The academic study of music in the United States has traditionally been dominated by the use of notated scores. Undergraduate students, in particular, are often told that they must perform a score exactly as it is written, as the score reflects the composer’s intentions. This approach is very useful for students who are using scores as tools for learning performance techniques. However, it can have a bad effect on the way that we approach scores when we study music history: it can lead us to imagine scores are unchanging and authoritative.
For example, suppose you find a score by the musician Dietrich Buxtehude (d. 1707), such as his Passacaglia in D minor. Here is a score excerpt to look at:
There is a grand staff consisting of treble and bass clefs, labeled “manual”; and a lower staff (bass clef) labeled “pedal.” This is traditionally understood as notation for the organ, with the “pedal” part being played by the organist’s feet, and the “manual” part with the hands. There is no key signature, but the regular presence of both C-sharp and B-flat supports the idea that the music is indeed in D minor.
Both these assumptions are essentially correct. Buxtehude spent over forty years (from 1668–1707) working as a church organist in the German city of Lübeck, and the piece was probably intended for the organ (although it could just as easily have been played on a pedal harpsichord). Similarly, although modern terminology such as “D minor” hadn’t yet quite emerged in Buxtehude’s time, the music does mostly seem to follow tonal procedures and to use a D minor scale. The music seems easy enough to understand and reconstruct accurately.
At the same time, everything I’ve just said relies on a number of unquestioned assumptions about scores. In what follows, I’ll break them down one by one.
1. Tuning. We don’t actually know precisely what Buxtehude meant when he tells us to play the note “A.” Modern tuning systems tend to treat A as a multiple of 440 Hertz vibrations per second. This was not established as international standard tuning until 1939, and many ensembles still deviate from it slightly (often going up to around 445). Historic evidence of tuning standards is hard to codify, but depending on the type of instrument, the time period, and the place, the note A could fall between 388 and 487 Hertz. This means that in modern terms, the notated pitch A could sound anywhere from modern G to modern B. In Buxtehude’s case, we can refine this a little bit, based on the tuning of historic German organs, which suggests that his A was somewhere between 416 and 487 (roughly G sharp to B natural), with 463 (roughly a B flat) being quite common.
2. Temperament. Modern keyboard instruments are tuned in an equal-tempered system, so that A-flat and G-sharp, like all other black keys on a keyboard instrument, are equivalent to each other, just as all major scales are exact transpositions of each other. This was not always the case. A variety of tuning systems existed in the past. The traditional basis for tuning instruments derives from the pure mathematical ratios associated with Pythagoras. If one tries to employ Pythagorean intervals with a chromatic scale, some intervals sound unpleasant. Mean-tone tuning provided an effective workaround, but again with problem – one interval (usually a fifth) would be sour, meaning that depending on your tuning, some keys would be unusable (especially keys with multiple sharps and flats—this is why early notated music often sticks to keys with less than four sharps or flats). Consequently, multiple tuning systems arose; modern transposing instruments (such as the clarinet) reflect these older tuning traditions.
At the same time, these systems had a unique benefit. Because each semitone was not necessarily the same size in mean-tone temperament, a C major and D major scale were not exact transpositions of each other. Instead, each key had a unique sound based on its intervallic content. Because temperament wasn’t standardized and could vary in different places, broad agreement on the exact difference between G minor and C minor was rare, although musicians left behind many contrasting descriptions of the emotional differences between each key. The modern system of equal temperament makes all the keys usable, but at the cost of each key’s unique intervallic and emotional qualities.
Scholars still debate whether Johann Sebastian Bach’s “Well-tempered clavier” (book 1: 1724) used an early form of equal temperament or another tuning system which made all keys used while retaining their unique qualities. Buxtehude’s organ was probably tuned in mean-tone, but it might have used an experimental tuning developed by the organist Andreas Werckmeister.
3. Timbre. This is a simple point, but worth considering: the sound of the organ itself has changed since Buxtehude’s time. The number of stops and the sounds they produce have changed. Some organs are now powered by electricity rather than the traditional bellows, and the amount of air pressure that it’s possible to exert through the bellows has also changed, affecting the volume of the instrument. Certainly, it’s possible to play the music well on a modern instrument, but a modern performance will sound different. Related issues arise when we perform music written for the harpsichord on a modern piano, or if we perform lute music on a guitar. While we would have to do more research to learn about the organs that Buxtehude played when he created this piece, in later life he played an organ with fifty-four stops.[1]
4. The notation itself. So far, everything here assumes that we are indeed looking at a score written by Buxtehude, which is a faithful representation of what he meant us to see. But in fact, Buxtehude’s surviving manuscripts are generally written in organ tablature, not in lined staff notation. He might have copied his piece into staff notation himself, or someone else might have. And there might be mistakes or errors in a transcription from one form of notation to another. So how did we wind up with this music written in staff notation?
If you check the score example above again, there’s a German annotation: “Herausgegeben von Philipp Spitta. Neue Ausgabe von Max Seiffert” [edited by Phillip Spitta. New edition by Max Seiffert]. I found this score by going to the International Music Score Library/Petrucci (imslp.org), where one can download music that’s out of copyright. They helpfully provide the publication date of 1903 for this edition, which is roughly 200 years after Buxtehude’s death. Spitta (1841–1894) published an edition of Buxtehude’s music in 1876–1877, and is best-known for his multi-volume biography of J. S. Bach (finished 1880). Seiffert (1868–1948) was another music historian (Spitta’s student), who issued his own edition of Buxtehude in 1903, including updated versions of his teacher’s editions of the music. Both Spitta and Seiffert were involved in the business of establishing what Buxtehude wrote and creating a performable, readable text that contemporary musicians could play and scholars could study (organ tablature has generally fallen out of use). That means that whenever Buxtehude’s own notation was contradictory or unreadable, they would correct it and resolve any ambiguity; and if there were two contrasting versions of the piece available, they would decide which one seemed more authentic.
While both Spitta and Seiffert were respected scholars—among the foremost authorities on Buxtehude’s music in their own time—the score they created reflects their impression of Buxtehude. The 1903 edition does necessarily not reflect the composer’s unfiltered original concept, but unless we do more research, we can’t tell how much their edition alters Buxtehude’s music.
Sometimes, editors can impose too much on a score, transforming it to meet their own agendas. Because of this, some performers prefer to use “urtext” scores, which are supposed to represent the composer’s original notation and to clearly indicate any editorial changes.
In conclusion, it’s easy to take a lot for granted when we look at a notated score. But in reality, there are a number of variables to consider. Historic scores are far less fixed, prescriptive, or reliable than they look.
[1] Malcom Boyd, Bach (Oxford University Press, 2000), 21.
Matthew Franke's Music History Course Notes 