Page:Language and the Study of Language.djvu/488

466 of the k-sound before u did not differ enough from its pronunciation before a and o to call for an independent notation. Of the remaining three Latin letters, the X is a Greek invention (used in some Greek alphabets also with its Latin value, or representing xi, instead of chi), and, as standing for the double sound ks, not less needless than Q; Y and Z are later importations out of the Greek alphabet, and used only in Greek words, to signify peculiar Greek sounds (the Greek upsilon having by this time changed its value of u for that of the French u, German ü).

The changes which we, in our turn, have introduced into the Latin alphabet, in adapting it to our purposes, are not insignificant, although far from being enough to make it represent our spoken language as fully and consistently as it formerly did that of the Romans. Besides the eighteen articulations of the early Romans, we have (as was shown above, in the third lecture  ) at least fourteen others which call more or less imperatively for separate designation. There are the a of cat and care, the a of what and all, and the u of cut and curl; there are the two semi-vowel sounds, y and w, the palatal nasal (which we commonly write with ng, as in singing), the three sibilants, z, sh, and zh (the z of azure), the two sounds of th, in thin and thine, and the v of valve; and, finally, the compound consonants ch (in church) and j (in judge). Some of these needs we have managed to provide for: we have turned the two forms of the Latin i, I and J, into two separate letters, with very different values; we have done the same thing with the two forms of u, V and U, converting the former into a sign for the sonant labial spirant; by doubling the same character, we have made one wholly new letter, w, for the labial semi-vowel; and we have utilized y and z, as semi-vowel and sonant sibilant. We have also brought k back into its old place—yet without perceptible gain, since its introduction makes c superfluous; k, c, and s having but two sounds to designate among them. The new characters which the Anglo-Saxons had devised for expressing the two th-sounds we have unfortunately suffered to go out of use again. And q and x are still as useless to us as they were of old to the Romans. Hence, we have virtually only