American Language Supplement 2

by H. L. Mencken

  1 Pitch Patterns in English, Studies in Philology, July, 1926, p. 372.

  2 Dr. Scripture, born in New Hampshire in 1864, had a Ph.D. degree from Leipzig and an M.D. from Munich. He became director of the psychological laboratory at Yale in 1892, and lectured at Columbia and the Johns Hopkins. Later he was professor of experimental phonetics at Vienna.

  3 The Acoustical Nature of Accent in American Speech, American Speech, Feb., 1937, pp. 49–56.

  4 Other phonologists have used the x-rays, the moving picture, the vibrograph, the resonator, the kymograph and, of course, the tuning-fork and the laryngoscope. Their reports bristle with talk of decibels, centroids of energy and other such things. So far, nothing comparable to the electrocardiograph and electrocephalograph has been devised to study speech, but no doubt it will come. See A Brief History of Palatography, by Elbert R. Moses, Jr., Quarterly Journal of Speech, Dec., 1940, pp. 615–25.

  1 These categories were suggested by Dr. Joseph Tiffin.

  2 An annual bibliography is in the Supplement to the Publications of the Modern Language Association. From 1934 onward American Speech printed one quarterly, prepared by Dr. S. N. Treviño, of the University of Chicago, whose own contributions have been numerous and valuable.

  1 The Doctor; London, 1834–47, Interchapter XXV.

  2 He called this alphabet Palaeotype. It is given in full in his Early English Pronunciation; London, 1869–74, Vol. I, pp. 3–12.

  3 The International Phonetic Alphabet, by John S. Kenyon, American Speech, April, 1929, pp. 324–27.

  4 Phonetic Transcription and Transliteration: Proposals of the Copenhagen Conference, April, 1925; Oxford, 1926. This report was prepared by Otto Jespersen and Holger Pedersen, both professors in the University of Copenhagen. There were twelve phoneticians in attendance, coming from Norway, Denmark, Sweden, Germany, Poland, Holland, England and France. The English representative was Daniel Jones.

  1 A Petition, American Speech, Oct., 1939, pp. 206–08.

  2 Fourth edition, revised and enlarged; London, 1937. On pp. xxxvii and xxxviii he gives a list of other phonologists using it.

  3 Cambridge (England), 1944.

  4 p. 169.

  1 pp. 11 and 14.

  2 New York, 1919.

  3 New York, 1925.

  4 p. vi.

  5 Knott died on Aug. 14, 1945.

  6 Problems in Editing an American Phonetic Dictionary, by Kenyon and Knott, American Speech, Oct., 1936, pp. 227–31. Suggestions by Miles L. Hanley, L. Sprague de Camp, C. K. Thomas, Lee S. Hultzén and Cabell Greet were printed in American Speech, Dec., 1936, pp. 319–26. The Pronouncing Dictionary was finally published in 1944.

  1 American Speech complained in a review, April, 1935, p. 140, that many of the pronunciations also go back to Noah. “Professor Kenyon’s splendid piece of work,” it said, “is in the preface; in the body of the dictionary are regularly the old, often provincial and unrepresentative pronunciations.”

  2 Dr. Kenyon tells me that he is not convinced that this system is simpler than the IPA, as I ventured to say in AL4, p. 320. “Some five years of dealing with it,” he says, “convinced me that it was more elaborate and complicated than that of the IPA with its one invariable symbol for every significant sound.” Perhaps I should have said “more familiar” rather than “simpler.” See Progress in Pronouncing Dictionaries, by Bert Emsley, American Speech, Feb., 1940, pp. 55–59; A Survey of English Dictionaries, by M. M. Mathews; London, 1933, pp. 90–92, and Pronouncing Systems in Eighteenth Century Dictionaries, by Esther K. Sheldon, Language, Jan.-March, 1946, pp. 27–41. The New Practical Standard Dictionary of 1946 uses a simplification of the Webster system, without any recourse to the IPA.

  3 Mispronunciations?, American Speech, April, 1936, pp. 137–41.

  4 American Dictionaries and Pronunciation, American Speech, Dec., 1938, pp. 243–54.

  5 American Dictionaries and Variant Pronunciations, American Speech, Oct., 1939, pp. 175–80.

  1 On Handbooks, American Speech, Feb., 1940, pp. 89–92.

  2 His speech was delivered in April, 1933. A stenographic report of it was printed in the Quarterly Journal of Speech, Feb., 1935, pp. 1–10.

  3 A table showing how seven standard dictionaries disagree as to the pronunciation of different words is in Webster 1934, pp. lix-lxxviii. The dictionaries covered are Webster itself; the New English Dictionary and its Supplement; Oxford, 1888–1933; Daniel Jones’s English Pronouncing Dictionary, second edition, London, 1924; H. C. Wyld’s Universal Dictionary of the English Language; London, 1932; the Century Dictionary; New York, 1911; Funk and Wagnalls’s Standard Dictionary; New York, 1931, and the International French-English and English-French Dictionary, edited by Paul Passy, George Hempl and Robert Morris Pierce, 1904. The usage of 200 “educated, native-born citizens of the United States, not professional radio speakers,” is shown in Radio Pronunciations, by Jane Dorsey Zimmerman; New York, 1946.

  1 Dr. E. H. Sturtevant, of Yale, in explaining his failure to use any sort of phonetic alphabet in his Linguistic Change: An Introduction to the Historical Study of Language; Chicago, 1917, p. v, said: “Such a notation would have required a long explanation, which some readers would have skipped, and which would have caused others to lay the book aside.” Dr. Leonard Bloomfield, also of Yale, agreed in The Stressed Vowels of American English, Language, June, 1935, p. 98: “Any transcription shocks and offends all but the few readers who have been inured to the free use of graphic symbols.”

  1 Phonetic Transcriptions from American Speech, American Speech Reprints and Monographs, No. I; New York, 1936.

  2 The text is in On Teaching Speech, by Greet, Baltimore Evening Sun (editorial page), Dec. 26, 1934. Under the title of The Young Rat it first appeared in Sweet’s Primer of Spoken English; Oxford, 1890, pp. 66–68. At Columbia the title was changed to Grip the Rat. Greet later changed it to Arthur the Rat, for Arthur allows for many more different pronunciations than Grip.

  3 A more detailed account of the beginning of this enterprise, with a description of the apparatus used, is in American Speech Records at Columbia University, by Ayres and Greet, American Speech, June, 1930, pp. 333–58. See also Diction of Roosevelt and New Deal Aides Recorded for Columbia Language Study, New York Times, July 12, 1934.

  4 In 1942 the Army and Navy, in association with the Linguistic Society of America and the Intensive Language Program of the American Council of Learned Societies, began teaching foreign languages to soldiers and sailors by the same method. Many (though apparently not all) of the recordings and manuals prepared for the purpose are now published by Henry Holt & Co. The linguistic theories embodied in the Intensive Language Program have been challenged by Ephraim Cross in Learning Foreign Languages: a Little Politics and Some Economics, Modern Language Journal, Feb., 1947, pp. 69–79.

  5 Linguaphone for Languages; New York, 1945, p. 3. The course ran to sixteen ten-inch records of two sides each, and included “thirty conversational lessons and two lessons in phonetics.”

  1 Talking Dictionaries, Quarterly Journal of Speech, April, 1941, pp. 274–81.

  2 The IPA, almost always with modifications, may be found in the standard words on phonology, e.g., A Pronouncing Dictionary of American English, by Kenyon and Knott, already cited. American Speech, with delicate humor, reprints it on the inside back cover of each issue – but without any explanation, understandable to a layman, of the significance of its symbols. They are used by all the philological journals, and are obtainable in various type faces. The Mergenthaler Linotype Company offers them in ten-point only but they come in both bold and Roman, and run to 202 characters. The Intertype Corporation offers but 38, but all of them come in both eight-point and ten-point, and some in other sizes up to fourteen. The Lanston Monotype Machine Company offers them in eleven faces, running variously from five-point to eleven. The American Type Founders, Inc., offers only a few of the characters, for it deals mainly in type faces that are not used in text composition. The Ralph C.
Coxhead Corporation, which manufactures a typewriter called the Vari-Typer, offers three phonetic alphabets, one of more than 80 characters. Some sense of the inadequacy of the IPA must have been in the mind of Dr. Kenneth L. Pike when he wrote his revolutionary Phonetics: A Critical Analysis of the Phonetic Theory and a Technic for the Practical Description of Sounds; Ann Arbor (Mich.), 1943. Unhappily, his method of representing sounds, though it tells a lot to a phonologist, is apt to be baffling to the layman, for it represents the ordinary sound of the letter f, for example, by MaIlDeCVoeIpfocAPpdaldtlfnransnfSrpFSs.

  2. THE VOWELS

  “Every vowel sound without exception,” said Hilaire Belloc in “The Contrast,”1 “has taken on this side of the Atlantic [i.e., the American side] some different value from what it has on ours [i.e., the English side]. And in many cases the change is so great that the exact setting down of it in an accurate transliteration would involve a totally different spelling.” Even “a totally different spelling,” I am convinced, would not suffice to indicate these differences, for they are almost infinite in gradation and hence virtually innumerable. Consider, for example, the much debated a-sound, a favorite gauge of the disparity between English speech and American. At one end of the scale is the broad, solid ah that speakers of the Received English Standard put into such words as fast, last, glass and dance, and at the other end is the so-called flat a, as in can and Daniel, used by speakers of General American. Between the two stands the compromise of the Boston-Hudson Valley dialect, first given countenance, I believe, by Joseph E. Worcester in his “Comprehensive Pronouncing and Explanatory Dictionary” of 1830. But what, precisely, is this compromise a? There is actually no way to record it in print, even with the aid of the IPA and its extensions, for the sound sought to be recorded is not fixed, but variable, and in one form or another it runs all the way from virtual identity with the most extreme form of the English a to something that is very hard to distinguish from the a of General American. I have heard a group of six Bostonians, all talking at once, use six different variations of it, and in the New York region more are easily to be distinguished. In one community in Virginia, according to George P. Wilson, ten different ways of pronouncing aunt are in common use, and elsewhere in the State he has detected two more.2 Most persons who attempt the compromise a, in fact, vary it considerably in their own speech, so that it is sometimes difficult to make out whether they use it or do not use it. So with the other vowels: they are all in a state of flux, and Arthur J. Bronstein has hinted that something analogous to the Great Vowel Shift of 1500, which separated Modern English from Middle English, may be in progress.1

  The old-time grammar-books were content to inform the young that there were five vowels in English, but this was true, of course, only of the letters used to represent them, not of the sounds. In 1791 John Walker was constrained to distinguish ten vowels and diphthongs,2 and in 1837 Isaac Pitman, the pioneer of modern English shorthand, went on to six long vowels,3 six short ones, and four diphthongs. The number has been growing ever since, but no two phoneticians seem to be in agreement as to what it is precisely. Richard Paget, in 1925, was content with “thirteen separate vowels,”4 but Daniel Jones, in the 1937 edition of his “English Pronouncing Dictionary,” went to fifteen vowels and twelve consonants, and Leonard Bloomfield, in 1935, reported seventeen “syllabic phonemes” among “educated speakers in Chicago” alone.5 In 1937 Oscar Browne reported in his “Normal English Pronunciation” that phoneticians distinguish seventy-two “variants of vowel sounds,” but some of these, he added, are not heard in English. All such estimates, of course, are unreliable, for what they classify should be called, not vowels, but vowel-groups, and the differences within each group are too numerous and too minute to be described in words or symbols. “The vowel,” says John W. Black, “is an ever-changing phenomenon during phonation.… Certain factors are constant and others are variable.… Each vowel is a succession of different structures.… Teachers … should not say that this is the vowel in top without adding that this is one of the possible vowels in top.”1

  The sounds of vowels are produced by columns of air going through the two resonator-spaces above and below the tongue, with modifications effected by variations in the position of the tongue itself and the lips. Obviously, no two human mouths are precisely alike, and equally obviously the mechanism of speech thus differs from individual to individual, and in the same individual from time to time. “These positions of the vowels in the mouth,” said Dr. Robert Bridges, “are like the places of the outfielders at cricket, whom the captain shifts about according to the idiosyncrasies of the bowler and batsman: their stations are named and relatively well established, but it cannot be foreseen on any occasion where any one of them will be standing; and in any case the accurate knowledge of the ideal vowel-position is of no more practical use to the speaker than the scientific millimetred analysis of the action of the complicated stops of a clarinet would be to the performer on it.”2 Nevertheless, Bridges believed that the five vowels of the old grammar-books sufficed “to represent the acme of the main distinctions of quality and timbre,” and that getting any closer to their sounds would be possible only “if all speakers had exactly similar organs.” As things stand, it is sometimes difficult to distinguish between a and o on the one hand and a and e on the other, or between a and the neutral vowel, but on most occasions they remain reasonably distinct, despite the wide range of sounds on both sides of the fence. “There can be considerable variation in the composition of the spoken vowel,” says Leroy T. Laase, “and the character of the vowel still be clearly recognizable.”3 The quantity, i.e., the length, of any vowel is conditioned mainly by the stress put upon it, and to a lesser extent by its position with relation to other sounds. Thus the exclamation ah, in isolation, shows a longer a than the one in father or palm, and the quantity thereof is further affected by the surrounding consonants and the rapidity of speech.1 “The pronunciation of a word in isolation,” says R.-M. S. Heffner, “represents an unhampered shot at the goal of an ideal or normal pronunciation, while pronunciation in context represents a more or less disturbed or jostled shot at the same goal.”2 The most active factor, however, is stress, and Heffner concludes that “strong stress with falling intonation produces a slightly greater duration of the vowel than strong stress with rising intonation.”3 All such differences in quantity tend to become differences in quality, and in the long run they may produce entirely new vowels, or, at all events, variations so marked that they have to be represented by different symbols.4

  That there is a movement in American English toward the shortening of vowels has been noted by various observers.5 The English authorities ordain the long e in evolution,6 and the long i in isolation and the words of the fragile class,7 but in the United States the short e and i seem to be dominant in these words. There is also a tendency to substitute the short a of radish for the long a of made in data, vagrant, aviator, Danish and even radio;8 the short e of pen for the long e of scene in penalize, economics, detonator, scenic and electricity; the short i of sin for the long i of idea in sinecure; the short o of fog for the long o of bone in mobilize, soviet, choral and voltmeter, and the short u of sum for the long u of cube in quintuplet.1 In both English and American usage there is a strong movement toward substituting the so-called neutral vowel for clearer vowels in unstressed syllables, especially in colloquial speech. Thus the a of about, the e of the, the i of habit, the o of hillock and the u of upon are all reduced to a grunt that has given considerable concern to the phonologists, for it varies somewhat in different situations, and differs in many of them in England and the United States.2 In this country the schoolma’am, influenced by her veneration for spelling-pronunciations, makes war upon it, and I was myself taught in primary-school to enunciate the e of the clearly, making it identical with the ee of thee, but it is far too firmly lodged to be disposed of. The English, says C. H. Grandgent, prefer to omit the mauled vowel altogether, thus producing the collision fo
rms that Americans always notice in their speech.3

  There is no need to add much to the discussion of the individual vowels in AL4, pp. 334 ff. As we have already seen in the present book, the broad a did not begin to flourish in England until the Eighteenth Century, though it was, of course, used before then, especially in dialects. C. Cooper, whose “Grammatica Linguæ Anglicanæ” was published in 1685 and who has been described by Wyld as “by far the most reliable phonetician among the Seventeenth Century writers,” recorded the flat a of what is now General American not only in bath, gasp and path, but also in car, tar, quality, barge, carp, dart, larch and tart. To this day, in fact, that a is retained by the English in a large number of words – perhaps in quite as many as show the broad a that Americans think of as so characteristic of England. Examples are manse, fancy, pants, vassal, pantry, lass, crass, paraffin, pariah, can, mandate, mannequin, pamphlet, ant, ass (the animal), parasol, avoirdupois, bas-relief, candle, passenger, mammal, palate, parrot, saddle, latch, handsome, quagmire and passive. “Some English Roman Catholics, mostly converts,” says H. W. Seaman,4 “insist on a long a in mass,” but when the word is used to designate a quantity of matter the American a is used.1 Wyld says2 that the change from the old (and still American) flat a to the broad a of the English past, bath and after was still hanging fire in the early Eighteenth Century, and that it was “difficult for Englishmen at that time.” James3 ventures the opinion – “but that,” he adds cautiously, “is only an opinion” – that “in the end the short vowel [i.e., the American vowel] will prevail.” “In the Sixteenth Century,” he says, “[it] was universal.”4