In the three previous chapters on India, I have discussed marriage and the transmission of property to women in one locality, then looked at the distribution of this and other family variables in the hierarchy and finally examined their regional spread in North and South. Before finally reviewing the status of theoretical approaches to these differences, I want to consider an attempt to discuss differences of a more specific kind that gives prominence (but not exclusively so) to modes of production. At the same time, I want to provide a link back to Tibet and forward to West Asia. To do this I turn to Gough's analysis of kinship in two areas of South India, Thanjavur (1981) and Kerala (1961), which is especially valuable as she looks at a region of the country on a comparative and historical basis in a way that emphasises the wider interdependence of castes in relation to state organisations embracing both town and country, subject and ruler.

There are broadly two types of possible explanations of differences: the historical and the sociological. I use these terms advisedly, since they are neither alternatives nor yet completely distinct; any enquiry needs to take into account both the past and the present. But, quite apart from the selection of the significant factors, there is the question of how far into the past one should search. Moreover, when we isolate historical factors as potential elements in an explanation, we still have to show their relevance to the ethnographic present, to the analytic moment.