CHAPTER I.

THE VOWELS.

§ 51. The Greek has five simple vowels, represented by seven letters, and fourteen diphthongs. They are exhibited, according to their classes and orders, in the following table.

NOTE. The simple vowels and diphthongs are divided below into classes, according to the simple sound, which is their sole or leading element; as A sounds, &c. They are divided into orders, according to the length of this sound, or its combination with other sounds; as short vowels, &c. Diphthongs are termed proper, when the leading sound, or the prepositive (§ 31), is short; improper, when it is long (§ 54). The classes are arranged according to the openness of the vowel from which they are named.

$52. 1. Of the five simple vowels, three are open (§ 28), α, ɛ, and o; and of these a is more open than the other two. The open vowels never take the second place in a diphthong, and are, therefore, termed the prepositive vowels (§ 31). The two remaining vowels, v and, are close; and of these more close than v. As the second vowel in a diphthong is

is

always one of the close vowels, they are termed the subjunctive vowels (§ 31).

There is reason to believe, from both internal and external evidence, that the five simple vowels were pronounced by the ancient Greeks nearly as given in § 27. Without aiming at minute accuracy, therefore, we represent the ancient sounds of these vowels, as follows; of a by ah, of & by eh, of o by oh, of by oo, and of by ee. Whether the sound of was ever as open as our oo, is a mere matter of conjecture. If so, its reduction must have commenced early; for at the time when the Roman orthography of Greek names became fixed, it was too close to be represented by the Roman u (our oo), while, at the same time, it had not yet become reduced, as it has since been (§§ 15, 29), to the sound of the Roman i (our ee). In writing Greek words, therefore, the Romans retained the Greek T (in the Roman form, Y,) as a letter which had no representative in their own alphabet.

2. The long sounds of two of the vowels, & and 0, were much employed, and were, therefore, distinguished by separate characters ( and w) from their short sounds. The long sounds of the other vowels were comparatively rare, and, consequently, received no distinction of this kind.

When speaking of the letters, and not of their sounds (§ 49), we say that the Greek has seven vowels; and calls and the short vowels, because they always represent short sounds, n and the long vowels, because they always represent long sounds, and a, 1, and u, the doubtful vowels, because their form leaves it doubtful whether the sound is long or short.

$53. 3. An initial v, in the Attic and common dialects, always received the rough breathing, to assist in its utterance; as in English an initial long u is always preceded by the sound of y; thus is, iμμεis, as, in English, use (pronounced yuse),

union.

4. The long vowel is regarded as the short vowel doubled (§ 47); that is, ū = ăα, n = εε, ∞ = 00, ₺ = vũ, and i = ii. Whenever, therefore, in the formation of words, a short vowel is lengthened, or two short vowels of the same class are united in sound, the corresponding long vowel ought to result. But by reason of an early precession of the long open vowels (§ 30), ă, unless it follows &,, or e, is usually lengthened to ŋ, instead of ; and ɛɛ and oo commonly form, not the corresponding long vowels and w, but the diphthongs η & and ov, which are closer in sound.

Hence is termed the corresponding diphthong of, and v of .. See §§ 59, 68, and compare the earlier contraction βασιλῆς with the later βασιλεῖς ($ 69),

II.

REMARKS UPON THE DIPHTHONGS.

$54. 1. In Greek, diphthongs always begin with a more open, and end with a closer sound. As the latter must always be or v (§ 52), it follows, that only fourteen diphthongs are possible in the language, eight ending in, and six in v. Of these, wv occurs only in the Ionic dialect.

2. A short prepositive left time for the full utterance of the subjunctive vowel, and the diphthong was then termed proper, as really combining two sounds; but a long prepositive nearly or quite crowded out the sound of the subjunctive, and the diphthong was then termed improper, as though diphthongal only in appearance.

η,

3. After a, 1, and w, the subjunctive lost its sound entirely, and came, at length, to be written beneath the prepositive, if this was a small letter (§ 9); thus Aidns (a), or ads, is pronounced Hādēs; "Hidn, or on, ēdē; Lin, or ad, öde.

4. In diphthongs, except the three just mentioned («, ?, and @), the breathings and accents are written over the second vowel, and thus often mark the union of the two vowels; as αὐτή, herself, but ἀϋτή, cry; ηύδα, but ἤϋσε· αἵρεσις (α), but “Αιδης (α).

If two vowels which might form a diphthong are pronounced separately, the second is marked with a diaresis (§ 22); as άürŃ, Küos.

$55. 5. ROMAN LETTERS. In transferring Greek words. into Latin, the diphthong a becomes a; oi, œ; &, i or ē; and ov, u; as Puidogos, Phædrus; Bolwria, Baotia; Neikos, Nilus; Mýdɛia, Medea; Movoa, Musa.

A few words ending in aia and oa are excepted; as Maia, Maia, Teoía, Troia or Troja; so also Alas, Ajax.

The improper diphthongs,, w, are written in Latin, simply a, e, o; as "Adns, Hades, ideov, Odeum. But, in a few compounds of ồn, song, be comes a; as rgayædía, tragœdia, Eng. tragedy.

$56. 6. PRONUNCIATION. In the English method, the diphthongs are, for the most part, pronounced according to the prevailing sound of the same combinations in our own language; & like ei in height, or like oi in boil, vi like ui in quiet, av like au in aught, ev and nu like eu in Europe, neuter, ov and wu like ou in thou. The diphthong at is sounded like the affirmative ay (ah-ee, the two sounds uttered with a single im pulse of the voice), and vi like whi in while,

By the modern Greeks, the diphthongs are sounded as follows; a, like &; e. g. Kairag, as if Kirag, o, and, like; e. g. ixeïvos, λomós, viós, as if ixivos, duós, iós ov, like oo in moon; e. g. vous, pronounced noos; av, sv, nu, wu, before a vowel, a liquid, or a middle mute, like av, ev, eev, ōv, respectively; in all other cases, like af, ef, eef, of; e. g. 'Ayaún, süλoyos, nüda, pronounced Agáve, évlogos, eévtha; airós, ßariasús, pronounced aftós, vasiléfs.

$57. Written characters become fixed, though the sounds which they represent are fleeting. Hence the orthography of a language, when once established, is permanent, while its pronunciation is continually changing; and the two may thus become widely separated from each other, as appears remarkably in our own language. But in the first reduction of a language to writing, its orthography is intended to be an exact representation of its utterance. So scrupulous were the ancient Greeks in this respect, that they even varied their orthography, to express dialectic varieties of pronunciation. We may assume, therefore, that they spoke as they wrote, and that their diphthongs were real combinations of the sounds represented by the component letters. The original pronunciation of the diphthongs, then, may be rudely expressed as follows (§ 52); as, ah-ee, 81, eh-ee, 01, oh-ee, vi, oo-ee, av, ah-oo, ev, eh-oo, ov, oh-oo, &c. From these sounds there had been an evident precession at the time when the Latin orthography of Greek names became fixed (§ 55). For the precession which has since taken place, see § 56. But, yet further, from the laws of contraction (§§ 66 - 70), we learn that there had been a precession of the diphthongs, even before the Greek orthography itself had become fixed. We have, therefore, to guide us in investigating the precession of the Greek diphthongs, four data; 1. their original elements, as furnished to us by etymology; 2. the Greek orthography; 3. the Latin orthography of Greek words; and 4. the pronunciation of the modern Greeks.

$58. Let us attempt, for the illustration of this subject, to estimate numerically the precession of the several combinations of the short vowels. For this purpose, we will assume as a unit the vocal power, or the amount of voice, required in the utterance of the closest short vowel, viz. Y. If we represent this by 1, then the vocal power required in the utterance of may be represented by 2; of, or o, by 3; and, of a, by 4. For the long vowels, these numbers should of course be doubled, and we must assign to 7, 2, to v, 4, ton and to w, 6, and to a, 8. The vocal power of a diphthong would be represented by the sum of the two numbers assigned to its constituent sounds. Thus the vocal power of a, according to its original pronunciation, would be represented by 5 (4+1), of av by 6, &c. The consonant sounds, ƒ and v, to which, by an extreme precession, has been reduced in some of the diphthongs, are still closer than ; but, as our object is the illustration of a general principle, and not a minute computation of its effects, we will represent their vocal power by 1, instead of the more exact fraction. We will likewise confine ourselves, for the sake of greater simplicity, to the combinations of the short vowels, two by two. As there are five short vowels, these combinations will be twenty-five in number (5 × 5). But three of these, vo, 10, and iv, are never united in the same syllable, and ", being already at the extremity of the scale, is of course incapable of precession. There remain twenty-one combinations, which are given in the following table.

$59. The first column in the table exhibits the elements which are combined, with the numbers denoting their vocal power. The second column exhibits the forms which these elements usually assume in combination, preceded by the numbers which denote the vocal power of the combined forms according to their original pronunciation, as inferred (§ 57), and followed by the numbers which denote their vocal power according to the modern pronunciation (§ 56). The third column exhibits the whole reduction of vocal power which has taken place in each combination, and is obtained by subtracting, from the numbers denoting the vocal power of the elements, the numbers denoting the present vocal power of the combined forms.

The table exhibits a reduction in every combination except two, as and as, in which the open prepositive has prevailed over the closer subjunctive vowel. The result shows, that in the whole twenty-one combinations there has been a reduction of forty-three degrees, that is, of more than one third of the whole, or, upon an average, of more than two degrees in the vocal power of each combination. The actual reduction in the amount of voice required is probably even greater than is given in the table.