1 A reticular approach to kinship Douglas R. White and Michael Houseman For special issue of l’Homme 2013 Regardless of now marriage is locally defined, most people in most places marry (and/or have children with) people who are neither too close (immediate family members) nor overly distant (total strangers). As a result, many members of a given community are linked to each other by multiple direct or indirect ties of consanguinity and/or affinity. A reticular or network approach to kinship is concerned with the nature of these interconnections, they way they combine to form patterns, and how these patterns relate to other types of social phenomena: kinship terminology, residence, religious or political formations, economic transactions, and so forth. Marriage relations, however they are locally defined, are taken to play an essential role. While consanguineal ties link individuals into open-ended family trees, marriage acts not only to link these trees together, but also to relink the individuals in them to each other in a variety of additional ways. Such marriage-based connections give rise to closed circuits of kinship and marriage ties, thereby generating cohesion within the genealogical field. Members of different families can thus come to be appreciated as part of a single kinship network unfolding through time. From this point of view, a network approach to kinship builds on Lévi-Strauss’ (1967) (and Tylor’s [1889]) insight into the structural properties of affinal ties: consanguineous groups, and relatively isolated individuals, are integrated into higher order totalities through marriage. However, whereas Lévi-Strauss sought to grasp the relational properties of matrimonial alliance synchronically, by appealing to reductive “elementary” structures deriving mainly from classificatory schemes and normative precepts, a reticular approach proceeds directly from the analysis of actual marriage choices. Indeed, perhaps the most distinctive aspect of a network perspective is the emphasis it lays on the systematic treatment of empirical kinship and marriage data. A reticular approach is inherently practice- based. The development of kinship network analysis With its emphasis on actual (marriage) behaviour, the study of kinship networks may be said to derive, within anthropology, from a number of different sources. One overarching antecedent is the work of researchers affiliated with Max Gluckman’s “Manchester school” of anthropology who, seeking to distance themselves from normative models of social structure, undertook detailed analysis of interaction in particular situations in order to infer their underlying rules and assumptions. John Barnes (1954) introduced the notion of “social network” to describe and analyze systems of relations (kinship, friendship and neighbourhood in a Norwegian parish) which do not cluster into clearly delimited groups. His idea of treating interacting individuals and assemblies as a set of nodes some of which are joined by lines, was adopted and developed by others: Elizabeth Bott in her treatment of kinship and friendship in Britain (1957), S.F. Nadel in his theoretical treatise on “field theory” (1957), and especially Clyde Mitchell and his colleagues who used network analysis to visualize and understand the workings of personal, organizational and inter-ethnic networks in urban settings in Africa (Mitchell 1965). This promising approach disappeared from anthropology in the 1970s. However, the network paradigm was taken up elsewhere to become the foundation for major field of inquiry in the social sciences, increasingly dominated by quantitative concerns and resulting in a slew of innovative concepts and analytical procedures such as structural cohesion, centrality, balance, block modelling, etc. Insofar as kinship studies are concerned, foremost among the contributions to this more mathematically oriented