On the Fly: Teaching Open Improvisation Through Play

Ethan Rubin, Independent Researcher

Theoretical Background

The purpose of On the Fly is to provide musicians with an enjoyable, engaging, and intentional process for practicing open improvisation. By “open improvisation,” we mean music that is invented by the player in the moment, without predetermined melodies or harmonies. On the Fly is particularly concerned with fostering cooperative improvisation, in which a group of players simultaneously improvise an original piece of music, and is best played in a group of three or four musicians. Improvisation is a notoriously difficult skill set to teach, for several reasons. First, it presupposes a baseline level of technical ability on an instrument, to the point of relatively unconscious execution of musical tasks; the more attention one must invest to produce a clear note or find the appropriate notes for a chord, the more difficult open improvisation will be. Second, it requires an ability to create new melodies and harmonies spontaneously. This often leads to a sort of paralysis among beginning improvisers, who feel they simply “don’t know what to play”, which must be addressed through a combination of music consumption, imagination, and a willingness to experiment and “put oneself out there.” Third, a cooperative improviser must be able to recognize and respond to the cues and inclinations of other members of the group.

In addition to these substantial challenges, which demand focus, creativity, technical skill, and a host of other capacities, improvisers tend to fall into a few glaring pitfalls, which can reduce the listener’s enjoyment of an improvised performance and tend to stymie the development of the musicians themselves. These pitfalls fall roughly into two categories:

Repetition: Most musicians have particular melodic lines they play often, devices that suit their instrument, and so on. The cognitive load of improvisation inclines players to their most familiar or comfortable motifs because they lack the bandwidth for more inventive lines. As a result, improvisers can fall into the trap of playing the same fragments or patterns because they come to mind most readily, especially if there is no pressure to go beyond them.

Aimlessness: On the other hand, some repetition is necessary for the listener to have something to grab on to, to gain “memory traction”. Similarly, most compelling music has some sort of overarching structure. Combining strategic repetition with intentional structure gives the listener the experience of a cohesive, dynamic whole. It is often difficult, however, for the player to remember improvised melody lines for the sake of repeating them, and it is similarly difficult to take a broader perspective of the entire improvisation’s contours to give it a deliberate structure.

The difficulty of improvisation can be mitigated by using predetermined structures, which is why improvisation is usually taught and performed within structured forms. Blues is a popular starting point for teaching improvisation, both because the AABA structure is straightforward and blues permits an unusual degree of melodic repetition. This is not to say that the blues lacks nuance or complexity, but rather that its structural features provide a manageable starting point for musicians new to improvisation. In jazz, more complex harmonic progressions, in the form of chord changes, provide a varied and challenging landscape to navigate, which lends an inherent level of complexity and diversity to the improvisations that result and draws out more inventive playing as musicians attempt to link up melodic lines that fit the chord structures underneath. Many folk music styles take a different approach, in which improvisations begin with increasingly ornament iterations of a melody, often through culturally typical or traditional ornaments; celtic music and American old time music tend to fall in this category. Several Middle Eastern musical traditions, most notably the Persian dastgah, build on yet another type of structure. In vastly oversimplified terms, dastgah consists of series of melodic motifs that are traditionally developed in a certain order and style. The motifs, known as gushehs, are infinitely manipulable, as they can be inverted or altered, combined, ornamented, played in various registers and at different speeds, and so on. This gives the player improvisational freedom while still providing enough structure to lighten some of the cognitive load of simultaneously improvising and making structural and aesthetic plans for the piece.

On the Fly is intended to work on a similar principle. By guiding a group of players to improvise according to a contour and focus on a particular variable of their playing, the game’s structure encourages more fully developed and enjoyable performances. At its best, it prevents boring repetition by providing “varied terrain”; having to navigate changes encourages one to play more inventively. At the same time, it provides a structure with direction and intention, which allows the musicians to focus on their playing and thereby helps prevent aimlessness.

This project addresses a nebulous skill set, which is difficult to define and therefore difficult to teach. To meet that challenge, I have borrowed concepts and techniques from a variety of fields. In addition to music education and various improvisation-heavy genres and styles, the field of experiential education has provided substantial insights, due to its mechanism of presenting small groups with dynamic challenges to practice interpersonal skills under strictures. The individual music teachers I have had, as well as influential musicians I’ve played with or listened to, are also resources in this process. Much of the foundation of this game, including the need for it, is anecdotal, but the concept behind it seems to resonate with many musicians and has been reinforced throughout my life as a musician.

Game Design

The basic mechanism of the game combines a contour, presented in the format of a line graph, with a symbol representing one of a set of variables. The “contour” approach is a useful way to think about structure for open improvisation for several reasons. Considering the shape and direction of an improvisation has functional advantages that make it reproducible for players in future improvisations. The songs or pieces to which musicians are typically exposed in the Western tradition have an overarching structure of alternating verses and choruses, perhaps with a bridge. In any case, these sorts of structures are difficult to create spontaneously, because they require repeating sections and establishing clear divisions between one section and the next. This is a significant ask for individual memory, and a still larger one for a simultaneous collective effort. A style like Dastgah is more in line with On the Fly’s concept, but still assumes immersion in a complex cultural tradition; a player must already know the sequence of melodic motifs in order to improvise. With a contour, musicians can practice creating an overarching structure that is more fluid and thereby lends itself to greater spontaneity. Likewise, the fluid and spontaneous aspects of a contour structure facilitate gentler “leading” of fellow improvisers and ease some of the jarring effects of missed cues or delayed reactions.

By pairing the contour with different variables of musicianship, On the Fly hopes to encourage musicians to think about and experiment with the tools at their disposal. Rather than matching the contour to a generalized sense of size or intensity, isolating variables demonstrates the many ways in which musicians can create engaging and useful effects and mechanisms in their improvisations. The idea is that practicing these things – creating a larger contour for a piece, deliberately using tools to create structure and contrast – will enable musicians who train using this game to apply those skills to their improvisations outside the context of On the Fly. Each of these variables has its own character, challenges, and opportunities:

Dynamics: Control of volume may seem to be an obvious tool, but it is also easily forgotten when a musician is distracted by other challenges. Even when playing structured music, inexperienced musicians often forget to manipulate dynamics in the face of technical difficulty. This may be the easiest or most familiar of the variables, but it is still worthy of intentional practice.

Tempo: Alterations in tempo are challenging in a group because musicians usually intend to be playing together. Gradual tempo changes are often considered flaws, “rushing” being a perennial problem among students. Tempo changes like double or half time are difficult to execute, particularly if unplanned between a group of players. Within open improvisation, tempo is a more value-neutral concept, and there is no inherent reason that it cannot be approached as a fluid attribute available for manipulation. Changes in tempo can be dramatic and noticeable, even to the untrained listener, but their capacity for nuanced mood creation when applied judiciously is often underrated.

Consonance: Consonance is perhaps better recognized by its antonym, dissonance. Deliberate dissonance is challenging because it is so counterintuitive: why play things that don’t sound good? In the development of Western classical music, it took centuries to develop an acceptance of and taste for dissonance as a musical device; medieval music avoided dissonance whenever possible, and in some cases forbade dissonance as a heresy. Dissonance is now recognized as a great source of tension and contrast, and a device worth developing. Moreover, modern tonal music is as reliant on dissonance as on consonance, and purely consonant music creates a surprisingly jarring effect.

Register: Some region of any instrument is bound to feel most comfortable for the player. Parts of the instrument may resonate more readily, and there are positions that are easiest or around which players orient themselves or organize the lines they like to play. Stretching register variation creates textural changes, both in the feel of the piece and the tone of the instrument itself; even beyond high and low notes, players benefit from learning to play with growly sounds and squealing and whistling and the like. In contemporary classical music, these sounds fall under the umbrella of “extended techniques”, meaning that they go beyond the methods of sound production that are typically considered “proper” on the instrument. In the context of open improvisation, no sound is off limits if it is situated in an effective context, and so the idea of extending technique is particularly apt here.

Density: Another common pitfall of improvisation is just playing all the time. Music educators in improvisational styles often counsel students not to play every idea they have, or to “get out of the way” of other players from time to time. On the other end, a tentative improviser might suffer a degree of the aforementioned “paralysis” and feel that they can’t find anything to play. One or more players may also feel that there is no room for them in the improvisation amid the cacophony of others in the group and choose not to play much as a result. Experimenting with density can help musicians better appreciate its effects, whether positive, negative, or neutral. Even outside of the context of this game, this variable can be a useful instantiation of the “step up/step back” principle referenced in experiential education: a player who tends to play all the time might make a conscious effort to step back, while one who avoids the spotlight can make a point of taking center stage more often.

In addition to the variable that pairs with the contour, several other variables have to be determined for gameplay. The most essential are mode and meter, which are decided by rolling two other dice. These aspects present an opportunity to give the player more control over the game’s difficulty; the basic versions emphasize common and comfortable meters and modes (from a Western standpoint), but players can choose to use “advanced” dice that include odd time signatures and more challenging modes. On the Fly also includes Wild Card challenges and narratives that add variety to the gameplay, provide opportunities to make the game more amusing, and promote more different kinds of engagement with the improvisation process. One set takes the form of “mini-games”, in which players are challenged to steal each other melodies or force time signature changes, for instance, while the other set provides narratives or scenarios that players must accompany in the manner of a film score. The Wild Card sets are presented as “booster packs”, which can be shuffled into the contour deck at the players’ discretion.

All of the materials for On the Fly are contained in PDF files, which allows it to be printed and assembled by anyone with basic access to computer technology. This method of production and dissemination increases access by allowing anyone to make their own set with which to play. It also decreases cost, both for the game designer, by rendering a manufacturer unnecessary, and for the players, who do not need to cover the cost of manufacture and can therefore have the game for the price of some cardstock paper and scotch tape.

Challenges and Alterations

In the process of planning, designing, and testing the game, several challenges arose that altered some mechanism or principle of the game’s materials and processes. Most of the obstacles conflicted with a broad goal of “minimal intrusion” in gameplay. Because the difficulty of improvisation arises in part from the many competing demands on the players’ attention, the structure and experimentation imposed by the game should feel as intuitive and natural as possible while continuing to stretch the players’ abilities; in short, the game should not be “in the way” of the improvisation. A few of the “intrusion” issues encountered during playtests are as follows:

In the process of design, the question of controlling for multiple variables created a decision point. What is the least intrusive way to select the multiple variables that need to be determined? Part of the goal of this project is to relieve some demands on players’ bandwidth, and a confusing system to determine variables would work against that goal. The two initial mechanisms for variable selection were the symbol die and the tray with keys. That said, several other variables needed to be determined, most notably mode and time signature. Should those variables be selected by dice, by tray, some combination thereof, or some other method? In the case of combining dice with trays, which dice should be matched with which trays? In short, these questions were preliminarily resolved through conversation with a graphic designer, fellow musicians, and playtesters. The solution arrived upon was to roll the die for variables in the box, because that die is the only one that is linked directly to the contour of the line graph and as such needs to be set aside as special. The mode and time signature are determined separately by dice to avoid any confusion or inputed relationship between those two variables. This arrangement seemed to satisfy playtesters as the most intuitive and least distracting way to determine all the variables needed.

The first hand-written prototype of this game used a die with six letters, which represented the initials of the variables (V for volume, D for density, and so on.) Players found the process of remembering which letter corresponds to each variable, or the task of referring to the instructions to check, cumbersome and therefore distracting. With the help of a graphic designer, the new die uses symbols that represent the variables, in the hope that they will be more intuitive and minimize the hassle of translating the die result into an actionable parameter for improvisation.

The original game used the variable “dissonance” rather than “consonance.” Although dissonance is a more familiar word, playtesters agreed to use “consonance” instead so that the peaks of the contours become the moments that are most harmonious, which they found to be a more intuitive orientation than the reverse.

In addition to these more conceptual challenges, alpha playtesting revealed some logistical questions regarding game format and gameplay. A few of those obstacles included:

When using the tray to randomize keys, the question of “true” randomization came up among several playtesters. Are the outer edges more or less likely results than the center squares? In either case, should more challenging or unusual keys be the more likely or less likely outcome? This question is still unresolved and continues to be a deliberate focus of playtesting.

Some contours are more challenging than others, and a few playtesters suggested that they should be identified as such. One solution would be to divide the deck into different levels, marked on the cards, to allow players to separate out tough contours and make the game easier. Some playtesters, however, argued in favor of the currently variable difficulty levels, and polling so far has been inconclusive.

Several playtesters asked an unexpected question: what do we do about an improvisation after it’s done? There was some consensus around wanting a metric or framework for saying whether or not an improvisation went well, or critiquing performance for the sake of improving future efforts. In response, On the Fly’s instructions now include some suggestions for “debriefing” an improvisation. Post-challenge debriefs are a common feature of experiential education curricula, which suggested a few potential approaches.