Henry Sweet (1800-1910) was the greatest authority on historical phonetics of his day.  His books on Anglo Saxon pronunciation are still in print.  His work on a new orthography for English influenced the defvelopment of the IPA (International Phonetic Alphabet) but is now largely forgotten. One of these important works is digitized below.

Sweet's publication comes after the pioneering work of Isaac Pitman and Henry Ellis.  Sweet critiques Ellis' Glossic notation which evolved into New Spelling and is the basis of countless new orthographies for English that are based on the most frequent English spelling patterns. This agreement with Nomic (the traditional orthography (TO)) is obtained at a great sacrifice of phonetic consistency. Sweet favored International spelling patterns.  He called his augmented Roman notation with Latin sound correspondents Romic.

Sweet favors using the Roman alphabet and the original Latin sound values.  For the sounds in English speech that are not represented, Sweet recommends augmenting the Roman characters with 2 Anglo Saxon symbols (ash and eth), 1 greek character (theta), and 2 turned letters (  ). Sweet employs q for /ng/ and c for /tsh/.  Sweet uses two base pronunciations for romic, RP and American. He wants the notation to be elastic enough to accomodate both.

THE absolute necessity of phonetic reform is now almost universally recognised, not only by practical teachers but also by scientific philologists. All the objections that prejudice and irrational conservatism have been able to devise have been successfully met, and the only question now is, What system shall we adopt?

 The great difficulty of arriving at any agreement is the multiplicity of possible systems. Any system, however clumsy and arbitrary, which clears away only a portion of the irregularities of the existing spelling, is an improvement on it. Any one, for instance, if he likes, can drop the silent w in such words as write, and make night into nite, thus getting rid of a large number of irregularities at one stroke. In fact, given a hundred human beings of average intellect who can read and write, it would be perfectly easy to turn out a hundred different systems of spelling, all of them more or less an improvement on the existing one.

 This was until lately the state of things - every man did what was right in his own eyes. But in the thirty years that have elapsed since Messrs. Ellis and Pitman first began to work on a phonetic alphabet, practically everything has been changed, especially within the last ten years. The labours of Messrs. Bell and Ellis have given us a thorough analysis of the sounds of English, the history of English pronunciation has been fully investigated by Mr. Ellis, and a variety of spellings have been practically tested.

 It is now possible from an examination of these various systems to deduce certain general principles, by which all reform must be guided. If there were no such principles, the problem would be a hopeless one. Nothing can be done without unanimity, and until the majority of the community are convinced of the superiority of some one system, unanimity is impossible.

 No one is qualified to give an opinion on spelling reform who has not studied these general principles, and has at least an elementary knowledge of the formation of the sounds of the English language and their relations to one another.

 The present remarks are intended to supply the necessary information in as clear and untechnical a form as possible, so as to enable the general reader to form an independent judgement without having to search through an indefinite number of scattered publications. 
 
 

THE object of an alphabet being to represent to the eye the sounds of a language by means of written symbols, it follows that in a rational alphabet -

(1) Every simple sound must have a distinct symbol, and
(2) There must be a definite relation between each sound and its symbol.

The immediate practical question of Spelling Reform resolves itself therefore into this - 
By what arrangement of the existing alphabet can the sounds of the English language be best represented?

If then we exclude new letters as impracticable, we are obliged to fall back on digraphs, which are already largely employed in English and most other languages. The obvious objection to them is that they violate the natural principle of denoting every simple sound by a simple sign. In a rational alphabet such as Visible Speech, this principle is carried out consistently, the consonants of she and the, for instance, being denoted by single letters just as that of see is. But with the Roman alphabet, which does not claim to be rational and consistent, this principle cannot be carried out: our business is to make the best use of the materials we have, and if we can make a convenient and unambiguous symbol for a simple sound by joining two letters together, we are clearly right in doing so. In fact we may consider the h in sh and th simply as a diacritic written for convenience on a line with the letter it modifies. It would be possible to write and print the h above the s and t, or to make some kind of tag, but the expense of casting new types and trouble of writing the new letters would not be repaid by any gain of ease or certainty in reading.

There is, however, one simple method of forming new letters without casting new types, which is very often convenient. This is by turning the letters, thus -  . These new letters are perfectly distinct in shape, and are easily written. The  was first employed by Schmeller to denote the obscure e-sound in the German gabe, &c. Mr. Ellis, in his "Palæotype," uses it to denote the allied English sound in but.

A great improvement would be to do away with capitals entirely. They greatly add to the difficulty of learning the alphabet, have a disfiguring and incongruous effect among the lower-case letters, and serve no useful purpose whatever. Proper names are always recognised in speech by the context, and do not require to be marked in writing either, whose exclusive function is to give a faithful representation of the sounds of language. Whenever general distinctions are required, they can be indicated by the use of a larger or smaller fount, or by thick (Clarendon) or thin type. We thus arrive at the general conclusion that a reformed alphabet must consist of the existing lower-case types, supplemented by digraphs, and, if necessary, by turned letters.

Employment of Letters.

This problem may also be stated thus - 
What values must be assigned to the letters that they may be most easily learnt, read, and written?
The obvious requisites are unambiguity and consistency, and that system which combines them in the highest degree (as far as the radical defects of the Roman alphabet will allow), while observing the practical considerations stated in the previous section, is the best.

It is clear that the defects of our present orthography are mainly due to its disregard of these fundamental principles.

Ambiguity is shown in the use of one symbol for several distinct sounds, as in man, lane, ask, salt, or of different symbols for one sound, as in why, wine, eye, lie. This fault is a violation of the fundamental principle of all rational spelling, viz. that of representing every sound by an invariable symbol (which may, however, be either a single letter or a digraph).

An alphabet is inconsistent when it fails to construct and apply its symbols on definite and uniform principles. It is, for instance, self-evident that a rational alphabet will indicate diphthongs by the juxtaposition of their elements, as in the oi of oil, which is really composed of o and i. But in English this simple principle is not carried out with the other diphthongs. In out, for instance, there is not a trace of an o, nor does its second element in the slightest degree resemble the u of but. Again, au, which would be the proper symbol of the ou in out, does not denote a diphthong at all.

The practical effect of inconsistency is not only greatly to increase the number of arbitrary symbols, but also to make their acquisition more difficult, because of the conflicting associations of ideas thus engendered.

Before going any further it will be worth while to stop and consider what are the causes of the ambiguity, inconsistency and complexity of the present English spelling. When we have a definite idea of the cause, we shall be better able to devise a cure.

Up to the sixteenth century English spelling was mainly phonetic, like the present German. At that time the words man, lane, care, father, water, were all written with the same vowel because their vowels all had the same pronunciation, viz. that of the Italian a in father. Similarly wine was written with an i because its vowel really was the long sound of the i in win, wine being pronounced as ween is now, which last, again, had a pronunciation agreeing with its spelling. However, as literature developed, and the printing-press began to assert its authority, the spelling became more and more fixed, till at last it became entirely stationary, while the pronunciation went on changing without intermission, so that the ee of ween came to be the long sound of the i in win, while wine itself changed its long vowel into a diphthong, as in the present English. The a in man, &c. changed also in various ways without any corresponding change being made in the spelling. In short we may say that our present spelling does not represent the English we actually speak, but rather the language of the sixteenth century. In other words, the present confusion in our spelling is due to the abandonment of the original Roman values of the letters, chiefly in the long vowelsª.

The only way of curing these evils is evidently to return to the original Roman values of the letters. If the beginner has once learnt to pronounce a, e, i, o, u, as in glass, bet, bit, not, full, he simply has to remember that long vowels are doubled, as in biit = "beat," and fuul = "fool," and diphthongs formed by the juxtaposition of their elements, as in boi = "boy" and hai = "high," to be able to read at once the majority of the vowel symbols. Of the consonants, whose original values have been mostly preserved, little need be said at present.

Of course, the Roman alphabet requires to be supplemented, and this is a problem that requires much thought, in order to attain the maximum of consistency and simplicity, so that the new symbols may, if possible, suggest any relationship they may bear to other known ones. Thus æ as the symbol of the a in man at once suggests a sound intermediate between the true a in father and the e in bet, which the a in man really is. Further details must be reserved till we come to the analysis of the sounds of English, for, until we know what the elementary sounds really are, it is impossible to symbolise them intelligently.

ª For a general sketch of the changes of English pronunciation and spelling, see my "History of English Sounds" (Trübner). 

Transition from and to the Present Spelling.

We have hitherto considered the question of spelling reform solely from the point of view of those who learn to read for the first time. But we have also to consider the question of the transition from and to the present orthography. The two points of view may be contrasted thus:-
(1) Which system of spelling will be easiest learnt by a child learning to read for the first time?
(2) Which will come easiest to an adult who has already learnt on the received system?

 The first of these two alternatives is, as we have seen, fully met by the simple principle of returning to the original Roman values of the letters. The second, on the other hand, requires that our new spelling should be based not on the original values of the letters but on some one of their present values. We may, then, distinguish two main classes of reformed spellings, (1) the Roman-value system, and (2) the English-value system. The only consistent and practical alphabet on the English-value system that has yet been produced is the "Glossic" of Mr. Ellis.

 Glossic is based on the principle of retaining the traditional means of expressing the sounds of English, but selecting one among the many symbols of each sound, and using it invariably to express that sound, rejecting, of course, all silent letters. Thus ee is taken as the sole representation of the sound of long i, being written not only in feel, but also in reed, skeem, = "read" and "scheme", peek ="pique", &c. ai is written not only in fail but also in naim = "name", rain = "reign", &c.

 It cannot be denied that from its own point of view this system has considerable advantages. It would certainly cause the adults of the present generation less trouble than any Roman-value spelling, for any one who has learnt to read on the present system can read Glossic at sight. Mr. Ellis also thinks that those who had learnt Glossic would easily acquire the ordinary or "Nomic" spelling, as he calls it. Before attempting to settle the relative merits of the Roman- and English-value systems, as regards ease of transition to and from the "Nomic" spelling, it will be well to weigh the following considerations.

 (1) In both systems a large number of words will retain their spellings entirely or almost unchanged. The following words, for instance, remain unchanged in both: best, bend, desk, fed, let, men; if, hit, fish, wish, in, gift; on, hot, god, dog, pot; oil, boil, loin, and many others.

 (2) Many, indeed most of the remaining words, will undergo great alterations under both systems. Let us consider, for instance, that most of our written words are practically hieroglyphs, which we recognise individually by the consonant skeletons without thinking of the sounds they represent. Thus, if we substitute a (-) for the vowels in such words as kn-ght, wr-ck, -n-gh, we still recognise them without any difficulty, which would not be materially increased even by the introduction of different vowels. Now, on any system whatever of phonetic spelling, these words, which all contain silent consonants, entirely alter the shape of their skeletons, so that whether we write nite, neit or nait, rec or rek, inuf or enf, the results are equally disguised to the eye, and can only be made out by an effort. Any possible superiority of one alphabet over another is thus very considerably reduced. To this may be added that, although in most cases where any superiority in point of resemblance to Nomic can be claimed by one system over the other, the advantage is naturally on the side of Glossic, yet the Roman-value system often has the advantage on its side. Thus the u in "full," "pull," "put," &c., and the i in "pique," "machine," "marine," &c., are preserved unchanged in the Roman-value system, while in Glossic u being used to represent the vowel in "but" cannot be retained in "full," and the i of "pique," &c. must of course be written ee.

 (3) Again, the very resemblance of Glossic to Nomic often causes very puzzling confusions. Thus "latter," "ridding," "supper," become later, riding, super, while the Nomic "later," "riding," "super(fine)," are represented by laiter, reiding, seuperfein. The Roman-value system, being more remote from Nomic, is much less liable to such cross-associations. In fact, the relation of Glossic to Nomic is very like that of two closely allied languages, such as Danish and Swedish, or Spanish and Portuguese. Although Danes and Swedes soon learn to understand one another's languages, they hardly ever, even after years of study, succeed in speaking each other's languages with real accuracy, the very nearness of the two languages, with their constant deviations from one another in matters of detail, causing constant confusion and cross-association. 

Vowels are formed by retraction of the back of the tongue, as in "father"; by advancing the front of the tongue, as in "bit"; or else they are mixed, as in "bird," in which the tongue is in a position half-way between back and front. By height they are high, as in "hit," mid, as in "hate," or low, as in "hat." The vowels of these three words are all front, but the distinctions of height apply to back and mixed vowels as well. Thus the u of "full" is high-back, just as that of "hit" is high front. All these vowels may be further modified by labialization, or rounding. Thus, if the ee of "feel" is pronounced with narrowed lip-opening, we obtain the French u in "lune" - the high-front-round. There are besides other modifications caused by the shape of the tongue itself.

 Of the large number of possible vowels only a small proportion is employed in each language.

 Again, among the special vowels of any one language we must distinguish between those differences which are distinctive, that is, to which differences of meaning correspond, and those which are not. Thus the first elements of the diphthongs in "by" and "out" vary considerably: some people sound them broad as in "father," some thin, as in "man," with various intermediate sounds. And yet the meaning of the words remains unchanged. The distinction between the vowels of "men" and "man," on the other hand, though really slighter than that of the different pronunciations of "by" and "out," is a distinctive one.

 It often happens that two sounds, though formed in different ways, have nearly the same effect on the ear. Thus the English vowel in "turn" is formed in a totally different way from the French one in "peur," the former being an unrounded, the latter a rounded vowel, and yet they are hardly distinguishable by an untrained ear. The consequence is that two such vowels are never employed together in the same language to distinguish the meanings of words, and for practical purposes they may be considered as variations of the same vowel. Hence we have to distinguish not so much between sounds as between groups of sounds. One of the most important distinctions of these groups is that of "close" and "open," the open vowels being generally formed by a "low" position of the tongue or by some other widening of the mouth passage.

 Disregarding special exceptions in individual languages, we may assume the following as the chief distinctive groups in language generally:
 
 

Of these groups the mixed (3) is, as remarked above, almost identical in sound with the close and open front (11, 12), with which latter the dull-back (1) is often identified, although in sound it is really intermediate between them and the clear-back (2). In practice, therefore, the symbols for 11 and 12 will also suffice for 1 and 3.

 a, i and u, at once supply symbols for 2, 4 and 7 respectively. For 10 we have only to restore y to its original Roman value, which it still retains in Danish and Swedish. If we assign e to the close-front (5) and o to the close-back-round (8), in accordance with the general European tradition, we must find letters for the corresponding open vowels. For the open-front (6) æ at once suggests itself, the a indicating openness. For the open o (9) there is no type ready to hand; I propose therefore to adopt the turned  used by Mr. Ellis in his Palæotype. This letter, which is really a turned c, is meant to suggest a turned o, which is impracticable. For the rounded e (11) the turned '  may be used, and for the open sound (12). We thus obtain the perfectly parallel forms i, e, æ, and y,, . The last two at the same time supply symbols for the special English u in "but" (1) and "turn" (3).

 Diphthongs are, of course, symbolised by the juxtaposition of their elements. The following are the English diphthongs:-
 
 

Diphthongs in all languages vary greatly in their constituents, and the above combinations must be understood as simply denoting general tendencies. Thus ai does not literally imply a combination of the a in "father" and the i in "bit," but merely a movement in that direction. We may start, not with a full-back vowel, but with a mixed one, which may move towards i, but without reaching it: in fact the commonest pronunciation of "aisle" may be represented by el. In the same way ei only implies a front vowel moving upwards, and, as a matter of fact, the starting-point may be either a close or open e or even the a of "man." Indeed ei often begins with a mixed vowel, in which case "veil" is confounded with "vile."

 Note that ei and ou in English supply the place of close long ee and oo, which most English people are unable to pronounce.

 ii and uu are often diphthongised in a peculiar way in English, by being made to end in the consonants y and w respectively, wiin (ween) and fuul (fool) becoming wiyn and fuwl.

 Having thus laid a general foundation, we may proceed to discuss some special modifications required in English.

 As there is no short or close e or o in English, it is superfluous yo use æ and  to denote the quality of sounds whose openness is always implied by their shortness. We can, therefore, discard  altogether in English, and employ æ to denote the peculiar a in "man," for which it would otherwise be difficult to find an appropriate letter.

 The longs of æ and  may be expressed, as with the other vowels, by doubling - ææ,. But as this is inconvenient, and as  is not used in English, it is better to denote the long of æ by ae, the separation of the letters implying length. Long  may, on this analogy, be denoted by ao.

 ª The vowel in "turn" is open-mixed, that in "gabe" close-mixed. 

The consonant r in English only occurs before a vowel, either in the same or the next word, as in "erring" (eriq), "far off" (faar aof). When not followed by a vowel, that is, either by a pause or a consonant, it is weakened into  - the er of "father." After aa and  the  is absorbed, as in "bar" (baa), "farther" (faadh), "her" (hoe), "heard" (hoed), the first two being indistinguishable from "baa" and "father."  is sometimes dropped after ao, especially before a consonant, as in "floor," "floored," although the full flao, flaod are most usual in careful speech, especially when the  is final. After other vowels  is preserved throughout, also when the r is sounded as a full consonant: compare "air" (ae), "aired" (aed), and "airy" (aeri) with "far off" (faar aof), "her own" (hoer oun), and "flooring" (flaoriq).

 The following table will give a general idea of these changes:-
 
 

faar aof (far off)	faa	faadh (farther).
hoer oun (her own)	hoe	hoed (heard).
fiiriq (fearing)	fii	fiid (feared).
aeriq (airing)	ae	aed (aired).
muuriq (mooring)	muu	muud (moored).
flaoriq (flooring)	flao	flaod (floored).
fairiq (firing)	fai	faid (fired).
flauri (flowery, floury)	flau	flaud (flowered).
leiriq (layering)	lei	leid (layered).
louriq (lowering)	lou	loud (lowered).

Note that ei(r) and ou(r) in rapid, especially in vulgar speech, often pass into ae(r) and ao(r).

 When r is preceded by a short vowel, as in "hurry" (hri), "merry" (meri), no  is generated. 

The two chief unaccented vowels in English are  and i, together with the rarer o. The former may be regarded as a shortened oe, as in "her," into which it always passes when emphasised or prolonged, but it is really nothing but a voice murmur without any definite configuration. The i is an intermediate vowel between i and e, and might as well be written e as i. It may be regarded either as a very open i or a very close e.

 The following are examples of :-

temt (attempt), pouz (oppose),pon (upon), tdei (to-day).
souf (sofa), menshn (mention), peishns (patience), kært (carrot).
faadh (father), on (honour), mezh (measure).
faowd (forward), shepd (shepherd).
feivrit (favourite), mezhriq (measuring).

is often dropped before l, n and m; always when the  is preceded by t or d and followed by l or n:-

 metl (metal), gaadn (garden), gaadniq (gardening), mtn (mutton).
iivl (evil), loukl (local), simbl (cymbal, symbol).

 When two or more unaccented s or is follow one another, one of them is often thrown out, as in -

 hist()ri (history), feiv()rit (favourite), vedzh{/i}tbl (vegetable).

 i is less common than . It is most usual as a weakening of front vowels, especially when i or y is written:-

 piti (pity), mndi (Monday).
divaid (divide), ditekt (detect).
ræbit (rabbit), fishiz (fishes), biliti (ability).

 It is the regular unaccented vowel before dzh, even when a is written:-

 vilidzh (village), kæridzh (carriage), kolidzh (college).

 In rapid speech i is apt to pass into , except when final.

 Unaccented o in ordinary speech is simply  rounded. When dwelt on it becomes ou. Examples are -

 pteito (potato), folo (follow), felo (fellow).

 In rapid speech this o passes into .

 These vowels occur also in unaccented monosyllables. Compare "a man" ( mæn) with "against" (genst), "to go" (t gou) with "to-day" (t dei), "for all" (fr aol) with "forgive" (fgiv), "of course" (v kaos) with "offence" (fens).

 the and to have two distinct unaccented forms. Before consonants they both have , while before vowels they assume the fuller forms dhi and tu:-

 dh mæn (the man), dhi enmi (the enemy).
t gou (to go), tu ent (to enter). It was, I believe, first noticed by Mr. Ellis that "that" as a demonstrative is always full dhæt, while as a conjunction and relative pronoun it becomes dht:- ai nou dht dhæt dht dhæt mæn sez iz truu (I know that that that that man says is true). 

List of English Symbols.

The following table gives a complete list of the English vowel symbols in the "Romic" system I propose, together with those
consonant ones which require elucidation, with examples.

aa: papa, far, glass, after, aunt. [Before s and f or before two (pronounced) consonants aa is sometimes shortened, and
sometimes becomes æ: glæs, ænt.]
æ: man.
ae: aerate, bear, fare. [Always followed by .]
ai: Isaiah, aisle, wine.
ao: extraordinary, broad, more.
au: Faust, now, noun.
e: red.
ei: they, veil, name.
i: ill, fishes.
ii, iy: machine, feel.
o: not, cloth, cross, soft. [Often becomes ao before th, s, and f: klaoth, kraos, saoft.]
oi: boy, boil.
ou: flow, soul, stone.
u: full, put, good.
uu, uw: truth, rue, fool.
  : up, come; father, here.
oe: her, turn, heard.

c: church, catch.
dh: then, with.
j: judge, gentle.
q: sing, finger.
sh: fish.
th: think.
x: six, wrecks.
y: young.
zh: rouge, pleasure.

As regards the use of the letters there can be no question about the values of the following:- b, d, f, g, h, k, l, m, n, p, r, s, t, v, w, z.

 This leaves c, j, q, x undisposed of. We also have y, which is not required as a vowel-symbol in English. If we allow y to retain its present value, we can also retain j as a convenient abbreviation of dzh. For tsh we have ch, which, by the omission of the superfluous h, can be reduced simply to c. We thus have c and j perfectly parallel. q may very well be taken to represent the back nasal ng, as Mr. Ellis has done in his Palæotype. x lastly, if employed at all, must in consistency be extended to all kss in the language, not only in such words as six, but also in rex (wrecks), cex (cheques), &c.

 These contractions fully counterbalance the necessity of retaining the digraphs th and sh, to which must of course be added dh and zh. wh is very generally made into w in Southern English, but it is well to keep up the distinction on the chance of its being afterwards revived. The breath yh (= German "ich") sometimes occurs in such words as "hue" (yhuu), more commonly, however, pronounced hyuu, with a separate h before the y.

 Consonants are often dropped in English. Thus the h of the personal pronouns is generally dropped when they come after a verb, and are unaccented, as in ai sao im (I saw him). Saw her and soar are both pronounced sao. The d of and is generally dropped before a consonant, as in ct n cmgen (cut and come again), where the vowel is dropped also on account of the t and n (p. 187 above).

 Assimilations also occur in rapid speech. Thus, many people who pronounce the q of "going," &c. quite distinctly in most cases, regularly change the back into the point nasal (n), when it is followed by a point consonant (t, d, n), as in gouin t ... (going to ...). In I can't go the t is generally dropped, and the point nasal is often assimilated to the g by being made into the back nasal q - ai kaaq gou. 

The chief accent or stress in each word may be marked by (·) following the letter on which the accent begins:- genst (against), fg·iv (forgive). To indicate the secondary accent, when necessary, (:) may be used:- i:nknv·iinyns (inconvenience), dist:rktb·iliti (destructibility). These very convenient marks were introduced by Mr. Ellis. In practice the accent need only be marked when it is on some other than the first syllable. Thus it need not be marked in foutogræf (photograph).

 Unaccented vowels are always shortened. Thus hii in hii gouz (he goes) is much shorter than in it iz hii (it is he), but its vowel is quite distinct from the regular short i in "hit." As this shortening is always implied by the want of accent, it need not be marked: hi gouz would imply that the i was pronounced as in "hit."

 Emphasis, or the accent of a word in a sentence as distinguished from that of a syllable in a word, is marked by a (·) before the word. Such subordinate monosyllables as "he," "she," "it," "and," "if," "to," "for," &c., are assumed to be unaccented unless they are marked. We thus distinguish between hii gouz and it iz ·hii, between hii hæz mai buk (he has my book) and it iz ·mai buk not ·hiz (it is my book, not his). Principal words, such as nouns, non-auxiliary verbs and adjectives, which regularly receive a full accent, may be marked in the same way whenever they are made exceptionally emphatic, thus ai ·fg·iv yu indicates that the second syllable of fg·iv is uttered with extra emphasis. 

Nomic.	Glossic.	Romic.	Spanglish
man	man	mæn	mæn
lane	lain	lein	lein
hare	hair	haer	her
ask	aask	aask	ask
wall	waul	waol	wol
salt	solt	solt	solt
nail	nail	neil	neil
Nile	Niel	nail	nail
air	air	aer	er, æer
saw	sau	sao	sao, s'o
ten	ten	ten	ten
he	hee	hii	hi
where	whair	whaer	wer
stern	stern	stoen	st'ern
see	see	sii	si
sea	see	sii	si
bear	bair	baer	ber
earth	erth	oeth	'erth
head	hed	hed	hed
break	braik	breik	breik
veil	vail	veil	veil
key	kee	kii	ki
eye	ei	ai	ei
few	feu	fyuu	fiu
grew	groo	gruu	gru
in	in	in	in
myth	mith	mith	mith
wine	wein	wain	wein
first	ferst	foest	ferst
pique	peek	piik	pik
thief	theef	thiif	thif
lie	lei	lai	lai
on	on	on	on
hole	hoal	houl	hoal
none	nun	nn	nan
more	moar	maor	mor
word	werd	woed	werd
no	noa	nou	no
do	doo	duu	du
soon	soon	suun	sun
good	guod	gud	g'ud
blood	blud	bld	bl'ad
door	doar	daor	dor
oath	oath	outh	oath
oar	oar	aor	or
woe	woa	wou	wo
shoe	shoo	shuu	shu
oil	oil	oil	oil
boy	boi	boi	boy, boi
boil	boil	boil	boil
out	out	aut	aut, out
now	nou	nau	nau
soul	soal	soul	soal
you	yoo	yuu	yu
four	foar	faor	for
up	up	p	'ap
tune	teun	tyuun	tun, tu'n
rule	rool	ruul	rul
burst	berst	boest	b'erst
full	fuol	ful	f'ul

Nomic.	Glossic.	Romic.	Spanglish
cat	kat	kæt	kaet
cease	sees	siis	sis
ocean	oashen	oushn	oacen
chin	chin	cin	chin
scheme	skeem	skiim	skim
get	get	get	get
George	Joarj	jaoj	jorj
ghost	goast	goust	goast
laugh	laaf	laaf	laf
through	throo	thruu	thru
see	see	sii	si
as	az	æz	æz
sugar	shuoger	shugr	sh'ug'ar
thick	thik	thik	th'ik
this	dhis	dhis	this
Thames	Temz	temz	temz
vex	veks	vex	vex, veks
example	egzaampl	egzaampl	egzampl
Xerxes	Zerkseez	zoexiiz	zerziz

The results of a detailed study of this table may be conveniently, though somewhat roughly, summed up in the following lists, in which, however, only the commonest groups are given, each represented by its typical word:-
 
 

We see that out of a total of thirty-three typical words more than a half either remain unchanged or else undergo equally violent changes under any possible scheme of reform. Also that only a third of the whole thirty-three remain unchanged in Glossic, from which the two in parentheses ought, strictly speaking, to be excluded, as their agreement with Nomic is obtained at a great sacrifice of phonetic consistency.

One very important result of a return to the Roman values of the letters would be the restoration of the original harmony of the English with the Continental values of the letters, which would much facilitate the acquisition of English by foreigners, and vice versa. At present, English people and foreigners have to learn each other's languages almost entirely by eye, unless thoroughly taught by a native, and consequently are utterly at a loss when brought face to face with the spoken language - in fact, they have to learn the same language twice over. Thus when a German sees the English written word right he easily associates it with his own recht, as also the English name with the German name, but when he hears the genuine rait and neim, he is thrown completely off the scent. Conversely, when an Englishman comes across the German knie for the first time, he at once thinks of his own knee, and naturally drops the k in the German word as well as in the English: if he were used to see the English word spelt nii, he would never think of dropping the k in German.

 It will, of course, be urged by the advocates of historical spelling that the silent letters in right and knee are really valuable helps in acquiring the language. All this really amounts to is, that sixteenth century English bears a much closer resemblance to German than nineteenth-century English does, consequently that a German will learn the former more easily than the latter, and that an Englishman who knows sixteenth-century English will thereby learn German more easily. The practical result is, of course, that English has to be learnt twice over both by the English themselves and by foreigners. The worst of it is, that instead of learning the older stage of our language on an intelligent and systematic plan, we have it forced on us - whether we really want it or not - in the shape of a garbled and imperfect orthography, which, instead of giving us clear ideas of the language of the period it represents, only serves to hopelessly confuse our notions of our present language.

 Of course the orthographies of most of the Continental languages require reform as well as English; French, especially, most urgently demands a thorough change. Indeed, there is no reason why foreigners should not learn French on a phonetic system, leaving the present French spelling to be acquired afterwards, even if the French themselves do not inaugurate a reform.

 There are many significant facts in the pronunciation and spelling of English which show that the return to the Roman values of the vowels would not be by any means so violent a change as is generally supposed. Even without going beyond the commonest words in our vocabulary we have whole classes of words like machine, marine, oblique, antique, &c., in which long i retains its Roman value. In geographical names, such as Alabama, Chicago, Granada, Medina, Messina, the accented vowels all have the Roman values. In such names as Isaiah, Achaia, Cairo, the diphthong also has its strict analytical value. Indeed, the tendency is becoming stronger and stronger to retain as much as possible the native pronunciation of foreign names. The definite adoption of the Romic principle by the Indian government, and the reformed pronunciation of Latin, are all most important moves in the same direction. 

One very important result of a return to the Roman values of the letters would be the restoration of the original harmony of the English with the Continental values of the letters, which would much facilitate the acquisition of English by foreigners, and vice versa. At present, English people and foreigners have to learn each other's languages almost entirely by eye, unless thoroughly taught by a native, and consequently are utterly at a loss when brought face to face with the spoken language - in fact, they have to learn the same language twice over. Thus when a German sees the English written word right he easily associates it with his own recht, as also the English name with the German name, but when he hears the genuine rait and neim, he is thrown completely off the scent. Conversely, when an Englishman comes across the German knie for the first time, he at once thinks of his own knee, and naturally drops the k in the German word as well as in the English: if he were used to see the English word spelt nii, he would never think of dropping the k in German.

 It will, of course, be urged by the advocates of historical spelling that the silent letters in right and knee are really valuable helps in acquiring the language. All this really amounts to is, that sixteenth century English bears a much closer resemblance to German than nineteenth-century English does, consequently that a German will learn the former more easily than the latter, and that an Englishman who knows sixteenth-century English will thereby learn German more easily. The practical result is, of course, that English has to be learnt twice over both by the English themselves and by foreigners. The worst of it is, that instead of learning the older stage of our language on an intelligent and systematic plan, we have it forced on us - whether we really want it or not - in the shape of a garbled and imperfect orthography, which, instead of giving us clear ideas of the language of the period it represents, only serves to hopelessly confuse our notions of our present language.

 Of course the orthographies of most of the Continental languages require reform as well as English; French, especially, most urgently demands a thorough change. Indeed, there is no reason why foreigners should not learn French on a phonetic system, leaving the present French spelling to be acquired afterwards, even if the French themselves do not inaugurate a reform.

 There are many significant facts in the pronunciation and spelling of English which show that the return to the Roman values of the vowels would not be by any means so violent a change as is generally supposed. Even without going beyond the commonest words in our vocabulary we have whole classes of words like machine, marine, oblique, antique, &c., in which long i retains its Roman value. In geographical names, such as Alabama, Chicago, Granada, Medina, Messina, the accented vowels all have the Roman values. In such names as Isaiah, Achaia, Cairo, the diphthong also has its strict analytical value. Indeed, the tendency is becoming stronger and stronger to retain as much as possible the native pronunciation of foreign names. The definite adoption of the Romic principle by the Indian government, and the reformed pronunciation of Latin, are all most important moves in the same direction. 
 
 

Consonant symbols       b [b] big, baby, mob ch [tS] chip, church, pitch d [d] did, dog, pinned dh [] the, then, with f [f] five, photo, tough g [g] good, giggle, big h [h] hit, have, behold j [dZ] job, ginger, bridge k [k] kill, kick, cat l [l] live, lolly, pool m [m] mad, mammoth, him n [n] nod, pain, reign ng [] sing, bringing, finger p [p] pit, pepper, shepherd r [r] rim, bury, hearing s [s] sip, missing, cross sh [S] ship, machine, motion t [t] tip, putting, heat th [] thin, wrath, pithy v [v] very, live, vivid w [w] win, which, bewitch y [j] young, yes, beyond z [z] zebra, rose, dizzy zh [Z] pleasure, vision, rouge

Vowel symbols       ä [æ] man, scatter, ant a [a:] aadvark, barn, past ai [ei] aim, lane, reign e [e] chair, pear, fare au [o:] automatic, lord, bought e [e] bed, head, said ee [i:] seed, meal, believe i [i] beer, fear, here ei [ai] height, light, sky i [i] bit, reform, luggage o [o] pot, wash, cough oa [u] boat, soul, though oi [oi] boil, toy, employed oo [u:] soon, truth, shoe ou [au] sound, clown, plough ü [] cut, other, blood u [u] put, could, hood oe [:] burn, stir, learn u [u] moor, tour, cure [] father, bottom, tenant

This is a system without caps. Altho caps are used for Z (zh) and S (sh). This is a compromise notation without a dh/th distinction which uses u for up and oo for /u:/.

dispyoot wüns roaz bitween dh wind nd dh sün, wich wz dh stronggr v dh too, nd dhai greed t put dh point on dhis ishoo, dht wichev soonist maid  trävl taik of hiz kloak, shud bee kountid dh mau pouful. dh wind bigän, nd bloo widh aul hiz meit nd main blast, koald nd fisz  thraishn staum; bt dh strongg hee bloo dh kloas dh trävl räpt hiz kloak round him, nd dh teit hee graspt it widh hiz händz. dhen broak out dh sün: widh hiz welkm beemz hee dispoest dh vaipr nd dh koald; dh trävl felt dh jeenyl waumth, nd z dh sün shon breitr nd breit, hee sät doun, oavküm widh dh heet, nd kast hiz kloak on dh ground. 

(Spanglish) A disput wans aroaz bitwin the wind and the sun, which was the stronger ov the tu, and they agrid tu put the point on this ishu, that whichever sunist meid the travler teik of hiz cloak, shud bi acounted the mor paurful.  The wind bigan, and blu with ol hiz mait and meiner blast, coald and fiers as a threishen storm; but the stronger hi blu, the cloasr the travler rapt hiz cloak araund him, and the taiter hi graspt it with hiz handz.  Then broak aut the sun: with hiz welcam bimz hi dispoazd the vapor and the coald; the travler flet the jinyal warmth and az the sun shoan braiter and braiter, hi sat daun, overcam with the hit, and kast hiz cloak on the graund.

B. Contracted. Simplified

(c = tsh, j = dzh, q = ng. a for aa, the length being implied. The combinations ii, &c. are regarded as diphthongs and written simply i, the length of the first element being implied, as in ei; on the same principle ae and ao are simplified into e and o.)

It must, however, be remembered that these results are not to be attained by the adoption of any system indifferently that may be proposed. What is wanted is a simple, consistent, and above all elastic spelling, which, within certain practical limits, will adapt itself to every change of pronunciation. Changes of pronunciation cannot be controlled by any spelling based on the Nomic values of the letters. There is, for instance, no reason why oo should represent the sound of long u any more than that of long i, nor consequently why the uu of "boon" should not change through byyn (with the French u) into biin without any change of spelling being thought necessary, and consequently without any control of such possible changes being exercised. 

In short, historical spelling destroys the materials on which alone history itself can be based. This is the case in the English of the last few centuries. The word "name," as its spelling indicates, was in Chaucer's time pronounced naam'a, or something like it. It is now neim, although still written "name." 

Although new types should be avoided at first, their exclusion is only a practical consideration, not a matter of principle, and there is no reason why they should not afterwards be introduced by degrees. Thus Mr. Pitman's  is unquestionably superior to q as a symbol of the back nasal ng, for its shape at once associates it with the other nasals n, m. Again the Greek  and  (or perhaps better the Anglo-Saxon ð) would do very well for th and dh, both being easily written. The long s and the tailed z of Pitman's Phonotypy are also excellent letters for sh and zh. We should thus avoid the ambiguity of such words as pothuk (pothook), which at present can only be avoided by writing pot-hook. 
 
 
 

Index http://yi.com/home/ChandlerJames/spelref.html