a rapidity which some of us deplore; there is nowadays an engineering tripos, an agricultural tripos, and we may even have a bootmaking tripos before long. But in anything which affects the spiritual welfare of her children it seems impossible to move her an inch.

One feels that an article of this kind, in which the writer has not hesitated to put his opinions plainly, must needs meet with many readers who will differ from its conclusions, although they can hardly dispute its facts. But at least all will agree that the question is one of considerable importance. At least, too, the writer hopes, all will credit him with sincerity and earnestness of purpose in writing it, and will share his real and trustful hope that before long the rulers of the Universities will awake to a sense of their responsibilities; that Christianity rather than agnosticism, that the Church militant rather than the Church dubitant, may hold sway there once more; and that Oxford and Cambridge may better fulfil their proud claim to be homes of religious education' as well as of sound learning' by fostering and cherishing amongst their younger members that one great Faith without which all human wisdom is but folly, all human knowledge but vanity.

ANTHONY C. DEANE.

THE PROPER PRONUNCIATION OF GREEK

I

No branch of the science of languages is of a more absorbing interest or of a wider scope than that which inquires into the relations of the spoken to the written word, the correspondence of the living voice to the lifeless image, the analogy between pronunciation and alphabet.

The invention of writing as a vehicle of thought and speech-that is, the transference of speech into script by means of conventional signs or symbols of sound, whereby communications are preserved for silent reading or audible enunciation-is an achievement of human ingenuity far excelling in importance and consequence the art of printing. For the transition from manuscript to type was but a means of multiplying what had already been recorded graphically.

Yet, writing is as inferior to the living word as it stands above printing. The primary object of writing-that of representing to the eye, with the greatest accuracy possible, the impressions created upon the ear by sound-can, in the nature of things, never be completely attained at the outset ; while, in course of time, the representation of speech, as originally fixed, ceases to respond always to sounds which are constantly subject to modification. The symbolism of the dead letter-the onμara λvypà-cannot keep pace with the mobility of quickened speech-the ἔπεα πτερόεντα. The former remains inert and lethargic; the latter is ever responsive to the varying influences of human sense and faculty, which react upon it and constantly modify its character and tone. As speech falls short of thought, so conventional script is necessarily but an imperfect rendering of articulate sound.'

There are two main reasons for the imperfections which are inseparable from the art of writing. The human voice is an organ of such delicacy and flexibility, it is able to note distinctions of sound so minute and subtle, so diversified in pitch and tone, that it is impossible to record adequately all its varied modulations by means of an alphabet. Hence, all alphabets are, in comparison to actual

· Λόγον ζῶντα καὶ ἔμψυχον οὗ ὁ γεγραμμένος εἴδωλον ἄν τι λέγοιτο δικαίως. (Plat. Phædr. 276 A.)

VOL. XXXVIII-No. 221

681

Y Y

pronunciation, insufficient and imperfect. Some are deficient in letters expressive of certain elementary sounds; some include letters superfluous, inasmuch as other letters convey the same sounds; some possess ambiguous letters-namely, such as serve to note two or three various sounds. These defects are often all present in one and the same alphabet, whether it consists, as in the Slavonic tongues, of forty-eight letters, or of twenty-six, as in the common alphabet with which the Romanic and Teutonic languages manage to represent widely divergent pronunciations. In every instance, however, each language employs a more or less limited number of vowels and consonants in order to express a great variety of sounds. The shortcomings, therefore, of the art of writing result both from an originally imperfect construction of the alphabet, and from the divergence, which grows out of usage and becomes gradually apparent, between articulate sounds and their graphic exponents.

Various causes tend to bring about such divergencies and modifications in pronunciation; the most prevalent being a natural tendency to abbreviate words, by contracting two syllables or blending several letters into one, and to render articulation easier and more fluent, by substituting soft sounds for harsh ones, or for such as may demand a greater effort to pronounce. Local usage, idiomatic peculiarities, fashionable innovations, foreign influence—all work in the same direction.2

Beyond this, however, all men do not pronounce alike, nor do they conceive the sounds of words in like manner. Consequently, before the orthography of a language becomes fixed by some common accord, one and the same word often occurs spelt variously, as it may have sounded to each writer at a time when no settled law served him as guide. The permanency of this natural diversity in the conception of word-sounds is manifest, not only in the errors of spelling of halfeducated persons, but more especially in the disagreement of the

2 Not only the several towns and counties of England have a different way of pronouncing, but even here in London they clip their words after one manner about the court, another in the city, and a third in the suburbs; and in a few years, it is probable, will all differ from themselves, as fancy or fashion shall direct all which, reduced to writing, would confound orthography.'-Swift's Proposal for Correcting and Ascertaining the English Tongue, 1711.

The pronoun it is spelt in Tyndale in eight different ways: hyt, hytt, hit, hitt, yt, ytt, it, itt. The word tongue is also met with as tung, tong, tunge, tonge, tounge; and head as hed, heede, hede. Archbishop Trench enumerates the following fourteen ways in which the word sudden is spelt in early writers: sodan, sodain, sodaine, sodayne, soden, sodein, sodeine, sodeyn, soddain, sodden, suddaine, suddein, sudden, sudeyn. To this day the Greek word ecstasy is often written also as extasy, extacy, ecstacy; and farther occurs quite constantly instead of further, the correct comparative of forth. The names of Shakespeare and Raleigh are written in a variety of ways; in Germany the very common surname of Meyer is met with also as Maier, Mair, Mayer, Mayr, Meir, Meier, and Meyr. Even more numerous are the variations of Morrel, Morell, Maurelle, &c.

In Notes and Queries, No. 147, it is stated that the postmaster of Woburn had

6

advocates of phonetic spelling, even when they use one and the same phonetic alphabet. There are also the inborn and inherited, or the studied and affected, peculiarities of individual speakers; and, finally, local usages seem to blunt people's perception of the niceties of articulation.7

These various influences ever tend to widen the chasm existing between the spoken and the written word. Partial reforms in the spelling of certain words or the disuse of certain letters, restore some measure of approximation between pronunciation and orthography. The obsolete Greek Vau or Digamma, for instance, though it has long disappeared from writing, had not yet ceased to be pronounced when the Homeric poems were composed, and they cannot now be correctly scanned unless it be observed in metre. The letter 8, having first become silent in the pronunciation of many French words, is now marked by a circumflex on the vowel which preceded it; while oi has noted no less than 244 variations of the spelling of the name of that town on letters there addressed. And in the Times of the 27th of September, 1876, we read that Mr. C. T. Townsend, the Danish and Norwegian Consul at Ipswich, has for the last three or four years received letters from Northern Europe, on the envelopes of which are some extraordinary variations of the Suffolk capital. Subjoined are no fewer than fifty-seven of these incorrect orthographies: Elsfleth, Epshoics, Epshvidts, Epsids, Epsig, Epsvet, Epsvidts, Epwich, Evswig, Exwig, Hoispis, Hvisspys, Ibsvi, Ibsvig, Ibsvithse, Ibwich, Ibwigth, Iepsich, Ie yis Wich, Igswield, Igswig, Igswjigh, Ipesvivk, Ipis Wug, Ips Witis, Ipsiwisch, Ipsovich, Ipsveten, Ipsvick, Ipsvics, Ipsvids, Ipsvidts, Ipsvig, Ipsvikh, Ipsvits, Ipsvitx, Ipsvoigh, Ipswch, Ipsweich, Ipswgs, Ipswiche, Ipswick, Ipswict, Ipswiech, Ipswig, Ipswigh, Ipswight, Ipswish, Ipswith, Ipswitz, Ispich, Ispovich, Ipswich, Ixvig, Iysuich, Uibsvich, and Vittspits. It is only due to the deciphering powers of the Postal authorities to say that these letters generally come to Mr. Townsend without delay.'

One of these advocates, Mr. Ellis, writes:-Mr. Bell's pronunciation, in many instances, differs from that which I am accustomed to give, especially in foreign words. Both of us may be wrong.' And Mr. Sweet says of Mr. Ellis: Mr. Ellis insists strongly on the monophthongal character of his own e's and e's. I hear his e (may) and o (no) as distinct dipththongs, not only in his English pronunciation, but also in his pronunciation of French, German, and Latin.'

I could mention the names of three bishops, one of whom pronounced the vowel in God like gaud; another like rod; a third like gad.'-Professor Max Müller On Spelling.

These peculiarities are fully set forth in Professor Max Müller's admirable essay 'On Spelling' (Chips from a German Workshop, vol. iii. pp. 285-7, new ed.). To convince people that the one pronunciation (of English) is right and the other wrong seems utterly hopeless. I have heard a highly cultivated man defend his dropping the h in the beginning of certain words, by the unanswerable argument that in the place where he was brought up no one pronounced these initial h's. What Scotchman would admit that his pronunciation was faulty? What Irishman would submit to laws of spelling passed in London? . . . I have heard Americans maintain in good earnest that there was much less of nasal twang in America than in England. People are not aware how they pronounce and how differently they pronounce one and the same word. . . . The confusion becomes greatest when it is attempted to identify the pronunciation, say, of a vowel in German with a vowel in English. No two Englishmen and no two Germans seemed to be able to agree on what they heard with their ears or what they said with their tongues; and the result in the end is that no vowel in German was really the same as any other vowel in English.' See also A. H. Sayce, Introduction to the Science of Language, 1880, vol. i. pp. 260, 299–301.

been replaced by ai in some substantives and in the imperfect and conditional of French verbs, where it had gradually come to be sounded as ai. In like manner chirurgeon and sacristan were first pronounced surgeon and sexton, and then so written in English.

Such readjustments, however, are, in all languages, only rare exceptions amidst the multitude of words, the pronunciation of which differs vastly from their spelling. They continue to appear in one shape upon paper, while they produce another sound from the mouth. And the written word will ever remain but an imperfect image of articulate speech; pronunciation being too subtle and intricate to be more than approximately represented.

No language offers more distinct and convincing instances of these facts than the English language, the pronunciation and spelling of which is so peculiar, as to constitute its one great and, in some cases, insurmountable difficulty.9 The foreign student, more especially, is puzzled by the singular and arbitrary custom of sounding similar combinations of letters in a dozen different ways, and different combinations of vowels in one and the same manner.10 There seems to be no reason for orthographic anomalies such as deign and disdain, scent and dissent, precede and proceed, high and height, tie and tying, &c.; rhyme is thus unorthographically written, instead of rime, as if it were derived from rhythm. Words like colonel, laugh, enough, victual are, as now pronounced, phonetic riddles; it remains inex

8 This, with a few other modifications, constitutes the so-called Orthographie de Voltaire, who maintained that 'l'écriture est la peinture de la voix; plus elle est ressemblante, meilleure elle est.' He considered it incongruous to write emploieroient, octroieroient, and pronounce emploieraient, octroieraient. This change in pronunciation dates back to the sixteenth century, having originated in the Italianised court of the Valois, in which the primitive French sounds came to be considered harsh. Racine substituted ai for oi in verbs; but it was Voltaire's authority which made the rule absolute, with all but the higher clergy and a few writers, e.g. Chateaubriand and Nodier. Both crthographs still subsist in harnais and harnois, but each with a special acceptation.

9 'If we compare English as spoken and English as written, they seem almost like two different languages; as different as Latin from Italian.'-M. Müller, loc. cit. p. 261. 10 Mr. Meiklejohn, late Assistant-Commissioner of the Endowed Schools Commission for Scotland, sums up as follows the perplexities of English pronunciation and spelling: 1. Out of twenty-six letters, only eight are true, fixed, and permanent qualities-that is, are true both to eye and ear. 2. There are thirty-eight distinct sounds in our spoken language, and there are about four hundred distinct symbols (simple and compound) to represent these thirty-eight sounds. In other words, there are four hundred servants to do the work of thirty-eight. 3. Of the twenty-six letters, fifteen have acquired a habit of hiding themselves. They are written and printed, but the ear has no account of them: such are w in wrong, and gh in right. 4. The vowel sounds are printed in different ways: a long o, for example, has thirteen printed symbols to represent it. 5. Fourteen vowel sounds have one hundred and ninety printed symbols attached to their service. 6. The single vowel has five different functions; it ought only to have one. 7. There are at least 1,300 words in which the symbol and the sound are at variance-in which the word is not sounded as it is printed. 8. Of these 1,300, 800 are monosyllables-the commonest words, and supposed to be easier for children.