Sonneck Society for American Music
Bulletin, Volume XXIII, no. 3 (Fall 1997)
Report on a Proposed Biliography of American Pocket Songsters of the Nineteenth Century
Norm Cohen, Portland, Oregon
Introduction
Anglo-American vernacular music has passed through many changes in the principal media of dissemination during the
five or more centuries since the first popular song was sold: handwritten manuscripts; broadsides and
broadsheets; chapbooks; garlands, budgets and songsters; sheet muisc; song folios; cylinder recordings and
phonographic records; magnetic tape cassettes and cartridges; video tapes; compact discs; digital tapes. Each
of these media has some unique aspects; each has played some special role in the social history
of Europe and America and the contribution of music to that history.
This essay is intended to serve as (1) a brief introduction to a fascinating area of 19th century American
printed songlore that stands at an important crossroads between folk/vernacular and popular music traditions;
(2) a progress report on a project begun last year, with the financial and moral support of a modest grant from
the Music Library Association (MLA); and (3) a request for advice and information, especially from
those readers affiliated with institutional libraries that might have holdings in the area under
discussion.
Songsters
The late D.K. Wilgus defined songsters as "nineteenth-century song collection[s], in America usually
pocket-sized (16mo or 24mo), containing the texts of music-hall, patriotic, religious and sometimes
traditional songs."1 One might amend this by removing the restriction to nineteenth-century;
that was certainly the era in which songsters flourished, but they were not confined to it. Irving Lowens
defined a songster, for the purposes of a bibliography, as "a collection of three or more secular poems intended to
be sung."2. The earliest American publication that fitted his definition was a 1734 reprint in Philadelphia
(by Benjamin Franklin) of a London publication of eleven years earlier. Songsters were cheaply printed and
distributed in large quantitites, often by manufactuerers of medicines, tonics, or elixers, or by distributors
of other consumable goods, or bhy popular stage entertainers. As such, they often contained advertisements,
testimonials, and short homilies in support of their company's wares. Sometimes they were produced by music
publishers who used them more or less as samplers of their other wares, primarily available in sheet music
format. Only rarely did publishers go to the expense of including musical notations; generally, instructions of
"[sung] to the tune of ..." had to suffice. Songsters printed in the 19th century generally had paper covers and
ranged from a few dozen pages to several hundred.
Musicologists, folklorists, and social historians have not been ignorant of the vast treasures contained in these
ephemeral pamphlets, many so cheaply printed that they fall apart today in the hands of the curious reader. Ewing drew
extensively on songsters in his study of the temperence movement.3 Jackson based a survey of American popular
music of 1825-1850 largely on songsters of the period.4 Cray used songsters to examine the role of cheap
print in the dissemination of tradition folk songs in his study of the ballad, "Barbara Allen."5 Richard
Lingenfelter relied heavily on songsters for sources for his collection of gold rush songs6 While the
contents of songsters have hardly been ignored, songsters as a class have been accorded little attention. Among the
few exceptions are Mile's bibliography of American presidential campaign songsters7 and Johannsen's
lengthy three-volume study of one particular publishing house.8 Two years ago, Honea published a preliminary
checklist of Masonic songbooks,9 a category that partially overlaps songsters per se;
and there has been some interest shown in the specific area of American circus songsters.10
The importance of songsters to reserach in various fields is thus easily defended. Other valuable aspects to this
body of popular literature are of interest. In the first place, when one considers the various sources for popular music
of the 19th century--broadsides and broadsheets, chapbooks, sheet music, and songsters--sheet music more than any of the
others represents the formal product of an established commercial music manufacturing industry. Sheet music songs therefore
start life furthest from the popular level of any form of popular music--though it may well happen that they do not
end life that way. Broadsides, broadsheets, and chapbooks were preemininent in the 17th and 18th centuries, but waned in
the middle of the 19th century, particularly after the Civil War. Therefore, for the last half of the 19th century,
songsters are the preeminent source for genuinely popular or vernacular musical material. Songs from the minstrel
stage in particular were exeedingly well-represented in songster collections, and the minstrel stage constituted our
first major black-white musical interchange.
Secondly, during this period, many songs were not individually copyrighted, but only as part of songster or
booklet collections. These songs are virtually impossible to find by the usualsearch methods through the files of the
Copyright Office, since they do not appear under their own titles. Their exhumation from archival collections
would be facilitated greatly by the eventual compilation of a complete songster index.
Thirdly, songsters contained many topical songs that can serve to illuminate social moods and issues of the day. The many
songsters delaing extensively if not exclusively with the Civil War constitute an obvious example; a more recondite one appears in The
National Clay Minstrel, all 84 songs of which related to Henry Clay's 1844 presidential campaign.
The song that arouses my curiosity is "Advance Whigs." It bears an illustration composed of the decorated pair
of letters, "OK"--an early printed example of the phrase which most (though not all) etymologists now agree originated in 1840
and referred to Martin Van Buren (nicknamed Old Kinderhook, from his place of origin in New York). Van Buren, the Democratic
president in 1836-40, did not win his party's nomination for president in 1844; why this reference to him in Whig campaign
literature of that year? Still a powerful figure in the Democratic party (he ran again for president in 1848), Van Buren's
nickname may have become an icon for the Party in general by 1844.
Fourthly, because of their vigorous role in popular literature, songsters, and the songs in them can provide etymologists with
valuable data for the early appearances of Americanisms, nonce words, slang, and neologisms in print. For example, Christy's
Bones and Banjo Melodies, an undated songster probably from the 1860s, contains the song, "Hunkey Dorey," the title of which
is a slang expression which, according to A Dictionary of Americanisms, was popularized through the Christy Minstrels' own use of
it in "Josephus Orange Blossom." This song, possibly of equal antiquity, may be an alternative source for the phrase's origins.
Fifthly, because songsters often represent a "second time around" for many of the songs they include, they may offer an additional
service as popularity indicators. This, though, is an application that cannot be exploited without more thorough understanding
of hte nature of the songster publication business. In fact, there are many questions that need to be answered about the songster
business. How and where did they advertise? To what classes of readers? Are there any business records extant that would tell numbers of
songsters sold? Which publishers used exclusively (or mostly) new songs, and which reprinted older publications?
What roles did performers play in the compilation and production fo songsters issued over their names? Because we probably will not
find anyone involved in the songster business of the 1956 century to interview, we will be forced to search through other
archives, newspapers, documents, diaries, etc., for any tidbits that are to be found.
Finally, while songsters are clearly of use to popular music historians, their value to folksong historians musc not be
overlooked. Many songs that in the 20th century found their way into recorded repertoires of hillbilly entertainers or
into the published field collections of traditional folksongs and ballads can be found in songsters, and may in
fact have first seen the light of day there. Keith Goldstein, who amassed two enormous collections of songsters and broadsides, both
now at the Center for Popular Music at Middle Tennessee State University in Murfreesboro, saw a knowledge of the two genres
as essential to the American folk musicologist: becausse of their nominal cost at publication and their lack of musical notation
(usually), they served mostly a working-class audience with minimal formal musical education -- in short, the social stratum
that has preserved most of our American folk music.
The prominent song collection in the U.S. is probably that of the Center for Popular Music at MTSU (1500 items). Other large collections
are at the Library of Congress (1200+), the Brown University Library (1000+), the American Antiquarian Society at Worcester (1000+),
the Morris N. Young Collection (ca 800), not at the University of Illinois, Lilly Library, Indiana University (ca. 600), Buffalo and Erie County
Public Library (450), and the New York Public Library (375), Harvard University Library (250), and the University of California at
Los Angeles (150+). These round figures are not necessarily comparable; inasmuch as the definitions of songster are rather vague,
the criteria for inclusion in a census differ at different institutions.
Progress Report
The MLA award, to which I alluded to earlier, is officially titled the "Dena Epstein Award for Archival and Library Research in American
Music," and was made possible by a gift from long-time MLA member and American music scholar Dena Epstein and her husband, Morton.
Although one ultimately wants musch more information, the modest goal of this project is the compilation of a bibliography of American
pocket-sized secular songsters published between 1860 and 1899, together with a list of libraries holding each individual songster.
I had no idea of the number of songsters with which I was dealing, but, based on the holdings of a few large
collections, estimated that there upwards of 2,000, and hopefully no more than 10,000 songsters fitting my description.
The restriction to "secular" songsters was made partly in the interests of manageability, but also because I felt one could make
a meaningful division between secular publications and religious ones, which generally have a completely different
repertoire. Secular songsters include several distinct types: political, minstrel stage, curcus, temperance, etc.,
and any others that defy neat categorization.
I have already alluded to the vagueness of the definition of "songster"; what about "pocket" songster? I had not given any
thought to a precise definition until I was queried by a librarian. I have provisionally limited the size to a height of 20 cm.
(8 inches), which allows for fairly generous-sized pockets, Is there any pratical consequence to this dimension? I think there are differences --
in repertoire, in audience, and in mode of distribution, at least -- between pocket-sized, soft-cover, text-only booklets and
full sized, hardcover, text/tune volumes, but distinctions blur at the boundaries. One needs to establish whether the same
collections of songs are available in pocket-sized collections as in full-sized harcover publications.
I sent a survey letter to about 130 institutions, which according to information in one of three references (listed in footnote 3, 8, or
12), had substantial holdings in the area of songsters.11 In addition to a question of approximately how many American
songsters published in 1860-99 the library held, the survey asked other questions about accessibility, cataloguing, utilization, etc.
Some 70 institutions replied identifying 16 with holdings of about 100 or more songsters.12. I found that for many librarians
my query was not easily answered, partly because "songster" is not a subject category that is commonly used, requiring
threfore searches in several possible different categories in a card catalog or electronic equivalent; furthermore, the date limitation
posed problems, since most collections are not catalogued by dates. Consequently, the patient librarians who responded to my questionnaire
often had to check card by card to tally the number of relevant entries. Many librarians simply sent me printouts form their
electronic catalogs where they had used the keyword "songster" and the appropriate date limitations.
The dimensions limitation presented another problem, since most sorting routines apparently did not allow for selections on this
parameter. The restriction to the years 1860-1899 was also problematic, in part because of the large number of songsters that were published
without dates. If one does a computerized search for songsters between 1860 and 1899, one loses all those without definite dates but clearly
published within that 40-year span.
Another problem was that some large institutions have songsters in completely separate departments--distributed among
literature, music, history, or Americana collections, and librarians in one department may have little or no knowledge
of holdings in another department. UCLA, for example, has a substantial songster collection in Special Collections, its rare
books division, but also has many in the Music Library and some in the general research library collection. The Library of
Congress has collections in music, folklife, and rare books.
I has originally assumed that the next step would be a visit to as many of the major institutions as I could manage, in order to
examine first-hand the songsters themselves. But it soon became apparent that one could do a great deal of preliminary bibliographic reseach
via modem using library catalogs that are available onlin in varying degrees of accessibility. More useful yet is WORLDCAT, the onlin
database maintained by OCLC of some 35 million items held in any one of several thousand libraries around the world. By judicious
use of WORLDCAT one can compile a good first cut at a bibliography.
A search of WORLDCAT using the key title word of "songster" and restricting language to English and date of
publication to 1860-99 netted about 1800 entries in various libraries that I cold easily download to my own computer via modem.
Of course, this fails to catch all songsters that don't actually have the word "songster" in their titles, so the next step was to
augment my database by searches for particular authors, keywords ("comic," "circus," "political campaigns" or other
searchable parameters).
A typical WORLDCAT entry, similar to many library card catalog entries, is shown here:
OCLC ACCESSION: 32125301
TITLE: Eugene Tompkin's Black Crook songster containing words and music.
PLACE: Chicago
PUBLISHER: W. Rossiter
YEAR: [1894?]
FORMAT: [16] p.: music; 31 cm.
NOTES: With some music. Issued in printed illustrated wrapper. Portrait on cover: Carline and Greppo.
Contains songs from Eugene Tompkins' ballet: The Black Crook. A ballet which takes place at the foot of the
Hatz Mountains, in the Castle of Wolfenstein, and in the heart of the Brocken.
SUBJECT: Carline and Greppo--Portraits.
Once all these have been dowloaded, there are several problems. First, these entries hold much more information than I
wish to deal with for now, so I have stripped them all down, eliminating some of the less useful details. Second is the
problem of errors; these entries were made, after all, by bibliographers many of whom had no special knowledge of
songsters or American music, and some of the comments are misleading, unnecessary, or just plain wrong. For example, how
has the bibliographer decided on date? Has he/she distinguished between publication and copyright date? If neither is
given, what other information was used? Third is the problem of deciding when two entries, describing songsters in two
different libraries, are really the same songster. In the world of cheap publications, publishers paid little attention to such niceties
as insuring the same title on cover, title-page, and spine. Which is the "real" title? In the above example, is the title "Eugene
Tompkin' sBlack Crook Songster..." or "Black Crook Songster..."? One also finds songsters with one "publisher" on the
title page and another on the cover. Sometimes, one is merely a distributor. Different bibliographers come to different
conclusions in these matters. More puzzling are the many instances where WORLDCAT contains two entries from the same institution
of what appear to be descriptions of the same songster (Brown University Library has microfilmed all of their songsters,
and each songster has two separate entries in OCLC's catalog; one for the real item and the other for the microfilm copy).
Worse yet, many instances exit of songsters that appear to be identical based on catalog description, or even title, page, yet
have difference contents. Some of the problems will persist until the songsters can be compared visually.
After some streamlining the WORLDCAT entries and adding the information on holding libraries, I end up with a format that
looks like this:
TITLE: Dick Sands' Irish character songster
PLACE: New York
PUBLISHER: A.J. Fisher
YEAR: 1877
OTHER: 18mo. dome song books; 113
OCLC LIBS.: RBN
ACCESSION: 30732705
OCLC LIBS: RBN, VHB
ACCESSION: 20742896
OTHER LIBS: MTSU
This particular example is a case of two separate entries in the OCLC catalog that appear, from the descriptions, to be
the same songster. Two libraries are listed for one OCLC accession number, and one of those same libraries is listed under
the other entry. The three-letter codes are OCLC's own abbreviations for various libraries. "Other libraries" designates
those institutions not entered in OCLC's catalog; in this case the information refers to MTSU's Center for Popular Music,
and the accession number is thei own. Some of these details may be dispensed with as the project progresses.
This armschair assembly of a bibliography is a good first step, but it is only that. Unfortunately, several major collections are
not available on WORLDCAT, including the American Antiquarian Society, the recently acquired collection at MTSU in Murfressboro, and, as
far as I can tell, much of the Library of Congress collection. Of course, ultimately, many songsters will have to be examined
directly in order to determine when different catalog entries refer to the same item. Direct examination will also help to fix
approximate year of publication (based on contents) for those numerous undated songsters catalogued simply as "[18--]." It is my
intention to carry out these various steps, hopefully in the present century--but I am reluctant to attach expected dates of completion
at this time. I would welcome comments from readers who know of other large collections that I have missed, or have other thoughts
on the direction this project might take. Please address comments to: Norm Cohen, 6507 SE 31st Ave., Portland, OR 97202-8627.
NOTES
1. D.K. Wilgus, Anglo-American Folksong Scholarship Since 1898 (New Brunswick, NJ: Rutgers
University Press, 1959), 437.
2. Irving Lowens, A Bibliography of Songsters Printed in America Before 1821 (Worcester,
MA: American Antiquarian Society, 1976), ix.
3. George Ewing, The Well-Tempered Lyre (Dallas: SMU Press, 1977).
4. George Stuyvesant Jackson, Early Songs of Uncle Sam (Boston: Bruce Humphries,
1933).
5. Ed Cray, "Barbara Allen: Cheap Print and Reprint," in Folklore International: Essays in
Traditional Literature, Belief, and Custom in Honor of Wayland Debs HandA, D.K. Wilgus, ed. (Harboro:
Folklore Associates, 1967), 41-50.
6. Richard A. Dwyer and Richard Lingenfelter, The Songs of the Gold Rush (Berkeley &
Los Angeles: University of California Press, 1964).
7. William Miles, Songs, Odes, Glees and Ballads: A Bibliography of American Presidential
Campaign Songsters (Westport, CT: Greenwood Press, 1990).
8. Albert Johannsen, The House of Beadle and Adams and its Dime and Nickel Novels; the Story of
A Vanished Literature, 3 vols. (Norman: University of California Press, 1950-52).
9. Sion M. Honea, "Nineteenth-Century American Masonic Songbooks: A Preliminary Checklist," in
Music Reference Services Quarterly 3, no. 4 (1995), pp. 17-31.
10. R.L. Parkinson, "Circus Songsters," in Circus World Museum Inc. Newsletter No. 20 (July
1989); J.M. Bonin, "Music from the Splendidest Sight: The America Circus Songster," Notes,
(20 June 1989) 699; F.D. Pfening, Jr., "Circus Songsters," The Bandwagon (Nov.-Dec. 1963), 10-12;
A.M. Smith, "Circus Songsters," Hobbies (Jan. 1947), 26.
11. D.W. Krummel, J. Geil, D.J. Dyen, D.L. Root, Resources of American Music History: A Directory
of Source Materials (Chicago and Urbana: University of Illinois Press, 1981). See also fn. 2 and 7.
12. These include, in addition to those institutions mentioned in the text, University of Pittsburgh; Frances
Willard Memorial Library, Evanston, Ill.; Johns Hopkins University; Hungtington Library, San Marino, CA; New
York Historical Society; Western Reserve Historical Society, Cleveland; and University of Michigan, Ann Arbor.
Norm Cohen is a retired physical chemist now teaching part-time in Portland, Oregon. He is author
of Long Steel Rail: The Railroad in American Folksong (U of Ill. Press, 1981) and Traditional
Anglo-American Folk Music: An Annotated Discography of Published Sound Recordings (Garland, 1994).
He has edited or annotated more than two dozen albums of folk/country music and was editor of
the John Edwards Memorial Foundation Quarterly (1966-1988).
Updated 1/06/97