Communicating with musicians interrogates the nature of controls and freedoms. Speaking from a composer’s perspective, we need to ask ourselves what exactly in the music do we wish to control, how will we control it, and how specific is the control?

In common practice Western classical music notation (which I will refer to from now on as tradition notation), we organize the information we wish to communicate on a 5-line musical staff. This system of organization is widespread and taught in music conservatories. But as with all communication systems, it privileges certain parameters of sound (e.g. synchronicity, discrete pitch) and disenfranchises others (e.g. timbre, randomness). This particular system came out of a musical practice needed to reference these parameters in order to communicate them on a large scale.

Would this system be appropriate for communicating with improvisors? Well, yes and no. The issue with being a composer (like myself) and communicating with improvisors is that I must determine what controls I want to exert and what controls I decide to cede to the musicians. While in traditional notation, the form of a phrase is dictated in the music. We know exactly when an event starts and ends. However, in freely improvised music, this is not so clear. Musicians often collectively feel the beginnings and endings of an event. Since this is felt, evaluated, and executed by the ensemble in real time, the realm of possibilities for phrase structure in an improvisation far exceeds those that are possible when using traditional notation. For this reason, we want to seek a balance for the amount of control imposed on the musicians, and the improviser’s freedom.

This brought us to the question of when the score is generated. Traditionally, the composer comes into the room with the score already prepared. The written music is fixed, and the musicians read their parts. What constitutes “correct” interpretation of the music is already determined.

In more recent work, a score can include very precise flexibilities. John Zorn’s piece, Cobra, includes a set of fixed cues notated on cards, along with rules corresponding to the cues that instruct the players what to do. Since there is no traditional notation (only the cards), each performance of the piece sounds radically different. The order in which these cards are presented is determined by the ensemble’s leader and how they perceive the performance in real time. In this way, the leader exerts controls over the form, and the musicians react to the leader, and the leader to reacts to the musicians, creating a live feedback loop.

In Ben Goldberg’s work Brainchild, the leader whispers instructions to individual musicians in the ensemble throughout the course of the performance. These instructions consist of rules that are precise in their language, but may be interpreted in wildly different ways. “Play what musician X is playing, but do it backwards”, or “Play exactly what they are playing, but add three.” Since the leader must physically approach the musician and whisper to them, this paradigm limits the amount of information conveyed and the speed of the message. It is impossible for several musicians to receive simultaneous cues unless the leader sets up the scenario ahead of time.

Thus, the problem here is how to generate a complex, dynamic, and responsive score in real time that can be communicated simultaneously to everyone, that at the same time empowers the improvisors to play freely. Our solution (for the time being) is to control five parameters in real time: stability, density, strength of transients, speed of activity, and spectral variety (these words will be thoroughly defined in a later journal entry). We will scale the highness and lowness of these parameters numerically, and the leader determine these values in real-time using electronic sliders. The five values (reinterpreted as a graphic) will be projected on a monitor which the musicians will read, interpret, and play. As the piece continues, the leader will respond to how the improvisors interpret them by issuing new values, thus creating a real-time feedback loop to guide the form.

We choose this approach because it communicates the perceptual character of the sound while still affording a large degree of freedom to the improvisors. The musicians themselves are still in control of the moment-to-moment details of the music, while the leader exerts vague control over the large-scale character of the sound. This system is descriptive rather than prescriptive, giving the improvisers complete physical agency over their instruments.

One caveat to this system is that the entire ensemble will receive the same data. Of course, one solution would be to communicate a separate set of values to each musician, but we feel that this approach would exert too much control over the improvisors and possibly stifle interaction within the ensemble itself. While we will define our parameters to the ensemble, we will not specifiy how the parameters are to be interpreted. We are confident that each musician’s individual and interactive interpretation of the data will be sufficient in inspiring and producing interesting musical results.