Improvisation: Responding to "The Bulletproof Musician"

The Bulletproof Musician [hereafter, TBM], one of those music-blogs that I've enjoyed reading over the last year or so, today came out with a post encouraging classical musicians to experiment with improvisation. Although TBM's articles are usually very stimulating, I felt that there were some gaps in this one. I've never been entirely comfortable publishing comments on other blogs, but today I decided to make an exception: classical improvisation is a topic that I've engaged with for many hours over many years, and I'd like to contribute a bit to the discussion.

First off, I have some philosophical differences with TBM. As a serious classical improviser myself, I don't agree that brain scans and memory lapses should be motivating factors in starting to improvise. (I know, I know, the memory-lapse discussion may be light comedy -- but the brain scans are a central part of TBM's article.)

The variations in brain activity between improvisers and non-improvisers is all very interesting, but far more important is the variation in musical, rather than mental, activity, both for performers and listeners. Our classical-music culture is saturated with performers who depend entirely on the printed page. To draw a comparison with language, imagine a world in which people could only recite sentences that they spent months practicing and repeating, but were unable to formulate any coherent, grammatically-correct (read: harmonically-sensible) thoughts on their own. This would not be considered a healthy environment: conversation topics would be severely limited, and people would get very bored very quickly. Yet this is the environment that we've fostered in classical music. In fact, I would go even farther: speakers who cannot spontaneously create a correct, if short, sentence, are not considered to be fluent in the language. Similarly, I think it's fair, if pessimistic, to claim that most classical musicians are not really "fluent" in the language of classical music.

Brain scans aside, improvisation liberates those musicians who do it, and engages those audience members who are there to witness it. As Robert Levin points out in the video that TBM cites, when an audience watches an improvising performer, they know that they are hearing something unique, something that belongs only to them, and something whose memory they'll cherish. They know that they won't be able to look it up on iTunes or Spotify the next day, and, accordingly, they'll listen differently. Performers in this situation benefit as well: huge portions of the music-related self-help material I've read involves ways to keep repertoire from going stale, or to keep performers from getting bored with pieces they've been playing. Well, those techniques are all fine, but if the performer is a habitual improviser, he may never come up against the problem of staleness or boredom in the first place. Improvising, whether cadenzas and ornaments or entire fantasies, keeps repertoire feeling new, and increases the risk and excitement of performing. This, too, benefits those on both sides of the stage.

Second, TBM's "Take Action" suggestion of doing some free improvisation is admirable, but ultimately, I think, off-the-mark. The great improvisers, from Bach and Mozart to Chopin (and Robert Levin!) don't improvise by playing random notes. Likewise, if we want to learn to do it, and do it meaningfully, riffing thoughtlessly on a scale isn't going to get us very far. To continue the comparison with language: David Foster Wallace pointed out that anybody who knows enough words can come up with a sentence like "Did you seen the car keys of me?", but that this is neither beautiful nor elegant. In music, of course, we need our sentences to be beautiful. A musician improvising a structureless melody may experience some illusion of improvisatory freedom, but, as scales, arpeggios and etudes teach us, true freedom in any pursuit comes only with technique and structure. I suspect that one of the reasons so few classical musicians improvise is that they believe improvising to be random -- that is, they think that the next note is a matter of intuition and emotion, and that if they don't "feel it", they must not "have it". Nothing could be further from the truth, though. As my old teacher once told me: "Improvising is easy, just like singing an opera, tuning a piano, or playing the violin; someone just needs to tell you how to do it."

For classical musicians who want to explore improvisation, there are a number of very useful resources, which I suggest perusing in the following order:

1. 1. A great starting point for anybody who plays a melody-instrument is the two-volume "Division Violin", available on IMSLP and also in a nice edition from Broude Brothers. These "divisions" are written-out improvisations over repeating ground basses (the classical equivalent to a jazzer's improvisation over a 12-bar blues), and they're very useful for developing freedom and a melodic vocabulary. These are important first-steps whether your ultimate goal is to improvise fugues, fantasias, or cadenzas. (They're also on 25% end-of-the-year sale from Broude Brothers right now!)
2. 2. From there, after using the "training wheels" of other peoples' written-out improvisations, those who can read German should try the excellent, very clear and complete, "Improvisation mit Ostinatobässen," published by Edition Walhall. This is a complete textbook for the beginning improviser. It offers suggestions for how to create mental frameworks for chord progressions, how to incorporate these frameworks into a variety of different forms, and how to start spinning out melodic figures from scratch. Sure, this book still focuses on repeating basslines, but this is an extremely useful way to build up an improvisatory vocabulary, and can then serve as a jumping-off point into other forms.
3. 3. Having worked through those ground basses, the next step is to free yourself from the repeating bassline pattern. If your intended destination is the classical cadenza, go straight to CPE Bach's "Essay on the True Art of Playing Keyboard Instruments". The final chapter contains many useful sample chord progressions, which serve as excellent models for fantasias and cadenzas of any length. And next to the extended chord progressions, he also shows many useful shorter progressions that go to very distant keys and back. These should be internalized, and are great for getting out of harmonic binds. (Note for violinists: there's another useful section, distilling some of the same information, in Baillot's Treatise.)
4. 4. Finally, for those who want to take the next step, Music in the Galant Style is an excellent resource. It is a compendium of partimenti -- stock phrases, chord progressions, melodic figures, etc. -- in a variety of styles, and in a variety of different contexts, from Bach's fugues to Mozart's Die Zauberflöte. This text is great for making the leap from CPE Bach/Baillot into coherent, larger-scale improvisation, and for improvising with/in counterpoint.

Bach Explored: By and For Pisendel

"Shall I program wonderful, known, popular works, or something unknown that I think my audience will love?" While musicians may have asked this of themselves at any point over the last few centuries, I imagine that the question has never been as pertinent as it is now. Audiences have access to more music, both famous and obscure, than they ever did before, and -- thanks to digitized archives from around the world -- performers are able to rediscover entire libraries of unknown works without even leaving their homes. But, in fact, our focus on what the audience will hear, and what music the audience will or won't know, may actually mask another, equally important difference between (some of) the popular and obscure repertoires.

Specifically, it strikes me that much of today's known, popular early-music repertoire consists of works written by the composer for other people, while a lot (though not all) of the still-unknown music was written by composers for themselves. Of course, this isn't always the case, but the correlation seems pretty consistent: sonatas by Handel, Bach, and Telemann, still the most-frequently-played baroque violin works, were either written for other violinists or for publication. At the other extreme, a violinist-composer like Johann Georg Pisendel published nothing during his lifetime, and composed primarily for his personal use. As one might expect, his surviving manuscript works are quirky and idiosyncratic, designed to suit his own styles rather than those of anonymous consumers.

Aside from the obvious pleasures of reconstructing obscure repertoire, one of the wonderful things about practicing the music of Pisendel is that it offers us a tantalizing glimpse into the mechanics of his hands and mind. Spend enough time with his unique, recurring technical demands, and you begin to feel what sorts of things he was good at, and what he liked to do on the violin. He quickly becomes a living, breathing, violin-playing personality. In Pisendel's case, this musico-physical understanding is particularly enticing: as a student of Westhoff, Pisendel was firmly rooted in the 17th-century polyphonic violin traditions; at the same time, however, he influenced countless composers of his own generation, and was the recipient of dedications by Albinoni, Bach, Tartini, Telemann, and Vivaldi. And, most useful for my current purposes: his violin sonatas are extremely challenging. (Although some performers might find Pisendel's technical demands gratuitous, there's no better way to get an idea of his abilities than to try to learn his hardest music.)

Technique was so central to Pisendel's craft that it sometimes seems each of his sonatas is "about" a different technical challenge, as though he thought first of whatever technique he wanted to showcase, and then structured the music around it. The C-minor sonata (come back for the second Bach Explored concert, in January!) is all about thirds high up on the D and A strings; the unaccompanied sonata is all about contrary-motion double stops; the D-major sonata, on my music stand for early next month, is all about micro-sequences in which a single hand-shape moves up the fingerboard in small increments. This isn't exactly like the standard baroque practice of writing a melody and then repeating it sequentially; this happens on a tiny scale, perhaps five or six times within a single phrase.

He uses this technique all over the sonata, but its most vivid manifestation comes during the last movement's cadenza: the performer strikes an exceedingly awkward left-hand pose and then takes it on a stepwise tour up and down the fingerboard:

In Pisendel's hands (as it were), mere mechanics can thus take on real musical importance.

If he was indeed the violinist for whom Bach wrote the Sonatas and Partitas, as popular legend holds, then it's tempting to view a passage like this -- from the C-major fugue -- as having been designed with Pisendel in mind. Note the similar incremental rise and fall:

Of course, in the pre-chinrest era, everybody was shifting in steps. Pisendel may have inspired parts of the C-major fugue, but ultimately we have no idea what was actually on Bach's mind when he came up with the above passage. Still, aside from the Sonatas and Partitas, in only one other violin work is Bach so obsessively sequential on such a small scale -- and, in this case, the manuscript (of BWV 1023) survives in Pisendel's own collection, so we can be quite sure that he played it. One particularly sequential passage in the first movement hardly requires any left-hand finger adjustment -- just Pisendel-style micro shifts:

After becoming acquainted with all of Pisendel's violin sonatas, however, one begins to wonder whether he would have been very good at the rather sophisticated bowing technique required for this passage. If we can judge by his surviving works (can we?), his brilliant left-hand technique far outstripped that of his bow arm…but that's a musing for a different post.