Small-scale convergence in the Hindu Kush mountain area

The Hindu Kush mountain area, situated at the crossroads of South and Central Asia, displays high language density as well as a high level of linguistic diversity (including Indo-Aryan, Iranian, Nuristani, Sino-Tibetan, Turkic and Burushaski). It has been described as a distinct linguistic area (Toporov 1970; Èdel’man 1980; Bashir 2016), alternatively as transitional between two larger linguistic areas (Tikkanen 2008), or as an accretion zone (Nichols 1992: 21). The present study attempts at shedding light on the finer details of the region’s linguistic geography. Primary data was gathered from a tight sample of 59 language varieties. The data set was coded for 80 structural properties, covering morphosyntax, lexico-semantics and phonology, and subsequently analysed for their phylogenetic and geographical distributions. Two observations were particularly significant: First, only a small number of properties are shared by most or a substantial part of the languages in the region. This poses problems for viewing it as a distinct linguistic area vis-à-vis surrounding areas, although certain properties cluster in the central parts of the region. Second, a pattern emerges of a handful of distinct micro-areas within the Hindu Kush, each characterized by substantial feature sharing. Some of those micro-areas are phylogenetically heterogeneous (e.g. an Eastern Hindu Kush micro-area involving the isolate Burushaski and a number of Indo-Aryan languages), while others are phylogenetically rather more homogeneous but diverge typologically from related languages outside the region. Preliminary analysis suggests that such micro-areas are the product of multiple, small-scale relationships between local communities (Koptjevskaja-Tamm & Wälchli 2001: 728), rather than being propelled by a major lingua franca or by any substantial level of structural bilingualism. The Hindu Kush micro-areas overlap surprisingly well with areas of comparable size, previously identified in anthropological-historical research as “geo-cultural regions” (Cacopardo & Cacopardo 2001: 17). The latter are characterized by a shared social, cultural and religious identity, often in contrast to surrounding lowland cultures. This correlation between linguistic feature sharing and a common cultural and geophysical context is especially evident in a Western Hindu Kush micro-area (Liljegren & Svärd 2017). This part of the Hindu Kush withstood for the longest time conversion to Islam and held on to a distinctly pre-Islamic world view and cultural practices, well into the 18th century, in some cases longer (Cacopardo & Cacopardo 2001: 25, 249–250; Klimburg 2008; Jettmar 1975: 394; Jones 1974: 168). The centre of diffusion appears to be Nuristan (situated in the remotest and most rugged part of northeastern Afghanistan) and the Nuristani languages, but in various ways involving adjacent language communities across phylogenetic boundaries. A salient feature is the use of fine-grained geomorphic spatial coding that takes features of the “alpine” environment as its primary frame of reference (Heegård & Liljegren 2018). Such a system is particularly complex and specialized in Nuristani Prasun with its 39 directional affixes that indicate location and direction of an entity in relation to the flow of the river, the degree of deviation from the horizontal plane, and relative distance (Buddruss & Degener 2017: 449–482)