The most central question raised by the Angel Project from my perspective is the degree to which data based upon listeners' experience with my work can be effectively generalized. This is not to suggest that my psychologist colleagues overstep their competence by claiming any undue scope for their findings. In my view they do not. I, in turn, am wary of making assertions that are outside my competence as an artist. But the creative process itself is so variegated, its facets so differently manifested from one person to the next, that neither the compositional process nor its assumed goals can be thought of as consistent. To imagine that immediate and wide-reaching generalization might be possible, therefore, would be to seriously misunderstand the realities of the world of contemporary, international concert music. There is an unprecedented profusion of stylistic and technical allegiances now, and the kind of extrapolations that were possible in 1800 are no longer.
If it is a fact that the desire to generalize regarding the details of procedure and stylistic consistency now is — as I believe — circumstantially doomed, this surely does not mean that the experiments reported here are without value. It has seemed urgent to me that there be a serious and broadly conceived examination of the relationship(s) between how composers make their music (including what they assume is heard by sympathetic listeners), and how their assumptions compare with the actual experience that listeners report. Though I took only a small part in suggesting what should be examined in this Project, I have been interested in and impressed by what was looked at and what was discovered.
It seems appropriate to me that an examination of the assumptions embedded in contemporary musical discourse begin by asking what composers believe to be their materials, the sonic elements which form the basis for their musical discourse. From an overview perspective, the second large question concerns whether the experiential envelope of the whole work produces in the listener the responses the composer imagines it will. Given the diversity of aesthetic perspectives referred to, the only "common ground" that I can imagine centers on these two factors: What does the composer believe she is working with and what is she trying to do with these materials? Thus, the fundamental point is not whether what has been observed in relation to my work is directly relevant to the work of other contemporary composers. It is rather that two essential initial steps have been taken: 1) the kinds of questions to be asked and ways of asking them have been formulated and exercised, and 2) an initial body of data has been assembled against which later, parallel studies can be compared.
I felt no little trepidation in embarking on the project described here: submitting to the objective testing of my assumptions, of beliefs that have arisen over decades of dedication and experience. During the 20th century, a number of the most powerful conventions developed through the course of Western music history were, in effect, set aside by composers. And these set-aside conventionalities were not superceded by comparably powerful alternatives. So it would not be surprising if the following question arose in the minds of many composers: What tools am I actually working with, and why? I have thought of myself as a "searcher" after the new, but not for the sake of rejecting tradition, or of embracing that which is novel because of its lack of connectedness to the past. From my perspective, composers write their music out of a need to interact with, and bend the world of sonic experience to immediate aesthetic purposes. But they are aware, increasingly during the last half century, of the context within which they live: enormous opportunity without an accepted or compelling base of conventionality. So they proceed, necessarily, in the presence of tenuous convictions and flickering doubt. I have grappled with this circumstance by attempting to forge a matrix of constraints within which I can be relatively confident that my elected freedoms will result in a principled (and therefore potentially graspable) outcome. It is crucial for me that the force of attractive ideas not lead me into a domain of behavior that is unrealistically remote from the realities of human perception and cognition. Thus I have become a lay reader about perception as it applies to music.
Much of my work has been in relatively large-scale forms, musical structures that consume 25 minutes or more. This has come about because I am interested in how one's impressions evolve over time, and whether the architecture of a work leaves one — even in small ways — changed. It seems to me that musical form requires considerable duration in which to make its optimum impact, particularly if its terms are unfamiliar and must be, at least to some degree, acquired during the course of the experience. And I do not refer to the experience of "noticing" a formal structure, but rather to that of living through the form. As immediate and pleasurable as the binary experience of a Scarlatti Sonata is, for example, I don't feel its kind of form or scale so much as I accept it as a framing device.
All of this suggests that I am motivated to understand what approaches will maximize a listener's grasp of and response to the shape of the musical experiences I create. I plan the formal proportions and consider their use at length before actually beginning to populate the opportunities I devise with specific content. If the ways in which I do things do not result in such grasping and responding on the listener's part, all is lost. In many, but not all, of my works, a certain variety of material is an essential component. But if one musical subject is like another, if one process is indistinguishable from another, if one ordering is as good as another, then the possibility of the music having optimal scope and impact is reduced. Thus, at the most basic level, the question I need to have answered is: Are my musical materials heard as I wish them to be? My UCSD colleague, G. J. Balzano, made a simple but powerful observation about the testing of thematic material identity. If there are 34 fragments that make up the subsections of my five themes and experimental subjects are asked into how many groups they would put them, the two most undesirable responses would be "one" (no differences are heard) and "thirty-four" (no connections are heard). But precisely where an ideal level of ambiguity is to be found is far from clear. This is not simply a matter of vagueness or indistinctness. Ambiguity implies that one registers the elements but remains uncertain about how they should be grouped, or about which of several possible relationships is being suggested. But both ambiguity (the thing is recognized but its contextualization is problematic) and vagueness (the thing is indistinct, the impression blurred) may be useful to the composer at times.
When I examine the results of the perceptual evaluations by experimental subjects that have been completed to date, I do not find the data difficult to understand. That is to say, the level of ambiguity implied in their responses (to which textural norm is an element to be related?) and the nature of the confusions or indecision (why is a particular subsection of one theme grouped with those of another rather than with its own?) regarding my thematic materials is easily understood. I can posit reasons for the reported profiles. While this was a ratifying result, it did not yet ensure that listeners would register the form of the whole in ways that would be equivalently satisfactory. Thus, I was interested to know the degree to which the subjects of the real-time, in-concert experiments would register the existence of formal divisions that I intended to be heard. A reassuring aspect of the curves plotted from these experiments is that the majority of their significant changes (of prominent breakpoints in listener response) coincide closely with moments in the performance at which intended formal divisions in the music occurred. So, in general, the data from the experiments to date suggest that my assumptions about how people hear are viable. Further, it does not seem that the level of listener sophistication is crucial to the standard of response. On the one hand, this is a welcome perspective, but, on the other, it suggests that studies on the effects of greater familiarity (with this particular piece, with my music or that of others with similar convictions) could be important to the larger significance of the Angel Project.
Several composer colleagues (particularly Philippe Manoury) were concerned about the possible implications of the process I underwent in the Angel Project. "What," he asked, "would you have done if the data showed that your assumptions were greatly at variance with the experimental results?" My response is that I will continue to look closely at the new data that arise as the psychologists probe the meanings in what they have examined. There appears to me, for example, to be a suggestion that, while listeners distinguish effectively between the material types that I have created, they are not as successful in carrying these identities with them under transformation. I speculate that the trend, viewed from a larger perspective, will be that both my materials and their transformations and combinations are more difficult for listeners to process effectively than I wish. Perhaps, then, further exploration of several questions — even in relation to the musical material already in hand — would be useful. One would be whether some ways of characterizing musical materials are more effective than others (assuming that something like a "thematic" basis is required for the composer's approach). That is to say: What are the most salient musical characteristics from both the perspectives of musically sophisticated and musically untrained listeners? A second, allied question would address the sense and the degree to which "initial" materials can undergo transformation without losing their identity. Both of these issues are broached within the current set of studies, but it may be important to extend them.
Should I consider simplifying the terms of my musical discourse, and, if so, how would such simplification be accomplished? What effect might such changes in behavior have upon the nature of my engagement with the process of composition? There is an inescapable conundrum here. The historical-social fact is that we live in a time in which the composer's premises, the issues that engage him or her, differ markedly from the experience and capacities of the majority of serious music-listeners. This is even true if one limits the audience to be considered to those with wide experience in contemporary music. (Compare, for example, the premises inferable from the music of Elliott Carter, Helmut Lachenmann, Toru Takemitsu, Iannis Xenakis, Morton Feldman, and Luigi Nono.) At the same time, the materials, strategies, and, therefore, prospects open to composers are of unprecedented richness. The essential ethic of the creative process will not allow one to act in violation of one's own aesthetic sensibility. (I presume that this is self-evident and does not require elaboration.) So one is poised between the (possible) lure of modifying one's ways so as to allow them to be better registered and more fully responded to by contemporary listeners, on the one hand, and the intellectual and emotional attractions of challenging one's own taste, experience, and capacities on the other. I would prefer to face these questions armed with more than informal, and in all probability deeply biased, positions to guide me. It is to the sort of studies reported in this document that I (and other musicians who share my concerns) can turn for useful perspective.
The persuasiveness of these studies will remain limited (in the senses described above) until similar examinations are made of the work of other composers. The rub here is not what can be done directly by listening to existing music. (There could be examinations, for example, of Schoenberg, Stravinsky, Varèse and Webern, composers of well-known but diverse bodies of music.) The difficulty (the failure of ideal parallelisms) would then be that experimenters would have to attribute intention to these composers, altering the degree to which the results could be reliably evaluated. The most desirable situation would be to enlist a varied sampling of accomplished composers, active now, and relatively clear in their own minds about what they are doing and how they do it. (This is no small matter.) These composers' work could then undergo the sort of exploration that occurred in the Angel Project. If this came about, the experimental strategies now initiated could be honed. Perhaps, then, a sufficient weight of evidence could be accumulated so that some influence on the course of musical practice and thought might be exerted. The Angel Project proposes the sorts of steps that might be made in order to improve the level of informed discussion as to the ways in which music can continue to evolve in an increasingly wealthy yet problematic landscape of potential.