Kinship and mate choice in a historic eastern Blue Ridge community, Madison County, Virginia.

Potential mates analysis is difficult to apply to small historic populations that lack clear boundaries or regular vital event registration. Here I analyze the actual mate pool as an alternative way to identify causes of nonrandom mating when unmarried members are unknown. Factors influencing mate choice within a historic eastern Blue Ridge community in Madison County, Virginia, are examined for four marriage cohorts: 1850-1879, 1880-1899, 1900-1919, and 1920-1939. These factors include nuclear kin avoidance, preferred age differences between mates, and preferences for more distant kin. A simulation is used to recombine members of the cohort-specific pools of married individuals to generate the probabilities of various types of kin marriages. The pedigree and vital statistics data are derived from first-time marriage licenses filled by community members in Madison County from 1794 to 1939. The numbers of marriages examined for each cohort are 88, 120, 132, and 132, respectively; the mate pools constructed from the samples are viewed from the female perspective. The results generated by simulation on the actual mate pools consist of mean kinship coefficients, numbers of marriages between "allowed" kin types, and probabilities of these values when marriage is random with respect to kinship. The results indicate significantly high levels of inbreeding in all four marriage cohorts, primarily because of high levels of first-cousin marriages in the first three cohorts and of first-cousin once-removed marriages in the 1920 cohort. The observed mating patterns are discussed in terms of the social history of the Blue Ridge community and restrictions of the data.