Serious Music

This dictionary of musical themes, by Harold Barlow and Sam Morgenstern, supplies an aid which students of music have long needed . . . We should now have something in musical literature to parallel Bartlett’s Familiar Quotations . Whenever a musical theme haunted us, but refused to identify itself no matter how much we scraped our memory, all we should have to do would be to look up the tune in Barlow and Morgenstern, where those ingenious dictionary‑makers would assemble some ten thousand musical themes, with a notation‑index or theme‑finder, to locate the name of the composition from which the haunting fragment came, and the name of the composer.

– John Erskine, 1948, in the preface to Barlow and Morgenstern’s A Dictionary of Musical Themes .

In the 1940s it must have been laborious to construct a dictionary of musical themes, but that’s what Barlow and Morgenstern went ahead and did. It is unclear whether anyone ever actually used it to identify the tunes that were haunting them, and, at any rate, it is obsolete today, given that our iPhones can tell us the name and composer of a song if you merely let it listen to a few bars. The iPhone software is called “Shazam,” a great advance over locutions such as, “Hey, can you Barlow‑and‑Morgenstern this song for me?” Now, in defense of Barlow and Morgenstern, Shazam does not recognize much classical music, which makes me the life of the party when someone’s Shazam comes up empty‑handed in the attempt to identify what the pianist is playing, and I pull out my 642‑page Barlow and Morgenstern and tell them it is Chopin’s Concerto No. 1 in E minor. And, I add, it is the third theme occurring within the second movement . . . because that’s how I roll.

The other great use I have found for Barlow and Morgenstern’s dictionary is as a test bed for the movement theory of music. Each of its 10,000 themes nicely encapsulates the fundamental part of a tune–no chords, no harmony, no flourishes. Most themes have around one to two dozen notes, and so, in movement terms, they correspond to short bouts of behavior. (Figure 18 shows three examples of themes from Barlow and Morgenstern.) There are at least two good reasons for concentrating my efforts on this data set.

Figure 18 . Example themes from the Barlow & Morgenstern dictionary. Top : A theme from Bach’s Partita, No. 1 in B minor. Middle : A theme from Beethoven’s Sonata No. 7 in D. Bottom : A theme from Sibelius’s Quartet Op. 56 “Voces Intimae.”

First, the dictionary possesses a lot of themes–10,000 of them. This is crucial for our purposes because we’re studying messy music, not clean physics. One can often get good estimates of physical regularities from a small number of measurements, but even though (according to the music‑is‑movement theory) music’s structure has the signature of the physical regularities of human movement, music is one giant leap away from physics. Music is the product of cultural selection among billions of people, thousands of years, and hundreds of cultures, and so we can only expect to see a blurry signature of human movement inside any given piece or genre of music. On top of that, we have the wayward ways of composers, who are often bent on marching to their own drum and not fitting any pattern they might notice in the works of others. Music thus is inherently even messier than speech, and that’s why we need a lot of tunes for our data. With enough tunes, we’ll be able to see the moving humans through the fog.

The Dictionary of Musical Themes is also perfect for our purposes here because it is a dictionary of classical music. “What’s so great about classical music?” you might ask. Nothing, is the answer. Or, at least, there is nothing about the category of classical music that makes it more worthy of study than other categories of music. But it is nevertheless perfect for our purposes, and for an “evolutionary” reason. We are interested in analyzing not just any old tune someone can dream up, but the tunes that actually get selected. We want our data set to have the “melodic animals” that have fared well in the ecology of minds they inhabit. Classical music is great for this because it has existed as a category of music for several centuries. The classical music that survives to be played today is just a tiny fraction of all the compositions written over the centuries, with most composers long dead–and even longer obscure.

Ultimately, the theory developed here will have to be tested on the broad spectrum of music styles found across humankind, but, for the reasons I just mentioned, Western classical music is a natural place to begin. And who is going to be motivated to analyze broad swaths of music for signs of human movement if their curiosity is not at least piqued by the success of the theory on a data set closer to home? As it happens, for many of the analyses carried out in the following chapters, we did also analyze a database of approximately 10,000 Finnish folk songs. The results were always qualitatively the same, and I won’t discuss them much here. At any rate, Finnish folk are universally agreed to be a strange and taciturn people, and they are (if just barely) in the West, so they don’t really broaden the range of our musical data.

With the Barlow and Morgenstern app installed in our toolkit, and with good Finyards slandered without reason, we are ready to embark on a quest for the signature of expressive human movers in music.

In this chapter we will successively take on rhythm, pitch, and loudness. As we will see, when we humans move, we have our own signature rhythm, pitch modulations, and loudness fluctuations. I will introduce these fingerprints of human movement, and provide evidence that music has the same fingerprints. I have at this point accumulated more evidence than can be reasonably included in this chapter, and so I have added an “Encore” chapter at the end of the book that takes up many other converging lines of evidence for human movement hidden inside music.