With programmatic and vocal music, the listener’s task is facilitated by the inclusion of a story; the composer tells us what has been injected into the music and lies there, waiting to be heard. But can a plot really be encrusted in the musical fabric, or are we simply following the composer or librettist’s instructions and superimposing the story over the music? Would we not enjoy the music without knowing its program? Can it be expressive of something other than its program?

Even when we think of text and tune as inextricably linked, we find that associations have changed over time. It is well known that most Lutheran chorales started off as secular songs; a 16th-century German peasant hearing the tune known nowadays as O Haupt voll Blut und Wunden would have been reminded of sorrow caused by his sweetheart, not of the suffering of Christ on the cross. Likewise, Renaissance composers who used “popular” tunes as scaffoldings in their masses were likely not trying to merge secular and sacred. The meaning of a given tune was thus entirely due to its text, not to the pitches themselves.

In the case of purely instrumental music—hearing, say, Rachmaninoff’s music as “highly sexually charged” (as one recent letter writer described it), deeply nostalgic, or profoundly mystical, does not mean that the music actually contains eroticism, nostalgia, or mysticism, but rather that this is the emotion or state that one chooses to project, consciously or not, into the music. Even if Rachmaninoff himself had meant his works to be expressive of a specific emotional state, how could he have achieved this? Associations with specific musical patterns are too related to cultural and historical boundaries to be truly reliable. (Consider, as an obvious example, that the tritone had a different meaning for Bach than for Schoenberg.)

We also love to think of music as expressing every aspect of the life of its creator. There is the popular image of Mozart, knowing that he was about to die, becoming obsessed with death and writing his own Requiem. Even if any of these stories were true—and they have mercifully been proven wrong in most cases—there is a certain danger inherent in such generalizations. Does Bach’s love of beer make his music expressive of alcoholism? Or does Chopin’s lifelong struggle with tuberculosis make his works sickly? The logical extension of this idea is that, since every aspect of a composer’s life influences each and every one of his works, therefore all of his pieces must expressevery aspect of his life—which is to say that, in fact, they do not express anything.

Where does this lead us? It seems to me that modern musicology is more and more concerned with extra-musical associations rather than with music itself; yet, labeling works according to what they seemingly express does not make them more accessible, nor does it enhance them. Music is no greater because we think of it as sad instead of happy, as feminine instead of masculine, and so forth. Such characterizations tend to reveal more about the writer than about the work under discussion. It may be, then, that a greater restraint is needed, so that music can speak for itself instead of being forced to serve our agenda.