The incidence of occurrence of 97 selected surnames from place names among the listings in the white pages of English telephone directories were recorded. Among the 13,109 listings of these surnames 328 (compared with 172.4 expected at random) occur in the exchange or group of exchanges with the place name that corresponds with the surname, that is, the place from which at least some instances of the surname putatively derived. Furthermore, for the three largest directories (London, Greater Manchester and Birmingham) the log of observed/expected incidence and the log of distance of the places from which each of the names derived are negatively correlated: r = —.412 (p < .01), r = —.265 (p > .05), and r = —.412 (p < .01), respectively. Although this set of surnames includes some only partially originating from the selected places, the distribution shows a tendency to some association even after the whole span of time since the surnames were established in the Middle Ages. The pattern contrasts with what would be inferred from studies of marital migration and with most studies of polymorphisms in the association and in the very splotchy and irregular distribution with barely distinguishable dines.
Kaplan, Bernice A. and Lasker, Gabriel W.
"The Present Distribution of Some English Surnames Derived from Place Names,"
2, Article 8.
Available at: https://digitalcommons.wayne.edu/humbiol/vol55/iss2/8