Monday, September 25, 2006

Simplicities and Their Complements

From a review at Scientific American of a spate of recent books on science and religion:

Richard Dawkins, in The God Delusion, tells of his exasperation with colleagues who try to play both sides of the street: looking to science for justification of their religious convictions while evading the most difficult implications--the existence of a prime mover sophisticated enough to create and run the universe, "to say nothing of mind reading millions of humans simultaneously." Such an entity, he argues, would have to be extremely complex, raising the question of how it came into existence, how it communicates--through spiritons!--and where it resides.


If the reviewer, George Johnson, has summarized Dawkins correctly, it shows Dawkins's ongoing inability to extricate himself from confusions about simplicity and complexity. He's had this problem at least since The Blind Watchmaker, which contains several arguments conflating several completely distinct notions of simplicity in explanation -- material simplicity, simplicity of efficient causes, and simplicity of the structure of the explanation. And it's a pretty seriously muddle, too, because they tend in completely different directions. Complete explanations, for instance, are always more complex than what they explain, because otherwise they wouldn't completely explain the explanandum (put roughly, the explanation has to have every element of the explanandum and its explaining factor, so the explanation will always be more complex than an adequate description of the explanandum); material simplicity/complexity and efficient simplicity/complexity aren't about the explanation itself, but about what you are appealing to in the explanation. (Relying on one or the other results in very different types of explanation.) In the above summary, we find Dawkins portrayed as muddling together yet more types of simplicity/complexity, in this case complexity of effects with internal structural complexity. Since things aren't constituted by their effects, and things that are extremely complex structurally sometimes are less sophisticated in their effects than structurally simpler things, any link between the two would have to be defended with considerable sophistication. Perhaps Dawkins has a viable argument, and Johnson just isn't giving us any idea of what the argument is; but I'm skeptical. And the questions of coming into existence, communication, and residence arise regardless of complexity; and they hardly go without answer, although I imagine Dawkins wouldn't be impressed with the answers usually given.

I also find slightly amusing the fact that Dawkins, at least as portrayed here, implies that creating and running the whole universe is a less sophisticated activity than reading the minds of all the intelligent little primates on a small planet in that universe.