Sonata as Interrogation

Some philosophers writing about music really know what they’re talking about. Here’s Jerrold Levinson, on Prelude 12 from The Well-Tempered Clavier Book 2:

The basic key is F minor, which maintains its sway, more or less, through measure 19, despite a passing flirtation with F major in measures 4-8 and harmonic ambiguity in measures 16-19. At measure 20, A-flat, the relative major of F minor, surfaces firmly and continues to the double bar…

While I do know what all that means, it makes my head hurt, and thankfully Levinson also gives us the good stuff:

..overall I would characterize the effect as one of chaste and mildly elegiac wistfulness. ...the main motive possesses a familiar sighing quality, the dominant rhythm an air of hesitancy and reserve, the open texture and soft dynamic level convey a certain delicacy, the alternation of dominant rhythm and faster arpeggiated one an effect of charm, while the basic F minor tonality and harmonic side-glance at F major generate a mild sense of tension or unease.

Still it’s never good when philosophers learn only from each other. What do musicians say when they write about music? They will of course do plenty of “analysis”—tonic and dominant, stretto and inversion—but how much will they go in for calling music wistful, hesitant, reserved, delicate, or charming? And: assuming they do find “human characteristics” in music, which ones do they find, and which go unmentioned? Sad or joyful music is to be expected, but will music be called grumpy, or introverted, or interested in your stamp collection? I read Charles Rosen’s Beethoven’s Piano Sonatas: A Short Companion to find out.

Rosen calls his book “practical,” “meant as a guide for listeners and performers,” but that is clearly a joke. Stuff like this

The expressivo poco ritenente of bar 99 [of opus 111], however, practically applies to only one chord, the dominant ninth of the sub-dominant F minor with the D flat in the melody. (In spite of the notation, the a tempo of bar 100 should start with the upbeat of the previous bar.)

loses 99.99% of all listeners when the first Italian word hits the page. Long stretches of the book are performance notes: don’t play this faster than 76, that note should be held for its full value. But here and there Rosen does apply to music words that have their “home use” in application to human beings. One motif is called “hesitant” and “momentarily indecisive.” The finale of Mozart’s piano concerto in C is “amiable” (when played at its intended tempo); Mozart’s piano sonata in D ends with a “dashing” and “flamboyant” finale.

Well what is it about the finale that makes it dashing, and what it is about that motif that makes it indecisive? On one proposal, when the way a piece of music sounds and “moves through musical space” resembles the way a dashing person looks or acts, then it is a dashing piece of music; to hear a piece of music as dashing, then, is then to be aware, when hearing it, of this resemblance (if only subconsciously). On another proposal, when a piece of music naturally prompts us to imagine that the music is (somehow) the “sonic” behavior in which a (fictional) dashing person expresses their personality, then it is a dashing piece of music. These proposals—the appearance theory and the persona theory—differ in what they say engagement with, and appreciation of, the human characteristics of a piece of music involves; but both seem to handle “hesitant,” “dashing,” and “flamboyant.”

Does Rosen’s book provide any data that casts one of the theories in a better light than the other? Rosen quotes another pianist referring to the “fiery temperament” of a passage in two-part counterpart, and Rosen himself says that the Largo of Beethoven’s Ghost Trio is “frighteningly spectral.” Now hearing a “fiery temperament” in music seems to require imagining something (a person) having that temperament—a point for the persona theory. The score, however, remains tied, since, on the other hand, “spectral” is a pure appearance term.

There are two more moments in the book that, I think, push the needle more toward the persona theory. About the first statement of the theme of opus 109 (second movement) Rosen writes that “to understand its meaning one must, as usual, imagine the soprano line sung.” If appreciating the theme means imagining it sung, then appreciating it means imagine someone singing it, and that definitely means imagining a persona, in whom the various human characteristics attributed to the music could inhere.

The second moment comes when Rosen discusses the motif that opens the scherzo from the sonata in A flat. He calls it “an affirmation followed by a scherzando comment,” but then observes that the motif becomes, in a later statement, “a question answered by an emphatic new motif.” If the appearance theory is right, then hearing the motif as an affirmation, and later hearing it (when repeated) as a question, just involves hearing its resemblance to assertoric and interrogative speech. This I find hard to believe. Surely to hear it as an affirmation I must imagine it as someone’s affirmation.

This last example is interesting also because with it we have moved beyond, or away from, music exhibiting emotions or character traits. The idea that a bit of music can be happy or sad, or jovial or flamboyant, is one that philosophers have worked hard to account for. The idea that a motif can be an affirmation, a comment, a question or an answer is not really on their radar. It should be.

---

A earlier version of this essay was published in October 2022, under the title Beethoven: or, on human characteristics in music.

See also: On Appearance Emotionalism.