J Biosoc Sci. 1992 Jan;24(1):113-21.

The influence of religious affiliation on surname repetition in marriages in Tiszaszalka, Hungary.

Koertvelyessy T, Crawford MH, Pap M, Szilagyi K.

Department of Sociology and Anthropology, Ohio University, Athens.

The repeated-pairs of surnames in marriages (RP) approach is applied to the population of Tiszaszalka in north-eastern Hungary. The results indicate that: (1) lineage-like behaviour in mate choice results in population subdivision in both the Catholics and the Protestants of the village; (2) unlike in some other Tiszahat villages, the isonymous and the repeating unions in Tiszaszalka occur in different lineages so, in neither of these subpopulations are isonymous and repeating unions monopolised by a few lineages; (3) religious affilitation influences the mating structure of the population as measured by RP summary scores.

PIP: Matrix methods have been devised to overcome the limitation of the isonomy method with high sampling errors. The repeated pairs (RP) of surnames in marriages approach proposed by Lasker and Kaplan in 1985 measures the effect of lineage in mate choice. Any excess of RP over the random occurrence (RP1) indicates lineage involvement via surnames. The degree of population subdivision from lineage influence on mate choice was estimated, lineage influence on repeating matings and isonomy was analyzed, and the effect of religion on mate choice in repeating and isonymous unions was explored. Marriage records were obtained in 1986 in Tiszaszalka and from the Genealogical Society in Salt Lake City, Utah (from 1806 for Protestants and from 1936 for Catholics). 1271 first marriages were analyzed: 972 Protestants and 299 Catholics. The RP and RP1 were larger among the Protestants than among the Catholics. The chi-square value for unique and nonunique surname pairs for the 2 religions was statistically significant: for total Protestant vs. incomplete Catholic and for partial Protestant vs. incomplete Catholic. The distribution of Isonymous and nonisonymous marriages of both religions was not significant indicating no inbreeding. After excluding isonymous matings the chi-square values for unique and nonunique surname pairs remained significant for both religious groups. The coefficient of isonomy indicated that both groups avoided isonymous unions since 1936, thus inbreeding was insignificant. In both groups lineage like behavior in mate choice resulted in population subdivision; religion influenced mating structure significantly and genetic variability decreased as a result. The incidence of isonomy was also low in several other villages in the region, and the RP approaches focus on husband's and wife's surnames was valuable regarding marriage patterns and genetic variability.