My friend Gerry Sussman, of MIT, called to ask if I’d anything to say on parallelism of art and engineering. (He was preparing a talk.) I did have a remark that I’d never shared with anyone. “A work of art,” said I, “is a machine with an esthetic purpose.” Gerry got a kick out of this, and used it in his talk, later commenting to me,
I was trying to point out that the thought processes of artists are not extremely different from the thought processes of engineers, or scientists, or mathematicians. [Gerry is all three of these, and a watchmaker too.]
In particular, an engineer designs a machine for a purpose. To do so he/she selects parts that will contribute to the desired behavior of the machine and interconnects those parts in such a way as to obtain that behavior. The interconnection may not be easy and the parts may have to be modified or interfaces constructed to fix various bugs that will appear.
Quite similarly, an artist, a poet for example, may design a poem to have a particular emotional impact on the reader. The poet may select certain patterns of words that have part of the effect and glue them together, with great precision, to build the desired impact. They may not go together easily, so there is an inevitable snipping and filing and sanding and polishing, and otherwise manipulating the interfaces to make the poem work.
This was forcefully brought to my attention by the essay “Philosophy of Composition” by Edgar Allen Poe. In this essay Poe explains the construction of “The Raven”. He says: “I select ‘The Raven’, as most generally known. It is my design to render it manifest that no one point in its composition is referrible either to accident or intuition—that the work proceeded, step by step, to its completion with the precision and rigid consequence of a mathematical problem.” There are similar comments by Baudelaire. —Gerald Jay Sussman.
I’ve discovered similar comments by the poets William Carlos Williams (“A poem is a small machine made out of words”) and Paul Valéry (“A poem is really a kind of machine for producing the poetic state of mind by means of words”).
It’s said that the painter Edgar Degas wanted to write poems, and complained to the poet Mallarmé that he couldn’t seem to write well, though he was “full of ideas.” Mallarmé replied, “My dear Degas, poems are not made out of ideas. They’re made out of words.” There we see the necessity for Sussman’s engineer to “select parts that will contribute to the desired behavior”: not ideas but words.
To some people, the idea of “a machine with an esthetic purpose” will be natural. A piano is a machine with an esthetic purpose. An audio recorder is another. So is a player-piano. Thinking of a piece of music itself as a machine is not a difficult step.
Other people are repelled by the idea; I ask them to notice that the “machine” point of view is useful, as it leads us to look for, and expect to find, understandable structures and processes, which are indeed there to be found.
Two examples in music:
Some fugues have strettos, passages where the “subject” of the fugue appears once in each voice, each time beginning before the previous one has finished. This overlapping will not have occurred earlier in the fugue, so the stretto is a formal intensification. Not all fugues have one; but when present, it comes at or near the end, to help create the climax. Thus a stretto has a purpose—intensification— and a mechanism appropriate to the purpose—overlapping of the subject.
Or consider the opening of Schubert’s Sonata in B♭, D. 960. Four of the five notes of the first chord (B♭,B♭,F,B♭) begin a harmonic series. The next note in the series, D, should appear above the fourth harmonic; instead it shows up two octaves lower. This displacement (to my way of thinking) creates potential energy which becomes kinetic, so that this voice moves twice as fast as the others: in 8th-notes, not quarters. This passage functions to get the movement under way (always an interesting compositional task) using an appropriate mechanism—harmonic displacement.
It would be fun to analyze an entire piece of music at this level of detail; but the proposition reminds me of Harvard mathematics professor Andrew Gleason telling me he’d once analyzed some math in terms of set theory—it’s a cliché that all math is reducible to set theory—and how very much harder it was than you’d expect from what a cliché it is. “Let them try it!” said Gleason.
One who did “try it,” and I think succeeded, was Deryck Cooke, in his brilliant book The Language of Music. Cooke’s thesis is that the main characteristic of music is to express and evoke emotion; and that tonal music uses the same, or closely similar, melodic phrases, harmonies, and rhythms to express and evoke the same emotions. Ranging from plainsong to Stravinsky, Cooke argues that music is a language in the sense that idioms can be identified and a list of meanings compiled; and he analyzes symphonies by Mozart and Vaughan Williams.”
In my class at California Institute of the Arts (CalArts) in 1973, we tested student perceptions of emotions attached to the degrees of the chromatic scale—for Cooke the basic units of meaning—and found excellent agreement with Cooke.
