Conlon Nancarrow (1912-1997) turned to player pianos because they could reproduce complicated rhythms that were impossible for a human to play precisely. He also used different types of symmetry to compose pieces such as this canon. These symmetries are visually evident in the rolls themselves.
Thanks to some perceptive comments on fasola-songwriters and elsewhere, I’m going to revisit my previous post on rhythm and meter. Two comments that intrigued me were Leah Velleman’s idea that there might be a generative theory of rhythm that applies to shape-note hymnody and Tarik Wareh’s observation that rhythm and the placement of bar lines are not independent phenomena. Another suggestion, emailed by a friend, was to look at higher-level accents.
I hope you’re not sick of LOUISIANA (SH 207), because I’d like to start there again. I had classified settings of the text as “even” if their accented syllables were evenly spaced. However, as Leah and others pointed out, I allowed some fudging at the end of lines, so my “even” rhythms weren’t strictly even. Here’s a rhythmic setting of “Come, little children” that is completely even. Each accented syllable starts a bar, with the unaccented syllable or syllables following it. The slurs indicate the structure of the poetry.
But this seems all wrong. There’s nowhere for the singers to breathe between phrases. Moreover, 2/4 is the only common choice of barring that preserves the musical symmetry we expect between the parallel two-line phrases of text. Here’s what happens if we bar in 4/4:
However, if we add a little time in the middle of the verse, order is restored:
This rhythm, though no longer “even” in terms of the pattern of accented syllables, may be grouped into higher-level units of two, four, or eight measures. Lengthening “morsel” (and “postle”) more or less gives a second accented syllable to that line, balancing the first line:
Come, LIT-tle CHIL-dren, NOW we MAY
Par-TAKE a LIT-tle MOR-SEL.
The “swung” version of this pattern, barred in 6/4, with each two-note grouping replaced by a half note followed by a quarter note, is exactly the rhythm of the first four lines of Walker’s BABE OF BETHLEHEM (ShH 103).
However, the rhythm in Example 3 is not actually the rhythm of LOUISIANA, which goes like this:
What I take from this is that giving more breathing space between lines (or at least pairs of lines) can take precedence over the binary symmetry of Examples 3 and 4—that is, it is acceptable and often desirable to have five-measure phrases. After all, we’re singing, not dancing. Moreover, making “Come” a half note creates a much easier entrance for the class and the leader.
Let’s turn now to the long-short-short-long rhythm of THE CHURCH’S DESOLATION (SH 89). I called this pattern “uneven,” but it’s actually uneven in a specific way: the distance between successive accented syllables alternates between two beats and four beats. I’ve put dotted lines before each accented syllable.
There is no obvious way to bar this rhythm. If I were not familiar with the song, I would bar it in 3/2, which emphasizes the parallel long-short-short-long rhythmic patterns and divides each line of poetry in half–a division that is also present in the poetry of the first and third lines.
As you know if you’ve looked at the song, this is not the way this rhythm is actually barred:
Is the rhythm, then, misbarred in The Sacred Harp? The “misbarring” football has been kicked around on fasola-discussions and elsewhere. And in fact, when arrangers “reform” earlier settings of a tune that has the long-short-short-long rhythm, they often change a 4/4 or 2/2 barring, which is typical in early shape-note sources, to a 3/2 barring—see my discussion of the history of BOURBON. An opposite phenomenon happens when a 3/2, reformed-harmony song like Lowell Mason’s GRAVITY (CB 266t / ShH 4b) or BOYLSTON (CB 447t) enters the shape-note world and is barred in duple time. Although it’s true that the accents of the poetry don’t line up as well with the text, Example 8 has more life in it than Example 7. There’s a polyrhythm—a rhythm produced when the same unit of time is simultaneously divided in two different ways—between the duple-time pattern of the barring and the triple-time pattern of the note values and text that is just soooo satisfying. And since Sacred Harp singers “beat time” with their arms while singing, the duple-time rhythm is embodied, as well (see this video, starting at 2:00). So I wouldn’t say the rhythm is mis-anything, any more than I would say that shape-note harmonies are “incorrect.”*
I can see that I’ll have more to say on the subject. I still haven’t addressed Tarik’s question of why I consider the settings of THE DYING CALIFORNIAN in the Denson and Cooper Sacred Harps to be fundamentally different. And I wanted to look at different meters and songs that have more complex interactions between musical and textual rhythms, though let’s not say “misbarring.” I’d love to discuss songs like Walker’s TENDER-HEARTED CHRISTIAN (ShH 270) as well. Let’s call this post “Part 1.”
*My preference for the barring in Example 8 raises the question of whether it is appropriate for a hymn to have “life,” at least in this obviously rhythmic way. If sung quickly, the rhythm of THE CHURCH’S DESOLATION does, I think, clash with the gravity of the words. The nineteenth-century reformers sought to make church music more reverential. One of the ways they did was to slow tempos and regularize rhythms and harmonies. “Unreformed” shape-note music is, mostly, not part of church worship.
Here’s a sequel to my previous post on tune families. After reading Charles Seeger’s article ”Versions and variants of the tunes of ‘Barbara Allen,'” I was intrigued by the idea of adding rhythm to my analysis of tune families. In this post, I’m going to explore the contribution of rhythm to a tune’s identity. Since settings of the same tune family can vary in four dimensions—pitch, time, text, and harmony—I’d like to incorporate rhythm into the study of tune families and also consider the existence of “rhythm families.”
First of all, let’s distinguish between meter and rhythm. Meter, in this context, is the pattern of stressed and unstressed syllables in the poetry. It is unaffected by how the text is set to music. Rhythm, on the other hand, is the pattern of musical note durations in a tune. Normally, there is a relationship between the meter of a text and the rhythm of the tune chosen for that text. However, the same meter may be expressed in several different rhythms in different songs, or even within the same song.
LOUISIANA AND IAMBIC 8s & 7s.
LOUISIANA, page 207 in The Sacred Harp, 1991 (SH), is a good illustration of how rhythm is incorporated into the identity of a tune. Its meter is 8, 7 D. iambic, meaning eight lines of iambic feet, alternating eight and seven syllables, with this stress pattern:
Come, LIT-tle CHILD-ren, NOW we MAY
Par-TAKE a LIT-tle MOR-sel.
My reduction of the song is below. To make a cleaner picture, I’ve used only one note per syllable. Often different instances of the same rhythm are barred differently, so I’ve chosen not to show a time signature or bar lines. The dark bars indicate phrase endings, which often do not correspond to musical barring. It’s common for lines of poetry to occupy the same amount of musical time, even if they do not have the same number of syllables. However, in order to set the song in a familiar time signature—not something like 9/4, as this appears to be—rests are normally added between phrases, as in The Sacred Harp’s setting of LOUISIANA. Since I don’t find the rests an essential part of the story, I’ve left them out.
I’ll call this example “even” because the syllables are set more or less evenly, with some correction at the ends and beginnings of phrases. Other 8, 7 iambic songs with almost the same rhythm are MECKLINBURG, Shenandoah p. 259, and LOOK OUT, SH p. 90.
Now let’s experiment with changing the rhythm, while keeping the notes of the tune intact. Here, I use the rhythm of BABE OF BETHLEHEM, page 103 in The Shenandoah Harmony (ShH).
This rhythm is derived by “swinging” the original—that is, two equal notes are replaced by a long-short pair, with the long note falling on an accented syllable. It doesn’t take much to convince me that, in this form, LOUISIANA is almost identical to ZION’S CALL (ShH p. 131).
These two rhythmic patterns, or variations of them, seem to be the most common for 8, 7 iambic poetry. However, they are not the only choices. Let’s try LOUISIANA with the rhythm of THE CHURCH’S DESOLATION, page 89 in The Sacred Harp. If we think of an accented syllable followed by an unaccented one as being a “unit” (a trochaic foot—the pattern that typically functions as a unit in musical settings), the lengths of units are uneven, though the pattern is quite predictable.
In fact, LOUISIANA is a relative of THE CHURCH’S DESOLATION. You can especially hear this in the B part of the music. There’s even a LOUISIANA in The Hesperian Harp that uses this long-short-short-long rhythm. Although this pattern seems syncopated when sung quickly, it has a long history in church music, going back at least to the mid-eighteenth century. Note that the words “morsel,” “union,” and “communion” have what’s called a “Scotch snap,” meaning a short-long rhythm where the shorter note gets the accented syllable. I’ll call that pattern a snap.
Whether I hear the same sequence of notes as more like LOUISIANA, ZION’S CALL, or CHURCH’S DESOLATION depends on which of these rhythms I choose. There is an additional rhythm for 8, 7 iambic that I’ve not seen with a LOUISIANA relative, though perhaps someone else will find one. CHRISTIAN’S DELIGHT (ShH p. 359) uses essentially this pattern. It is a swung version of Rhythm 3, with the two long notes replaced by a long-short pair. There are snaps in the same places as Rhythm 3. JESUS DIED FOR ME, page 511b in the Cooper edition of The Sacred Harp, has the same rhythm.
Rhythm 4: Uneven, with swing.
Oddly enough, Rhythm 4 is very common in Seeger’s transcriptions of “Barbara Allen,” while Rhythm 3 is not found in the oral recordings. Perhaps the need to write the song in musical notation, coupled with the preference for duple time signatures like 4/4 and 2/4 rather than triple time signatures like 3/2, meant that Rhythm 3 was preferred to Rhythm 4 in shape-note books (Rhythm 4 must be set in triple meter, while Rhythm 3 can be duple or triple). For example, contrast the Cooper Sacred Harp‘s 1902 setting of WAYFARING STRANGER in duple meter, using an analog of Rhythm 3, to the Denson Sacred Harp‘s 1935 setting of the same tune in triple meter, this time with a pattern like Rhythm 4.
The four rhythms above and their close variants exhaust the possibilities I’ve seen for 8, 7 iambic in the Denson and Cooper editions of The Sacred Harp and The Shenandoah Harmony. There is a metrical index for The Sacred Harp, 1991 HERE and for the Cooper book HERE. Our eBook has a metrical index as well.
A side note—George Pullen Jackson and others have traced the LOUISIANA group of tunes, including also THE FLOWER (ShH 41) to a Scottish song Wae’s me for Prince Charlie (also HERE), which in turn is related to a 1615 tune, Lady Cassille’s Lilt. See Bronson, The Traditional Tunes of the Child Ballads, 1959, vol. III, page 198 ff.
TROCHAIC 8s & 7s.
It seems that we might be able to describe “rhythmic families,” meaning collections of rhythm patterns that have similar characteristics, in the same way that we can attempt to classify tune families. Let’s turn now to the 8, 7 trochaic family, a larger grouping that includes the “Come, thou fount” text that I discussed HERE. The trochaic foot is a stressed-unstressed pattern: “COME thou / FOUNT of / EVE-ry / BLESS-ing.”
Example 1: Even.
ShH 380 AUTUMN, 1787.
Example 1 is the most common rhythmic setting of 8, 7 trochaic. I found the following in The Shenandoah Harmony: 248b TRIUMPH, 76b MOUNT WATSON, 432t NEW MONMOUTH, 264t ANIMATION, 244 CONCERT, 334 GETHSEMANE, and 36 OUR JOURNEY HOME. Examples in The Sacred Harp are 56t, 59, 80t, 117, 145t, 148, 166, 312b, 333, and 458.
Example 2a: Even, with swing.
ShH 55 BALL HILL, 1844.
Other examples of 2a are also ShH 226 THORNY DESERT (A part), ShH 118 LOCHLEVEN, and SH 370 MONROE.
Example 2b: Even, mixed swing and straight.
ShH 246t PALMS OF VICTORY, 1854.
The “swung notes” can be other places in the tune—see ShH 293 CELESTIAL WATERING. This is also a popular choice. Instances of Example 2b in The Sacred Harp are 30t, 54, 135, 144, 145b, 323t, 332, and 385b.
Example 2c: Even, with swing and snaps.
226 THORNY DESERT (B part), 1835.
Snaps normally come at the end of a line of text. ShH 241t SINNER, CAN YOU HATE THE SAVIOR is similar. VILLULIA, 56b in The Sacred Harp, has an additional snap in the middle of the first line. Other settings like Example 2c in The Sacred Harp are 52b, 118, and 154.
Example 3: Uneven (short-short-long-long).
ShH 253b MY NATIVE LAND.
Example 4: Uneven, with swing (almost).
SH 410t THE DYING CALIFORNIAN, arr. Howard Denson, 1935.
Example 4 is Howard Denson’s 1935 arrangement of the song, which is now in The Sacred Harp, 1991. The rhythm is close to regular, but doesn’t fit the “uneven, with swing” pattern perfectly because the phrase “limbs are” is not swung. Another example of the “uneven, with swing” rhythm—this time, a perfect match—comes from the 1927 song SHE IS SLEEPING, p. 540 in The Sacred Harp, Revised Cooper Edition, 2012.
The original 1859 version of DYING CALIFORNIAN is in the Cooper book (CB) and on page 410 in The Original Sacred Harp, 1911:
Example 5: Mixed even and uneven.
CB 398b THE DYING CALIFORNIAN, 1859.
I would classify the first phrase as even and the second as uneven. I wonder whether Denson regularized the rhythm in 1935 to match how people actually sung DYING CALIFORNIAN, or if he was imposing his own idea of how the song “ought” to be.
In another category are three songs from the Cooper book: 449 BEAUTIFUL RIVER (1864), 454 THE BLESSED LAMB (arr. 1902?), and 483 WHERE THE SAINTS ARE PASSING OVER (1882). The composers draw out the phrase expressively in some places and compress it in others, in a way that seems to me more typical of gospel music than earlier shape-note repertoire. Example 6 is a phrase from BEAUTIFUL RIVER. The first line has the same rhythm as “I’ve been working on the railroad,” a glee from the 1890s.
Example 6: Expressive, regular.
CB 449 BEAUTIFUL RIVER.
See the Original Sacred Harp, 1911, page 460, for this tune from 1869. The fact that THE BRIDE’S FAREWELL is one of the fifteen least popular songs chosen (out of 557) by Sacred Harp singers is, I’m sure, partly due to its maudlin text and angular melody, but the awkward rhythm doesn’t help either.
The alert reader may have noticed that I omitted text settings from fuges or part songs. There are three, all from the twentieth century: SM Denson, TJ Denson, and JS James’ THE GREAT ROLL CALL (1909), CB 229, Paine Denson’s PEACE AND JOY (1959), SH 532, and Hal Kunkel’s TEN THOUSAND CHARMS (1996), ShH 140b. These songs are all popular in their respective books. I won’t analyze them here, but it is informative to study how each composer effectively varies the “traditional” rhythmic patterns for this meter.
The classification system I have developed for 8s & 7s—even, uneven, swing, straight, snap—can be applied to different meters. What other terminology is necessary to describe the full catalog of meters? Certainly there are circumstances we have not encountered yet, such as “units” of stressed-unstressed-unstressed (dactyls) that occur in meters like 11s. Do the vast majority of plain tunes fall under the same general categories?
The rhythms I classified as “expressive” in that the durations of individual notes vary widely are associated with the gospel era (late 1800s and early 1900s). It also seems that the “uneven, with swing” rhythm appears relatively late. What is the chronology of rhythms? I find it particularly interesting that the rhythms of several songs in The Sacred Harp have been changed in the early twentieth century.
Finally, how have shape-note composers of the past set text rhythmically in fuges and part songs? Have they more or less stuck to established models, or have they made innovations? What can modern-day composers learn from this?
Another 8, 7 “farewell” song with irregular timing, arranged by William Walker in 1854:
The song appears in The Christian Harmony, 1866 with the same rhythm. The editors of the Deason-Parris revision of The Christian Harmony (1958) “corrected” the rhythm by using a rhythm similar to the pattern chosen by William Hauser in his 1848 setting of the tune. The Shenandoah setting (p. 14) by Dan Hunter is based on Hauser’s arrangement.
Steve Reich’s “Clapping Music” uses a single repeated rhythm pattern. The two players start by clapping the 12-beat pattern together in unison. After eight measures, one performer shifts the pattern forward by one beat and they clap the new pattern for eight measures. This process is continued until the two patterns align again. Here’s a visualization. Reich says that the piece was inspired by the clapping patterns in traditional flamenco music of the Roma people in Andalusia, Spain.