American Language Supplement 2

by H. L. Mencken

  1 Middle English ū and Related Sounds: Their Development in Early American Speech, Supplement to Language, Oct.-Dec., 1939. Some objections to his conclusion are stated in Development of Middle English u, by Eilert Ekwall, American Speech, Oct., 1940, pp. 306–10.

  2 p. 159.

  3 The English Language in America; New York, 1925, Vol. II, pp. 155–64.

  4 A History of Modern Colloquial English; London, 1920, pp. 242–44.

  5 Modern English Usage, lately cited, pp. 335–36. This discussion was first published in S.P.E. Tract No. XXII; Oxford, 1925, pp. 47–48.

  1 Vol. II, pp. 163–64.

  2 The NED shows that in the Fourteenth Century, when the word was just coming into English, both forms were in use in England. Luftenand and leeftenaunt are cited from 1375 and 1387 respectively. But lutenand also dates from c. 1375.

  3 Lif, Lef, or Loo, Edinburgh Evening News, Sept. 27, 1943.

  4 London, 1917; revised in 1924, 1926 and 1937.

  5 Transactions of the Albany Institute, Vol. X.

  1 p. 109.

  2 Quoted in Godey’s Lady’s Book, Jan., 1854, p. 86, with the implication that American actors were doing likewise.

  3 The English Language in America, Vol. II, p. 207. An excellent discussion of the history of the sound follows.

  4 Dr. L. L. Barrett tells me that he once knew a professor at the University of Virginia who even introduced it into French, as in regyarder for regarder.

  5 Southern English, in Culture in the South, edited by W. T. Couch; Chapel Hill (N.C.), 1934, p. 610. See AL4, p. 364.

  6 All of us, of course, insert y before the u in union and the like, but yarb for herb and yere for here are definitely rustic and vulgar. Wyld shows that yere for here and yearth for earth were used by Bishop Hugh Latimer in the sermons he published in 1549, and that yearl for earl was in use at the same period. It was probably a fear of such vulgarisms that knocked the y off yeast. It was pronounced east in Maryland in my boyhood, but the y has now been restored.

  1 See AL4, pp. 345–46.

  2 I take this from A Modern English Grammar, by Otto Jespersen; Heidelberg, 1922, Vol. I, pp. 329–30. Isaac Watts, in The Art of Reading and Writing English, 1721, listed jice as correct for joist, but Robert Nares in Elements of Orthoepy, 1784, denounced hist for hoist as a “low vulgarism.”

  3 Normal English Pronunciation; London, 1937, p. 91.

  4 Mr. William R. Bradley, of New York City (private communication, Dec. 1, 1943), sends me a tearsheet from Science containing an article entitled Medical Orthoepy, by Dr. B. N. Craver, of the School of Medicine, Wayne University, Detroit, which says in part: “A very common error in the pronunciation of medical terms is to render as diphthongs vowels which should be sounded separately.… Thus, protein, correctly a three-syllable word, has been accorded but two; so also with caffeine, rabies and others. However, for such words as oubain, sparteine, codeine, caries, facies and others correct speech demands the pronunciation of all three syllables. Syndrome, analogous to epitome, should have all vowels sounded, but it has so long been mispronounced as a two-syllable word that lexicographers remark that pronunciation in medicine.”

  5 There was a contrary exchange in the Philippines after the promulgation of civil government on July 1, 1902. Thereafter the refined began speaking of the days before that event as those “of the empire,” but the less tutored resident Americans preferred umpire. I am indebted here to Mr. Hartford Beaumont.

  1 Chaw must have been accepted in England during the Eighteenth Century, for Walker in 1791 described it as having “grown vulgar,” apparently recently. It was permitted by Nares in 1784.

  3. THE CONSONANTS

  “In London and some parts of the South [of England],” said R. J. Lloyd in 1894, “the r following a vowel at the end of the word or syllable has disappeared, but there is no other part of the English-speaking world except Eastern New England where this is quite the case.”2 Lloyd might have excepted also the Tidewater South, but everywhere else in the United States, including even the Hudson Valley area, the r is usually sounded. The late C. H. Grandgent of Harvard (1862–1939) once estimated that, in the West, it appears before consonants, as in card, north, part and farm, 81 times out of 100, in the Middle States 64 times, in New England 36 times, and in the South 24 times.3 Bernard Bloch, one of the collaborators in the Linguistic Atlas of New England, has since shown that it is now conquering even New England. In the Western third of the area he has found it prevailing in more than 75% of the cases, and even within the Boston territory there are speechislands in which it is clearly sounded. The older speakers, he says, still omit it; the young ones insert it. Its eastward extension, he concludes, “reflects not merely the spread of a single feature from Southwestern New England, but a gradual victory of the chief type of American English over a specifically provincial dialect.”1 There are, to be sure, some neighborhoods in which a contrary tendency seems to be showing, but Bloch inclines to think that even there the r will finally conquer.

  Archibald A. Hill has shown that the loss of r after vowels and before consonants is frequent in the English dialects,2 and has produced examples from as long ago as the Fifteenth and Sixteenth Centuries, e.g., hoss for horse (1473–88), assenycke for arsenic (1530) and cott for court (1552).3 Wyld says that it was lost earliest before s and sh.4 It is sounded in England when it is followed by a vowel, as in red, ride and rode, but it is omitted when it stands at the end of a word, as in car, fair and fur. In the latter case, however, it is commonly restored when the word following begins with a vowel, as in “The car is at the door.”5 But this restoration is not invariable, and there are situations in which many speakers seem to find it difficult to decide whether they should sound the r or not. As a result, some of them, eager to be correct, insert it where it has no place, as in “the idear is” and “vanillar ice-cream.” This confusion is promoted by the fact that many quite dissimilar words, e.g., law and lore, are pronounced precisely alike in Southern English. People so afflicted, says A. Lloyd James, “will not infrequently talk about the lore of Moses when they mean the law of Moses.”6

  The best discussion of the American r that I am aware of is in a paper by John S. Kenyon.7 He describes at length the vocal mechanism whereby the various r-sounds, ranging downward from the trilled r of German, French and Scots, are produced, and distinguishes between the mere muffling of the sound and its complete extinction. In many cases, in the middle ground, he says, r is reduced to a sort of vowel. He observes, like Ring Lardner before him,1 that the literary custom of representing the vulgar pronunciation of fellow, window and their like by feller, winder, etc., is misleading, for the final syllable in most cases does not show r at all, but is simply the neutral vowel.2 Why has r survived in General American? Kenyon rejects the theory that the schoolma’am, egged on by Webster, preserved it by insisting on spelling-pronunciations, and points out very wisely that there were more of her clan in Eastern New England, where it vanished, than to the westward, where it persisted. He rather inclines to believe that the character of the immigration into the West was mainly responsible. It was largely made up of Scotsmen, of Irishmen and of Englishmen coming from regions outside the influence of London speech, and “they brought their r’s with them.” “There is much reason to think,” he concludes, “that the Western treatment of r … is parallel to the Western pronunciation of words like half, which belongs to an ‘older family’ than Eastern hahf.” Even more than the use of the flat a, the sounding of r is the chief hallmark of General American speech; indeed, Leonard Bloomfield says that this General American, or, as he calls it, “Central-Western type of American Standard English,” may be defined as “the type which preserves old r in final position and before consonants.”3 Its sound, one may admit without cavil, is very far from lovely,4 but as the late Frank H. Vizetelly was fond of pointing out, it at least makes for intelligibility.5 – and the desire to convey ideas is the chief purpose of speaking at all.1

  The dro
pping of the final g in words ending in -ing seems to be more widespread in England than in America, and is tolerated, if not exactly recommended, by most of the English authorities on speech. Kenyon says2 that, in the United States, it “appears to be more common among the educated in the South than in the North and East.” “The spelling-pronunciation,” he goes on, “is now so general that it is in excellent usage, but it must not be hastily concluded that the pronunciation -in is necessarily a mark of ignorance or lack of cultivation. It is still commoner than most people suppose. It is a good illustration of the muddling through by which forms and usages regularly become established in standard use. Hundreds of people have religiously practised saying coming instead of comin without ever intelligently considering the facts, or whether the effort was worth while.” Krapp shows3 that -in is to be found plentifully in the early American records, and that it must have been general in the Seventeenth Century. He ascribes the prevalence of -ing to the rage for spelling-pronunciations, and notes that, among the innocent, “the analogy of words” has produced such forms as kitching and garding. Lardner noted that in the common speech the final g is commonly dropped in nothing and something, but retained in anything and everything.4 Lieut.-Col. F. G. Potts5 accounts for this on the plausible ground that “anything and everything have strong secondary accents on their last syllables, and are pronounced as if those syllables were separate words.” Americans, like Englishmen, seldom give a clear sound to the g in such words as length and strength. It becomes, at best, k, and this sound is recorded without comment by Jones in his “English Pronouncing Dictionary.” Dr. Alfred D. Schoch argues that this substitution is quite rational. “The g in these words,” he says, “is only an orthographic expedient; they really have no g in them – that is, no g sound. What they do have is a velar nasal consonant like the n in ink, which stands between two other sounds that are articulated forward in the mouth, and so stands to have its articulation shifted forward to the n-position. I don’t remember, though, that I have ever heard these words with a plain n-sound. What is more likely to occur in ordinary talk is that the ng nasal consonant may disappear and leave the nasality of the e to take its place.”1

  It is as grievous to an Englishman of tone to be accused of dropping his h’s as it is to a white Southerner to be accused of using you-all in the singular. Nevertheless, both are guilty to some extent. Daniel Jones, in prescribing the usages of what he calls Received Pronunciation (RP), lists hospital with a clear h but allows both hostler and ostler, and when he comes to hotel says that “some use the form otel always; others use it occasionally, when the word is not initial.”2 All Americans believe that they sound the h invariably, but when they use an before hotel, which happens sometimes, they actually say an ’otel, for sounding h after consonants is phonologically unhandy, as the cases of on his and to kiss (or neck, or shoot) her sufficiently show. They are saved from cockneyism by the fact that, as a practical matter, they seldom use an before hotel and its allied words, despite the assumed influence on their speech habits of the King James Bible, which gives it before haven,3 hair,4 host,5 hedge,6 helmet,7 herb,8 hidden,9, high,10 hand,11 hole,12 holy,13 horn,1 horse,2 house,3 householder,4 hundred,5 hypocrite, etc.6 In late years the more popular American prints of Holy Writ have quietly substituted a for an before all these words, though in a few aberrant cases an is retained. In the only instances in which hen, hind, hot and huge appear in the text with an indefinite article the King James Version itself uses a, which is also used before horrible. Hunt never appears as a noun and hurt never with an indefinite article.

  It is hard to make out whether the use of a or an influences the pronunciation in more cases than the pronunciation influences the use of the articles. So far as I know the only study of English and American practise has been one reported by Louis N. Feipel in 1929. He investigated three hundred contemporary books by authors of decent standing on both sides of the water and found that the English used an much oftener than the Americans. In the case of hallucination, for example, the score ran three to one, in that of horizon four to one, and in that of hysterical five to one. An historical was found in six English writers and four Americans, but a was likewise used by six Englishmen and four Americans, which left a sort of stalemate. An heroic was used by eight Englishmen and four Americans – including Waldo Frank and Lewis Mumford –, but two Americans used a heroic, which all the English avoided. No author of either country used a hour, a heir or a honest, which seems to indicate that the h is dropped in all of them in both countries. No American used an hereditary, an hermaphrodite, an hermetically, an hydraulic, an hyena, an herculean, an hypnotic, an hypocrisy or an hilarious, but a few Americans used an heraldic, an habitual and an hiatus. Such an one was used by twelve Englishmen and seven Americans, and such a one by six of the former and four of the latter.1 Mark Twain was noting these differences in 1879, when he said to an Englishman encountered on a German train:

  Your educated classes say humble now, and heroic, and historic, etc., but I judge that they used to drop those h’s because your writers still keep up the fashion of putting an an before those words, instead of a. This is what Mr. Darwin might call a rudimentary sign that an was justifiable once, and useful – when your educated classes used to say umble, and eroic, and istorical.2

  “In the American pronunciation,” wrote Noah Webster in 1789, “h is silent in the following: honest, honor, hour, humor, herb, heir, with their derivatives. To these the English add hospital, hostler, humble. But an imitation of these, which some industriously affect, cannot be recommended, as every omission of the aspirate serves to mutilate and weaken the language.”3 Perhaps the best present-day American practise is set forth in the Style-Book of the Atlantic Monthly, as follows:

  Before words beginning with h use a with monosyllables and words accented on the first syllable: a hat, a habit, a hurricane. In such cases one bears heavily on the aspirate, so that it is equivalent to a consonant. Before polysyllables accented elsewhere than on the first syllable use an: an habitual, an historical, an heretical. In such words the h is naturally so slurred in pronunciation that its presence is scarcely apparent, and a distinct effort is required to pronounce it distinctly, as one must if a is used before it. With those words beginning with hu in which the combination is pronounced almost like yu, a should always be used, without regard to the accent: a humane, a humility.4

  The leading English authority, F. Howard Collins’s “Authors’ and Printers’ Dictionary,”5 ordains a before hope, horse, hospital and humble, and also before honorarium, which last is somewhat puzzling, for Thomas R. Lounsbury, in 1904, listed honor as one of “four words beginning with h in which the initial letter is not pronounced by educated men anywhere,” the others being heir, honest and hour. “This usage,” he said, “extends of course to their derivatives.”1 “Whether,” he went on,

  they will continue to hold out forever against the stream of tendency which is bringing about the resumption in speech of letters once silent must be left to the prophets to announce. In this instance their predictions can be uttered with perfect safety. None of those now living will survive to witness their fulfilment or non-fulfilment. So far no one has ever advocated the pronunciation in them of the initial letter save Walter Savage Landor. He may have been led to take this course by the irritation he felt at having his own usage criticized, for when he came to the employment of the h he is reported to have frequently exhibited distinct orthoepic frailty.2

  When it appears in any save the initial position h is frequently dropped, even by speakers of General American. No one, for example, sounds it in exhaust and exhort, and many also omit it in exhibit.3 The English long ago dropped it from forehead, which is forrid or forred in their speech. There was a time when they also dropped it in blockhead, hothouse, hedgehog, greenhouse, abhor and adhere.4 The compensatory insertion of h in situations where it does not belong is purely dialectical in English and does not occur in Standard Southern English. In American it is
quite unknown, save only in such vulgar forms as hit for it and overhalls.1 But Americans sometimes retain the h where English usage does not sound it, especially in proper names, e.g., Northampton.2

  The elision of other sounds from vulgar American speech is discussed in AL4, pp. 352–54. Such forms as bound’ry, comf’table, and prob’ly are fit matches for the English secret’ry and extr’odn’ry. The sound most often dropped is that of medial r, and the late George Hempl (1859–1921), professor of Germanic philology at Stanford University, long ago published a formidable list of examples, e.g., pa’tridge, su’prise, qua’ter, co’ner, the’mometer, pe’formance, lib’ary, yeste’dy,3 sa’sparilla, pu’sy (for pursy, usually encountered in pussy-gut), pa’lor and Feb’uary.4 Some of these are to be found also in English usage. Other sounds that are likewise dropped on occasion are those of k, as in e’cept;5 n, as in kill for kiln;6 th, as in scythe;7 l, as in a’ready and cert’n’y;8 v as in fi’ cents;9 d, as in We’nesday, kin’ness and tole; t, as in of’n,10 apos’le and Chris’mas, and s as in some of the almost innumerable deteriorated forms of yes.1 Sometimes, in careless speech, one consonant is substituted for another, as in grampa and robm (robin), or two for two, as in sebm (seven);2 sometimes a cluster of consonants is omitted, as in gra’ma; and sometimes there is elision of a combination of consonant and vowel, as in pro’bition, guv’ment and o’n’ry.3

  Hilaire Belloc, in 1924, alleged that th, in American speech, was becoming d, “even in carefully pronounced words, traditional and in the mouth of a highly-educated man.”4 “The,” he went on, “has not yet become de, but it is on the way.” This change, so far as I know, has never been acknowledged by any American phonologist, but many of them have studied the parallel change of t to a kind of d that they call the voiced t, as in water, butter, battle, twenty, etc. An Englishman commonly pronounces the t in pity clearly, but in colloquial American speech the word often comes close to piddy. This voiced t, according to Kenyon,5 “occurs most commonly between vowels, sometimes between a vowel and certain of the voiced consonants when it is at the end of an accented syllable before an unaccented one (twenty), or sometimes when it is at the beginning of an unaccented one where there is some doubt which syllable the t is pronounced with (want to go).” It also occurs between two unaccented syllables, as in join us at eleven. It does not occur at “the beginning of syllables initial in the phrase, whether accented or unaccented (table, today), nor at the end of syllables final in the phrase, whether accented or unaccented (repeat, rivet), nor at the beginning of accented medial (Miltonic) or final syllables (retain).” In 1942 Dr. Victor A. Oswald, Jr., of Columbia, made an attempt to discover the extent to which this d-like sound was substituted for t in ordinary literate speech, using students of the Hazleton, Pa., Senior High School as laboratory animals. He found that in bitter, betting, plotting, and sorted the overwhelming majority of them sounded a clear t, but that in bleating, waiting, hearty, hurting and writing most of them used a consonant that sounded like d.1 An ingenious correspondent2 tells me that he hears this sound even in street: “most Americans say sdreet.”3 He also hears g for k in score and b for p in sponge. The use of s for sh before r, noted in many English dialects, seems to be common in the South. It was denounced by a Baltimore orthoepist, so long ago as 1856, as “the affected pronunciation of over-refined school-girls who cannot bring themselves to utter the homely English sound of sh when combined with an r, for fear apparently of distorting their faces,”4 but it survives below the Potomac in speakers of all ages, though it is far from universal. Sir Richard Paget believes that both s and sh, along with f and the sound of th in both, should be “thrown out of our language” and replaced by their voiced equivalents, to wit, z, zh (as in pleasure), v and the th of thy. “These unvoiced or whispered sounds,” he says, “are in every way inferior to the voiced sounds – as inferior, in fact, as whispered speech is to voiced speech. Their carrying power is one-tenth to one-twentieth of that of the voiced sounds; they are incapable of being sung or of carrying vocal inflection – they are the prime cause of all verbal misunderstanding. On the telephone they are practically inaudible.”1