Mark Summers (Sheffield University): Human-Machine Creative Collaboration and Copyright: performing with improvisation-driven music systems
Copyright can be problematic in areas unforeseen by legislators. One such area is that of interactive music systems which combine an improvising instrumentalist with live electronics to create a collaborative performance. The capabilities of such systems are as varied as the number of systems that have been created. These range in a continuum from the instrument paradigm, where an acoustic instrument is extended electronically (for example, MIT’s Hyperinstruments), to the player paradigm, where the system provides an improvising partner that feeds off the instrumental improvisation but is independent of it (Lewis’s Voyager or IRCAM’s OMax).
This paper will discuss these systems and their performance, raising questions about works (in the improvised input, in the output of the system, the mixed output of system and improviser or even in the system itself?) and authorship (is the improviser performing or creating, is the creator of the system the author of its output, are the creator and improviser co-authors of the mixed output?). In the light of these questions and based on legislation and case law, suggestions will be made as to where copyright may subsist and who might own it.
Machines don’t do anything other than what they are told to do, so this presentation is about the human-machine interaction rather than the agency/intention of the machine.
Assumptions for this presentation:
- UK music made by UK musicians
- The improviser and system creator are two different people
- Any works are original
- Audio recordings are made
An improviser plays an instrument which plays into microphones – into a system – which then puts out a sound through a speaker. The improviser and the speaker simultaneously reach the listener.
What is a work? Is there a difference between music and a musical work? Is there a ‘work criterion’ in law? Tension between the musical work of copyright law and that of musicians. There are ‘thick and thin’ works. How does something qualify to be considered a ‘work’ – in music or in law? Music exists in time. A musical work needs time.
Mark’s own music does not follow the pattern of a composer’s ‘made work’ interpreted by performers. Creativity is shared – performer and composer can be the same person.
Copyright has a bias toward work-concept music. IDMS is a work to be performed; it is a tool used to perform work; it is an agent in music making simpliciter. Precise sound structure is only defined in real time in the performance. Systems are not complete without an improviser. Resulting works should be judged on sound hear in performance.
Is the improviser ‘mere’ performer or author. Mark cites case law – Hadley v Kemp; Beckingham v Hodgens; Fisher v Brooker. These judgements variously accept the performer, or not, as a co-author. Sound is anot a musical element protected by law (Barron 2006), but Sawkins v Hyperion suggests that it is important.
Creative collaboration – good improvisers are valued for their inventiveness, but they are expected to improvise – that is their function. IDMS is often ‘something to play with’ not just ‘something to play’. Joint authorship? Mark now speculated about possible permutations of reward (economic rights). He looks at binaries of creation – the system creator creates a work (getting copyright) but does not get copyright if they only make the tool system. The performer may or may not be rewarded in these scenarios depending on their input and the relationship to the ‘system creator’. He then applies this to computer-generated works.
Mark now develops a more complex system of possible permutations – easier to show here in a photo of his slide (right).
He now shows us some case studies including audio and discusses the level of ‘creativity’ contributed by the performer. He attempts to define different pieces as ‘works’ or ‘non-works’ and speculates as to how PRS declarations should be applied for these works.