114 SOCIAL SCIENTIST As must be clear from the above song, work songs are an index not merely of economic activity but of the social situation. It is therefore possible to argue that work songs especially those which are particular to women can provide the best perception of women within patriarchal (or for that matter matriarchal) systems. Secondly work songs by women pertain to an identifiable strata of society and a clearly distinguishable economic sector. More than two thirds of these work songs originate in the rural sector and pertain to women involved in agricultural activities. Besides these there are stray songs which refer to the cotton picking, spinning or basket— weaving activities of women. Secondly, the majority-of these work songs relate to women of the lower strata of society although one does occasionally come across songs by upper class rural women engaged in domestic chores in what is now classified as 'unpaid labour' in feminist economics. In terms of the class structure identifiable in the majority of these work songs as clearly lower class, this genre becomes crucial to the study of women's history. This is because what is normally available to us for the reconstruction of the social history of women is high literature by and large authored by men, especially of the brahminical upper caste. In the evolution of the brahminical patriarchal order, women have always been looked upon as 'the other'. These texts on which our understanding and reconstruction of our past is based either marginalise women and the lower classes of society or integrate them into an overarching patriarchal structure. Therefore in most of what is classified as high literature, women lacked access to the means of creating, disseminating or preserving their history. In contrast these work songs provide us with a grass roots perception of women's place in historical societies as perceived by these women themselves. Certain fundamental problems arise in the use of folk traditions, in the present context folk songs, for historical re-construction. Jan Vansina (Vansina: 1961) in discussing the historical methodology— involved in the use of oral traditions classifies folk songs as 'second-hand' traditions since they are narrative in form and go through mutations and interpolations in being handed down through generations. Folk songs are therefore categorized as 'soft evidence' in contract to archaeological or inscriptional evidence which can be located fairly accurately in space and time and therefore constitute 'hard evidence'. Nevertheless, anthropologists and historians are increasingly making use of such sources to construct a social history of such neglected social categories as women, craftspersons and peasants, with women of course also present in the latter two categories. E.P. Thompson, in a thought provoking paper on 'Folklore, Anthropology and Social History', talks about the neglect of women's history by historians due to their pre-occupation with 'becoming', a proce^ in