Monday, September 03, 2007

Intelligent Design Credo (1)

(Some commenters are keen for me to be more committed in what I write. So here's a couple of posts which seek to address in some depth my actual thoughts on some aspects of intelligent design, and how it relates to the Bible.)

I believe in Intelligent Design. What do I mean by this? I mean that life and the universe isn't simply the product of blind forces, but the product of deliberate agency. To an extent, this is a matter of presuppositions – all Christians must believe in intelligent design (the concept, not the movement), and the propositions that define an intelligent design worldview are a subset of those which define the Christian worldview (including the young-earth creationist one). If Christians don't believe in intelligent design, the God in which they believe has little to do with the God who has revealed himself in the Bible, and it's hard to know in what sense they can describe themselves as Christians.

So my belief in ID is a matter of presuppositions. But I would argue that rejection of the idea of external agency is also a matter of presuppositions. I am still convinced that it isn't possible to argue that belief or lack of belief in God is a consequence of empirical observation. In my opinion, the evidence of such observation is just as consistent with belief in God as it is with an atheistic worldview.

I think that it will always technically be an option to believe that there is no god – that the universe is self-supporting, and somehow appeared without the existence of external agency. It is worth noting how this position has been forced to retreat (or at least, redefine itself) over the last century, however. The presumption that the universe is eternal (“Steady State”) - and thus all outcomes will inevitably occur eventually – has given way to the Big Bang model, which says that the universe has an origin in space-time. The presumption that a universe like ours is inevitable has given way to an awareness of how finely tuned conditions and constants have to be for the universe to develop as it has. The presumption that complex, intelligent life is common is giving way to the “rare earth” model, which suggests that the number of places in the universe where complex, intelligent life could possibly develop is very small. And so on. Of course, none of these changes to the scientific worldview in themselves represent knock-down evidence of a designer. However, the idea that life and the universe is designed certainly has no less force than it did at the time of (say) Darwin, and probably has more.

If there is a designer, it is not necessary for that designer to have left direct visible evidence of its work – see Del Ratzsch, “Nature, Design and Science”. It is possible that such a designer is able to manipulate initial conditions (perhaps the initial distribution of matter after the Big Bang), or apparently natural processes (specific sequences of mutations), in such a way that the eventual state is designed, though it looks like the outcome of natural processes (Ratzsch describes such things as “pronatural”, if I have understood him correctly). Ratzsch argues that it is possible to infer the presence of design in such cases, although it isn't simple. Conversely, the fact that there is no primary evidence of design says little about whether or not design has actually occurred. I'm going to read his book again, and try and distill some of his argument into some blog posts – almost every page has something thought-provoking and useful on it.

However, my hunch is that there will prove to be more direct evidence of design than this, and that this evidence will be visible no matter how deep into the science you get. My hunch is that design will be apparent in systems that lie realistically beyond the scope of a naturalistic process. I don't think that random mutation (coupled with natural selection) is a sufficiently powerful engine to generate the complex biochemical systems that we see in life today, and that such systems could only have arisen through a highly contrived, non-random sequence of mutations. I suspect that it would be possible to say that such events could have occurred by chance, but that the chance hypothesis will not prove to be satisfactory. As I have pointed out before, Richard Dawkins accepts that there is a limit to the acceptability of the chance hypothesis. He argues, in effect, that if there is less than one in a billion billion (1018) chance of a particular specified outcome, then a chance explanation is not really satisfactory. (He was talking about abiogenesis.) He made a similar point in both “The Blind Watchmaker” and “The God Delusion”. From the ID side, Michael Behe demonstrates in “The Edge of Evolution” that in 1020 tries, new biochemical systems haven't arisen in either the malarial parasite (for example, to overcome the sickle cell mutation which confers some protection to humans from the parasite) or HIV. This suggests that there is a limit on the competence of random mutation to generate new systems. William Dembski errs on the side of caution, and sets an absolute probability boundary of 10-150. However, in principle, the point is that there is an acceptance on both sides of the debate that an excessive reliance on improbable outcomes strengthens the suggestion that there is a designer.

It's important to note that we are not talking about “general” improbability, here. If I shuffle a deck of cards, and lay the cards out one by one, the particular sequence that I get will probably never have been seen before. That doesn't mean that there is anything significant about it, though. In the same way, it isn't simply the sequence of DNA bases in humans that is of relevance. But in the context of biochemical systems, proteins require binding sites and other chemically active regions. In principle, it is likely to become possible at some stage to determine the actual specification of a particular biochemical solution. Also, once a minimally specified solution has been arrived at, natural selection provides a means whereby the solution can be improved – but if a minimally specified solution is itself beyond the probabilistic resources of an organism, then this is again indicative of design.

As I have discussed elsewhere, Dawkins argues that however improbable life is, a designer is much more improbable. This is a silly argument, more worthy of a character in a Douglas Adams book than an Oxford professor. Suffice it to say that it has been refuted. Also, it is worth pointing out that neither the fact that a designer's identity is unknown, nor that its intentions are unknown, nor that the design is not “optimal” (according to some other person's concept of what is optimal) nor that the design is “harmful” (according to some other person's concept of how things ought to be) represent refutations of the proposition that something is designed.

Study into the improbability of systems and what this implies about design is still at a relatively early stage. However, the darwinist assertion that natural processes are on their own able to generate systems and organisms that look designed is far from being shown to be true. In fact, other than for relatively small-scale examples, a significant number of which I have talked about on my blog already, there are few systems where it can be conclusively shown that they have arisen through ateleological means.

The fact that there are lots ways to solve most biochemical problems doesn't completely get round the problem of the specific outcome that we see. People argue that the formula “come up with any solution to biochemical problem P” allows an indeterminately large number of answers, and therefore the task is a lot easier in evolutionary terms than would be suggested simply by looking at how improbable the eventual solution is. It is important not to exaggerate how well-specified a system is, and this is an area where more work needs to be done. However, if a solution requires multiple steps (which is conceptually the case for so-called “irreducibly complex” systems), then having granted even a most improbable first step, we can then look at the probability of subsequent steps. If these are too low, then the solution will still look designed. In this area, it is important for the sake of the discussion between proponents of ID and proponents of naturalism to move away from “just-so” stories of how the mammal got its cascading blood-clotting system, or how the bacterium got its flagellum, and to start to determine how feasible hypothetical developmental pathways are. This is something that is important for people on both sides of the debate. People with a naturalistic worldview will be keen to demonstrate that naturalistic processes could reasonably achieve what they need to. People who are keen to demonstrate design will want to show that what is required of natural processes is just too improbable. At the moment, there is little to show in terms of bridges between the conceptual evolutionary processes and the real world. This ought to change.