Thursday, March 23, 2006

Sober continued

This is a continuation of the discussion started in my previous post.

We have three options that explain the appearance of what we see – Dembski's triad: chance; design and regularity. There are limitations in our use of these, which have been considered elsewhere in this blog – for example, we don't know every regular process. However, the logical completeness of these three has not been challenged.

We can't vaguely wave our hands at “everything” and consider these three options. We need to address a specific feature. Sober, in his paper, considers the vertebrate eye. Different people have suggested other things.

The key class of events that we need to look at to consider the reasonableness of chance versus design explanations are specified random events – for example, the appearance of a new protein which provides specific additional functionality in a biochemical system. This would be an example of a “novel trait” - an event which Sober and others acknowledge is the chance heart of the evolutionary process. We need to be careful in thinking about this: this isn't an argument that a new protein needs to appear fully specified – just that a protein with some new functionality becomes available to an organism. The objection by opponents of design theory is often related to this key point: “But evolution isn't a random event,” they maintain. This is true, but irrelevant – because even opponents of design theory do acknowledge that the starting point of evolution (the appearance of novel traits; the random production of a new protein to fill some role in an organism) is a random event.

In principle, it is possible – though difficult – to calculate the probability of such an event. For example, if a new trait relies on the appearance of a single new protein with enzyme functionality, it would be possible to experimentally determine how specified that protein has to be to provide this functionality. The assumption would be that in its present form, natural selection has ensured that the protein is highly specified, and well adapted to its role. When this enzymic functionality first appeared, it would not have been well specified, the protein having arisen from another one with different functionality, or from a random polypeptide. By substituting a progressively more damaged protein for the well specified one, it would be possible to determine how specified a protein has to be before it can express its required functionality. If we can establish how specified the proto-protein was (proposed random event) before natural selection got to work on it (regularity), we can come to a conclusion as to how reasonable the chance explanation is. As Dawkins is prepared to concede, too great a reliance on the chance hypothesis at long odds falsifies evolutionary theory.

It is possible to look at this from a different perspective, as well. Better understanding of the biochemistry of protein reactions, coupled with better imaging techniques, allows us to visualise what actually happens in these reactions. As we understand the functional components of a protein, it ought to be possible to determine what range of amino acid sequences would permit the required reactions to be carried out. It seems likely that most reactions will require components such as a sensing mechanism, an active site and a mechanism which uses biochemical energy. If we can understand the nature of what is happening at the molecular level, we can again come to conclusions about the likelihood of such proteins arising at random.

Andrew Rowell argues that, in general, evolutionary theory would predict that the gradients from unspecified proteins to well-specified proteins should be quite gentle (allowing the evolutionary ascent of Mount Improbable). What this means is that, if evolutionary theory is true, it shouldn't be unusual to see existing proteins co-opted for new functions, and random polypeptide sequences expressing functionality. In general, if we observe that such events are unusual, this would provide more support for design theory – though I have no doubt that people who prefer evolutionary theory would dispute this analysis, as indeed they dispute any reasonable falsification of their theory.

Opponents of design theory argue that this analysis only demonstrates how improbable evolution is as an explanation, and that with no prior evidence about whether there is a designer, this does nothing to show that design theory is more probable. Sober says:
To evaluate the design hypothesis, we must know what it predicts and compare this with the predictions made by other hypotheses. The design hypothesis cannot win by default. The fact that an observation would be very improbable if it arose by chance is not enough to refute the chance hypothesis. One must show that the design hypothesis confers on the observation a higher probability, and even then the conclusion will merely be that the observation favors the design hypothesis, not that the hypothesis must be true. (p.7)
Sober is correct to say that the issue is one of small probabilities. For example, when Behe argues that the bacterial flagellum can't evolve, what he actually means is that the probability of it evolving is too small to be considered a reasonable explanation. Design theorists argue that the amount of complex, specified information in organisms, in the form of the DNA genetic code, is too great for it to be a reasonable explanation that it arose by chance either one-off or through a naturalistic process (Dembski's No Free Lunch).

I suspect that Sober is missing the point, though. Whilst this may not be a sound philosophical argument, I think that at some point, common sense has to come into play. Let's go back to dice-rolling. The first five times I toss a coin, it comes up heads. That's not improbable. But I keep on tossing, and it keeps on coming up heads. Sober's argument is that, no matter how many times I come up with heads – even if I toss heads every second from now until the end of the universe – the chance explanation (this is a fair coin, and I have hit a lucky streak) is more powerful than the design explanation (this coin is loaded) – because I don't have any prior evidence about a designer.

Whilst technically this looks like a solid argument against design, in practice nobody lives their life on this basis. One has to doubt the fundamental soundness of an argument when the conclusions that it leads one to are so counter-intuitive, and against our experience. I suspect that the flaw is in the issue of low probability. The argument that you can't distinguish between chance and design in this way is fine when there isn't much to choose between the probability of each. But when the probability of chance gets too low and yet nonetheless something has happened, it is unreasonable to discard the design hypothesis simply for lack of knowledge. I accept that this wouldn't constitute a formal proof of design. But since formal proofs simply don't exist in experimental science, this should hardly be considered surprising.

The next issue is that of assumptions about the designer's goals and abilities. This probably has to be approached theologically. It has already been argued (for example, by Dave Heddle) that Christians should not have a problem with saying who they identify as the “intelligent designer.” It is possible to at least suggest a candidate for a designer, and then see what impact this has on the analysis.

So let's say, for the sake of argument, that the intelligent designer is the Christian God. Sober writes in objection to design:
One needs independent evidence as to what the designer's plans and abilities would be if he existed: one can't obtain this evidence by assuming that the design hypothesis is true. (p.10, I think)
... is continuous with the precepts of 'negative theology,' which holds that God is so different from us and the world we already know about that it is imposisble for us to have much of a grasp of what his characteristics are.
This fails adequately to deal with the claims of orthodox Christian theology, for one thing. Within Christian theology, the Bible is considered to be a revelation by God of himself. Francis Schaeffer, in He is There and He is Not Silent, includes an appendix called “Is Propositional Revelation Nonsense?” This addresses precisely these objections to Christianity – that we have no reason to believe that God will make himself known, and that God is so different from us that we can't possibly hope to understand what he is like.

Of course, this leads us into all sorts of odd areas – such as the nature of biblical texts, their reliability, their authenticity and so on. However, the point is this: Sober is objecting to the design argument on the basis that we don't know the designer, or his intentions or his characteristics. At least from a Christian point of view, those questions have been answered. There is something circular in Sober's objections here. “We can't believe in design because we don't know the designer,” is the heart of this objection – the “Achilles' Heel” of the design argument. But then the rejection of the idea of the Christian God is considered reasonable because: “we don't know that things are designed.” By starting from an atheist/deist/agnostic position, this argument then uses the fact that you are starting from this position as evidence in support of the position!

This is from a Christian perspective. I have no doubt that other religions would also argue that they have a revelation from their own deities about their nature. I don't believe, then, that the objection that we don't know about the intentions of the designer can be reasonably considered as part of this argument. It would be better to say that the argument is: “We aren't prepared to accept the evidence we have about the nature of the purported designer or designers.” This makes clear the fact that the rejection of the design argument is also a rejection of what certainly the Christian God, and possibly gods in other religions, considered to be sufficient revelation of themselves to support the proposition that they are the designer. In other words, choosing agnosticism is effectively anti-theism/atheism (as it involves discarding evidence that God claims is sufficient), and anti-theism/atheism is, unless the evidence from God is not available for evaluation, a conscious, moral choice.