Wednesday, October 24, 2012

Simon Conway Morris at Norwich Cathedral. Part 2

Getting ready for the lecture

In spite of its billing Conway Morris’ lecture didn’t contain a great deal about the Christian implications of his content. However, I myself was very interested to hear his remarks on convergent evolution and the high optimization of functionality. These two facts alone suggest that whatever the engine driving evolution may be it has a startlingly high probability of generating an extremely unrepresentative class of structures. Perhaps Morris was using extraordinary facts like this to quietly argue for themselves.

I was also interested in Morris’ brief excursion into mysticism toward the end of his lecture when he referred to the mysteries of traits like the love of music and the mythopoeic nature of human beings. (In this connection see this post of mine on William Irwin Thompson) Although Morris didn't explain the significance of these traits, in some ways his was a classic response. If the world of scientific logos doesn't argue explicitly for the existence of God then one is very much thrown back on the mood of mythos that the findings of sciences invoke.

Scientific establishment figures like Conway Morris stand for a mechanism of evolution that the homunculus IDers will claim does not work. Whether or not the science behind homunculus ID is correct, this claim appears to have an underlying motivation based on a misreading of evolution as a “blind and mindless process”, incapable of creating anything. Under this assumption an intervening homunculus is needed to make good the failure of evolutionary theory. The view of evolution as a “blind and mindless process” is, ironically, also endorsed by many an atheist: That the setting up a physical regime where evolution works is in itself a task that is highly computationally complex appears not to be factored in by either party. Although I have lot of respect for the opinions of some homunculus IDers and feel that their work is often under-appreciated,  polarization has spoilt their chances of making a worthwhile contribution to the debate.

At the heart of the dispute between the evolutionists of the academic establishment and the homunculus IDers is, I suspect, a humanly insoluble problem.  To illustrate let’s start by taking a single organism. Are we able to enumerate and evaluate all the possible heritable mutations it might undergo and determine what percentage of these mutations on a per environment basis will give rise to a stable population that has the mutant gene? Even if we were able to do this for a single organism we still have to face the wild card of a changing environment. Some otherwise viable mutations may not work in a changed environment. Even with a single known organism the possible evolutionary increments it could undergo are very difficult to catalogue and evaluate. So now imagine trying to catalogue the huge open ended vistas of configuration space and attempting to determine whether the juxtaposition of possible configurations fall into relationship with one another in a way that facilitates evolution.*1

My own guess here is that exploring configuration space is all but beyond our current computational technology with its use of linear computational streams. In fact I suspect that the problems of exploring configuration space, especially when we factor in possible environments, is computationally irreducible*2 and therefore there is no quicker way of analyzing this space apart from actually designing a simulacrum with the task of trying to get evolution to work (or not work as the case may be).*3. Until then we are very much dependent on the observational sciences of molecular and paleo biology; in this sense the simulation has been done for us! 

Thanks from the clergy at the end of a fascinating and stimulating evening.

*1 There is also another question that considerably complicates this picture even further. This is what I can best describe as “informational cross fertilization”; almost a kind of “interspecies sex” if you like. In the above text I have imagined an organism undergoing an incremental change. However, if this organism exists in an environment of other organic innovations, it is conceivable that it could make a step change by somehow receiving information (via virus infection?) about these innovation, information that could be used to take its development forward by a quantum leap. This co-option process certainly happens in technological development but as I'm not a biologist I've no idea how important it is in biological evolution. If this kind of cross fertilization exists then we will have a non-linear system of change which may account for the jumps we see in evolution.
*2 The idea of computational irreducibility comes from Stephen Wolfram.
*3 Relevant to question of whether something can be rendered using analysis is the work of John Holland. See here

No comments: