What a Place Name Can and Cannot Prove

A settlement whose name means the clearing by the oaks is evidence of something, and the argument is about what. The name records that speakers of a particular language once described the place that way. It does not record when, how many of them there were, or what happened to whoever was there before.

The last point is where the reasoning most often fails. A region may fill with names from an incoming language without a large population arriving, because names spread with administration, with the written record, and with whoever is in a position to draw a map. The reverse error is equally common: a name surviving from an older language is sometimes taken as proof that its speakers remained, when a name is precisely the thing a newcomer keeps, having no reason to rename a river everybody can already find.

This is why names are strongest as evidence when they are used negatively. A language that has left no names at all in a region is hard to argue was ever administratively dominant there, and that single negative conclusion is more secure than anything the same body of names can establish in the other direction. Positive claims about numbers or replacement need what the names cannot supply. The safest procedure is to treat a body of names as a question for other evidence to answer.