IntroductionThis paper focuses on internal migration and intermarriage patterns between Kurds and non-Kurds who moved between regions of Turkey in the last four decades of the 20th century. We aim to assess the spatial and consanguinal boundaries between Kurds and Turks in recent history. Kalmijn (1998) argued that intermarriage is an indication of interaction across group boundaries in that members of different groups accept each other as equals while more recent studies also claim that intermarriage is a sign of boundary blurring (Black, 2013; Horowitz, 2015). Studies as early as Drachsler (1921) or Gordon (1964) have viewed intermarriage as an indication of assimilation while Smits (2010) in his recent review argues that relatively low levels of intermarriage might be a warning sign of violent conflict and/or future separation. Conversely, relatively higher levels of intermarriage might indicate the continuation of a peaceful coexistence or a non-violent secession.Due to an evident language shift and widespread bilingualism among Kurds in Turkey (Polat & Schallert, 2013; O'Driscoll, 2014; Zeyneloglu et al., 2015; Zeyneloglu et al., 2014) we have introduced, for the first time, birth-region (re: territorial association) as a measure of ethnic origin instead of language as overt questions on ethnic identity or ethnic origin are absent in Turkish censuses and surveys. Our hypothesis is that in a universalistic setting intermarriage of both the minority (Kurds) as well as the majority group (Turks) to each other will correlate with education in line with modernization theory. We refute earlier claims of Gunduz-Hosgor and Smits (2002) who had concluded that the Turkish-Kurdish intermarriage pattern rather fits exchange theory since neither the dataset (TDHS-Turkish Demography and Health Survey) nor the ethnicity marker (mother tongue) used in that study are suitable for the analysis of intermarriage as is elaborated in the following sections.The paper is organised as follows: We first introduce the concept of universalism (as opposed to particularism or differentialism) and discuss why and how Kurdish individuals may switch to Turkish identity. We, then discuss the shortcomings of using language as a proxy for ethnicity in the Turkish context drawing on an analysis of the 2003 TDHS data demonstrating language shift among the Kurds. Data and methods are described in the fourth section before general trends of homogamy and in-marriage at the national level are examined. Following the conceptual framework and the context, we discuss the relationships between inter-regional migration, education and intermarriage among Turkey's Kurds and Turks. We conclude with recommendations.1. Birth region versus ethnic identity in a universalistic environmentIn his seminal work on the ethnic origins of nations, Smith (1986) observed that some ethnic groups were able to vastly extend in size and territory by incorporating other groups though rather with fluid boundaries, while some other ethnie have established sharper cultural markers and physical boundaries between them and the outside world, albeit in much smaller or even peripheral geographies. He labelled this dichotomy as 'lateral' versus 'vertical' (Smith, 1986, pp. 76-79). While his observations on the lateral versus vertical dichotomy seem precise, his arguments about its reasons remain case-based, are not exhaustive, and contain transitions of the same ethnie between the two main categories, thus straining the initial theory (1986, pp. 82-98, 105-119).Todd (1985), on the other hand, suggested a bijective link between the anthropological family type prevalent in a region or a nation and its political structure. The concept of equality in Todd's work is similar to the lateral/vertical dichotomy of Smith. Todd (1985, pp. 7, 28, 55-65) argues that groups where parents treat all brothers as equal and the inheritance is shared equally between the brothers as a sign of this equality have a tendency to assimilate subjugated ethnie. …