More than fifty years have passed since John Tasker Howard's biography Stephen Foster, America's Troubadour and his article "Stephen Foster and His Publishers" appeared.¹America's Troubadour remains—half a century after it was written—the authoritative biography of nineteenth-century America's most renowned songwriter, and his article is still among the fullest treatments of the dealings between an American composer and his publishers. Perhaps Howard's greatest contributions to Foster biography were his deft reconstruction of the composer's financial affairs and the attendant dismantling of several popular misconceptions about the composer, particularly the pernicious myth that Foster was victimized by unscrupulous and avaricious publishers. This return visit with Foster and his publishers focuses not on the financial, but on the artistic—principally on publishers' attitudes toward the musical content of Foster's compositions: How faithful were Foster's publishers to the texts that they received from him? Did they edit or arrange his works? How much artistic control did they allow the composer once his manuscript was in their hands? Was Foster able to correct proofs? Was he allowed to make post-publication changes or revisions to his songs? Renewing our acquaintance with Foster and his publishers furnishes new perspectives on the workings of the American music publishing trade, on Foster's working habits, on his image of himself as composer, and dispels several remaining myths about Stephen Foster.

The autograph of "Open Thy Lattice, Love," Foster's first published composition (Philadelphia: George Willig, 1844), offers the earliest glimpse of his dealings with publishers.² The manuscript is not a Stichvorlage (a fair copy used as a model by the engraver), but an early sketch, covered with deletions and revisions.³ Despite these, however, the young composer's piano accompaniment remains awkward and amateurish: the right-hand figuration changes nearly every other bar, the left hand is filled with thick, closely spaced chords, and the voice-leading is often rough. Willig's published piano part bears little resemblance to Foster's sketch; it has been considerably reworked into an idiomatic, if conventional, accompaniment. The melody, however, is virtually unchanged.

The contrast between manuscript and print suggests that the accompaniment was rearranged by someone other than the young Foster, a fact that tends to support an unfortunate and commonly held view of the composer: that his fluent genius for uniting memorable melody with resonant poetic images masked a limited technical mastery of musical composition.⁴ One persistent theory, for which there is little corroborative evidence, is that Foster regularly submitted his songs to Henry Kleber, a prominent Pittsburgh musician, for editing and revision. In fact, Louis P. Kleber, Henry Kleber's nephew, claimed to know (more than eighty-five years after the fact) the specific note that his uncle changed in "Old Folks at Home."⁵ Another tradition holds that Foster's publishers made revisions and improvements to Foster's music.⁶ Several challenges to the conventional wisdom have been mounted recently, based on an increasing understanding of Foster's place in the American popular scene of mid-century. Charles Hamm has suggested that Foster had a broad knowledge of the song literature of his day, and that the simplicity of his style was a matter of deliberate choice, not the result of insufficient technical skill, while Kathryn Reed-Maxfield has noted an "overriding concern and skill for the craft of songwriting" in Foster's minstrel songs.⁷ As we shall see, there are other reasons to question the received wisdom on Foster; while the youthful composer of "Open Thy Lattice, Love" may have received some editorial assistance, the songs of the mature Foster needed no such rewriting.

At times publishers offered Foster practical suggestions, as Firth, Pond & Co. did in a letter of 12 September 1849 marking the beginning of the firm's long and fruitful contractual relationship with Foster.

From your acquaintance with the proprietors or managers of the different bands of "minstrels," & from your known reputation, you can undoubtedly arrange with them to sing them & thus introduce them to the public house, it is safe to hand such persons printed copies only, of the pieces, for if manuscript copies are issued particularly by the author, the market will be flooded with spurious issues in a short time.

It is also advisable to compose only such pieces as are likely both in the sentiment & melody to take the public taste. Numerous instances can be cited of composers whose reputation has greatly depreciated, from the fact of their music becoming too popular, & as a consequence they write too much & too fast, & in a short time others supersede them.⁸

Foster, incidentally, heeded both pieces of advice from his new business associates. He stopped giving copies of songs to minstrel performers before publication (a practice which had indeed resulted in a number of unauthorized editions of his early songs), and in composing was carefully attentive to sentiment and melody—those very qualities which seem the source of his songs' enduring ability to "take the public taste."⁹

Foster, in turn, offered his publishers suggestions, even demands, concerning artistic matters. During the 1850s he vigilantly oversaw the publication of his works, down to the smallest detail in their presentation.¹⁰ He may, in fact, have learned about the need for such oversight from the printing of "Open Thy Lattice, Love." Willig's earliest copies of the song name the composer as L.C. Foster; later copies printed by W.C. Peters from the original plates correctly identify the then-obscure songwriter as "S.C. Foster."

Two communications from the early 1850s confirm the composer's involvement in the publication process early in his career. A communication in the "To Correspondents" column of the February 1850 issue of the Baltimore Olio, a musical periodical published by W.C. Peters, reveals that Foster asked for changes to songs even after they had been printed:

"S.C.F" Pittsburgh.—It is too late to make the alteration as you requested. It is very popular in its present state. The Soiree P—a will appear next week.

The alteration was probably intended for "Summer Longings," the only Foster song printed by Peters in the year and a half before the appearance of the February issue of the Olio.¹¹ ("The Soiree Polka," the work which the notice says "will appear next week," was entered for copyright on 12 February 1850.) Ironically, "Summer Longings," which Peters had deposited for copyright on 21 November 1849, was reprinted in the February 1850 issue of the Baltimore Olio —the same issue that contained his communication to Foster. The reprint, despite the publisher's note to Foster that he could not accommodate the requested change, shows several significant musical differences from the original version of the song. However, on stylistic grounds the changes seem improbable for Foster,¹² and the numerous changes in the Olio do not accord with the specific reference to a single alteration in Peters's "To Correspondents" column. Furthermore, although there is no direct evidence as to who undertook the alterations, such changes by Peters, made without the composer's consent, would not be surprising given his house practices and temperament.¹³

Foster was more successful with another request for changes, though non-musical ones, made at about the same time. On 23 February 1850 Foster wrote to the minstrel performer E.P. Christy:

Dear Sir:

Herewith I send you copies of two of my late songs "Gwine to run all night," and "Dolly Day." I regret that the title-page had been ordered, and probably cut before I was informed of your desire that your name should not be used in connection with other bands. I have accordingly ordered my publisher in Baltimore to have a new title page cut bearing the name of your band alone like that used by Messrs. Firth, Pond & Co. N.Y. . . .¹⁴

Foster's order was carried out; later copies of title pages on Foster songs published by F.D. Benteen, his publisher in Baltimore, refer exclusively to the Christy Minstrels, dropping references to the Campbell Minstrels and to the New Orleans Serenaders.¹⁵

Further information about Foster's demands for such changes can be gleaned from his autograph Stichvorlagen. Of the twenty known musical holographs of Stephen Foster, seven (see Table 1) are Stichvorlagen; the others are sketches, presentation copies for friends, or copies of songs which Foster gave to minstrel performers.¹⁶

 

Table 1: Holograph Stichvorlagen of Stephen C. Foster

TITLE	PUBLISHER	DATE OF COPYRIGHT ENTRY
1. Oh! Boys, Carry Me 'Long 2. Where Has Lula Gone? 3. Sadly To Mine Heart Appealing 4. Linda Has Departed 5. For Thee, Love, For Thee 6. The Wife 7. Where is Thy Spirit, Mary?	Firth, Pond & Co. Firth, Pond & Co. Firth, Pond & Co. Firth, Pond & Co. Firth, Pond & Co. Firth, Pond & Co. George Mercer, Jr.	25 July 1851 7 September 1858 28 December 1858 1 March 1859 10 June 1859 9 February 1860 9 December 1895

There are two surviving autographs for "Oh! Boys, Carry Me 'Long": the first, the Firth, Pond & Co. Stichvorlage, the second, a copy that Foster sent to minstrel performer E.P. Christy.¹⁷ Comparison of the Stichvorlage and print based on it suggests that by 1851, Foster needed no help from his publishers in preparing his music. Firth, Pond & Co.'s earliest edition follows the manuscript faithfully—even slavishly observing its often idiosyncratic punctuation and capitalization. (There was little standardization in such matters for dialect songs.) In offering a nearly identical copy of the song to E.P. Christy for performance,¹⁸ Foster noted that "this song is certain to become popular, as I have taken great pains with it." Writing to Christy again one week later, he demanded that the minstrel performer "preserve the harmony in the chorus just as I have written it, and practise the song well before you bring it out."¹⁹ Foster, despite his lack of extensive first-hand knowledge of the minstrel stage,²⁰ displays an astonishing degree of self-confidence and artistic surety, pressing his demands for the harmonies of his score to be preserved exactly. Soon Foster would be pressing similar demands on his publishers.

No Stichvorlagen are known to survive from the early to mid-1850s, the years of Foster's rise to preeminence among American popular songwriters, yet we have evidence of Foster's continuing involvement in the publication process in the form of two early printed copies of "My Old Kentucky Home." The first was deposited—atypically for Firth, Pond & Co.—without a title page, on 11 January 1853. About three weeks later, on 31 January 1853, they deposited another copy of the song, complete with title page.²¹ The reason for this exceptional double copyright deposit doubtless lies in the substantial differences between the two songsheets, even though both were printed from the same plates. The caption title on the earlier copy reads "My Old Kentucky Home," while on the 31 January copy, it is "My Old Kentucky Home, Good-Night!" In the later copy dotted rhythms lend a sharper profile to the rhythm of the final line, and a significant revision appears in the piano coda.

 

Figure 1a. "My Old Kentucky Home," mm. 30-36. Proof copy deposited 11 January 1853 (original partially illegible)

 

Figure 1b. "My Old Kentucky Home," mm. 30-36. Later copy deposited 31 January 1853

 

Both the rhythmic changes and the new piano coda are almost surely by Foster. The attempt to observe the length of the vowels in the later reading is certainly far more typical of the composer than the earlier version. And the coda follows a pattern that became something of a formula in Foster's songs published between 1852 and 1855: a suddenly active left hand, frequently an Alberti bass if the meter is duple, accompanies a right-hand melody which begins with a variation on the concluding phrase of the vocal part, or (as here) a variation on one or more of the preceding vocal phrases. The motive for the changes is not entirely clear. The earlier version was probably never intended for public sale; it lacked a title page, several notes in the coda are illegible, and the rhythm in measures 30 and following gives the impression of being incomplete. Moreover, Foster, whatever his limitations as a composer, never offered for public consumption anything like the brutally exposed parallel fifths in the first measure of the original piano coda. Perhaps a preliminary proof of the first version was rushed to the district court merely for the purpose of securing a quick copyright (Foster may have committed another indiscretion in handing out autograph manuscripts), or possibly the first version was based on a sketch which Foster himself intended to replace. Though Foster's precise role in the revisions to "My Old Kentucky Home" are cloudy, he surely had a part in the changes.

In the fall of 1858, when the manuscript of "Where Has Lula Gone?" reached Firth and Pond, Foster was emerging from something of a compositional slump, having published only a handful of compositions in the previous two years; yet he had nearly one hundred compositions in print, including most of his commercially successful and best-remembered songs. He had published almost exclusively with the firm since 1853, and was familiar and comfortable with its working methods; often he prepared manuscripts so as to expedite the publishers' work.²² For instance, if he submitted additional verses on a separate sheet, he sometimes cued in words from these above the music. About midway through "Sadly to Mine Heart Appealing" Foster cued in "Hush'd have been for many a year" just above the words to the first verse as an aid to the engraver—insurance that the text underlay was proceeding correctly. Similar markings are found in the Stichvorlage of "Linda Has Departed." Foster had learned through experience the need for such hints. The first two verses of "Mother, Thou'rt Faithful to Me" end in a rhymed couplet, yet in the earliest printed copies of the song the third verse concludes without the expected rhyme:

But, warm in affection of sacred hue,
Mother, mother, thou'rt faithful to me.

The error, which arose when the engraver mistakenly punched the final words of the second verse at the end of the third, was corrected, and in later copies the final line reads "Mother, mother, thou'rt faithful and true."

Foster also meticulously dictated the items in the caption title, down to the period that Firth, Pond & Co. usually added after the title. Usually the caption title in the early prints follows the layout of his manuscript exactly, even observing the wording Foster had chosen to credit himself, alternating among such different phrases as: "Poetry and Music by Stephen C. Foster," "Words and Music by Stephen C. Foster," or simply "Stephen C. Foster."

Foster even felt comfortable enough to forward an incomplete manuscript for "Sadly to Mine Heart Appealing" to Firth, Pond & Co. His manuscript lacks the piano introduction and coda; instead, the first empty system contains only clefs, key and time signatures, and the tempo indication, "Andante". Similarly, he allotted a blank half-system at the end of the song for the piano's coda. The mystery of the missing measures revolves around Foster's amusing idea of incorporating a Scottish melody into the introduction of this song, a setting of Eliza Sheridan Carey's poem "On Hearing an Old Scottish Melody." A paragraph from Foster's letter to his brother Morrison dated 22 October 1858 explains:

If you have the book containing scotch melodies I wish you would send it to me, I will return it to you. I have sent to F.P. & Co. the song "Sadly to mine heart appealing" (Lines suggested on hearing an old Scottish melody) and would like to select an old tune for the introductory symphony. If you have not the book probably you can tell me where to find one.²³

The published version of the song incorporates "Robin Adair," the old Scottish melody Foster selected and arranged to complete the missing introduction and coda.

The engraver added bar lines to divide the manuscript's empty opening system into eight measures, and similarly divided the blank half-system at the end of the song into four measures—doubtless according to instructions Foster had enclosed with the song. These empty measures were numbered and marked for casting off (determining where line and page breaks would occur). He probably proceeded to cut the plates for the main part of the song, and added the "Robin Adair" introduction and coda when a second manuscript containing them arrived. This later manuscript is lost, but if the engraver had had it when he began his casting off, he would hardly have troubled to rule and number the empty measures on the earlier manuscript. In any event, the song was entered for copyright without significant delay, on 28 December 1858, despite Foster's belated search for a suitable Scottish melody (see Table 3 below).

None of Foster's Stichvorlagen show evidence of editing or arranging by a house editor. Other than indications for plate numbers, a few instructions to the engraver, and engravers' markings for casting off, they bear no writing not in the composer's hand. Not only do the Foster Stichvorlagen lack signs of editorial revision, they show no indication that the music was even checked for mistakes in the early stages of the publication process, even though several of the manuscripts contain careless errors and oversights by Foster.

Despite this lack of marking of the manuscripts, there are, in fact, numerous differences between three Foster manuscripts and the Firth and Pond prints. (The other two, "Where Has Lula Gone?" and "Sadly To Mine Heart Appealing," follow Foster's autograph faithfully.) Some of the differences seem to be engravers' sins of omission, usually omissions of performance indications such as slurs, staccato dots, and dynamic markings. Lack of space on the plate may have necessitated some of these. Others may have been due to carelessness or oversight; engravers regularly engraved the music and words first and added performance indications later.²⁴ Others are minor corrections, evidently made only after proofs had been pulled. Yet there are also more significant differences—differences in notes, rhythms, and words of songs. Most of these changes, including the most significant, were not made by an editor immediately after receiving the manuscript, nor during the initial engraving, but later, after a set of proofs had been pulled from the plates.²⁵

"Linda Has Departed" shows several such proof changes. Foster's autograph has the following reading in measure 6 of the vocal part:

 

Figure 2a. "Linda Has Departed," m. 6, vocal part, reading of Foster's autograph

 

Variations in the thickness of the staff lines make clear that this version was engraved on the plates, but that some time after the initial engraving this bar was changed to:

 

Figure 2b. "Linda Has Departed," m. 6, vocal part, reading after proof changes

 

A similar alteration to the plates can be seen in measure 23 of the same song.

 

Figure 3a. "Linda Has Departed," m. 23, vocal part, reading of Foster's autograph

 

Figure 3b. "Linda Has Departed," m. 23, vocal part, reading after proof changes

 

The change in measure 6 was probably an attempt to clarify the original notation of the rhythm, and the ossia later in the song may have been added to avoid an obligatory high f "—making the chorus more singable (thus salable). These small changes may have been initiated by the publishers for commercial reasons. Another such change, perhaps introduced by the publishers, occurs at the beginning of "The Wife," where Foster's autograph tempo marking "Moderato tempo justa" has been changed to the more conventional "Moderato" in the print. Yet before we conclude that the publishers forced these alterations on an unwitting composer, we need to examine several more examples of proof changes.

In measures 14-17 of the temperance song "The Wife," two words differ between the publishers' manuscript and the earliest print: "And" and "To" (see Table 2, below). These changes were made at the proof stage, after the plates were engraved, since in the earliest known state of the first edition (Foster Hall Collection copy 634.1) these words, unlike other words on the page, are slightly misaligned, and at one spot a small portion of a serif from the letter "T" remains visible beneath the word "And."

The key to the reason for these seemingly insignificant changes lies in Foster's sketches for these lines in his manuscript book.²⁶ The choice between these same two words lay at the heart of an earlier revision to the words that Foster had made between his manuscript book sketch and the holograph which he later sent his publishers.

 

Table 2: Foster's Three Versions of "The Wife"

MANUSCRIPT BOOK SKETCH	PUBLISHER'S STICHVORLAGE	FHC 834.1 (AFTER PROOF CHANGES)
He's gone to sup The deadly cup And while the long night through He's gone to laugh To talk and quaff And while the drear night through	He's gone to sup The deadly cup, To while the long night through, He's gone to quaff, To talk and laugh To while the drear night through	He's gone to sup The deadly cup, And while the long night through, He's gone to quaff, And talk and laugh To while the drear night through:

The first reworking may stem, in part, from Foster's attempt to establish a stronger connection between the verbs "talk" and "laugh" by placing them in the same line. Yet his first revision gave rise to an unpleasant threefold repetition of the hard "t" in "to" at the beginnings of lines, and this may have occasioned his subsequent proof changes. Whatever Foster's motivation for the changes, they are all the more revealing precisely for their seeming triviality. In Foster's mind the lowly preposition and conjunction must have loomed large indeed.²⁷

More proof changes can be traced directly to Foster. In "For Thee, Love, For Thee" a major difference between Stichvorlage and print occurs in measure 25 (and in the analogous bars of the second and third verses); again the change was made to the plates after the original reading had been engraved.²⁸

 

Figure 4a. "For Thee, Love, For Thee," mm. 25-28, reading of Foster's autograph

 

Figure 4b. "For Thee, Love, For Thee," mm. 25-28, reading after proof changes

 

Again Foster's manuscript book provides the clue to the reason for his changes; the sketches there show a concern with lending variety to the lines beginning with the words "for thee." Variations tried and rejected in the manuscript book include:

For thee love, for thee love
For Thee only thee—
Of thee love, of thee love
With thee only thee
In thee only thee
of thee ever thee—

Foster's change to the words also led to the rhythmic change in the vocal line. Foster's requests for changes may sometimes have held up the publication of some songs. Although the manuscript of "The Wife" was marked "rec'd Aug. 1/59," the song was not entered for copyright until some six months later, on 9 February 1860.²⁹ Comparison of Firth, Pond & Co. plate numbers, the numbers which they assigned songs in their "Foster's Melodies" series, and dates of copyright entry indicates that several Foster songs encountered delays in publication.

 

Table 3: Plate Numbers, Foster's Melodies Series Numbers, and Dates of Copyright Entry for Foster Songs Published by Firth, Pond & Co., April 1858-May 1860

	TITLE	PLATE #	SERIES #	DATE OF COPYRIGHT ENTRY
* * *	Lula is Gone Linger in Blissful Repose Where Has Lula Gone? My Loved One and My Own Sadly To Mine Heart Appealing Linda Has Departed Parthenia to Ingomar For Thee, Love, For Thee Fairy Belle Thou Art the Queen of My Song None Shall Weep a Tear For Me The Wife Poor Drooping Maiden Cora Dean Under the Willow She's Sleeping The Glendy Burk	4395 441630 4443 4484 4492 4554 4452 4566 4604 4709 4677 4705 4754 4824 4825 4861	33 34 35 36 37 39 38 40 41 44 42 43 45 47 46 48	19 April 1858 13 July 1858 7 September 1858 24 November 1858 28 December 1858 1 March 1859 4 April 1859 10 June 1859 19 August 1859 21 December 1859 9 February 1860 9 February 1860 9 February 1860 no record of copyright 3 May 1860 29 May 1860

"The Wife" is one of three songs in Table 2 that bear plate and series numbers that seem conspicuously out of line with the date of copyright entry. In addition to Foster's requests for changes, there may have been other complications during the publication of "The Wife." There are two sets of markings for casting off; the first set shows that the song was originally to have been spread over four pages. The second set of marks corresponds to the way the song was actually cast off, squeezed economically onto three pages. Copies from the earliest edition of "None Shall Weep a Tear For Me," like those of "The Wife," show passages where the staff lines vary in straightness and thickness. Foster's requests for corrections may have been at least contributing factors in these delays.

There is another type of change that Stephen Foster was able to extract from Firth, Pond & Co. When he arranged his own songs for voice with guitar accompaniment he often made small changes to the melody from the original piano-vocal version.³¹ In two cases the changes in the guitar arrangement seem to have so pleased Foster that he asked Firth and Pond to insert them into his previously published versions for voice and piano. The earliest copies of "Little Ella" in the original arrangement for voice and piano have the following reading in measure 23:

 

Figure 5a. "Little Ella," mm. 23-24, vocal part, reading in the earliest editions of the piano-vocal arrangement

 

When Foster arranged "Little Ella" for voice and guitar almost one year later, this passage was changed slightly:

 

Figure 5b. "Little Ella," mm. 23-24, vocal part, Foster's arrangement for voice and guitar

Figure 5c. "Little Ella," mm. 23-24, vocal part, reading of Foster Hall copy 333.3, a later edition of the piano-vocal arrangement

 

Later copies of the original piano-vocal arrangement incorporate Foster's changes to the cadence figure. The changes were made by punching and burnishing the plates and were made in two stages. In the first stage (represented in Foster Hall Collection copy 333.2) the cadence figure was changed only in verses two and three. A still later copy (Foster Hall Collection copy 333.3) has that change, and rectifies the oversight in the earlier copy, extending the melodic revision to the first verse as well.³² Foster may have preferred the revised cadence figure since it avoids the mild harmonic clash of the earlier reading, and because it preserves the repeated-note motive that serves as a pick-up to nearly every phrase of the melody. A similar post-publication revision to the melody of "Willie We Have Missed You" arose as the result of the preparation of an arrangement for voice and guitar.³³

As might be expected, Foster's relations with publishers changed during his final years of near destitution in New York. After 1860 few of his newest creations were commercial successes.³⁴ No longer did he have contracts guaranteeing a royalty interest in his compositions; he probably sold most of the works of the 1860s outright for cash in hand or against debts. There was little financial incentive for Foster to remain directly involved in the publishing process after receiving his fee, and little reason for publishers to bow to his artistic whims.

The hymns published by Horace Waters, one of Foster's principal publishers from these last years, provide an extreme example. Many of Waters's hymnbooks were produced by "recycling hymns"—reprinting previously published hymns using the same plates employed in an earlier printing. Yet Waters was rarely content simply to reprint a hymn without alteration. He reprinted ten Foster hymns that had first appeared in Waters' Golden Harp in his Athenaeum Collection.³⁵ The Athenaeum Collection exists in two bibliographically distinct states; among the differences between the two are changes to some of Foster's hymns.³⁶ Six of these Athenaeum hymns were then reprinted in Waters's Heavenly Echos.³⁷ Of the ten original hymns from Waters' Golden Harp, only three survived these reprintings without changes. The text of "The Angels Are Singing unto Me," for example, was revised for its appearance in Waters's Heavenly Echos; Foster's original words, "And soft winds are roving on the lea," are changed in the later hymnbook to "And soft winds are blowing o'er the lea." Waters's changes were by no means restricted to hymns' texts, more often they involved the notes: there are a dozen such differences between the two versions of "The Beautiful Shore" that appeared in two states of the Athenaeum Collection.

Yet unlike earlier post-publication changes, there are a number of reasons to suspect that the publisher, not Foster, was responsible. As we have noted, there was little financial incentive for Foster to press for such changes. Furthermore, Waters's alterations to hymns were not confined to works by Stephen Foster; they also characterize reprintings of hymns by other composers whose works appeared in his hymnbooks.³⁸ Most of Waters's changes are unlike those we can trace directly to Foster. While Foster's changes involve text, text setting, or the melody of a song, Waters's are usually changes to the pitches of one of the three lower voices. For instance, Foster's "Tell Me of the Angels, Mother" was revised between the printing of the two states of the Athenaeum Collection; the single change to the electrotyped plates is an octave displacement of a note in the bass part.

Waters's tampering with Foster's compositions did not cease with Foster's death; some changes to Foster's hymns occur for the first time in Waters' Heavenly Echos, published in 1867—three years after Stephen Foster's death. Horace Waters's protégé and successor, Charles M. Tremaine, carried on this questionable tradition. Among the changes he introduced into his issue of Foster's "Merry Little Birds Are We," published using Waters's plates, are changes of pitches, additions of fermatas and trills, and the following advice to performers at the head of the chorus: "In singing this chorus a fine effect will be produced by having a few boys whistle in imitation of birds."

One final, though quite different, example of posthumous revision of a Foster composition will complete our visit with Foster and his publishers. One other holograph that reached a publisher has survived, though strictly speaking it was not a Stichvorlage, the autograph of "The Voices That Are Gone," entered for copyright more than a year after Foster's death, on 13 January 1864, by William A. Pond, one of the principals in Firth, Pond & Co. before its dissolution in 1863.³⁹ The title page of Pond's edition confirms that the music is by Stephen C. Foster, but with "The Symphonies and Accompaniments by John P. Cooke." Cooke altered the harmony (Foster's original unfolds almost entirely over a C pedal point) and added octave doublings to Foster's piano part. He did not, however, tamper with the melody. The reason Pond engaged Cooke to arrange the accompaniment is unclear; Foster's simple accompaniment is serviceable enough. It does not seem to have been the firm's usual practice; of the five Foster songs published by this firm after the dissolution of Firth, Pond & Co. none of the others has an indication that an arranger was involved. Ironically, the earliest surviving musical manuscript of Foster and his last surviving musical autograph suffered the same fate—revision of the accompaniment by someone other than the composer.

John Tasker Howard showed that Foster was not exploited financially by publishers. The same holds true, at least for much of Foster's career, regarding artistic matters. Stephen Foster, the first American musician to earn a living exclusively through his activity as composer, showed an ability to prevail upon his publishers artistically and to retain a degree of control over the publication of his works that was perhaps unprecedented for a writer of American popular song. The nature of Foster's dealings with publishers, both financial and artistic, reveals a self-assured craftsman, deeply concerned and involved with the form in which his works reached the marketplace. These dealings also suggest that a complex and little understood interaction between publisher, composer, and consumer influenced the form in which popular song reached the parlor pianos of nineteenth-century America; these interrelationships determined, in part, how such music was circulated, performed, and, ultimately, woven into the fabric of American musical life.

---

¹John Tasker Howard, Stephen Foster, America's Troubadour (New York: Crowell, 1934); page numbers in citations below refer to the revised edition (New York: Crowell, 1953). John Tasker Howard, "Stephen Foster and His Publishers," Musical Quarterly 20 (1934): 77-95.

²The present location of the manuscript is unknown; it was reproduced in Evelyn Foster Morneweck, Chronicles of Stephen Foster's Family (Pittsburgh: University of Pittsburgh Press, 1944), facing p. 267. Willig's earliest edition of the song is reproduced in the Foster Hall Reproductions: Songs, Compositions and Arrangements by Stephen Collins Foster, 1826-1864 (Indianapolis: Josiah Kirby Lilly, 1933).

³The exact date of the sketch is not known. Morrison Foster, perhaps following Robert Peebles Nevin, claimed that "Open Thy Lattice, Love" was composed when his brother was sixteen; see Morrison Foster, Biography, Songs & Musical Compositions of Stephen C. Foster (Pittsburgh: Percy F. Smith, 1896), 13; and Robert Peebles Nevin, "Stephen C. Foster and Negro Minstrelsy," Atlantic Monthly 20 (1867): 612. The song was actually composed somewhat later; the source of Foster's text is a poem by George Pope Morris published under the title "Serenade" in The New Mirror 2 (October 14, 1843): 32. Thus Foster, who was born July 4, 1826, was at least seventeen when the song was written, and the sketch cannot predate the publication of the song by much more than a year.

⁴See for example Howard, America's Troubadour, 116, 188-89, 194-95; Harvey Gaul, The Minstrel of the Alleghenies (Pittsburgh: Friends of Harvey Gaul, [1952]), 18-19; Harold Vincent Milligan, Stephen Collins Foster: A Biography (New York: G. Schirmer, 1920), 96-97, 110-14; Grove's Dictionary of Music and Musicians, 5th edition, ed. Eric Blom (New York: St. Martin's Press, 1955), 3: 456.

⁵Letter from Louis P. Kleber to Fletcher Hodges, Jr., February, [1939], now in the Foster Hall Collection, University of Pittsburgh.

⁶Morneweck, Chronicles, 451-52.

⁷Charles Hamm, Yesterdays: Popular Song in America (New York: Norton, 1979), 201-27, and Music in the New World (New York: Norton, 1983), 231-35; Kathryn Reed-Maxfield, "Emmett, Foster, and Their Anonymous Colleagues: The Creators of Early Minstrel Show Songs," paper read at the national meeting of the Sonneck Society, Pittsburgh, 4 April 1987.
Foster had at least a minimal knowledge of music theory; see for example his letter in the 26 February 1853 issue of the Musical World, quoted in Howard, America's Troubadour, 220-22, which discusses the resolution of the seventh and the correct doubling of chord tones. His academic musical training, though limited, was usually more than sufficient to the genres in which he worked. He demurred at demands for more complex scores, refusing, for example, his sister's request for an organ work (see Howard, America's Troubadour, 131, and Morneweck, Chronicles, 286).

⁸Letter now in the Library of Congress, ML 95 .F8 Case.

⁹Indeed these two facets of musical creation occupied him most as he revised and polished his songs; on Foster's working methods, see note 22 below.

¹⁰In his financial dealings with publishers, too, Foster seems to have retained considerable control. John Tasker Howard pointed out that Foster may have been able virtually to dictate the terms of his contracts with Firth, Pond & Co. See America's Troubadour, 244.

¹¹Peters entered "Away Down South," "Susanna," "Uncle Ned," and "Santa Anna's Retreat from Buena Vista" for copyright on 30 December 1848. Peters's advertisements in Louisville and Cincinnati newspapers for 1847-49 show, however, that these songs had been issued by the summer of 1848. See the critical reports for these songs in The Complete Works of Stephen C. Foster, ed. Steven Saunders (Washington: Smithsonian Institution Press, in press).

¹²See the critical report for this song in The Complete Works of Stephen C. Foster.

¹³Richard D. Wetzel, private communication to Steven Saunders, 25 August 1985. I am grateful to Dr. Wetzel for information about W.C. Peters's activities as music publisher.

¹⁴Letter from Stephen C. Foster to E.P. Christy now in the Library of Congress, ML 95 .F8 Case.

¹⁵For a fuller discussion of this incident and descriptions of the title pages before and after the changes, see Howard, America's Troubadour, 179-82.

¹⁶This total of twenty autographs does not include a handful of musical sketches in Foster's manuscript book—a book usually reserved for reworking words of songsnor musical autographs which are lost.

¹⁷Both manuscripts are in the Huntington Library, San Marino, California; they share a single call number, HM 16511.

¹⁸The only significant difference between the two manuscripts is in the words of verse 4. Foster's Stichvorlage begins "Farewell to de hills," while the manuscript he sent to Christy begins "Farewell to de fields."

¹⁹Both letters, the first dated 12 June 1851, the later dated 20 June 1851, are in the Huntington Library, San Marino, California.

²⁰For instance, Foster confessed in his letter to E.P. Christy of 23 February 1850 that he had never heard the minstrel performer's band. More than a year later, in the letter of 20 June 1851, he admitted not knowing whether Christy's minstrels used a piano, or had a tenor who could sing up to g with ease.

²¹Both copies of "My Old Kentucky Home" are among the early copyright deposits in the Library of Congress. The earlier copy is in M1 .A12 Vocal, v. 54; the later copy is in M1 .A12 Vocal, v. 57.

²²Foster's music script in these Stichvorlagen, while not approaching the calligraphic penmanship of his presentation copies, is always neat and legible. There are occasional revisions in these manuscripts; examples include changing the spelling of notes (from to , for instance), realigning left- and right-hand notes, and correcting copying errors. Only a few of these, however, are significant enough to be called compositional changes. His manuscript book, which he used to rework words of songs, shows that song texts were painstakingly revised and polished; sketches for the words of a single song sometimes were spread over as many as a dozen pages. The few surviving musical sketches suggest that Foster was similarly critical of his music, and revised it carefully before beginning his fair copy for the publisher.

²³Letter from Stephen C. Foster to Morrison Foster, 22 October 1858, original in the Foster Hall Collection.

²⁴A comparison of proofs of Foster's "Where is Thy Spirit Mary," published in 1895, and the proofs of James Hewitt's "The Music of Erin," published in 1807, shows how little publishing practice changed in this respect during the nineteenth century. The proofs of "Where is Thy Spirit Mary" are in the Foster Hall Collection, University of Pittsburgh; those of "The Music of Erin" are reproduced in Wolfe, Early American Music Engraving and Printing (Urbana: University of Illinois Press, 1980), illustration 12.

²⁵At places on the plate where a note or symbol does not touch another sign (for instance, dynamic signs or music that lies above or below the staff) a skilled engraver can make such changes undetectable. However, if a change involves symbols on the staff, then the punching and burnishing process will usually require that some staff lines be redrawn. Variations between tools or difficulty in aligning the redrawn line results in staff lines that differ in width, or are not precisely parallel—telltale signs that a plate has been burnished and a change made. For a description of the process see Wolfe, Early American Music Engraving and Printing, 78.

²⁶Foster's manuscript book is part of the Foster Hall Collection.

²⁷Foster made a similar last-minute change directly on the manuscript of "The Voices That Are Gone," this time involving the words "once" and "then". On this manuscript, see below.

²⁸Traces of the earlier reading can be seen clearly even in the Foster Hall Reproduction.

²⁹Backlogs at Firth, Pond & Company could cause such delays, too. Foster wrote to E.P. Christy in 1851 that "I have just received a letter from Messrs. Firth, Pond & Co. stating that they have copyrighted a new song of mine ("Oh! boys, carry me 'long") but will not be able to issue it for some little time yet, owing to other engagements. This will give me time to send you the m.s. and allow you the privilege of singing it for at least two weeks, and probably a month before it is issued. . . ." The letter suggests that a delay of even a month in printing a song was hardly typical. Indeed, Foster's publishers could be extraordinarily rapid in bringing out his compositions. Foster made a marginal notation in his sketchbook "sent Laura Lee July 19th." Just over three weeks later, on 7 August 1851, F.D. Benteen deposited a printed copy of "Laura Lee" for copyright.

³⁰The plate number on the earliest copies of "Linger in Blissful Repose" is 4116; this was apparently an error, and was changed on later copies to 4416.

³¹In contrast, arrangements of Foster songs known to have been prepared by other arrangers nearly always confine their melodic changes to transposing the song into a key suitable for the guitar.

³²Unfortunately copy 333.2, the first piano-vocal copy to contain the change, does not have any changes in the publishers' address or name or any other revealing changes that would indicate definitely whether the change to the plate was made before or after the preparation of the guitar arrangement. It seems likely, however, that the change came after the guitar arrangement, as a result of Foster's reworking of "Little Ella's" melody. Firth and Pond could scarcely have had a motive for introducing such a minor change to the plates unless it were called to their attention by the composer.

³³Another type of change to plates which Firth and Pond made is worth noting: post-publication corrections of errors. At least seven Foster songs published by Firth, Pond & Co. had mistakes in the first published copies which were corrected in subsequent impressions.

³⁴See the reminiscences of Foster's friend and collaborator George Cooper, recounted in Harold Vincent Milligan, Stephen Collins Foster: A Biography (New York: G. Schirmer, 1920), 104, and Howard, "Stephen Foster and His Publishers," 93.

³⁵Both New York: Horace Waters, 1863.

³⁶The title pages of the two states of the Athenaeum Collection are identical. The Library of Congress has both states of the hymnbook: the call number of the earlier state is M2117 .W31 A72, of the later, M2117 .W31 A7.

³⁷NewYork: C.M. Tremaine, 1867. Although Tremaine is listed as the publisher of his collection, Waters was its editor, and the copyright was entered in his name.

³⁸As examples of Waters's changes to other composers' works compare the following titles in his Athenaeum Collection and Golden Harp: "Come to Me," "God Speed the Right," and "Christmas Carol."

³⁹The manuscript (ML 96 .F8 Case) was presented to the Library of Congress in about 1914 by the William A. Pond Company.