Imágenes de página

they are proved from the grammarians and by euphonic changes to be the sonants of P, T, and K, respectively.*

When the living languages are not unanimous, their testimony is not invalidated, if the want of agreement can be referred to well-established laws of change. One law of change, for example, is the substitution of so-called aspirates for smooth and middle mutes, in other words, of continuous for explosive sounds. The three gates of the breath, labial, lingual, and palatal, are left ajar, instead of being perfectly closed. Sonants suffer this change more frequently than surds. Thus the modern Greek B and D, and the Spanish medial B and D, have become V (nearly), and Th sonant. The modern Greek G before e and i, as well as the English, French, Italian, and Spanish C and G before e and i, have


[ocr errors]

* The distinction between sonants and surds is second in importance to none in phonology Though familiar to scholars, it is often overlooked or misunderstood. Every letter is a sonant or a surd. Every sonant has its surd, and every surd its corresponding sonant. Both are given with the orgins of speech in precisely the same position. The Welsh l in Llewellyn, commonly regarded so difficult, is only the surd of the English l in loud. The Welsh rh is only the surd of the English r. The German ch in mich is the surd of the English consonant y. A deep Arabic guttural is the sonant of the German ch in Buch. The distinction is as clear and simple as it is important : sonant, sounding; surd, soundless, except as mere articnlate brenth can be enrd. Exump'es will make this clearer than any definition. Z and s are pronounced with the organs of speech in the same position, but z is articulate sound, s is articulare breath ; z is the sonant of 8, and s is the surd of 2 The Greek Digamma (English w in woo) has its surd in woh in when. The vowels of course are all sonants, and have their surils in the various surd aspirates representid by the letter h. The surd of a, as in father, is the h which precedes that sound of a, viz., h in harp. The surd of į, as in machine, is the n hard in heul, and so of the rest. The difference between h in home and h in heal is as marked, if not as obvious, as between o and ee

We give the consonants in order.


Rh (Welsh) Z с






G (soft) Ch (in chant)

W (digamma) Wh (in when)
Ch (in chant)


X G (hard) K

Y (consonant) Ch(in German mich) L Ll (Welsh)


s M H, with closed lips, in Z (in azure)

Sh hm (in contempt) Th (in this) Th (in thin) N

H in hn (in contempt) An Arabic guttural Ch (German G (hard) Q

in Buch)




become continuous sounds in the place of the old explosives K and G hard. The medial consonant is often altogether dropped in French and English, as parabole, paravole, parole ; regen, reyen, rain. The general testimony, then, of living languages as to B, D, and G is unimpaired by any merely apparent counter-evidence.

Of the Greek vowels, four are determined by living languages. The fifth, U, was originally u in rude (rood), afterwards the French u. U has passed through this change in France, Germany, and Greece.

The Greek diphthongal sounds can not be learned from

modern usage.

Chi does not appear in the alphabets of Western Europe. The modern Greek and Sclavonic languages give it the sound of the German and Scotch ch, as in loch.

Phi was near the Roman and modern F, for the Greeks spelled words like Fabius with an initial Phi. It was not identical with it, for the Romans never spelled a Greek word like philosophia with an F. Quintilian refers to the difference between them. He says of F: “Inter discrimina dentium efflanda est." Phi was uttered “fixis labris." F

. was a labial-dental. Phi seems to have been strictly a labial. The sound of Theta, like that of Chi and Phi, can not be determined by the usage of living languages. The modern Greek sounds it as the English th in thin. The remaining consonant, Zeta, will be considered presently.

The statements of the old grammarians and the incidental remarks of such authors as Plato and Aristotle, our second source of knowledge in Greek pronunciation, are given at length in Sophocles's “ History of the Greek Alphabet." These men were familiar with the classifications of Greek letters, and indeed the inventors of them. Our modern terms are little more than translations of the words of the Greek orthoepists. Their comments touch upon almost every point of phonology, and though sometimes vague and obscure, they are usually sufficiently explicit. They leave no doubt upon two important points, viz., that when the Greeks introduced their new alphabet, in 403 B.C., they pronounced as they wrote, without silent letters or orthographical expedients, and that they gave to each of their letters a single sound (differing in three vowels only as long and short), and assigned each sound to a separate letter. The only exception was the palatal nasal ng. Had there been silent letters or different sounds for the same letter, their numerous and minute criticisms would not have suffered the fact to

pass unnoticed. Were the Greek a derived and composite language, like our own, which employs the Roman alphabet, but has seven consonantal sounds unknown to the Latin, which abounds in silent letters and assigns several sounds to a single letter, and several letters to the same sound, there would be little hope of recovering its pronunciation. But we learn from the grammarians that its vowel sounds were but five; that its diphthongs, like the Welsh, retained for centuries each of the two primitive sounds; that its consonants were simple as well as few, viz., the four liquids; three smooth, three middle, and three aspirate mutes; two sibilants — the one a sonant, and the other a surd; 8, and the palatal ng.

The third source of evidence, which is found in the relations of the different letters, is especially satisfactory, as it comes from the language itself, without the intervention of interpreters. Never was there a language framed by man more musical than the Greek, or where euphonic changes were more universal. They are seen in the inflexions of nouns, adjectives, and verbs ; in compound words; in elisions, contractions, crases, and assimilations. Now, if we had not the true sounds of the letters, if we mistook labials for linguals and surds for sonants, no philosophical explanation could be given of these various changes. But philology refers them to well-known laws of the organs of speech, which are the same in all ages; and if the Greeks, with a nicer sense of harmony than ourselves, rejected, as discordant, certain combinations of letters which give no offense to modern ears, and hence made changes where change seems to us unnecessary; on the other hand, we certainly find nothing in the combinations of sound which we attribute to the Greeks that is at variance with our highest conceptions of harmony in speech. Illustrations of these laws are found in every grammar. The most familiar are those which unite sonant with sonant, surd with surd, labial with labial, aspirate with aspirate; laws which give us in Latin imponere, not inponere; lectus, not legtus; but which find their fullest illustration only in Greek. Curtius has well shown that certain contractions and crases prove that both the vowels of Greek diphthongs were sounded; that they were true, and not apparent diphthongs. Thus, the contraction of Kalya to Kayo proves that a in Kai was sounded, and report from trudovo shows that o in ou was sounded.

We have endeavored to convey some idea of the nature of the evidence that the true sounds of Greek letters are known, and not merely conjectured. It may be stated here that Professor Sophocles, himself a Greek, and the first philologists of Germany, have, after independent inquiries, arrived at the same general conclusions. They differ as to Theta and Zeta. Mr. S. assigns to both, their English sounds, and to Chi, the German ch.

We briefly present these results : ā=a in father, ēza in hate, ii in machine, õ=in note, ū=00 in rood (later French u). Short vowels differ from long only in length. In diphthongs both vowels had their primitive sound: ai-ai in aye (a short), ei=ey in they, oi=oi in voice (o as in note, but short), ui=ui in suing (nearly), au=ow in now, ou=ow in blow (later=00 in rood), eu=eu in feud (nearly), āi==ay in aye, ēi-y-ai in hail, õiew=oy (in buoy, as pronounced by Jameson, böe), zu=ā00, õu=ow in blow (prolonged).

Consonants as in English, except Chi, which - German ch in bach. The characters Phi, Chi, and Theta do not appear in most European languages, and their sounds are still discussed in Germany. Phi must have been near the English f. Theta was probably identical with th in thin. In the latter part of the classic period, Upsilon became the Frenchu, German ü, “dulciter spirans,” says Quintilian, and ou took the sound of oo in rood. Iota subscript becaine silent as early as the age of the Emperor Augustus. The later

sounds of the u and ou (Upsilon and Omicron-Upsilon) recommend themselves to Americans, as distinguishing more clearly the Omega from the diphthong ou.

What was the sound of Zeta? It was not ts, sd, or ds, for Quintiliar tells us that it was unknown to the Roman language. Ts was a combination forbidden by Greek euphonic laws, and the Hebrew ts is represented in the Septuagint by Sigma. Zeta, we know, was a sibilant and a sonant. It was developed, as Curtius has shown, from ye (Gamma Iota) and Ôe (Delta Iota). It interchanged with Sigma in words like Smyrna. It was a litera jucundissima, in the phrase of Quintilian. Plato speaks of its noble sound. It was said by the grammarians to be compounded of sd, and called a double consonant.

Was Zeta the English and Modern Greek z? Our z answers the description in every point but one: it is not a double consonant. The English z is a sonant sibilant. It would naturally be developed from ye and ò, as its surd 8 was developed from ze and Tl. It easily interchanges with s in English and

τι. French. S and d are, so to speak, blended in it. S is a sibi- . lant, d is a lingual-sonant; our z is a lingual-sonant-sibilant. No other two letters could represent it. So, ng represents our nasal palatal in ring, because n is a nasal and g a palatal; but neither n or g is separately heard in ring. So, th (in thin) represents a lingual-surd-aspirate, though neither t nor h is separately heard in thin, because t is a lingual-surd and h is an aspirate. When it was said by the grammarians that Zeta was compounded of sd, it could not be meant that they were separately heard, for then the sound would be Roman as well as Greek, nor could it well have interchanged with s in Smyr

But why was z called a double consonant, and why did it sometimes make a long syllable? It was the only sonantaspirate in the Greek language, and had a strong, continuous sound. It might make a long syllable, as the sonants bl made always a long syllable, while pl did not. The Modern Greek, Polish, Russian, and Lithuanian give to Zeta the English sound.

Some distinguished German philologists suppose Zeta to



« AnteriorContinuar »