Emergence and the Computational Sublime
Jon McCormack (jonmc@csse.monash.edu.au)
School of Computer Science and Software Engineering
Monash University, Clayton, Victoria 3800 Australia
As an artist, I am interested in exploring computation as a medium — as a raw material in which to define an artistic vision. However, to truly adopt this position involves a critical re-evaluation of many of the ideas that dominate computing and its related techno-scientific disciplines. It also requires a certain amount of faith: faith in computing as a concept and faith as to the range and scope of its descriptive power. Perhaps even, faith as to what might become possible that is not possible now — making this seem like a perilous position to be in.
Even on the simplest computers, the phase space of possible programs is vast ‘beyond imagination’. To explore this phase space completely, even simultaneously on billions of the fastest computers ever built, would take more time than the universe is old. This vast space suggests a new kind of ‘computational sublime’. But within such a vast space, most of the set of possible programs will be inoperative or uninteresting. The questions we might ask about this are: how do we find the interesting ones? How do we define the term ‘interesting’ and why should it be that only programs that fit such a category should be the subject of our enquiry? Might there be something to be gained from looking at the uninteresting ones as well? Can the concept of the computational sublime be different from other romantic narratives of the sublime?
Artists are looking to create works that are ‘more than the sum of the parts’, that lead to outcomes that they did not anticipate or expect. This led me to consider a long-standing debate in philosophy regarding the concept of emergence: the creation of fundamentally new properties. The ‘emergence debate’ has been renewed in recent times due to results from the ‘complexity sciences’ where computer experiments reveal ‘emergent patterns and behaviours’. If artists want to create things that exceed the designers expectations, does the medium of computation (in theory at least) permit this?
People have used computers for many different purposes and in many different ways, but undermining all interpretations is the idea of representation — computers are symbol-processing machines that obey a deterministic set of rules. Symbols are ‘given meaning’ by the programmer, a process that is so transparent in programming it is rarely consciously understood or deeply considered. If we are to consider computation as a medium in which to work critically, then this issue of representation or mimesis becomes paramount.
I would like to draw open the question of thinking about computing in ‘Kantian aesthetic’ terms (as a matter of sensuous beauty or when a sensuous object or process stimulates our emotions, intellect, and imagination). I can see many ways in which this view may be limiting and problematic. If the concept of ‘aesthetic computing’ is not just to service or glorify the ‘positive self-image of modern Western culture’, nor to dwell as a techno-romantic adjunct to the rationalist ideology, then it needs be considered within a broader context. As an example of an alternative model, I would like to introduce some of Heidegger’s thinking on aesthetics and being into the discussion.
Some References: