Thee facts are well illustrated by English spelling, which in its existing standard form presents an extreme case of phonetic corruption or dealphabetization. The earliest English (Anglo-Saxon) spelling was almost purely pho netic, that is, the scribes sought to indicate the actual sounds of the words they wrote, using— with a high degree of precision—for this purpose the Roman alphabet (with certain modifications) and giving to its letters the values assigned to them in Ihe pronunciation of Latin. with which they were familiar. But as the sounds gradu ally changed, confusion, due largely to imperfect adaptation (sec above), slowly set in; certain symbols were not accurately distinguished, or were employed with different values in different combinations; the significance of some was ir regularly altered and new ones were introduced; others were lost ; and the accentual marks over the long vowels used in Anglo-Saxon disappeared, increasing the difficulty in distinguishing the long from the short vowels—a defect which has never been remedied except (in part) by non alphabetic devices. But the most impor tant irregularities were due to the influence of Norman. and later of literary (Parisian) French, tinder which, in the thirteenth and four teenth centuries. English was practically re spelled in accordance with Anglo-French meth ods. Notwithstanding this and other sources of confusion, however, English spelling remained for a considerable period in a notable degree phonetic. The change effected by the invention of printing has been indicated above; but it should be added that Ihe printer's choice of spellings not only was not guided by any princi ples, but was often determined by ignorance, English having suffered, in this particular, more than any other language. Alteration in response to phonetic change was not, indeed, wholly stopped—for it has continued in some measure until the present day—but it was made more and more difficult, and from the sixteenth century the power of effecting any substantial phonetic improvement of English spelling has been lost. Notwithstanding many later modifications of de tails (there are still over 3000 words whose orthography is unsettled), the development of English orthography ends with that period, the subsequent life of the spoken language being practically unrepresented in it. In direc tion—that of sound—the alteration has been great, but we still retain what is essentially the Elizabethan spelling, in total disregard of the fact that the pronunciation which it embodies has very largely disappeared.
Another important source of confusion in Eng lish spelling, closely associated with this process of fixation, must also be mentioned. Under the in fluence of the revival of learning, which brought words of classical origin into special prominence, the idea was developed that (regardless of their pronunciation) such words should be made to conform in spelling as exactly as possible to the Latin and Greek terms from which they were ultimately derived. This etymological theory of spelling gained strength, both from the weakness of the contemporary feeling for phonetic accu racy and from the practical difficulty found by the printer, in his quest of uniformity, in mak ing a selection among existing forms, which offered a wide opportunity for the activity of pedants; and from the sixteenth century until the present lime it has kept its hold upon our orthography. There was much tinkering, not only with words directly borrowed from the class ical tongues, but also with words of Romance origin, while the native English and Scandi navian elements of the language were but little interfered with. A result was the introduction
of many forms erroneous from both the phonetic and the true philological points of view; for apart from the indefensibility of this use of spelling as a means of rendering derivation obvious to the eye the method was dangerous in the hands of men whose knowledge of the history of the lan guage was necessarily inadequate. A familiar example is the present English debt (from Early Modern English and Middle English det, dette, from Old French dette), in which the b was ety mologically inserted both in French (though later abandoned) and in English to make the spelling more directly suggest the original Latin debita, though it has never been pronounced. On the other hand, this process in some cases actu ally corrupted pronunciation. For example, the 1 in fault (Middle English and Old French lance), which was inserted in the same way, to suggest the Latin fallcre, has actually come to be pronounced, though the correct pronunciation (in this particular) survived as late as the time of Pope, who makes the word rhyme with ought, thought, and taught. Qf pure etymological blunders, again, an instance is the s in island, which was inserted to indicate derivation from the Latin insula and connection with English isle (also corrupted from ilc), to neither of which it is at all related. The corruptions of this kind form a long and instructive list.
The results a the above-indicated history, em bodied in modern English orthography, are briefly as follows: (1) While all of the other principal European alphabets have retained with comparatively small variations the Roman or 'Continental' values of the letters, that of English has to a very large extent abandoned them. This is especially true of the 'long' vowels—at least of their common or name-giving values: Thus, ap proximately, 'long a' (say) = Cont. é; 'long e' (mete) = Cont. i; 'long (isle) = Cont. diphthong ai; 'long o' = Cont. 6; 'long a' (duty) = Cont. diphthong in. Great confusion thus exists in the vowel-nomenclature, as well as an unfortunate divergence from the common usage of those languages which are most closely connected with English. both historically and practically. The quantity of vowels, also, as. was noted above, is not (as in oldest English) distinguished by any alphabetic means, while the various orthographic devices employed for the purpose (doubling of consonants, etc.) are clumsy and are not used with uniformity.
(2) The letters of the alphabet are used with a great diversity of sound-values: Thus, of the vowels, accented and unaccented, a has 9 (as in name, bare, man, father, water, want, ask, village, data), c 9 (be, here, there, acme, met, alert, English, sergeant, prudent) ; i, 8; o, 10; u, 9; y, 3-48 in all; of the consonants, b, 2 (counting the silent ones) ; e, 6; d, 4; f, 3; g, 4; h, 3; j, 5; k, 2; 1, 3; in, 3; a, 3; p. 2; q, 3; r, 2; s, 5; t, 5; v, 2; w, 2; 5; y, 2; z. 4— 70 in all. To a certain extent this multiplica tion of values is due to the inadequacy of the alphabet, which has but 26 sym bols to represent the 40 elementary sounds of cultivated English speech (44 with the diph thongs), but it is far in excess of what is neces sary and results in manifold ambiguities.