Page:Language and the Study of Language.djvu/337

VIII.] further that the language of this race shows it to have been Scythian, a member of the westernmost, or Finno-Hungarian, branch of the family. By others the Scythian character of the dialect is explicitly denied. The discussion is at present in the hands of too few persons, and those too little versed in Scythian philology, to admit of a definite and satisfactory conclusion; and meanwhile we are justified in regarding with extreme incredulity any theory which puts Scythian races in the position of originators of an independent civilization, and teachers of Semites and Indo-Europeans. Such a position is wholly inconsistent with what is known of their history elsewhere, and would constitute a real anomaly in ethnology; while we are not authorized utterly to deny its possibility, we certainly have the right to demand full and unequivocal evidence before we yield it our belief. The fact—if fact it be—is of a revolutionary character, and must fight its way to acknowledgment.

The linguistic tie, now, which binds together the widely scattered branches of this great family, is a somewhat loose and feeble one, consisting less in the traceable correspondence of material and forms, the possession of the same roots and the same inflections, than in a correspondence of the style of structure, of the modes of apprehension and expression of grammatical relations. Each great branch forms by itself a group as distinct as is, for instance, the Germanic or the Slavonic in our own family; but there is no such palpable and unmistakable evidence of kinship between Ugrian, Turkish, Mongol, and Manchu, as between German, Russian, Greek, and Sanskrit. It is, to no small extent, those who know least in detail respecting the languages of the family who are most ready to assert and defend their historical connection: and, on the other hand, Castrén, himself a Finn, and whose long and devoted labours have taught us more respecting them than has been brought to light by any other man, ventures to assert with confidence only the demonstrable linguistic relationship of Ugrian, Samoyed, and Turkish, and regards the inclusion of Mongol and