Again, a very large and various field of proof lies in those instances of the direct transference of the sounds of one language into those of another, which literary composition sometimes requires, and which are sure to occur very frequently in an extensive literature like the Greek. Examples of this are most common in the case of proper names, and occur especially in translations, as in the ancient translations of the Hebrew Bible and of the New Testament, which have been admirably used for the illustration of Greek orthoepy in the work of Seyffarth. When Strabo, for instance, (p. 213,) in the case given by Pennington, (p. 73,) says of the inhabitants of the newly colonized town of Como in Upper Italy,—Νεο κωμῖται ἐκλήθησαν ἅπαντες· τοῦτο δὲ μεθερμηνευθὲν Νο ϐουμκώμουμ λέγεται, we learn that the diphthong ου was considered by an intelligent scientific man in the time of Augustus, as being either the exact equipollent of the Latin u, or the nearest approximation to it within the compass of Hellenic vocalization; and when we are told further that the modern Greeks and the modern Italians pronounce the same vowels the same way even now, we cannot for a moment doubt that the method of pronouncing that Greek diphthong now practised in Scotland (as in boom) is the correct one. From the same passage we may legitimately draw the inference, with regard to the second letter in the Greek alphabet, that it was in all probability pronounced softly like our v; for our b is no representative whatever of the Latin v, whether we suppose that letter to have been pronounced like the corresponding letter with us, or like our w. The modern Germans, in the same way, who have not our sound of w, substitute for it in their language the sound of v regularly, as in wasser, which they pronounce VASSER, and many such words. If, therefore, an ancient Greek wished to express the letter v, and does so by his own B, the inference is irresistible, either that his B was pronounced like our v, and was viewed as the exact expression of the Latin letter so pronounced, or as an approximation to it, if pronounced like our w; or, on the other hand, that the Greek organ being utterly incapable of pronouncing the soft sound of the Latin v, and having no letter or combination of letters capable of expressing it, gave up the attempt in despair, and wrote the soft Latin v with a hard Greek B. But this supposition is improbable, for three reasons: First, because the general character of the Greek language, as contrasted with the Roman, was not that of blunt hardness but of liquid softness, (see Quinctilian and Cicero, passim;) Secondly, the ancient Greeks, in fact, had a combination of letters by which they could express in an approximate way the Latin v, namely, ου, and by which they actually did so express it on many occasions; Thirdly, the modern Greeks likewise do pronounce the second letter of the alphabet like the Latin v; and the burden of proof lies on those who assert that the ancients pronounced it otherwise.
A fourth method of proof lies in the remarks made on the identical or cognate sounds of syllables, either incidentally by general writers, or specially by grammarians in treating orthography and orthoepy; and in the accidental interchange of letters in inscriptions and coins. Of the strictly grammatical kind of evidence a very valuable fragment has been preserved in the Ἐπιμερισμοί of Herodian, the Priscian of the Greek grammarians, published by Boissonnade in 1817. In this work are alphabetically arranged large classes of words, which, while they are pronounced with the same vowel to the ear, are differently spelt to the eye; as if I should say in English that the vowel-sounds in the words fair, fare, heir, there, have the same or a similar orthoepy, but a very different orthography. Of the other, or incidental kind, may be mentioned those plays of sound with which epigrammatic writers sometimes amuse themselves, and of which the echo-poems found in some of the collections of modern Latin, are the most notable example. Thus, Erasmus, in ridicule of the Ciceronians, wrote two lines, of which the first, a hexameter, ends with Cicerone, the ablative case of the great orator’s Latin name, while the second line, a pentameter, striking the ear as a sort of echo of the first, ends with the Greek word ὄνε, O you ass! from which significant jingle the inference is ready enough, that the penultimate syllable of both these words, in the classical pronunciation of Erasmus, was accented, and that the sound of the vowel in both was the same. The proof, of course, in such a case would have been equally complete if the word in the second line had been spelt with a different vowel instead of with the same.
Fifthly, In determining the pronunciation of any language at any past period of its history, its presently existing pronunciation, though furnishing no absolute proof, is entitled to be taken into account along with other circumstances, and in the absence of any distinct evidence to the contrary, must be taken as conclusive. Erasmus appealed with great success to the vanity of academic men, when he said, with reference to the common Greek pronunciation in his day, “Pronuntiationem, quam nunc habent eruditi, non aliunde petunt quam a vulgo, scis quali magistro;” but to this a learned advocate of the existing Itacism very wisely replies, that even supposing it were true that the vulgar pronunciation of Greek comes to us only from the vulgar, the common people, as is well known, are generally far more tenacious of hereditary national accent than the upper classes of society;[12] of which we have a familiar English example in the case of the stout Yorkshiremen, who have preserved for two thousand years the deep hollow sound of u, (saying Ool, for Hull, &c.,) which is the normal sound of that vowel in all the European languages. In this view it is passing strange to note, that the slender sound of the first syllable of ἡμέρα, as if written heeméra, which is the rule with the modern Greeks, is the precise sound, that in a passage of Plato is noted as the ancient sound, compared with the fuller sound, haiméra, fashionable in his day;[13] while Aristophanes[14] in one of his plays, introduces a conservative old Spartan lady saying ἵκει, instead of ἥκει; a distinct proof both that η was not considered identical with ι in his day, and that it was then sounded as it is now, by one of the most ancient people in the Pelasgic peninsula.
Such appear to me to be the methods of proof that lie open to an inquirer into the orthoepy of any language, living or dead, at any given period of its history. With these, of course, the student must combine such general rules on the philosophy of language, and on the habits of human speech, as a little experience of practical philology will readily supply. I now proceed to state the results to which I have arrived, by a thorough study of the existing evidences. After that we shall make our practical inference, and answer a few natural objections.
In the shape of results, therefore, all that my present purely practical purpose requires me to lay down, with regard to ancient Greek vocalization, may be combined in the following two propositions—
Proposition I.—It is demonstrably certain that the method of pronouncing the vowels and diphthongs generally practised in England and Scotland, especially in England, since the days of Sir John Cheke,—that is from about the middle of the sixteenth century—is doubtful in many points, and in not a few most important points directly opposed to the whole stream of ancient authority and tradition. It is in fact in a great measure conjectural, arbitrary, and capricious.
Proposition II.—It is equally certain that the modern Greeks have declined in several most important points from the purity of Hellenic orthoepy, as practised in the most classic times; but many of the striking peculiarities of the modern pronunciation can be traced back, with more or less uniformity, to a period not far removed from the most flourishing period of Greek literature, a period certainly when pure Greek was both a spoken and a written language, and preserving such a living organic power, as entitled it by a spontaneous impulse from within to modify the laws of its own orthoepy.
Both these propositions, so far as the vowels are concerned, are proved by a single glance at the passage of Dionysius (περὶ συντάξεως) already referred to, and which I shall now translate:—
“There are seven vowels; two long, η and ω, and two short, ε and ο; three both long and short, α, ι, υ. All these are pronounced by the wind-pipe acting on the breath, while the mouth remains in its simple natural state, and the tongue remaining at rest takes no part in the utterance. Now, the long vowels, and those which may be either long or short, when they are used as long, are pronounced with the stream of breath, extended and continuous; but the short vowels, and those used as short, are uttered by a stroke of the mouth cut off immediately on emission, the wind-pipe exerting its power only for the shortest time. Of all these, the most agreeable sounds are produced by the long vowels, and those which are used as long, because their sound continues for a considerable time, and they do not suddenly break off the energy of the breath. Of an inferior value are the short vowels, and those used as short, because the volume of sound in them is small and broken. Of the long again, the most sonorous is the α, when it is used as long, for it is pronounced by opening the mouth to the fullest, while the breath strikes the palate. The next is η, because in its formation, while the mouth is moderately open, the sound is driven out from below at the mouth of the tongue, and keeping in that quarter does not strike upwards. Next comes the ω, for in it the mouth is rounded, and contracts the lips, and the stroke of the mouth is sent against the extreme end of the mouth, (ἀκροστόμιον, the lips, I presume.) Inferior to this is the υ, for in this vowel an observable contraction takes place in the extreme region of the lips, so that the sonorous breath comes out attenuated and compressed. Last of all comes ι, for here the stroke of the breath takes place about the teeth, while the opening of the mouth is small, and the lips contribute nothing towards giving the sound more dignity as it passes through. Of the short vowels, neither is sonorous; but o is the least agreeable, for it parts the mouth more than the other, and receives the stroke nearer the wind-pipe.”
Now, while every point of this physiological description may not be curiously accurate,[15] there is enough of obvious certainty in it to settle some of the most important points of Greek orthoepy, so far as the rhetorician of Halicarnassus is concerned; and his authority in this matter is that of a man of the highest skill, which, as the daily practice of our law courts shows, is worth that of a thousand persons taken at random. That the Itacism of the modern Greeks did not exist, or was not allowed by good speakers[16] in the time of this writer, so far as the single vowels are concerned, is abundantly manifest; for not only do η, ι, υ, which the modern Greeks identify, mean different sounds, but the sound of the η in particular is removed as far from the ι as it could well be in any scale of vocalization, which sets out with the supremacy of the broad a. And if these sounds were distinguished by polished ears in the days of Augustus Cæsar, it is contrary to all analogy of language to suppose that in the days of Alexander the Great, Plato, or Pericles, they should have been confounded. Provincialisms, indeed, and certain itacizing peculiarities, such as that noticed by Plato, ([page 24 above]), there might have been; but that any language should confound its vowel-sounds in its best days, and distinguish them in its days of commencing feebleness, is contrary to all that succession of things which we daily witness. Different letters were originally invented to express different sounds, and did so naturally for a long time, till fashion and freak combined with habit, either overran the phonetic rule of speech by a rank growth of exceptive oddities, (as has happened in English,) or fixed upon the organs of articulation some strong tendency towards the predominance of a particular sound, which in process of time became a marked idiosyncrasy, from which centuries of supervening usage could not shake the language free. This is what has taken place in Greece with regard to certain vowel-sounds. But before pursuing these observations further, let us see distinctly what the special points are, that this remarkable passage of the Halicarnassian distinctly brings out. The ascertained points are these,—