Tin Pan Alley's Contribution to Folk Music
by Norman Cohen
Western Folklore, Vol. 29, No. 1 (Jan., 1970), pp. 9-20
Tin Pan Alley's Contribution to Folk Music
NORMAN COHEN
University of California, Los Angeles
(Introductory Note:) A preliminary version of this paper was read at the Seventeenth Annual Meeting of the California Folklore Society (Northridge, April 1969). I am, as usual, grateful to my wife Anne for many helpful discussions of the contents of this paper. D. K. Wilgus and Albert Friedman were kind enough to read it and offer their criticisms and comments. I am indebted to Joseph C. Hickerson of the Library of Congress Archive of American Folk Song for continual aid in ferreting out early sheet music references to the pop songs that have entered oral tradition.
--------------------------------------
SEVENTY YEARS AGO, in an address at the first meeting of the English Folk Song Society, Sir Hubert Parry declared:
there is an enemy at the doors of folk music which is driving it out, namely, the common popular songs of the day; and this enemy is one of the most repulsive and most insidious. If one thinks of the outer circumference of our terribly overgrown towns ... where one sees all around the tawdriness of sham jewellery and shoddy clothes, pawnshops and flaming gin-palaces... all such things suggest to one's mind the boundless regions of sham. It is for the people who live in these unhealthy regions who think that the commonest rowdyism is the highest expression of human emotion; it is for them that the modern popular music is made, and it is made with a commercial intention out of snippets of musical slang. And this product it is which will drive out folkmusic if we do not save it....[1]
Later in the evening at that same meeting, Mrs. Kate Lee, one of the founders of the Folk Song Society, delivered a paper on her experiences as a folksong collector. At one point she recalled having heard one singer tell her he knew a song entitled.
The Wreck of the Princess Royal. I did not take down this song, because the title had a modem sound, but I afterwards found out that I had made a great mistake, it being a well-known folksong sung by sailors. Moral: You should never go by titles of songs.[2]
It is seventy years too late to suggest to Mrs. Lee that there might have been another moral to draw from her experience, and had the views of these early pioneers of folksong study died with their generation, there would be little purpose in exhuming them at this late date. This has not been the case, however. Denigrations of urban or modem or sentimental or parlor songs still accompany discussions of folksong. Of course, not all folklorists exhibit discomfort in the presence of ink smudges from commercial sheet music. One needs look no further than Herzog's definition of folksong to realize that some scholars have readily accepted the importance of urban influences on folk music.[3]
I am not now concerned with Sir Hubert's fear that urban music would drive out pure folk music; seventy years of fruitful collecting render that prophecy nugatory. What I do propose to discuss is the nature and extent of the popular urban song material which grew in a specific period in American history and which has entered folk tradition, and also how scholars have treated this material. Once I have demonstrated that this is not an insignificant portion of folksong, I can turn to more speculative questions: Why did this material enter folk tradition? and, What will be the future relationship of urban pop music and folk music?
The period of American pop music I wish to consider lies approximately between 1860 and 1910. Part of the industrialization and urbanization that followed the Civil War was the rise and development of a national popular song business. The dissemination of pop songs in this five-decade span of popular music was characterized by the predominant role of the theater (including vaudeville and variety) and sheet music. In earlier decades, the distribution media were primarily the broadside, the street hawker, the songbook, and the tavern busker.[4]
The use of sheet music rather than broadside, which typically printed only words and no tune, meant that text and tune were born and disseminated together. The usual broadside instruction "to be sung to the tune of... ," ignored the possible currency of more than one tune and thus was very kind to oral tradition. The street hawker peddling his own compositions usually sang the tune several times for the purchaser to memorize-a practice that encouraged the rapid separation of original text and tune. Sheet music, on the other hand, left nothing to chance. For this reason, as I have noted elsewhere,[5] they form a valuable corpus with which to work. For no other body of folksongs do we have both original text and tune at hand; thus we can study the parallel effects of oral transmission on words and melody. No hypothetical archetypes need be constructed; the material is all there.
The music of 1860 to 1910 was hardly a homogeneous product. The 1860s and 1870s, once preoccupation with the war had diminished, abounded with the type of verse George S. Jackson has called "Della Cruscanism" -the silly and superficial creations of female poetasters who dominated American literature in the early nineteenth century.[6]
Most of the songs of this era have been forgotten, but a few have achieved remarkable lasting popularity among the folk, such as "Wildwood Flower" (originally "I'll Twine Mid the Ringlets" [1860]) and "Little Rosewood Casket" (1870). As the 1880s wore on, the narrative ballad began to supplant the lyric song. This was the era of Tin Pan Alley.
Prior to 1880 music publishers were scattered throughout the country. In the years that followed, however, Union Square in New York gained ascendancy as the locus of song publishers-in effect, a latterday Seven Dials. The magnet which drew them there was the presence of a major entertainment center boasting music halls, theaters, dance halls, and burlesque houses. During this decade several publishers discovered that songs could be marketed like any other commodity. This meant manufacturing them to meet prevailing taste and "plugging" them, that is, prevailing on the singers to feature them in their public performances. In the next decade theatrical activity moved further uptown and the music publishers followed it, making Twenty-eighth Street the center of the music-publishing world for a quarter-century or so. This was the street that was dubbed "Tin Pan Alley" by journalist-songwriter Monroe Rosenfeld about 1903 in an incident which has been recounted in numerous variants.[7]
Most of the songs discussed in this paper came from the Union Square and Tin Pan Alley periods. However, I shall use the term Tin Pan Alley (TPA) in a broad sense to include products of both the Union Square and Twenty-eighth Street periods, as well as of the late 1860s and 1870s when mass circulation of sheet music was just becoming important.
The post-Civil War years were still witness to the popularity of the blackface minstrel shows. But I have excluded from consideration all pseudo-Negro songs of that provenance because they are of a different style even though chronologically they overlap with songs included in this discussion. [8] My purpose is to focus attention on songs that were indisputably of professional origin and written for a mass market with little or no folk background. Many of the songs of the minstrel shows were drawn from an earlier folk repertoire. Similarly, hymns and gospel songs and other folk-derived material, such as some of the Irish music hall pieces (e.g., "Boston Burglar") are likewise excluded.
Two important developments early in the twentieth century justify a closing date of 1910 for this survey. The first was the rise of the phonograph industry, which radically altered the means by which songs were popularized and disseminated. The second was the publication in 1912 of Irving Berlin's "Alexander's Ragtime Band," which changed significantly the main course of pop music. Critics and social commentators of the day unleashed their scorn and fury against the dance craze and its attendant wave of snappy tunes with meaningless lyrics, using language reminiscent of Sir Hubert's condemnation of urban music quoted earlier. [9]
Edward B. Marks, the publisher and song-writer, summarized the changing taste succinctly: "The public of the nineties had asked for tunes to sing. The public of the turn of the century had been content to whistle. But the public from 1910 on demanded tunes to dance to." [10] There were of course forewarnings of this trend prior to Berlin's smash hit. Throughout the 1890s a scattering of danceable pseudo-Negro songs, designated in the parlance of the times "coon songs," followed by the more genuinely Negro derived ragtime and jazz melodies, caught the public ear even while the sentimental ballads were at their peak.
In summary, if we ignore the evanescent products of the Della Cruscan genre, the period 1860 through 1910 was bounded at either end by periods of strong Negro influence in American pop song-influence that made rhythmic elements the primary ones. Songs such as "Buffalo Gals" or "Old Dan Tucker" from the earlier, minstrel period, and "Preacher and the Bear" or "Oh, Dem Golden Slippers" from the later coon song period, easily found their way into the academic collections. The other musical style-the sentimental parlor ballad-which features the narrative and/or melodical element rather than the rhythmic one-has been given short shrift in folksong scholarship.
Here one should interject that the term ballad in the context of pop music has a meaning different from that used by folklorists. A ballad was any slow narrative or descriptive song. In more recent years, with the virtual disappearance of narrative songs from pop music after 1920, the term has come to be applied to any slow song that is not a blues.
How much of the music of Tin Pan Alley (in the broad sense defined earlier) has entered folk tradition? It is a question fraught with methodological problems. How, after all, does one assay the extent of TPA songs in tradition in the face of the assertion that many collectors systematically exclude such material from their collections and publications? I have used D. K. Wilgus's Western Kentucky Folklore Archive as the standard representative collection. This archive, currently housed at UCLA, contains some fifty-two hundred ballad and song variants (excluding instrumental pieces) and includes the collections of Perrow, Combs, and Halpert, as well as Wilgus's own. Thus, although the bulk of the material was collected after about 1950, some was obtained as early as 1913. Because Wilgus has been favorably disposed toward the unorthodox folksong material, and also perhaps more importantly-because much of his and Halpert's collectanea were gathered by students, whose lack of discrimination here becomes a positive virtue, I am confident that the archive is an honest representation of what the folk actually sing (at least, in one geographic region).
An item-by-item examination of the fifty-two hundred songs in the Western Kentucky Archive reveals that a minimum of 11 per cent are demonstrably of TPA origin and an additional 6 percent probably are but have not yet been traced to their sources. Of the different titles, 5.5 percent are demonstrably of TPA origin and 6.5 percent more probably are. The former variable measures TPA's contribution to the repertoire of the typical folksinger, whereas the latter number is a measure of TPA's contribution to the total number of traditional songs. The former seems to me the more useful parameter, since it gives more weight to those titles which are more widely known. It is interesting to compare the analogous figures for Child ballads in the archive. A total of 223 variants of thirty-one different titles have been deposited. This is equivalent to about 1.5 percent of the titles or 4.5 percent of the total reported texts.
My purpose in making this comparison is not to fire any sort of broadside at the Child canon; however, the statistics of the occurrences of those venerable pieces are particularly useful since one can be confident that every Child ballad encountered by a collector will be archived and recorded. Thus if the Western Kentucky Archive is indeed typical, one should expect that any representative collection will contain about three times as many TPA songs as Child ballad texts. Needless to say, this is far from the case. An analysis of two dozen primary collections of various types indicated that in only three instances do the figures approach these predictions." [11]
In all the others, the proportion of TPA material is less than 10 percent and in many instances less than 2 percent. There seems to be no systematic variation in these data: there is no chronological trend, nor do any geographic factors seem important, with the signal exception that every New England or Canadian collection is virtually barren of TPA material. The only explanation for the deficiencies of the various sources is the bias of the individual collectors. One particularly revealing case is the Southern Michigan collection of Gardner and Chickering.[12] Of the 327 texts printed, none is of TPA origin. However, an appendix to the book, listing another 180 texts collected but not printed, contains 39 TPA songs, representing an overall fraction of 8 percent of the total collection.
As for the apparent absence of TPA material from northern collections, I have no corpus similar to the Western Kentucky archive at hand to use as a reference standard. However, one recent datum is instructive. In 1968 Folk-Legacy Records issued an LP by Sara Cleveland of upstate New York Songs. [13] Kenneth Goldstein, who wrote the brochure notes, includes a list of all the songs in Mrs. Cleveland's repertoire. Of the 189 titles listed, 6 percent are apparently Child ballads (granting some uncertainty in identification by title alone) and 14-17 percent are TPA songs. If there is something in the upper latitudes that is uncongenial to the songs of TPA, it must not make its appearance until north of the state of New York.
Admittedly the entire burden of the argument could be dismissed by the assertion that songs composed professionally and circulated by mass media cannot ipso facto be folksongs.[14] A folksong here is taken to be any song whose survival is not entirely dependent on commercial media. It is apparent that the usual criteria of a folksongantiquity, oral transmission, anonymity of author, variation, etc.- are all simply pieces of evidence pointing to an existence outside commercial media.[15]
In spite of outspoken denials by early collectors of the appropriateness of including the modern popular songs in their collections, a great number managed to survive the weeding-out process.[16] Louise Rand Bascom, in her 1909 collection, had many interesting if not accurate observations to offer about the mountaineers' musical tastes: "Certainly On the Banks of the Wabash, Just One Girl, and other socalled popular airs never reach the mountains," she stated,[17] although five texts of the former ballad are on deposit in the Western Kentucky Archive. Bascom included in her collection two texts of "Kitty Kline," perhaps the earliest occurrence in an American scholarly collection of a pop song which was essentially postminstrel in character if not in actual date of appearance.[18] Shearin and Combs, in their 1911 syllabus of Kentucky folksongs, commented:
Amid the flotsam and jetsam of popular parlor-songs every where current the following have come to hand. They are hardly worth preserving, even by title, save for the fact that in spite of their pseudo-literary tang they are fellow travelers by oral tradition with the true folk-songs and songballads. [19]
They then enumerated some forty titles, many of which have since been collected by other workers, including "After the Ball," "I'll Be All Smiles Tonight," and even some which, like "Fond Affection," may not have been mass distributed. However, in spite of this halfhearted censorship, over a dozen songs, mostly from TPA in the 1880s and 1890s, appear elsewhere in their syllabus evidently with the complete approval of the compilers.
One song, "The Preacher and the Bear," was published a mere seven years before the syllabus was printed. Apparently Shearin and Combs were not familiar with the sheet music of the day.
The case of Louise Pound's 1915 Nebraska syllabus is more puzzling. In her introduction, she notes that she is excluding from the syllabus "popular songs fairly recent in origin ... and especially sentimental songs of the parlor or drawing room type." [20] Yet elsewhere she says that "in many cases it was possible to determine the composer... of the piece, or the year of its original appearance...." [21]
Thus she included and recognized the authors of "After the Ball" (1892) and "Grandfather's Clock" (1876) but could not give authors for similar pieces in the syllabus such as "In the Baggage Coach Ahead" (1896) or "Two Drummers" ("Mother was a Lady" [1896]). If these are not sentimental parlor ballads then we are indeed talking about an empty set.
In his brochure notes to the Sara Cleveland LP, Goldstein observes:
So little repertoire study has been carried out with American folk singers that we are hard pressed to comment on the large number of sentimental songs and homiletic pieces known by Sara. Certainly many traditional singers, North and South, knew many such songs... but few collectors have considered these worthy of notice and have chosen to publish only those pieces which they considered more traditional.[22] In spite of this significant statement and in spite of the fact that, as mentioned earlier, some 14-17 percent of Mrs. Cleveland's repertoire was born in Tin Pan Alley, none of the selections included on the phonograph record is of that category. Goldstein and Folk Legacy Records, like other collectors before them, could not resist displaying Child No. 52, previously unreported in America, but passed up an opportunity to issue the first traditional recordings of any of a dozen different TPA songs.
There is no need to dwell further on the treatment of TPA songs in the scholarly collections beyond the observation that no collection published so far has listed a single copyright or precise date. It would seem that collectors were unaware of the copyrighting phase of our government's activities. Leslie Shepard does not hesitate to term Rollins' Analytical Index to the Ballad Entries in the Registers of the Company of Stationers of London the "indispensable research tool for the study of broadside ballads of the sixteenth and seventeenth centuries." [23] I have little doubt that the U.S. Copyright records, if properly used, could serve a parallel function for native American folksongs of the post-Civil War decades.
Several interesting questions remain which I can only touch upon. For example, why did these songs of Tin Pan Alley make such a dent in the repertoire of the American folksinger? One can venture many hypotheses based on sociological factors. Perhaps the urbanization and industrialization of rural America left many persons with a sense of longing for a lost way of life so that the sentimental ballads of mother and home found a receptive audience in the hills after they had been driven out of the towns.[24]
However, another factor less often considered is the predominant style of TPA songs of the 1880s and 1890s. The language used was simple and effective poetry, never rising to levels of great artistry but never sinking to the convoluted awkwardness of the broadside hacks of a century earlier. On the other hand, American popular songs of the pre-Civil War period were, as indicated earlier, often flowery and stilted, and marked the unsuccessful attempts of their creators at romantic poetry. There were numerous sentimental songs of mother and home in the 1850s and 1860s, but they tended to be more descriptive and lyrical than the narrative ballads of the later years. These antebellum songs were, in effect, not memorable, and for this reason the songs of the last decades of the century achieved a place in oral tradition that their predecessors could not attain.[25]
In this connection it is interesting that the men of the musical world who composed the sentimental ballads regarded the words as of secondary importance, whereas "the music is the thing, first, last and all the time." [26] Furthermore, it was usually the case that the single name associated with a pop song was that of the composer rather than the lyricist.
It is thus somewhat ironic that when the songs of Tin Pan Alley entered oral tradition, the words generally survived intact, whereas melodies and harmonies often suffered extensive truncation and simplification. In some hillbilly or field-recorded versions, all traces of the original gilded age melody have vanished, and the ballad is rendered in the rubato parlando style usually reserved for our oldest folksongs.[27]
Throughout this paper the two noncomparable terms Tin Pan Alley song and sentimental parlor song have been used almost interchangeably. Of course, not all the TPA songs of our period were sentimental ballads; however it is the opinion of some of the pop music historians that "sentimental balladry dominated the stages of the minstrel show, variety, burlesque, and musical comedy." [28] Furthermore, these were the TPA songs that survived best in oral tradition, although it would be foolish to deny the popularity of some other types. In contrast, the songs of the mauve decade that flourish in such modern urban settings as pizza parlors and summer camps are the novelty songs and those compositions best suited for harmony singing. It is tempting to speculate on the future effect of urban pop music on folk music. The preceding comments suggest that folk music will borrow from pop music only when the styles of the two come close enough together to permit interaction. The converse is not true: urban music is more subject to abrupt changes of temperament and style, so that at any time it might borrow esoteric elements. This is not to deny that there is a constant pressure exerted by urban music on folk music which leads to steady but gradual changes in styles. However, the wholesale adoption of several hundred urban songs in a period of three of four decades bespeaks more than just a trickle of influence.
Finally, it should be noted that these remarks provide the basis for another view of hillbilly music. The economic and sociological factors that set the stage for recorded hillbilly music in the 1920s have been examined in detail by Green, Wilgus, and others.[29] I should now like to suggest another factor that contributed to the success of that new branch of the phonograph industry in the years following 1923. The ballad had been driven out of pop song by the dance. While persons of urban tastes followed this change of regime readily, the more conservative must have longed to hear the old favorites of Gussie Davis, Paul Dresser, Charles Harris, and others. Hillbilly music answered this need admirably. Although the basis of hillbilly music in the older Anglo-American traditional folk music cannot be denied, the fact is that on hillbilly discs sentimental TPA songs far outnumbered the items tabulated by Child and Laws. Furthermore, Tin Pan Alley bequeathed to Tin Pan Valley the right to create new pieces in the same style. "There's a Little Box of Pine on the 7:29," though written in 1931, could easily be mistaken for a composition of the 1890s. The 1963 million-selling hit, "Detroit City," was an updated replay of "Take Me Back (to the Sweet Sunny South)." And Jimmie Davis's "How Far is Heaven?" recorded in 1950 by Kitty Wells, is thematically reminiscent of Charles K. Harris's "Hello Central, Give Me Heaven" (1901).
The work in this area is far from complete. Dozens of songs that bear the dusty but still unmistakably lavender-and-rosewater odor of Tin Pan Alley have appeared in various folksong collections or archives and have not yet been traced to their sources. Material from the earlier years following the Civil War is particularly elusive. When all these are properly accounted for, the percentage figures on TPA's contribution to folk music will doubtless be revised upward. Furthermore, we need more detailed studies of the effect of oral transmission on both tune and text over the past sixty to one hundred years. Such work will surely provide new insights into native American folk music and the folk processes. And it may even tell us something interesting about American pop music.
1. Journal of the Folk Song Society 1 (London, 1899): 1.
2. Ibid., p. 7.
3. Dictionary of Folklore, Mythology, and Legend, ed. M. Leach and J. Fried (New York, 1949), pp. 1032 ff.
4 For background on the early stages of American pop music, see Sigmund Spaeth, A History of Popular Music in America (New York, 1948); Isaac Goldberg, Tin Pan Alley (New York, 1930); and David Ewen, Songs of America (Chicago, 1947).
5 WF 27 (1968): 146.
6 Early Songs of Uncle Sam (Boston, 1933), p. 126.
7 Goldberg says the name came to Rosenfeld while he was visiting Harry Von Tilzer; Goldberg names Von Tilzer as his source of information (p. 173). Ewen recounts essentially the same story (Panorama of American Music [Englewood Cliffs, N.J., 1957], p. 165). Maxwell F. Marcuse gives a similar anecdote but claims Von Tilzer actually thought of the phrase (Tin Pan Alley in Gaslight [Watkins Glen, N.Y., 1959], p. 10). Hazel Meyer relates both versions and also a third story which gives the credit to 0. Henry; the latter supposedly gave the phrase to lyricist Stanley Murphy, who occasionally collaborated with Von Tilser (The Gold in Tin Pan Alley [Philadelphia; 1958] p. 40)
8. It should perhaps be pointed out that the minstrel shows hardly confined themselves to pseudo-Negro songs, but featured a great deal of the sentimental parlor songs as well.
9 For examples see Mark Sullivan, Our Times, vol. 4 (New York/London, 1932), chap. 10.
10 They All Sang (New York, 1934), p. 156.
11. Earl Stout's Folklore From Iowa (New York, 1936) has the highest proportion of TPA songs of any collection examined: 26-30%, compared with 8% Child ballad texts. For Louise Pound's Folk-Song of Nebraska and the Central West: A Syllabus (Lincoln, Nebr., 1914) the figures are 11% and 4%, respectively. For Mellinger Henry's Songs Sung in the Southern Appalachians (London, 1934), the figures are 12% and 5%.
12. Ballads and Songs of Southern Michigan (Hatboro, Pa., 1967; reprint of 1939 ed.).
13 Sara Cleveland of Brant Lake, N.Y. (Folk Legacy Record FSA-33).
14 This point of view is taken by G. M. Laws in Native American Balladry (Philadelphia, 1964), and has been frequently criticized, e.g., in D. K. Wilgus, Anglo-American Folksong Scholarship Since 1898 (New Brunswick, N.J., 1959), pp. 251-253.
15 This is not the only definition which will admit to the class of "folksong" the TPA songs considered here. Ed Kahn, in a paper "Folksong: An Operational Definition" (14th Annual Meeting of the California Folklore Society, Davis, April 1966), proposed a recursive definition which states in essence that folksongs are those songs sung by folksingers. This definition would also be satisfactory for the purpose of this paper.
16 Wilgus reviews the fate of the sentimental parlor ballads in many early academic collections.
17 JF 22 (1909): 238.
18. This song was originally titled "Kitty Clyde" and published in 1854.
19. Hubert G. Shearin and Josiah H. Combs, A Syllabus of Kentucky Folk-Songs (Lexington, Ky., 1911), p. 28.
20. P. 6.
21. Pp. 4-5.
22. P.5.
23. Foreword to An Analytical Index to the Ballad Entries (1557-1709) in the Registers of the Company of Stationers of London, comp. Hyder E. Rollins (Hatboro, Pa., 1967), p. xv.
24. "See, e.g., Bill C. Malone, Country Music, U.S.A. (Austin, Texas, 1968), p. 25.
25 Wilgus suggests that hillbilly music borrowed those songs similar in theme to older traditional ballads ("Country-WesternM usic and the Urban Hillbilly," paper delivered at the Symposium on the Urban Experience and Folk Tradition, Detroit, May 1968).
26. Charles K. Harris, After the Ball (New York, 1926), p. 19.
27. E.g., cf. Fiddlin' John Carson's recording of "In the Baggage Coach Ahead" (Okeh 7006, 1924), or Charles Rose's version recorded in the field by Sam Eskin (AAFS LWO 2995, 11712 A2; 1947), with the original melody.
28. Ewen, Songs of America, p. 181.
29. Archie Green, "Hillbilly Music: Source and Symbol," JAF 78 (1965): 204; Wilgus, "the Hillbilly Movement," in Our Living Traditions: An Introduction to American Folklore, ed. Tristram P. Coffin (New York, 1968), p. 263.