•m iP- !ru ... Joint Committee on a phonetic English alphabet Report of a Joint ^Commit I representing the National Educational Association PE 1135 J64 £A^ REPORT OF A JOINT COMM ITlfiEK % Representing the National Educational association, the American Philological Association, «a>d';the Modern Language Association of America , d, p, b, m, f, v, w. DiphtJiofigs: ai, au, oi. Arranged as nearly as possible in the familiar abc order, the letters and their names are as indicated in the table below. The name of a vowel is its phonetic value as uttered separately, and of course with stress; the always unstressed [9] being called "weak a" to distinguish it from the stressed [u]. The name of an explosive consonant, and also of [h], [y] and [w], is its pho- netic value followed by [i] ; of any other consonant, its phonetic value preceded by [e]. etter Name Key-word Letter Name Key- word a art 0 north a artistic 0 august' a air 6 note a at 0 poetic a ask P Pi pit b bi be r er rat d di do s es set e mate 1 el ship e met t ti ten f ef fee P ej) thin g gi go d ed that 9 en sing u mood h hi he u push i marine 0 urge i tin u hut j = d3 ji jaw 0 weak d about c = tj ci chew V ev vat k ki kin w wi win 1 el let y y* yes m em met z ez zest n en net 3 e3 azure 1 8 REPORT OF A JOINT COMMITTEE It will be observed that the alphabet consists of thirty-four let- ters, which number is increased to forty-two by the use of the cir- cumflex— the only diacritic mark required. The circumflex over a vowel-sign denotes primarily length, but in some cases also a concomitant closeness or roundness. This point is of impor- tance for the proper understanding of the notation. Between [e] and [e], [i] and [i] , [o] and [6], [u] and [u], there is a differ- ence of quality as well as quantity. Were we proposing a notation of maximum precision, it would be important to use a quantity- mark which should be nothing else ; but in that case we should need, for example, four signs for the i-sounds : one for the open short in pit, a second for the open long heard sometimes in fear, serious, a third for the close long in mari?te, and a fourth for the close short in react. A similar need would arise in connection with the e-sounds, the o-sounds and the u-sounds. But, as has been seen, the alphabet here described does not aim at maximum precision ; and in a notation where simplicity, economy, and read- ableness are very urgent considerations, the open long i, heard in serious, as pronounced by many with an approximation to Sirius, may very well be merged with the close long [i] heard in seen ; while the close short in react may be merged with the open short in pit. In this way we reduce the four i-sounds to two and are enabled to denote the two by means of familiar and instantly recognizable signs. It should be borne in mind, then, that each letter, and espe- cially each vowel-letter, denotes not a single invariable sound, but a type of sound, or a certain range of quite similar sounds. The pronunciation of each letter is indicated by means of key-words, — a rough method but the only one available. The difficulty of finding good key-words arises from the fact that the words chosen may be pronounced differently in different parts of the English- speaking world, or by different persons in the same part. For example, if an American from one of the Northern States wishes key-words for the shortened a of palm, he thinks perhaps of what, yacht, watch, squad, quality, and the like ; or he may select examples from the very numerous words written with so-called " short o," as hot, not, odd, cob, box, copper. But both these classes of words are uttered by Englishmen and by some Americans with a slightly rounded vowel (like that heard usually in august'), and this rounded vowel, which we denote by [e], is regarded as "standard" ON A PHONETIC ENGLISH ALrilABET 1 9 English. The dictionaries tell us to pronounce what and quality as [hwet] and [kwoliti]. Thus altho wkat and not, as pronounced by the majority of Americans, furnish good illustrations of the shortened a of palm, we can not use them as key-words without running the risk of misleading certain readers and seeming to put forward our own "dialect" as "standard English." To give another illustration of this difficulty. When a key- word is needed for long 0, it is natural for an American to take such words as boat, note, gold. But an Englishman, at any rate a Lon- doner, pronounces these words with a diphthong, which the English phoneticians write variously as on, ow, or o11. Shall we then say that this diphthongal utterance is normal English, or that it is London dialect ? It is not for the committee to decide such questions, since our concern is with the facts, and not with theories or ideals of pronunciation. On the other hand, we must take note of such important and characteristic variations (whatever we may regard as "correct"), and provide a notation adequate to represent them. These are examples of the difficulty that constantly arises in the choice of key-words. It should be understood, therefore, that the words given are often only an approximation ; but whenever it be- comes necessary to use one that is particularly ambiguous, it is preceded by an asterisk referring the reader to explanatory mat- ter in the ensuing notes. As for the arrangement of the letters, we begin with [a] and proceed with the palatal or front vowels of the series ending in [i] ; then we return to the " low-back " position and proceed with the vowels of the labial or rounded series ending in [u]. In enume- rating the consonants we begin with those produced farthest back in the mouth and proceed outward, considering first the explosive consonant of each contact-region. This seemed, on the whole, the most convenient of the many arrangements possible. It is presumed that the letters will generally be used in the perpendicular or " Roman " form, but they can be cut as italics if it is desired.1 As we use Roman forms in this report and must 1 As some of the needed types were not to be had on the market in the size and style required, the committee ordered a number of matrices made specially for this report. Great care was taken with the drawings, but not- withstanding that fact certain letters are not exactly what they should be. The general intention is, however, clear in every case ; and the reader is requested to judge by that and not to assume that the exact forms, as they appear in the text, are put forward as unimprovable finalities of typography. 20 REPORT OF A JOINT COMMITTEE use them in connection with other print, we have adopted the ex- pedient of bracketing the letters of the Revised Phonetic Alphabet and all words or groups of words respelled in them. Italics are employed for words cited in the usual spelling. The letters are numbered that they may be referred to conveniently, but the diph- thongs, being combinations of other letters, are not given num- bers of their own. They are discussed in sub-sections of the num- bers relating to their first element. The letters are given only in lower-case Roman type. It was not thought necessary, for the purposes of this report, to go to the expense of having a large number of new matrices made for capitals and script letters. These can easily be provided when the need arises. i. a art, calm, palm, father, ardor, retard, aha, mama ; * ask, *aunt, *path. In American English [a] is heard chiefly before r, but Lon- doners and some Americans pronounce it also before /, s, th, nt, ns, nch, in a number of words such as aunt, ask, path, basket, mas- ter, pass, past, last, haunt, vaunt, taunt, lance, France, launch, staunch, bath, after, quaff, lattgh. Most Americans pronounce [a], not [a], in the majority of these words, but [e] in certain book- words such as taunt, vaunt, haunt. As to the use of the " tran- sition" [a], see No. 5. The rounded Scottish [a] — "fawther" for father— -may best be written [e] . 2. a *aha, *mdma; * what, * watch, * yacht, * squad, * qual- ity ; * not, *hot, * cob, *fop, * rock, * copper. The sound meant is the shortened [a] of palm. When un- stressed it tends toward [a] ; aha, mama, becoming [aha, mama] . The [a] is heard in these words only when they are uttered care- fully with the same quality of vowel in both syllables. In British speech [a] is a rare sound, but in the Northern United States it is very common, being written a (mostly after a labial) in several words represented by what, watch, etc., and ON A rilONETIC ENGLISH ALPHABET 21 as so-called "short o" in a much larger number, like not, hot, etc. The sound [a] is also heard very frequently, in the speech of those who always pronounce r (see No. 2j), in words containing [r] followed by a voiceless consonant ; thus [a] in barter, marker, while barber, larger, have [a]. 2\ ai by, die, dye, eye, height, fine, high, sign, find, 2S.sk. This very common diphthong consists of [a] and [i] uttered in one syllable. In aye, as spoken deliberately in voting, the [a] can be more plainly heard than in eye. The combination [ai] is normal, but the first element varies a good deal in the dialects. The Cockney makes the diphthong [ui] or [ai], while the Irishman converts it into [ei], his fine time becoming [fein teim]. 2\ au now, how, mouse, fiour, sauerkraut. Here again there is some dialectic variation of the first ele- ment. A familiar American (chiefly Southern) pronunciation of house is [haus] or [haus], while the Yankee drawl makes it [haus] . 3. a air, heir, care, bare, bear, /air, /air, flare, pear, there. In standard English [a] occurs only before r, but many Ameri- cans pronounce it before other voiced consonants, particularly in monosyllabic words, such as bad, cab, have, man, lamb. See No. 4. In its origin [a] is usually either a prolongation of [a], as in pair, from Latin/wrthru Old French paire, and pare, from French pare?-, or else it has resulted from the diphthongization of [e], as in pear, Anglo-Saxon pern. The old diphthongal quality of the ea in pear, bear, ivcar, etc., was lost long ago, but the close affin- ity of [a] and [e] is seen in the varying quality of [a]. According to Storm, an excellent observer, the English (i.e. London) vowel in air, eare, is between the [a] of sat and the [e] of set. He de- 22 REPORT OF A JOINT COMMITTEE notes it by the Greek e. In American English [a] and [a] vary little except in quantity. 4. a a/, back, batch, bang, cash, manner, matter, tap, arrow, fallow ; * add, * bag, * cab, * man, * lamb * has ; * abhor, * accept, * Atlantic. The first group of words with asterisk represent a class (mostly monosyllabic words ending in a voiced consonant) in which the [a] is converted into [a] by many American speakers. This lengthening does not belong to universal English, but is part of the so-called " American drawl." The second group of key-words with asterisk represent a very numerous class in which [a] occurs in an unstressed initial syllable. In such case the vowel retains its proper sound only when the word is uttered deliberately ; in the flow of connected talk it be- comes [9] (see No. 18). Two expedients are open to the dictionary- maker in dealing with such words : ( 1 ) either to adopt a simple diacritic mark for obscuration (such as a dot over or under the letter); or better (2) to repeat for the colloquial or "familiar" pronunciation [aksept, oksept]. The discrimination of [a] from [a] in the manner here recom- mended was proposed in the phonetic alphabet of 1877. From the international point of view it would be preferable, perhaps, to interchange the signs, and use [a] for the vowel of art, [a] for that of at ; the reason being that [a] is the sign generally associated in other languages with the vowel of art. But the point is not very important, and there are two countervailing arguments. (1) The usage recommended has already gained some currency, having been adopted by the Standard Dictionary and, virtually, by the International Phonetic Association. (2) It leaves us the familiar print-form [a] for the a-sound of most frequent occurrence in stressed syllables, and devotes the new sign [a] to a sound that oc- curs much less frequently. OX A PHONETIC ENGLISH ALPHABET 23 5. 6 * as£, * ?/ass, * au///, * IdMnch, * France. * bath. The sound is between [a] and [a]. It occurs before a voice less fricative, that is, before/, s, th; less often before a nasal fol- lowed by a consonant ; rarely in other situations. For additional examples see No. i. The committee were at first divided as to the need of a letter for this sound, some being disposed to regard it as a negligible va- riation of [a] or [a]. The facts are briefly as follows. During the seventeenth century the old short [a] and [e] became long in cer- tain situations, so that [plant], [pas], [pa]}], [laf], became [plant], [pas], [pa)3], [laf], just as the older [klojD], [soft], [les], became [kle)3], [soft], [los]. Late in the eighteenth century this new [a] became [6.] in the speech of London, tho conservatives resist- ed the innovation and clung to [a]. In time the feeling arose that [pas] was vulgar, and [pas] affected; so that refined speakers made a virtue of pronouncing an intermediate sound. We quote from Smart's edition of Walker (1838): " But Walker is a bigot ; he allows of no compromise between the broad a with which a vulgar mouth pronounces ass [a], and the sound, narrower if possible, than the a in at [a], with which an affected speaker minces the same word. Surely in a case like this there can be no harm in avoiding the censure of both parties by shunning the extreme which offends the taste of each." Since Smart's time the [a] has largely prevailed in the usage of Londoners, tho his "mincing sound" [a], which has all the claims of antiquity in its favor, is also heard. But the "refined transition," as Sweet calls it, that is [a], is likewise in use. Good observers have noticed that English ladies tend to. say [ask], [glas], while the men say [ask], [glas]. In the United States the refined transition has gained large accep- tance, especially in deliberate speech. Many Americans would certainly object to the necessity of choosing between [mastor] and [mastar], [Frans] and [Frans], [ba|D] and [bo]}], and would say that neither well represents their habitual, deliberate pronuncia- tion. The great dictionaries all recognize the refined transition and provide a notation for it. 24 REPORT OF A JOINT COMMITTEE All this being so, it seemed best to the committee to include a letter for this sound in the Revised Alphabet, altho the sound has no real existence in the ordinary natural speech of the majority of English speakers. The character chosen [a] is an easy modi- fication of [a] and has already some currency. The quantity of the sound is regularly long everywhere, what- ever its quality may be. Its short would be hardly distinguish- able from [a]. After the old short sound had become long the dictionaries continued to mark the vowel of ask, path, etc., as short, because its quality suggested that of the vowel in cat, and the notions of quantity had become mixed up with those of quality. It is time for our dictionaries to recognize that this bothersome sound is at any rate regularly long, however we may shade its quality. 6. 6 day, fate, great, tain, rein, reign, veil, they, taste, way, weigh, savior, danger, chamber, angel. This is the close long e-sound. A very precise notation would require a different sign for the open long, which is heard only before r ; for example in Mary, as often pronounced with the vowel like that of merry, only long. But in an alphabet of medium precision this variety may be merged with [e]. In fact many pro- nounce a real [e] in such words as Mary, vary, chary, without approximation to the vowel of merry, very, cherry. So too a sep- arate sign would be needed for the close short, which is heard only before a vowel in certain book-words such as chaotic, aeriah aorta. But the close short occurs only in situations where its closeness will take care of itself, even if we write [keetik], [eirial], [eerto]. 7. e ebb, led, dead, said, left, men, meant, shepherd, merry, bury, never, bless, leather, measure, bell, wealth ; ^except, ^example, ^enclose, *embody, *esteem. This is the open short e-sound. In an initial unstressed syl- lable (the words with asterisk) it usually becomes an obscure [i] ON A PHONETIC ENGLISH ALPHABET 25 or [&], tho it retains the distincter sound in more deliberate utter- ance. Thus one would read: [eksept yi bl born agen] ; but one might say: [ol do boiz iksept mi]. In such case, as in that of [a], the dictionary has either to adopt a diacritic mark for obscura- tion (such as a dot over or under the vowel affected), or else to repeat for the less formal pronunciation; thus [eksept, iksept]. See above, page 13, and below, page 45. 7'. ei *<^ay, *fatc, great, *ra\n, etc., as in No. 6. In the speech of Southern England, notably London, a stressed [e] is regularly converted into a diphthong. Beginning with the position for open [e], the tongue is raised during the act of utterance toward the position of [i]. To denote this diphthongal sound Sweet uses ei, the Oxford Dictionary e*. The small superior letter has the advantage of marking the second vowel as subordi- nate, but for the convenience of the printer we prefer [ei]— a sign already familiar in such words as veil, deign, skein, weigh. The sign [ei] will not be needed for the transcription of world-English but is necessary to mark the usage of a very important minority. The Cockney tends to make the first element still more open than [e], so that his lady becomes [laidi]. 8. I police, manne, lean, lien, seen, scene, Peter, believe, re- ceive, people, Caesar; year, *serious. This is the close long i-sound. The open long, as heard in /car, serious, when pronounced with the vowel similar in qual- ity to that of lyric, Sirius, is also to be written [I]. The close short, heard only before a stressed vowel in such words as reac- tion, theistic, Aeolian, is to be written [i] . Like the close short e - und, it occurs only in situations where its closeness is a mat- ter of course. 26 REPORT OF A JOINT COMMITTEE 9- 1 bit, bitter, city, kid, kill, build, biscuit, hymn, business, lyric; *added, *foxes, ^goodness, *princess, *messa.ge, ^pref^ce, *prefate ; *eject, *evade, *refer, *debate, react, beatitude ; *pififul, *vanity, *difficult. The open short i-sound. In the first group of words with asterisk — they represent a numerous class — we have a sound which is between [e] and [i], but nearer to [i]. Many speakers make no distinction between the two vowels of business, pronouncing it [bizniz]. Palace is made to rime with Alice, smartest with artist, basket with ask it, prelate with sell it, and the last syllable of beverage is not distinguished from the last syllable of Cole- ridge. A very precise notation would require a sign for an un- stressed vowel between [e] and [i] ; in the absence of it, [i] comes nearest to what is desired. See above, page 13, and be- low, page 45. In the second group with asterisk the starting-point is [i], but owing to the lack of stress the vowel is shortened. At the same time it becomes more open if a consonant follows, as in debate, but retains its closeness before a vowel, as in react. It is thus a close approximation to the truth to write [ijekt], [ived], [rifur], [dibet], [riakt]. In the third class of words with asterisk the medial [i] retains its quality in careful or deliberate utterance, but in ordinary talk tends to become [9]. 91. iu, iu tube, new, few, feud, Tuesday, lute, mule, pure, Pui'itan, puristic, mulatto. This diphthong consists of [u], or [u], preceded by a more or less audible [i]. Initially and generally after a consonant, the vowel [i] is normally displaced by the consonant [y]. As the discrimination of [iu] and [iu] from [yu] and [yu] is difficult and of no great practical importance, it is recommended that [iu] and [iu] be dispensed with and that [yu] and [yu] be written in their place. In the pronunciation of many, especially in the United ON A PHONETIC ENGLISH ALPHABET 2J States, the [u] or [u] is somewhat unrounded thru diminished lip-action, and the [i] or [y] is often omitted before the dentals t, d, u, I See No. 14. 10. 0 awe, awful, /aw, a//, halter, stalk, haul, lord, form, remorse, dbg% off, clotJi. This convenient and necessary letter was recommended in the phonetic alphabet of 1877. It is an o with the breve dropped into the circle. In the common spelling it is often written a, as in all, call, but is to be classed as an o-sound because its production is attended by a certain rounding of the cheeks or lips. 11. o audacious, autumnal ; *hot, *rob, *rocfc, *copper; *what, *watcJi, *szuap, ^quantity. The short [0] is best heard unstressed in such words as au- tumnal, autocracy. The key-words with asterisk are equivocal because, in the greater part of the United States, they are pro- nounced with a completely unrounded vowel usually identical with [a]. See No. 1. 111. oi oil, boil, point, boy, Troy, oyster, enjoy, enjoin . That the first element of this very common diphthong is [0] and not [o], may be perceived by comparing coin and coin-cide. 12. 6 oh, owe, go, boat, note, loan, bone, flown, snow, hoe, roe, though, bestow, toll, behold, mould; *four, *^oar, *port, * story, * glory. This is the close long o-sound. The open long is heard in the words with asterisk, as pronounced by most Englishmen and many Americans. They are words in which the vowel, with primary or secondary stress, is followed by [r]. This "opening" of [6] makes it hardly distinguishable from [0] ; so that four and mounting become virtually identical with for and momitig, and 28 REPORT OF A JOINT COMMITTEE soar with saw 'er. Many Americans, however, distinguish these pairs as [for] and [for], [mornirj] and [mernirj], [sor] and [ser]. Strictly speaking the open sound is between [6] and [e], so that a very precise notation would require three signs for the three vowels of boat, soar, and saw 'er. The open long [6]i is here dis- regarded; it may be written [6] or [©], as may seem best. 13. o poetic, cooperate, bohemian, window, borrow, furlough, donation, rotation ; ^potato, apolitical, *obey, * solicit. The difference between [o] and [0] is that the former is more rounded ; note the action of the lips in passing from the initial syl- lable of co-operate to that of causation. In standard English [o] does not occur except in unstressed or weakly stressed syllables. The words with asterisk indicate a class in which [o] tends to become [a] in ordinary talk. It remains close only before a vowel, as in poetic. 13'. ou *oh, *owe, *go, *boat, note, etc., as in No. 12. In the speech of Southern England, notably London, a stressed [6] is regularly converted into a diphthong. Beginning with the position for open [o], the lips are rounded during the act of utterance to the position for [u]. To denote this diphthongal sound the English phoneticians use on, ow, and ou. On the whole [ou] seemed preferable. Its use is already familiar in such words as soul, mould, dough, though, pour, etc. Like [ei], the sign [ou] will be needed only in transcribing the speech of Southern England. 14. u ^•00^, roof, booth, moon, choose, lose, pool, mood, 00^; *poor, *moor, *boor; *rule, *rude, * 'rheumatism, *true, *fm\t, *blue, *bloom. This is the close long u-sound. Before r, as in poor, moor, it is slightly more open, being the long of No. 15, but the varia- tion does not require a distinct sign. ON A rilONETIC ENGLISH ALPHABET 29 The second group with asterisk — [u] after r or / — present a difficulty. In this class of words many Americans pronounce an [u] that is slightly unrounded thru diminished lip-action. In this pronunciation rule does not rime exactly with school, nor fruit with boot, nor rude with rood, nor blue with do, nor is the first part of rheumatism exactly identical with room. Moreover this slightly unrounded [u] is heard also very commonly in the class of words in which the dictionaries generally prescribe [yu] (No. 91), especially those in which the [y] is reduced nearly or quite to the vanishing-point. The result is that the untrained British ear hears an American say Noo York, Toosday, and confound loot with lute. In the pronunciation referred to, however, the differ- ence depends not so much upon the presence or absence of the faint [y] in [y u] as upon the quality of the [u] . The committee differed as to the need of a letter for this par tially unrounded American [u]. There is as much need of it, perhaps, as for [a] ; many Americans will say, no doubt, that their natural pronunciation of music, feud, Tuesday is not well repre- sented by [myuzik], [fyud], [tyuzdi], and certainly not by [muzik], [fud], [tuzdi]. A modification of the letter [u], consisting of a minute curve or loop in the left-hand branch, somewhat as in [a], was considered at length, but it was finally decided not to in- clude it in the Revised Phonetic Alphabet. 15. u push, pull, fool, good, hood, nook, stood, could, zvould. This is the open short u-sound. The corresponding close short, which is also to be written [u], may be heard in whoever, Louise. It occurs only unstressed and before a vowel. 16. 0 urn, urge, her, fur, sir, earn, fern, bum, bird, herd, heard, murder, girl, pearl, whirl, word, churl, dirge, merge, splurge, myrrh. This vowel, which is so variously represented in the traditional spelling, is more nearly the long of [a] than of [u]. It occurs only before rand only in stressed syllables. Unstressed it reduces to [o], as in murmur. Were the circumflex, as used in the Revised 30 REPORT OF A JOINT COMMITTEE Phonetic Alphabet, a sign of quantity only, the better sign for this sound would be [a] with the circumflex. But the use of that sign, while theoretically preferable, would result in giving to every word containing the sound an entirely strange aspect. On the other hand, the use of [0], notwithstandingN the fact that [u] and [0] differ in quality, as do [e] and [e], or [i] and [i], leaves many words with a very "natural" appearance; as [Orn], [bOrn], [murdar]. On the whole the vowel is associated in the general mind rather with u than with any of the other letters — e, i, ea, o, y — which are used to represent it. \y. u up, but, cub, must, puff, tough, enough, fun, much, touch, crutch, love, come, some, above, rustle, hubbub, puzzle, dull, gull, multiply ; *hurry, ^courage, ^flourish. This vowel occurs only in stressed syllables. Unstressed it becomes [&]. Before [r] its quality varies between that of [u] in up and that of [0] in urn ; but the variation may be disregarded. 18. Q sofa., idea,, equa.1, sepa.mte, embassy, ma.ture, facility, bal- loon, celery, cover, moment, ebony, doctor, guttura.1, na- tion, nat\ona.l, lonesome, glorious, a man, the man; *enemy, *hom\ny, * 'effort, ^record. This is the famous " obscure " vowel. Its tongue-position is nearly the same as that of [u] — it is really between that of [u] and that of [a] — but [u] is always stressed, [o] never. This gives them a difference of quality easy to be heard if one compares, say, cup and hiccup. The obscure [o] is the goal to which the most of the other vowels tend when not supported by the stress. It is thus of very frequent occurrence. The alphabet of 1 877 made no provision for this vowel, leaving it to be identified, in any particular case, with that one of the other vowels which it most resembled, or from which it had de- rived; thus: sofa; dialect; moment; neshun; ballot; gamut; in dhi weter. By this means a text printed in the reformed spell- ing was made to look rather more like ordinary print than would ON A PHONETIC ENGLISH ALPHABET 3 I otherwise have been possible. But it is not phonetic writing to use five or six different signs for the same sound. There is really no better reason for doing so in the case of this vowel — which has a definite character of its own — than in the case of any other. That the use of a separate sign for it leads to the spelling of the same word in different ways is not an objection, but a point in its favor; since it teaches observation of and fidelity to the facts. The word //^pronounced [di], is a rarity; in actual use it is either [di], as in the ocean, or [da], as in the man. In the sentence / know that that is so, we have two thats, [dat] and [dat], which must be distinguished if one makes a pretence of phonetic writing. In some of the key-words given above many speakers avoid the obscuration; saying, for example, [rekerd], [efort], [mom- ent], [enimi], and not [efart], [rekard], [m6mant], [enami]. Let it be remembered that we are not here inculcating views or set- ting up a standard, but providing a notation for the facts of English speech. The tendency to weaken an unstressed vowel is such a fact, and a very important one. The sign generally employed by phoneticians is an inverted e, either italic or Roman. But in accordance with our general prin- ciple of avoiding turned letters we have preferred the form [a], an easy modification of [a]. It looks very like the a of the phone- ticians, while in a good many words, such as [separet], [daialekt], [normal], [sofa], it has a not unwelcome suggestion of [a], with which it is closely related. 19. h hat, hen, his, hole, aha. The combination wh in what, white, when, etc., is to be writ- ten [hw] in accordance with the earlier and more correct practice. 20. k Veep, kit, kick, call, cat, clock, flicker , tobacco, quit, box. The choice here is between k and c, which originally had the k-sound only, but in English has the sound of s or sh in many words. Each letter has its advocates whenever the question is 32 REPORT OF A JOINT COMMITTEE discussed, and however we decide it there will very certainly be many who will wish that it had been decided the other way. The argument in favor of [k] is that it is never ambiguous and is the sign now almost universally employed by phoneticians to repre- sent the sound under consideration. If we use [k] there is no danger that kit, kite, kettle, curtain, cane will ever be confused by any one with sit, site, settle, certain, sane. There is a slight difference of quality between the back k of call and the front k of kettle. The front or palatal k has de- veloped into ch in many words ; compare cool and chilly, speak and speech. But while the difference between the guttural and the palatal k is historically of some importance, it is disregarded in the Revised Phonetic Alphabet. 21. g gap, gate, dog, again, aghast, ghost, dagger, blackguard. As in the case of [k], the front or palatal modification of g, heard in geese as compared with goose, is disregarded. 22. n sing, young, longing, singer, finger, banker, anchor. The sign here proposed for ng is the one now most generally employed by phoneticians. In banker, anchor, we have a combina- tion of [rj] and [k] ; in finger, longer, a combination of [rj] and [g] . 23. y yes, yet, ewe, union, bilious, collier, minion, beauty, pure, cure. The choice here is between y and j. The argument for y is that we readers of English are used to it. The argument for/ is that it is the letter generally employed in phonetic literature for the sound under consideration — the so-called palatal semivowel— j/ being used for the German ii, the French u. It may also be said that j is intrinsically the better sign for the sound, because the form shows its close relation to i, of which at first it was only a ON A PHONETIC ENGLISH ALPHABET 33 calligraphic modification. The sound of [y] is that of an i made so high and close as to cause a faint consonantal rustle. On the other hand, the objection toy is that it might easily be mistaken t or they of jaw. Thus the choice between y and j for the semivowel turned finally upon the question of retaining/ as a possible notation for the two consonants, or the composite consonant, heard in jaw. After much discussion the committee decided to recommend the retention of [j] as an alternative to [d3] for the composite conso- nant ; and thus it seemed to follow as a matter of course that a form of y must be used for the semivowel. To distinguish this consonantal [y] from the y employed by phoneticians for German ii, French u, it is suggested that for the former the familiar italic or script letter cut as a perpendicular type might be used. 24. t tin, bit, bitter', thyme, phthisic, ptomaine. 25. d do, sad, ladder^, dry, land. 26. 1 let, live, bell, elbow, shallow. Like [n] and [r], [1] may perform the function of a vowel and form a syllable. The endings spelled -al -el -le 41, are pronounced [al] with the [9] more or less distinctly heard. It is most clearly audible after a vowel, as in Baal, towel, fealty. After a consonant, as in battle, nibble, devil, shovel, the vocal murmur reduces to almost zero, leaving syllabic [1]. Shall we then write [batal], [nibal], [deval], [Juval], or simply [batl], [nibl], [devl], [fuvl] ? By the latter spelling we lose the distinction heard in [batl] as compared with [batlirj]. In general it will be better to write the [a], even if it is very faint. 27. r red, rat, trick, arrow, stretch ; *far, "father, * bur den. The proper denotation of the r-sounds is a matter of some diffi- 34 REPORT OF A JOINT COMMITTEE culty. Before a vowel, as in red, trick, very, the printed r is pro- nounced, and pronounced almost uniformly by speakers of English everywhere ; the tip of the tongue being raised far enough to pro- duce an audible consonantal sound, but not made to vibrate. It is this sound which we propose to denote by the familiar print- form [r] . If the tip of the tongue is raised high and made to vi- brate, we have the trilled r, heard in Scottish dialect and in forced utterance, for example sometimes in elocutionary efforts. This trilled r might be denoted when necessary by an italic. But the untrilled [r] may be weak or strong according as the tongue-tip is raised less or more. In the speech of Southern England, to some extent also in that of the eastern and southern United States, an [r] not followed by a vowel, as in far, farm, farthing, other, other day, is not pronounced at all, or is replaced by o. In this pronunciation no distinction is made between father and farther, laud and lord, sought and sort, boa and bore. Other day is heard as [udo de], but other end as [udar end]. For this dropt [r] of course no sign is needed. Between the dropt [r] and the trilled r there are various grades depending on the vigor with which the tip of the tongue is brought into play. It would be useless to attempt to distinguish them graphically. The sign [r] may be used to denote any r-sound that is audible as a consonant, however weak it may be. A still different variety of r-sound, the so-called "Northum- brian bur," is not produced with the tip of the tongue at all, but be- tween the back part of the tongue and the uvula. It may be de- noted when necessary by r. This " velar r " is a very common sound in French and German. A preceding [o] does not disappear before r to the same ex- tent as before / and n. Such words as butter, never, gunner, will be best written with [or] ; [butor], [never], [gunor], rather than [butr], [nevr], [gunr]. 28. n no, ten, manner, sign, kmfe, inch, engine. Syllabic [n], as in fasten, written, b?itton, boatswain, will be best written [on] ; [fason], [riton], [buton], [boson], rather than [fasn], [ritn], [butn], [bosn]. ON A PHONETIC ENGLISH ALPHABET 35 29. S sit, house, this, missing, cent, peace, scene, psychology. 30. z zest, lazy, buzz, his, was, houses, scissors, pansy, palsy. 31. f s\\ip, dis\\, issue, passion, nation, function, gracious. The letter here proposed for the sh-sound, the once familiar "long s," is already in very general use among phoneticians. See No. 33. 32. 3 azure, seizure vision, leisure , pleasure . The only noteworthy objection to this letter, which is also in common use, is its close resemblance to [z] in the script form. This difficulty can be obviated by giving to the upper loop of the script [3] a flat top. 33- e=tf chip, chest, choose, church, rich, bach, e^tch, voucher. The ch-sound presents a perplexing problem. The familiar sign ch is objectionable because it is hopelessly ambiguous, and because, whatever else the sound may be, it is certainly not a com- bination of c and h. The majority of the best phoneticians analyze it as a combination of [t] and [J], an analysis which is accepted by the Oxford Dictionary. Following that we should write the word church as [tjurtj], while Judge Churchill would appear as [D3ud3 .TfOrtfil] . This looks odd, but much in phonetic writing must necessarily look odd at first. If the mere oddity were the only objection it would be our plain duty to accept the situation and make the best of it. But the difficulty is that our ch-sound is not a combination of ordinary t, such as is heard in tin, with ordinary sh, such as is heard in skip, — no such obvious combina- tion as we have, for example of k and s in fox. It must be re- membered, however, that each of our consonants varies more or less under the influence of neighboring sounds. The / of ten is 36 REPORT OF A JOINT COMMITTEE not like the t of written, nor is the sh of ship like the sh of nuU shell. What we have in our ch is really a variety of / combining with a variety of sh. But the elements do not combine in se- quence ; they blend almost simultaneously, giving rise to a com- posite sound, in which most people can hear neither a / nor an sh. Storm calls the sound a combination of a peculiar palatal t and the German ch of ich ; or a cross between tsh and ty. When a highly expert phonetician must resort to such language in de- scribing the sound, the layman can hardly be blamed for assert- ing— with a good deal of vehemence, as he often does — that ch is not a combination of / and sh. It may be remarked in this connection that the sound does not often develop out of t+sh, but usually from a palatalized k ; com- pare bench and speech with bank and speak. At the same time it is not likely that the trigraph tch, as in watch, would have gained acceptance so easily had it not been felt that a / is actually heard. One may say, then, that the sound is a duality, but a duality in which the parts lose something of their separate charac- ter and form a new unity. To many people the ch in chin will always seem no less simple and indivisible than the sh in ship. The dilemma is, then, that a notation in accord with the best scientific authority will offend the eye of many English speak- ers and seem to them finicky and absurd. This being so, it seemed best to treat this sound as a peculiar one-in-two, which might properly have a sign of its own as an alternative spelling to the preferable [tf]. Of several forms discussed a c with an oblique cross appeared the least open to objection. 34- j=d3 ]aw, jet, jelly, gin, gist, congeal, judge, ridge, pigeon. The case for [j] = [d^], which is simply the voiced [c] = [tf], is implicitly disposed of under No. 33. 35- \> thin, thick, thro7igh, bath, breath, ether, Luther. The choice here is between the Anglo-Saxon or Runic thorn []f\, the Greek 0, and a ligature of / and h — i.e. h with the top of a t. The committee were divided in their preference, but voted in ON A niONETIC ENGLISH ALPHABET 37 favor of the old thorn, which is the letter now most commonly used by phoneticians to denote the sound in question. 36. d this, thus, that, with, loathe, breathe, mthcr, either. For this sound, the voiced [J>], a crossed d seemed preferable to the Greek d, with or without cross. 37- P pit, tip, upper, spend, plead, pulp. 38. b hat, tuh, huhhle, blue, member, cupboard. 39. m mat, Java, lamb, summer, elm, phlegm. 40. f fat, ii, stuff, buffer, laugh, phj/sic. 41. v vex, veal, love, river, oi, Stephen. 42. w was, way, wit, woe, word, once, quest, persuade. For wh in what, see No. 19. IV Specimens (0 While the Revised Phonetic Alphabet is designed primarily to be used in the respelling of words for pronunciation, rather than in the printing of literary matter, it was thought desirable to give a few examples of its applicability to the graphic representation of connected discourse. The specimens are of two classes : first, dig- nified prose and noble poetry, in which one pronounces more care- fully ; second, familiar talk, in which one gives a freer rein to the natural tendencies of the language. It must be remembered that 38 REPORT OF A JOINT COMMITTEE the object in view is not to teach orthoepy, but to illustrate the use of a notation. Numbers (i) and (2) are transliterated in accordance with the approved dictionary pronunciation as modi- fied by varying conditions of stress. Number (3) gives a sonnet of Keats as read by an Englishman — a graduate of Oxford — whose utterance may be taken as fairly representing that of edu- cated Londoners. Number (30) gives the same matter as read by a New Englander— a member of the Harvard Faculty. Number (3#) gives the same matter as read by a member of the committee who is a Michigan man some time resident in New York. Num- ber (4) is transcribed as accurately as possible from Sweet's primer of spoken English. Number (4a) gives an average American version of the same matter. Number (5) is Yankee dialect, transcribed from the Biglow Papers, with a few emenda- tions. Number (6) is Central New York dialect from " David Harum." 1 Co hwail do rivaizd fonetik alfebet iz dizaind praimerili to bi yuzd in do rispelirj ov wurdz for pronunsiejen, rador dan in do printirj ov litereri mator, it woz J^ot dizairobol to giv o fyu igzampolz av its aplikobiliti to do grafik reprizentejen ov kanektid diskors. do spesimenz or ov tu klasiz : furst, dignifaid proz end nobol poitri, in hwic wun pronaunsiz mor karfeli ; sekand, fomilyor tek, in hwie wun givz 0 frior ren to do nacorol tendansiz ov do langwij. it must bi rimembord dot di objikt in vyu iz not to tie ©rjDoepi, but tu ilustret di yus ov 0 note Jon. numborz (1) ond (2) or trans- literetid in ekordens wid di epruvd dikjoneri pronunsiejen az modifaid bai veriin kendijenz ov stres. number (3) givz 0 sonit ov kits oz red bai on inglijrnen — 0 gradyuet ov oksford — huz utorons me bi teken oz farli reprizentin dat ov edyuketid lundonorz. number (3^) givz do sem mator az red bai 0 nyu inglender, — 0 membor ov do harvard fakalti. numbar (3^) givz da sem matar az red bai a membar av da kamiti hu iz a mijagan man sum taim rezident in nyu yerk. numbar (4) is transkraibd az akyuritli az posibal from swits primor ov spokon irjglij. number (4a) givz on avorij emereken vurjen ov do sem mater, number (5) iz yanki daielekt, transkraibd from do biglo peperz, wid e fyu imen- dejenz. number (6) iz sentrel nyu york daielekt, from "devid harom." 1 On the absence of capitals see above, page 20. The reader is by no means to infer that the committee disapprove of capitals, or think lightly of their use in print. ON A PHONETIC ENGLISH ALPHABET 39 (2) From Lincoln's Gettysburg Address. Four score and seven years ago our fathers brought forth on this continent a new nation, conceived in liberty and dedicated to the proposition that all men are created equal. Now we are en- gaged in a great civil war, testing whether that nation, so con- ceived and so dedicated, can long endure. We are met on a great battle-field of that war. We have come to dedicate a portion of that field as a final resting-place for those who gave their lives that that nation might live. It is altogether fitting and proper that we should do this. (2) for skor and sevan yirz ago aur fadarz brot forJD on dis konti- nent a nyu nejan, kansivd in libarti and dediketid tu da propazijan dat ©1 men ar krietid ikwal. nau wi ar ingejd in a gret sivil wer, testirj hwedar dat nejan, so kansivd and so dediketid, kan Ion endyur. wi ar met on a gret batal-fild av dat wer. wi hav kum tu dediket a porfan av dat fild az a fainal restirj-ples for doz hu hir gev der laivz dat dat nejan mait liv. it iz eltugedar fitirj and propar dat wi Jud du dis. (3) A Sonnet of Keats. Much have I traveled in the realms of gold, And many goodly states and kingdoms seen; Round many western islands have I been, Which bards in fealty to Apollo hold. Oft of one wide expanse had I been told That deep-browed Homer ruled as his demesne ; Yet did I never breathe its pure serene Till I heard Chapman speak out loud and bold : Then felt I like some watcher of the skies When a new planet swims into his ken ; Or like stout Cortez when with eagle eyes He stared at the Pacific — and all his men Looked at each other in a wild surmise, Silent upon a peak in Darien. 40 REPORT OF A JOINT COMMITTEE (3) mue hav ai travald in da relmz av gould, and meni gudli steits and kirjdamz sin ; raund meni westan ailandz hav ai bin, hwie badz in fialti tu apelo hould. eft av wun waid ekspans had ai bin tould dat dip-braud houma ruld az hiz dimin ; yet did ai neva brid its pyua sirin til ai hud eapman spik aut laud and bould : den felt ai laik sum weear av da skaiz hwen a nyu planit swimz intu hiz ken ; 0 laik staut ketez hwen wid igal aiz hi stad at da pasifik — and el hiz men lukt at ie uda wid a waild samaiz, sailant apan a pik in darien. (3«) mue hav ai travald in da relmz av gold, and meni gudli stets and kindamz sin ; raund meni westan ailandz hav ai bin, hwie badz in fialti tu apelo hold, eft av wun waid ikspans had ai bin told dat dip-braud homa ruld az hiz dimin ; yet did ai neva brid its pyua sirin til ai hod eapman spik aut laud and bold : den felt ai laik sum weear av da skaiz hwen a nyu planit swimz intu hiz ken ; e laik staut ketez hwen wid igal aiz hi stad at da pasifik — and el hiz men lukt at ie uda wid a waild samaiz, sailant apan a pik in derien. (3*) mue hav ai travald in da relmz av gold, and meni gudli stets and kirjdamz sin ; raund meni western ailandz hav ai bin, hwie bardz in fialti tu apelo hold. eft av wun waid ikspans had ai bin told dat dip-braud homar ruld az hiz dimin ; yet did ai nevar brid its pyur sirin til ai hurd eapman spik aut laud and bold den felt ai laik sum weear av da skaiz ON A THONETIC ENGLISH ALPHABET 4 1 hwen 6 nvii planit swimz intu hiz ken; or laik staut kertez, hwen wid fgel aiz hi stard ot do posifik — ond ©1 hiz men lukt ot ie udor wid o vvaild sormaiz, sailont opon o pik in derien. (4) Colloquial English of London. How do you do ? Quite well, thank you ; how are yon, ? How are they all at home ? All well except the baby. What is the matter with the baby ? I do not know exactly. She was crying all night long ; there must be something wrong with her. Poor thing ! I am sorry to hear it. You do not look well yourself. No, I am ill from want of sleep. What is the French for " I do not understand " ? I want to let this Frenchman know I can not understand what he is saying. It is rather odd, I can talk French myself, but I can not under- stand it when it is spoken. You should tell them not to speak so fast. I do not believe they can speak slow ; they are too excitable. What do you think of young Mortimer ? I think he is the most conceited young fool I ever saw ; but there is no great harm in him. At any rate he is a gentleman; he would not do any- thing mean or dishonorable. I am afraid that can not be said of his elder brother. Oh, he is a regular cad; you know he was turned out of his club for cheating at cards. (4) hau dyu du ? kwait wel, )xirjk yu ; hau o yu ? hau o dei el ot houm ? el wel, iksept do beibi. hwets do mato wid do beibi ? ai dount nou igzaktli. Ji woz kraiirj el nait leg ; do must bi sum- })in rerj wid o. puo J3in ! aim seri to hir it. yu dount luk wel yuself. nou, aim il from went ov slip. hwets do frene for ai dount undostand ? ai went to let dis frencmon nou ai kdnt undostand hwet iz seiirj. its rador ed, ai kon tek frene maiself, bot ai kant undostand it hwen its spoukon. yu Jud tel dom net to spik sou fast, ai dount biliv dei kan spik slou; de o tu iksaitobol. hwot do \\u fcirjk ov uun metimo ? ai fcink iz do moust kon- sitid nun ful ai evo se; bot doz nou greit ham in im. ot eni reit hiz o jentolmon; hlwudnt du eni}Dirj min odizenrobol. aim ofreid dat kant bi sed ov iz eldo brudo. ou, hiz o regyulo kad; yu nou i woz tund aut ov iz klub fo citin ot kadz. 42 REPORT OF A JOINT COMMITTEE (4*) hau da ua du (hau di du) ? kwait wel, ]Darjk yu; hau ar yu? hau ar de §1 at horn ? ©1 wel, iksept da bebi. hwats da matar wid da bebi ? ai dont no igzaktli. Ji waz kraiirj el nait lerj ; dar must bi sumJDin reng wid ar. pur [Dirj ! aim sari ta hir it. yu dont luk wel yarself . no, aim il fram want av slip. hwats da frene far ai dont understand ? ai want ta let dis frencman no ai kant understand hwat iz seing. its radar ad, ai kan tek frene maself, bat ai kant understand it hwen its spokan. uu Jud tel dam nat ta spik so fast, ai dont biliv de kan spik slo; de ar tu iksaitabal. hwat da uu jDirjk av uun mertimar ? ai f)irjk iz da most kan- sitid yun ful ai evar se; bat darz no gret harm in im. at eni ret hiz a jentalman; hi wudnt du eni}}in min ar disanarabal. aim af red dat kant bi sed av iz eldar brudar. 6, hiz a regyular kad ; yu no i waz turnd aut av iz klub far eitirj at kardz. (5) From the Biglow Papers. But wen it comes to beiri killed, — I tell ye I felt streaked The fust time 't ever I found out wy bagonets wuz peaked ; Here's how it wuz : I started out to go to a fandango, The sentinel he ups and sez, " Thet's furder 'an you can go." " None o' your sarse/' sez I ; sez he, " Stan back ! " Aint you a buster ? " Sez I, " I'm up to all thet air, I guess I've ben to muster; I know wy sentinels are sot ; yo aint agoin' to eat us ; Caleb haint no monopoly to court the seenoreetas ; My folks tu hum air full ez good ez his'n be by golly ! " An' so ez I wuz goin' by, not thinkin' wut would folly, The everlastin' cus he stuck his one-pronged pitchfork in me An' made a hole right thru my close ez ef I wuz an in'my. (5) but wen it kumz ta biin kild, — ai tel yi ai felt strikid da fust taim tevar ai faund aut wai baganets waz pikid ; hirz hau it wuz : ai startid aut ta go tu a fandango, da sentanal hi ups n sez, " dets fOrdar an yu kan go." "nun a yar sdrs," sez ai; sez i, "stan bak! " "ent yu a bustar? " sez ai, " aim up tu el det ar, ai ges aiv ben ta mustar ; ai no wai sentanalz ar sat ; yi ent agoin ta it as ; kelib hent no monapali ta kort da sinaritaz ; ON A PHONETIC ENGLISH ALrHABET 43 mcii f6ks ta hum ar ful az gud az hizn bi, bai gali ! " an so, bz ai waz goin bai, nat ]3inkin wut wud fali, di evdrlastin kus i stuk iz wun-prerjgd pieferk in mi on mctl 8 hoi rait j^ru ma cloz, az ef ai waz an in'mi. (6) From "David Harum" "I did smell woolen some," said David, "but I had the hoss an' they had the money, an', as fur's I c'd see, the critter was all right. Howsomever, I says to 'em: ' This here's all right, fur's its gone, but you've talked putty strong 'bout this hoss. I don't know who you fellers be, but I c'n find out,' I says. Then the fust feller that done the talkin' 'bout the hoss put in an' says, 1 The' hain't ben one word said to you about this hoss that wa'n't gospel truth — not one word.' An' when I come to think on't afterward," said David with a half-laugh, "it mebbe wa'n't gospel truth, but it was good enough jury truth. I guess this aint over 'n' above int'restin' to ye, is it ? " he asked after a pause, looking doubtfully at his sister. (6) "ai did smel wulin sum," sed devid, "bat ai had da hes an de had da muni, an, az furz ai kad si, da kritar waz ol rait, hausam- evar, ai sez tu am : ' dis hirz ol rait, furz its gen, bat yuv tekt pati strerj baut dis hes. ai dont no hu yu felarz bi, bat ai kn faind aut,' ai sez. den da fust felar dat dun da tekin baut da hes put in an sez : ' de hent ben wun wOrd sed ta yu baut dis hes that want gespal tru|3 — nat wun wurd.' an hwen ai kum ta }Mnk ant afterward," sed devid wid a haf-laf, "it mebi want gespal tru)D, bat it waz gud anuf juri tru)}. ai ges dis ent ovor n abuv intrastin tu yi, iz it ? " hi askt Mtar a pez, lukirj dautfali at hiz sistar. V Proposal of a Shorter Alphabet (for ordinary phonetic writing and practical spelling reform) Some of the distinctions which must be made by any good dic- tionary in the respelling of words for pronunciation, and are accord- ingly provided for in the Revised Alphabet, may be safely ignored 44 REPORT OF A JOINT COMMITTEE in rapid phonetic writing. See above, page' 16. In general it is not necessary to mark a vowel as long if the sound differs only in quantity from that of the corresponding unmarked vowel, nor if the long sound occurs only before [r]. In the latter case the [r] itself becomes a sufficient sign of length. We may merge [a] in [a], and [a] in [a], without serious loss, omitting the circumflex since the long sounds occur only under restricted conditions such as preclude the short sound. No confusion can arise between [kam], calm, and [kam], cam. We may also dispense with [a], writing the words that contain it either with [a], or with [a], as may be preferred ; while [Cr] may be discarded as occurring only before [r]. On the other hand, we can not do without [e i, e, u], because the result would be an identical spelling of numerous pairs of words such as met — mate, sin — seen, not — naught, pull — pool. Can we do without [6] ? The natural answer would seem to be yes, seeing that the short sound occurs but rarely and then only in unstressed syllables, where its quantity would take care of itself. On the other hand the dropping of [6] would make it easy to confuse not and note, since [o] and [e] are apt to resemble each other in script. It will therefore be best to keep [6]. Some might wish to dispense with [o], but that is a very convenient letter. Dropping the signs that can be most easily dispensed with in rapid writing, we have the following thirty-eight letters for the Shorter Alphabet : a, a, b, d, e, e, f, g, rj, h, i, i, j, c, k, 1, m, n, e, ©, 6, o, p, r, s, }, t, }>, d, u, u, u, 9, v, w, y, z, 3. VI Suggestion of a Differentiated Alphabet (For purposes requiring very great precision) While it is believed that the Revised Phonetic Alphabet will prove equal to all ordinary demands, it falls short of the precision which will occasionally be required. Cherishing the hope that the notation described in Section III might gradually become the stand- ard for popular-scientific purposes, the committee were reluctant ON A PHONETIC ENGLISH ALPHABET 45 to include any new letters that were not incontestably necessary. But they found it difficult to draw the line. Certain letters were discussed and finally rejected, of which there is as much need, perhaps, as there is of the letter [a], which was included. Take, for example, the high unstressed vowel between [e] and [i], which is heard in added, honest. It is not ideally represented by [i], tho [adid], [enist] are better than [aded] [onest]. The maker of a large dictionary might very properly choose to use a distinct let- ter, say a modification of the small capital 1, for that sound. A similar remark might be made of the partially unrounded American ;/, heard long in music and short in mutation. Again, it was pointed out in Section III that in the case of most of the vowels we have to do with an open and a close sound, each one of which may be long or short. Of the four varieties the Revised Alpha- bet provides signs for two only ; but a dictionary aiming at maxi- mum precision must in some way denote all four. This can be done, without the use of confusing diacritic marks, by employing systematically for the open sound a slightly dipt or otherwise in- complete form of the fundamental letter. As for the consonants, the letters of the Revised Alphabet will suffice for all needs that are likely to arise in connection with the spelling of standard English, but there are certain dialectal and foreign sounds for which provision should be made. And here again the use of slightly modified forms of the familiar Roman letters is to be recommended. As this report is not directly concerned with international phonetics, there was no occasion to have types cast for these differentiated letters, but a member of the committee undertook to prepare tables which should illustrate what is here meant, and at the same time indicate the line along which a satisfactory solu- tion of the problem of an international phonetic notation is most likely to be reached. These tables are presented, with explana- tory matter, in Appendix II. Herewith the committee bring their report to an end and submit it to the public of scholars for criticism and amendment. They are aware of its imperfections, but they have done their best with a difficult and perplexing problem, and they hope that the result of their labors is at least good enough to arouse the interest of the competent in making it better. There were dif- ferences of opinion on various points, but none on any matters of 46 REPORT OF A JOINT COMMITTEE fundamental importance. The signatures below indicate an ap- proval of the report as a whole, but not necessarily an approval of everything in it. The chairman, as final redactor and proof- reader, requests that he, and not the other members of the com- mittee, be held responsible for any errors that may be discovered. Calvin Thomas George Hempl Charles P. G. Scott O. F. Emerson E. O. Vaile Committee Rwi sed All abet iioh a a £ Verry ma a Sheridan \no Walker 1 751 a fis-is- a Websterli?z Swiss ch ; 8 the uvular r of France and Germany, the Northumbrian bur; 9 an untrilled variety of 8, common in Germany; 10 g in good ; 11 c in cook ; 12 the open North-Germany in Tage ; 13 ch in German ach ; 14 Russian / mpalka ; 15 ng in long ; 16 g in gig ; 17 k in kick ; 18 y in ye ; 19 German ch in ich ; 20 Italian^/, Spanish //, etc.; 21 Italian gn. Nos. 22 and 23 are palatal stops made even farther forward than 16 and 17, and with a more fricative quality, so that they suggest the English/ in jet and ch in chip. Nos. 24 and 25 are the palatal z and s of Polish, etc. ; 26 the z in azure, French/ inje and g in gentil ; 27 sh in she, s in sure, French ch ; 28 English d made at the alveolar process (see 35) ; 29 English alveolar / (see 36) ; 30, 3 1, the usual z and s ; 32 r as made by trilling the tip of the tongue ; 33 an untrilled variety of 32, the usual r in red, very, etc.; 34 n in nun. Nos. 35 and 36 are the truly dental t and doi Sanskrit, French, etc., in making which the tip of the tongue touches the base of the upper teeth, or the adjacent gums; 37 and 38 English th in this APPENDICES 53 and tJiin ; 39 the usual / (there are varieties which may be easily differentiated); 40, 41 the usual b and / ; 42 South German wt Spanish b in saber ; 43 German w in zwei, schwarz, etc. ; 44 the usual m ; 45 English w in war ; 46 the same sound voiceless, a common pronunciation of wh in what ; 47 French it in bias ; 49 French » inputs ; 49, 50 English v and/. G. H. Guttural Lingua] Labial Velar Palatal Dental Bl- Velo- Palato- Denti- labial labial labial I 9 GJ cU b 10 16 22 28 35 40 1 4 6 k II c c 17 23 t ? 29 36 P 41 s B3 bzh w w 3 V 12 18 24 26 30 37 42 45 47 49 lhn Q h fif W £ fi £ f 2 3 4 7 IS 19 25 27 31 38 43 U 4a 50 L R It -1 f rrl 5 8 9 14 20 32 33 39 0 0 n m IS 21 34 44 PLATE :i. Table Of Differentiated Consonants PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY PE Joint Committee on a 1135 phonetic English alphabet J6*+ Report of a Joint Committee represents the National Educational Association