American Language

by H. L. Mencken

  The historical study of American pronunciation was put on a solid basis by the publication of the second volume of the late George Philip Krapp’s “The English Language in America” in 1925. There had been scattering investigations on the subject before then, but Dr. Krapp was the first to undertake an exhaustive examination of the available material — the early dictionaries, grammars and spelling books, the attempts at devising phonetic alphabets, and the records of the Massachusetts, Connecticut and New York towns, many of them made by unlearned men and written phonetically. One of his conclusions was that most of the peculiarities of American pronunciation have historical precedents in England, and that many of them may be found to this day in the English dialects. Even the nasalization which Englishmen always mark in American speech “is by no means exclusively American.” It was charged to the English Puritans by their critics, and is denounced in “Hudibras,” Part I, Canto III. Dr. Krapp believed that “differences of practise among standard American speakers, that is, among members of good standing in the community, were formerly much more numerous than they are today,” and that they “continue to show an increasing tendency to disappear in an all-embracing uniformity.” He demonstrated that many forms now confined to isolated speech-islands, for example, in rural New England or the remoter parts of the South, were once almost universal. He showed that the type of American prevailing in the Boston area and in the tidewater regions of the South is closely related historically to the Southern type of English, and that Western America is derived, at least in large part, from Northern English.44

  2. THE VOWELS

  One of the most noticeable differences between Standard English and Standard American lies in the varying pronunciation of a in about 150 words in everyday use. The English, in general, prefer the broad a of dark before f, ft, m, nch, nd, nt, sk, sp, ss, st and th, as in laugh, draft ca(l)m, branch, command, chant, ask, clasp, grass, last and path, whereas most Americans use the flat a of that.45 But these preferences are not invariable. There are many Englishmen, otherwise quite orthodox in their speech habits, who incline toward the flat a before n and f and before s followed by a consonant,46 and nearly all of them use it in fancy, despite a general American belief, promoted by haw-haw types on the stage, that they always say fahncy. They also use it in stamp, though they cling to the broad a in example. Again, they prefer the flat a of rack in amass, elastic, gas, lass, mass, massive, masticate and plastic. Finally, they pronounce the verb can just as we do, though they sound the broad a in can’t.47 Contrariwise, most Americans use the broad a before (l)m, as in palm, and virtually all use it before r, or r and a consonant, as in bar, cart, park, harm, etc., and before th in father. Moreover, it is used in most of the English situations in the Boston area, though with a slight change in value, and in a number of words in the South, e.g., master, aunt, tomato, tassel.48

  It used to be believed that the broad a was historically the more respectable, and that the flat a had come into American and into some of the English dialects as a corruption, but the exhaustive researches of Krapp have disposed of that notion.49 During most of the Eighteenth Century, in fact, a broad a was regarded in both England and America as a rusticism, and careful speakers commonly avoided it. When Thomas Sheridan published his “General Dictionary of the English Language” in London in 1780 he actually omitted it from his list of vowels. He had room for an a approximating aw, as in hall, but none for the a sounding like ah, as in barn. He gave the pronunciation of papa as if both its a’s were that of pap, and even ordained the same flat a before r, as in car and far. Benjamin Franklin, whose “Scheme for a New Alphabet and a Reformed Mode of Spelling” was published in Philadelphia in 1768, was in complete accord with Sheridan. He favored the flat a, not only in all the words which now carry it in American, but also in calm, far, hardly and even what, which last was thus made to rhyme with hat. Franklin’s pronunciations were presumably those of the best circles in the London of his time, and it seems likely that they also prevailed in Philadelphia, then the center of American culture. But the broad a continued common in the folk-speech of New England, as it was in that of Old England, and in 1780 or thereabout it suddenly became fashionable in Standard London English. How and why this fashion arose is not known, nor is it known what influence it had upon the educated speech of New England. It may be that the New Englanders picked it up, as they picked up so many other English fashions, or it may be that they simply yielded to the folk-speech of their region. Whatever the fact, they were using the broad a in many words at the time Noah Webster published his “Dissertations on the English Language” at Boston in 1789. In it he gave quality, quantity and quash the sound of a in hat, but he gave advance, after, ask, balm, clasp and grant the a of arm. In subsequent editions of “The American Spelling Book” he favored the broad a before a final r or before r followed by a consonant, e.g., bar, depart; before (l)m, e.g., embalm; before a final s or s followed by a consonant, e.g., pass, ask; before f, e.g., staff, half; before th, e.g., path; before lv, e.g., salve, calves; before n followed by ch, s or t, e.g., blanch, dance, ant; in words spelled au before s, e.g., sauce; in words spelled au before n followed by ch, d or t, e.g., staunch, jaundice, aunt; and in a number of other words, e.g., chamber, slander, gape.50 He even advocated the broad a in bracelet, though in his “Dictionary of the English Language Compiled for the Use of the Common Schools in the United States” (1807) he abandoned it for the flat a.

  Webster’s immense authority was sufficient to implant the broad a firmly in the speech of the Boston area. Between 1830 and 1850, according to C. H. Grandgent,51 it ran riot, and was used even in such words as handsome, matter, apple, caterpillar, pantry, hammer, practical, Saturday and satisfaction. Oliver Wendell Holmes protested against it in “The Autocrat of the Breakfast Table” in 1857, but it survived his onslaught. It has been somewhat modified in sound with the passing of the years. Says Grandgent: “The broad a of New Englanders, Italianate though it be, is not so broad as that of Old England.… Our grass really lies between the grahs of a British lawn and the grass of the boundless prairies.” In the New England cities, he adds, it has been “shaken by contact with the Irish,” and is now restricted, in the main to

  a few specific classes of words — especially those in which an a (sometimes an au) is followed by a final r, by an r that precedes another consonant, by an m written lm, or by the sound of f, s, or th: as far, hard, balm, laugh, pass, rather, path. In the first two categories, and in the word father, ah possesses nearly all the English-speaking territory; concerning the other classes there is a wide divergence, although flat a appears everywhere to be disappearing from words like balm. Yankeedom itself is divided over such combinations as ant, can’t, dance, example, in which a nasal and another consonant follow the vowel; aunt, however, always has broad a.52

  The imprimatur of the Yankee Johnson was not sufficient to establish the Boston a outside New England. His rival and bitter critic, Lyman Cobb, whose “Just Standard For Pronouncing the English Language” appeared in 1821, allowed it before th and lv and in words wherein it appeared as au, but ordained the flat a in class, clasp, fast, ask, asp, branch, dance, chaff, raft and their congeners. Webster’s other great antagonist, Joseph E. Worcester, whose “Comprehensive Pronouncing and Explanatory Dictionary of the English Language” appeared in 1830, set up a distinction between the true British broad a and the modified New England a described above by Grandgent, and frowned upon the former. “His hesitation with respect to words like ask, dance, chaff, etc.,” says Krapp, “was due not to the fear that the sound which he advocated might seem too near [the a of hat], but too near [the a of bard]. In other words, the vulgar extreme which was to be avoided was [the latter and not the former].” William Russell, who published a number of popular textbooks during the second quarter of the century, advocated the flat a in all words of the glass, grasp, past, graft, grant, dance, branch, chant, rather and bland classes. By 1850 it was dominant everywhere west of the Berkshires and so
uth of New Haven, save for what Grandgent calls “a little ah-spot in Virginia,” and its sound had even got into such proper names as Alabama and Lafayette.53 “In the United States beyond the Hudson — perhaps beyond the Connecticut,” says Grandgent, “the flat a prevails before f, s, th, and n.” Nevertheless, the broad a has got into a few words, if not many. Those in which it is followed by lm I have mentioned. They were once pronounced to rhyme with ram and jam, but their pronunciation that way has begun to seem provincial and ignorant. Krapp says that the a has likewise broadened in alms, salmon and almond,54 but it is my own observation that this is not yet generally true. The first syllable of salmon, true enough, does not quite rhyme with ham, but it is nevertheless still very far from palm. The broad a, by a fashionable affectation, has also appeared in vase, drama, amen and tomato — in the last case probably helped by the example of Southern speech. This intrusion has been vigorously denounced by an Englishman, Evacustes A. Phipson. He says:

  It is really distressing to a cultivated Briton visiting America to find people there who … follow what they suppose to be the latest London mannerism, regardless of accuracy. Thus we find one literary editress advocating the pedantic British pronunciation tomahto in lieu of the good English tomato, rhyming with potato, saying it sounds so much more “refined.” I do not know whether she would be of the same opinion if she heard one of our costermongers bawling out “ ’Ere’s yer foine termarters, lyde, hownly tuppence a pahnd.” Similarly, we sometimes hear Anglomaniac Americans saying vahz for vase. Why not bahz, and cahz?55

  Amen, with the broad a, seems to be making progress. E. W. Howe tells a story of a little girl in Kansas whose mother, on acquiring social aspirations, entered the Protestant Episcopal Church from the Methodist Church. The father remaining behind, the little girl had to learn to say amen with the a of rake when she went to church with her father and amen with the a of car when she went to church with her mother.56 In Canada, despite the social influence of English usage, the flat a has conquered, and along the Canadian-New England border it is actually regarded as a Canadianism, especially in such words as calm and aunt. The broad a, when heard at all, is an affectation, and, as in Boston, is sometimes introduced into words, e.g., stamp, which actually have the flat a in England. In the United States, save in the Boston area, one never hears it in gather, lather and blather, and even in Boston it is often abandoned for the flat a by speakers who are very careful to avoid the latter in palm, dance and aunt. Krapp says that it is used in “some words of foreign origin,” notably lava, data, errata, bas-relief, spa, mirage and garage, but this is certainly not true of the first three, all of which, save exceptionally, have the flat a. So has piano, though the Italian i is preserved, and pyano is now only a vulgarism. Patent, in American, always has the a of cat, but in English the a of late is often heard when the word is used in the sense of a license or monopoly. In England mater, a common synonym for mother, has the same a, but in the American alma mater, which is seldom used in England, the second and third a’s are that of bard, though the first is commonly that of pal. In English the second a of apparatus is always that of late, but in the United States it is often that of cat. The same difference is to be noted in the pronunciation of data, gratis, status and strata. In phalanx it runs the other way, with the English preferring, for the first syllable, the a of rack, and Americans that of mate. In radio the usual American pronunciation shows the a of mate, but the plain people of New York City apparently prefer the a of rack, and Alfred E. Smith’s use of it in 1928 attracted nation-wide attention, and inspired some imitation. There were pundits at the time who argued that Al was right, and cited the analogy of radical. The English use a broad a in the final syllables of charade and promenade, but most Americans prefer the a of mate. In the second syllable of asphalt the English always use the a of rack, but Americans sometimes use the aw of bawl. In England the a of patriotism is always the a of rack, but in the United States it is often that of late.57 Larsen and Walker58 say that the latter a is used by Americans in ignoramus, tornado and ultimatum, but I often hear the a of dram in ignoramus, and either that of rack or that of bar in tornado and ultimatum. In radish the a is sometimes that of cab and sometimes a sort of e, hard to distinguish from that of red. In such proper names as Alabama, Alaska, Montana, Nevada and Colorado the flat a of has is often heard, especially in the States themselves, but a broad a is certainly not unknown. The usual pronunciation of again and against gives them a second a indistinguishable from the e of hen, but there is also a spelling pronunciation employing the a of lawe. In the years before the Civil War the plain people converted the a of care into the a of car in bear, dare, hair and where, into a short i in the verb can, into a short e in catch, and into a long e in care, scarce and chair, thus producing bar, dar, har, whar, kin, ketch, keer, skeerce and cheer. They flattened the long a of marsh and omitted the r, making the word rhyme with lash, and they reduced sauce to sass, saucy to sassy and because to becuz or bacaze. The e of learn, serve, mercy, certain and eternal became a broad a, producing larn, sarve, marcy, sartin and tarnal; the a of caught was flattened, producing catcht or ketcht; and the a of drain was turned into a long e, producing dreen. Some of these toyings with a survive, but not all. The rest have been exterminated by the schoolmarm, or forced into exile among the remoter dialects.59

  There are some differences between the pronunciation of e in English and American, but not many. The English use a long e, like that of bee, in the first syllable of evolution, which is always short in American. They also prefer a long e in epoch, but in the United States it is often short. Contrariwise, the English use a short e in penalize, lever and egoist, whereas most Americans prefer a long one. The English are always careful to make the first syllable of seamstress sem and that of cleanly clen, but Americans commonly stick to the e of the stems. In the United States a spelling pronunciation often appears in pretty, making the first syllable rhyme with set; it always rhymes with sit in English. The use of the long e in deaf, though historically very respectable and ardently advocated by Noah Webster,60 has disappeared from cultivated American speech; it persists, however, in the vulgate. In the same way the i-sound, as in sit, has disappeared from get, yet, general, steady, chest and instead; Benjamin Franklin defended it, but now even the vulgate is losing it: This pronunciation, according to Menner,61 was correct in Seventeenth Century England, and perhaps down to the middle of the next century. It is probable that the colonists clung to such disappearing usages longer than the English. The latter, according to Webster, were unduly responsive to illogical fashions set by the exquisites of the court and by popular actors. He blames Garrick, in particular, for many innovations, most of them not followed in the colonies. But Garrick was surely not responsible for the displacement of mercy by marcy, which Webster ascribed somewhat lamely to the fact that the letter r is called ar: he proposed to dispose of it by changing the ar to er. In his time the a of lame was generally heard in egg, peg, leg and so on, but it too is now confined to the vulgate, as is the a of bag in keg. As Krapp shows, the neutral e that has taken its place, and toward which all our vowels seem to be tending,62 shows signs of itself disappearing. This is particularly noticeable, in American, in such words as moral, quarrel and real, which become mor’l, quar’l and reel, each a single syllable. In the vulgar speech this neutral e is also dropped from other words, notably poem, diary, violet and diamond, which become pome, di’ry, vi’let and di’amond. Even in Standard American it grows shadowy in the second syllables of fertile, hostile, docile, servile, agile, reptile, etc. In Standard English these words are pronounced with the second syllables rhyming with vile, but the English use a short i in fragile and facile. They also use it in senile, which commonly shows the long i in American. The long e-sound in creek and creature is maintained in Standard American, but changed to the short i of sit in the vulgate. Sleek has divided into two words, slick and sleek, the former signifying cunning and ingratiating, and the latter referring especially to
appearance. Of late there has been a strong tendency to abandon the old e-sound in such terms as bronchitis and appendicitis for an ai-sound, as in pie and buy; this is a senseless affectation, but it seems to be making progress. A contrary movement to abandon the old ai-sound in iodine, quinine, etc., for an e-sound, as in sleep, has better support in etymology, but is apparently less popular. Chlorine and vaccine are always pronounced with the e-sound, but iodine continues to be iodyne, strychnine is still usually strychnyne, and kin-een for quinine still sounds strange. In two other familiar words the ai-sound has been supplanted in American: in sliver by the short i of liver, and in farina by an e-sound. Both have the ai-sound in Standard English. Dynasty, in American, has a first syllable like dine, but in English it is din. Isolate is always eye-solate in English, but sometimes it is iss-solate in American. Tribunal and simultaneous have the y-sound in American, the short i in English. Misogynist has the short i in American but the long one of mice in English. Been, in American, is almost always bin or ben; bean never appears save as a conscious affectation. But in England bean is preferred.