Page  202 ï~~Formal Processes of Timbre Composition Challenging the Dualistic Paradigm of Computer Music A study in Composition Theory (II) Agostino Di Scipio Laboratorio Musica & Sonologia, L'Aquila (Italy) lms@vxscaq.aquila. infrh. it Abstract The utilisation of computers has fostered a variety of approaches to music composition. It is possible to gather them in two general categories, which reflect separate cognitive features in the compositional process, i.e. algorithmic composition and timbral design. These are different attitudes peculiar to a profound cognitive dualism. The most innovative and aesthetically refreshing potential of electroacoustic and computer music composition lies in the possibility of blurring that distinction, thereby questioning the dualistic paradigm of music composition. 1. The Cognitive Dualism of Electroacoustic and Computer Music Skimming through the pages of computer music publications, either research paper collections like the ICMC Proceedings or text books [Dodge & Jerse, 1985], it seems evident that the utilisation of computers in the compositional process has fostered a number of different approaches and attitudes. It's not oversimplifying to gather all approaches into two main categories: algorithmic composition and timbral design (or timbre composition). Following a former study [Di Scipio, 1993aj, this paper explores such distinction, as well as the possibilities of computer music by which that distinction can be significantly questioned. In very general terms, algorithmic composition involves a strong emphasis on the formal properties of the musical architecture, and on the generative rules of music at different levels of structure. In timbre composition, the emphasis is especially on the perceptual and semantical properties of the sound materials and their associated auditory images. However, far from being specific to eletroacoustic and computer music, these two attitudes reflect cognitive dispositions which have run parallel with the development of musical thought along this century. Computer music amplifies the potential of the many possible approaches in between them, and, even more importantly, makes us particularly aware of this dualistic set of our musical mind. The distinction is also one that reflects two instances of, respectively, formalism - a radical stance of rationalism - and that which has been sometimes called (at least in Europe) materism - meaning a special care taken by artists (including composers of electroacoustic and instrumental, especially orchestral music) for the instrinsic aestehtic values of their material, in a radical stance of naturalism. 1.1 Sound Form Since its inception, electroacoustic music has been the preferential terrain of an unprecedented approach to music composition, one in which the process of composing-the-sound is an activity of invention and design distinct and, at least in principle, equally relevant as the ability of framing the sound materials in an overall structure - i.e. distinct from composing-with-sounds. In a thoroughly-going music of timbres, the new creative environment opened up by electronic and computer technology is one that forces us to consider sound as something composed, something stemming from personal designs, the very result of an act of intent. Sound becomes a result of composition, a artifact rather then a pre-existent object. This is central in what we call timbre composition. So much so that theoretically we should consider any single sound of a musical work capable per se of triggering subjective processes of apprehension and appreciation, before (or even: rather than - it depends on how radical is the stance taken) being understood as a partial component of an overriding structure. Sound becomes the very object of an appreciator's aesthetic judgement. Philosophers would tell us that whatever we exert acts of appreciation upon cannot but be named form. Thus, taken per se, any sound which is the intended artifact of a creative, individual action must be conceived as a kind of (musical) form. At least, we must say sound aspires to such category, it aspires to being self-consistent form. 1.2 Sound Material This condition notwithstanding, the tradition of electroacoustic and computer music also shows Composition, Composition Systems 202 ICMC Proceedings 1994

Page  203 ï~~that, once composed, a sound is then put in some relation with other sounds. A radical view of timbre composition would be one in which any single sound event is conceived as an individual musical object deliberatley designed and constructed. But in any other (less radical) view, sound becomes a musical event if it is contextualised, if it stands in some relationship with other sounds. And if such relationships (syntax), psychologists would say, have some degree of coherence and cognitive reality in the subject's mental representation. Thus composing-the-sound appears as a preliminary step followed by composing-withsounds. This latter takes sounds which were the object of invention, design and appreciation (forms) as partial components of a whole gestalt, whose consistency is determined by coherent syntactical links among those constituents. Hence, sound returns to the condition of material. Indeed, the act of composing-with-sounds essentially negates the sound's own internal tension towards the status of a composed, formed totality. Within the logics of the final artifact, the sound material disappears and vanishes. The formative act which gives substance to perceivable relations among sounds, to a "musical discourse", is overwhelming with respect to the innermost tension of the individual sound object (the tension which makes it form). The logics of the musical discourse obliges sound to the status of material. Sound material must be neutral enough [Boulez, 1987] not to prompt eccentric and centrifugal forces of its own, thereby disturbing the clear recognition of the very logics of relation that the composer imposes to sounds. Thus composing-with-sounds - which is primarily a matter of syntax, of symbolic manipulation - makes "material" that which, by composing-the-sound, aspires to become "form". 1.3 The Dyalectics of Sound and Structure in Computer Music For the traditional dyalectics of art, the material stands as a representant of Nature in the artist's process of invention. In the logics of the artifact thus created, the material is neutralised by the formative process. Its specific strenght and features are dominated, or artisticaly exploited. Finally the trace of Culture is imprinted in the material. Within the structure of the artifact, the material is "not-existent" [Lukdcs, 1971], its internal forces - the forces of a natural phenomenon - are mediated by the artist's techniques and expressive means, hence negated in themselves. The characteristic potential added to this dyalectics by technologically bansed compositional processes stems from the fact that now sound materials, too, can be intended as cultural artifacts, not only their set up in the overall configuration of the musical form. Though standing as Nature in the process of art, we should not forget that the notion itself of material is a historical notion. Materials have always been the outcome of human knowledge, hence have always been cultural in essence. So much so that - if one accepts the philosophical definition of "material" as "objectified consciousness" [Adorno, 1970] - we can argue that By struggling with material, composers struggle with society" [Zuidervaart, 1991, p.97]. Before the inception of electroacoustic music, however, materials had never been an integral part of the compositional process, the outcome of a single individual's knowledge and invention. Today, sound can be form, rather than mere material, exactly because its very existence and morphology are caught within the individual compositional project. Adorno's "objectified consciousness" is now "locally distributed", inherently fragmented. In this sense, technologically based composition makes to sound material what the technique of the early avantguardes of this century had made to the tonal system and the conventional rules of composition: the breaking of a common language and the liberation of individual rules and approaches. However, whatever the syntactical principles which give substance to the act of composing-withsounds, these push the sound back to its older status of something given. By composing-withsound we try to re-integrate - but faking - the material's older role of the representant of Nature in the process of art. What remains a distinct feature of technologically based composition, then, is the possibility of tighter conceptual links between the structure of sound materials (microstructure) and the structure of music (macrostructure). This possibility, in fact, has been the proper region (in part a terra nova) explored by most electroacoustic and computer in the last 4 decades. Ultimately, this dyalectics internal to the music reflects a profound cognitive separation, in which sound and structure are experienced as different, if not contrasting aspects. Most of today computer music conforms to such profoundly dualistic action frame, which is in fact overtly manifested in the clear-cut separation that musical system design makes between "instrument" - an operating "model of sound material" - and "score" - an operating "model of musical structure" (see below). n practice, however, the distinction between algorithmic composition (where sounds are symbols manipulated by explicit rules, thereby remaining material outside the focus of the ICMC Proceedings 1994 203 Composition, Composition Systems

Page  204 ï~~compositional task) and timbre composition (where sounds aspire to be self-consistent objects, individual forms) is a matter of degree. Or even a matter of where we enter the compositional process, as Pierre Schaeffer noticed in the mid 60ies:...poser une separation aussi nette sarait oublier l'implication essentielle qui articule les structures du simple au compose, et qui ne fait pas apparaitre forcement le simple au point de depart: on entre dans de telles relations a n'importe quel niveau, et on accede alors aussi bien aux ttages suptrieurs qu'aux inferieurs [Schaeffer, 1966, p.35-36] In computer music the dualism "sound/structure" and the implicit (pseudo)-ontological opposition "form/material" - plus their relative aesthetic attitudes, "formalism" and "materism" - are captured by the different level at which computation is utilised. In the disposition of algorithmic composition, computation takes place primarily at a level as high as the level of note-tonote relations (note-time and event-time). Then, computed data and relationships among data are given concrete shape with independent, preexisting sonic material to become audible. (This may be done via instrumental scoring, the control of MIDI tone generators or the call to a software "orchestra"). computational Differently, in the attitude of timbre composition computation concerns the lowest scales of time (control- and audio-time); it yields audible results which are then submitted to autonomous higherlevel organisational strategies. computational In the former case the emphasis is on the composer's model of musical design, defined as the composer's instantiation of a procedural description for the musical structure. In the latter the emphasis is on the composer's model of sonic material, his/her procedural description aimed at the generation of sound materials, possibly in terms of timbral morphology. Because of this fundamental cognitive dualism and its pendant in the dyalectics proper to the process of composing, computer music research has always followed two distinct directions, often largely independent one from the other. At one level - analysis - scientific research has been addressed to the representation of musical processes and the representation of sound signals. At an other level - synthesis - those scientific efforts have given rise to automated tools for computer-aided composition and to sound synthesis and processing techniques. So we are back to the initial observation concerning the separation between the cognitive competences of algorithmic composition and those of timbre composition, a separation that in fact computer music publications usually mirror faithfully. 2. Challenging the Dualistic Paradigm The use of computer, however, can allow us to break the barriers of our musical imagination within which we, as composers and theorists, still work. One of the most relevant challenges of today computer music implies a profound re-working of the modus operandi just described, and inevitably results in a different perspective concerning the relation of material to form, of sound to structure. In this second section, [would like to describe some examples of this alternative perspective. For a discussion bearing on its parallel transformations in the very notion of timbre, musical time and musical form outside the dualistic paradigm, the reader is referred to [Di Scipio, 1993a; 1993b; Dufourt 1991, part H, chap.III] and [Di Scipio, 1991; 1994 (only for Italian readers)]. Important related issues are also treated in [Truax, 19921. 2.1 Models of microstructural composition Indeed, since the inception of electroacoustic music, we have faced clear signals of new paradigms of music composition. This can be seen, for example, in Xenakis' work with analog and, later, digital means. In some of his musique concrite production (Concret PH, 1958) as well as in successive work with analog (Analogique B, 1960) and digital synthesis (La Ltdgend d'Eer, 1977), Xenakis' approach is one in which relevant features of the compositional project are concerned with the musical structure at the level of the minimal scales of temporal granularity. At the time of Concret PH, similar paths were being explored, with different means and aesthetic aims, by Henri Pousseur (Scambi, 1958) and, partially, by Composition, Composition Systems 204 ICMC Proceedings 1994

Page  205 ï~~Stockhausen (Kontakte, 1960). Pousseur himself devoted an entire chapter of his Fragments Theoriques sur la Musique Experimental to microcompositional strategies [Pousseur, 1970; p. 138-157]. Later followed the work of Herbert Brun, with his early Infraudibles (1969) and the SAWDUST project. And G.M.Koenig's Ssp (Sound Synthesis Program). More recently, a similar vein is found in Paul Berg's PiLE program [Berg, 1985], in several approaches of granular synthesis [Truax, 1988; Roads, 1991; Di Scipio, 1990; Di Scipio & Tisato, 1993] and the closely related use of FOF synthesis [Rodet et al., 1984]; and, finally, in renewed approaches of non-standard synthesis [Manzolli, 1993; Chandra, 1993], including Xenakis' dynamic stochastic synthesis [Xenakis, 19911 and methods of functional iteration synthesis in course of development by this writer and colleagues [Di Scipio & Prignano, 1994]. What does bind together all such different experiences in composition and sound synthesis? The common point is the fact that the composer's model of sonic material behaves like a model of micro-time sonic design [Di Scipio, 1993b]. To say it with different words, models of sound material and of musical design become inseparable: the compositional process is applied at the scale level of the microstructure of sound and yelds the structured organisation of myriads of minimal units (each of which, taken per se could not be perceptually significant). 2.2 Algorithmic approach and microcomposition The task of microcompositional strategies can be described as one of letting global morphological properties of musical structure emerge from the local conditions in the sonic matter. One must use a microstructural representation of sound signals, either in the discretized 3D space At4fAg of granular representations, or in the discretized 2D space AtAg (the simple time/sample grid) of nonstandard synthesis. The minimal discrete units of such representations are seen as symbols submitted to manipulation by formal processes of the composer's own invention. These embody an arbitrarily devised process and reflect an abstract model corresponding to no generalisable (acoustic) theory. In the computer implementation of such processes, the control-structure features primarily compositional - rather than physical or psychophysical - parameters. The fundamental point in such approaches lies in the fact that algorithms implementing a model of sound material actually implement ideas concerned with musical form, too. This equals to rule-based approaches where, at least in principle, form is operationalised by audio-time processes [Laske, 19891. Models of standard synthesis are instances of known theories of sound; one utilizes them in his/her own model of musical design. In contrast, non-standard models instantiate a possible theory of sound; one explores them, and learns how they can mediate the sonic structure. So, while for the dualistic paradigm of computer music composition, the coherence of the sonic structure stems from the coherence of the underlying acoustic representation, for the holistic paradigm such a coherence can only be dependent on the concretization of the strategy captured in specially designed algorithmic processes. It becomes the coherence of a musical structure, wrenching sound out of the realm of "natural" phenomena where it still stood according to the former conception. 2.3 Form as the formation of timbre Indeed, the opposition of composing-the-sound and composing-with-sound, we have seen, resolves a separation in competences and actions bearing on sound materials from competences and actions bearing on the musical construction; blurring the separation between models of materials and of musical design also means questioning this apparently inavoidable opposition. In fact, each approach of algorithmic micro-structural design proposes a Theory of Composition as a Theory of Sonological Emergence: how to determine a ground-level system's or process' quantitative organisation capable of bringing forth a metalevel system or process of peculiar qualitative, morphological properties. The resulting objects of such perspective of compositional design cannot be named sound material. On one hand, they may appear as extended, sometimes internally articulated sonic gestures - rather than single, discrete events (the discrete symbols of a "musical discourse"). On the other, we have seen they are the direct result of a creative process extended to a level that we usually conceived of as a "composer-independent" task of sound synthesis. These observations lend themselves to view the sounding results of composition as musical form, but in the special sense in which timbre - the qualitative emerging properties of the sonic structure - can be conceived as form. Thus, in this medium form can be described as a process of timbre formation through time. Rule-based (algorithmic) timbre formation is the proper realm of invention of this sonic art, where timbre - the phenomenal appearance of sound composed - comes to be an attribute of the musical structure rather than of the material components of that, the phenomenal attribute of the musical artifact composed in a holistic approach to sonic design. ICMC Proceedings 1994 205 Composition, Composition Systems

Page  206 ï~~3. Composing as the exploration of possible formal Theories of Sonological Emergence One may not need to address his/her algorithmic models towards the composition of timbre in explicit terms; indeed one may not be interested at all in thinking of his/her design task as one concerning timbre. Admittedly, many approaches of non-standard synthesis, up to date, can be situated in the more general context of algorithmic composition, with aesthetic results relatively rich in timbral properties (sometimes really dull, indeed). But this has nothing to do with the dynamics of the creative process of microstructural composition: whether a composer address conceived of his work as timbre composition or not, each local configuration of sonic units that he/she composes has its own morphological reality and brings forth emerging properties (those we say timbre) which reveals a a higher level of existence and can be observed in terms of phase states of a dynamical process. Thus, a composer's model of micro-time sonic design is a possible instance of a Theory of Sonological Emergence (TSE), that by which he/she imagines and explores possible links between the patterning of atomic details - a ground-level process (glp) - and the sound forms which emerge from them - a meta-level process (mlp). 3.1 Auhebung The passage of a system or process from a given structural organisation to a new state of order which is recognised as a function of the qualitative properties of the former, is what we call here a phenomenon of emergence - aufhebung [Wilden, 1972]. Similar phenomena can be described with rules of morphostasis (conservation of coherence, identity) and morphogenesis (dynamical behavior, change) which together capture the main peculiarity of social and living systems: selforganisation. We have seen that in the holistic paradigm of music composition, where (the materials') timbre and (the music's) form are experienced as one and the same qualitative, emerging conditions in the sonic structure; it may be argued that this process be describable by morphogenetic models, microtheories of music cognition (each concerned with one work or a set of works by the same or diverse composers). The subject of cognition (a composer, a listener) would be an integral part of this model, even though in actuality he/she always tries to hide his/her active, inventive role and simulate a (virtual) seif-organisation of the musical structure. Here we are only concerned with the properties which are to be expected in a mlp arising from a lower glp. Among them we'll recognise: " new local configurations of order/disorder " significative structural innovations (semantics) " an increased adaptive range (learning) " the possibility of changing goal-state " a transformation in the code of communication " a positive sensitivity to noise For an external observer of the artifact, the glp may be simply inexistent, since it remains in the background of its synchronous mlp. It is the composer's task to let him/her perceive the epiphenomena in the mlp both as structural qualities brought forth by the glp behind it (i.e. to reveal the his/her techniques of art) and, simultaneously, as a comment to that (metacommunication). 4 g lp." 44 s~uckx- exi hghewshuctL rdorder) 3.2 Constraints and mathematical models A most relevant set of extra-compositional constraints which an instance of TSE can be expected to take into account is the one prompted by perception and cognition, particularly as far as issues of auditory scene analysis are concerned [Bregman, 1990]. It's not by chance that dynamical auditory percepts seem to be better characterized in terms of "granular" representations and of "textural constructs" - maybe borrowing the notion of texton - the atomic element of texture - from studies in visual field perception [Julesz, 1981]. (For a preliminary application of Julesz' theory to music, see [Gabel, 1993]). Other constraints on the operationalisation of such a model include the composer's aesthetic and ethical beliefs, as well as other socio-cultural issues which cannot be discussed here. Finally, a major obstacle could be represented by technological limitations of a particular task environment where the compositional process takes place. Much of the problem is a question of intelligent software design in the implementation of user interfaces and, especially, the controlstructure among musical objects and data existing at various scales of time and structure. Except for very few cases (the work of [Scaletti & Habel, 1991] can be viewed as one such exception), it is quite difficult to find computer music systems which let the composer decide where - at which scale of structure - to enter the compositional process. Still today, when this kind of problem has been largely and repeatedly recognised by many Composition, Composition Systems 206 ICMC Proceedings 1994

Page  207 ï~~composers, designers of musical systems and musicians themselves seem to lack this sort of ecological awareness. My opinion is that though everybody will accept that computer programs can be conceived as theories [Simon & Newell, 1970], in actuality very few are prone to understanding music composition as a kind of creative modeling of (prompting a theory of) musical experience. Fundamental in implementing an instance of TSE is the algorithmic procedure and its mathematical form. Indeed, the core model in this kind of musical undertaking is expected to " match the requisites put forth by the constraints; " show qualitative responses involving at least some of the major features of a phenomenon of emergence (see above); " be computationally feasible and conceptually clear to the composer. In my experience as a composer, models of nonlinear dynamical systems have proved to be of real use with respect to such expectations. Starting in 1989, 1 have used simple iterated equation systems as a central feature in the control-structure of granular synthesis [Di Scipio, 1990; Di Scipio & Tisato, 19931. Presently, the same kind of theoretical perspective is taken in order to enlarge the scopes of nonstandard methods of sound synthesis to timbre composition; this research is now coming to some results with the implementation and analytical characterization of functional iteration synthesis [Di Scipio & Prignano, 1994] (that I hope will be the object of a specific presentation to the computer music community as soon as possible). The term functional iteration was drawn from terminology introduced by one of the fathers of chaos theory [Feigenbaum, 1980], and indicates the iterated applications of difference equation systems including nonlinear transformations and mapping. There is a whole world of multi-levelled structural behavior to be studied in such simple models, and its application in the digital generation of sound seems to give rise to sonorities which fall outside the classes of sounds obtainable by any standard method of synthesis. (An important point: in such kind of musical research one should avoid to use models of deterministic chaos to get a sort of superoscillator or enhanced unit generator. The essential philosophy in this kind of application aims, after all, at making practicable a renewed conception of the compositional work; it would only be misleading to exploit such means within a conceptual paradigm extraneous to them.) 3.3 Subsyinbolic strategies of composition A final observation refers to the fact that in all models which embody some instance of a TSE, the level of the glp, at which the microcompositional process operates, corresponds to the rate of mental information processing proper to a subsymbolic stage of cognition. A sound grain, a digital sample, a texton, etc... are not in themselves significant as percepts, and can hardly be experienced as discrete symbols of a musical construction (exceptions may come to mind, though). Presently, my definition of subsymbolic stragies of composition can only stand for this correspondence with scale or dimension in the level of musical structure directly addressed by microcompositional models. It could be possible, however, to establish a closer link between different cognitive rates of music information processing, the subsymbolic included, and compositional models embodying a possible TSE - for the notion of emergence is of central relevance also for theories of cognition and concept formation [Edelman, 19891. In much the same way as subsymbolic paradigms of cognition try to capture how sensorial data are mentally pre-processed to constitute a symbol, and then how symbols are treated as components of higher forms of organization, a holistic approach to composition - understood as a theory of sonological emergence - may find it appropriate to describe its models of sonic design in terms of subsymbolic processes yielding the musical structure we experience by listening. 4. Summary In the first part of this paper, I have discussed the relation of forms to material, of sound to structure within the dualistic paradigm of music composition - which, in the computer age, reflects into the distinct experiences of algorithmic composition and timbre composition. As shown in the second part, computer music composition can go beyond the boundaries of that paradigm; by blurring the separation between models of sound material and models of musical design, one can approach composition from a holistic perspective, questioning the very opposition (once reputed ontological) of form to material, of structure to sound In the third part, I have attempted a clarification of what kind of theory of composition may be put forth within the holistic paradigm just described. The definition of Theory of Sonological Emergence has been prompted, together with a short discussion of the constraints and the prerequisites for instance models of such theory. 4.1 Acknowledgment Thanks are due to Ignazio Prignano; our common work, along with discussions with other friends at LMS and elsewhere, have been a major ispiration for the above reflections. ICMC Proceedings 1994 207 Composition, Composition Systems

Page  208 ï~~References [Adorno, 1970], T. Wiesengrund Adorno, 4Astetische Theorie. Suhrkamp, 1970 [Berg, 1985] P.Berg, 'ILE. A language for sound synthesis". In Roads, C. (ed) Foundations of computer music. MIT Press, pp.64-108, 1985 [Boulez, 1987] P.Boulez, "Timbre and composition. Timbre and language", Contemporary Music Review, 12(1):pp.161-172, 1987 [Bregman, 1990] A.Bregman, Auditory Scene Analysis. MIT Press, 1990 [Chandra; 1993] A. Chandra, "COUNTERWAVE", Proceedings of the 10th C.I.M., University of Milan, pp.151-158, 1993 [Di Scipio & Prignano, 19941 A.Di Scipio & I.Prignano, "Sintesi per iterazioni funzionali." LMS Technical Report, 1994 [Di Scipio & Tisato, 19931 A. Di Scipio & G.Tisato, "Granular synthesis with the INTERACTIVE COMPUTER Music SYSTEM", Proceedings of the 10the C.I.M., University of Milan, pp.159-165, 1993 [Di Scipio, 19901 A.Di Scipio, "Composition by exploration of nonlinear dynamical systems", Proceedings of the ICMC, ICMA:pp.324-328, 1990 [Di Scipio, 1991] A.Di Scipio, "La musica di due culture. Tracce di una mutazione". In Boschi, C. (ed) Musica e Scienza. Ii margine sottile. ISMEZ:pp.17-70, 1991 [Di Scipio, 1993a] A.Di Scipio, "Models of material and of musical design become inseparable. A study in composition-theory", Proceedings of the International Conference on Cognitive Musicology, University of Jyviskyla, pp.330-316, 1993 [Di Scipio, 1993b] A.Di Scipio, "Micro-time sonic design and the formation of timbre", Contemporary Music Review (in press, summer 1994) [Di Scipio, 19941 A.Di Scipio, "Kair6s. Sulla morfologia dinamica del suono e del tempo", Sonus (in press, summer 1994) [Dodge & Jerse, 1985] C.Dodge & T.Jerse, Computer Music. Schirmer, 1985 [Dufourt, 1991 ] H.Dufourt, Musique, Pouvoir, Ecriture. Bourgois, 1991 [Edelman, 1989] G.Edelman, The remembered present A bilogical theory of consciousness. Basic Books, 1989 [Feigenbaum, 1980] M.Feigenbaum, "Universal behavior in nonlinear systems", Los Alamos Science, 1(1):pp.4-27, 1980 [Gabel, 1993] G.Gabel, "Julesz' Texton Theory applied to Music", Proceedings of the 4th Arts & Technology Symposium, Connecticut College, New London: pp.87-105, 1993 [Julesz, 1981 ] B.Julesz, "The elements of Texture Perception and their Interactions", Nature, 290: pp.91-97, 1981 [Laske, 1989] O.Laske, "Towards an epistemology of composition", Interface, 20(34): pp.235 -269, 1989 [Lukscs, 19711 G.Lukcs, Fruhe Schriften zur Asthetik (1912-1918) II Band, 1971 [Manzolli, 19931 J.Manzolli, "Musical Applications from FRACWAVE Sound Synthesis Method". Preprints of the 94th AES Convention, Berlin, 1993 [Pousseur, 1970] H.Pousseur, Fragments Theoriques sur la Musique Experimentale. I. University of Bruxelles, 1970 [Roads, 1991] C.Roads, "Asynchronous granular synthesis". In De Poli, G., Piccialli, A. & Roads, C. (eds) Representations of musical signal. MIT Press, pp.143-186, 1991 [Rodet et al., 1984] X.Rodet, Y.Potard & J.B. Barriere, "The Chant project: form the synthesis of the singing voice to synthesis in general", Computer Music Journal, 8(3): pp.15 -31, 1984 [Scaletti & Habel, 19911, C.Scaletti & K.Habel, "An Object-based Representation of audio signals". In De Poli, G., Piccialli, A. & Roads, C. (eds) Representations of musical signal. MIT Press, pp.371-390, 1991 [Schaeffer, 1966] P.Schaeffer, Trait des Objects Musicaux. Le Seuil, 1966 [Simon & Newell, 1970] H.Simon & A.Newell, "Information Processing in computer and man". In Z.Pylyshyn (ed) Perspectives on the Computer Revolution. Prentice-Hall, pp.256 -273, 1970 [Truax, 1988] B.Truax, "Real-time granular synthesis with a digital signal processing computer", Computer Music Journal, 12(2): pp.14-26 [Truax, 1992] B.Truax, "Musical Creativity and Complexity at the Threshold of the 21st century", Interface, 21(1): pp.2942, 1992 [Wilden, 1972] A.Wilden, "La scrittura e it rumore nella morfogenesi del sistema aperto". In E.Morin (ed) Teorie dell'evento. Bompiani, pp.77-112, 1972 [Xenakis, 1991] I.Xenakis, Formalized Music. Pendragon Press, 1991 (extended version of the 1971 edition) [Zuidervaart, 19911 L.Zuidervaart, Adorno's Aesthetic Theory. The redemption of illusion. MIT Press, 1991 Composition, Composition Systems 208 ICMC Proceedings 1994