the Khasi-Karbi-Tiwa triangle
Thursday 25 September 2008, by
The border between Ribhoi district of Meghalaya and Karbi Anglong district of Assam is characterized by the coexistence of different languages (Khasi, Karbi, Tiwa, Nepali), different cultural and social features and different ethnicities. This is why in our attempt to study the interactions between these three levels we chose this area as one of our main cases. It is a border region, not only because different administrative units intersect here but also because several cultural zones converge at this point. Thus, it is an ideal place for studying how people mix, or don’t mix, intermarry, exchange tools, techniques, customs, myths... Our primary aim is to look beyond the generally accepted association between one language, one ethnicity and one fixed set of cultural features. For example, Khasi speakers are traditionally supposed to bear Khasi clan names and to all be matrilineal; on the contrary, Karbi speakers are said to bear Karbi clan names and to all be patrilineal. To what extent are these assumptions correct, where exactly do they prove to be inaccurate and what are the characteristics of such places?
Here, we have chosen to analyse the distribution of a particular social feature: descent, i.e. the rule defining how an individual belongs to a group claiming a common ancestry. As a preliminary step before field investigations, we used electoral data to get a broad idea of the general distribution of descent types. For each voter, the electoral rolls give the name of his father, mother or husband. When the last name of the father is different, we may assume that the person took the name of his mother, according to a matrilineal principle. Generally speaking, the same naming principle is followed at household level, but it is not always the case, and this is a discovery in itself: we have found dozens of cases where in the same household individuals took either the name of their father or the name of their mother. This particular point needs to be investigated on site for a proper understanding of the reasons for this: is this the sign of a change in social practices, is it only the result of isolated individual choices or is it due to the particular context of the electoral census?
The map given here displays the proportion of matrilineal relationships for each village in Mawhati Constituency (local assembly). Of a total of 114 villages, altogether comprising 21,300 voters, some 21,000 relationships were recorded.
As this area is very largely matrilineal, the statistical classes were defined so that the occurrences of patrilineality remain visible.
Please have a look at the resulting map.
Our first findings and preliminary interpretations as inferred from this map are as follows:
certain continuities clearly appear in the distribution of descent types.
matrilineality is more prevalent in the south than in the north. This may correspond to the Khasi-speaking area, but this will have to be checked.
a patrilineal belt extending from the northeast to the southwest is also visible. Here again, our first guess is that it corresponds to the Karbi-speaking area in the foothills above Guwahati and Dimoria.
the northeastern zone around the Umiam-Umsiang river confluence somehow forms a continuity as most villages are either mixed or moderately matrilineal. This zone is where the Tiwa language is spoken together with Karbi and Khasi.
any discontinuities, however, will have to be scrutinized. What do "moderate matrilineal" and "mixed" villages stand for? What concrete social reality do they correspond to? Are they ethnically/linguistically/culturally "mixed" villages? And if this is the case, to what extent do the different components interact in terms of descent? And if these villages are culturally homogenous, how can we explain that their inhabitants simultaneously follow different descent rules?
The same type of work has been completed for the Hill Tiwa area around Umswai and is now being done for the northern Jaintia Hills. Our next step is to compare the distribution of clan names with the distribution of descent types.