Information Technology Reference
In-Depth Information
of the material space, but when we carry out the generative procedure, a single point
will be the result. That point may or may not be a part of what we expected, possibly
requiring a revision of the conceptual representation.
8.3.3 Interplay Between Representations
The philosopher Daniel Dennett has said (Denton
2004
):
The purpose of the brain is to predict the future, to make plans and hopes, and in following
these predictions, we partially make the future.
The brain is good at prediction, because that is what it is evolved to do. The
musician and writer Stephen Nachmanovitch (
1990
) said that life
is
improvisa-
tion. But creative processes also mimic what life is about—predicting, pursuing,
acting, adjusting, etc. in a continuous circular process. So, in describing how
we form our world, Dennett also gave us a good description of how we create
art.
As a composer, I use generative processes to project my ideas beyond my pre-
dictive horizon (Dahlstedt
2001
). I may understand the conceptual network in the
immediate neighbourhood, and apply the algorithm or process to get further away,
hoping that the interestingness will carry over to distant parts of the space. Or I
may understand the broad paths in the conceptual network of the process, and
apply it, leaving the details to the process. I may use generative processes that
are too complex for my predictive capacity, in a trial-and-error fashion: adjust-
ing parameters as I go, based on temporary results, and possibly, at the same
time, adjust the actual algorithm itself. This amounts to the reiterated interplay
between material and conceptual representation, through development and pars-
ing.
This interplay is crucial to the proposed model. An idea expressed in a concep-
tual representation is realised by searching for a suitable material representation,
either by gradually shrinking the set of points covered by the conceptual represen-
tation in an iterated process between idea and tools, or by searching for a unknown
pleasing result by trying a sketch, evaluating it and modifying it until something
interesting is found. Once again, this is an iterated process between ideas, tools and
material, and can be illustrated in terms of these networks (tool networks, concep-
tual subspaces, etc.) that coexist as different organisational principles in the material
space.
There has to be a path from the material representation back to the conceptual
representation, to carry interesting coincidental results back into the conceptual rep-
resentation, and to provide for feedback from temporary results to affect the con-
ceptual representation. How do we recognise pregnant ideas and interesting coin-
cidence? What we need is a kind of reverse development process: the parsing of a
material representation into a conceptual description. This is a central part of the
creative process; our brains do it all the time, but computationally it is a non-trivial
