Page:Language and the Study of Language.djvu/119

III.] by nothing but individual caprice, so in the traditional transmission of language—which is but the same process of teaching children to Speak, carried out upon a larger scale—we must look for similar cases of arbitrary phonetic transitions.

So important a part of the history of a language are its special methods of phonetic change, that, in investigating the relations of any dialect with its kindred dialects, the first step is to determine to what sounds in the latter its own sounds regularly correspond. Thus, on comparing English and German, we find that a d in the former usually agrees, not with a d, but with a t, in the latter; as is shown by dance and tanz, day and tag, deep and tief, drink and trink, and so on. In like manner, the German counterpart of an English t is s or z: compare foot and fuss, tin and zinn, to and zu, two and zwei, and the like; and a German d answers to our th, as in die for the, dein for thine, bad for bath. What is yet more extraordinary is the fact that, if we compare English with the older languages of our family—as with Latin, Greek, and Sanskrit—we discover the precise converse of this relation: as German t is English d, so English t is Latin d (compare two and duo); as German d is English th, so English d is Greek th (compare door and thura, daughter and thugatr); as German s or z is English t, so English th (the lisped letter instead of the hissed, the spirant for the sibilant) is Latin, Greek, and Sanskrit t (compare three and tres, treis, tri; that and -tud, to, tad). In short, taking the series of three dental mutes, surd, aspirate, and sonant, t, th, and d, we find that the Germanic languages in general, including the English, have pushed each of them forward one step, while the High-German dialects, chiefly represented by the literary German, have pushed each of them forward two steps. Thus, in tabular form: