> The image we have in our heads of a composer sitting at the piano and noodling almost certainly doesn't apply to a Beethoven or a Bach or a Schumann.
There's a mountain of evidence of Beethoven literally sitting at a piano and noodling. He started his career and was widely known as a professional piano noodler!
Furthermore, his early and mid-period music is dominated by the kinds of themes and figuration you'd expect from a professional piano noodler. Try singing op. 2 no. 1 piano sonata vs. fitting your hand to the keys.
It actually took Beethoven a long time and a lot of work to become proficient at writing things like lyrical melodies and non-clunky, non-pianistic counterpoint (and I'd argue he gave up on the latter and just figured out the best contexts for clunky counterpoint to be most effective).
Schumann also had some material in his symphonies that suspiciously fits the hands rather nicely on the piano. So it isn't so cut and dry as you make it out.
Bach was a bit different. At least for quite a stretch he was writing weekly cantatas where he'd do a performance on Sunday, then on Monday start writing the cantata for the next week. When you factor in the logistics of copying all the parts, practicing with the choir, etc., it doesn't leave much time for composing an entire service of music for small orchestra and choir.
And these are typically around 30 minutes of music, pieces full of sophisticated polyphony. Even if you memorized an entire cantata and time-traveled back to take his place I think you'd still be cutting it close just to copy it onto paper in the daylight you had available. And last I studied this, there was no evidence of a notepad or anything that he used to sketch out his ideas.
Anyhow, Bach and Mozart are the only composers I know of for whom the evidence suggests "transcribing from the head."
Edit: actually, even if the choir were one person to a part, Bach's cantatas are so full of fugal textures and virtuosic solos that it would still be challenging just to write everything down, copy the parts, and rehearse the music. I.e., it's not like writing 30 minutes of single-melody songs, scribbling some tab notation above it and letting the accompanists fill everything in accordingly.
In fact, I might be downplaying it. In addition to what most people think of as baroque polyphony, the cantatas also have a more-or-less simple "chorale tune" flowing either on top of that, or possibly in the organ pedals. And Bach wrote out the organ part that I believe he would have been playing himself.
So imagine a modern Church service with hymns and stuff, and paste that on top of what you think of as Bach's music to get a Bach cantata. :)
Really good response, thank you. If I could edit mine to reflect some of your points I would. The Grove Dictionary supports most of what you say. I think I need to concede that the early Beethoven might not have been able to do this stuff straight out of his head.
I stand firm that the latter day Beethoven would have had no trouble with it (technically; obviously not being able to hear is torture for any musician). I am a moron and a third-rate musician. I was able to compose from a desk after a couple of years of study in my early 20s. He spent his whole workday in music for decades.
> I stand firm that the latter day Beethoven would have had no trouble with it (technically; obviously not being able to hear is torture for any musician).
If "it" means "transcribing from one's head" the evidence-- fascinating, fun to read and analyze evidence-- points in the other direction. Composition didn't come easy to composers like Beethoven and Brahms, and they spent years working on it.
For Bach and Mozart, there is a variety of evidence that all points not only to their ability to "transcribe in their heads," but also to somehow imagining fully-formed phrases, sections, and even full pieces, all while incorporating new musical sounds, in their heads. Their music seems to be more or less sophisticated based on context and requirements, and not so much (or at all) on time allotted to write. There's really no other way to explain things like the triple dance (in 3 different meters!) in Don Giovanni being written in about a day, again with no pre-compositional scribbles. There are many more examples in the manuscripts.
In fact, for a long time scholars assumed a "compositional evolution" with Bach to arrive at a chronology for the cantatas that expressed that evolution. Scholars in the 1950's used forensic evidence to show that long-standing chronology was wrong. Apparently, this caused an enormous controversy because it left Bach scholars with no way to explain Bach's compositional development.
Finally, among other things, this distinction is a reason Beethoven's music dominated in the latter half of the 19th century. As a composer would you rather follow a path of using small, self-contained groups of notes to divide and conquer complex forms, developing slowly and surely toward mastery as you practice? Or would you rather try holding all the music you've ever heard in your head, attempting to re-organize and write variations in your memory, then then transcribe the most fitting parts you came up with for the task and hand and blindly hope what you came up with is at all coherent?
There's a mountain of evidence of Beethoven literally sitting at a piano and noodling. He started his career and was widely known as a professional piano noodler!
Furthermore, his early and mid-period music is dominated by the kinds of themes and figuration you'd expect from a professional piano noodler. Try singing op. 2 no. 1 piano sonata vs. fitting your hand to the keys.
It actually took Beethoven a long time and a lot of work to become proficient at writing things like lyrical melodies and non-clunky, non-pianistic counterpoint (and I'd argue he gave up on the latter and just figured out the best contexts for clunky counterpoint to be most effective).
Schumann also had some material in his symphonies that suspiciously fits the hands rather nicely on the piano. So it isn't so cut and dry as you make it out.
Bach was a bit different. At least for quite a stretch he was writing weekly cantatas where he'd do a performance on Sunday, then on Monday start writing the cantata for the next week. When you factor in the logistics of copying all the parts, practicing with the choir, etc., it doesn't leave much time for composing an entire service of music for small orchestra and choir.
And these are typically around 30 minutes of music, pieces full of sophisticated polyphony. Even if you memorized an entire cantata and time-traveled back to take his place I think you'd still be cutting it close just to copy it onto paper in the daylight you had available. And last I studied this, there was no evidence of a notepad or anything that he used to sketch out his ideas.
Anyhow, Bach and Mozart are the only composers I know of for whom the evidence suggests "transcribing from the head."
Edit: actually, even if the choir were one person to a part, Bach's cantatas are so full of fugal textures and virtuosic solos that it would still be challenging just to write everything down, copy the parts, and rehearse the music. I.e., it's not like writing 30 minutes of single-melody songs, scribbling some tab notation above it and letting the accompanists fill everything in accordingly.
In fact, I might be downplaying it. In addition to what most people think of as baroque polyphony, the cantatas also have a more-or-less simple "chorale tune" flowing either on top of that, or possibly in the organ pedals. And Bach wrote out the organ part that I believe he would have been playing himself.
So imagine a modern Church service with hymns and stuff, and paste that on top of what you think of as Bach's music to get a Bach cantata. :)