In many societies and communities, the choice of names for children was determined by local or cultural custom. Through the study of naming patterns, a genealogist may establish parentage for a child, or even hypothesize the existence of unknown children.
This is especially true of societies where naming patterns were followed consistently and with few exceptions. Researchers of the early Dutch families of New York, for example, have been aided by the stringent naming pattern followed by Dutch parents. First and second sons were usually named for paternal and maternal grandfathers, and daughters for their grandmothers. With the births of succeeding children, the names of great-grandparents would often be used, though variations may be expected to arise. Newborns in most cases would be named for a close relative or friend of the family.
Even in families or communities which did not follow stringent patterns, other naming practices may be discovered—such as using the mother's surname before marriage as a child's given or middle name, or repeating one or more given names each generation—which may prove useful to the genealogist.
Use of a naming pattern should not be assumed except in a community where several instances of the pattern have been, or may be, documented.