaydullah of Nehri which he understood to be an unambiguous expression of Kurdish nationalism. "The Kurdish nation ... is a people apart", he translated the key phrase of this letter into English. "Their religion is different, and their laws and customs are distinct... The chiefs and rulers of Kurdistan ... one and all are united and agreed that matters cannot be carried on in this way with the two [Ottoman and Qajar] governments, and that necessarily something must be done... We also are a nation apart. We want our affairs to be in our hands." 

We only have the English translation by Cochran of this letter, and Cochran may have read more nationalism in the shaykh's words than there was. The key phrase probably is "Their religion is different". The shaykh's letter was a response to fears that the Russians were going to help the Armenians establishing a Christian state in eastern Anatolia. (Two decades before, when Shaykh Shamil was waging a jihad against the Russians in Daghistan, Shaykh Ubaydullah's father had dispatched a number of Kurdish warriors to support him.) The shaykh saw his people probably not so much apart from the Ottomans (to whom he continued pleading loyalty even while his tribal followers invaded western Azarbayjan) as apart from the Armenians and Assyrians. 

The establishment of the Hamidiye regiments (1892) allied parts of the (Sunni) Kurdish tribal elite even more strongly to the Sultan than they had been before. Until the early 1920s, most members of this stratum rejected the idea, unsuccesfully propagated by a handful of modern-educated men, that they belonged to any 'national' entity apart from that defined by loyalty to the Ottoman dynasty. The Hamidiye regiments could plunder and kill with impunity, of which not only Armenians were victims but other tribes and especially peasants as well. Ziya Gökalp, incidentally, led demonstrations of the citizens of Diyarbakir that led to the abolishment of the Hamidiye in 1909. (They were, however, soon revived as aşiret alayları.) 

Fredrik Barth's ethnic boundaries and the Kurds 

Taking a cue from Barth, it may be helpful to ask which social boundaries involving the Kurds were and are most carefully maintained. Prior to the First World War, the most conspicuous of all boundaries was that between Muslims and Christians (or Jews), which in northern Kurdistan practically meant a boundary between Kurds and Armenians. Turkish and Arabic-speaking Sunni Muslims were much less alien to the (Sunni) Kurds, and the boundaries between them were fuzzy. In fact, Ottoman documents contain numerous references to mixed nomadic tribes, consisting of Kurdish and Turkish-speaking sections, which indicates that there was not a hard boundary separating Turkish and Kurdish tribesmen. At least one Arabic-speaking tribe, the Mahallami, apparently considered itself as Kurdish and was considered as such by other Kurds. (Presently, however, most Mahallami insist on not being Kurds.) The large Milli tribe was predominantly Kurdish but incorporated some Arabic-speaking sections. 

In pre-Republican times, the speakers of Zaza, an Iranian language related to Kurdish, were to my knowledge never mentioned as a distinct group but always considered as Kurds (even by Ziya Gökalp in his 1922 study). It was Hasan Reşit Tankut, the Türk Ocakları’s inspector for Eastern Turkey, who pointed out the linguistic and other cultural differences between Zaza and Kurmanji speaking groups and advocated driving a wedge between the two in order to turkicize them more efficiently. (The emergence, in the 1980s, of a Zaza nationalist movement that declared the Zazas to be an ancient nation that had always been different from the Kurds was to be welcomed and financially supported by certain circles in Turkey’s intelligence establishment.) 

The Yezidis were in a different position. On the one hand, they were not considered as Muslims by their neighbours, and they emphasized their different identity by various external signs; on the other hand, being tribespeople and Kurdish speakers they were much closer to the Muslim Kurds than most Christians were. There was a clear boundary between Sunni Kurds and Yezidis but its importance varied from time to time. By the beginning of the present century several tribes in the Tur Abdin, the mountains stretching east from Mardin in present Turkey, appear to have had Yezidi as well as Muslim segments (which probably was due to gradual conversion of Yezidi tribes to Islam). 

It is not entirely clear how sharply defined the boundary between Alevi and Sunni Kurds was. Several 19th-century sources (including some Ottoman salname) speak of the Kizilbash (Alevi) as if they are a separate ethnic community, irrespective of language. Language, in other words, did not play an important part as a boundary marker, but religion did; and the boundaries separating Christians from Muslims were much more unambiguous than those between heterodox and orthodox. 

Perhaps the most important boundary of all, however, was that between tribal and non-tribal populations. The Muslim-Christian boundary was especially sharp where it coincided with that separating tribesmen and non-tribal peasants or craftsmen. Where Christians were tribally organized and militarily strong, as the Nestorians of Hakkari and the Jacobites of the Tur Abdin still were for most of the 19th century, they were treated as equals by Kurdish tribesmen. There were even Kurdish tribes that incorporated Christians as members (Gökalp still lists several of these tribes). The non-tribal populations of the region included speakers of Kurdish, Zaza and Gurani as well as Armenian, Aramaic, Arabic and perhaps Turkish, and there were  Sunni and Alevi Muslims among them as well as Christians. The tribesmen made no sharp ethnic distinctions among these non-tribal groups, referring to them by the blanket term of ra`yat ("subjects"), by slightly more precise terms such as feleh (for Christian peasants, especially Armenians) and kurmanc (for Muslim peasants in northern Kurdistan), or by terms of local scope that differed from region to region. The tribesmen referred to themselves simply as `ashiret ("tribe") or as kurd

One other relevant boundary should be mentioned: that between the representatives of high Ottoman culture (military-bureaucratic officials, the higher religious functionaries, part of the urban notables) and the various local populations. The former constituted a quite distinct group, maintaining its distance from the vulgus by its use of an artificial language, Ottoman Turkish, and an elaborate etiquette. It is similar to, but probably not identical with the boundary that Ziya Gökalp signalled in 1913 when he blamed the ehrî for having 'erased' their ethnic affiliations. 

By taking these boundaries as the primary criterion of ethnicity, we gather that, by the early 20th century, the core of the Kurdish ethnie consisted of the Kurdish-speaking Muslim tribes. To them one should add the small but important group of urban notables, who often acted as the tribal chieftains' partners, representatives and brokers. The Zaza and Gurani-speaking tribes, or at least the Sunnis among them, living in similar ecological environments and sharing a common history with their Kurdish-speaking neighbours, were to most purposes also part of the core. Alevi, Yezidi, Shi`i and Ahl-i Haqq tribes were more peripheral, and the non-tribal peasantry, whatever their language or religion, were not considered, and did not consider themselves, as part of the ethnie

Kurdish nationalists and the Kurds 

This composition of the Kurdish ethnie was reflected in the membership of the first Kurdish nationalist associations in Istanbul. These were mostly educated members of the tribal or religious elite, and overwhelmingly Sunnis. We find speakers of the northern and southern Kurdish dialects as well as a few Zazas among them, very few Alevis and no Yezidis. Not surprisingly, there were no members of peasant or urban craftsman backgrounds, and in the Kurdish rebellions of the 1920s and 1930s these social strata remained aloof. 

It was a thin stratum of educated Kurds who took it upon themselves at first to 'educate the Kurds' (the most common term used in Kurdish print media from 1898 to 1918) and later to 'awaken' the Kurdish nation (an expression that regularly occurs in the journal Jîn, which appeared in 1918-1919 and in the memoirs of the activists of those years). 

The first Kurdish journal, Kurdistan, published initially in Cairo in 1898, emphasized its educational mission. Each issue contained articles on contemporary world politics, of which the editor wished the Kurds to be more aware. There were frequent injunctions to read and study more, and repeated appeals to Kurdish chieftains to make efforts towards the education of their subjects — an indication perhaps that not only the tribal elite but also subjected peasantry are included in its idea of the Kurds as a people. The journal otherwise still represents an unreflective traditional conception of Kurdishness, quoting freely from Ahmad-i Khani and writing with nostalgia of the Kurdish emirates of the past. 

In Jîn we find more conscious efforts to 'construct' Kurdish ethnic or national identity, and efforts to go beyond the accepted view. Khalil Khayali asks, perhaps in response to Ziya Gökalp's contempt of the şehrî, "Are the (Turkish-speaking) townspeople of Kurdistan Kurds?", and gives an affirmative answer. To do this, he has to ask what it is that makes a person Kurdish or Turkish; in the case of these townspeople, it is their assumed relationship with the unambiguously Kurdish people from the surrounding countryside. Memduh Selim (who later was to play an active role in the Ararat rebellion) in several articles deliberately attempts to create 'national' symbols, and calls upon the Kurds to respect their 'national holidays'. (It was to take several decades before the Kurds 'rediscovered' Newroz as their chief national holiday — and a few decades more before pan-Turk nationalists 'rediscovered' that it was really the commemoration of their ancestors' escape from Ergenekon.) Abdülaziz Yamülkizade writes that with the coming of Islam and the Shi`i-Sunni scission, the Kurds became divided, and is one of the first to state implicitly that the Alevi Kurds are just as Kurdish as the Sunnis.[1] (It has only been a part of the Alevis who allowed themselves to be convinced, and the Alevi identity is to this day a major rival to Kurdish ethnic identity.) 

The Turks and their virtual absence from Kurdish nationalist discourse 

It is perhaps surprising that the Turks have never been the ‘Other’ in opposition to whom the Kurds defined their own identities — at least not the Turks as an ethnic group. Throughout the 20th century, Kurdish nationalism has not placed Kurds against their Turkish neighbours (as it has at times opposed them to Armenians). It was state policies and state discourse, including the official view that in Turkey everyone is a Turk, against which Kurdish nationalism reacted in the first place. 

In Kurdish, there are two terms that can be translated as ‘Turkish’: Tirk is the ethnic Turkish villager or tribesman, Rom represents the state — Byzantine, Ottoman or Republican Turkish. Tirk could be a neighbour, and it is relatively rare to find instances where a sharp opposition of Tirk and Kurd is made; Rom was the quintessential alien, who implicitly defined Kurdishness by contrast. Well into the second half of the 20th century, Kurdish shepherds would shout to each other “Rom hat!”, “the Rom are coming”, when they saw Turkish soldiers approaching. ‘Romê reş’, ‘the black Rom’, has remained a common expression by which villagers refer to the state in its most oppressive aspects — black probably referring to the colour of late Ottoman military uniforms as well as carrying other negative symbolic associations. 

It was not ethnic Turks but the state in its nation-building phase against whom the Kurds defensively constructed their ethno-national identity. As long as the Kemalist movement was one of Muslims against non-Muslim enemies and the leadership spoke of Turkish-Kurdish brotherhood, it could mobilize many more Kurds for the common goal than the small circle of Kurdish nationalists could on behalf of separate Kurdish interests. The substitution of Turkish nationalism for Islam as the (would-be) uniting factor ended what had appeared to many Kurds as a natural alliance, and even this took some time to take effect. 

Leading circles in the Republic deliberately confounded civic and ethnic definitions of the Turkish nation: every citizen of Turkey, irrespective of his or her ethnic background, was declared to be a Turk, but soon all were told that they were in fact the descendants of Turkish tribesmen that had come from Central Asia. By the 1950s, the official claim that the Kurds, Laz and Circassians were also ethnically Turkish found belief among many citizens, including members of these groups. The new Kurdish movement that gradually emerged from the 1960s onwards had to dissociate itself from this definition of Turkishness and to that extent was anti-Turkish. In the 1960s and 1970s, Kurdish nationalists blamed the national oppression of the Kurds on the Turkish ruling classes, whom they declared colonial oppressors. These ‘Turkish ruling classes’ were in their view not necessarily ethnically Turkish but were the contemporary equivalent of the Ottoman ruling elite. Since the late 1980s I have frequently heard Kurdish nationalists explicitly exonerate the average ethnically Turkish citizen from accusations of national oppression. 

In the 1990s, when Turkey’s ethnic heterogeneity was more freely discussed and people had become aware that there were not just Turks and Kurds in the country but numerous others as well, some influential Kurdish personalities began spreading the notion that it was not (ethnic) Turks but especially the mühacir, the ‘Macedonians’ and the Circassians and Chechens, who were responsible for national oppression of the Kurds. (This view was especially popular in the dark days under Tansu Çiller and Doğan Güreş, who appeared to exemplify the domination of the state by non-Anatolian elements.) There was a subtle shift in Kurdish nationalist discourse (especially in circles close to the PKK) from the colonization of Kurdistan by Turkish ruling classes to an (as yet implicit) perception of colonization of the peoples of Anatolia by the Macedonian and Caucasian military-bureaucratic elites who had taken control of the state. From here it was in fact only a small step to the position adopted by the PKK (and quite a few other Kurds as well) in the wake of Öcalan’s arrest, that Kurds and Turks had common interests, that the Kurds had been a kurucu unsur of the Republic (alongside the Turks) and had a right to their legitimate share in it — as opposed to demanding the right to separate from it. 

The Kurds and Turkish nationalism 

The Kurds appear to have made a much greater contribution to the rise of Turkish nationalism in the past two decades than the other way around. The guerrilla war waged by the PKK was probably the single most important factor causing the electoral victory of the two most nationalist parties, MHP and DSP in 1998. Funerals of soldiers fallen in the struggle against the PKK offered the MHP excellent propaganda opportunities, which it used expertly. From the 1980s on, members of the ‘special teams’ engaged in non-conventional warfare in Eastern Turkey were recruited in ultra-nationalist circles; in the early 1990s, civilians of various stripes, including radical Islamists (Hizbullah) and ultra-nationalist (‘ülkücü’) Turks, were used for covered operations including assassinations. The rise of a virulent and sometime aggressive Turkish nationalism was, it seems, not a spontaneous reaction to Kurdish guerrilla violence but was to a large extent engineered. 

Inevitably there was a backlash: Kurds responded to the involvement of Turkish ultra-nationalists and the rise of Turkish nationalism with anti-Turkish feeling: hatred of the ultra-nationalists and also distrust of the intentions of even Turkish democrats. Although it is risky to generalize from my personal observations — serious sociological investigations that could refute or support them are lacking — I have the impression that in the 1980s and 1990s, Kurds who had previously had many non-Kurdish friends, colleagues and associates, tended to move in more limited, almost exclusively Kurdish circles, and many Turks gradually lost their Kurdish friends. In the left labour and student movements, this had already happened in the 1970s, when Kurdish activists left Turkish (i.e., non-ethnic) political organizations to establish their own Kurdish associations. This social separatism was accompanied (and probably reinforced) by the surprising return of Kurdish as a vehicle of communication. 

Several observers (including myself) have predicted that the rise of Turkish (ultra-) nationalism and the presence in western Turkey of large numbers of recent Kurdish migrants, living in compact and relatively closed communities and no longer attempting to hide their Kurdish identities, might easily lead to communal violence, especially if the PKK were to carry the struggle to the cities as it repeatedly threatened to do. It was, obviously, anti-Kurdish pogroms that I feared, not anti-Turkish pogroms carried out by Kurds. With relief I note that my predictions did not come true, and I should like to believe that they were unfounded. 

The Kurds have suffered much violent oppression in Republican Turkey, but by and large this violence was exercised by the state, in the name of its civilizing mission. Ethnocide, the effort to eliminate Kurdish ethnic identity, was a constant element in Turkey’s policies towards the Kurds from the late 1920s on, and on at least one occasion (Dersim 1937-38) these policies were followed through to the ultimate consequence of genocide. Although the view of world history as a permanent struggle between competing nations enjoys popularity in right-wing nationalist circles in Turkey, this violence cannot be understood as part of such a struggle between the Turkish and Kurdish ethnies; it was part of the modernizing project carried out by Turkey’s self-appointed Kemalist elite.



[1] Turkish nationalists of course had a similar line. At the 1995 Hacı Bektaş festival I heard Bülent Ecevit deliver a speech in which he appealed to Sunni-Alevi reconciliation in the name of a common Turkishness: “Şah Ismail Sultan Selim kadar Türk’tü”.