From Tetradic Society to Dispersed Alliance: Notes Arising from a Chapter by N.j. Allen

I N.J. Allen has recently investigated the possibility of modelling transformations from tetradic society – which he devised as a model of prehistoric human kinship and is intimately linked to his scholarship – to what are conventionally known as Crow-Omaha systems, themselves the subject of a recent collection reappraising the problems associated with them (Trautmann and Whiteley 2012). Allen’s chapter appears in this volume and is immediately followed by a typically sceptical chapter by R.H. Barnes (2012), which should remind us of the controversy surrounding Crow-Omaha not only over what this hyphenated category of analysis actually means, but also over whether it actually means anything at all, even for the Crow and Omaha themselves (two Native American peoples of the USA). However, I am not concerned with these controversies here, which broadly speaking revolve around two more general concepts: kinship terminology and systems of affinal alliance. These two concepts can, of course, be linked in certain circumstances, and included routinely in the latter are the prohibitions on marriage into the social groups, however defined (clans, lineages, even families), of certain relatives like ego’s four grandparents. It is these prohibitions that are my concern here, coupled with the consequence that they ban repeating affinal alliances between such social groups for a number of generations. This results effectively in a dispersal of alliances that may also happen with cross-cousin marriage, though the formal model thereof does not show this or indeed require it. The temporary ban on re-alliances between alliance groups in subsequent generations may actually be expressed indigenously as such, as by many Munda, a linguistically related group of tribes and low castes in east peninsular India who were the subject of part of my doctorate (1984, but see also 1992: Chapters 7, 8). However, there is nothing formally ‘Crow-Omaha’ about the Munda, since their terminologies are entirely lacking in the characteristic vertical equations.