1.
We first met as three: Rhodri, Gary, Sarah.
Rhodri explained that when improvising, he responds to the environment around him: the sounds made by other musicians, ambient sound in the room, sounds encroaching from outside the room, sounds made by the audience.
Gary realised that when writing, he does something very similar. He plays music – often music by Rhodri – he responds to noises in his house, he responds to noises encroaching on his study from nearby houses, he responds to things he has heard or recorded on his dictaphone.
Sarah suggested we start by seeing if Rhodri and Gary could work together. Seated near enough that each couldn't help but be aware of the other, Rhodri would improvise musically – responding to the sounds of the rehearsal room, which might include sounds made by Gary – while Gary would come up with text – responding to the environment around him, which would include Rhodri's musical improvisation.
This we did for twenty minutes.
Sarah watched and listened.
2.
We next met as four: Rhodri, Gary, Sarah, and the actor Lynn Hunter.
We repeated the basic process – Rhodri improvising musically while Gary improvised text, but now with the addition of Lynn. Lynn was given the text which Gary had generated at the first session, and invited to use that text to respond to what was going on around her – the room she found herself in, Rhodri's musical improvisations, Gary's activity or inactivity, Sarah as audience. Lynn could deliver all, none or some of the text, in whatever order or manner seemed interesting to her.
Now Gary's improvised text, and Lynn's improvised performance of that text, formed an unavoidable part of the environment Rhodri was responding to, and could feed directly into Rhodri's musical improvisation.
3.
We repeated this process several times, each time generating more text, which would then be added to the text Lynn had at her disposal in the next session.
Sarah guided us in forming the broad rules of the improvisation – for example, as non-actors, both Rhodri and Gary were keen that Lynn did not draw them into her performance. But Sarah observed that the non-actors' embarrassment made for more interesting results both muscially and in terms of the text generated; and so Lynn was directed to include Rhodri and Gary in her performance whenever she felt like it.
On a similar theme, Sarah noted that when we performed before an audience, Lynn and Rhodri's improvisations would be witnessed directly, while the audience would only have access to text Gary had improvised in previous sessions (which, Sarah suspected, Gary was editing before handing it over to Lynn). To correct this imbalance, Sarah suggested that Lynn should end the improvisation by taking whatever text Gary had produced in that session and, without any preparation, reading it to the audience.