Maybe the structures can be designed to frame a certain kind of interaction.
I am working on an interactive installation with a fine artist which is
designed to encourage a _dialogue_ between the installation and the
participants (who are not performers, by the way, but users. The computer
'reads' the behaviours of the participants, and responds to those - but not
on a "click and go" basis. It reads patterns of behaviour, rather than
specifics of behaviour, and produces its own behaviours, which the
participant then responds to - and so on.
The dialogue is in the hands of both installation and spectator. In a
sense it is an improvisation between a) participants and installation and
b) between the participants themselves. In order to improvise with some
kind of focus participants need to learn how the installation is responding
to them, and learn how to interact with it productively (a feature of all
interaction). Another analogy might be that of team games. You learn how
to respond to team mates when playing soccer, for example, or baseball,
within the strcutures of the rules of play. Once you know the structures
of the game you can play the game.
We are still working on the programming of the structures, attempting to
set up the possibilites for a sophisticated dialogic interaction (a far
more difficult enterprise than a one-one interaction between user and
installation or between installation and an ensemble of performers, whose
behaviours are designed as a choreography of the space in some way and
interactions pre-planned.) If anyone is in London 25th-31st August we will
be researching the interactivity at The Place Theatre please come along a
play.
So - the specifics of interaction probably can't be designed. However, the
structures of play of the interaction can (indeed must be) be designed.