Document Type



To the extent that persons having the same surname can he assumed to he descended from the same progenitor through the male line (in patrilineal societies), and assuming that relationships through female lines and mixed lines are proportional to those through the male line, then the frequency of shared surnames between two communities samples the genetic lineages and measures the degree of biological kinship between the communities. Specifically the number of possible pairs of individuals with identical surnames in two samples, divided by the number of all possible pairs would be just twice the coefficient of relationship (the proportion of genes identical by descent). I use the term coefficient of relationship by isonymy (Ri) for the average coefficient of relationship estimated in this way. Ri = Z(Nsl • Ns2)/2N ,N2 where Nsl equals the number of instances of surname s in the first sample and NS2 the incidence of the same surname in the second sample, and N, = SsJMsl and N2 = 2<,Ns2 and are the respective sample sizes. Random samples of 178 to 632 surnames of individuals in each of 5 communities within 51 km of each other on the north coast of Peru give the 10 relationship coefficients between communities of.0010 to.0024 compared with the five relationship coefficients between the sexes within the communities of.0018 to.0200