Pat your head, rub your tummy, improvise a fugue...

This is my third attempt at explaining this concept as I keep getting sidetracked with other ideas in the field of improvisation. Therefore, I won't try to define Improvisation yet, not talk about differing approaches, but just stick to the topic:

Can one 'reverse engineer' an existing (recorded) improvisation by assigning the different elements of the improvisation to either automatic or thought-out processes, and use this information for one's own improvisations?

It's a simple question! Yes or no?!

What does it all mean?

I should define some terms; it's really easier than it sounds.

1. I refer to 'reverse engineering' as the process that seeks to discover how something works by taking the completed object, taking it apart, and figuring out how it works and how it was made. Think of a crashed space-ship. What you do with your standard Earth-side crashed spaceship from outer space is to take it apart and fiddle with it and try to see if you can see how it works, so you can make your own. It's kind of the opposite of designing a spaceship from scratch. You know that this one already works (or did work) so if you can figure out how it was made, you're set!

The idea of the above question is if you can take apart an existing improvisation and figure out how it works and just maybe make one of your own using those blueprints.

2. There's also the automatic processes. When you are playing parallel fourths, or random notes, or a repeating ostinato, these are all techniques that, if you practice them enough, you can do without thought. At least, that is the goal. If you can do these things without thought, just running on autopilot, then you can spend your valuable and finite brain power on things that deserve it.

"This autopilot is great! Now I can concentrate on this delicious fish!"

As William Porter, one of the heads of the organ department at the Eastman School of Music, said, "One of the goals as we practice here is that you don't think. One of the things that can get in the way of improvising well is when you think too much about things that don't really require your thought."

"We can only think about a few things at once. I discover this every morning that I try to make breakfast. Something gets burned."

"But if you arrange this so that you can do it as automatically as you can, then you can think about something else."

3. Finally, there are the thought-out processes. Obviously, these ideas do require your thoughts. In this realm are the techniques of finer melodies, figuring out how to modulate to where you want to go, how the stretto will appear in your fugue, etc. These techniques, with some training can take less and less thought to accomplish but will probably never become entirely automatic.

The rest of your brain-power can then be devoted to the most important goal of music: expression, natural unimpeded expression.

The percentage of how much brain power is being used by automatic process, and thought-out processes, and natural expression, is impossible to tell, and can only be ascertained by experience. However, there is a natural question to be asked...

How much can we handle?

There are limits to human body as well as the mind. No matter how much drive, ambition, and positive thinking you have, flapping your arms isn't going to get you off the ground (not without some assistance). It would be great to have an idea of how much can be handled in an improvisation, but it's difficult to ascertain.

Have a listen to Marcel Dupre improvising a 5 voice double fugue on the Orbis Factor chant:

It was a fairly common occurrence to hear Dupre improvise such a fugue after mass each sunday; it was something he had much experience doing. This practice actually makes the whole process more automatic than if it was his first time, but that doesn't mean it doesn't require a lot of thought. Is he considering and shaping 5 independent musical lines the whole time? Not really. (For one thing, a five voice fugue often doesn't have 5 voices playing at the same time.)

You see, a carefully conceived musical line is one that you can hear before you play it. There are no surprises in it; it sounds just as you imagined it (mostly, anyway); every note matters. With some of the automatic processes, you don't know precisely how it will sound, but only have a general idea; the individual notes don't matter, but the texture and the effect do.

As Dupre introduces the first subject in the fugue, he knows exactly how it will sound. When he introduces the second voice with subject in the dominant, he has attention spent on what kind of line can accompany the subject, using contrary motion and independence of rhythm as a guide. Because of the stepwise nature of the second half of the subject, he opts for a string of suspensions. He uses those suspensions often in the course of the fugue. The third voice comes in, then the fourth, and fifth finally in the pedals. When listening, I don't hear any clear instances of more than two or three active voices at a time. The other voices merely walk from harmony to harmony.

Later when the two fugal expositions and developments are complete and he combines the two subjects, you can hear the other voices holding long note values while the feat of combining the two themes is accomplished. You can almost hear the brain power being diverted to where it is needed more.

Then we hear the stretti, where the subject is tossed from voice to voice, overlapping itself, which is a feat of mental gymnastics because each voice in turn gets the brain's full attention and is then shifted into the background.

One can see that with some knowledge and listening we can take Dupre's feat from the realm of superhuman divinely inspired magic, to the realm of a mature experienced and well practiced master with great clarity and organization of thought. Master though he was, there was nothing outside the range of 'ordinary' human ability; he was merely extremely well trained and experienced. (We should also keep in mind that Dupre was getting along in years when this recording was made; in his mid 70's and suffering from a form of arthritis that had warped his fingers making it necessary for him to adapt his technique to continue playing.)

Another drop in the bucket is in his treatise on improvisation. He gives attention to and prescribes exercises for completing melodies and harmonizing them in the soprano, pedals, and left hand tenor but nowhere does he have any mention or exercises for improvising simultaneous melodies. The closest thing is canon, but that is not two melodies, it is rather two of the same melody, at different times.

So What?

So where do we stand with our original question?

Based on the discussed information, I believe the answer is yes. Assuming that this is true (because assuming never gets one into trouble, does it?), we can observe and see new avenues for practicing.

One of the most helpful results of answering this question is that we can tell if the improvisation we are trying to perform is actually impossible to conceive!

As an improvisation, that is. Anything is possible on paper.

(A quick aside: One shouldn't be too hasty in calling something impossible. I recall that Art Tatum grew up listening to three handed piano arrangements of popular tunes. Not knowing that they were for more than one pianist, he thought that was what music was supposed to sound like and that's the way he learned to play, in a style that wasn't supposed to be possible for a single pianist. )

When listening to an improvisation, we should try to identify where the organist's attention is centered at any given moment. Of what's being played, what is the foreground and what is the background? Remember that the foreground should be thought-out material that is conceived before being played, while the background is less carefully thought-out and more of an automatic character; it's a texture, an effect.

Not all improvisations being equal, you may see that an improvisation is sometimes more background than foreground!