The Analysis of Computer Music Works with the Help of Code: Six specific features of the code to consider

International audienceIn the electroacoustic community, it is well known that one specificity of this music is that it had no score. We know that electroacoustic music is not made and composed with traditional notation and instruments, and we know that this music does not match the classical music theory framework. Therefore, the analysis of this music has an essential issue: how can we analyse a music that we can only access via listening? As François Delalande pointed out, the arrival of electroacoustic music led to “disorder if not revolution” in the field of musical analysis (Delalande 2013: 7). My research focus on a subfield of electroacoustic music, the computer one. I use this term to refer to a music which is composed and produced with programming language, like the descendants of Music N (Mathews 1969) and Patcher (Puckette 1988) for examples. In comparison with the electroacoustic music the computer one is particular because it lead to another object, something which have been made by the composer and which is made of discrete signs, the code. My research field is the analysis of computer music based on the latter code, and the general question of my research is how to analyse computer music with the help of their code? The presence of the code in the analysis of computer music is not new. For example, at a practical level lets consider the pioneer work of (Lorrain, 1980) and (Di Scipio 2000). At a theoretical level one can consider (Stroppa 1984), (Battier 2003), Zattra (2015) and (Clarke 2012). However, I found that the presence of the code in the theory of computer music analysis has not yet been fully expressed. In a previous work (Larrieu 2019) I have shown that the situation for analysing computer music is different from the general electroacoustic one, because the computer music is not a ‘non written music’. I have concluded that in computer music the analyst can both listen to the music and read the computer code, and moreover that a theory for analysing computer music work is relevant if it does allow a circulation of meanings, between the code and what is listen. Subsequently in (Larrieu 2020) I have suggested an organisation of the well known existing methods to analysis computer music – i.e. the faktura, the genetic criticism applied to computer music and the Interactive Aural Analysis – according to their proximity with the code and the listening. In this presentation I would like to pursue the question of musical analysis in the computer music context. I will describe 5 specific features of the computer music code when considering analysis. Those features are based on a work of my PhD, where I have analysed Jupiter (1987) by Philippe Manoury and assisted by Miller Puckette, for flute and live electronics. I will point out the following features: (i) code contains knowledge (ii) code contains semantic information (iii) code result from an activity of design, (iv) code contains a lot of information and (v) code is obscure. The two last points makes sense in comparison to the analysis based on traditional score. They are here to point out that ‘something has changed’, but of course they are not at the same level of the three first three. My presentation will be a description of those 5 specific features. I understand them as a first framework where one can think of the analysis of computer music works with the help of their code. The goal of my presentation is first to consider the code in musical analysis, seeing it more in relation to the composer than to the computer, and then to engage the discussion with the community in presenting those specific features of the code in computer music context