Information Technology Reference
In-Depth Information
4.1 Introduction
This chapter documents a conversation on the prospects for computational evalua-
tion of art, aesthetics and creativity. The dialogue began in July 2009 at the Dagstuhl
seminar (Boden et al.
2009
). At the seminar a small group of participants decided to
explore the problem of evaluation of creative works. Here “evaluation” included the
decision process during artwork production and that following an artwork's com-
pletion, including evaluation by others such as audiences and critics.
Following the seminar, the dialogue continued via email over a period of many
months (from July to November, 2009) and covered a variety of fascinating issues.
What follows here is an edited version of this correspondence, in chronological
order. As the reader will understand, the dialogue is relaxed and conversational—
points are not always justified, and the unplanned and improvisational nature of
the conversation reveals different ideas than would be found in a formal authored
chapter. But we hope readers will appreciate the sincerity and openness of all the
contributions, the value of candid personal opinions, and the shared sense of trying
to explore the complexity of the issues raised.
There is much here that philosophers may be familiar with (and possibly even
aghast at). But it does serve as an important historical record, particularly from the
perspective of a number of pioneering artists who have been working in this area
for decades. Their wisdom and experience brings a compelling perspective to the
conversation. The collective insight of these pioneers provides an important point of
reference for the next generation of researchers and artists entering the field.
4.2 Background: Evaluation of Artistic Artefacts
Before presenting the edited dialogue, a short background is first provided, in order
to establish the context from which these discussions began.
1
The discussion is cen-
tred around the idea of computational evaluation of creative artistic artefacts. There
are a number of points to be made to flesh out this idea. Firstly, how something
is evaluated depends on the evaluator's perspective and role. This role may be as
creator or designer, viewer, experiencer, or interactive participant.
This leads to some initial questions:
•
What are the main features of human creative and aesthetic evaluation?
•
How do these features (and the methods that are used) change according to the
evaluator's role in the process?
•
What aspects of evaluation can be made computational?
•
Is it necessary for computational evaluation to mimic the evaluation methods of
humans?
•
Does it make sense to automate a task that is so especially human?
1
Elements of this section are based on the initial Dagstuhl group discussions (Boden et al.
2009
).
