The metaphysics of improvisation

TL;DR: In this paper, the authors present a taxonomy of the most common types of choices available to an improviser at the time of performing an improvised piece, including the most important ones from a phenomenological perspective.

Abstract: and general in order to be able to be performed on a wide variety of new inputs. The inputs may be an inventory of notes based upon key, style, et cetera. (Pressing’s cognitive model calls this the “referent.”) My taxonomy is a way of making those inputs clear. 184 Since improvisation (and composition) is fundamentally cognitive and motor selection, then this presupposes a set of things from which to select. One can only select if there are options, choices. Now, it seems that whether a person (agent) is aware (conscious or cognizant of the options (all or even some)), one may always post facto reconstruct the set of choices which were available or present to the agent at the time of selection. By this I mean the set of choices that were available to the agent from an objective point of view. This set has little to do with the actual, individual conscious states of the agent; however, it does involve many specific conditions of the agent and the agent’s environment. For example, a musician S may say that “it never occurred to me to play that B-flat after the A,” even though objectively that choice was available to S. Sometimes, however, we describe others, and even ourselves, as just doing something—no other options presented themselves to consciousness. So, “selection” may seem like an inappropriate term or concept for what is going on in improvisation. It may, however, seem more accurate in composition. When humans perform actions in quick succession, consciously it does not seem like a choice or decision is being made for each separate action. In fact, in some cases it may be difficult to individuate the rapid succession of actions into discrete units. It seems to be a unitary flow of movements. These are half-intentional actions. Beside the (SCI) case, examples of this kind of phenomenon are playing sports, talking, and just mundane actions like walking to the market. From a phenomenological perspective, in some moments the choice or decision aspect can be discerned, while other moments “feel” automatic. Consequently, it is in these seemingly automatic moments that selection may be an erroneous description. But there are several pieces of evidence that suggest that in both cases similar or the same processes are realized. First, it would be impossible to account for talent and skill if some sense of choice or selection or decision was not involved. Indeed, psychologists and others 185 indicate that some people are better than others (usually in some specific domain of behavior) in their speed of thinking, choosing, and moving in situations that require rapidity. In other words, if we cannot attribute responsibility to selection or choice, even in an attenuated sense, training and effort would be diminished or demolished. Why would one train if one could not control the automatic thinking or moving? What would be our understanding of talent and expertise? Second, there is reductio ad absurdam argument that can be given here, analogous to the one Thomas Nagel presents in the classic “Moral Luck” article. One could argue that artistic agents are never responsible for anything they do; they have no agency because all novel thoughts impinge. Humans do not cause their thoughts and selection. My only response is that creativity is still a mystery, and we are not yet epistemically entitled to run this argument to the point of absurdity. Furthermore, cognitive science has informed us that even in these moments sub-conscious motor and kinetic programs or mechanisms are running. Some of these were delineated for improvisation above. This is one reason why a phenomenology needs to be appended to cognitive models and the like. One should also be interested in what is present to the consciousness of the agent, and what is consciously occurring while playing (if anything), not only the underlying processes posited by a cognitive theory, nor what could be going on as argued for in a philosophical theory. David Sudnow is perhaps the best example of a phenomenological approach to improvising. By introspecting on his improvised piano playing and his learning how to play 62 For example, see Sian Beilock, Choke: What the Secrets of the Brain Reveal about Getting It Right When You Have To (New York: Free Press, 2010). 63 Thomas Nagel, “Moral Luck” in Mortal Questions, Canto Classics Series (New York: Cambridge University Press, 1979), 24-38. 64 David Sudnow, Ways of the Hand: The Organization of Improvised Conduct (Cambridge, MA: Harvard University Press, 1978), and David Sudnow, Ways of the Hand: A Rewritten Account, foreword by Hubert L. Dreyfus (Cambridge, MA: MIT Press, 2001). 186 paino and improvise, Sudnow gives excellent descriptions of the process and actions. One of the most important insights he gives is that selection in jazz piano improvisation is in large part about fingering and the way one’s hands and fingers move across the keys. I can attest that the same is true for stringed instruments, like guitar. Often, when I improvise, my attention is on finger patterns that I know work (with embellishments) over certain “changes.” How strongly the phenomenology of playing an instrument comes to play in thinking about creativity and improvisation in particular comes to the fore in this extraordinary account of a conversation with the famous, brilliant pianist Bill Evans. ... I [Gene Lees] kidded him [Bill Evans] about his rocking a finger on a key on a long note at the end of a phrase. After all, the hammer has already left the string: one has no further physical contact with the sound. ‘Don’t you know the piano has no vibrato?’ I said. ‘Yes,’ Bill responded, ‘but trying for it affects what comes before it in the phrase.’ Evans reveals that there are motor selections that do not enter into the perceivable product (in this case sounds) but yet affect properties of that product. Not all selections will be perceivable in the final product (e.g., performance, recording). One should also be aware that selection may be coerced in both a literal and metaphorical sense. External factors such as authorities may constrain what artists do, thereby eliminating or reducing choices. I may only have the resources to learn one instrument. If I only know how to play saxophone, I am not going to pick up a trumpet. At any given time t, the agent (improviser, player, performer) has twelve pitches available in the range physically determined by the instrument. This range is vague because 65 Gene Lees, “The Poet Bill Evans,” in Reading Jazz: A Gathering of Autobiography, Reportage, and Criticism from 1919 to Now, ed. Robert, Gottlieb (New York: Pantheon, 1996), 424. 187 given certain techniques, which some musicians are able to do and others not, and physical enhancements to instruments, the range can be extended, both to the top and bottom of the frequency or pitch range. But it would wrong to suppose that this complete selection options set is fully available every time, in every context to an improviser. What decreases the possibilities of the selection options group are the constraints that are given and/or accepted by the player, the genre, context, et cetera. Now, the agent may at any point deviate from these constraints (intentionally or otherwise), but she may not deviate from the complete selection options group, unless she changes instrument or technique. The idealized selection options group is coextensive with the set of all physically realizable pitches and all possible durations. This set may be expressed in many ways. For instance, one could simply give the Hertz (Hz) cycles (frequency) of the pitch indexed to a timed duration, such as eight seconds or two seconds. Obviously, this is an infinite set, because the duration of a produced pitch could be infinitely long, and the sound waves, although severely limited by human audibility capacity (even non-human animal audibility) could be infinitely low or high, although there are frequencies which we cease to call sounds. Practically, in Western music theory, the accepted range of pitches is the human audition range (called audio or sonic), 66 I am assuming the agent is using the Western Equal Temperament (ET) tempered system. On the drawbacks of the exclusive use of the ET system that was more or less codified in the eighteenth century, see Ross W. Duffin, How Equal Temperament Ruined Harmony (and Why You Should Care) (New York: W. W. Norton, 2007). Scholars have identified at least 150 different temperament tuning systems in Western art music. Of course, ET does not apply to many non-Western music systems. The locus classicus is J. Murray Barbour, Tuning and Temperament: A Historical Survey, Dover Books on Music Series (n.d., n.p., 1951; reprint ed., New York: Dover, 2004). 188 which is approximately from 20 50 Hz (the lowest pipe organ sounds) to 20,000 30,000 Hz; while the accepted range of durations caps out at 128 notes. Selection is the process of choosing an output. The output may be physically realized or produced sound, or a notation for a realizable sound, or both. A single selection is actually an array of various factors as explicated in Pressing’s cognitive model. In using the term “choosing,” or “choice,” again I make no commitment to a theory of free will. This theory and taxonomy may remain neutral. If free will is false, then the selection process will be a product of some set of causal laws. Those causal laws will still have to operate within the taxonomy. Moreover, ideally a selection may be viewed as a choice of each discrete unit with relevant arrays, even though phenomenologically one may not be aware of all of the arrays. A musical phrase or lick may be played wherein the agent chose to play the lick as a whole. The entire phrase, then, which may consist of several pitches of different durations, dynamics, rhythms, and attacks, is the unit of selection—not each discrete pitch et cetera. Following are the selection options sets for musical sound generation (