American Language Supplement 2

by H. L. Mencken

  The compensatory change of oi into er, as in erster for oyster and erl for oil, is equally mysterious. In 1941 Dr. Lou Kennedy, chairman of the faculty committee on speech testing at Brooklyn College, devised a nonsense jingle to illustrate its occurrence, as follows:

  “O Father,” cried the Erster Boy

  To the Erster in the sea,

  “When I grow up to be an Erster Man

  What will become of me?”

  The Erster’s verse (voice) was choked with grief

  And he shook till the water rerled (roiled)

  At the awful fate of ersters

  For whom men fished and terled (toiled).

  “They berl (boil) us, Boy, or brerl (broil) us

  And fry us deep in erl (oil),

  Or the ranks we’ll jern (join)

  If they lack the kern (coin)

  Of those who are caught to sperl (spoil).”

  In 1926, as I have noted in AL4,5 Henry Alexander, a Canadian phonologist trained in England, suggested that oi in thoid and er in erster are really the same sound – that the hearer, expecting er in the first place, hears it as oi, and expecting oi in the second place, hears it as er.6 He said:

  This phenomenon … is found in several cases in the history of English.… It probably lies behind the puzzling substitution of w for v recorded in Cockney English of the Victorian era, a peculiarity familiar to all readers of “Pickwick.” The explanation is that Mr. Weller probably used a bilabial voiced fricative, i.e., a v sound formed by the two lips instead of the lip and teeth, for both v and w. This sound is heard today in certain German dialects; it is the voiced form of the Japanese f-sound. Acoustically, it is a compromise between v and w.… A similar explanation may account for the misuse of the aspirate … [as in] ’am and heggs (ham and eggs). If we assume that in all cases such speakers have a very weak h-sound, it is quite possible that this would impress itself on the ear as h when the hearer was expecting no h at all, and on the other hand would be insufficiently strong to sound like a real h when the hearer was prepared for a full h-sound.

  But this ingenious theory has been attacked by various correspondents – for example, Dr. Roger A. Johnson, professor of mathematics at Brooklyn College.1 “Contrary to Alexander,” he writes, “there is nothing half-way about these sounds: they are definitely and clearly interchanged.” He goes on:

  I am not an expert in these matters, but is it not possible to formulate a rule which might be called the Law of Reversal, to be stated somewhat as follows: When the replacement of one sound by another has for some reason become prevalent, there ensues a reversal whereby the second is replaced by the first?… The lower-class English acquired the habit of eliding the h at the beginning of a word. In a misguided attempt to correct this error they succeeded in putting h’s in all the wrong places.… Similarly, the attempt to correct foist leads to erl and erster. The result is not confusion between the two sounds, getting them sometimes right and sometimes wrong, but an out-and-out interchange.

  A somewhat similar theory was launched by Howard K. Hollister in 1923,2 and both got some support from Robert J. Menner, of Yale, in 1937.3 “The New York pronunciation of bird with oi,” said Menner, “is ridiculed as much as any pronunciation in the country, and in attempting to correct it New Yorkers sometimes lean over backward and pronounce oil with the er-sound.” But other authorities reject this notion – for example, A. F. Hubbell, of Columbia.4 The diphthong in thoid, he says, is actually variable, and sometimes it comes closer to ui than to oi. “These diphthongs,” he goes on, “do not occur in all the words in which … General American has er. They are heard only in syllables in which the written r is followed by one or more consonants which do not constitute an inflectional ending,” e.g., woik and woild,1 but not stoid (stirred). There is a like difference in the other direction. The change from oi to er is never made in words which end with the former, e.g., toy, boy, enjoy, or in their derivatives, or in those in which oi is followed by al, z or t, e.g., loyal, noise,2 loiter, but it does occur in those in which it is followed by l, n, nt, s, st and d, e.g., oil, join, joint, voice, oyster3 and avoid. Hubbell rejects the Menner-Hollister-Johnson theory on the ground that “speakers who use r-colored vowels in words like coil do so only in these words and not in words like curl,” but he confesses that, for the present, he lacks “any satisfactory explanation” of the change.

  Various other peculiarities of New York vulgar speech have been noted by other observers, e.g., a final ng is sometimes changed to nk, as in singink; an intrusive second g appears in a syllable ending with g, as in Long Giland; s and sh are voiced, so that acid becomes azid and assure becomes azhure; voiced consonants are unvoiced, so that village becomes villitch and hills becomes hilce; a glottal stop is substituted for medial t, as in le’er (letter);4 the two sounds of th are confused, or converted into t or d, as in t’row and wid; final r often disappears, as in cah (car) and fah (far); an intrusive r is frequent, as in I sawr;5 the long i and the short u are converted into ah, as in tahm (time) and cahm (come), and the flat a into e, as in kesh (cash). Even more curious changes have been reported. Hit becomes hitth, dead is indistinguishable from debt, and trip acquires an f-sound, making it something on the order of tfip. Many of these forms are ascribed to “a common habit of holding the tongue nearly flat in the mouth during the articulation of t, d, s, z, sh, zh, n and r, instead of placing its tip somewhere upon the gum above the upper teeth.”1 It is a strange fact that a man born and bred to this dialect later became one of the most adept practitioners of Oxford English known to linguistic pathology. He was William Joyce, who alarmed the English during World War II in the character of Lord Haw Haw.2 The Dutch loans in the New York vocabulary, some of them long since taken into the common speech of the country, are discussed in AL43 and Supplement I.4

  North Carolina

  One of the first Northerners to leave an account of his observations in the South was William Attmore, a Philadelphia merchant who visited North Carolina in 1787 to collect some lethargic accounts. In his diary for December 6 of that year he wrote:

  It sounds strange to my ear to hear the people in Carolina, instead of the word carry or carried, commonly say toat or toated. I asked a boy what made his head so flat; he replied “It was occasioned by toating water.” This is the usual phrase. I am told the joiner charges in his bill for “toating the coffin home” after it is finished.1

  To tote, of course, was not a North Carolinaism, nor was it new, for it had been recorded in Virginia in 1677, and there were many other records of it before 1787. Wentworth finds it in use from Maine to Oregon, but the DAE marks it “chiefly Southern,” and Noah Webster called it “a word used in slave-holding countries, said to have been introduced by the blacks.”2 But it is highly characteristic of all the dialects spoken in North Carolina today – Appalachian, Tidewater and Lowland Southern.

  The earliest study of the State’s speech that I am aware of was contributed to Dialect Notes in 1918 by J. M. Steadman, Jr.3 Nearly all the words and phrases on the list might have been found in a dozen other States, but there were nevertheless a few that have not, so far, been recorded elsewhere, e.g., frensy (or frency), the withered, dry leaves of tobacco or cotton; high-bob and scoots, a high-chair; to nullify (or nellify), to balk; shoe-round, a dance; to see monkeys, to be overcome by the heat; swamp-root, moonshine whiskey, and sore-back, a Virginian.4 Four later word-lists from North Carolina were published in 19445 and a fifth in 1946.6 Like the Steadman list they consisted largely of locutions common to the whole South and a good part of the North, but again there were some exceptions, e.g., foot-pie, an apple turnover; goochy, goosy; lazy-gal (or lazy-wife), a bucket running along a rope, used to bring water from a spring; to softmouth, to wheedle; to talk short, to speak angrily; aboon, above; to belch back, to rebound; cha-cha, the katydid; dogwood Winter, a spell of cold after dogwood is in blossom;1 lap-baby a child in arms; sanky-poke, a traveling-bag; tourer, a tourist; and to unfeed, to defecate. These came
from all over the State – some from the mountains to the west, some from the low-lying and isolated coast country, and some from the Yadkin region, settled largely by Germans. The latter began to filter in in 1709, and their descendants continued to use German, e.g., in church services, until c. 1850. The language is now forgotten, but it has probably left some sediment in the English of the area.

  The dialect of the North Carolina coast differs considerably from the Tidewater speech of Virginia and South Carolina. It shows a number of archaic terms that seem to be survivors of the first settlements. An early report on it, published by Collier Cobb in 1910,2 was chiefly devoted to these archaisms. A number of them have not been reported from other places in the United States, e.g., acre, a furlong; may, a sweetheart; kelpie, a water-sprite; to scoop, to run away; cracker, a boaster; to bloast, to brag; bloater, a chubby child; cant, gossip; to abrade, to sicken; birk, a smart young fellow; and fause, a tidal creek; but nearly all of them are to be found in the English dialects. Cobb also listed some curious pronunciations, e.g., buer for butter, egal for equal, leuch for laugh, plead for pleased, fant for infant, fole for fool, and wharrel for quarrel. He described the dialect as the language of “the better classes, or at least the middle classes, in England in the days of Elizabeth.” He noted that, even in 1910, improving communications with the mainland – there called the country – were wearing it down.3

  Elizabeth Jeannette Dearden, in her study of the materials accumulated by the late Guy S. Lowman, Jr., for a projected South Atlantic States section of the Linguistic Atlas, shows that the division between Appalachian and General Southern speech lies more to the eastward in North Carolina than in Virginia. This she explains on the ground that “the Virginia Piedmont was settled by expansion from the coast,” whereas that of North Carolina “was settled to a large extent by the Scotch-Irish and Germans who pushed out into the uplands from the mountains.” “This area in North Carolina,” she continues, “has acquired nowhere near the linguistic prominence and uniformity of the Virginia Piedmont.… There has been a spreading of mountain terms down the Cape Fear river.” The isoglosses separating the mountain speech from that of the Piedmont enter North Carolina somewhere between Surrey and Warren counties. They are not very clearly marked, and a large area shows mixed speech. The extension of Piedmont speech eastward is blocked by the so-called pine barrens, where new terms “are stopped short because of the sparsity of the population and the high rate of illiteracy.” There are many speech-pockets on isolated peninsulas along the coast: they show archaisms that have disappeared further inland, and no two of them are quite alike. Plans for a more comprehensive and scientific survey of North Carolina speech are now being furthered by George P. Wilson, secretary-treasurer of the American Dialect Society, and Hans Kurath. Wilson is accumulating word-lists, and Kurath is seeking to augment and analyze the material gathered by Lowman.1

  North Dakota

  I have been unable to find any report on the speech of North Dakota. Apparently it differs but little, if at all, from that of the adjoining States, especially Nebraska and Minnesota.

  Ohio

  Ohio, with an area of 41,222 square miles (nearly that of England) and a population of more than 7,000,000, shows a number of diverging speech areas, though most of its people speak General American. In the Western Reserve, comprising thirteen counties in the northeastern corner of the State, there is still some evidence that most of the early settlers came from New England, and in the extreme south, along the Ohio river, there are equally plain tracks of Southern influence. Also, there are areas in which foreign immigration has left its mark, notably that of Cincinnati, which was settled largely by Germans. Some surviving New England pronunciations in Hudson township, which straddles the Cuyahoga river just below Cleveland, were noted in the first issue of Dialect Notes (1890) by N. P. Seymour,1 but he had to add that many that had been familiar in his youth were beginning to disappear, e.g., sneck for snake, bury to rhyme with fury, put to rhyme with hut, deestrict for district, and cheer for chair. In 1917, John S. Kenyon, a highly competent phonologist born in 1874 in Medina county, also just south of Cleveland, contributed some valuable observations on the pronunciation of that country in his youth,2 and in 1921 he followed with a few corrections.3 Nearly all the terms he listed were of New England origin, and he himself traced the genealogy of most of them, but there were also a small number borrowed from non-Yankee immigrants, e.g., shillalah from the Irish and wampus from the Pennsylvania Germans. In 1890 J. M. Hart and other members of the Philological Society of the University of Cincinnati contributed some local notes to Dialect Notes,4 but only one of the words they offered showed German influence, to wit, allerickstix, which was described as a schoolboys’ term meaning all right (Ger. alles richtig).

  In 1916 W. H. Parry dealt with the dialect of six counties in the southeastern part of the State, three of them fronting on the Ohio river and facing West Virginia.5 This area, he said, was settled by two streams of immigration, the first coming from New England by way of the river and the second from Maryland, Virginia and southern Pennsylvania over the mountains. The result was a mixed speech with some curious oddities, e.g., the insertion of r between u or e and sh, as in rursh (rush) and frersh (fresh). “Occasionally,” said Parry, “there is a sentence arrangement peculiar to German, such as the use of once as the German einmal and the use of the verb at the end of the sentence.”1 His list showed only a few terms not reported from other parts of the country, e.g., tucks, rheumatism; bone-eater, a dog, and dry-hole, a stupid person (from the oil field term).2

  There has been a natural and considerable interchange of words and phrases between one part of Ohio and another, and the well-known Germanist, Dr. R-M. S. Heffner, who was born in 1892 at Bellefontaine, in the west central part of the State, testifies that New Englandisms were common there in his boyhood, though the area lies “well out of the sphere of influence of the Western Reserve.3 He reports, indeed, that of the thousand-odd forms from the Maine coast listed by Dr. Anne E. Perkins in 1927 and 19294 no less than 630 are “entirely familiar to my ear from the usage of my father and his friends.” The elder Heffner was born in Logan county, in which Bellefontaine is situated, and “his parents came over the mountains to Ohio from Pennsylvania.” The speech of Athens, O., the seat of Ohio University,5 has been reported by Lewis A. Ondis.6 Athens is in the southeastern part of the State, adjoining the area discussed by Parry in 1916. Ondis says that the peculiarities he lists “are found on all social levels, including town officials, business people and even native school teachers,” and that they “are quite general over an area of about thirty or forty miles about Athens, reaching the vicinity of Lancaster and Chillicothe.” “The continual flow of students from all parts of the country and the permanent faculty of Ohio University,” he continues, “have had hardly any influence on the natives.” I quote from his paper:

  -ish. This combination is invariably heard as -eesh.… A native will pronounce fish, ignition, official, commission, vision, fissure and issue as feesh, igneetion, offeecial, commeesion, veesion, feessure and eesue. Short i in any other position is normal.

  -ush. Generally pronounced -oosh, especially in bush, bushel, push, which are invariably boosh, booshel, poosh. Long and short u, as in fuse and tub, are normal.

  a. In this region and generally throughout Ohio the short a approaches the sound of short e as in met, bet, though somewhat prolonged. Such words as calf, half, land, pass, past, salve are usually pronounced keff, heff, lend, pess, pest, sevv.

  au, aw. These digraphs approach the sound of o in come, stop, done, so that caught, daughter, automobile, lawn are pronounced cut, dotter, ottomobile, Ion.

  Ohio, of course, is in the area being investigated by Albert H Marckwardt and his collaborators for the projected Linguistic Atlas of the Great Lakes and Ohio valley regions. They have already gathered material from the towns of Hiram, Medina, Ottawa, Mt. Vernon, Richmond, Bear’s Hill, Worthington, Reynoldsburg, Marietta and We
st Union. In 1930 Hans Kurath, then of Ohio State University and later editor-in-chief of the Linguistic Atlas, made a detailed report upon the speech of a young woman junior at the university.1 The subject was a native and lifelong resident of Columbus, born of a father of German ancestry (though he did not speak German), and a New England mother. Henry Sweet’s “The Young Rat” was used to test her speech, which was recorded in a modification of the IPA alphabet. My impression is that this record shows a good specimen of General American, with a few Southern influences.2

  Oklahoma

  The educated Oklahomans speak General American, and in the speech of the lowly Okies there is little to distinguish it from that of the adjacent wilds, especially the Ozark regions of Missouri and Arkansas. As far as I know, there has been no published study of the local vocabulary. A brief glossary appears in “Oklahoma: A Guide to the Sooner State,” one of the WPA’s series of State guidebooks,1 but it is confined to oil-field terms. In 1938 Floy Perkinson Gates made a report on some of the words and phrases used in the Dust Bowl,2 but offered no evidence that they were peculiar to Oklahoma. They included black blizzard, a very severe storm; black roller, of the same meaning, and dust pneumonia, apparently a variety of silicosis. Dean E. H. Criswell, of the University of Tulsa, has collected material from a dozen counties, and has interested some of his students in English in the work, but it will take a number of years to cover so large a State.3

  Oregon

  Save for a few terms borrowed from the Chinook jargon,4 the speech of Oregon does not depart noticeably from the vocabulary of General American, and in phonology it sticks to the same normalcy. There are, of course, some pockets of different speech, mainly created by accidents of immigration. One such pocket is in the hills of Wallowa county, in the northeastern corner of the State. This inhospitable region was settled by fugitives from the Ozarks, and they brought with them the speech of Appalachia, with some traces of Lowland Southern. Says T. Josephine Hausen, the only observer to report upon it:5