In the domain of music for performers and electronic sounds (whether fixed or live) there are various paradigms of interaction: the performer(s) may be situated in an electroacoustic ‘environment’; there may be a primarily responsorial or ‘proliferating’ relationship; or the relationship may be closer to the traditional one between soloist and accompaniment. These paradigms preserve a relatively unproblematic dichotomy between performer, whose sound is inextricably linked to a sense of action, presence and spontaneity, and fixed or treated sound, which is more or less de-coupled from this presence. Once one tries to create a continuous, intimate relation between the two, so that one is dealing with an extension of the instrument rather than an emulated ‘other’ or environmental context, one is confronted with a fundamental difference between a sounding body whose physical properties transparently determine its sonic possibilities, and the loudspeaker, which can produce practically any sound at all. This paper interrogates this dichotomy between the fallible-corporeal and the fixed-disembodied, activating questions both about the social fact of live performance and about the compositional practices which give rise to a sense of extended instrumentality.