American Language Supplement 2

by H. L. Mencken

  The Board on Geographical Names, save in very rare cases, does not concern itself with the pronunciation of place-names, and when other governmental agencies venture to do so it is seldom to much edification. The Legislature of Arkansas decided solemnly in 1881 that the name of the State “should be pronounced in three syllables, with the final s silent, the a in each syllable with the Italian sound, and the accent on the first and last syllables,”1 but it will be noted that the a’s in the second and third syllables, as one now hears them in Árkansaw, are actually anything but Italianate. Moreover, the name of the Arkansas river is the Arkánsas along its course through Kansas and so is the name of Arkansas City, which is in Kansas just over the Oklahoma line. “The Arkansaw Traveler,” the national hymn of the State, is always, however, Árkansaw, and evidence assembled by the DAE shows that the Indians who infested the region in the early days were called Arkansaws so early as 1772, and that Congress thus spelled the name in 1819. Indeed, there are Arkansawyers who argue spitefully that Kansas itself should be Kansaw.2 Meanwhile, Arkansaw for the State has been accepted by the British Broadcasting Corporation, though it makes the Kansas town Arkánsas City.3 The same authority now ordains O-high-o for Ohio and Massachóosets for Massachusetts, though the English schoolboys of past generations were taught to say O-he-ó and Massátchusetts.4 The Arkánsas pronunciation, so abhorrent to all patriotic citizens of the State, may have arisen by assimilation with that of Kansas, and perhaps it was helped on by the Eastern schoolma’ams who once tried to substitute Gloucéster for Gloster and Worcéster for Wooster.5 Similarly, the presence of an a instead of an i at the end of Missouri, Cincinnati, Miami,1 etc., is probably the end product of a schoolma’amish war upon an early tendency to turn every terminal a into y, e.g., Indiany, Ithacy, Floridy, Uticy, Alabamy, Caroliny, Susquehanny, Philadelphy, Nebrasky, sody, opery, asthmy, balony.2

  The pronunciation of Missouri has been under debate for many years and has produced a large literature, some of it marked by anything but scholarly calm.3 Allen Walker Read has published a characteristically comprehensive review of the whole matter.4 In that review he rejects the theory just mentioned, first launched by E. S. Sturtevant,5 that the -a-ending represents a fastidious effort to get rid of the apparent vulgarism of the -y-ending, though Sturtevant has since been supported by such accepted authorities as George Philip Krapp6 and John Samuel Kenyon.7 Instead he seeks an explanation in the disinclination of the carnivora of a proud and once bloodthirsty State to let it pass under a name which suggests a diminutive.8 But he overlooks the unchallenged presence of the same diminutive in Mississippi,9 one of the least infantile names on the American map, and in the names or pet-names of such testo-steronic towns as Boise, Idaho; Tulsy, Okla.;10 Hickory, N. C., and Corpus Christi, Texas. The early authorities show that Missouri, not Mizzoura, was first in the field, and that it apparently remained in favor until the Civil War era. At that time a craze for elegance seized the nascent intelligentsia of the State, and the -a-ending was urged upon the plain people with such fervor that the overwhelming majority of them adopted it and have continued to use it to this day.1 But in the early 90s or thereabout a new wave of pedagogues, chiefly, it would appear, from the East, launched a counterattack in behalf of the -i-ending, and a bitter battle was soon joined. The gogues might have had some chance of success if they had been content to argue only for the -i-ending; unhappily, they also tried to unvoice the two z’s in the middle of the name, and so convert the manly Roman sound of buzz and whizz into the puny Phoenician cheep of kiss and bliss. This was a fatal blunder, for even those Missouri sophisticates who were willing to accept -i revolted against -ss- in disgust and indignation. Today even the dictionaries and encyclopedias, which are usually at least a generation behind-hand, prefer the -zz- to the -ss-, and most of them have also surrendered to the -a-ending.2

  Three other conspicuous American place-names whose pronunciation has kicked up controversy are Iowa, Los Angeles and San Antonio. Iowa is borrowed from the name of an Indian tribe, and the history of that name was thus told by the late Dr. Frank H. Vizetelly:

  The Iowas were a branch of the main group of a southwestern Sioux tribe which received the name of Pahoja, or Gray Snow,3 which they retained, but they were known to the white people by the name of Ioways or Aiaouez. This name was spelled phonetically in both cases – one for the guidance of the English-speaking people, and the other for the guidance of the French-speaking people.

  In 1689 the spelling Aiaoua was that used in American literature, and is to be found in Perrot’s Memoirs, p. 196. The form varied at that time, and the plural was given as Aiaouais and Aiaouez. In 1702 Iberville used Ayooués (the acute accent gives the word the sound of e in they). By 1731 Ayoouais was used by Beauharnois and Hocquet, but varied in 1761 to Aiauway, a form to be found in the Journal of Lewis A. Clark. In the following century (1804) Aieways was used, and Aiowais was used by Pike in his travels (1811), but in 1810 we had loway as a variant. In 1824 Iowas and Iaways were used, and in 1825 Ihoway, which latter form is to be found in Senate Document No. 21 of the Eighteenth Congress, Second Session. The forms current in 1848 were Ayavais for the plural and minus the s for the singular. And loway. By 1858 the name took the form Ayeovai in the singular, with an s in the plural, and in 1905 Iaways.1

  Ioway seems to be preferred by the plain people of the State, and the name so appears in the State song, but the Geographical Board long ago declared for Í-o-wa, and it is supported not only by the majority of outsiders but also by a formidable faction within the State. There is even a body of opinion in favor of putting the accent on the second syllable, but it is apparently feeble. Nearly all the accepted authorities ordain I-o-wa, but the Thorndike Century Senior Dictionary2 gives I-o-way as an alternative, and the Funk and Wagnalls New Practical Standard Dictionary3 notes it as used locally. The controversy over the pronunciation of Los Angeles has been going on for years, and will probably never end. As I noted in AL4, p. 542, the Los Angeles Times, the chief guardian and glory of the local culture, advocates what it renders in type (somewhat darkly) as Loce Ahng-hayl-ais, but a dozen other forms are to be encountered in the town. The gabbers of the British Broadcasting Corporation are instructed to use Loss Anjileez4 and those of the National Broadcasting Company are encouraged to attempt Lawsan-j’l-uhs,5 but most other authorities wobble. Even Frank H. Vizetelly, usually so sure of himself, could not decide between Los Anggeles, with the o as in go, and Los Anjuhliz, with the o as in not,6 and one local expert has actually recommended different pronunciations in successive editions of the same book.1 Webster 1934 gives both Los Anggeles and Los Anjeles, the New Practical Standard gives Los Anjilus, Los Anjileez and Los Ang-guh-lus, Robert Shafer gives four forms in a phonetic alphabet that I refuse to try to interpret,2 and Kenyon and Knott give half a dozen pronunciations without deciding between them, and record despairingly that “a resident phonetician says: ‘The only one I’ve never heard is Los Angheles.’ ”3 Many of the inhabitants abbreviate the name to L.A., and others use Los.4 The controversy over San Antonio has to do with the question whether the people there ever call the town Santóne. The Hon. Maury Maverick, its best-known citizen, assures me that they do not. “The average person,” he reports, “says Santónyo, although the well-informed generally say San An-tón-i-o. The Mexicans, of whom there are some 90,000 in San Antonio, pronounce the word with a broad a and sound every syllable.”5 But other observers insist that Santóne is in common use, and Charles H. Hogan says that the first syllable is frequently stressed.6 Dr. E. G. Reuter, a resident, tells me that he has also heard San Antone.7

  As I have recorded, the Geographical Board follows local usage, whenever possible, in the spelling of place-names. The same policy with respect to pronunciation was advocated by Noah Webster so long ago as 1803.8 “The true pronunciation of the name of a place,” he said, “is that which prevails in and near the place.” But he defended and indeed advocated changes in Indian names to bring them in accord with “the g
enius of our language, which is accommodated to a civilized people.” This is the line taken by Allen Walker Read in his excellent review of the subject.1 He rehearses and answers the argument that the pronunciation of place-names should follow the spelling or be decided by self-constituted authorities on language or by legislative action, and concludes that it should be determined “simply by impartial observation of selected speakers in the locality of the places named.” Even this rule, of course, cannot be followed slavishly, for sometimes there is a sharp difference of opinion, even among natives, as to the true pronunciation of a given name, and more often an accepted local form is challenged by another prevailing somewhere else, or even generally. We have just seen examples in the cases of Arkansas, Los Angeles and Iowa, Another is provided by Chicago, which is Chicawgo in the city itself but Chicahgo in the rest of the country.

  Read raises the question as to what is to be done about a name which, while designating the same place or object, differs in pronunciation in two regions, e.g., that of the Arkansas river. Is it, he asks, “to change its name for different parts of its course? ” Well, why not? If the people of Kansas prefer Arkánsas for that part of the river within their boundaries, then it is the Arkánsas there, and if those of Arkansas prefer Árkansaw for their own share of it, then it becomes the ÁRkansaw as the border is crossed. The name of Beaufort is pronounced differently in the two Carolinas – and both forms are “correct.”2 The best pronouncing dictionary of American place-names, that of Alfred H. Holt, first published in 1938,3 avoids all vain speculations as to how names ought to be pronounced, and is content to record accurately how they are pronounced. Holt’s authority is always local, and he is careful to find out if there are variations on different cultural levels. His method of indicating pronunciations is so simple and effective that it puts to shame the phonetic alphabets invented by learned men, and he relieves the tedium of what might otherwise be a depressing subject by cracklings of pawky humor.

  The study of place-names is comparatively recent in the United States. Washington Irving printed some observations upon them in the Knickerbocker Magazine in 1839 and Henry R. Schoolcraft discussed the Indian names of New York before the New York Historical Society in 1844,1 but it was not until 1861 that a separate work upon the subject appeared. This was a pamphlet of thirty-two pages by a surgeon named Usher Parsons, entitled “Indian Names of Place in Rhode-Island.”2 Parsons’ interest was mainly in the Indian names in use in the State “when civilization commenced” and his stated purpose was to provide a supply “for the convenience of those who may hereafter wish to apply them to their country villas, factories or institutions,” but within those limits he made a good job of it, and it was more than thirty years before a better study of Rhode Island place-names appeared. In 1870 James Hammond Trumbull followed with a work which remains one of the classics of the subject,3 and by the 80s Henry Gannett had begun an investigation which was to result in gazetteers and “geographical dictionaries” of Connecticut, Massachusetts, New Jersey, Rhode Island, Kansas, Utah, Texas, Delaware, Maryland, Virginia, West Virginia, Colorado and the Indian Territory, and an omnibus volume entitled “The Origin of Certain Place Names in the United States.”4 It was this book which set the study of American place-names on its feet.

  Unhappily, that study is yet to be organized on a national scale, and in consequence the work done so far is spotty and incoördinate.1 In some States, e.g., Oregon, Arizona, South Dakota, California and Missouri, the record is substantially complete, and in others, e.g., Pennsylvania, Nebraska, Minnesota, Wyoming and West Virginia, admirable progress has been made, but in yet others, e.g., Maryland and New Jersey, there is hardly a beginning. The excellent discussion of place-names in George Philip Krapp’s “The English Language in America,” published in 1925,2 awakened new interest in the subject, and in 1927 Dr. Robert L. Ramsay, professor of English at the University of Missouri, undertook a systematic effort to investigate the nomenclature of that State. He had the invaluable assistance of Allen Walker Read, then an instructor on his staff, and Read presently launched the enterprise with a paper in the Missouri Historical Review.3 Thereafter, Ramsay’s graduate students began a study of Missouri place-names that went on until 1945, by which time eighteen reports, covering all the 114 counties of the State, had been completed. The material thus accumulated, running to 32,324 names, was entered upon three sets of cards, one of which was deposited with the Board of Geographical Names in Washington. At the beginning of this huge enterprise Ramsay published an “Introduction to a Survey of Missouri Place-Names” which amounted to a treatise upon the whole technic of place-name research, and remains the best American handbook on the subject.4

  Meanwhile, Lewis A. McArthur, a Portland business man, had been carrying on an inquiry into the place-names of Oregon, and in 1928 there appeared the first edition of his “Oregon Geographic Names,” a truly admirable work.5 It is not based upon local tradition alone, but represents a diligent and thorough examination of all the available records, and shows an extraordinary resourcefulness. How much time and labor went into it is beyond estimate, but certainly very few works of American scholarship have ever enlisted more.1 The only compilations that rival it and Ramsay’s in scientific value are “West Virginia Place Names,” by Hamill Kenny;2 “Arizona Place Names,” by Will C. Barnes;3 “South Dakota Place Names,” sponsored by the department of English of the State university,4 and “California Place Names,” by Erwin G. Gudde.5

  The literature of the subject is extensive, but a large part of it is concealed in pamphlets and papers published locally, and hence difficult of access. A bibliography listing 195 titles was presented to the Modern Language Association in December, 1938, by Harold W. Bentley and M. Robert Snyder, and since then Richard B. Sealock, librarian of the Gary, Ind., Public Library, and Pauline A. Seely, of the Los Angeles County Public Library, have completed a much more extensive one covering the United States, Alaska, Canada and Newfoundland. It is in press as I write.6

  There is room here for no more than brief notices of the more important contributions to the subject. To take them alphabetically by States, the place-names of Alabama have been reported on by Peter A. Brannon, Frank H. Elmore and others in various issues of Arrow Points, the monthly bulletin of the Alabama Anthropological Society, and the Indian names among them have been studied at length by William A. Read7 and H. S. Halbert.8 The place-names of Alaska were first investigated by Marcus Baker, whose report was published by the Geological Survey in 1900.1 In 1940 John Drury Clark and L. Sprague de Camp printed a brief but illuminating note upon them, listing about seventy,2 and three years later the WPA brought out a report on the geographical nomenclature of the coastal areas.3 The map of Alaska is sprinkled with names borrowed from the Eskimos, the Russians and various Indian tribes, and includes such strange forms as Aniahwagamut, Metlakatla, Ouzinkie, Egegik, Chickaloon, Iditarod and Bogoslof. Clark and de Camp show that there is a tendency among the American settlers to give English values to vowels encountered in print and to place accents according to English patterns. “A feature of Alaskan idiom,” they say, “is the use of the names of rivers, without the word river, to indicate general regions, or, more accurately, watersheds. Sometimes the word country is added; sometimes not.… A man will say ‘I’ve been up in the Chandalar,’ meaning the Chandalar river watershed.”

  Will C. Barnes’s “Arizona Place Names,” before mentioned, is a painstaking and excellent compilation; indeed, it comes close to exhausting its subject. Barnes complains of a tendency common all over the country, to wit, to ascribe to “Indian origin” every place-name for which no other etymology is available. There are, he says, more than twenty tribes in the State, and running down such vague ascriptions is seldom rewarding. He finds some mellifluous names of Spanish origin, e.g., Huerfano (orphan), Nogales (walnut), Tortolia (little dove) and Tinja (an earthen water-jug), and others authentically Indian, e.g., Huachuca (Chirchua-Apache: thunder), Vekol (Papago: grandmother), Topock (
Mohave: bridge), Parish-hawampitts (Piute: boiling water), Cochibo (Papago: kochi, pig, and bo, pond), and Shato (Navajo: sunny side), but most of the pioneers were matter-of-fact men and were content with such banal names as Milltown, Maryville, New London, Shultz, Sunset City and Smithville, with an occasional ascent to Grasshopper, Frog Tanks, Tombstone and Total Wreck.1 In Arkansas the place-names of French origin have been studied by Branner and Renault,2 and there is a discussion of the State names in general by Fred. W. Alsopp.3

  Those of California, now dealt with at length in Gudde’s study, before mentioned, have been listed by many other fanciers, beginning with C. M. Drake.4 As everyone knows, the map of the State is adorned with a large number of charming Spanish names, many of them worn down considerably by American speech habits, e.g., Los Angeles, but the rest fairly well preserved. Walt Whitman, in “An American Primer,” objected to the saints’ names among them on the ground that they had “a tinge of melancholy and of a curious freedom from roughness and money-making” and hence knew nothing “of democracy – of the hunt for gold leads and the nugget or of the religion that is scorn and negation.” “Chase them away,” he exclaimed, “and substitute aboriginal names.” Fortunately, his fatuous counsel went unheard for years and when it was heard at last it was not heeded.5 The sonorous Spanish names were described romantically in 1914 by Nellie Van de Grift Sanchez6 and more soberly by various other writers afterward, e.g., Joseph B. Vasché,7 Gertrude Mott,8 Martha A. Marshall,9 Archibald A. Hill,10 Laura K. McNary1 and Robert Shafer.2 Most of these authors make some effort to expound the true pronunciations, but usually without much effect, for there is a difference of usage and opinion in and about many names beside Los Angeles. The commonest error says Vasché, is the substitution of the a in cat for the Spanish a in father, but almost as bad are the substitutions of the a in father or the u in up for the Spanish o and of the i in it for the Spanish long l, as in machine. The California place-names of Indian origin have been studied by the anthropologist, A. F. Kroeber.3 One of them, Pasadena, is commonly assumed to be Spanish, but it actually comes from a Chippewa term said to mean “crown of the valley” and was suggested to the first settlers by a missionary to Missouri who was a relative of one of them.4