My original thinking in regard to the ordering of the S and D versions of Angel's material assumed that there would be advantages for the listener in hearing the thematic materials clearly at the outset, in knowing what was at work as the composition unfolded. Also, because of the role reversals dictated by the change from S to D (piano?ensemble, ensemble?piano), I expected the latter to be more transparent, that it would work better with the computer overlay when it appeared second in performance. Together, these two considerations led me to expect that S-D would be the preferable ordering, the most engaging from an overall perspective. The significance of these assumptions, while they remained objectively true, faded under concert experience. The oceanic darkness of D's opening is simply more gripping than the more didactic shape of Theme 1 in S. Further, the long and engaging evolution of D's TR2 to 4 in the solo piano is more persuasive than its ensemble parallel in S. So, unexpectedly, the logical and strategic merits of S-D were swamped, in the event, by the more persistent emotional tugging of D-S. This is true even though, in the latter case, the identity of the basic materials comes after much of their exploration, and, further, their clear and complete exposition is interfered with by the co-existence of the computer images.
Much has already been said about the unanticipated growth in duration of the Other element. I had thought of it as a "sport", a nonconforming element that would act as a counterweight to the orderliness of the Sectional laying out of thematic content. Because of the performance difficulty of this material, however, its proportional weight in the structure of the two parts expanded. In addition, I had expected to compose an ensemble version of this material as I had for the themes. Had this happened, the recurrence of the structure of ostinatic overlays would have been less perceptually marked. But as I considered the content and processing of computer images, I felt that it was not feasible to disturb the ambiguous tranquility of this element, so only the most subtle of transformations was possible. I decided that, whichever part was performed first, the soloist would play Other live. So the second presentation was inevitably carried by the computer. As a hedge against the repetition being too "familiar", particularly given its almost three-minute length, I decided to add 11 brief extracts from the piano or ensemble recordings of the themes. This lessened the degree to which the repetition is felt as literal, and also added variegation to the impression left by its appearances. All in all, the weight of Other became greater than had been intended. Its position in the formal experience is made more unique (more nonconforming) by the fact that its materials appear nowhere outside its boundaries. It casts few shadows. It is in some degree compensatory that Other is more tranquil and expansive than intended and, thus, provides a refuge for the listener from the ongoing and interwoven nature of the music before and after it. Its nondevelopmental, lulling character encourages the possibility of "stock-taking", of stolen mid-course evaluations of the music's progress.
Theme 5 ends both of the parts in a similarly elegiac mood. This allows a natural turning back towards a re-initiation of the opening materials as the complementary part begins. However, it is not a satisfying way to conclude such a large work with an iterative essence. The lyric impulse is the intended, and, I think, the appropriate outcome for the succession of thematic characters that each part traverses. Still, it is gentle rather than soaring, an arrival of modest assertiveness. While working on framing the nature of the set of computer images, I realized that I could use the final S7 to bring a more satisfying close to the juxtaposition of the work's two large and parallel parts. As it belonged to the Sectional type, its materials were to be explicit, independent, and delineated. I made it a summation of what had come before. But, there was no powerful, low-register progression in the piece to this point (as a result of the gradual achievement of a linear character in Themes 1-5, and the fact that, when line finally does become dominant, it is in the middle register). It was therefore necessary to devise one by connecting in a fresh way various earlier events that had the appropriate sonic character. Many-layered spatialization was also used to give a unique gestural animation to this final image. In performance, its subtle emergence from the last section of Theme 5, marked by the insistent iterations of the SPIRLZ algorithm, allows the final cadence of the instrumental music to conclude effectively at the same moment as one's attention is shifting towards the computer's accumulating force. I feel that the weight of this final computer section is well-gauged in relationship to its purpose. It must achieve the sense of "cresting" while not either overbalancing (and thereby trivializing) the role of the preceding instrumental music, nor being so conclusive that no space remains for the gentle, ruminative world of the final piano Epilog. The Epilog embodies the chorale nature of Theme 5's core. But rather than using the anacrusis feature to lend desired weighting to linear elements within successive chords, here the emphasis that allows crucial dyadic voice leading is achieved more subtly, by dynamic balance within verticalities.