Note to users. If you're seeing this message, it means that your browser cannot find this page's style/presentation instructions -- or possibly that you are using a browser that does not support current Web standards. Find out more about why this message is appearing, and what you can do to make your experience of our site the best it can be.

Science 19 January 2007:
Vol. 315. no. 5810, p. 330
DOI: 10.1126/science.1134163
Prev | Table of Contents | Next

Technical Comments

Response to Comment on "The Geometry of Musical Chords"

Dmitri Tymoczko

The basic sonorities of traditional Western tonality divide the octave nearly evenly and are found near the center of the orbifolds Formula and Formula. Many common musical patterns exploit this fact, which permits efficient voice leading between structurally similar chords. In actual music, these patterns sometimes appear incompletely or are accompanied by additional notes. Using orbifolds in musical analysis therefore requires interpretive skill.

Department of Music, Princeton University, Princeton, NJ 08544, USA.

E-mail: dmitri{at}princeton.edu

Theorists agree that when analyzing music, there are two distinct phenomena to consider: the actual notes (the surface) and the common musical patterns those notes may imperfectly embody (the background). Orbifolds can be used to represent either phenomenon, although they are typically most useful when modeling background patterns.

Triads and seventh chords are the conceptual building blocks of Western tonality, the only complete harmonies recognized by traditional theory. These sonorities are found near the center of the orbifolds Formula and Formula. Theorists agree that in actual music these harmonies are sometimes incomplete and are sometimes accompanied by additional notes (doublings and "nonharmonic tones"). Thus, we cannot expect the surface of every tonal piece always to inhabit the center of some orbifold.

Figure 1A, reproduced from (1), is typically accompanied by an additional voice, as in Fig. 1B. The full progression lies on the singular boundary of Formula. Directly plotting Fig. 1B on this orbifold is not maximally informative, as its lowest voice operates according to what are generally recognized to be distinctive musical principles. My report therefore separates the progression into two parts: the upper voices, which exhibit efficient voice leading between structurally similar chords, and the bass, which adds harmonic support by leaping to the root of each chord (1). The upper voices exploit the geometry of Formula; the bass plays a different musical role.


Figure 1 Fig. 1. (A) A common classical upper-voice pattern that exploits the geometry of Formula. (B) In actual music, the three-voice pattern is often accompanied by an additional bass voice, whose function is to provide harmonic support by sounding the root of each chord. (C) A two-voice passage evoking (B). [View Larger Version of this Image (19K GIF file)]
 

Incomplete chords pose a related challenge. Most theorists would understand Fig. 1C to imply a succession of triads, as in Fig. 1B. We can represent the musical surface by plotting the incomplete chords of Fig. 1C on the orbifold Formula; we can represent the background by plotting Fig. 1A or 1B on the appropriate orbifold. Again, the upper three voices of the background pattern (Fig. 1A) make the most interesting use of orbifold geometry.

Brown and Headlam (2) observe that tonal phrases sometimes cadence on unisons. We can typically model these cadences as incomplete manifestations of a prototypical five-voice background pattern (Fig. 2). The pattern's top four voices use maximally efficient voice leading to connect nearly transpositionally related multisets, interestingly exploiting the geometry of Formula. If, alternatively, we are interested in the musical surface, we can again use orbifolds. As Headlam and Brown note, when a tonal piece cadences on a unison it moves from the center of an orbifold to its singular boundary. Thus, these cadences do possess a distinctive geometrical signature, even if we restrict our attention to the musical surface.


Figure 2 Fig. 2. Many tonal cadences, including all of the examples in (2), can be understood as variants of a prototypical five-voice pattern (A). The pattern's prototypicality is illustrated by the fact that it concludes several of the first pieces in Bach's Well-Tempered Clavier (WTC)(C to F), as well as several of Beethoven's first piano-sonata movements (G to K). By contrast, Headlam and Brown's cadence (B) concludes only one of the first 20 Beethoven piano-sonata movements and none of the first 20 pieces in the WTC. (A) The prototypical five-voice pattern. (B) The cadence from figure 1 in (2). (C) Bach, WTC Prelude 3. (D) Bach, WTC Fugue 3. (E) Bach, WTC Prelude 6. (F) Bach, WTC Prelude 10. (G to H) Beethoven, Piano Sonata No. 1, movements 1 and 2. (I to K) Beethoven, Piano Sonata No. 2, movements 1, 3, and 4. [View Larger Version of this Image (28K GIF file)]
 

Given these two analytical possibilities, it is unclear why Headlam and Brown claim that orbifolds cannot effectively model such progressions. Perhaps they have not clearly distinguished the musical surface from the background or have misinterpreted my remarks about the background (that the basic tonal sonorities inhabit the center of their respective orbifolds) as remarks about the surface. It is also possible that they have misinterpreted my Report (1) as an attempt to model all the style-specific norms of 18th- and 19-century composition (3). Instead, it was an effort to model a few general musical principles common to a wider range of Western musical styles.

Orbifolds are tools that can be used for a variety of musical purposes. Like any tools, they have limitations, but these limitations are considerably less constraining when the tools are skillfully deployed.


References and Notes

  • 1. D. Tymoczko, Science 313, 72 (2006).[Abstract/Free Full Text]
  • 2. D. Headlam, M. Brown, Science 315, 330 (2007); www.sciencemag.org/cgi/content/full/315/5810/330b.
  • 3. Headlam and Brown's fourth footnote suggests that this may be so. My report asked when structurally similar chords can be linked by efficient voice leading. This question admits a trivial answer, because one can always move all of a chord's notes in the same direction by the same small amount. My remarks about voice independence, beyond reflecting a general Western musical value, were meant to exclude this trivial solution from the discussion. (I did not suggest that any composer used independent voice leadings exclusively.) Headlam and Brown, however, interpret me as attempting to model the avoidance of parallel perfect intervals, a style-specific convention that was not very important in either medieval or modern music.
Received for publication 29 August 2006. Accepted for publication 20 December 2006.

To Advertise     Find Products

ADVERTISEMENT

Featured Jobs

Science. ISSN 0036-8075 (print), 1095-9203 (online)