U
U holds the twenty-first place in our alphabet. The corresponding place in the Greek alphabets was occupied by Y (with some slight variations of form). The form in the Italian alphabets was generally V. These three are only modifications of one original; but they are independent symbols with us, though Y does not represent any sound otherwise unrepresented. It will be most convenient to describe the three forms once for all.
With T we reach the end of the original Phœnician alphabet. The remaining symbols—no fewer than six with us, four in the completed Latin alphabet—are accretions, either modifications of old symbols for greater exactness or old symbols themselves which had fallen out of their proper place and were added again. The first new symbol was needed to represent the important vowel sound u. We have already seen that the Greeks employed the Phœnician symbols for the breaths which they did not want as symbols for the vowels which they did want. Thus we should have expected that the Phœnician vau would have been used for u. But vau was already employed for w, which was a living sound in early Greek; the form used was ꓝ (the so-called digamma), the origin of our F. What then was the origin of the symbol for u? In the earliest Greek we find the two forms Y and V, e.g., in inscriptions of Thera. Now the Moabite form of vau is Y, which resembles the Y more than F. It is difficult to see why the presumably oldest Phœnician form should appear, not in the sixth place of the Greek alphabet, but at the end, where it must have been an arbitrary addition; and, although the Y form could be derived from F (middle steps are found), it is not easy to get F from Y. We may suppose that the two symbols, F and Y, were obtained by the Greeks from independent Phœnician alphabets, the first being kept by those Greeks who required a symbol for w, and did not at first need any special sign to distinguish u (which in the earliest Greek times known to us had the value of German u) from o; while the others took the form Y to express the modified u, and probably never really adopted the F, except as a numeral; it does not appear even in the very old Abu Simbel inscription written by the Ionian mercenaries of Psammitichus. It is noteworthy that in that inscription both the forms Y and V appear, whereas in those of Thera and Melos we have sometimes the one sometimes the other, but not both in the same inscription, and a study of the writing leads to the conclusion that the Y was felt to be the true (i.e., the older) form, but that V was used more frequently for the sake of simplicity.
At Rome the single form V denoted both the vowel u and also the consonantal w. F retained its place as sixth in the alphabet, but with the value of f which was unknown to the Greeks; a peculiar form, C, in which the middle stroke has gone to the bottom, seems to have been affected by its neighbour E; this is found in Etruscan, Umbrian, and Samnite inscriptions; it has, however, the value of w; while a curious symbol 8 appears at the end of the Etruscan alphabet, and is also used in the Eugubine tables, with the value of f; the origin of this is uncertain. It may be a rounded form of the second symbol in the digraph FB (i.e., FH) by which the sound F is indicated in a very old inscription (see Rhein. Mus., xlii. 317); if this is so, the Latin alphabet has the first member of the digraph, the Etruscan has the second. Next, the symbol Y was added (together with Z) in the 1st century B.C. to represent more exactly, in borrowed words, the sound of Greek upsilon.
Lastly, the form U was differentiated from V. It is the uncial form, and so belongs to the general transition from the pointed to the rounded character which conduced to greater convenience of writing. Examples of it may be seen in the article on Palæography; see the specimen of Latin uncial of the 5th or 6th century (vol. xviii. p. 153), and the half-uncial of the Lindisfarne Gospels, about 700 A.D. (ibid., p. 159). It was clearly a matter of convenience to have separate symbols to represent sounds so distinct as u and v; but the application of the two symbols seems to have been due to chance rather than design. The form V remained in use at the beginning of words, whereas u, which was the uncial and cursive form, naturally was used rather in the middle; by degrees the initial form was appropriated to consonants, perhaps, as Dr Taylor suggests (Alphabet, ii. 189), because the consonant is commoner at the beginning of words, or for some other reason, while the medial form was connected with vowels.
The sound which U denotes is produced by "rounding" the lips to the furthest extent consistent with a clear vowel-sound, and by raising the back of the tongue higher than for any other rounded sound. It has two varieties (like all other vowels) according as the position of the tongue is more or less tense, producing thereby a narrower or a wider aperture for the voice to pass through; whence the sounds are technically called "narrow" and "wide" respectively. The narrow sound is heard in English only when the vowel is long, as in "book," "rule," but in northern English (Scotch) "book" may be heard short. The wide sound is heard in "full," "good." The digraph oo is commonly used for the u sound, and attests the fact that the original sound of ō has frequently passed into ū, as in "good," "food," &c., written "gode," "fode" in Middle English; sometimes, however, the oo has come by analogy into words where ū is the original sound, as in "room," M.E. "roum," O.E. "rūm." Original ū has commonly passed into the au sound, spelt in English ou or ow, as in "how," "house," "mouse," "bower," for O.E. "hū," "hūs," "mūs," "būr." According to Mr A. J. Ellis, words derived from the French had in Chaucer's time the sound of French u; and Sir John Cheke's statement "cum duke, tube, lute, rebuke, δυκ, τυκ, λυτ, ρεβυκ dicimus, Græcum υ sonaremus," seems strong for the same practice in the 16th century. In the 17th century the modern pronunciation of u as iu in "muse," "duke," "mute," "pure" had come in. Hence also we may explain the substitution of u for y in some genuine English words, as "busy " (orig. bysig). At the same time begins the corruption of u to the (so-called) ŭ sound in "but," "shut," &c.; this is not a u sound at all, but the neutral vowel as heard indifferently in "but," "sun," "son," "blood"; it is often confounded by writers with the true u heard in "pull" and in the northern pronunciation of "but," "shut." For the history of the German "modified" u (spelt ü, but originally ue) see under Y.
