By Antti Leino
Nordiske navnes centralitet og regionalitet (2007)
Abstract: The Finnish basic map, and the database used by the National Land Survey to produce it, contains over 300 000 dicerent toponyms and over double that amount of named places. It is impossible to study the distributions of all these toponyms at once, and thus a large-scale onomastic analysis requires an overview of the material.
Modern computational methods can extract the most significant directions from the overall variation into a few components, and these components can subsequently be used as a basis for onomastic analysis. A brief analysis of the names of two types of natural features results in components that have interesting parallels in the history of Finnish settlement and in language contacts.