TO THE EDITOR OF THE CLASSICAL JOURNAL.

THE rules given for the pronunciation of the Hebrew letters,

have in several respects been different from each other, and so uncertain as to their authority, that it is of no trivial advantage that an inquiry, as to the true mode of pronunciation, should have been so well begun, as that relative to the power of the letter gnain, inserted in your last No. As a specimen of correct and elegant critical discussion, that inquiry merits a distinguished place; and it is only as offering, with no less deference, some observations, which may possibly contribute to the determination of the question as to this, and some other letters of the Hebrew alphabet, that the following are submitted to your notice, and if you think proper, to that of your readers.

The writer of the inquiry having, to my apprehension, sufficiently proved that the power of the gnain is that of some kind of spiritus asperrimus; the only remaining question is, can it be ascertained with any degree of precision farther, as to the kind, whether it be simply a hard aspiration, or an articulate sound, together with such an aspiration? As the Arabic, a dialect of the Hebrew, is still a living language, a reference to the powers of the Arabic letters seems to be the most obvious and certain method of forming a just idea of the powers of the Hebrew letters; and the more so, as the Arabic seems to have suffered little or no change for many centuries. Now in the Arabic alphabet, we find two characters, the ain and gain, retaining a similar name to that of the Hebrew gnain, and holding the same place in words of the same sense, and the same form; so that there can be no doubt that the Arabic ain or gain and the Hebrew gnain, originally designated the same sound. The sound of the Arabic ain is an aspirated sound formed in the throat, and may be represented by gh; if the gh be supposed to bear such a relation to g as the Welsh ch, or Greek x, does to the hard c or k. The sound of the gain is that of the Northumbrian burr, and may be represented by ghrr. Whether the Hebrews employed both these sounds, cannot perhaps now be known; though, as a more cultivated people, it may be presumed they rejected the harsher sound; and that the milder is the proper one of the Hebrew gnain. The Arabs read Js and J, the same root lactare, the former with ain, the latter with gain, and with very little difference as to the sense. The former is translated sustentare familiam; the latter lactare; which to a people, whose principal sustenance is milk, is nearly the same thing. This seems more like a distinction of dialect than of sense; but whether this is generally the case or not, I am not sufficiently conversant with the Arabic to determine, though I am inclined to think it is.

as

עול

If it be permitted to assume, that the true sound of the Hebrew gnain is such as I have represented it, the reason why no correspondent character is assigned to it, in the Greek expression of Hebrew words, will be evident, viz. that the Greek alphabet afforded none, and the Greeks probably could not pronounce it, or consider the sound otherwise than as a harsh aspirate. And for a similar reason it may be inferred, that the sound of ng or gn; or more properly ngn, was made to represent that of the gnain by some one whose native language had no gutturals; and was strongly nasal. Such a person, in the attempt to pronounce the guttural gh, would infallibly articulate the sound of the ng in strong instead of it.

I suspect that it really was so done by Petrus Martinus, in his Hebrew Grammar.

The Hebrews having so long been dispersed among other nations, the colloquial language, which they learned when young, must necessarily have affected the pronunciation of their own; and therefore from European Jews we cannot expect to learn any very accurate method of pronouncing it; and as to the supposed correspondence of the letters of the Hebrew with those of the Greek alphabet, in this and three more instances, because they hold a similar, place in alphabetic order, this is by no means a necessary consequence; because they probably would substitute a necessary letter for one which was not so, and which they could not pronounce; though they had their alphabet from the Hebrews, as the very name of Cadmeans, which signifies people of the East (and they probably were a colony of the Canaanites) evinces. With respect to the gnain, this supposition is a most unfortunate one, as it makes every verb, in which this letter occurs, irregular; and, where it is the middle radical, the root forms only one syllable instead of two. Thus, 727 ghabad, becomes obad; you shamagh, shamo; and Paghal, the old example of the regular verb, Pol. This is of itself a serious objection; and, should the Hebrew student proceed to study the Arabic, would be much more so, as it would impede his recognising the same word differently pronounced. If he does not, and dislikes the guttural, the common rules of expressing the guain by gn when it begins, and by ng when it terminates a syllable, may, for distinction's sake, be found the most useful.

I presume that the same idea of a supposed relative similarity of the powers of letters, similarly disposed in the Hebrew and Greek alphabets, has induced the authors of some Hebrew grammars to give to the samech the power of sh; and to the schin that of s; to the teth that of th, and to the thau that of t; and to the he that of the vowel e; contrary to the rules of the old grammarians. I take them in this order, because it admits of proof, that the innovation is decidedly erroneous in the first instance, and therefore the principle also, on which the innovation has been founded. For the proof it is only necessary to refer to Judges chap. xii. v. 6. "Then said they unto him, say now Shibboleth (a) and he said Sibboleth (bao)." In these words the schin, has indubitably the power of sh and the samech that of s; because he, who can pronounce the sh, can pronounce the s; whereas the articulation of the sh is found difficult and almost impossible by whole nations, to whose language it is not proper. In this instance, therefore, the reference to the Greek alphabet is wrong, and the principle is therefore at least questionable in the

".(סבלת) (שבלת)

others.

From the places of the teth and theta, each being the ninth letter of its respective alphabet, the assumption, that their powers were the same, is certainly plausible; but however plausible it may

be, the single word л, Táxia, Talitha, proves the reverse, as do also the proper names , 9, Seth, and , AT, Lot, in which the corresponds to t and the to the theta or th. And though the German Jews cannot pronounce the th, yet the manner, in which they attempt it, shows a traditional idea of the original power of the letter; for they give it the power of s (unless the aspiration be abated by dagesh, when it has the power of t) as they who cannot articulate the th generally do. In expressing the by s, Buxtorf has also generally complied with the custom of the Jews, and written Toras for A Torath &c. when the ends a word. In a few instances he gives it the power of th, as in л, schabbath, which the German Jews pronounce shabbus.

That the heth had the power of an aspirate, has not, I believe, been disputed; and that the he, was a simple aspirate, may also be inferred from the Greek expression of WW, "None, Hosea, my, Tála, Gaza, NT, Aages, the earth, NT, "Iov or 'Inov, Jehu. In this last expression, which is that of the Codex Alexandrinus, the seems to have been used as the character of aspiration, which it is known to have had; (for were it that of the long e, it would be difficult to account for its omission in other copies) and if so, this must be the reading of a very ancient copy.

However this be, the instances here adduced are sufficient to show that the had no determinate representative in the Greek alphabet, when the septuagint was written; and that sometimes a word, of which it is a constituent letter in the Hebrew, is in its Greek expression written without any alphabetic representative of it whatsoever, which can scarcely be accounted for, unless it be admitted that the was the character of a simple aspiration.

It may still be urged, that this admission increases the number of aspirates to an unusual degree. I grant it does; but not so as to cause any confusion. On the contrary, they may thus be accurately distinguished; the y guain being pronounced as the guttural gh; the caph as the Welsh ch, the heth as the German ch or Spanish j; and the he as the English h or simple aspirate. It should here be observed that the ch, or aspirate of c is guttural in the Welsh, and palatical in the German. The English language employs none of these except the simple aspirate; and therefore to one, who speaks the English language only, it is almost impossible to make the necessary distinctions in pronunciation. To such, the most useful rules will probably be such as are found in Bythner's Grammar; or to give to the gnain the power of gn or ng, according as it begins or ends a syllable, to the caph that of kh, and to the heth and he that of h.

If what has now been advanced be correct, it will follow, that the Hebrew alphabet has in reality only three characters, viz. "N for vowels. The Arabic has the correspondent elif, waw, and ye,

and no more in its alphabet; from whence it may be inferred, that both these alphabets were originally constructed on the same system; and in fact the Arabic consists of little else than cursive characters derived from the Chaldaic, or common Hebrew alphabet.

This scanty provision of vowel characters indicates some peculiarity in the principle, on which these alphabets were originally constructed, arising from the nature of the languages to which they were adapted; and the more so, as in a great, if not the greater, number of words, none of the three vowel characters is introduced. Did the distinct signification of different words depend chiefly on the vowels, it is evident that a particular attention must, in constructing an alphabet, have been paid to vowel characters, and a sufficient number have been provided to mark the necessary distinct sounds; and therefore, when the alphabet has but three such characters, it may be inferred, that, though some vowel sound is necessary to be added to a consonant, in order to form a syllable, the sound annexed is obscure, and that no great precision is required. And this seems to be the nature of the Persian and Arabic languages; as the rule given in the grammars of these languages, for the supplementary vowel is, that it should be an obscure sound, such as a short a, or as the u in the English word but. The same inference may also be drawn from the very mode in which additional characters or points are used to supply the deficiency; as there are but three, fatha for a or e, kesra for i, and damma for o or u; and if the difference between the sounds of the a and e, or the o and u, produce any material ambiguity as to the meaning of the words, undoubtedly distinct characters for each of them would have been too necessary not to have been assigned to each. This inference will appear the more just, when it is observed, that the roots of all regular verbs in the Hebrew and Arabic consist of three, and in some instances of four, consonants; and that additional consonants, except in a few cases in which the person is easily determinable by reference to the subject, concur in forming the personal inflexions. The irregular verbs consist of three letters, one at least of which is a vowel, or the simple aspirate; if a vowel, it is one of the three which has an alphabetic representative. For these, such a character was indispensable, and they are all, except the aspirate, the characters of long vowels; and as such only seem to have been absolutely necessary, I conceive that no great stress was laid on the supplementary sounds, and no great precision as to these of much importance. Perhaps also, where the nature of the language and a regularity of the inflexions would admit of it, the advantage of abbreviation, by omitting the short vowels in writing, might have contributed to the custom of so doing, where the omission was not likely to render the sense ambiguous; but for ge