So, how can music compositions be structured and how does that relate to bridge building? The metaphor occurred to me because I was thinking about the problem of long compositions. How you structure a two or three minute piece is pretty clear, at least we have a lot of examples that are easy to analyze. Usually it is something like one phrase (8 measures or 16 measures as a period or sentence) followed by a different phrase, followed by the first phrase repeated. In Baroque binary dance movements we can label the first phrase A and the contrasting one B. This gives us a form A repeated, B A, then B A repeated. This is often shown as AABB, but since the A is often repeated after the contrasting B, this leads to a form called "rounded binary" which is better labeled AABABA.
That's not much like bridge building! By the way, I never took engineering in university, so I know almost nothing about bridge building, just so you know. But I have a feeling that the structure of something like this:
is based on factors like the strength of the materials used, the distance spanned, the weight of the bridge and basic principles of physics. In music the analogous factors might be the instruments used in the composition (orchestra, string quartet, solo harmonica, whatever) and their ranges and timbral qualities, the duration of the piece and? Ah, there is the rub! Because what comes next is where it gets very complicated. The composer has a myriad of confusing choices. Say you are writing a symphony. You have historical models in three, four and even five or more movements (Beethoven's Symphony No. 6 is in five movements). Each of these movements might be structured in various ways using forms as varied as "first movement sonata form", minuet and trio, scherzo, aria form, rondo form or something entirely different. But there are other symphonies in eleven movements, each of which is a setting of a poem for voice and orchestra--this is what Shostakovich chose for his Symphony No. 14. Or you could go in a completely different direction and write a lengthy single movement as Allan Pettersson did in several symphonies. His Symphony No. 9, for example, is in a single, uninterrupted movement about 70 minutes long. (Oddly enough, Shostakovich's Symphony No. 14 was written in 1969 and Pettersson's Symphony No. 9 in 1970.)
So how are these structured? I don't think that question has ever been answered with any thoroughness. When it comes to genuine theoretical understanding of the principles underlying the structure of a lot of music, we just don't know. It makes sense when you listen to it, but just how and why is hard to answer. We have a pretty good theoretical understanding of how music was structured up to, say, 1830, but from then on, the answers are rather tentative.
Unlike engineers, composers work from intuition, not a clear set of physical laws. Sometimes I think there are some basic musical principles, like tension and release, like dissonance and consonance, like pulse and sustain, but how these can be used in a composition seems to have so many applications, that general principles just don't seem to be evident. Still, it is tempting to think that musical structure can be as evident as engineering.
In fact, there was a school of composition that tried very hard to make this happen. It is a school of what Taruskin calls "maximal complexity" and that is also called "total serialism". The most well-known practitioners are Milton Babbitt and Pierre Boulez, who looked back at the music of Anton Webern for inspiration. These composers create a musical structure whose entirety is dictated by a set of basic principles used to create a tone row that is processed in various ways. Two interesting things came to pass with this method: first of all, audiences did not find the results very palatable and secondly, the principles and processes used became more and more complex and ultimately obscure so one begins to wonder if there is much difference between this method and intuition.
So, the upshot is that music composition is not very much like bridge building after all. But it would be very interesting to read what music theorists might have to say in, oh, a century from now...
Pettersson's Symphony No. 9 doesn't seem to be on YouTube, so here is the much shorter, at 43 minutes, Symphony No. 7:
UPDATE: I replaced the clip with one with better sound.
(One final thought: Pettersson's symphonies tend to be monothematic in the sense that the whole large form is spanned with a single theme or motif such as a scale segment or even a simple melodic cell like E to F. One is reminded of the simple, soaring arc of a suspension bridge...)