The Making of The Shenandoah Harmony

The Shenandoah Harmony.

At 469 songs on 480 pages, The Shenandoah Harmony is the largest new four-shape tunebook published for more than 150 years. The music committee includes John del Re, Kelly Macklin, and their daughter Leyland del Re of Virginia; Nora Miller of Maryland; Daniel Hunter and me, Rachel Hall, of Pennsylvania; and Myles Louis Dakan and Robert Stoddard of Boston.1 The book is now in its second printing since its release in mid-February this year. We held our first all-day singing on June 2nd in the Harrisonburg, Virginia area, near Ananias Davisson’s grave. Over one hundred singers from at least sixteen states attended. Recordings, photographs, and videos of the singing are on our web page.

The original inspiration for The Shenandoah Harmony was to create a collection of songs compiled, printed, and published by Ananias Davisson from 1816 to 1826 in the Shenandoah Valley of Virginia. These works include five editions of the Kentucky Harmony and three editions of A Supplement to the Kentucky Harmony. Davisson’s innovative books, which combined European congregational hymns and New England class songs and anthems with the frontier sound of arranged folk hymns and camp meeting songs, had a profound influence on later tunebooks. Although Davisson did compose and arrange music himself—”Idumea” (p. 47b in The Sacred Harp), for example—his greatest talent lay as a tunebook compiler, and, in particular, in his selections of Mid-Atlantic folk hymns and composed songs. Many popular songs, including “Pisgah” (p. 58), “New Topia” (p. 215), and “Exit” (p. 181), came into the Southern shape-note repertoire through his publications. The Kentucky Harmony and its Supplement, in addition to two other early nineteenth century Mid-Atlantic sources, the two-part Wyeth’s Repository (Harrisburg, Pennsylvania) and the multiple editions of The Easy Instructor (Philadelphia and Albany, New York), were critical in determining which of the vast number of New England compositions were eventually sung in the South.

In addition to being avid Sacred Harp singers, the del Re family has been singing from Davisson’s books for over twenty-five years. Inspired by Judy Hauff and Ted Mercer’s recordings of songs from Wyeth’s Repository and Kentucky Harmony, they sought out songs associated with their own region. Through these recordings, they also became aware of the wealth of shape-note music found in out-of-print and inaccessible tunebooks. The Shenandoah music committee formed in late 2010 to early 2011 and continued to meet for a full weekend once a month for two years. Although the initial plan was for a more modest book, by summer 2011, we had decided to make a book suitable for all-day singing. The group has considered about 1500 songs together; thousands more were reviewed by individual committee members but not brought to the group.

Shenandoah Harmony source list.

In all, we used seventy-five tunebooks as sources (see our source list). In addition to Davisson’s books, in which we found over a quarter of our songs, the principal sources of The Shenandoah Harmony are Hauser’s Hesperian Harp (1848); McCurry’s Social Harp (1855); Walker’s Southern Harmony (1835-54), Southern and Western Pocket Harmonist (1846), and Christian Harmony (1866); the two parts of Wyeth’s Repository (1810, 1813); Ingalls’s Christian Harmony (1805); Mansfield’s American Vocalist (1848-9); and Stone and Wood’s Columbian Harmony (1793). Other books such as The Virginia Harmony (1831) were chosen because of their association with the Shenandoah Valley.  Moreover, several scholars—chiefly, Nikos Pappas—generously helped us locate obscure tunebooks, such as George Miller’s Methodist Camp-Meeting Song Book, published in Dayton, Ohio in 1841.  We also sought different versions of the same song in several sources, finding that sometimes an original song had interesting features that were lost or confused in later editions, but other times later versions were improvements.

Our experience singing from The Sacred Harp has profoundly influenced every aspect of The Shenandoah Harmony. In order to make our book suitable for all-day singing, we sought a variety of sentiments, levels of difficulty, and compositional techniques, while still maintaining the distinctive style of part writing and mix of eighteenth- and nineteenth-century texts typical of four-shape tunebooks. Despite our focus on Davisson, only 130 of the songs we selected are found in his publications in some form. We added sixty-one songs originating in other tunebooks of the Mid-Atlantic and its western frontier, mostly from the period 1800–1850; 109 songs from New England, 1770–1810; eighty-two Southern songs from 1835–1911; and nineteen British Isles or European songs, mostly pre-1800. The Shenandoah also includes sixty-eight songs written since 1950, with an emphasis on Mid-Atlantic and Midwestern composers, including thirty-six compositions never before published in a book. Seven songs derive from oral traditions, including those of Hoboken, Georgia, Sand Mountain, Alabama, and Glen Rock, Pennsylvania. We did not include any of our own compositions.

Not wanting to overwhelm singers with too many unfamiliar songs, we chose some pieces appearing in other modern tunebooks such as The Northern Harmony and An American Christmas Harp (the video shows singers in Ireland singing “Pennsylvania” [p. 254 in The Shenandoah Harmony], which we first learned from The Northern Harmony). However, The Shenandoah Harmony does not duplicate songs in The Sacred Harp, 1991 Edition, although there are a handful of familiar Sacred Harp songs that have different texts or arrangements. Some songs appeared in previous revisions of The Sacred Harp. In comparison with The Sacred Harp, the Shenandoah has fewer twentieth century compositions, a higher proportion of minor songs, a different geographical emphasis, more songs with fewer than four parts, more secular songs, and some alternate texts in other languages (German and Polish).

Editing, with the committee and friends. In time, the book grew too big to print.

Next to song selection, editing was the music committee’s most time-consuming and delicate job. The Sacred Harp, which has shaped our expectations as singers, is very much an edited collection, not a historical sourcebook. Its editors have not only added and subtracted songs over the years, but also, as tastes changed, revised older songs to suit the singing style of the day. Adding alto parts in the early twentieth century is the most obvious example of revision, but there are other situations in which old songs were changed—to relieve dissonance, for example. In compiling The Shenandoah Harmony, our motivation was more to contribute to the living tradition of shape-note publishing than to preserve the past. If we found a song that clearly had compelling features but seemed awkward or incomplete or had fewer than four parts, we first sought different versions in the old sources. Although the decision to change a song was a difficult one that we did not take lightly, we did edit songs, add parts, substitute texts, and arrange songs. All added parts and substantial rearrangements are indicated. Singers may omit the added parts if desired. Scholars and anyone interested in the history of shape-note music are encouraged to use the source abbreviation and page number on each song to find its original version.

The song “Zion’s Walls” (p. 109 in The Shenandoah Harmony) from The Social Harp, which may be familiar from Hugh McGraw’s recording The Social Harp: Early American Shape-Note Songs (Rounder, 1994) or the classical composer Aaron Copeland’s 1952 setting, is a good example of the committee’s editorial process. The original version (see the transcription, left) has clear errors—the text doesn’t fit the music, the bass part has an incomplete measure, and the repeat is ambiguous.

“Zion’s Walls,” original setting from The Social Harp (click to enlarge).

“Zion’s Walls,” as edited for The Shenandoah Harmony (click to enlarge).

In addition to correcting these errors, the committee made some subjective decisions (see the edited version, right). We decided that 6/4 would be a more appropriate time signature for the song, according with common practice of how songs are paced. We also omitted the fermatas, resolved the repeat by rebarring the beginning portion of the song, and added a second verse. As the original text has only one verse, we added a verse with a similar sentiment and meter from another text. One change we did not make, but considered, was adding an alto part. All four of the women on the committee sing alto, at least occasionally. However, our general policy was to add an alto only if the bass part felt awkward for an alto to sing, which was not the case here. The reception of the edited version has been positive—according to our current minutes, it is one of the more frequently called songs in the book.

Researching the songs and texts was an important part of the project—not only because we wanted to credit the composers, but also because knowing a song’s history helped us to locate different versions and make more informed decisions. Many of the old tunebooks have incorrect composer, text, or source information, or none at all. Attributions are a moving target as better information becomes available—for example, I recently traced “Vienna” (321b in The Shenandoah Harmony) from Wyeth’s Repository, Part II (1813) to a 1727 German publication. We are particularly fortunate to have had the assistance of many scholars, in particular, Nikos Pappas, Warren Steel, and John Martin, and numerous online resources, such as the Hymn Tune Index, IMSLP, and hymnary.org.

Page layout.

The physical construction of the book was a massive project in itself. After years of singing from The Sacred Harp, we all found the typography and layout of it and other older books most effective for singing and leading. In particular, we preferred a compact format, in which up to four songs can be displayed on one opening, even if the occasional misalignment between text and notes makes sight-reading more difficult. Compact formatting also reduces page turns and allows for the inclusion of more songs without making the book too heavy to hold. I led the typesetting team, which also included Robert Stoddard, Peter Golden, and Adrian Mariano. We typeset the music in Lilypond 2.14, with some modifications: we made the shapes bigger and the staff lines thinner, to reflect the fact that most singers look at the shape of the notes more than their placement on the staff. I used LaTeX, a scientific typesetting program, to compile and design the entire book with input from the committee. We used fonts inspired by early twentieth century typefaces. Dan took particular care in setting the order of songs. Although not strictly divided into three parts, as the 1844 Sacred Harp was, The Shenandoah Harmony is roughly ordered like an old tunebook, with more straightforward and accessible songs at the beginning. The committee decided on a cover design and color. John and Kelly found a printer, Bookmasters of Ashland, Ohio, and formed the Shenandoah Harmony Publishing Company to handle the business of publishing and distribution.

In response to requests from singers, the entire Shenandoah Harmony is now also available for sale as a hyperlinked PDF. LaTeX, which is more like a computer programming language than a word processor, allowed me to create and update multiple indices and add hyperlinks. Without the size and weight restrictions of a print book, the electronic edition has thirty-five additional pages, including, in addition to the standard title and first line indices, fuller composer and source indices, a metrical index, an index of texts organized by author, a chronological index of songs, indices of fuge entrances and choruses, an index of songs with fewer than four parts, and a bibliography with web links. All page numbers in the indices are hyperlinked to the songs.

Singers at Christ Church, Philadelphia.

Although this article mainly details the music committee’s contributions to producing The Shenandoah Harmony, the book would not have been possible without the support our singing community, who contributed countless hours helping us select, edit, proofread, typeset, and research the songs and texts, who gave financial assistance to the project, and who lent their voices and hearts in singing. Please see our web site for details on upcoming singings. We hope to sing with you all, and soon!

1. Myles moved from Washington D.C. during the course of the project and Robert served as an adjunct member. [↩]