Page  00000001 THE COMPOSITIONAL PROCESSES OF ELECTROACOUSTIC COMPOSERS: CONTRASTING PERSPECTIVES Catherine Upton Royal Northern College of Music Library, UK Barry Eaglestone University of Sheffield UK Nigel Ford University of Sheffield UK ABSTRACT This paper explores composers' approaches to composition, organising their work, and attitudes to the tools and technology. This qualitative naturalistic study involved in-depth interviews with four composers, and inductive data analysis. It concludes that individual differences have software implications, and a challenge for composition software designers therefore lies in the heterogeneity of their user population. 1. INTRODUCTION This paper presents a study of working methods, attitudes and use of the tools and technology of electroacoustic (e/a) composers.. Full documentation of related work is given in [5]. Here we note the main conclusions. There is currently a lack of a research base for e/a composition systems. Studies have been mainly restricted to 'traditional' tonal Western Art Music. The search for objective models of composition is also exacerbated by sparcity of insightful research into creativity, particularly from a software perspective. Appropriate methodology is also problematic because of the difficulty of determining cognitive processes of composers. Further, derivation and validation of models requires empirical studies, but these have rarely been used to research composition. The case study-based and experimental studies that exist fail to address interaction with computer systems, or to address complexities of professional composition. Few go beyond trivial composition exercises using crude and simple sound sources. A number of studies have focused on e/a composers' attitudes towards user interfaces, software environments, and user requirements. However, understanding gained is fragmentary and biased towards a software development perspective. The present research therefore seeks to expand upon these studies by seeking to enrich understanding of the compositional process. It does not attempt to define a rigid model of the composition process, or discover underlying cognitive processes that occur when creativity occurs. Rather, it aims to develop a holistic picture of facets involved in composition, to gain a general impression of how individual composers work, and the attitudes and issues which impact. 2. METHODOLOGY A qualitative and naturalistic approach was taken to address human and divergent elements of creativity in composition. Being qualitative, the study does not attempt to generalise or come to categorical conclusions. Instead, deeper more detailed understanding may be gained from this less prescriptive intuitive approach. Composers were selected to provide a 'theoretical' or 'purposive', rather than representative sample. Four male composers, A-D, were interviewed, all attached to academic institutions, and compose from a Western Art music bias, as opposed to 'popular' e/a music. Data were collected using qualitative semi-structured interviews to produce holistic impressions of how each composer works, and their attitudes to their work and the tools that they use. All interviews followed as a basic outline, questions about; background information, the compositional process, and tools and techniques used. Interviews were recorded, transcribed and analysed inductively, applying the exploratory approach of grounded theory [2] to identify concepts suggested by the data. These were coded and drawn out of the data, and then cross-referenced and linked together to form conceptual models and categories. 3. RESULTS AND ANALYSIS Notions that emerged from our analysis concerned: composers' backgrounds and styles; composition processes; approaches to form and structure; how composition work is organised; and use of, and sometimes frustration with, technology. Here we discuss A and D only, since these represent the two extremes. Full analysis of all four subjects, including the qualitative data, can be found in [5]. 4.1 Musical Background and Compositional Style A and D, both well-established composers with teaching and research positions, have composed e/a music for many years with portfolios including prize-winning pieces. Both have traditional music backgrounds, i.e., undergraduate music degrees and specialised postgraduate music courses. Consequently, they have written instrumental music. However, they mainly focus on e/a composition now which is a distinctive genre with its own tradition, technological aspects of which are implicit in its creation. A and D both have strong backgrounds in this tradition, having worked with analogue techniques before computer technology predominated. Composer A described his work as "acousmatic", which has an idiosyncratic aesthetic. The focus is on timbral aspects, and performance and medium is always a fixed format. Diffusion of the sound into threedimensional space is an important and central aspect of his work. In contrast, D is more interested in technical and theoretical aspects of sound, rather than its aesthetic timbral qualities. He refused to be categorised, but indicated more technical research of IRCAM is an influence. D's interest is in technical components of a sound, and manipulating its physical properties. Thus, he works more theoretically than the other subjects,

Page  00000002 being more interested in techniques such as FOF synthesis: D's music is therefore quite pitch-oriented, although not necessarily based on the tempered pitches of the chromatic scale. He is concerned with the nature of sound, but perhaps focuses more on integral internal structures, and theoretical aspects of e/a composition. D indicated he composes predominantly for "mixed media". However, he has also composed pure instrumental and pure tape pieces in the past. Thus, each composer defines his work uniquely, with quite different compositional and aesthetic emphases. 4.2 The Compositional Process Participants described basic, fundamental stages which they often work through in composition. Composer A often starts by recording fresh sounds from which to formulate musical / sonic ideas. However, initial treatment of sounds is different in each case. He organises sounds through making value judgments, grouping them together and considering where in the framework of the piece each will work, before processing them. In contrast, D felt he lacks a standard process, but despite this, D's interest in, and use of, formal techniques does to some extent dictate the way in which he works with sound material. He often works from a pitch basis, and also, when composing 'mixed' pieces, sometimes uses timbral qualities of live instruments as a basis for the composition. D therefore often takes a physical property of sound material as a basis for his composition, as opposed to the other subjects whose work seems more influenced by extramusical aspects, even if they are at an abstract level (i.e. emotional, conceptual characteristics of the sound). Ideas from previous compositions, experiments, and simply 'doodling' in a software tool, also were important sources of inspiration in initial stages. This indicates two major concepts, also identified in other studies; the role of serendipity and experimentation, and the role of the composer's memory (discussed later). However, creative impetus can be seen as a result of divergent thinking [3]. Random thought processes and fortuitous 'accidents' experienced during development of one composition can result in fresh and new sound material for another. Once initial ideas for a piece are established, musical, or sonic, material can begin to be developed and shaped into a composition. This is seemingly a lengthy and demanding process, which is as equally creative as the preliminary phase. Testing and comparison forms a major part of the development process. All subjects talked about testing sounds out with others to see which work best together and how best to develop the musical language of the composition. Even D, whose work is more motivated by specific techniques and theory, suggested that experimentation with sonic material played an important part in development of his compositions, by providing creative impetus for the piece. All subjects indicated that much evaluated material will not be used, but this would seem to be an important part of the compositional process since through comparison and refinement of material the composition can begin to take shape and make sense musically. Also, the composer's memory has a significant role, since it is responsible for divergent thought processes that might lead to creative connections with other musical material that has been created by the composer: Thus, the development phase can be seen as a combination of divergent and convergent thought processes. Convergent and more methodical processes come into play as material requires formulation and refinement. Even though e/a music exists outside the language of Western tonality, composers will build their own sonic languages through combinational and structuring strategies, and interaction with sound material. This cannot be completely random, and must make sense musically and artistically, thus requiring convergent actions and thought processes. However, divergent thought processes during this phase can lead to creative connections and deviations, needed in any genre to make music original and interesting. All subjects alluded to this combination of methodical and intuitive approaches. The exact compositional processes may vary between compositions. However, this mixture of convergent and divergent thought processes, and methodical and intuitive actions, will always feature in the evolution of a composition. There are evident differences between ways subjects unify and structure the musical material, depending on style or aesthetic. For subject whose music is more timbral based, including A, this will come through connecting and putting together sounds to make sense artistically. For D, who is more interested in theoretical aspects of the sonic material, refinement might come through use of specific techniques which give the composition musical coherence. However, these approaches are not mutually exclusive. The composers who work more acousmatically, or timbrally, might use specific techniques to achieve desired effects, though they may be less physically theoretical than the techniques employed by D. Similarly, D may make small aural modifications to colour or embellish sounds, which are superfluous to techniques being employed. It could be said that A, who seems more focused on timbral aspects, shapes. a composition through intuitive connecting and moulding of the sounds and refines the music through specific techniques to achieve certain effects. In contrast, D shapes compositions through use of specific techniques and theories, and refines his music through intuitive aesthetic judgements. The compositional process could therefore be seen as a complex interplay of divergent and convergent thought processes, influenced by composers' individual sound languages, and those of an established e/a tradition. However, the divergent thought processes, the idiosyncratic, and the unintentional, are what makes a composition creative, spontaneous and innovative.

Page  00000003 4.3 Form and Structure and Organisation A and D expressed strong views on musical structure. A does not plan a piece's structure, but is conscious of the form it takes and has made attempts recently to work differently, giving this more consideration. He talked of past works being "through-composed", perhaps a result of not planning structure. Therefore his compositions might be structured more intuitively according to the nature of the sound material. He also mentioned that recently he has been trying to compose more thematically, working in sections, although this would not necessarily be akin to the narrative thematic dichotomy of Western tonality-based forms (e.g. Sonata form). Thus, for him, structuring is an inductive process. D, in contrast, tends to have more predetermined plans, in which techniques employed determine a composition's structure. D's interested is in composers, such as Stockhausen and Boulez, whose work stems from the German Electronische Musik tradition. Their music often focuses on specific techniques and algorithms to achieve 'total serialism'. D's work is of a similar aesthetic in which theories, algorithms, and techniques such as FOF and FOG synthesis, are used to create music in a precise way, and precisely structured. Thus, structure and technique are integral to his composition process. The subjects were asked about how they organise their work, since a composer's random ideas they may want to later re-use or develop further, must be organised in some way. Also, composers must keep track of progress of a particular piece, particularly when not working linearly, which this study shows, is often the case. The impression given by all subjects was of difficulty in organising work efficiently. A and D explicitly expressed a desire to be able to document their work better. Specifically, they must provide evidence of research as part of their work. But also, perhaps, they might simply want to keep better focus over what they are doing while composing. A main difficulty is devising ways of managing sound files for easy retrieval of contents. Composer A proposed audio-retrieval based on spectral density similarities as a partial solution. Query-by-example would find similar sounds to aid composition development, particularly of acousmatic/timbre-based music, when experimenting with sounds. However, audio retrieval research has yet to devise technology sufficient for extensive use in this type of work. Subjects were also asked about "pen and paper" notemaking. A and D made lots of notes, sometimes extensively, which forms an important part of their working processes. However, it seemed note-making was not always efficient or systematically organised. A commented that notes quickly became irrelevant, because they are often cryptically written on the spur of the moment, lacking details needed later. Also, the inherent sequential structure of a notebook was inadequate. Divergent and random thought processes tend not to be sequential. Therefore a notebook, as A and D mention, a notebook does not support arbitrary creative processes and flashes of inspiration particularly effectively. Also, the nature of e/a music makes it difficult to describe. It is rarely notated using the traditional five-line stave of Western music and graphical representation is not standardised and is personal to the composer. However, even if graphical representation is used, composers themselves must remember what it means, and it is unlikely to be accurate. Physical properties of sounds or a command line from a synthesis program may be written down, but this is unlikely intuitively to communicate information about the aesthetic nature of a sound, for example, and therefore might also be limited in organisational value. Despite these drawbacks, note-making appears to be an important part of the compositional process for which no better alternative has yet been found. All subjects indicated reliance on memory to organise work. So between note-making, memory, and personal systems for organising sound files etc., each composer finds an individual way of overcoming the difficulties in organisation. Also, it is possible that chaotic organisation of work can act as a creative impetus. Perhaps ineffective organisation, and resulting loss, due to memory laps, of information, prevents information overload of ideas from swamping composers during the creative process. If an efficient organisation system could be devised, would that perhaps result in too many creative choices? 4.4 Use of Technology All subjects use software quite differently, and use different software tools according to their objectives. None use analogue anymore (aside from sound recording), although D would not rule it out as an option. All use various tools for processing sounds, and a mixing program to construct the final composition, although their approaches to using these tools contrast. Musical background and experience of A and D shaped different ways of using technology. Both use software more extensively than the younger subjects, being more aware of its potential and limitations. Also, both observed that since techniques used often work in similar ways on different tools, it is not what you use but how you use it that matters. A's use of tools appears influenced by the acousmatic tradition. He uses a two-track editor, Adobe Audition, for recording and editing sounds. A often uses Bidule to understand whether more detailed work is required (in software such as Csound), as it allows him to investigate and develop sounds more specifically, but is more straightforward to use. He uses MAX/MSP, a graphical programming language, as a "major piece of software", and found artistic value in GRM tools (Groupe de Recherches Musicales), which are born out of the French Musique Concrete / Acousmatic tradition. He views the latter as "quite aggressive plug-ins" that "really pull the sound apart in a major way" allowing

Page  00000004 composers to be "very hands-on with the sound". These plug-ins allow timbral development of sounds. However, he also identifies their tendency to clich6. A also commented on the richness of choice of plug-ins currently available to composers. In D's music, specific techniques and theory are often integral to a composition. He is highly interested in manipulating physical properties of sounds and often needs lower levels of abstraction, to analyse properties and implement techniques he is using. He therefore requires high levels of control and precision. Consequently, D does much programming, using programming languages such as Csound and MAX/MSP, to devise what he needs to realise particular ideas or techniques. D does not consider technical aspects of his work to be isolated from the creative process, but integral and necessary. Being more influenced by research at IRCAM, D tends to use IRCAM tools, such as Audiosculpt. For mixing he uses ProTools, and MAX/MSP for live work, because it is then only really possible to use one program at a time. D does not use plug-ins, since they offer insufficient control, and are superfluous to the aesthetic of his music. They have no part in realisation of a technique since they are primarily for development (and distortion) of timbre. He describes plug-ins, and other commercial software, as high level "black boxes". Thus, composers' compositional processes are to some extent determined by the software they use and how they use it, but the software is in turn selected to support their particular perspectives and approaches. When questioned about frustration with software, subjects were generally positive. D was most positive. For him, problems with early software and analogue have largely disappeared. Decreasing costs and improved performance and capacity have made functions accessible, which previously were confined to large studios. However, D does not use commercial software, and is a competent programmer, thus able to overcome many potential limitations. Composer A has frustration with specific features, e.g., imprecise graphical sliders when processing a sound. However, his main concern was "peer pressure" to keep at the cutting edge of technology, and consequential time needed to master new software, and continual honing of already acquired skills. He felt that value is placed on use of tools, over creativity and musical skill. Two notions that strongly emerged were the need to keep composition software programs open-ended, and tensions caused by rigidity of audio sequencers. 4. DISCUSSION AND CONCLUSIONS It is the nature of this area that it is not possible to generalise. We should not build standard software around stereotypical composers' requirements, since individuality and idiosyncrasy are inherent in creativity. Composers A and D represent polar extremes in terms of approaches and consequential working methods, but each difference has software implications. It is apparent that composers select tools from a wide and evolving range, consistent with their compositional approaches and aesthetics. Also, some generic tools have limitations for all the composers. Software environments which host tools provide generic services which are clearly inadequate with respect to the subjects' needs to organise, annotate and record the composition process. For example, the subjects clearly perceive a need for database-like functionality within which to accumulate, organise and facilitate meaningful retrieval of audio resources. More interestingly, subjects exhibited different cognitive style; A and D conform, respectively, to global and analytic stereotypical extremes [4, 6]. Such differences are fundamental and intrinsic to individuals and require different functionalities from tools and the environments within which they operate. Thus, for example, A requires support for what is largely an inductive composition process, whereby structure, content and aesthetics of compositions will emerge from the process, while for D the process is largely deductive, since conception and theories will be realised and elaborated. To our knowledge, compositions systems have not been designed to address potential tensions caused by differing cognitive styles. Thus, rather than proposing technological solutions, we provide better understanding of differences and idiosyncrasies that can exist between composers, and hence challenges for designers of computer music software. 5. 'REFERENCES [1] Clowes, M., "An investigation of compositional practices in the field of electroacoustic music", MSc,University of Sheffield, 2000. [2] Ellis, D., "Modelling the information-seeking patterns of academic researchers: a grounded theory approach", Library Quarterly, 63 (4), 469-486, 1993. [3] Nuhn R, Eaglestone BM, Ford N, Moore A, Brown:G, "A qualitative analysis of composers at work", Proceedings of the ICMC 2002, Gothenburg., ICMA, pp 527-599, 2002. [4] Pask, G., "Learning strategies, teaching strategies, and conceptual or learning style", in R.R. Schmeck (Ed.), Learning Strategies and Learning Styles, Plenum Press, New York, pp. 83-99, 1988. [5] Upton, C., "An investigation into the compositional processes of Electro-acoustic composers", MSc Dissertation, University of Sheffield, UK, 2004. [6] Witkin, H.A., Moore, C.A., Goodenough, D.R. and Cox, P.W., "Field-Dependent and FieldIndependent Cognitive Styles and their Educational Research", Review of Educational Research, 47(1), 1-64, 1977.