23 July 2007


Recently there has been a good deal of debate about the existence, or otherwise, of God. Christians face the problem that the God of the New Testament is normally presumed to be both omnipotent and supremely good, which seems hard to square with the existence of suffering. It should be noted that moral "goodness" is notoriously difficult to define with regard to human behaviour, and is presumably even more problematic applied to a deity.

If we consider the more general idea of a creator or other supreme entity, and leave the idea of "goodness" aside, then we are faced with two key questions:

(A) What characteristics would such an entity have?

(B) What would count as "evidence" of the existence of such an entity?

Much of the debate seems to have blurred the distinction between the Christian God and the idea of God in general, and also failed adequately to consider questions A and B above. By not properly defining terms or considering the fundamental issues, the debate has effectively had the character of political argument rather than intellectual analysis. If a Christian claims that God has particular characteristics definable in ordinary language, or that God has performed particular actions comprehensible to human minds, then it is relatively easy to shed doubt on this. The effect is to undermine the concept altogether.

In a discussion with Alister McGrath, Richard Dawkins says "we have to talk about probabilities", and McGrath agrees. McGrath seems to think there is an inherent improbability in the complexity of living things, which makes the existence of a creator god plausible. Dawkins responds "I want to say exactly the same thing of a designer … Any being capable of designing a universe, or an eye, or a knee would have to be the kind of entity which would be statistically improbable in the same kind of way that the eye is". But both seem to be misusing statistics: we cannot really assign probabilities to hypotheses as abstract as that.

The strongest argument against the existence of God is probably that of superfluity (a version of Ockham's Razor): the concept is irrelevant because it is neither useful nor needed in explaining any facts we wish to account for. However, this argument is not as strong as it may appear.

Within a closed and self-contained system of explanation, e.g. physics, a concept at the biological level such as 'sexual reproduction' is not required to explain all the observed phenomena: we merely need to know the initial conditions and the laws of particle interaction. Yet when we move to a more macroscopic level, we find that it is useful and necessary to invoke certain additional higher-level concepts to provide satisfactory explanations. (This is sometimes expressed using the idea of 'emergent properties'.) The fact that we find it perfectly possible to explain phenomena without recourse to additional theoretical entitites is not strong evidence that no other concepts would be required at a 'larger' level of significance.

Other articles
Wikipedia on emergent properties
Discussion between McGrath and Dawkins
Stephen Law on the probability of God

No comments: