Corpus statistics show a great variation in derivation of the names of state inhabitants. Rarely a found name can be regarded as ungrammatical and false.